Respiratory Medicine

Series Editors: Sharon I. S. Rounds · Anne Dixon · Lynn M. Schnapp

Gregory A. Schmidt Editor

Extracorporeal
Membrane
Oxygenation
for Adults
Second Edition
Respiratory Medicine
Series Editors
Sharon I. S. Rounds, Brown University
Providence, RI, USA
Anne Dixon, University of Vermont, Larner College of Medicine
Burlington, VT, USA
Lynn M. Schnapp, University of Wisconsin - Madison
Madison, WI, USA
Respiratory Medicine offers clinical and research-oriented resources for
pulmonologists and other practitioners and researchers interested in respiratory
care. Spanning a broad range of clinical and research issues in respiratory medicine,
the series covers such topics as COPD, asthma and allergy, pulmonary problems in
pregnancy, molecular basis of lung disease, sleep disordered breathing, and others.
The series editors are Sharon Rounds, MD, Professor of Medicine and of
Pathology and Laboratory Medicine at the Alpert Medical School at Brown
University, Anne Dixon, MD, Professor of Medicine and Director of the Division of
Pulmonary and Critical Care at Robert Larner, MD College of Medicine at the
University of Vermont, and Lynn M. Schnapp, MD, George R. And Elaine Love
Professor and Chair of Medicine at the University of Wisconsin-Madison School of
Medicine and Public Health.
Gregory A. Schmidt
Editor

Extracorporeal Membrane
Oxygenation for Adults
Second Edition
Editor
Gregory A. Schmidt
Division of Pulmonary Diseases, Critical Care, and Occupational Medicine
Department of Internal Medicine, University of Iowa
Iowa City, IA, USA

ISSN 2197-7372     ISSN 2197-7380 (electronic)

Respiratory Medicine
ISBN 978-3-031-05298-9    ISBN 978-3-031-05299-6 (eBook)
https://doi.org/10.1007/978-3-031-05299-6

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2016, 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Humana imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To the Core ECMO Specialists at the
University of Iowa. Through an incredibly
challenging 2 years, you have continued to
lead, guide, educate, negotiate, muster
resources, and otherwise advocate for
excellent ECMO care. Our patients are
fortunate to have you. I’m grateful to work
alongside you.
Preface to the 2nd Edition

Only 3 years ago, extracorporeal membrane oxygenation (ECMO) was considered

an extreme measure, limited to a small number of dedicated practitioners in highly
specialized centers. In the interim, the pandemic due to SARS-CoV-2, and its asso-
ciated illness, COVID-19, ushered ECMO into mainstream medicine. Lung failure
related to COVID-19 is often so profound, even in patients in previously good
health, that ECMO is in increasing demand. The medical community responded by
building new ECMO teams and disseminating the knowledge and skills to deliver
ECMO to a far greater number of patients. These heightened capabilities are likely
to help not only patients with COVID-19 but also those with ARDS of other causes;
diseases dominated by ventilatory failure such as status asthmaticus and COPD;
cardiogenic shock; following cardiothoracic surgery complicated by circulatory or
gas exchange failure; and as a bridge to lung transplant.
Alongside the impetus of the pandemic, technological improvements in mem-
branes, pumps, circuits, and cannulas have led to more efficient and safer ECMO. An
increasing evidence base supports the role of ECMO in improving patient outcomes.
Finally, knowledge of the adverse consequences of conventional management of
lung failure, including ventilator-induced lung injury, ICU-acquired weakness, and
nosocomial infection, continues to accumulate. Indeed, the future of ECMO
is bright!
The aim of this book is to deliver a concise, evidence-based review of ECMO,
emphasizing the use of veno-venous ECMO for respiratory failure in adults.
Chapters are devoted to describing the complex physiology and technology; the
evidence base in varied clinical conditions; how to obtain vascular access; daily
management of the circuit and patient; guidance regarding the weaning and decan-
nulation process; and recommendations for crisis management and rehabilitation
related to ECMO. New chapters on managing the systemic circulation (especially
volume status and right ventricular dysfunction), anti-thrombotic therapy, recogniz-
ing membrane failure, and ethical considerations have been added. The section on
daily management newly emphasizes the signal role for point-of-care ultrasound in
guiding care and is intended for a broad range of clinicians.

vii
viii Preface to the 2nd Edition

This book has been written for practicing physicians, nurses, perfusion special-
ists, therapists, and critical care trainees who are considering whether to refer their
patients for ECMO, debating whether to offer ECMO capability to their patients, or
are already providing ECMO but seek a practical reference to best practices and
updated information. It could never have been completed without the inspiration
from my colleagues at Iowa who strive daily to save the sickest patients; the trainees
whose curiosity makes us all want to know more; my contributors who are at the
forefront of a truly challenging field; and our publisher at Springer-Link who pushed
for this important revision. Finally, I recognize all those who do the hard work: the
ECMO specialists, bedside nurses, and therapists who dedicate their lives to the
critically ill. This is an exciting time, ripe with change and new opportunities.

Iowa City, IA, USA Gregory A. Schmidt

Contents

1 
Physiology of Extracorporeal Life Support (ECLS)����������������������������    1
Matthew J. Brain, Warwick W. Butt, and Graeme MacLaren
2 
Circuits, Membranes, and Pumps����������������������������������������������������������   63
Bradley H. Rosen
3 Modes of ECMO��������������������������������������������������������������������������������������   81
Jonathan Eaton, Christopher Trosclair, and L. Keith Scott
4 Vascular Access����������������������������������������������������������������������������������������   97
Steven A. Conrad
5 
Hypoxemic Respiratory Failure: Evidence,
Indications, and Exclusions�������������������������������������������������������������������� 115
Kathleen E. Melville, Cara Agerstrand, Daniel Brodie,
and Darryl Abrams
6 
Ventilator Management During ECLS�������������������������������������������������� 125
Antonio Pesenti, Giacomo Bellani, Giacomo Grasselli,
and Tommaso Mauri
7 
Managing the Systemic Circulation: Volume
Status and RV Function�������������������������������������������������������������������������� 147
Sundar Krishnan and Gregory A. Schmidt
8 Antithrombotic Therapy for ECMO������������������������������������������������������ 159
Usha S. Perepu
9 Membrane Dysfunction �������������������������������������������������������������������������� 173
B. D. Warren, M. J. Sobieszczyk, and P. E. Mason
10 ECCO2R in Obstructive Diseases: Evidence,
Indications, and Exclusions�������������������������������������������������������������������� 187
Lorenzo Del Sorbo, V. Marco Ranieri, and Vito Fanelli

ix
x Contents

11 
ECMO as a Bridge to Lung Transplantation���������������������������������������� 205
Christian Kuehn and Ruslan Natanov
12 
Daily Management of Patients on VV ECMO�������������������������������������� 217
Charles Rappaport and Kristina Rappaport
13 Crises During ECLS�������������������������������������������������������������������������������� 229
Purnema Madahar, Dana A. Mullin, Meaghan Flatley,
Darryl Abrams, Phillipe H. Lemaitre, Daniel Brodie,
and Cara Agerstrand
14 Mobilization During ECLS �������������������������������������������������������������������� 253
Gregory A. Schmidt
15 ECMO Weaning and Decannulation������������������������������������������������������ 265
Sharon L. McCartney and Sundar Krishnan
16 
Venoarterial ECMO in Respiratory Failure������������������������������������������ 277
Avery Tung and Tae H. Song
17 
Ethical Challenges in Extracorporeal
Membrane Oxygenation�������������������������������������������������������������������������� 293
Elizabeth Sonntag and Meera Pahuja

Index�������������������������������������������������������������������������������������������������������������������� 303
Contributors

Darryl Abrams, MD Division of Pulmonary, Allergy, and Critical Care Medicine,

Department of Medicine, Columbia University College of Physicians & Surgeons,
New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New
York, NY, USA
Cara Agerstrand, MD Division of Pulmonary, Allergy, and Critical Care
Medicine, Department of Medicine, Columbia University College of Physicians
and Surgeons/NewYork-Presbyterian Hospital, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New
York, NY, USA
Giacomo Bellani, MD Department of Pathophysiology and Transplant, University
of Milan, Milan, Italy
Department of Anestesia and Critical Care, Fondazione IRCCS Ca’ Granda
Ospedale Maggiore Policlinico, Milan, Italy
Matthew J. Brain, MBBS, Ph.D., FRACP, FCICM, DDU Department of
Medicine, Launceston General Hospital, Launceston, TAS, Australia
Daniel Brodie, MD Division of Pulmonary, Allergy, and Critical Care Medicine,
Department of Medicine, Columbia University College of Physicians and Surgeons
/ NewYork-Presbyterian Hospital, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New
York, NY, USA
Warwick W. Butt, FRACP FCICM Intensive Care, Royal Childrens Hospital,
University of Melbourne Department of Paediatrics, Melbourne, VIC, Australia
Steven A. Conrad, MD PhD, MCCM, FCCP, FELSO Departments of Medicine,
Emergency Medicine, Pediatrics and Surgery, Louisiana State University Health
Sciences Center, Shreveport, LA, USA

xi
xii Contributors

Ochsner LSU Health Academic Medical Center, Shreveport, LA, USA

LSU Health Shreveport, Shreveport, LA, USA
Lorenzo Del Sorbo, MD Interdepartmental Division of Critical Care Medicine,
University Health Network, University of Toronto, Toronto, ON, Canada
Jonathan Eaton, MD Critical Care Medicine Division, LSU Health Shreveport,
Shreveport, LA, USA
Vito Fanelli, MD, PhD Dipartimento di Anestesiologia e Rianimazione, Azienda
Ospedaliera Città della Salute e della Scienza di Torino, Università di Torino,
Torino, Italy
Christian Kuehn, MD Department of Cardiothoracic, Transplantation, and
Vascular Surgery, Hannover Medical School, Hannover, Germany
Meaghan Flatley, MD Department of Surgery, Columbia University College of
Physicians and Surgeons/NewYork-Presbyterian Hospital, New York, NY, USA
Giacomo Grasselli, MD Department of Pathophysiology and Transplant,
University of Milan, Milan, Italy
Department of Anestesia and Critical Care, Fondazione IRCCS Ca’ Granda
Ospedale Maggiore Policlinico, Milan, Italy
Sundar Krishnan, MBBS Department of Anesthesiology, Duke University School
of Medicine, Duke University Medical Center, Durham, NC, USA
Divisions of Cardiothoracic Anesthesia and Critical Care, Department of
Anesthesiology, Duke University Medical Center, Durham, NC, USA
Phillipe H. Lemaitre, MD Center for Acute Respiratory Failure, New York-­
Presbyterian Hospital, New York, NY, USA
Division of Cardiothoracic Surgery, Department of Surgery, Columbia University
College of Physicians and Surgeons/NewYork-Presbyterian Hospital, New
York, NY, USA
Graeme MacLaren, MBBS MSc FCICM FRACP FCCM Cardiothoracic ICU,
National University Hospital, Singapore, Singapore
Purnema Madahar, MD, MS Division of Pulmonary, Allergy, and Critical Care
Medicine, Department of Medicine, Columbia University College of Physicians
and Surgeons/NewYork-Presbyterian Hospital, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New
York, NY, USA
Phillip E. Mason, MD Department of Surgery, Brooke Army Medical Center, San
Antonio, TX, USA
Tommaso Mauri, MD Department of Pathophysiology and Transplant, University
of Milan, Milan, Italy
Contributors xiii

Department of Anestesia and Critical Care, Fondazione IRCCS Ca’ Granda

Ospedale Maggiore Policlinico, Milan, Italy
Sharon L. McCartney, MD, FASE Divisions of Cardiothoracic Anesthesia and
Critical Care, Department of Anesthesiology, Duke University Medical Center,
Durham, NC, USA
Kathleen E. Melville, MD Division of Pulmonary, Allergy, and Critical Care
Medicine, Department of Medicine, Columbia University College of Physicians &
Surgeons, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New
York, NY, USA
Dana A. Mullin, MS, CCP Department of Clinical Perfusion and Anesthesia
Support Services, NewYork-Presbyterian Hospital, New York, NY, USA
Ruslan Natanov, MD Department of Cardiothoracic, Transplantation, and
Vascular Surgery, Hannover Medical School, Hannover, Germany
Meera Pahuja, MD, MSc Department of Internal Medicine, Virginia
Commonwealth University School of Medicine, Richmond, VA, USA
Usha S. Perepu, MD Department of Internal Medicine, University of Iowa,
Iowa City, IA, USA
Antonio Pesenti, MD Department of Pathophysiology and Transplant, University
of Milan, Milan, Italy
Department of Anestesia and Critical Care, Fondazione IRCCS Ca’ Granda
Ospedale Maggiore Policlinico, Milan, Italy
V. Marco Ranieri, MD Department of Medical and Surgical Sciences (DIMEC),
Alma Mater Studiorum, University of Bologna, Bologna, Italy
Charles Rappaport, MD Division of Pulmonary, Critical Care and Occupational
Medicine, University of Iowa Hospital and Clinics, Iowa City, IA, USA
Kristina Rappaport, MSN, CCRN University of Iowa Hospital and Clinics, Iowa
City, IA, USA
Bradley H. Rosen, D.O. Division of Pulmonary, Critical Care, and Occupational
Medicine, Department of Internal Medicine, Roy J. and Lucille A. Carver College
of Medicine, University of Iowa Hospitals and Clinics, Iowa City, IA, USA
Gregory A. Schmidt, MD Division of Pulmonary Diseases, Critical Care, and
Occupational Medicine, Department of Internal Medicine, University of Iowa, Iowa
City, IA, USA
L. Keith Scott, MD MSc FCCM Critical Care Medicine Division, LSU Health
Shreveport, Shreveport, LA, USA
xiv Contributors

Michal Sobieszczyk, MD Department of Pulmonary and Critical Care Medicine,

Brooke Army Medical Center, San Antonio, TX, USA
Tae H. Song, MD Section of Cardiac Surgery, Department of Surgery, University
of Chicago, Chicago, IL, USA
Elizabeth Sonntag, MD Department of Internal Medicine, Virginia Commonwealth
University School of Medicine, Richmond, VA, USA
Christopher Trosclair, MD ECLS Program, Critical Care Medicine, Louisiana
State University Health Shreveport, Shreveport, LA, USA
Critical Care Medicine, LSU Health Shreveport, Shreveport, LA, USA
Avery Tung, MD, FCCM Department of Anesthesia and Critical Care, University
of Chicago, Chicago, IL, USA
Bryce Warren, MD, PhD Department of Internal Medicine, Brooke Army Medical
Center, San Antonio, TX, USA
Chapter 1
Physiology of Extracorporeal Life Support
(ECLS)

Matthew J. Brain, Warwick W. Butt, and Graeme MacLaren

Introduction

Extracorporeal life support (ECLS) and related implantable circulatory assistance

devices describe several advancing technologies with broadening scope that are
being increasingly incorporated into the management of critically ill patients.
ECLS may be provided in several configurations to support or replace cardiore-
spiratory function (Fig. 1.1). In veno-venous extracorporeal membrane oxygenation
(VV-ECMO), the objective is maintaining systemic oxygen delivery by oxygenat-
ing venous blood returning to the right heart. In veno-arterial mode (VA-ECMO),
systemic blood flow is augmented by the extracorporeal blood pump, while veno-­
pulmonary artery (VPA-ECMO) describes augmentation of pulmonary arterial flow.
Both of the later configurations can also incorporate support of oxygenation. A vari-
ant of VV-ECMO is ECCO2R where CO2 removal occurs at blood flows too low to
provide appreciable oxygenation.
The site of vascular access can also classify configurations, with cannulas being
either peripherally placed via the great vessels or centrally placed via thoracotomy.
Miniaturisation of rotary pumps (without oxygenators) has also occurred, allowing
development of left and right ventricular assist devices (LVAD and RVAD),

M. J. Brain (*)
Department of Medicine, Launceston General Hospital, Launceston, TAS, Australia
e-mail: [email protected]
W. W. Butt
Intensive Care, Royal Childrens Hospital, University of Melbourne Department of
Paediatrics, Melbourne, VIC, Australia
e-mail: [email protected]
G. MacLaren
Cardiothoracic ICU, National University Hospital, Singapore, Singapore

© The Author(s), under exclusive license to Springer Nature 1

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_1
2 M. J. Brain et al.

VA-ECMO
VV-ECMO

VPA-ECMO LVAD

Fig. 1.1 Schematic of ECMO configurations, circles represent pumps and diamonds represent
oxygenators. VA-ECMO: veno-arterial extracorporeal membrane oxygenation demonstrating
cavo-aortic flow. VV-ECMO: veno-venous cannulation demonstrating cavo-atrial flow from the
inferior vena cava to the right atrium via the oxygenator and pump. VV-ECMO may also require a
second cannula taking blood from the superior vena cava or dual lumen cannulas that access blood
from the inferior and superior vena cava, while returning blood to the right atrium. VPA-ECMO:
veno-pulmonary artery cannulation may be configured as atrial to pulmonary artery flow with or
without oxygenation support. LVAD: left ventricular assist device (usually implanted) taking left
ventricular blood and returning it to the proximal aorta. RVAD: a right ventricular assist device is
not shown but may be implanted or external and can be configured identically to VPA-ECMO
without an oxygenator, or may directly drain the right ventricle as per the LVAD. Extracorporeal
carbon dioxide removal (ECCO2-R) is commonly performed with a VV-ECMO configuration, usu-
ally with a single dual lumen catheter. Intravascular membrane oxygenators have also been devel-
oped [1] but are not currently in clinical use

respectively. This chapter will focus on VV-ECMO and only cover other configura-
tions when relevant.
A basic ECMO circuit consists of a blood pump and oxygenator connected by
conduits (Fig. 1.1). Other components may be added to this basic configuration,
most commonly renal replacement therapy connected in parallel via side ports.
Minimising the number of connections and maintaining simplicity is important for
safety, infection control and troubleshooting.
Each configuration creates a unique interaction with the cardiorespiratory sys-
tem. Sound understanding of the physiology and limitations of each mode is
required to prescribe, manage and wean this support and recognise evolving com-
plications of the therapy. Although designed primarily to replace cardiorespiratory
function, the interaction of ECLS with several other physiologic systems must be
1 Physiology of Extracorporeal Life Support (ECLS) 3

considered. For example, most patients who require ECLS will have sustained a
major insult such as severe sepsis, trauma or surgery, or have suffered from progres-
sive cardiac or pulmonary disease. A systemic inflammatory response syndrome
(SIRS) and coagulopathies may arise from the underlying pathology, as a reaction
to the non-biological material of the ECLS circuit, or from the secondary develop-
ment of sepsis due to the invasive nature and immunosuppressive consequences of
treatment. The varying metabolic responses to critical illness and chronic pathology
have direct implications for oxygenation and CO2 production, as well as nutritional
supplementation to facilitate later weaning.
At this time, ELCS is a bridge to either recovery of a failing cardiorespiratory
system or a destination therapy such as an LVAD or organ transplantation. In rapidly
progressing pathologies ECLS is sometimes a bridge to decision, buying time to
assess the disease prognosis and implications of any destination therapies [2].
In order to comprehensively understand ECLS and its effects on human physiol-
ogy, it is necessary to first review cellular metabolism and oxygen transport.

Cellular Metabolism

The fundamental role of tissue perfusion is to provide sufficient substrate delivery

to match the metabolic demand of aerobic cellular metabolism. While anaerobic
metabolism can support cellular energy requirements for brief periods, only oxida-
tive metabolism can sustain cellular and organ function.
Cardiorespiratory physiology and any mechanical support must provide an ade-
quate hydrostatic pressure gradient across capillary beds to support blood flow and
thus maintain concentration gradients by which substrates, including oxygen, dif-
fuse into the immediate environment of cells. Likewise, a concentration gradient
must be maintained from the cell to the blood path for the waste products of metabo-
lism, primarily CO2, or lactate in the case of anaerobic metabolism. These functions
are interlinked, as the waste products of energy production are generally weak acids
and influence local perfusion and oxygen carriage.
The quantities of substrate required per unit time will depend on the supported
cell mass and its level of metabolic activity as influenced by demand (or stress),
temperature, inflammation and hormonal regulation.

Glycolysis and Aerobic and Anaerobic Metabolism

Glucose (and other simple carbohydrates) enter cells down a concentration gradient
through glucose transporters that allow tissue-specific behaviour such as preferen-
tial basal uptake by the brain, concentration-dependent uptake by the liver,
concentration-­
sensing by the insulin-secreting pancreatic ß-cells and insulin-­
dependent uptake in skeletal muscle and fat [3].
4 M. J. Brain et al.

Intracellular glucose is rapidly phosphorylated in the cytosol by hexokinases,

before becoming the primary substrate for energy production or biosynthetic reac-
tions including glycogen storage (Fig. 1.2). Utilisable intracellular energy is stored
in the phosphate bonds of adenosine triphosphate (ATP), and it is the breaking of
chemical bonds within glucose that powers ATP regeneration from adenosine
diphosphate (ADP) and inorganic phosphate (Pi).
Glycolysis describes the fracturing of the six-carbon glucose molecule into two
three-carbon pyruvate molecules with the next generation of two ATP molecules.
For glycolysis to continue, oxidative power (NAD+ concentration) must be continu-
ally restored. Under anaerobic conditions, this occurs by conversion of pyruvate to
lactate. Under aerobic conditions, pyruvate loses a carbon dioxide molecule to yield
acetyl coenzyme-A. This two carbon acetyl group can be incorporated into fatty
acids for storage or can enter the tricarboxylic acid (TCA) cycle to complete the
chemical breakdown of glucose to CO2. The TCA cycle yields 38 ATP molecules
per glucose, significantly more than glycolysis however generates NADH in such
quantities that a powerful electron acceptor is required for efficient restoration of
NAD+ so that the cycle can continue. This electron acceptor is oxygen.
Oxidative phosphorylation describes the process of restoring NAD+ to perpetuate
the TCA cycle. Although oxygen is utilised as an electron acceptor in many enzyme
systems, its highest consumption is in this process. Oxidative phosphorylation
occurs in the inner mitochondrial matrix, and it is to this intracellular destination
that oxygen must diffuse in sufficient quantities to sustain ATP generation for nor-
mal cellular processes.
When oxygen is not available in sufficient quantities, ATP generation from ADP
can only continue in the cytosol by anaerobic glycolysis. This process is inefficient,
as not only is less ATP produced but the resulting lactic acid is not as readily cleared
from the tissues or body as carbon dioxide. Lactic acid is thus a marker of glycolysis
activity in a hypoxic environment and usually indicates inadequate tissue perfusion
and/or global hypoxemia of the organism.

Carbon Dioxide Production and the Respiratory Quotient

The respiratory quotient (RQ) describes the ratio of the amount of carbon dioxide
(VCO2 ) produced per unit time to the amount of oxygen consumed (VO2 ).

VCO2
RQ = (1.1)
VO2

The respiratory quotient is dependent on the sources of fuel being used. For glu-
cose metabolism, the six-carbon atoms result in production of six molecules of car-
bon dioxide while consuming six molecules of oxygen; it thus has a respiratory
quotient of one. The oxidation of fatty acids (lipolysis) and some amino acids pro-
duce less CO2 per O2 (by not producing NADH in the pyruvate to acetyl-CoA reac-
tion, Fig. 1.2) and hence have respiratory quotients of less than one.
1 Physiology of Extracorporeal Life Support (ECLS) 5

2-
CH2 OPO3

OH
OH OH

OH (+) Glucagon
ATP ADP Glucose-6-P
Glucose Glycogen
(6 Carbons)
ATP (+) Insulin
ADP (-) Epinephrine
Fructose-6-P

(+) Insulin (+) Glucagon

(-) Epinephrine (+) Citrate

Fructose-1,6-bis-P

2-
O3 PO
HO
Dihydroxyacetone Glyceraldehyde-3-P
Phosphate
(3 Carbons) 2-
(3 Carbons) O OPO3
NAD+
NADH 2,3-Diphosphoglycerate
1,3-Bisphosphoglycerate Rapoport-Leubering
Glycerol ADP Shunt(Erythrocytes)
ATP
Triglycerides 3-Phosphoglycerate

Phosphoenolpyruvate
Fatty Acid Synthesis Amino Acids
(+) Insulin ADP (+) Glucagon
(-) Acetyl-CoA ATP (+) Acetyl-CoA
After transfer to Liver
Acetyl-CoA Pyruvate
Ketone Bodies Lactate
(2 Carbons) NADH NAD+ (3 Carbons)
+ CO O OH O OH
2 NADH NAD+

H3C O H3C OH
(4 Carbons) Oxaloacetate Citrate (6 Carbons) 3-5 ADP
O2 2 NAD+
Tricarboxylic Acid
3NAD+ + FAD Oxidative Phosphorylation
Cycle
3NADH + FADH + 2CO HO 2 NADH
2 2 2
3-5 ATP

Fig. 1.2 Key intermediates in intracellular metabolism: After entering cells, glucose is phosphory-
lated (-P) and can then be incorporated into glycogen, enter synthetic reactions (not shown), or be
metabolised to two three-carbon pyruvate molecules (glycolysis). The conversion of pyruvate to
acetyl-CoA, the subsequent tricarboxylic acid (TCA) cycle and oxidative phosphorylation only
occur in mitochondria and depend on oxygen to restore nicotinamide adenine dinucleotide to its
oxidised form (NAD+) for continued cycling. The number of ATP generated depends on the source
of reduction power; a single mitochondrial NADH produces 2.5 ATP; however, electrons from
cytosolic NADH must be transferred to mitochondrial FADH2 which yields only 1.5 ATP each [4].
Amino acids can enter or be synthesised at several points. Acetyl-CoA is a key junction molecule
providing the TCA cycle with two carbon acetyl groups, not only from glycolysis but also from
fatty acids and some amino acids. In glucose excess, acetyl-CoA is the starting point for fatty acid
synthesis and, in the starvation state, of ketone body production when insufficient oxaloacetate
exists for acetyl groups to enter the TCA cycle. Production of ketone bodies occurs predominantly
in the liver from fatty acid breakdown and constitutes a glucose-sparing fuel for the brain and
heart. Humoral promoters and inhibitors of reactions are shown
6 M. J. Brain et al.

In contrast to oxidation, each acetyl-CoA molecule utilised for fatty acid synthe-
sis (lipogenesis) results in production of a molecule of CO2 (from pyruvate to
acetyl-­CoA, Fig. 1.2) without increasing mitochondrial NADH. As the rate of oxy-
gen consumption depends on the mitochondrial concentration of NADH, lipogene-
sis results in CO2 production which exceeds oxygen consumption. Some oxygen
consumption still occurs as the synthetic reaction also consumes ATP; however, the
RQ will be greater than 1. Examples of respiratory quotients based on theoretical
stoichiometry include [5, 6]:

C6 H12 O6 + 6O 2
Glucose Oxidation : RQ = 1
→ 6CO 2 + H 2 O
C57 H110 O6 + 80.25 O 2
Lipolysis of glycerol triestearate : RQ = 0.667
→ 57 CO 2 + 55.5 H 2 O
NC2 H 5 O 2 + 2.25 O 2
Amino Acid ( Glycine ) Oxidation : RQ = 0.88
→ 2 CO 2 + 2.5 H 2 O
4 C6 H12 O 6 + O 2
Lipogenesis from Glucose∗ : RQ = 8
→ C16 H 32 O 2 + 8 CO 2 + 8 H 2 O
13.83 C6 H12 O6 + 5 O 2
Lipogenesis∗ : RQ = 5.6
→ C55 H104 O6 + 28 CO 2 + 31 H 2 O

Note that the RQ of lipogenesis is dependent on the fatty acid being produced
*

and the carbohydrate that is utilised. C16H32O2C16H32O2 palmitic acid.

C55H104O6C55H104O6 palmitoyl-stearoyl-2-oleoyl-glycerol.
A normal adult has a whole body RQ measured by indirect calorimetry of around
0.8, reflecting utilisation of mixed fuel sources. This value will alter in critically ill
patients, depending on the nutrient availability and humoral control of metabolism.
While glycolysis reflects enzymatic processing of glucose, complete aerobic metab-
olism is coupled to TCA intermediate availability, and when carbohydrate loads are
excessive, such as glucose supplementation exceeding 4 mg/kg/min (5.8 g/kg/day,
parenteral or gastrointestinal), lipogenesis occurs with a resulting higher respiratory
quotients [6–8].

Metabolism in the Stressed State

Key hormones coordinate the response of energy metabolism to stress. Insulin indi-
cates the fed state promoting hepatic glucose uptake, glycogen and amino acid syn-
thesis and diversion of acetyl-CoA to free-fatty acid production while in the
peripheral tissues stimulating myocyte synthesis of contractile elements and adipo-
cytes to triglyceride deposition.
1 Physiology of Extracorporeal Life Support (ECLS) 7

Glucagon is secreted by pancreatic α-cells in response to low blood glucose lev-

els and promotes glycogen breakdown and conversion of amino acids (from muscle
breakdown), lactate and glycerol. Glycerol results from adipocyte triglyceride
metabolism, and the released free fatty acids are converted to ketone bodies by the
liver for use as a secondary fuel source when glucose is scarce.
The catecholamines epinephrine and norepinephrine are released in response to
physiologic stress. By increasing intracellular cyclic AMP, they promote glycoge-
nolysis in muscles and catabolism of protein to release amino acids. In the liver,
epinephrine promotes gluconeogenesis and glycogenolysis and inhibits glycolysis.
These responses result in the hyperglycaemia that characterises the stress state and
is exacerbated by exogenous administration of catecholamines and glucose. The
adverse effects of hyperglycaemia include osmotic diuresis, electrolyte disruption,
fat deposition in the liver and impaired immune function.
The metabolic profile of patients receiving ECLS is typical of the stressed state,
but the differences between this and the starvation state are important. In starvation
there is an overall decrease in energy expenditure with maximal use of triglycerides
and ketoacids promoting conservation of muscle bulk; the brain, heart and renal
cortex adapt to utilising ketoacids for significant proportions of their metabolic
requirements. In contrast, the chronic stressed state is characterised by increased
resting energy expenditure, accelerated catabolism of lean body mass – primarily
muscle catabolism to amino acids [4] – and the immunosuppressive effects of
hyperglycaemia and persistently elevated humoral mediators, including catechol-
amines and cortisol.
In those requiring ECLS, particularly those needing prolonged periods of heavy
sedation, the combination of muscle catabolism, disuse atrophy, critical illness
myopathy and myopathy associated with muscle relaxants can result in profound
weakness. The respiratory musculature is not spared from this process with the
result being prolonged ventilator weaning, increased requirement for tracheostomy
and risk of secondary infection.

Erythrocyte Metabolism

Being a specialised organ for oxygen transport, erythrocytes are nearly 90% haemo-
globin by weight, with very few other organelles. Nevertheless, they require an
ongoing energy source for maintenance of membrane integrity, cytoskeleton struc-
ture, intracellular electrolyte and osmotic equilibrium and to maintain the ferric
moieties of haemoglobin in a reduced state (Fe2+).
Erythrocytes lack mitochondria and do not store glycogen and thus depend on
anaerobic glycolysis of plasma glucose to lactate for ATP production. Glycolysis in
erythrocytes is also utilised for reactions that do not produce ATP, such as reducing
power to correct oxidised haemoglobin (methaemoglobin carrying a Fe3+ iron atom
that cannot carry oxygen), glutathione production (protecting the cell membrane
against oxidative damage) and the production of 2,3-diphosphoglycerate (2,3-DPG)
that modulates the affinity of haemoglobin for oxygen [9].
8 M. J. Brain et al.

The Rapoport-Leubering shunt (Fig. 1.2) describes 2,3-DPG synthesis from

the glycolytic pathway. In most cells, 1,3-DPG is rapidly converted to
3-­phosphoglycerate with the phosphate molecule transferred to ATP; however, in
erythrocytes up to 20% of glycolytic flux occurs through the shunt, with the value
dependent on ATP requirements [10]. Oxygen depletion (resulting in less haemo-
globin binding sites for 2,3-DPG), acidotic conditions that inhibit 2,3-DPG syn-
thesis and the accumulation of inorganic phosphate which increases 2,3-DPG
breakdown [9] result in decreased intracellular 2,3-DPG concentrations. This is
most relevant under conditions of red cell storage for later transfusion where
lower glycolysis rates and accumulation of lactic acid can result in minimal 2,3-
DPG concentrations at the time of transfusion. Transfused red cells do not restore
normal 2,3-DPG concentrations for some time, and given the relatively high
transfusion requirements of patients receiving ECLS, this effect may have signifi-
cant implications for oxygen carriage. The role of 2,3-DPG will be further dis-
cussed when considering oxygen carriage.

Biophysics of Membrane Gas Exchange

Mitochondrial coupling of ATP production to NADH oxidation can only occur if

sufficient oxygen exists in the environment of cells. Similarly, carbon dioxide dif-
fuses from the mitochondria, through intracellular membranes and away from the
cell. The flux of oxygen into the environment of cells and the reverse movement of
carbon dioxide can be divided into two components:
1. Diffusion of gas molecules into and between liquid phases
2. The carriage of gas molecules in blood
When considering pulmonary gas exchange, a third component must be consid-
ered: the convective transport of the gas to the alveolar epithelium. Exposure of the
extracorporeal membrane to fresh gas flow is somewhat simpler than pulmonary gas
transport and will be considered with carbon dioxide transport.

Membrane Oxygenator Construction

Extracorporeal membrane oxygenators consist of a high surface area blood path

separated by a membrane from a path for fresh gas flow (sweep gas). The devices
are in continual evolution to optimise the efficiency of gas transfer, minimise untow-
ard host biological interactions, reduce priming volumes, minimise plasma leakage
and improve simplicity and integration into care; however, a brief introduction to
their design is important to understand their operation.
Membranes may be arranged in folded sheets or, more commonly, as tubes
known as hollow fibre oxygenators (Fig. 1.3). The pores of earlier polypropylene
1 Physiology of Extracorporeal Life Support (ECLS) 9

Fresh Gas Flow Diffusion Path

Hollow Fibres ~ 300 microns l = 75 microns Boundary Layer

δ ε
O2 τ CO2
C
ε

O2 Path ~ 100 microns (τ) RBC 8 microns

Pore size < 1 micron

Blood
Flow

Heated Water
Fig. 1.3 Schematic detail of hollow fibre oxygenator construction demonstrating extra-capillary
flow of blood around the gas-carrying hollow fibres. Cross current flow exists between gas and
blood. Heated water tubules are also demonstrated. The diffusion path for gas exchange is shown
(top-right) consisting of the porous membrane and the boundary layer of adsorbed proteins.
Parameters of effective diffusivity from Eq. 1.8 are demonstrated with ε being the porosity – the
area of membrane occupied by gas, τ the tortuosity an index of effective path length for gas to
traverse the membrane (a path length is shown but in reality will be unknown), and δ the constric-
tivity – the resistance to gas passage [11]

microporous membranes theoretically allow contact between plasma and the sweep
gas; however, more recent materials such as poly-4-methyl-1-pentene utilise closed
fibres and are thus considered true membranes [12]. Many systems utilise the lumen
of the hollow fibres for fresh gas flow, with blood flowing between fibres; this is
termed extra-capillary flow. Overall characteristics such as total surface area for gas
exchange, resistance to flow and trauma to cellular blood components are deter-
mined by factors such as membrane material, fibre diameter and length, fibre den-
sity and the velocity of the blood [11].
Heat loss over the extracorporeal circuit from blood to the environment can be
substantial, and heat exchangers are commonly incorporated into oxygenator
10 M. J. Brain et al.

design. Figure 1.3 demonstrates one such design where the microporous membrane
fibres are laid perpendicularly to impermeable capillaries that circulate heated
water, allowing a large surface area for heat transfer to blood.

Diffusion of Gas Molecules into a Liquid Phase

Concentration of Gases in Solutions

Unlike most solutes dissolved in body fluids that are quantified in moles, gas con-
centrations are reported in units of pressure. The universal gas law describes the
relationship between the partial pressure of an ideal gas and its container, with ideal
gas molecules best summarised as having minimal mass and absent intermolecular
attraction:

P = nRT (1.2)
V

The universal gas equation: P = the partial pressure, n = number of molecules of

gas measured in moles, T is temperature in degrees Kelvin and V is volume of the
container in litres. R is the ideal gas constant which in SI units is 8.314 J.K−1.mol−1
or in conventional units: 62.36 mmHg.K−1.mol−1.
At constant temperature, Eq. 1.2 simplifies to P ∝ n/V. Concentration is defined
as moles/unit volume, that is, n/V hence pressure is proportional to concentration. In
other words the greater the number of gaseous molecules in a given volume (con-
centration), the more force those gas molecules will exert on the walls of the con-
tainer (pressure). The physical reaction of dissolving in solution is also proportional
to the partial pressure of the gas above the solution, such that for oxygen dissolu-
tion [13]:

K Forward
O 2 ( gas )  O 2 ( dissolved ) (1.3)
K Reverse

[O2 ](gas) × K Forward = [O2 ](dissolved ) × K Reverse a.

(1.4)
[O2 ](dissolved ) = SC × [O2 ](gas) b.

The rate constant KForward in Eq. 1.4a describes the proportion of oxygen gas that
dissolves per unit time, while KReverse describes the proportion of dissolved oxygen
that leaves the solution to the gas phase. When the system described in Eq. 1.4a is
at thermodynamic equilibrium, the concentrations in the gas and liquid phases are
stable, and the constants may then be combined, resulting in Eq. 1.4b, also known
as Henry’s law. SC is the Bunsen solubility coefficient resulting from
SC = KForward/KReverse and is gas and solvent specific. SC is affected by other dissolved
1 Physiology of Extracorporeal Life Support (ECLS) 11

solutes and falls with increasing temperatures (i.e. KForward becomes smaller and
KReverse larger).
Due to difficulties in measuring the molar concentration of oxygen compared to
the ease of measuring a volume of 100% oxygen at standard conditions (STPD:
0 °C, 760 mmHg, dry gas), it is customary to report oxygen content in ml/dL. Under
these conditions, oxygen approximates an ideal gas such that 6.02 × 1023 gas mol-
ecules (i.e. one mole) occupies 22.414 litres at 0 °C. Quantification of human oxy-
gen consumption is performed using STPD rather than BTPS (body temperature
and pressure, saturated: Eq. 1.6) [14] as water vapour in the latter partially con-
denses with increasing pressure. This results in a significant deviation from an ideal
gas and invalidates the relationship between the number of molecules and volume
defined in Eq. 1.2.

The Solubility of Respiratory Gases in Solution

The Bunsen solubility coefficient of oxygen is 0.003082 ml.dL−1.mmHg−1 and

describes the measured solubility corrected to STPD. Utilising this conversion, the
solubility coefficient of oxygen in normal plasma is 1.38 × 10−3 mmol.L−1 mmHg−1,
while carbon dioxide is nearly 22 times greater at 3.08 × 10−2 mmol.L−1.mmHg−1
[15]. Thus, using Henry’s law (Eq. 1.4b) in normal arterial blood, the concentration
of dissolved carbon dioxide is nearly ten times that of dissolved oxygen:

[O2 ] = 1.38 ×10−3 × 90 mmHg

= 0.1242 mmol.L-1 Or 0.0278 mL.dL-1
(1.5)
[CO2 ] = 3.08 ×10−2 × 40 mmHg
= 1.232 mmol.L-1 Or 2.7 mL.dL-1

It should be noted that this does not include oxygen combined with haemoglobin
and CO2 in reaction with water as bicarbonate. The significantly higher plasma con-
centration of dissolved carbon dioxide (Eq. 1.5) resulting from its greater solubility
allows for more rapid elimination by gas exchange membranes when compared to
oxygen under the same flow conditions.
While in the gaseous phase, summation of each individual gases partial pressure
will equal the total ambient pressure that the gas mixture exerts on its container,
allowing each individual gas to be reported as a fraction of the total. For example,
the partial pressure of oxygen in inhaled 37 °C air that is fully saturated with water
vapour at 1 atmosphere (i.e. BTPS) is:

PIO = ( 760 mmHg − 47 mmHg ) × FiO 2

2
(1.6)
= 149.7 mmHg for an FiO 2 of 21%

This summative requirement is only met when the solution is exposed to a gas
phase. As solubility coefficients vary between gases, the number of moles of
12 M. J. Brain et al.

dissolved gas in a given quantity of solution that is in contact with a gas phase has
no such equivalent summation.
As at equilibrium, the partial pressure of a gas is proportional to the concentra-
tion in solution, it may be used as a substitute for concentration even when no gas
phase is present, as is the case for body fluids. In this case it represents the partial
pressure that would be required of a gas phase to maintain the existing concentra-
tion of dissolved molecules in solution.
If a dissolved gas is consumed by chemical reactions in solution (e.g. aerobic
metabolism of oxygen), the equilibrium partial pressure required of a gas phase
falls. Upon exposure to a gas phase with a higher partial pressure, such as in the
lungs or an oxygenator, gas molecules will dissolve, increasing the solution concen-
tration until equilibrium is again reached. This is the primary advantage of express-
ing concentrations of dissolved gases in body fluids as partial pressures – apart from
measurement practicalities, it allows easy quantification of the concentration gradi-
ent from the site of gas exposure to the site of usage. Its inconvenience comes when
considering stoichiometric relationships such as the respiratory quotient.

Biophysics of Membrane Oxygenation

The equations and constants introduced thus far describe a steady-state where a
fixed quantity of gas is in equilibrium with a solution and assumes instantaneous
reactions occurring in stationary homogenous mediums. However, both in the
human body and in the oxygenators used for ECMO, an exchange membrane is
always interposed between the gas phase and the body fluids it dissolves in. Even in
the case of porous membranes, a direct gas/blood interface is usually prevented by
formation of a biofilm comprised of adsorbed blood proteins that adhere after a
short period of operation.
Membrane properties impose additional time constraints over which gas
exchange can occur and requires consideration of the mass transport of molecules
into the body as flux (J), defined as the passage of a quantity of solute per unit time.
The membrane flux of solute down a concentration gradient is described by
Fick’s law of diffusion [16]:

∆C
J = − D. A (1.7)
∆l

This expression says the flux (J, mmol.s−1) of a solute over the thickness of a
membrane (Δl, cm) is proportional to the diffusivity coefficient, D, the concentra-
tion gradient across the membrane (ΔC, mmol.mL−1 or mmol.cm−3) and the area of
the diffusion front (A, cm2). The minus sign is mathematically required to describe
flux from a high concentration to a low concentration [16]. This universal statement
of mass transport is applicable not only to the extracorporeal oxygenator but also to
gas transport from plasma to interstitial fluid and into cells.
1 Physiology of Extracorporeal Life Support (ECLS) 13

The diffusivity coefficient, D, is expressed as area over time (cm2.s−1) and

describes a unique constant for the specific gas, barrier and solution under steady-­
state conditions. Higher numbers represent greater diffusibility with coefficients in
gases being orders of magnitude greater than coefficients in liquids. Low molecular
weight gases diffuse more quickly than higher molecular weight gases, and higher
temperatures provide gas molecules with greater kinetic energy, increasing diffu-
sion rates [11]. This is in contrast to the solubility of gases which decreases with
higher temperature – however, it must be recalled that diffusivity specifies a transfer
rate whereas solubility describes concentrations at equilibrium.
Describing diffusion in porous media – such as membranes in hollow fibre oxy-
genators – requires more parameters to be incorporated into the constant, resulting
in effective diffusivity, DEff, that for an isolated membrane has the following
parameters:

Dδ
DEff = ε (1.8)
τ

D is the diffusion coefficient described above for the gas liquid interface within
the pores, and ε is the porosity – the fraction of the interface occupied by gas. τ is
the tortuosity – a geometric description accounting for the increased length of diffu-
sion within the membrane. δ is constrictivity, which describes resistance to mole-
cules traversing pores due to their size relative to the pore diameter (Fig. 1.3)
[17, 18].
Under real conditions the value determined for the effective diffusion coefficient
is inseparable from the properties of any biofilm of adsorbed proteins or stationary
layer of blood [11, 17]. Furthermore the greater part of oxygen traversing the mem-
brane immediately undergoes a chemical reaction with haemoglobin until the latter
is saturated. This chemical reaction sustains the concentration gradient and is
described as an enhancement factor. After incorporating haemoglobin, the DEff for
oxygen becomes not only dependent on the membrane characteristics discussed but
also a function of haematocrit (%Hct). Equation 1.9 demonstrates an example of a
term for the effective diffusivity of a membrane exposed to a turbulent bovine
bloodstream [19, 20]:

DEff = ( 2.13 − 0.0092% × Hct ) × 10−5 ( cm 2 .s −1 ) (1.9)

The Driving Force for Diffusive Transport of Gases

By combining Eq. 1.4b (Henry’s law) for solute concentration and Fick’s law
(Eq. 1.7), an equation for diffusive membrane flux where the driving force is
expressed as a partial pressure gradient can be derived where k is the permeability
constant – the product of the solubility coefficient (SC, mmol.L−1.mmHg−1) already
14 M. J. Brain et al.

defined and effective diffusivity (DEff, cm2.s−1) having the units mol.cm−1.s−1.
mmHg−1 [21]:

PO2 ( Gas ) − PO2 ( Plasma )

J = − kO2 . A (1.10)
∆l

After a brief period of operation, any extracorporeal gas exchange membrane

exposed to blood will develop a film of adsorbed blood proteins and clotting factors
(Fig. 1.3). This is unlikely to be of uniform consistency and may be thicker in areas
of the oxygenator exposed to lower flows. Similarly the total membrane area, A,
slowly decreases over the membrane’s operational life due to macroscopically vis-
ible fibrin deposition.
In spite of this gradual decline in performance, over short periods of operating
time, the average membrane thickness, l, can be considered to be constant and,
under steady-state conditions, can be combined with −kO2 into a single constant,
resulting in a quantitative statement that the flux of a material is proportional to the
driving force and is opposed by certain resistances [16] and should be familiar as
equivalent to statements relating cardiac output to blood pressure and systemic vas-
cular resistance:

Driving Force
Mass transfer per unit area =
Resistance to Transport

J (
PO2 ( Gas ) − PO2 ( Plasma ) )
( mmol.s −1
.cm −2 )
A
=
RTotal
(1.11)

∆l
where RTotal =
−kO2

From Eq. 1.11 it is apparent that maintaining the partial pressure gradient
between the gas phase and the blood phase is important in maximising oxygen flux.
Maintenance of the local pressure gradient at any point along a hollow fibre is influ-
enced by three design factors and one operational factor: the time the blood is
exposed to the membrane, relative flow direction, local turbulence and
haematocrit.

Membrane Exposure Time

For a static liquid below a static gas, the rate of diffusion will decrease exponen-
tially until equilibrium when the conditions of Eq. 1.4b are met, that is, the rate of
gas dissolving into the liquid is equal to the rate of molecules leaving the liquid
(Fig. 1.4 left panel). It is clear from this figure that too short an exposure time will
result in submaximal oxygenation.
1 Physiology of Extracorporeal Life Support (ECLS) 15

14
Partial Pressure vs Time for a static gas/liquid
O2 Content vs Membrane Exposure Time

12

Partial Pressure of Gas Phase

10
Partial Pressure Liquid Phase
Partial Pressure Gas

O2 Content mL/dL
8
Oxygen Content (ml/dL) for Hb
10 g/dL pH 7.44 Temp 37.5° & BE 0
6
Dissolved O2 in solution
4

0
0 0.5 1 1.5 2 2.5 3 3.5 4
Time
Time (seconds)

Fig. 1.4 Partial pressure vs. time for a gas dissolving in a liquid: Left panel depicts partial pres-
sures approaching equilibrium for a static solution below a gas phase. The slope of the curve (i.e.
the rate at which equilibrium is approached) is proportional to the concentration difference. Right
panel describes the oxygen content of serum containing red blood cells vs. time exposed to 100%
oxygen across a membrane. Time has been adjusted to approximate transit times in current oxy-
genators. (Plotted from equations of Katoh and Nickalls [19, 22])

The more complicated relationship when haemoglobin is present is displayed in

Fig. 1.4 right panel. Here the oxygen content of the blood displays a plateau due to
the oxygen haemoglobin dissociation curve (discussed below).
In ECMO, the time of blood exposure to the membrane is proportional to the
length of membrane/hollow fibre traversed and inversely proportional to the blood
flow rate. As the length of microtubules adds to resistance to blood flow, determin-
ing the optimal length for oxygen flux over the physiological range of blood flows
while minimising resistance to flow is an important design parameter.

Relative Flow Direction

Figure 1.4a depicts a stationary blood and gas phase; however a similar pattern
exists for two phases moving in the same direction (co-current flow), and thus at low
flow rates, equilibrium will occur, and diffusive flux will cease. Inspection of the
figure makes it apparent that replacing gas partly depleted of oxygen (where CO2
also contributes to the total partial pressure) will maintain the concentration gradi-
ent. Utilising countercurrent flow where blood and gas flow in opposite directions
decreases the maximum concentration gradient at the blood inlet end of a hollow
fibre but increases the gradient at the outlet, thereby maintaining a gradient over the
entire fibre and allowing flux to continue along the membrane relatively indepen-
dent of flow rates. Cross current flow is also utilised (demonstrated in Fig. 1.3)
resulting in differing gradients across the bloodstream.
16 M. J. Brain et al.

Effect of Turbulence and Haematocrit on Local Concentration Gradients

In an environment where oxygen exchange is occurring, the uptake of oxygen by

haemoglobin maintains diffusion in plasma toward red cells [15], and elevated hae-
matocrit increases the flux of oxygen into the blood (Fig. 1.4b). Creation of turbu-
lent flow vortices more effectively purges a pipe of its contents than laminar flow
(discussed below) and brings erythrocytes into closer proximity to membranes,
increasing the local concentration gradient.

Resistances to Diffusive Transport

The total resistance to mass solute movement in Eq. 1.11, RTotal, is the sum of com-
ponent resistances, which can be divided into gas phase resistance (RG), membrane
resistance RM and blood side resistance, RB: RTotal = RG + RM + RB. Of these RG is
negligible and the factors influencing RM have been discussed. The most variable
and often most prominent component is RB due to the formation of a stationary film
on the blood side of the membrane (Fig. 1.3 boundary layer).

Area of the Gas Exchange Membrane

It is not practical for manufacturers to specify a membrane surface area (A, Eq. 1.11)
that encompasses the complex microscopic geometry of oxygenator tubules and
pores, the latter being incorporated into DEff as discussed [17]. Furthermore, it can-
not be assumed that any oxygenator design utilises the entire membrane area evenly.
Thus, while flux per unit area (J/A) is a useful description of isolated membrane
performance, oxygenators are better characterised by their total flux incorporating
area into the equation for resistance:

(P − PO2 ( Plasma ) )
( mL.min ) J
−1 O2 ( Gas )
=
RTotal
Oxygenator
(1.12)
∆l
where RTotal =
− kO2 . A

Although lacking the precision of Eq. 1.11 in defining properties of the mem-
brane, Eq. 1.12 can be incorporated into monitoring gas exchange efficiency of an
individual oxygenator over time and will be discussed after oxygen carriage is con-
sidered. The value of JOxygenator is usually reported in oxygenator product specifica-
tion sheets at varying blood flows.
1 Physiology of Extracorporeal Life Support (ECLS) 17

 ltrafiltration of Plasma Water Over the Oxygenator

U
Membrane [16]

Analogous to the membrane flux of oxygen down a concentration gradient is the

movement of water from the plasma to the gas partition of the oxygenator. The driv-
ing force is hydrostatic pressure and is generally defined as a conductance (the
inverse of resistance) termed the coefficient of ultrafiltration (KUF, mL.min−1.
mmHg−1):

water flux
K UF =
transmembrane pressure
(1.13)
QUF
=
PBlood − PGas

QUF (mL.min-1) is termed the ultrafiltration rate and describes the appearance of
fluid within the hollow fibres, while the terms for pressure (PBlood and Pgas) describe
the heights of a fluid column relative to atmospheric pressure in each compartment.
Many of the factors already described for resistance to diffusion will be contained
in the ultrafiltration coefficient and will not be discussed further. The ultrafiltrate
represents a homogenous fluid that will contain dissolved solutes from plasma pro-
portional to the size of membrane pores, which are typically smaller than 1 micron.
A major drawback of early microporous membranes was significant plasma
leakage, as the open porous structure allowed water flux. This has been significantly
alleviated by newer closed-fibre membranes; however, some water flux still occurs
under normal operating conditions. The water evaporates in the fresh gas flow and
leads to an insensible water loss proportional to the fresh gas flow and may reach
significance when supporting low bodyweight patients [12].

Membrane Oxygen Flux Matches Demand

Oxygenation has been described as a cascade of partial pressure gradients from the
external environment into the arterial blood and from capillaries into the extracel-
lular space, before finally diffusing across cell membranes into mitochondria.
The quantity of oxygen entering the body at steady state equals its consumption.
If no consumption occurs, no arterial-venous oxygen difference will exist, and the
circulating blood will be in equilibrium with the oxygen in the gas phase as per
Eqs. 1.3 and 1.10, abolishing further mass transfer. Oxygen will continue to diffuse
across the oxygenator; however, the movement will be in both directions at an
equal rate.
18 M. J. Brain et al.

Oxygen Transport

As outlined above, oxygen has a limited solubility in plasma of only 1.39 ×

10−3 mmol.mmHg−1 or 0.0031 mL.dL−1.mmHg−1 at 37°. For a normal arterial oxy-
gen partial pressure of 100 mmHg, this equates to about 3 mL of dissolved oxygen
per litre of blood. At that oxygen content, maintaining a nominal body oxygen con-
sumption of 250 mL.min−1 would require a cardiac output of 80–120 litres per min-
ute. Even breathing 100% oxygen at normal atmospheric pressure would not sustain
aerobic cellular metabolism, providing only 20 mL of oxygen per litre of blood [15].

Oxygen Carriage

Oxygen carriage in blood is augmented by the presence of haemoglobin, a complex

metalloprotein consisting of four subunits that alters quaternary structure in response
to physiologic stimuli including its own ligand, oxygen [4]. This rapid structural
alteration changes the affinity of haemoglobin for oxygen, resulting in higher affin-
ity in areas where oxygen is abundant and lower affinity in areas of oxygen con-
sumption. A normal adult has a haemoglobin concentration of between 12 and 16 g/
dL. However, it is quite common for anaemia to be present in critically ill patients,
and those receiving ECMO may have haemoglobin concentrations below 10 g/dL or
0.155 mmol.L−1 (assuming a molecular weight of haemoglobin of 64,458 g/
mol [23].
Haemoglobin is a spherical molecule consisting of four globin subunits (2 α and
2 β chains) with each globin containing a heme group in a peripheral molecular
crevice. Each heme molecule consists of a central iron atom in the ferric (Fe2+) state
between two histidine amino acids. This structure allows the iron atom to bind oxy-
gen without being oxidised to Fe3+, a change that would prevent further oxygen
binding. Haemoglobin demonstrates cooperative binding whereby the binding of
oxygen to the ferric moieties is enhanced if another binding site on the same mole-
cule is already occupied by oxygen. As each haemoglobin molecule has four bind-
ing sites, it can exclusively be 0%, 25%, 50%, 75% or 100% oxygenated.
Haemoglobin saturation refers to the fractional occupancy of all the oxygen binding
sites in a solution and due to cooperative binding results in the sigmoid haemoglo-
bin dissociation curve (Fig. 1.5) [4].
Apart from the conformational change induced by oxygen itself, four other major
factors influence the conformational state of haemoglobin – carbon dioxide, hydro-
gen ion concentration (pH), 2,3-DPG and temperature. By altering the affinity of
haemoglobin for oxygen, each of these factors affects how saturated the haemoglo-
bin in a given quantity of blood is with oxygen at any partial pressure and thus the
oxygen content of that blood. If all haemoglobin binding sites are occupied by oxy-
gen (100% saturation), the maximum oxygen carrying capacity is 1.39 mL per gram
1 Physiology of Extracorporeal Life Support (ECLS) 19

Oxygen Hemoglobin Dissociation and Oxygen Content

1
20
0.9

0.8

15
0.7

Blood Oxygen Content ml/dL

Hemoglobin Saturation %

0.6

0.5
10

0.4

0.3
SaO2 at pH 7.44 Temp 37.5° & BE 0
5
0.2 SaO2 at pH 7.32 Temp 37.5° & BE 0

Oxygen Content (ml/dL) for Hb 10 g/dL pH 7.44

0.1 Temp 37.5° & BE 0
Oxygen Content (ml/dL) for Hb 10 g/dL pH 7.32
Temp 37.5° & BE 0
0 0
0 20 40 60 80 100 120
Partial Pressure Oxygen (mmHg)

Fig. 1.5 The oxygen haemoglobin dissociation curve as calculated by the Thomas modification of
the Kelman Eq. 1.9. Also shown is the oxygen content for haemoglobin concentration of 10 g/dL
after applying Eq. 1.14. The effect of pH is demonstrated

of haemoglobin in adults [14] or 1.312 mL per gram of foetal haemoglobin [15].

Thus, the total oxygen content of adult arterial blood (CaO2) can be described by:

CaO 2 = Hb × SaO 2 × 1.39 + 0.0031× PaO 2 ( mL.dL−1 ) (1.14)

The total oxygen content can then be multiplied by blood flow (cardiac output)
to give oxygen delivery, DO2 which has the units of flux: mL.min−1. Note the scaling
factor of 10 to convert the units of oxygen content to ml.L−1.
⋅
DO2 = QB × Ca O2 × 10 (1.15)
20 M. J. Brain et al.

Modulation of Haemoglobins Affinity for Oxygen

The Bohr effect describes alterations in haemoglobin oxygen affinity due to carbon
dioxide and hydrogen ion concentrations. Carbon dioxide binds to amino acids in
the outer chains of haemoglobin to form carbaminohaemoglobin, stabilising the
molecule with the ferric elements in deeper crypts and facilitating release of oxygen
from haemoglobin. Similarly, increasing temperature and increasing hydrogen ion
concentrations stabilise haemoglobin in the deoxygenated state. Similar to oxygen,
carbon dioxide binding is reversible, and in compartments with a low CO2 concen-
tration, the effect is reversed, promoting oxygen uptake (the Haldane effect). This is
of importance when considering oxygenation in systems designed primarily for
CO2 removal and is discussed below.
The glycolysis product, 2,3-DPG also decreases haemoglobin affinity for oxy-
gen. 2,3-DPG binds to deoxygenated haemoglobin, lowering the apparent affinity
for oxygen by altering the electrostatic bonds that maintain the quaternary configu-
ration [10]. This has the most significance in stored blood where after 2 weeks,
2,3-DPG levels become negligible. After transfusion, DPG levels do not return to
normal until nearly 48 h [24]. In the absence of 2,3-DPG, the affinity for oxygen is
increased resulting in relatively lower oxygen release in tissues for the same pO2.
This phenomenon is only likely to be a factor in the most severe oxygenation prob-
lems, and the benefit of increased oxygen binding sites from transfusion probably
outweighs the transiently lower delivery (see Figs. 1.6 and 1.7: Venous oxygen satu-
ration vs. haemoglobin and cardiac output in a patient receiving VV-ECMO. VO2 is
held constant [25]).

ECMO Return Pulmonary Artery Arterial Oxygen

Cannula Saturation Saturation Saturation

Deoxygenated
Portion of
Venous Return

ECMO Fraction Total Venous Return (Mixed Venous Saturation)

of Venous Return = Cardiac Output

Fig. 1.6 Schematic of a VV-ECMO circuit. A mixture of fully oxygenated venous and deoxygen-
ated venous blood enters the right heart and perfuses the pulmonary circulation before the left heart
distributes it systemically (hence end-tidal CO2 will be not reflect true mixed venous pCO2).
Varying with the cannula position and venous return rate, some amount of recirculation is very
common, reducing the amount of oxygen delivered. Depending on recirculation, it may not be
possible to sample true mixed venous blood
1 Physiology of Extracorporeal Life Support (ECLS) 21

VV-ECMO Saturations vs Cardiac Output. VV-ECMO Flow 3

VV-ECMO Saturations vs Cardiac Output. VV-ECMO Flow 5 L/min, Required VO2 250 ml/min and Hb 10 g/dL
L/min, Required VO2 250 ml/min and Hb 10 g/dL
120% 400.00 120% 400.00

SvO2 SaO2 VO2 Supplied SvO2 SaO2 VO2 Supplied

100% 100%
350.00 350.00

80% 80%

300.00 300.00
Hb Saturation

Hb Saturation
VO2 ml/min

VO2 ml/min
60% 60%

250.00 250.00

40% 40%

200.00 200.00
20% 20%

0% 150.00 0% 150.00
2.00 3.00 4.00 5.00 6.00 7.00 2.00 3.00 4.00 5.00 6.00 7.00

Cardiac Output L/min Cardiac Output L/min

VV-ECMO Saturations vs Cardiac Output. VV-ECMO Flow 4 VV-ECMO Saturations vs Cardiac Output. VV-ECMO Flow 4
L/min, Required VO2 250 ml/min and Hb 10 g/dL L/min, Required VO2 250 ml/min and Hb 7 g/dL

120% 400.00 120% 400.00

SvO2 SaO2 VO2 Supplied SvO2 SaO2 VO2 Supplied

100% 100%
350.00 350.00

80% 80%

300.00 300.00
Hb Saturation
Hb Saturation

VO2 ml/min
VO2 ml/min

60% 60%

250.00 250.00

40% 40%

200.00 200.00
20% 20%

0% 150.00 0% 150.00
2.00 3.00 4.00 5.00 6.00 7.00 2.00 3.00 4.00 5.00 6.00 7.00

Cardiac Output L/min Cardiac Output L/min

Fig. 1.7 Arterial (SaO2) and central mixed-venous oxygen saturations (SvO2) vs. cardiac output in
a patient fully supported by VV-ECMO. Graph titles show physiologic and flow conditions along
with haemoglobin. See text for discussion [25]

The Oxygen-Haemoglobin Dissociation Curve

The oxygen-haemoglobin dissociation curve is characterised by an upper plateau at

higher partial pressures of oxygen where haemoglobin is between 90% and 100%
saturated. Below this a steep shoulder occurs, and the saturation of haemoglobin
rapidly decays as the partial pressure of oxygen falls. Physiologically the factors
that shift the curve to the right (i.e. to a lower affinity at a given partial pressure) all
derive from local tissue metabolism and hence facilitate increased delivery where
higher local oxygen consumption occurs (Fig. 1.5).
Several equations exist to model the normal oxygen haemoglobin dissociation
curve. One of the most informative is the Thomas modification [26] of the Kelman
[27] equation and its inverse [22], which calculate the haemoglobin saturation for
22 M. J. Brain et al.

any partial pressure of oxygen and allow for shifts of the curve due to temperature,
[H+] and carbon dioxide.
For convenience, the oxygen-haemoglobin dissociation curve is frequently
described by the partial pressure at which 50% of haemoglobin is saturated. A nor-
mal p50 for arterial blood is 26.3 mmHg. Values higher than this describe a ‘right-­
shifted’ curve, that is haemoglobin affinity for oxygen is less. Two curves of
haemoglobin saturation are demonstrated in Fig. 1.5, the only difference being the
hydrogen ion content reflecting the higher carbon dioxide concentration in venous
blood. At high oxygen partial pressures consistent with arterial blood, the difference
in haemoglobin saturation between curves with differing p50 is minimal. In con-
trast, there is a significant difference in the saturation of Hb at a partial pressure of
40 mmHg commonly found in venous blood.
The implications of the oxygen Hb dissociation curve become clearer when oxy-
gen content is also plotted on the same chart (Fig. 1.5). At haemoglobin of 10 g.
dL−1, the oxygen content at a partial pressure of 100 mmHg is 13 mL.dL−1 and mini-
mally affected by the arteriovenous pH difference. However, at an oxygen partial
pressure of 40 mmHg, a 1 mL.dL−1 difference becomes apparent between the two
content curves, being 10 mL.dL−1 at pH of 7.44 and 9 mL.dL−1 at pH of 7.32. In the
tissues where oxygen is utilised in metabolism, this ‘right shifting’ of the curve as
the products of metabolism acidify capillary blood serves to bolster the partial pres-
sure gradient for oxygen diffusion from capillary to cell.
Unfortunately, ‘right-shifted’ oxygen haemoglobin dissociation curves, while
advantageous for unloading oxygen in acidotic tissues, may be counterproductive at
sites of oxygen uptake if abnormally low alveolar oxygen partial pressures exist as
in the adult respiratory distress syndrome (ARDS). Inspection of Fig. 1.5 reveals
that if oxygen uptake were to occur at a partial pressure of 60 mmHg, a right shifted
curve (pH 7.32) will carry 0.5 mL.dL−1 less oxygen. This gap widens if uptake
occurs at even lower partial pressures.

Mixing Blood of Differing Oxygen Partial Pressures

The effects of the oxygen haemoglobin dissociation curve is of importance when

considering mixing bloodstreams with differing oxygen concentrations. If the volu-
metric flow rate of both streams is similar, then the haemoglobin saturation of the
two streams can be averaged as a reasonable approximation of the resulting mixture.
However, at differing flows, accurately calculating both the resultant oxygen ten-
sion and saturation requires conversion to oxygen content and measurement of the
flow rate of the two streams. As mixing oxygenated blood is fundamental to ECMO,
the steps of this process will be worked through (Table 1.1).
As demonstrated in Table 1.1, the average of the venous and arterial blood satu-
rations approximates the complete solution suitably when the flows are similar but
overestimates saturations when the volumetric flow differs. The approximation
worsens if the venous oxygen tension is reduced further. In contrast, taking the
1 Physiology of Extracorporeal Life Support (ECLS) 23

Table 1.1 Mixing blood streams with differing oxygen concentrations

Volume Hb O2 tension Arterial saturation O2 content
Equal flows (mL) (g) (mmHg) (%) (mL)
Venous blood 100 10 40 76% 10.07
Oxygenated 100 10 100 98% 13.09
blood
Mean (for comparison) (70) (86.8%)
Mixed blood 200 20 23.17
(= 2 dL) Final O2 content (mL/dL) 11.58
Final %saturation 87%
Final partial pressure 51.7
Unequal flows Volume Hb O2 tension Arterial saturation O2 content
(mL) (g) (mmHg) (%) (mL)
Venous blood 3000 300 30 58% 229.12
Oxygenated 2000 200 300 100% 262.90
blood
Mean (for comparison) (170) (88.0%)
Mixed blood 5000 500 492.02
(=50 dL) Final O2 content (mL/dL) 9.84
Final arterial saturation (%) 73%
Final partial pressure (mmHg) 38
Two examples of mixing blood streams are given, one with matched input flows and one with dif-
fering input flows. For simplicity only the volume of the bloodstream is described; however, it can
be assumed that the blood flow rate is this volume per minute. In all calculations the haemoglobin
is assumed to be 10 g/dL and is multiplied by the blood volume to give the total mass of haemo-
globin. In the first example the flows have been set to 1 dL/min so that values for haemoglobin and
oxygen content equate to values shown in Fig. 1.5. The mean saturation and tension is shown for
comparison to the result after converting to content

mean O2 tensions of the unmixed samples massively overestimates the final partial
pressures after mixing.
Understanding this concept is important as it highlights a physical limitation on
systemic oxygenation: utilising a saturable oxygen carrier (haemoglobin) makes
oxygen delivery flow limited. Even if supranormal oxygen tensions are achieved via
an extracorporeal circuit, an inadequate ratio of circuit flow to cardiac output results
in suboptimal oxygen delivery.

Venovenous ECMO and Oxygen Transport

In venovenous (VV) extracorporeal membrane oxygenation blood is accessed and

returned to the venous system/right atrium (Fig. 1.6); systemic oxygen delivery thus
remains dependent on cardiac output, making this system relatively straightforward
to quantitatively analyse.
24 M. J. Brain et al.

Various access configurations are utilised (Fig. 1.1); however, one of the most
common techniques is to cannulate the common femoral vein with a cannula which
has side and end fenestrations that allow blood to be drained from multiple points.
So called ‘multistage’ cannula where the side holes extend over 20–25 cm can be
placed via the inferior vena cava, across the right atrium and into the superior vena
cava to maximise harvest of venous return. If the access cannula only has openings
near the tip, then the tip is typically placed 5–10 cm below the cavo-atrial junction.
The return cannula has a single terminal orifice which terminates in the right atrium.
If higher extracorporeal circuit flows are required, the inferior vena cava may col-
lapse around the multistage cannula, intermittently restricting flow and causing the
external circuit to ‘shudder’. If higher circuit flows are necessary to achieve ade-
quate oxygenation and this negative access pressure cannot be resolved by giving
fluid, a second access cannula may need to be placed in the superior vena cava via
the internal jugular vein.
Figure 1.6 demonstrates a basic VV-ECMO circuit configuration. It can be appre-
ciated that the ECMO circuit is in parallel to venous return and thus a mixture of
oxygenated blood from the ECMO circuit and deoxygenated blood from the venous
return will enter the right heart. To appreciate the implications of this parallel circuit,
consider a young adult with severe acute respiratory distress syndrome fully sup-
ported by VV-ECMO with both femoral and internal jugular access (Fig. 1.6). Chest
x-ray demonstrates bilateral ‘white-out’, and it will be assumed the lungs are not
contributing to systemic oxygenation.
If the patient has the following parameters: cardiac output of 4 L/min, ECMO
flow of 4 L/min, a mixed central venous haemoglobin oxygen saturation of 52% and
an arterial saturation of 99%. What will be the effect on his saturations if his cardiac
output were to rise to 7 L.min−1 or were to fall to 2.5 L.min−1?
Apart from the non-contribution of the lungs, for this analysis other assumptions
include an oxygen-haemoglobin dissociation curve with a normal p50 and good
cannula position with minimal recirculation between the access and return lumens
in the inferior vena cava. To fully develop this model and illustrate several key
points, two other parameters are required: the haemoglobin concentration and the
patient’s total oxygen consumption (VO2). Initially, this hypothetical patient has a
haemoglobin concentration of 10 g/dL and is at steady state consuming 250 mL/min
of oxygen with no markers of tissue hypoxia. Thus, there are four independent vari-
ables: the cardiac output, the ECMO flow rate, the haemoglobin concentration and
the target VO2.
While providing circulating oxygenated blood is the goal of ECMO, it must be
highlighted that the VO2 required to avoid anaerobic metabolism is a parameter that
can only be achieved if the amount of oxygen delivered matches consumption. If
this required VO2 exceeds tissue delivery, VO2 is then supply-limited, initially
resulting in higher oxygen extraction with mixed-venous saturation decreasing first,
followed by clinical and biochemical markers of hypoxia such as confusion, oligu-
ria and rising lactate as anaerobic metabolism ensues.
Figure 1.7a demonstrates that under these conditions, increasing the cardiac out-
put from 4 L.min−1 to 7 L.min−1 will cause a drop in the arterial saturations from
1 Physiology of Extracorporeal Life Support (ECLS) 25

99% to 84%. In contrast dropping the cardiac output to 3 L.min−1 won’t affect the
arterial saturation but will cause the central mixed-venous saturation to fall from
52% to 25% as increased tissue extraction occurs.
The equations required to generate this model are the content and delivery equa-
tions (Eqs. 1.14 and 1.15) from which the principle of conservation of mass is
applied to determine the oxygen content at each of the following points: the ECMO
return cannula, the venous return and the pulmonary artery (Fig. 1.6). In VV-ECMO
the pulmonary artery is not the correct sampling site for mixed-venous saturation,
instead the pre-oxygenator blood tubing is the closest approximation.
The methods described in Table 1.1 for mixing bloodstreams are used to deter-
mine the pulmonary artery oxygen content and saturations. In the absence of lung
function, this represents systemic arterial oxygen delivery as well. Modelling the
solution requires an iterative approach to determine the highest achievable VO2 (if
the required VO2 cannot be met) by altering the tissue oxygen extraction – this
determines the venous oxygen flux to the right heart. Example values from Fig. 1.7a
are shown in Table 1.2.

 rterial Saturations Are Dependent on the Fraction of Cardiac

A
Output Captured

To explain the fall in arterial saturations with increasing cardiac output, it should
be appreciated that the required VO2 did not change. What did change was the
total venous return, which increased by 75% with the increased cardiac output.
The ECMO flow remained at 4 L.min−1, so the ‘shunt’ bypassing the ECMO oxy-
genator via the great veins increased resulting in the fraction of deoxygenated
venous blood in the right ventricle increasing from 3% to 43%. The shunt is cal-
culated as (cardiac output – oxygenated blood flow)/cardiac output. The oxygen-
ated blood flow equals the cardiac output if the cardiac output is less than ECMO
blood flow.
In summary, the resulting arterial saturation is analogous to the example of mix-
ing bloodstreams of different content, and the same effect can be achieved if cardiac
output is left constant and ECMO flow is decreased (Fig. 1.7b). In practice, increases
in cardiac output are likely to be accompanied by increased VO2, and this will result
in a fall of mixed-venous saturation (Fig. 1.7c).

Mixed-Venous Saturations Are Determined by VO2 and Cardiac Output

To appreciate why the mixed-venous saturations fall at low cardiac outputs in

Fig. 1.7a, recirculation between the access and return cannula must be considered.
Several factors can influence recirculation; however, our described model incor-
porates only cardiac output: If the heart stops, there will be no venous return, and
recirculation might approach 100% (in reality collapse of the great vessels will
occur first), that is blood will enter the ECMO circuit from the IVC and return via
 26

Table 1.2 Model output for solutions in Fig. 1.7a

Flux of fully Cardiac Total O2
Oxygenated blood oxygenated blood output/ Venous O2 Fraction of venous flux
ECMO flow (factors delivered by venous flux return (deoxygenated) leaving CaO2-­
set rate recirculation) ECMO return to heart CvO2 SvO2 blood leaving RV RV CaO2 SaO2 CvO2 VO2
ml/
L/min L/min ml/min L/min ml/min ml/L % % ml/min ml/L % ml/L min
4.0 1.5 201.0 1.5 0.0 0.0 0 0 201.0 134.0 100 134.0 201.0
4.0 2.5 335.0 2.5 85.0 34.0 25 0 335.0 134.0 100 100.0 250.0
4.0 3.9 522.6 4.0 279.6 69.9 52 3 529.6 132.4 99 62.5 250.0
4.0 4.0 536.0 7.0 500.5 71.5 53 43 750.5 107.2 80 35.7 250.0
Calculations performed with haemoglobin of 10 g.dL−1. Oxygenated blood flow refers to the volumetric flow of blood oxygenated blood into the right heart.
Note for calculation purposes, oxygen content is expressed in ml.L−1 rather than ml.dL−1
M. J. Brain et al.
1 Physiology of Extracorporeal Life Support (ECLS) 27

the right atrium, flowing in a retrograde fashion back down the proximal IVC while
no systemic oxygenation occurs. Similarly at low cardiac outputs, there will be
atrio-caval recirculation; the haemoglobin however is saturated with oxygen on the
first pass and will be unable to carry more. Thus when ECMO flow exceeds cardiac
output, the saturation is cardiac output limited. To supply the bodies required VO2
at low cardiac outputs, oxygen extraction has to increase, and the mixed central
venous oxygen content and venous saturation fall.
At high cardiac outputs and hence high venous return, recirculation will be sig-
nificantly lower as the oxygenated blood will be pushed through the right atrium
into the right ventricle. Figure 1.7c demonstrates falling mixed-venous saturations
with increasing VO2 requirements at higher cardiac outputs. This is caused by a
combination of falling arterial oxygen content by the reducing ECMO oxygenated
proportion of the venous return and increased peripheral extraction to achieve the
higher VO2 requirement.

The Effect of Oxygen-Carrying Capacity

The parameters in Fig. 1.7d are identical to Fig. 1.7a except for a lower haemoglo-
bin. The lower oxygen-carrying capacity means that at any cardiac output, the oxy-
gen extraction must be greater and hence the venous saturation will be lower. At
lower cardiac outputs, the combined effect is enough that no further oxygen extrac-
tion can occur and the achieved VO2 falls lower than the target VO2 – under these
conditions signs of tissue hypoxia will occur. Thus in the setting of low cardiac
output and borderline oxygenation, increasing the haemoglobin by transfusion may
alleviate hypoxia, while consideration is given to augmenting circulatory support.

Modelling VV-ECMO

The above analysis can be combined to display informative mixed central venous
and arterial saturations for any cardiac output/haemoglobin concentration (Figs. 1.8
and 1.9).
The surface plot for arterial oxygen saturations (Fig. 1.8) indicates that bolster-
ing haemoglobin to improve arterial oxygen saturations will only be of significant
benefit at higher cardiac outputs and even then cannot fully compensate for the
shunt past the ECMO circuit.
In contrast, Fig. 1.9 demonstrates venous saturations always increase with more
oxygen-carrying capacity. The left side of the venous saturation surface plot further
emphasises that oxygen delivery is dependent on cardiac output and falls when
output is less than ECMO flow rates.
28 M. J. Brain et al.

Fig. 1.8 Arterial oxygen saturation vs. haemoglobin and cardiac output in a patient receiving
VV-ECMO. VO2 is held constant [25]

Fig. 1.9 Venous oxygen saturation vs. haemoglobin and cardiac output in a patient receiving
VV-ECMO. VO2 is held constant [25]
1 Physiology of Extracorporeal Life Support (ECLS) 29

Further Consideration of Recirculation

In the above analysis, recirculation is primarily considered in terms of the cardiac

output and is assumed negligible when cardiac outputs are greater than ECMO flow.
However, recirculation in VV-ECMO occurs for other reasons and is almost always
present to some degree. Technical factors include correctable issues such as cathe-
ters being in too close proximity to each other or suboptimal positioning (e.g. jugu-
lar placement into a hepatic vein rather than the superior vena cava). However, other
factors such as the return jet impacting the atrial septum rather than being directed
at the tricuspid valve are more difficult to correct and will result in some retrograde
flow in the great veins through part of the cardiac cycle.
In general the tricuspid valve will be closed for most of ventricular systole, and
in the period of tricuspid closure, the continuous ECMO flow can only recirculate
through the right atrium and vena cava (Fig. 1.10).

The Importance of Mixed-Venous Saturation

The ability to conveniently measure the arterial oxygen saturation non-invasively in

real time via finger oximetry is of considerable clinical utility. What is being mea-
sured with a transcutaneous pulse oximeter is the difference in optical density of
transmitted light in pulsatile blood which is usually arteriolar rather than capillary.
Under some conditions such as low cardiac output, poor peripheral perfusion or
VA-ECMO, the peripheral circulation may lack detectable pulsatility, and these
monitors may struggle to provide accurate readings. In these situations, blood gas
monitoring (which usually provides a calculated saturation based on the partial
pressure, temperature, pH and CO2) is essential.

Fig. 1.10 Recirculation due to tricuspid closure. The left hand image demonstrates recirculation
in the atria after tricuspid valve closure at the beginning of ventricular systole (right ventricle at
top, atrium containing colour flow that is impacting the tricuspid valve). The right hand image
shows the return cannula in the superior vena cava (bicaval cannula) with blood flowing across the
tricuspid valve into the right ventricle. (Image M. Brain by permission of The Alfred Intensive
Care Unit, The Alfred Hospital, Melbourne)
30 M. J. Brain et al.

In the ECLS setting, the primary concern is optimising oxygen delivery within
the limits of the available system. Equations 1.16 and 1.17 describe oxygen content
and delivery. From these equations and Fig. 1.8, it can be appreciated that increasing
the haemoglobin concentration accommodates lower arterial saturations while aug-
menting mixed-venous saturation by increasing oxygen content and thus delivery.
This is important when other limitations such as ECMO flow rates or problems with
recirculation prevent achievement of higher saturation readings.

Venous Oxyhaemoglobin Saturation

The pulmonary arterial mixed-venous oxygen saturation (or in VV-ECMO the pre-­
oxygenator saturation) is an important measure of oxygen uptake. By utilising
blood gas analysis sampled from one of these sites, the central venous oxygen con-
tent (CvO2) can be calculated, again using Eq. 1.14. This concept has led to classic
descriptions of supply-dependent and independent oxygen uptake, and it is informa-
tive when interpreting mixed-venous saturations to calculate the content difference
with an approximation of cardiac output.
A critical point to recall when interpreting mixed-venous oxygen saturations is
that it does not represent the lowest peripheral saturation (see mixing blood of dif-
fering content above). Different organs (and in cases distinct parts of organs) will
have different metabolic activity and hence oxygen extraction. Those organs with a
high extraction ratio may still become supply-limited and thus hypoxic even when
the total organism DO2 exceeds VO2. What is actually important is the partial pres-
sure of oxygen in the capillary bed furthest from an arteriole in the relevant organ,
as this defines the lowest local oxygen gradient from blood to cells. Thus clinical
markers of hypoxia such as neurological status and renal function, along with lac-
tate trends, become as important as mixed CvO2. In the absence of practical ways to
measure either capillary oxygen tension or organ oxygen uptake, the aim should be
to target a venous oxygen content that supports higher oxygen extraction by some
organs, and this can be achieved by supplying more oxygen, increasing cardiac
output or increasing the haemoglobin.

Carbon Dioxide Physiology

An adult consuming 300 mL.min−1 of oxygen with a respiratory quotient of 0.83

will produce 250 mL.min−1 or 360 L.day−1 of carbon dioxide via anaerobic metabo-
lism, and this volume must be removed from the body. The properties of carbon
dioxide make its clearance more favourable than oxygen uptake.
1 Physiology of Extracorporeal Life Support (ECLS) 31

Carbon Dioxide Transport in Physical Solution

The solubility of carbon dioxide in plasma is described by Henry’s law (Eq. 1.4b)
with a value for SC at 37 °C of 0.0308 mmol.L−1.mmHg−1 [15]. At a pCO2 of
40 mmHg, this equates to 1.232 mmol.L−1 or 2.7 mL.dL−1 of dissolved carbon diox-
ide. The solubility decreases to 2.88 × 10−2 mmol.L−1.mmHg−1 at 40 °C.
Like oxygen, only a small proportion of carbon dioxide is transported as dis-
solved gas, the remainder being in chemical equilibrium with the gas phase through
reactions with water for a total blood content of nearly 500 mL.L−1. In blood, these
reactions largely occur in erythrocytes where the enzyme carbonic anhydrase (CA)
is abundant [28]; however, CA is distributed in many other tissues and in pulmonary
and renal capillaries [15].

Reactions of Carbon Dioxide in Solution

The following reactions describe the hydration of dissolved carbon dioxide with
arrows, indicating where the majority of reactants are at equilibrium. Equilibrium
concentrations are independent of the rate at which equilibrium is achieved. Though
not a hydration reaction, the reversible reaction of dissolved CO2 with amino acids
on haemoglobin is also listed here as part of CO2 storage:

CO 2 ( gas ) + H 2 O 
 CO 2 ( dissolved ) + H 2 O a. Dissolving in water
C. A .
CO 2 ( dissolved ) + H 2 O  H 2 CO3 b. Formation of carbonic acid (1.16)
C. A .
C. A .
CO 2 ( dissolved ) + OH −  HCO3− c. Reaction with hydroxyl ion
C. A .
+ −
H 2 CO3  H + HCO 3 d. Dissociation to bicarbonate
CO 2 ( dissolved ) + R − NH 2  R − NH − COO − + Η + e. Formation of carbamino groups

The equilibrium constant for the dissociation of carbonic acid to bicarbonate is

so small (6 × 10−11 mol.L−1) that plasma carbonic acid concentration is near unde-
tectable. No gas phase exists in blood, so all available carbon dioxide is dissolved
(the pCO2 of blood is the hypothetical value that would occur if a gas phase existed),
and thus the hydration of dissolved carbon dioxide (Eq. 1.16b-d) in the body can be
summarised into a single equation and the equilibrium constants combined (K′)
such that [13]:
32 M. J. Brain et al.

CO 2 ( dissolved ) + H 2 O  H + + HCO3− a. Summary Equation

K′
[CO2 ] × + =  HCO3−  b. Equilibrium Reaction
Η 
(1.17)
 HCO3− 
pH = pK + log
'
c. Logarithmic Form
[CO2 ]
Where [ CO 2 ] = 0.0308 × PCO2 at 37°. ( Henry 's Law )

K′ describes the compound of multiple equilibriums in Eq. 1.16 and is derived

experimentally. Notably, [H2O] is excluded from Eq. 1.17b1 because it is incorpo-
rated into K′. This is permissible because water concentration is proportionally so
much greater than the other molecules that any consumption of H2O in chemical
reactions produces a negligible effect on its total concentration.
Although less quoted than the logarithmic form of the Henderson-Hasselbalch
equation, Eq. 1.17b is most useful in understanding carbon dioxide in body fluids.
For plasma with a [H+] of 3.98 × 10−8 mol.L−1 (pH of 7.4) at 37°, the apparent K′ is
8.13 × 10−7 Eq.L−1.mmHg−1 (pK’ 6.09), and thus K′/[H+] is around 20, that is the
bicarbonate concentration is 20 times the dissolved carbon dioxide concentration.
Over the physiologic range, K′/[H+] varies with temperature and electrolyte bal-
ance. However, it is always greater than 1; thus the total CO2 contained in interstitial
fluids, blood plasma and erythrocyte water is largely stored as bicarbonate ions
(Fig. 1.12).

Effect of the Hydrogen Ion Concentration

Being lipid soluble carbon dioxide can be considered to diffuse through all mem-
branes with local production determining regions of higher concentrations, while
areas of gas exchange have the lowest concentration. This is in contrast to HCO3−
which cannot easily cross membranes unless being exchanged for another anion
(Fig. 1.11).
Of the molecules in the Henderson-Hasselbalch equation (Eq. 1.17a), CO2 is the
only independent variable with the amount in the body being determined by the

1
Occasionally the term αPCO2 is used in this equation to describe the total concentration of CO2
and carbonic acid however the concentration of the latter is orders of magnitude smaller than the
former. Thus [CO2] can be calculated accurately from Henry’s Law without alteration of the solu-
bility constant (Eq. 1.4b).
1 Physiology of Extracorporeal Life Support (ECLS) 33

CO2

ADP
ADP

Energy Glucose
Expenditure
ATP
ATP

O2

CO2 CO 2 CO 2
+ + +
H 2O H2O H 2O

C.A.

HCO3 - + H + HCO3 - + H + HCO3 - + H +

Lactate Lactate

- (I)
O O
- (I)
+ H+ O O + H+

H 3C OH H 3C OH
Lactic Acid Lactic Acid

Fig. 1.11 Dissolved carbon dioxide diffuses across all tissue planes and dissociates in each com-
partment to HCO3− with the ratio of CO2 to HCO3− depending on the [H+]

balance of production and flux out of the body [13]. The amount of hydrogen ions
and bicarbonate in any compartment depend not only on the reactions of carbon
dioxide with water but also on the concentrations of other strong and weak electro-
lytes with the final balance being determined by the need to maintain electrical
neutrality and the ionisation constant of water [13, 34]. The simplest summary of
this complex interaction is that bicarbonate behaves as an electrical ‘spacer’ in the
following equation for electrical neutrality:
34 M. J. Brain et al.

30 35
Oxygenated Blood [CO2]Total (mmol.L-1)
AV Difference (mmol.L-1)
Deoxygenated Blood [CO2]Total (mmol.L-1)
30
25

Components of Total CO2 by compartment (mmol)

25

20
[CO2]Total (mmol.L-1 whole blood)

Carbamino CO2 (mmol) 20

15

Erythrocyte HCO3- 15
(mmol)

10

10

5 Plasma HCO3- (mmol)

5

Dissolved CO2 (mmol)

0 0
10 20 30 40 50 60 70 80

PCO2 (mmHg)

Fig. 1.12 The CO2 dissociation curve for whole blood at 37° and haematocrit of 45%. Upper lines
represent dissociation of total CO2 in mmol.L−1 in deoxygenated and oxygenated whole blood; the
line AV Difference demonstrates increased carriage by carbamino groups as the oxygen saturation
falls. To convert mmol.L−1 to ml.dL−1 multiply by 2.226. Shaded lower areas demonstrate the
amount of CO2 (mmol) in erythrocyte water (325 mL per L blood at pCO2 40 mmHg) and plasma
water (509 mL per L blood at pCO2 40 mmHg). The total water content of 1 L of whole blood is
836 mL. Carbamino CO2 represents the maximum amount of CO2 that fully deoxygenated haemo-
globin can carry. Summation of the amount of CO2 in mmol in shaded areas at any pCO2 divided
by whole blood water (836 mL) yields the concentration of CO2. (After [29–33])

 Na +  +  K +  + 2 ×  Ca 2 +  + 2 ×  Mg 2 +  +  z + 
=  Cl−  +  HCO3−  + 1.8 ×  PO 4 2 −  + 0.28 ×  Alb −  +  lactate −  +  x − 
 (1.18)
CO 2
1 Physiology of Extracorporeal Life Support (ECLS) 35

This occurs because of the large supply of dissolved CO2 throughout the body
and the ready reversibility of its reaction with water. In contrast, concentrations of
the other charged species in Eq. 1.182 are all tightly regulated by homeostatic pro-
cesses and cannot be allowed to rapidly change (Fig. 1.11). Thus if lactic acid pro-
duction were to rise, the immediate effect is a shift of [HCO3−] to CO2 to maintain
electrical neutrality. A secondary increase in minute ventilation facilitates clearance
of the elevated CO2, and over time renal chloride loss increases to restore balance.
These mechanisms facilitate understanding the alkalotic effects of hypoalbumin-
emia and dilutional acidosis from relative hyperchloraemia commonly encountered
in critically ill patients. The reader is referred to other sources for a more detailed
discussion [13, 34].

Carbonic Anhydrase

Carbonic anhydrase is present in many tissues, including erythrocytes and pulmo-

nary capillaries, but not in blood plasma [15]. As a consequence most of the above
reactions occur at a faster rate within erythrocytes, and a membrane-bound Cl−/
HCO3− exchange protein allows the bicarbonate to diffuse through plasma water.
CA catalyses the reaction of carbon dioxide with water which otherwise has a
half-time for achieving equilibrium of around 15 s [13, 35]. Inhibition of carbonic
anhydrase does not affect the final equilibrium concentrations of the Henderson-­
Hasselbalch equation; however it affects dynamic equilibrium that occurs in a body
fluid compartment before diffusion of dissolved carbon dioxide into surrounding
compartments occurs (Fig. 1.9). This is reflected in higher tissue partial pressures of
carbon dioxide if a total blockade of carbonic anhydrase is imposed [15].

The Arteriovenous-CO2 Difference and Carbamino Carriage

Figure 1.12 demonstrates the total carbon dioxide content of whole blood at increas-
ing partial pressures. Total CO2 (often confusingly termed total bicarbonate) is
obtained by measuring the volume of CO2 gas produced after the addition of a
strong acid to a blood sample which shifts Eq. 1.17a fully to the left [36]. In contrast
the bicarbonate reported on an arterial blood gas sample is calculated from the mea-
sured pH and pCO2 using Eq. 1.17c and is typically 2–4 mmol.L−1 lower than total
CO2. This difference is due to carbon dioxide carriage on amino acids and particu-
larly carbamino formation on haemoglobin (Eq. 1.16e).
Inspection of the AV difference in Fig. 1.12 at a haematocrit of 45% demon-
strates the pCO2 increases from 40 mmHg to 47 mmHg as tissue metabolism
consumes oxygen. In this example the total CO2 rises correspondingly from

2
Being electrical neutrality, the units are available charge (milli-equivalents per litre mEq.L−1)
hence the concentration of double valent ions is multiplied by 2 to account for their charge density.
The multiplier of 1.8 for inorganic phosphate and 0.28 for albumin are approximations as the
charge density for these weak acids varies slightly with pH. [z+] and [x−] refer to other unmeasured
exogenous or endogenous cations and anions.
36 M. J. Brain et al.

19.6 mmol.L−1 (43.5 mL.dL−1) to 22.4 mmol.L−1 (49.8 mL.dL−1) the difference
being due to tissue CO2 production and is proportional to the respiratory quotient
[29]. The fraction of CO2 carried as carbamino groups increases in this process as
oxygen unloading increases the affinity of haemoglobin for carbon dioxide (haemo-
globin is more basic when deoxygenated, shifting Eq. 1.17b toward bicarbonate
[15]). It should be emphasised that whole blood bicarbonate is carried by both intra-
cellular and extracellular water [30].

Determinates of Carbon Dioxide Partial Pressure

Carbon dioxide presents a classic clearance problem. At a steady VO2 and fixed RQ,
a constant amount of CO2 will be produced. If the plasma partial pressure is also
stable, then the amount eliminated must equal production and also be constant.
Consider two oxygenators with identical membrane characteristics but differing
surface areas with the second membrane half the area of the first. Fresh gas and
ECMO flows, CO2 production and elimination are all constant, with a membrane
flux of 250 mL.min−1 in each system. The resistance to diffusive transport (RTotal) is
identical for the two membranes so that from Eq. 1.11:

∆l
(
J × RTotal = Area × PCO2 ( Plasma ) − PCO2 ( Gas ) ) where RTotal =
−kO2
= Area1 × ∆P1 Oxygenator 1 (1.19)
 Area 2 
=  × ( x × ∆P2 ) Oxygenator 2
 x 

The numerals denote the two oxygenators, and x is a scaling factor for membrane
area. Halving the area for available transport (making x = 2) will double the required
transmembrane pressure gradient to achieve the same CO2 flux.
Due to the lower solubility of oxygen, a much larger membrane area is required
for oxygenation than is necessary for carbon dioxide removal. The corollary is stan-
dard oxygenators that are capable of excessive CO2 removal resulting in hypocap-
nia. Adding CO2 to the sweep gas or reducing the sweep gas flow so that the
convection of carbon dioxide away from the membrane is slowed may alleviate this.

Extracorporeal Carbon Dioxide Removal (ECCO2R)

ECCO2R refers to the support of hypercapnic respiratory failure, usually in the set-
ting of acutely decompensated pulmonary disease. Modern systems utilise a scaled
down VV-ECMO circuit allowing significantly smaller vascular access catheters.
1 Physiology of Extracorporeal Life Support (ECLS) 37

While achieving clearance of carbon dioxide, the limited membrane area and lower
blood flow significantly reduce any oxygenation effect from these systems.

Blood Flow Requirements of ECCO2R

Lower blood flows in ECCO2R are made possible by the higher solubility of carbon
dioxide and its more linear dissociation curve across the physiologic range com-
pared to the sigmoid curve for oxygen saturation of haemoglobin (Fig. 1.13).
Complete saturation of haemoglobin at a concentration of 10 g.dL−1 with oxygen
results in a maximum oxygen carriage of approximately 13 mL.dL−1. If returning
venous blood has a saturation of 70%, its oxygen content is 9.3 ml.dL−1 (Eq. 1.14
and Fig. 1.5) allowing only 37 mL of oxygen to be added per litre. Thus an adult
consuming oxygen at 300 mL.min−1 with a mixed-venous saturation of 70% would
require an ECMO flow and cardiac output over 7 L.min−1.
In contrast to the saturable uptake of oxygen, nearly all the carbon dioxide in
venous blood could be removed with sufficient exposure to the gas exchange mem-
brane and a constant supply of fresh gas to maintain the trans-membrane gradient.
From the dissociation curve for carbon dioxide (Fig. 1.12), a mixed-venous pCO2 of
60 mmHg equates to 52 ml.dL−1. Hence from 500 mL of blood at this pCO2, 250 mL
of CO2 could be removed by an efficient gas exchange system allowing an extracor-
poreal circuit running at 500 ml.min−1 to clear the minutely CO2 load of adult
patients.

Recirculation with Smaller Access Catheters

The example above assumes no recirculation; however recirculation is an increased

risk with the smaller integrated vascular access devices used for ECCO2-R. Significant
recirculation of blood that had already been completely cleared of CO2 will limit
overall removal, and increasing the blood flow rate may not influence the mem-
branes exposure to carbon dioxide-containing blood. Optimal placement of vascular
access catheters in vessels with sufficient blood flow past the catheter tip is required
to minimise this phenomenon.

Oxygenation in ECCO2-R: Integration with Pulmonary Oxygenation

Despite limitations imposed by haemoglobin saturation with oxygen, some improve-

ment in systemic oxygenation may occur with ECCO2-R which can be explained by
three mechanisms [37]:
1. Some flux of oxygen to the venous blood still occurs across the membrane as
outlined in the section on oxygen transfer.
38 M. J. Brain et al.

Oxygen and CO2 Content vs Partial Pressure

Partial Pressure (kPa)

0 2 4 6 8 10 12

60

25

50

20

40
Content in Blood ml/dL

Amount in blood mmol/L

15

30
Oxygen Content for Hb 10 g/dL pH 7.44 Temp
37.5° & BE 0

Total CO2 ml/dL

10
20

5
10

0 0
0 10 20 30 40 50 60 70 80 90 100
Partial Pressure (mmHg)

Fig. 1.13 Comparison of oxygen and carbon dioxide content in blood vs. partial pressure

2. Facilitation of mechanical ventilation strategies that maximise oxygen exchange

without a need to focus on carbon dioxide clearance.
3. Improving the partial pressure gradient for pulmonary oxygen exchange.
1 Physiology of Extracorporeal Life Support (ECLS) 39

Ventilation Strategies with ECCO2-R

A significant mismatch between pulmonary ventilation and perfusion characterises

many forms of respiratory failure, with a true pathologic shunt developing in more
severe conditions such as acute respiratory distress syndrome or severe pneumonia.
In general, oxygenation in these pathologies is optimised by strategies that increase
the pulmonary surface area for gas exchange and maximise the proportion of pul-
monary blood flow through the ventilated areas.
Applying relatively high positive end-expiratory pressure (PEEP) recruits pul-
monary surface area, particularly in dependent lung areas that receive more blood
flow [15]. Avoiding barotrauma to the pathologically non-compliant lung necessi-
tates reduced tidal volumes (6 ml.kg−1) targeted to achieve safe inspiratory plateau
pressures (<30 cmH2O). This strategy improves oxygenation and decreases mortal-
ity compared to high tidal volume ventilation [37, 38].
The high PEEP and pulmonary hypertension resulting from hypoxic pulmonary
vasoconstriction place a significant load on the right ventricle which often has
impaired contractile function from hypoxia and the acidosis imposed by hypercar-
bia. Right ventricular dilation occurs and impairs left ventricular filling by shifting
the interventricular septum, and right heart output falls (Fig. 1.14). This overall
reduction in cardiac output reduces systemic oxygen delivery (Eq. 1.15). Inhaled
pulmonary vasodilators can be added to improve matching of pulmonary blood flow
to ventilated areas and may reduce the load imposed on the right heart.
Increasingly, prone positioning in severe lung injury has been shown to improve
mortality [39, 40] by improving ventilation-perfusion matching and recruitment of
dorsal lung regions (relative to supine positioning) and facilitates secretion clear-
ance [41]. The benefits of prone positioning a fully ventilated patient come at a cost
of more labour-intensive care to minimise pressure related injury and avoid device

Fig. 1.14 Acute right ventricular (RV) enlargement with secondary impairment of left ventricular
(LV) filling in hypoxia. Note the intrusion of the basal septum into the LV in the long axis view and
the flattened septum and enlarged RV in the short axis view. (Image M. Brain by permission of The
Alfred Intensive Care Unit, The Alfred Hospital, Melbourne)
40 M. J. Brain et al.

dislodgement [42]. Studies delineating whether prone positioning with ECCO2R is

of added benefit are ongoing [43].

Passive Oxygenation and Low Tidal Volume Strategies

If alveoli are open to fresh gas flow, then ongoing oxygen consumption by the body
maintains a pressure gradient between oxygenated alveoli and venous blood in pul-
monary capillaries allowing diffusion to occur. This passive removal of oxygen
from the alveolus permits convection of further fresh gas into the alveolus without
tidal ventilation occurring – so-called passive oxygenation; however the lack of
tidal ventilation results in poor pulmonary CO2 clearance. Maintaining sufficient
mixing of oxygen with expired CO2 in the ventilated alveoli maintains this process
and can be achieved with low tidal volumes if sufficient pulmonary surface area
exists. Though supporting oxygenation, this high PEEP, low tidal volume ventila-
tion strategy leads to significant hypercapnia as convection of carbon dioxide out of
the lung is reduced.
Obstructive small airways disease imposes a separate (but often co-existent)
issue. Here, prolonged expiratory times are required to avoid excessive pulmonary
hyperinflation and barotrauma, and this strategy again decreases alveolar minute
ventilation and carbon dioxide clearance.
In both settings, allowing ‘permissive’ hypercapnia is a well-documented strat-
egy; however, in some patients the resulting arterial pCO2 is so high that significant
acidosis ensues. Institution of ECCO2-R in these settings improves carbon dioxide
clearance and may alleviate the hypercapnic acidosis. In other settings such as head
trauma, evidence suggests that maintaining normal pCO2 is optimal, and this may
not be achievable with coexistent lung injury without strategies such as ECCO2-R.

ECCO2-R and Alveolar Oxygen Concentrations

Oxygenation utilising high PEEP low tidal volume strategies can be augmented by
maximising the oxygen partial pressure gradient between perfused alveoli and pul-
monary capillary blood to drive diffusive flux. As discussed, each gas in a container
contributes to the total partial pressure, which in an alveolus is usually atmospheric
pressure saturated with water vapour (PIO2, Eq. 1.6). The partial pressure of carbon
dioxide in perfused alveoli will be close to the mixed-venous partial pressure, and
the alveolar gas equation describes the resulting alveolar oxygen tension (PAO2):
The alveolar gas equation and its variations. PA denotes alveolar partial pressure,
and Pa denotes arterial partial pressure. Expressions a. and b. require knowledge of
the respiratory quotient and assume that other soluble gases in the alveolus (mostly
nitrogen) have reached equilibrium between the alveolus and plasma. In critical
care environments where end-tidal CO2 monitoring is available, Eq. 1.20c is the
most precise expression of alveolar gas and makes no assumptions about the respi-
ratory quotient [15]:
1 Physiology of Extracorporeal Life Support (ECLS) 41

Pa CO  1 − R.Q. 
PAO = PIO − 2
+  Pa × FIO ×  a.
2 2
R.Q.  CO2 2
R.Q. 
Pa CO (1.20)
PAO ≈ PIO − 2
b.
2 2
R.Q.
 PIO − PEO 
PAO = PIO −  Pa CO × 2 2
 c.
2 2
 2
PEO 
 2 

It is clear from inspecting Eq. 1.20b that the alveolar PAO2 and thus the pressure
gradient for gas exchange fall with increasing PaCO2. Table 1.3 demonstrates this
effect at different FiO2 levels. At a low PiO2, removal of CO2 with ECCO2-R signifi-
cantly increases the partial pressure gradient for oxygen diffusion from the alveolus
into plasma. However, the effect rapidly decreases with increasing PIO2 to the point
of being negligible.
In summary then any improvement in systemic oxygenation with ECCO2R is
most likely to be the result of facilitating ventilation strategies that optimise recruit-
ment and surface area for oxygenation by allowing alternative strategies for CO2
removal.

Biophysics of the Extracorporeal Blood Path

In the simplest configuration, extracorporeal circuits for ECMO consist of a pump

and the membrane oxygenator connected in series by tubing. Access and return can-
nula remove and return blood to the patient, and it is the vessels which are accessed

Table 1.3 The proportional effect on alveolar oxygen tension of CO2 removal at increasing FiO2
Atmospheric pressure (mmHg) 760 760 760
Water vapour (mmHg) 47 47 47
Calculated inspired oxygen partial pressure
PiO2 (mmHg) 150 428 570
RQ 0.85 0.85 0.85
Alveolar oxygen tension
PaCO2 (mmHg) 90 90 90
PAO2 (mmHg) 47 331 477
Alveolar oxygen tension with ECCO2-R
PaCO2 (mmHg) 40 40 40
PAO2 (mmHg) 106 390 536
Proportional effect of CO2 removal on alveolar PAO2
%change 125% 18% 12%
42 M. J. Brain et al.

(venous to venous or venous to arterial) that determine whether the mode of support
is primarily oxygenation and/or circulatory as well (Fig. 1.1).
This section will consider the path of blood in plastic conduits initially with only
a pump and then with an oxygenator interposed. Several of the parameters that
describe the extracorporeal blood path also apply to flow within the systemic circu-
lation, and interactions with the native circulation will be considered where appro-
priate. Various coatings may be applied to the circuitry to enhance biocompatibility,
reducing the inflammatory response to the circuit and the activation of the clotting
cascade in blood traversing the circuit.
Two principles, conservation of energy and conservation of mass, describe the
flow of blood through the extracorporeal circuit. The circuit flow will be considered
initially as conduits with a pump such as is found in a ventricular assist configura-
tions (LVAD in Fig. 1.1) before also discussing the oxygenator.

Types of Blood Pump

Two types of pump are currently used for ECLS, positive displacement pumps
(roller pumps) and velocity pumps, with the latter becoming more common. LVAD
devices generating pulsatile flow are no longer in use in adults and will not be
discussed.

Positive Displacement Pumps

Occlusive roller pumps provide a positive displacement of a fixed volume of blood

as a roller sequentially compresses the tubing. Two to three rollers are mounted on
a rotating cylinder while the tubing is encased around this cylinder with the geom-
etry such that when one roller ceases contact with the tubing, the next roller begins
compression, providing continuous flow. The negative pressure created by re-
expanding tubing behind the roller draws more blood from the patient. The tubing
diameter and length of the compressed path dictate the volume displaced by each
roller pass, while revolution rate (rpm) of the pump determines the output.
Roller pumps are relatively resistance-independent across the operating range of
the driving motor and are thus able to generate considerable positive pressure ahead
and negative pressure behind the pump. The higher pressures may cause haemolysis
of red cells under conditions of flow restriction; furthermore these pumps can also
propel air. Compared to centrifugal rotor designs, roller pump output is less affected
by pulsatile flow generated pressure gradients across the circuit.

Velocity Pumps

Velocity pumps increase the kinetic energy (velocity) of the flowing fluid by high
speed rotation of an impeller rotor. This kinetic energy is converted to potential
energy (pressure) by the containment of the flow stream.
1 Physiology of Extracorporeal Life Support (ECLS) 43

Centrifugal Constrained Vortex Pumps

These devices consist of an encased rotor magnetically coupled to a motor drive

external to the blood path. Compared to roller pumps which run in the 80–150 rpm
range, these devices routinely operate at 1500–4000 rpm, and considerable engi-
neering goes into rotor designs that minimise trauma to blood components and heat
generation at the rotor axis. The centrifugal force of the rotating mass of blood
generates negative pressure at the central pump inlet and positive pressure perpen-
dicularly against the periphery of the pump along which the outlet is located. These
devices are non-occlusive and will rotate freely whether the pump is occluded or
de-primed by entrainment of air. Loading conditions and the physical properties of
blood thus significantly influence the output.

Axial Flow Pumps

These pumps consist of a turbine design that accelerates blood in the line of flow,
typically operating at even higher speeds (up to 9000 rpm). Currently they are only
implemented in ventricular assist devices and, compared to centrifugal devices, are
sensitive to loading conditions.

Monitoring Pump Output

The physics behind the sensitivity of centrifugal and axial flow devices to loading
conditions will be discussed after consideration of flow. In ECMO axial flow pumps,
the relationship between pump speed and blood flow is load dependent and has thus
resulted in Doppler monitoring of the velocity of blood in the tubing. In implanted
devices (ventricular assist devices), it is not currently possible to accurately provide
a durable long-term flow measurement system, so flow is approximated from inter-
preting variations in rotation speed. Accurate haematocrit measurements are impor-
tant in this estimation as will be discussed. In practice, exact determination of flow
may be less important than clinical evidence of adequate forward output (perfusion,
exercise tolerance) and ventricular unloading (pulmonary congestion), while echo-
cardiography can measure flow and demonstrate ventricular collapse.

Physical Properties of Blood Relevant to Flow

The basic hydraulic principles outlined below describe the flow of a uniform incom-
pressible fluid in solid tubing. The conduits utilised in ECMO can be considered
solid tubes; however, blood only approximates a uniform fluid due to the presence
of suspended cells and chemical interactions of suspended proteins with the con-
duit walls.
44 M. J. Brain et al.

Two fluid characteristics are relevant to flow, density and viscosity. As a descrip-
tion of mass per unit volume, density (ρ) particularly affects the momentum of mov-
ing fluid and resistance to acceleration. For blood, density is slightly greater than
water at 1.06 g.mL−1 at 37 °C.

Viscoelasticity

Viscosity (η) is a measure of a fluids resistance to flow. The viscosity of plasma is

around 1.8 times that of water; however, the viscosity of whole blood is much
higher, at four to five times that of water for a haematocrit of 40%. Viscosity varies
with temperature, increasing by about 2% for each degree Celsius. More relevant to
ECLS is the effect of suspended cells and proteins, particularly erythrocytes – vis-
cosity increases in a non-linear fashion relative to the haematocrit as shown graphi-
cally in Fig. 1.15.
The effect of hematocrit on viscosity is particularly relevant to the acceleration
of blood by vortex centrifugal pumps (but not volume displacement roller pumps)
and needs to be considered in two clinical situations:

Relative Viscosity of Blood

14

12

10
Relative Viscosity

0 60%
25 50%
a

75
Cav

150 40%
Aor t ing

300
Aor t ing
a

t
d

500 30% cri

Ven

a
cen

ato
end
a

700
20% m
Des

900 ae
Asc

1100 H
1300 10%
s
illarie

1500
Shea
s

r Rate
riole

(s -1)
Cap

Ar te

Fig. 1.15 Relative Viscosity of Blood compared to water vs. hematocrit and shear rate. Viscosity
rises exponentially with increasing hematocrit but falls with increasing shear rates, as aggregating
forces between red cells become less prominent (back panel). Viscosity data plotted from equa-
tions in [44] and shear rates of vessels from [45, 46]
1 Physiology of Extracorporeal Life Support (ECLS) 45

1. Implanted ventricular assist devices where blood flow is estimated from the rotor
rpm – the software of these devices often requires a current haematocrit.
2. In ECMO when interpreting a falling haematocrit. This may mask other increases
in circuit resistance such as clotting within the oxygenator and similarly must be
considered when pressure gradients across the oxygenator increase after transfu-
sion (discussed under ‘the oxygenator as a resistor’).
Figure 1.15 also displays decreasing viscosity with increasing shear rate. Shear
rate is a measure of the rate at which adjacent fluid streamlines move with respect
to each other (see laminar flow below). At low blood velocities, aggregates compris-
ing erythrocytes and plasma proteins contribute to adhesive forces. As blood veloc-
ity increases, aggregates become smaller as erythrocytes move with respect to each
other. At high velocities, laminar layers of erythrocytes and plasma move with sig-
nificantly less intracellular adhesive forces [45]. Vessel size contributes to the shear
rate in combination with velocity. Arterioles and capillaries having the highest shear
rates and the great veins the lowest (Fig. 1.15) [46, 47]. Centralisation of erythro-
cytes also occurs in small capillaries, decreasing the effective haematocrit contribut-
ing to lower relative viscosities [46, 48].
To accelerate blood, force is required to overcome the viscous adhesive forces
between erythrocyte aggregates and layers of plasma, along with drag forces cre-
ated by red cells moving at different rates to the surrounding plasma [44]. These
components of force will be lost as heat (viscous friction). Some of the accelerating
force will also deform erythrocytes and in that respect imparts potential energy that
is released as the erythrocytes relax back to their normal shape (elasticity).

Flow of Ideal Fluids in Solid Conduits

Analogous to the concentration gradient driving molecular flux across membranes,

flow through a level tube will occur from a high pressure region at one end to a low
pressure area at the other, while certain resistances will oppose this flow:

⋅ PInlet − POutlet
Q= (1.21)
R

This is a statement of energy transfer: The region of high pressure has potential
energy, and the conduit allows this to be converted to kinetic energy by accelerating
a mass of fluid toward the low pressure region.
The resistances describe several properties of the fluid and conduit that limits the
rate of energy transfer. However, before discussing these resistances, it is important
to note this equation does not describe conditions within the conduit but only the
mass transfer between the two regions of pressures. Most importantly, most of the
flow acceleration, that is conversion of potential to kinetic energy, occurs before the
conduit orifice, and there is minimal pressure change within the conduit itself
(Fig. 1.16). Such flow convergence and divergence can be seen at the inlet and out-
lets of ECMO cannulas in vessels using colour flow Doppler techniques.
46 M. J. Brain et al.

Total Energy in System

Velocity in Flowing Fluid

Pressure in Flowing Fluid

High Pressure Low Pressure
Compartment Compartment
Laminar Flow in Conduit
(Velocity is Constant)

Static Static
Fluid Fluid

Entrance
Length
Flow Flow
Acceleration & Deceleration &
Convergence Divergence

Fig. 1.16 Graphic description of the Bernoulli equation for laminar flow in a rigid level conduit
from a high pressure compartment to a low pressure compartment for an ideal fluid. Although the
driving pressure gradient for flow is between the two compartments, pressure (potential energy) is
actually lower in the region of flow than in either compartment due to conversion to kinetic energy.
This lower pressure also exists in the regions of flow acceleration and deceleration where potential
and kinetic energy interchange; however, the total energy (potential + kinetic) at each point along
the system is constant

Figure 1.16 also displays the total internal energy in the system, that is sum of
potential and kinetic energy. This is depicted as constant though in reality falls
slightly across the system due to heat loss from viscous friction. Within the conduit
the velocity (for laminar flow) is also constant and does not vary from one section
to another without cross section variation. The fluid in the tube acquires momentum
in the zone of flow acceleration. This momentum (kinetic energy) raises the pres-
sure (potential energy) in the receiving container until no further pressure gradient
exists and flow ceases.
This system is described by the Bernoulli equation stating that the internal energy
of a fluid is the sum of potential energy due to its column height, gravity, con-
strained pressure and any kinetic energy [49]:

ρ × velocity 2 where ρ is density, g is acceleration of

ρ gh +
gravity ( 9.807 m.s 2 ) , h is fluid column height
2
+ Pressure = Constant
ρ × velocity2 (1.22)
+ Pressure = Constant For neglible gravity
2
Kinetic Potential
+ = Constant
Energy Energy
1 Physiology of Extracorporeal Life Support (ECLS) 47

The constant is the total energy of the system, and this equation simply says that
at each point along the tube conservation of energy will dictate the velocity and
pressure. For tubes of varying cross section, flow must be constant; thus the velocity
will increase as diameter decreases, leading to drops in pressure; however, the total
internal energy will still remain constant such that for any two points along the tube
the sum of momentum and pressure must balance [50]:

ρ × V12 ρ × V22
+ P1 = + P2
2 2 (1.23)
∆P = 1 ρ (V22 − V12 )
2

Inspection of Fig. 1.16 and consideration of viscoelasticity reveals why the

Bernoulli equation only strictly applies to an incompressible Newtonian fluid in a
solid conduit [49] – energy lost to viscous forces or stored as elastic potential energy
is not included in Eq. 1.23. Nevertheless the concepts of conservation of energy
from potential (pressure) to kinetic (velocity) provide the foundation for under-
standing conduit flow, and Eq. 1.23 is used clinically in Doppler calculations.

Resistance to Flow

Equation 1.21 implies that resistance simply alters the ratio between the pressure
gradient and blood flow. However, due to viscous friction, any constrictions to flow
create small turbulent vortices in their wake that create vibration and increased
shear velocities at the vessel wall, this energy being lost as heat and sound. For a
uniform diameter flow path with a fixed resistor, these factors tend to result in a loss
of potential energy (i.e. a pressure drop) across the resistor in addition to any pres-
sure drop from altering velocity.

Resistance to Laminar Flow

Laminar flow describes a parabolic speed profile (Fig. 1.16) with a maximum blood
velocity in the centre of the cylinder and a near stationery blood film in contact with
the conduit walls at which biofilms form. Laminar flow is the least traumatic on
formed blood elements and the most energy efficient, being directly proportional to
the pressure gradient and inversely proportional to resistance defined as:

8 × length ×η
Resistance to laminar flow =
π × radius 4
( mmHg.L−1.min −1 ) (1.24)

Combining Eqs. 1.21 and 1.24 and converting radius to diameter, d, yields
Poiseuille’s equation:
48 M. J. Brain et al.

⋅ ( PInlet − POutlet ) × π × d 4
Q= (1.25)
128 × length ×η

The effect of viscosity (η) has been discussed, and while length of the tubing is
important, it is far outweighed by the radius to the fourth power, whereby the flow
rate can be doubled by 20% increase in tube diameter. This becomes important in
selecting vascular access catheter size for a desired flow rate, which can be limited
by vessel size and technical factors relating to insertion.

Turbulent Flow

In contrast to laminar flow in straight tubules, at circuit constrictions, pump rotors

and around the complex geometry of the oxygenator, turbulent flow will occur. In
turbulent flow, more of the impulse energy goes to random kinetic motion, eddy
currents and friction between the fluid streamlines, suspended cells and the walls of
the tube. In comparison to Eq. 1.25, the pressure gradient across a conduit demon-
strating turbulence increases exponentially with increasing blood flow [15].
Unlike laminar flow there is less likely to be a stationary layer of blood in the
proximity of tubing walls or membranes. This is an advantage in an oxygenator as
it more effectively purges tubules, maintaining concentration gradients. These ben-
efits are offset by increased trauma on formed blood elements and the activation of
inflammatory mediators by the increased contact with the foreign surface.

Reynolds Number

The most important parameter in determining the character and average rate of flow
in a conduit is the ratio of fluid momentum to viscous forces known as Reynolds
number [50] where for a conduit of diameter d and a fluid with a mean velocity v :

ρ ×d ×v
Re = (1.26)
η

At Re numbers less than 100, pure laminar flow is apparent; from 100 to 1000
laminar flow occurs, but an increasingly wide boundary layer appears against the
stationary conduit; above 1000 there is a transition to turbulent flow streams and
eddy currents; Re numbers >10,000 are purely turbulent flow [50]. The linking of Re
to velocity is critical in determining the maximal flow through extracorporeal cir-
cuits and will be discussed below. The distance blood that must flow before laminar
flow can be established is also related to Re and is known as the entrance length: Re
× 0.03 × tube diameter (Fig. 1.16).
1 Physiology of Extracorporeal Life Support (ECLS) 49

A Reynolds number can also be calculated for the resistance (drag) that a particu-
late experiences locally in a viscous fluid. For erythrocytes suspended in plasma, the
cell diameter (dRBC) and a coefficient, ϕ, for cell deformity Replasma is defined as [44]:

ρ plasma × d RBC × ( vplasma − vRBC ) × φ

Replasma = (1.27)
ηplasma

Replasma contributes to the overall viscous friction of blood by describing local

turbulence around the moving erythrocytes. Local turbulence also contributes to
shear stress and deformation on erythrocyte membranes that contributes to hae-
molysis in areas such as rotor blades and flow constrictions.

Optimising Blood Flow for Gas Transfer

The membrane oxygenator, with its complex network of hollow fibres designed to
maximise the surface area available for oxygenation, also significantly increases the
cross-sectional area of the blood path (Fig. 1.17). By conservation of energy and
mass, flow (Q, ml.s−1) through any two points of the extracorporeal circuit must be
constant, and flow can be described as the product of blood velocity, V (cm.s−1), and
cross-sectional area, CSA (cm2):

Q1 = Q2
(1.28)
CSA1 × V1 = CSA 2 × V2

The cross section of the conduit is calculated as π.radius2 and an ECMO flow of
5 L.min−1 in 1 cm diameter tubing yields a velocity of 1.06 m.s−1. Oxygenator cross-­
sectional areas vary, but an approximation can be made by dividing the oxygenator
priming volume by the oxygenator width in centimetre, yielding a volume per cm.
This can be divided by the volume in 1 cm of conduit yielding a ratio of the cross-­
sectional areas. A value of 80 would imply the axial velocity of blood within the
oxygenator is around 1.3 cm.s−1 or around 100 times slower than the conduit veloc-
ity, allowing significantly more time for gas exchange.
Oxygenators are characterised by the surface area of their membrane, but there
is no guarantee that the blood flow distributes evenly over this area. In particular, for
hollow fibre oxygenators, the blood must spread across the opening face of the
tubules. Clot deposition over time will gradually decrease the number of paths that
blood can take.
50 M. J. Brain et al.

Patient Access Cannula Pre-Oxygenator Pressure Post-Oxygenator Pressure

Patient Return Cannula

Centrifugal Vortex Pump

Cross sectional area of conduit (cm2)

Cross sectional area of oxygenator (cm2)

Fig. 1.17 Schematic of a simplified ECMO circuit with a centrifugal vortex pump, a membrane
oxygenator and conduits between the patient access cannula. Systems consisting of centrifugal
vortex pumps (as opposed to roller pumps) are typically valveless and can flow in either direction
if the pump is off. The pressure drop across the oxygenator is depicted as the differing heights of
the blood columns. Cross-sectional area of the conduit and oxygenator (red areas) is shown

The Oxygenator as a Resistor

The design of an oxygenator is a compromise between maximising the surface area

for gas exchange and minimising flow resistance and priming volumes. The hollow
fibres/membrane of the oxygenator equate to an extracorporeal capillary system and
considerably impede the flow of blood, which in the case of adult ECMO is in the
order of 3–6 L.min−1. The resistance to flow within the oxygenator will depend on
design, flow rates and the formation of thrombus.
For patients fully dependent on ECLS for oxygenation or circulatory support,
failure of blood flow or gas exchange is life-threatening, and early warning of cir-
cuit clotting is important. It is common practice to continuously monitor the oxy-
genator inlet and outlet pressures by placing pressure transducers at these points
(Fig. 1.17). The flow generated by the blood pump is ultrasound monitored, and if
kept constant, Eq. 1.21 can be simplified to: pressure drop is proportional to resis-
tance. At constant flow then, a rising pressure gradient signifies increased resistance
to blood flow and is usually due to fibrin deposition in the oxygenator or alterations
in haematocrit.
1 Physiology of Extracorporeal Life Support (ECLS) 51

Conservation of Energy Across the ECMO Circuit

Figure 1.18 depicts total energy exchange across an extracorporeal circuit running
at steady state. As in Fig. 1.16, the total energy of blood is made up of potential
energy (pressure) and kinetic energy (velocity); however, energy loss to the system
from viscous friction (most notably across the oxygenator) and addition of energy
by the pump are now depicted. As we are now using blood, a full statement for
stable energy transfer must now replace the Bernoulli equation where h denotes
energy added or lost to the system [49, 50]:

ρ × v12 ρ × v22
ρ gh1 + + P1 + hpump + hRBC = ρ gh2 + + P2 + hfriction + hRBC (1.29)
2 2

Here the internal energy of the fluid at the intake comprises the fluid column
height, intake velocity and intake pressure and elastic energy stored in the erythro-
cyte. Energy is added by the pump (hpump), and for forward flow to occur, the sum of
pump energy and intake internal energy must be greater than the sum of energy lost
as viscous friction and the internal energy (fluid pressure and momentum) of the

Access Cannula Cannula to Pump (Axial Oxygenator Return Cannula

in Vessel wider ECMO Configuration) ‘Resistor’ in Vessel
Conduit

Total Energy in System

Pressure in Flowing Fluid

Velocity in Flowing Fluid

Fig. 1.18 Energy exchange across the extracorporeal circuit; after [51]. Refer to Fig. 1.16 for
discussion of energy transfer from pressure to velocity. The total internal energy of the fluid is
depicted with sequential losses of heat generated from viscous friction at constriction points. Thus
the total energy falls at all points other than the pump, most prominently in the low flow region of
the oxygenator. Narrow spacing between laminar flow lines represents high velocity at low pressure
52 M. J. Brain et al.

receiving vessel. Gravity and the fluid column height are now important, as the
suspended red cells have mass and will sink to the bottom of a stationary container.
It should be noted that ‘lost’ energy is contained as heat [50].
At start-up, the high velocity vanes of the rotating axial flow pump add kinetic
energy to the column of blood. This kinetic energy progressively accelerates the
blood column until a Reynolds number is reached where the amount of turbulent
flow is sufficient for energy losses from viscous friction to balance the left and right
side of Eq. 1.29 and a constant velocity results. This does not mean turbulent flow
will exist throughout the system; rather the energy loss from small areas of turbu-
lence will summate to oppose the applied energy of the rotor.
As the conduits before and after the pump are of equal cross-sectional area and
all blood must come via the pump, the pre- and post-pump velocity is constant (con-
servation of mass). Instead, the applied kinetic energy is converted to pressure
against the pump housing, against the post-pump conduit walls and against any
resistance (the oxygenator) or pressure load downstream (Fig. 1.18).
The power (watts) required of the pump is related to the required force exerted
on the blood (torque) and the rotation speed (revolutions per second, converted from
rpm by dividing by 60).

rpm
P = torque × 2π × (1.30)
60

Torque represents force multiplied by angular distance and has SI units of

Newton metre (N.m). It should be noted in Eq. 1.30 that torque is the force required
to move the rotors to both propel the fluid column and overcome viscous friction at
the rotor blades.

Occlusion of the Flow Path

It is important to appreciate that the total system energy in a velocity pump as

described above is dependent on flow occurring. The energy imparted by a rotating
pump will paradoxically fall if flow is occluded. In a system occluded downstream,
the pump will still maintain some pressure against the occlusion, but in the absence
of forward flow, the velocity term in Eq. 1.29 will be zero at the pump inlet, and no
work will be done maintaining blood velocity against viscous friction at any point
in the circuit other than viscous friction at the rotor blades themselves, that is the
rotors and blood around the rotors will continue moving, but blood in the rest of the
circuit will be stationary.
The only moving blood now is in circular motion around the rotors, and the rotor
velocity will accelerate until sufficient local turbulence exists around them that a
Reynolds number is reached where the resistance from this viscous friction again
opposes the applied energy and a new higher steady-state rotor velocity results. In
other words, by Eq. 1.30, if the same power is applied to the rotors as before the
occlusion event, then rpm will increase despite no forward flow occurring. Conversely
if the rotors were set to maintain a continuous rpm, power consumption must fall.
1 Physiology of Extracorporeal Life Support (ECLS) 53

This effect of occlusion increasing the rpm in velocity pumps is in contrast to the
effect on a volume displacement (roller) pump. If a roller pump continues against an
occlusion, then its continual volume displacement will steadily increase the pres-
sure in the distal conduit until mechanical failure occurs, that is the work the roller
pump performs on the blood increases with occlusion.

Flow Regurgitation

It is apparent from Eq. 1.29 that a high opposing blood velocity or pressure at the
pump outflow opposes forward flow and the energy imparted by the pump must be
sufficient to overcome this. This is relevant in pulsatile circulations where transient
peak pressures in the receiving vessel may exceed the energy imparted by the pump,
causing not only flow deceleration but occasionally flow reversal. In this situation,
the viscous resistance to circuit flow is now being overcome by the driving pressure
beyond the pump, and the energy imparted by the rotors will again result in viscous
friction at the blades from extremely turbulent flow (Eqs. 1.26 and 1.27).

Cavitation

Turbulent flow describes rapidly varying random inverse fluctuations of pressure

and velocity that sum to the total internal energy of the fluid [50]. The amplitude of
these fluctuations increases with the internal energy of the fluid. Not only may high
local velocities induce erythrocyte haemolysis from shear stress, a second mecha-
nism causing haemolysis known as cavitation may also occur.
If the peak negative pressure fluctuation generated in regions of turbulent flow
falls below the vapour pressure of dissolved gases in blood (usually at the tips of the
rotor blades), then bubbles will form and then rapidly collapse as the fluctuating
pressure increases again. Cavitation describes the implosion of these small bubbles
in blood; as each bubble collapses, a shockwave occurs as the walls of the cavity
collapse and enough force may be generated to rupture cell membranes.
The positive and negative pressure fluctuations in turbulent flow occur around the
average pressure in the region of flow, whereas the vapour pressure is an absolute
property of the fluid at a particular temperature and can be influenced by hydrostatic
pressure. Therefore, cavitation is more prominent if low pressures exist at the inlet of
the circuit and can be avoided by maintaining central venous pressure, keeping the
pump head below the level of the access cannula and avoiding excessive pump speed.

Interaction of Velocity Pumps with the Systemic Circulation

Understanding and recognising patterns of interactions between the native circula-

tion and the continuous flow velocity pump of an ECMO circuit are important for
ongoing smooth operation. For VA-ECMO where the native heart remains pulsatile,
the interaction is complex; however, for VV-ECMO the considerations come down
54 M. J. Brain et al.

to three things: adequacy of overall adequacy cardiac output, right ventricular func-
tion and interaction with pulmonary haemodynamics, and adequacy of venous
access (above).

Adequacy of Cardiac Output

As extensively discussed the flow of oxygenated blood to peripheral organs in a

VV-ECMO configuration is dependent on cardiac output. If cardiac output is insuf-
ficient and available interventions are not anticipated to augment it, then an alterna-
tive therapy such as VA-ECMO may be appropriate. The effect of ventilation
strategies common in ARDS on the circulation need considering as institution of
VV-ECMO will often improve haemodynamics as considered below for the right
ventricle.
As described previously, very high cardiac outputs relative to the available
ECMO circuit flow will result in lower achievable arterial oxygen saturation. This
may necessitate additional large access canulae.

 ight Ventricular (RV) Function

R
and Pulmonary Haemodynamics

The right ventricle dysfunction is common in the population at risk of requiring

VV-ECMO support. Pulmonary disease may increase pulmonary vascular resistance
and induce pulmonary hypertension. In addition primary contractile impairment
may occur particularly in association with ARDS and lower respiratory tract infec-
tion, PaO2 to FIO2 ratio <150 mmHg, hypercapnia and excessive positive pressure
ventilation (plateau minus total positive end-expiratory pressure >18 cm H2O [52].
Similar to the systemic circulation, pulmonary vascular resistance is defined as
the difference between pulmonary artery pressure and left atrial pressure divided by
the right ventricular cardiac output. Isolated modulation of pulmonary vascular
resistance can often be achieved by reducing ventilator pressures and the use of
pulmonary vasodilators such as inhaled nitric oxide or epoprostenol. Right ventricu-
lar performance may also be improved by correcting hypoxia and acidosis, but it is
sometimes necessary after LVAD insertion to temporarily support the right heart
with an RVAD or VPA-ECMO depending on the level of hypoxia (Fig. 1.1).
In VV-ECMO determining RV function and its improvement is a key decision at
the outset. In the setting of significant oxygenation failure as in ARDS, right ven-
tricular function is often significantly impaired by the direct effects on the myocar-
dium of hypoxia, acidosis and the transmitted effects of hypoxic pulmonary
vasoconstriction and elevated ventilatory pressures. Correcting oxygenation by
instituting VV-ECMO may ameliorate many of these processes and improve RV
function, where as if RV function is deemed unlikely to improve than full circula-
tory support with VA-ECMO may be more appropriate.
1 Physiology of Extracorporeal Life Support (ECLS) 55

Thrombus Formation and Anticoagulation

Thrombus formation within the components of the extracorporeal circuit is a com-

mon occurrence, and anti-coagulation is routine. Excessive extracorporeal circuit
thrombus poses several risks:
• Decreased efficiency of circuit components, particularly oxygenator surface area.
• Increased turbulence at circuit components with associated haemolysis and/or a
hyperfibrinolytic coagulopathy.
• Increased risk of sudden catastrophic circuit failure (which may be life-­
threatening at high levels of support).
• Increased embolic burden on the patient from emboli generated within the
circuit.
Minor thrombus formation in low flow areas of the oxygenator and circuit com-
ponents (Fig. 1.19) is common, and inspecting the volume of clot is a standard
operational observation along with assessment of its impact on pressure drop across

Fig. 1.19 Images of thrombus within a pump head and oxygenator before and after removal of
circuit from patient and washing out blood. (Image M. Brain. Launceston General Hospital)
56 M. J. Brain et al.

the oxygenator. Interval measurements of circuit and patient plasma-free haemoglo-

bin are also performed to detect excessive haemolysis that could result from oxy-
genator or pumphead thrombosis that is causing excessive erythrocyte trauma from
shear forces. As free haem is degraded into biliverdin, free iron and carbon monox-
ide, it has been proposed that carboxyhaemoglobin could be used as an indirect
marker. Measuring carboxyhaemoglobin has advantages of being readily available
on blood gas analysers, whereas plasma-free haemoglobin measurements are prone
to artificial elevation if the sampling is traumatic, shaken in transport or laboratory
analysis delayed [53–55].
Decreasing the risk of thrombus formation by anticoagulation is always balanced
against the risk of excessive and/or critical site bleeding that can occur at sites asso-
ciated with the large bore vascular access used for ECLS, surgical or trauma sites or
spontaneous bleeding at sites of septic emboli or unrecognised vascular pathology.
The enzymatic cascade that promotes the deposition of clot as well as the path-
ways that inhibit and break down formed thrombus is well described as is the modi-
fying effect of common anticoagulant therapies (Fig. 1.20). Of considerable
complexity however is the variety of interactions of the severe pathologies that are
common indications for ECLS with the host coagulation pathways. In studies of
extracorporeal circuits used in renal replacement therapy, sepsis is the most com-
mon factor associated with decreased filter life [56, 57], and this applies to ECLS as
well with VV-ECMO for severe pneumonia often observed to develop earlier sig-
nificant thrombus compared to VA-ECMO circuits for primarily cardiac failure.
Unfractionated heparin remains the mainstay of anticoagulation as it is readily
titratable and cheap to monitor with the activated partial thromboplastin time
(APTT). Heparins main action is to augment the effect of endogenous antithrombin
inhibiting both the conversion of prothrombin to thrombin as well as inhibiting
thrombin from potentiating fibrinogen conversion to fibrin and platelet activation
[58] (Fig. 1.20).
When unfractionated heparin is contraindicated due to suspected or confirmed
heparin-induced thrombocytopenia (HIT) without or with thrombosis (HITT), alter-
native agents are necessary. Short acting direct thrombin inhibitors such as lepirudin
or argatroban are preferred as they can be reach steady state over a few hours and
affect several coagulation tests including the APTT in a dose-dependent fashion
[59]. Agents that require monitoring using anti-factor Xa activity such as low
molecular weight heparin can be more problematic if used for ECLS as some mea-
surement techniques are prone to interference by free haemoglobin from by hae-
molysis [60, 61].

Reparative Properties of the Lung

ELCS is currently used to support a failing cardiorespiratory system until either

recovery occurs or the patient is stable enough to tolerate a destination therapy such
as an LVAD or organ transplantation [2]. An expanding body of literature also exists
1 Physiology of Extracorporeal Life Support (ECLS) 57

Surface Activation
XIIa
High Molecular Weight Kininogen
Prekallikrein
(+)
Tissue Factor
XI XIa VIIa

Ca2+

Hep:AT (-)
IX IXa

Degranulation
(+)
VIII VIIIa
Phospholipids
Ca2+ (+)
V
X Xa Va Platelets Aggregation
Hep:AT (-)
LMWH

Argatroban
Clot
Phospholipids (-)
(+)
Ca 2+
Ca2+
XIII
Prothrombin (II) Thrombin (IIa) XIIIa

Ca2+
Argatroban Hep:AT (-)
(-)

Fibrinogen Fibrin Cross linked Fibrin

Fig. 1.20 The coagulation cascade and inhibition by common anticoagulants (red). Black arrows
depict major steps where zymogens are converted to active clotting factors, and major potentiators
are listed above the steps in black. The positive feedback loops amplifying thrombin production are
depicted in grey. Both fibrinogen and fibrin molecules can bind with the GPIIb/IIIa receptors on
two adjacent platelets allowing aggregation. Hep:AT denotes the major inhibitory actions of
unfractionated heparin when bound with antithrombin. LMWH low molecular weight heparin

on the pathogenesis of fibrosis and reparative properties of the lung in both normal
aging, acute inflammatory and chronic pathologies.
Beginning at 4–5 weeks gestation as two primary lung buds of endodermal epi-
thelium within a mesoderm and vascular plexus [62], later the tips of the elongating
branches differentiate to alveolar type I and II cells. Development of alveoli in
humans is maximal in the first 2–3 years of life but continues into adolescence
[63–65]. ARDS is initially characterised by damage and increased permeability of
the pulmonary endothelium and alveolar epithelium resulting in increased fluid,
proteins, neutrophils and red blood cells first in the lung interstitium and then the
alveolar spaces. This is followed by fibrin deposition and alveolar type II cell hyper-
plasia which later differentiate into the gas exchanging type I cells that line normal
alveoli [66].
Understanding of the interaction between underlying stroma, alveolar type I pro-
genitor cells, collagen deposition from fibroblasts, and inflammatory cells in restor-
ing functional alveoli for gas exchange is still evolving. Distinct regions of the lung
contain populations of epithelial cells that function as adult progenitor cells, with
58 M. J. Brain et al.

some phenotypic plasticity as noted by their ability to undergo long-term self-­

renewal and give rise to different cell types such as alveolar and surfactant produc-
ing cells [62]. The reparative capacity of progenitor cells may also decline with age
or chronicity of the disease [66, 67]. Less well understood is the molecular signal-
ling that regulates the behaviour of these cells and the process of alveologenesis [62].

Conclusions

ECLS is an effective replacement of pulmonary ventilation and/or haemodynamic

function. Although seemingly complex, it can be comprehensively understood when
framed in a physiological and biomechanical context. The practical use of the tech-
nology will be explored in subsequent chapters.

References

1. Kim G-B, Kim S-J, Kim M-H, Hong C-U, Kang H-S. Development of a hollow fiber mem-
brane module for using implantable artificial lung. J Membr Sci. 2009;326(1):130–6.
2. MacLaren G, Brain MJ, Butt WW. ECMO in acute and chronic adult respiratory failure: recent
trends and future directions. Minerva Anestesiol. 2013;79(9):1059–65.
3. Scheepers A, Joost H-G, Schürmann A. The glucose transporter families SGLT and
GLUT: molecular basis of normal and aberrant function. JPEN J Parenter Enteral Nutr.
2004;28(5):364–71.
4. Stryer L. Biochemistry. 4th ed. WH Freeman and Company; 1995.
5. Beals M, Gross L, Harrell S. Efficiency of ATP production [Internet]. Metabolism for energy
and the respiratory quotient. 1999 [cited 2012 Aug 15]. Available from: http://www.tiem.utk.
edu/~gross/bioed/webmodules/respiratoryquotient.html.
6. Elia M, Livesey G. Theory and validity of indirect calorimetry during net lipid synthesis. Am
J Clin Nutr. 1988;47(4):591–607.
7. Wolfe RR, Allsop JR, Burke JF. Glucose metabolism in man: responses to intravenous glucose
infusion. Metabolism. 1979;28(3):210–20.
8. Hellerstein MK, et al. De novo lipogenesis in humans: metabolic and regulatory aspects. Eur J
Clin Nutr. 1999;53:53–65.
9. Mulquiney PJ, Kuchel PW. Control analysis of 2,3-bisphosphoglycerate metabolism [Internet].
Biochem J. 1999 [cited 2012 Oct 11];342:597–604. Available from: http://www.biochemj.org/
bj/342/bj3420597.htm
10. Mulquiney PJ, Bubb WA, Kuchel PW. In situ kinetic characterization of 2,3-­bisphosphoglycerate
synthase/phosphatase [Internet]. Biochem. J. 1999 [cited 2012 Oct 11];342:567–580. Available
from: http://www.biochemj.org/bj/342/bj3420567.htm.
11. Drummond M, Braile DM, Lima-Oliveira AP, Camim AS, Oyama RS, Sandoval
GH. Technological evolution of membrane oxygenators. Braz J Cardiovasc Surg.
2005;20(4):432–7.
12. Lawson DS, Holt D. Insensible water loss from the Jostra Quadrox D oxygenator: an in vitro
study. Perfusion. 2007;22(6):407–10.
13. Peter S. Strong ions plus carbon dioxide. In: Kellum JA, Elbers PW, editors. Stewart’s Textbook
of Acid-Base. 2nd Edition. Lulu Enterprises; 2009. p. 111–32.
1 Physiology of Extracorporeal Life Support (ECLS) 59

14. Anaesthsia Correspondence Web Site [Internet]. [cited 2012 Sep 17]. Available from: http://
www.anaesthesiacorrespondence.net/Correspond3.asp?articleid=3598&archive=1.
15. Lumb AB. Nunn’s applied respiratory physiology. 5th ed. Oxford: Butterworth-Heinemann.
16. Sargent JA. Principles and biophysics of dialysis. In: Jacobs C, editor. Replacement of renal
function by dialysis. Springer; 1996. p. 34–145.
17. Cussler EL. Diffusion: mass transfer in fluid systems. Cambridge University Press; 1997. 606 p.
18. Grathwohl P. Diffusion in natural porous media: contaminant transport, sorption/desorption
and dissolution kinetics. Kluwer Academic Publishers; 1998. 230 p.
19. Katoh S, Yoshida F. Rates of absorption of oxygen into blood under turbulent conditions.
Chem Eng J. 1972;3:276–85.
20. Matsuda N, Sakai K. Blood flow and oxygen transfer rate of an outside blood flow membrane
oxygenator. J Membr Sci. 2000;170:153–8.
21. Weibel ER. The pathway for oxygen: structure and function in the mammalian respiratory
system. Harvard University Press; 1984. 448 p.
22. Nickalls R. Inverse solutions of the Severinghaus and Thomas equations which allow Po2 to
be derived directly from So2 [Internet]. [cited 2012 Sep 3]. Available from: www.nickalls.org/
dick/papers/anes/severinghaus.pdf.
23. Epstein FH, Hsia CCW. Respiratory function of hemoglobin. N Engl J Med.
1998;338(4):239–48.
24. Heaton A, Keegan T, Holme S. In vivo regeneration of red cell 2, 3-diphosphoglycerate
following transfusion of DPG-depleted AS-1, AS-3 and CPDA-1 red cells. Br J Haematol.
1989;71(1):131–6.
25. Brain MJ. VV-ECMO oxygen saturation tool [Internet]. [cited 2014 Nov 4]. Available from:
http://www.irenabyss.com.au/Equipment/VVECMO/.
26. Thomas LJ. Algorithms for selected blood acid-base and blood gas calculations. J Appl
Physiol. 1972;33(1):154–8.
27. Kelman GR. Digital computer subroutine for the conversion of oxygen tension into saturation.
J Appl Physiol. 1966;21(4):1375–6.
28. Arthurs GJ, Sudhakar M. Carbon dioxide transport. Contin Educ Anaesth Crit Care Pain.
2005;5(6):207–10.
29. Mochizuki M. The CO∼ 2 dissociation curve at steady state in vivo. Yamagata Med
J. 2004;22(1):25–8.
30. Beilin LJ, Knight GJ, Munro-Faure AD, Anderson J. The sodium, potassium, and water con-
tents of red blood cells of healthy human adults. J Clin Investig. 1966;45(11):1817.
31. Mochizuki M. Analysis of bicarbonate concentration in human blood plasma at steady state
in vivo. Yamagata Med J. 2004;22(1):09–24.
32. Mochizuki M, Takiwaki H, Kagawa T, Tazawa H. Derivation of theoretical equations of
the CO2 dissociation curve and the carbamate fraction in the Haldane effect. Jpn J Physiol.
1983;33(4):579–99.
33. Tazawa H, Mochizuki M, Tamura M, Kagawa T. Quantitative analyses of the CO2 disso-
ciation curve of oxygenated blood and the Haldane effect in human blood. Jpn J Physiol.
1983;33(4):601–18.
34. Fencl V, Jabor A, Kazda A, Figge J. Diagnosis of metabolic acid-base disturbances in critically
ill patients. Am J Respir Crit Care Med. 2000;162(6):2246–51.
35. Gutknecht J, Bisson MA, Tosteson FC. Diffusion of carbon dioxide through lipid bilayer
membranes. Effects of carbonic anhydrase, bicarbonate, and unstirred layers. J Gen Physiol.
1977;69(6):779.
36. Centor R. Serum total carbon dioxide - clinical methods - NCBI bookshelf. In: Walker H,
Hall W, Hurst J, editors. Clinical methods: the history, physical, and laboratory examinations
[Internet]. Butterworths; 1990 [cited 2012 Sep 22]. Available from: http://www.ncbi.nlm.nih.
gov/books/NBK308/.
37. Baker A, Richardson D, Craig G. Extracorporeal carbon dioxide removal (ECCO2 R) in respi-
ratory failure: an overview, and where next? 2C04 3A13. [cited 2012 Sep 22]; Available from:
http://journal.ics.ac.uk/pdf/1303232.pdf.
60 M. J. Brain et al.

38. ARDS Network. Ventilation with lower tidal volumes as compared with traditional tidal
volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med.
2000;342(18):1301–8.
39. Guérin C, Reignier J, Richard J-C, Beuret P, Gacouin A, Boulain T, et al. Prone positioning in
severe acute respiratory distress syndrome. N Engl J Med. 2013;368(23):2159–68.
40. Abroug F, Ouanes-Besbes L, Dachraoui F, Ouanes I, Brochard L. An updated study-level
meta-analysis of randomised controlled trials on proning in ARDS and acute lung injury. Crit
Care. 2011;15(1):R6.
41. Perier F, Tuffet S, Maraffi T, Alcala G, Victor M, Haudebourg A-F, et al. Effect of positive end-­
expiratory pressure and proning on ventilation and perfusion in COVID-19 acute respiratory
distress syndrome. Am J Respir Crit Care Med. 2020;202(12):1713–7.
42. Messerole E, Peine P, Wittkopp S, Marini JJ, Albert RK. The pragmatics of prone positioning.
Am J Respir Crit Care Med. 2002;165(10):1359–63.
43. Cove M. Ultra-protective pulmonary ventilation supported by low flow extracorporeal carbon
dioxide removal (ECCO2R) and prone positioning for ARDS; a pilot study. ClinicalTrials.gov
Identifier: NCT02252094 [Internet]. Home - ClinicalTrials.gov. 2014 [cited 2021 July 23].
Available from: https://clinicaltrials.gov/ct2/home.
44. Jung J, Lyczkowski RW, Panchal CB, Hassanein A. Multiphase hemodynamic simulation of
pulsatile flow in a coronary artery. J Biomech. 2006;39(11):2064–73.
45. Lipowsky H. Microvascular rheology and hemodynamics. Microcirculation. 2005;12(1):5–15.
46. Papaioannou TG, Stefanadis C, et al. Vascular wall shear stress: basic principles and methods.
Hell J Cardiol. 2005;46(1):9–15.
47. Lipowsky HH, Kovalcheck S, Zweifach BW. The distribution of blood rheological parameters
in the microvasculature of cat mesentery. Circ Res. 1978;43(5):738–49.
48. Ercan M, Koksal C. The relationship between shear rate and vessel diameter. Anesth Analg.
2003;96(1):307–8.
49. Walker J. Fundamentals of physics. 9th ed., extended ed. Hoboken: Wiley; 2011. 1248 p.
50. Bahrami M. Introduction to fluid mechanics [Internet]. Simon Fraser University: Lecture
Notes for Introduction to Fluid Mechanics: ENSC283. 2011 [cited 2012 Oct 13]. Available
from: http://www.sfu.ca/~mbahrami/ENSC%20283/Notes/.
51. Bloomfield LA. How everything works: making physics out of the ordinary. Wiley; 2007. 736 p.
52. Zochios V, Parhar K, Tunnicliffe W, Roscoe A, Gao F. The right ventricle in ARDS. Chest.
2017;152(1):181–93.
53. Hoffman KR, Burrell AJC, Diehl A, Butt W. Elevated carboxyhaemoglobin as a novel indicator
for extracorporeal membrane haemolysis and oxygenator exchange. Crit Care. 2021;25(1):159.
54. Burns J, Hurtado-Doce A, Lees N. Carboxyhemoglobin associated with hemolysis as a marker
of impending oxygenator failure in VA ECMO. Crit Care Med. 2015;43:38.
55. Omar HR, Mirsaeidi M, Socias S, Sprenker C, Caldeira C, Camporesi EM, et al. Plasma free
hemoglobin is an independent predictor of mortality among patients on extracorporeal mem-
brane oxygenation support. Frati G, editor. PLoS ONE. 2015;10(4):e0124034.
56. Ghitescu I, Copotoiu SM, Toma RS, Ghitescu VT, Copotoiu R. Mean filter life span in con-
tinuous veno-venous hemofiltration for septic patients. Jurnalul Roman de Anestezie Terapie
Intensiva/Rom J Anaesth Intensive Care. 2009;16(1):17–22.
57. Brain M, Winson E, Roodenburg O, McNeil J. Non anti-coagulant factors associated with filter
life in continuous renal replacement therapy (CRRT): a systematic review and meta-analysis.
BMC Nephrol. 2017;18:69.
58. Jackson SP. The growing complexity of platelet aggregation. Blood. 2007;109(12):5087–95.
59. GlaxoSmithKline. Argatroban injection prescribing information [Internet]. 2008 [cited
2021 July 23]. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/
label/2008/020883s014lbl.pdf.
60. Ruby K, Salvatore L, Lomachinsky R, Ornstein D, Khan J. Hemolysis interference in the anti-
­Xa heparin activity assay - ISTH congress abstracts. Res Pract Thromb Haemost [Internet].
1 Physiology of Extracorporeal Life Support (ECLS) 61

2021 [cited 2021 July 23];5 (Suppl 1). Available from: https://abstracts.isth.org/abstract/
hemolysis-­interference-­in-­the-­anti-­xa-­heparin-­activity-­assay/.
61. Khan J, Chandler WL. Interference in the anti-Xa heparin activity assay due to hemolysis and
icterus during pediatric extracorporeal life support. Artif Organs. 2019;43(9):880–7.
62. Hogan BLM, Barkauskas CE, Chapman HA, Epstein JA, Jain R, Hsia CCW, et al. Repair and
regeneration of the respiratory system: complexity, plasticity, and mechanisms of lung stem
cell function. Cell Stem Cell. 2014;15(2):123–38.
63. Dunnill MS. Postnatal growth of the lung. Thorax. 1962;17(4):329–33.
64. Narayanan M, Owers-Bradley J, Beardsmore CS, Mada M, Ball I, Garipov R, et al.
Alveolarization continues during childhood and adolescence: new evidence from helium-3
magnetic resonance. Am J Respir Crit Care Med. 2012;185(2):186–91.
65. Thurlbeck WM. Postnatal human lung growth. Thorax. 1982;37(8):564–71.
66. Matthay MA, Zemans RL, Zimmerman GA, Arabi YM, Beitler JR, Mercat A, et al. Acute
respiratory distress syndrome. Nat Rev Dis Primers. 2019;5(1):18.
67. Mora AL, Rojas M. Adult stem cells for chronic lung diseases: B-MSCs aging and lung repair.
Respirology. 2013;18(7):1041–6.
Chapter 2
Circuits, Membranes, and Pumps

Bradley H. Rosen

Introduction

Modern ECLS is based on highly efficient, low-resistance, gas-exchanging mem-

branes. In order to couple the patient and artificial lung, vascular access is required
(see Chap. 4), along with tubing, a pump, and assorted means for monitoring, safety,
and infusing medications. Clinicians caring for these patients require a working
knowledge of the circuit so as to understand its clinical implications, recognize
when something goes awry, and know how to intervene. This chapter describes the
components of the circuit, providing the practitioner with an understanding of how
they function and interact. It is divided into two large sections: The first describes
the anatomy of the overall ECLS circuit, and the second the physiology and normal
operation of each of the components.

Circuit Anatomy

Overall Circuit Considerations

Circuit designs all attempt to balance efficacy, safety, convenience, and simplicity.
There is no one-size-fits-all solution, however, since varied patients, circumstances,
and clinician preferences may necessitate that safety override simplicity or that

B. H. Rosen (*)
Division of Pulmonary, Critical Care, and Occupational Medicine, Department of Internal
Medicine, Roy J. and Lucille A. Carver College of Medicine, University of Iowa Hospitals
and Clinics, Iowa City, IA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 63

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_2
64 B. H. Rosen

portability trump efficacy. For example, inserting multiple stopcocks into the circuit
can allow easy access for renal replacement therapy: one circuit carries out gas
exchange and dialysis. This is convenient, but each additional connector represents
an opportunity for failure (leak, thrombosis, air entrainment, rupture). In an indi-
vidual circumstance, whether to conduct renal replacement using the ECLS circuit
may depend on the ease of obtaining alternate venous access, the expected duration
of renal failure, or the ECLS physician’s experience and preference with regard to
circuit complexity. Thus, circuits range from rather complex designs incorporating
many safety and monitoring functions (see Fig. 2.1) to minimalistic, simpler layouts
that lack the various bells and whistles (Fig. 2.2).

Fig. 2.1 Schematic of a complex circuit design that implements all optional features, including
separate non-integrated blood analyzers, a manifold with access sites (closed unless being
accessed), and renal replacement circuit integration options. The dotted line indicates where the
manifold may be inserted between the pump and the oxygenator, if there is sufficient tubing to
allow for this

Fig. 2.2 Similar to the previous Fig. 2.1 but simplified with only necessary components: pump,
oxygenator, and flow probe/bubble sensor. The blood analyzers are internalized within the pump
and oxygenator. Renal replacement circuit is included here along the manifold, which derives post-­
oxygenator and returns pre-pump. Some centers do not include a manifold, simplifying the circuit
even further
2 Circuits, Membranes, and Pumps 65

In designing a circuit, simplicity is one of the paramount concerns. While com-

ponents and connectors can be cut into a circuit after purchase, each modification
produces a weak point susceptible to rupture or fibrin accumulation due to turbu-
lence. Any such alteration should be performed while the circuit is “dry” (prior to
priming, see below) and with regard for sterility. Additionally, each Luer lock is a
site of potential air entrainment, blood leak, or microbial contamination. The major-
ity of connections and access ports are located on the venous side of the circuit,
ideally between the pump and the oxygenator. This is intentional: Limited connec-
tors on the arterial side reduces the potential for accidental exsanguination, while
the similar reduced number of connectors proximal to the pump inlet limits the risk
of air entrainment and gas embolism due to negative pressures.
All circuits should involve as little tubing as possible while allowing adequate
spacing of components and facilitating mobilization of the patient. Greater tubing
lengths incur more resistance to flow (proportional to length according to Poiseuille’s
law), necessitating higher circuit pressures and leading to more damage to blood
elements.
Circuit length and complexity also relate to the degree and duration of time to
which blood is exposed to plastic surfaces, and this interaction elicits an inflamma-
tory response. It is believed that induced inflammation may further compromise
lung function, leading to further gas exchange deterioration. It has been hypothe-
sized that heparin coating of polymethylpentene (PMP) oxygenators serves to
reduce this response [1]. An added consequence of circuit-induced inflammation is
excessive fibrinogen production leading to increased fibrin deposition on circuit
surfaces. Further, this inflammation promotes platelet adherence, elevating the risk
of thrombosis which impairs oxygenator function [2]. Cellular deposition along the
membrane surface (on the blood side) correlates with a rising resistive pressure
across the device [3]. Some fibrin deposition within the oxygenator and circuit is
unavoidable (apparent first on the venous side), but excessive deposition is deleteri-
ous to circuit function. For this reason, sides of PMP oxygenators are clear to facili-
tate monitoring of fibrin deposition over the course of an ECLS run.
There are several additional implications of circuit length. The greater the sur-
face area, the more that medications commonly used in the care of critically ill
patients are subject to adsorption. Antibiotics (meropenem, cefazolin, and vanco-
mycin), sedatives (midazolam), and analgesics (morphine, fentanyl, and acetamino-
phen) are all meaningfully adsorbed, to a degree related to the lipophilic nature of
the drug. Antibiotics are only moderately affected (65–85% recovered after
180 min), but midazolam and fentanyl are severely adsorbed with less than 1%
recovered [4]. Even if an agent is not adsorbed, the extracorporeal circuit expands
the volume of distribution of any pharmaceutical due to the volume of blood within
the circuit itself (up to 1 L). Tubing length will also contribute significantly to the
volume required to prime the circuit, producing hemodilution and increasing dwell
time, the latter of which can affect anticoagulant needs during ECLS support.
Finally, tubing surface area also relates to the degree of heat loss. This can be sub-
stantial, such that ECLS circuits must incorporate a means for temperature control
(see section “Heat Exchanger and Heater-Cooler” below).
66 B. H. Rosen

Circuit Priming

Priming refers to the process by which the gas (ambient air present at manufacture)
is replaced with a physiologically compatible fluid. For ECLS in adults, the circuit
is primed with crystalloid fluids, such as normal saline, Ringer’s lactate, or propri-
etary mixed electrolyte solutions (e.g., Plasma-Lyte® or Normosol®). Purchased
circuits come attached to a large, empty priming bag. The bag is filled with sterile
crystalloid, clamps are opened to the venous and arterial limbs, and the priming bag
is raised to allow gravity to move the fluid into the circuit components while air
moves to the priming bag. The volume necessary to prime a given circuit depends
on the priming volume of each component (oxygenator, heat exchanger blood
phase, pump, manifold, and tubing) and directly relates to the degree of hemodilu-
tion that follows. For pediatric and neonatal ECLS, hemodilution is prevented by
priming the circuit with blood, but this is not necessary for adults where the typical
priming volume averages 500–1000 mL for a complex circuit design and for a sim-
ple one as low as 300 mL. In fully primed condition, a circuit can be stored for a
period of at least 30 days, although each institution has its own policies regarding
shelf life. Priming with a colloid may shorten the shelf life of a primed circuit,
another reason many programs choose a crystalloid prime. A simplified ECLS cir-
cuit can be fully primed in less than 10 min due to the microporous nature of the
membranes in use. Once the circuit is primed, the heat exchanger can be turned on
to raise the temperature to 37 °C before connecting the patient, as long as time per-
mits. Cardiopulmonary bypass circuits are often primed first with carbon dioxide
(to displace oxygen and nitrogen), hastening the subsequent fluid priming process,
but this is not generally done for ECLS circuits.

Orientation to the Circuit

ECLS circuits can appear intimidating, especially when one realizes that 5 L of
blood can rush through it each minute. A systematic approach to the intricacies of
the circuit and its components keeps the clinician from becoming overwhelmed, so
we begin with a brief, general tour. Figure 2.1 represents a comprehensive sche-
matic of a circuit, whereas Fig. 2.2 shows a greatly simplified design with few
extraneous components. In each instance, we describe the circuit beginning with the
drainage (outflow) cannula, proceeding through the gas-exchanging membrane, and
ending back at the patient through the return (inflow) cannula. Dual-lumen cannulas
allow blood to exit and enter at the same site but; for illustration purposes, we’ve
separated these in the figures.
Starting with the drainage cannula at its exit from the patient (internal jugular or
femoral vein, or right atrium), the distal end is connected to large-diameter conduct-
ing tubing. This is an important point for inadvertent disconnection, especially
immediately following the initiation of ECLS if the tie bands were not securely
2 Circuits, Membranes, and Pumps 67

fastened. In addition, like other areas of the circuit where there is turbulence or
stasis, this is a common site for thrombus to form. Careful examination of this con-
nection is an essential part of the regular circuit check (see Chap. 12). The conduct-
ing tubing should be kept relatively short in order to reduce resistance to blood flow,
surface area of contact with blood, and the priming volume. The conducting tubing
leads to a centrifugal pump before entering the membrane oxygenator. As there is
considerable heat loss as the blood traverses the circuit, a heat exchanger is neces-
sary to rewarm the blood to body temperature (this may be incorporated into the
oxygenator and hidden from direct view). If added separately, the heat exchanger is
placed proximal to the oxygenator. The oxygenator also receives the sweep gas
(usually medical oxygen), being joined to wall oxygen through a blender or to an
E-cylinder through a flow meter. The newly arterialized blood completes its extra-
corporeal course through the return cannula, delivering oxygenated and warmed
blood to the vascular system.
While a bridge between the outflow and inflow cannulas is often used as a shunt
to conduct weaning trials during veno-arterial (VA) ECLS, such a bridge is not
needed for veno-venous (VV) support since weaning can be conducted by reducing
or eliminating gas flow to the membrane while leaving circuit flow to the patient
undisturbed.
In order to monitor circuit function and prevent complications, devices to mea-
sure pressure and flow and to detect bubbles are included, and information is relayed
to a console (see Fig. 2.3 for an example). Typically, fluid pressure is measured on
the venous side of the pump (“PINLET”), providing information about how much suc-
tion is required to draw the needed circuit blood flow. Two additional pressure trans-
ducers (“PPRE” and “PPOST”) flank the oxygenator so that the membrane pressure
drop (ΔP) can be calculated. In addition, PPOST displays the pressure that drives flow
back to the patient. Circuit blood flow is monitored using an ultrasonic flow probe,
since centrifugal pumps do not guarantee a fixed relationship between revolutions
per minute and volume displaced, as was true for roller pumps. The flow probe may
be integrated within the pump or added as an aftermarket device. Ultrasound probes
are also used to identify bubbles, so some circuit designs utilize the same sensor for
both flow measurement and bubble detection. Bubbles distal to the oxygenator can
produce systemic embolism and are especially dangerous in VA modes but less so
in VV as the patient’s lungs will provide protection from air embolism.
Spectrophotometric sensors allow real-time measurement of such values as PO2,
PCO2, pH, inlet saturation (SPREO2), outlet saturation (SPOSTO2), and hemoglobin
concentration, among others. These sensors must be calibrated periodically by com-
paring the displayed value against a blood sample analyzed simultaneously using
conventional laboratory methods. The console receives data from various devices
along the circuit, displaying pump speed, flow, pressures, temperature, and other
physiological information. The console also may display alarm notifications and
allows the user to adjust the pump, heat exchanger, and other functions. The console
generally is integrated with the power supply and battery.
Ports are included in the circuit so that blood can be sampled and agents can
be infused. These are often collected in a manifold consisting of a series of Luer
68 B. H. Rosen

Fig. 2.3 ECLS console showing relevant pressures within the circuit (PINLET, PRE, POST, same as in
Figs. 2.1 and 2.2) along with pump speed (rpm) and flow (L/min) along the top of the screen, and
SPREO2 in the lower right of the screen with a battery charge indicator in the corner. The dial and
buttons along the right of the panel are used to adjust parameters, in addition to the touch screen
itself. At the top of the figure is a colored bar that can be used for rapid visualization of the
pump speed

lock connections with a three-way stopcock controlling flow to each. The mani-
fold derives from a region either after the oxygenator or between the centrifugal
pump and the oxygenator (a “safe zone” of interruption) and re-infuses proximal
to the pump so that small amounts of entrained air can be eliminated by the oxy-
genator. With more compact systems, the pump and oxygenator are often adja-
cent without tubing or connectors between the devices, and the manifold will
generally originate post-oxygenator and return pre-pump. Various infusions of
medications and anticoagulant agents may be connected to the circuit through
these ports, as sufficient flow can be drawn from the circuit so as to combine
renal replacement therapy (RRT) and gas exchange simultaneously, avoiding the
need for invasive vascular access solely for dialysis. Integration of RRT is often
through the manifold, as shown in Figs. 2.1 and 2.2, although there are variations
that may involve keeping the system completely pre-pump or interface directly
from ports on the oxygenator itself. Any ports along the circuit tubing itself tend
2 Circuits, Membranes, and Pumps 69

to begin as pigtails that end in a stopcock, since breakage of this can be clamped,
while integrated Luer locks that break cannot be managed without a circuit
exchange.

Function of the Circuit Components

Cannulas and Tubing

Single- and double-lumen cannulas are described more fully in Chap. 4. Cannulas
tend to be wire-reinforced to limit kinking and occlusion. They are attached to the
circuit tubing by means of adaptors, and these connections are secured by tie bands.
Tubing is clear, medical grade polyvinylchloride (PVC) allowing the clinician to
recognize blood color (as a clue to circuit function and recirculation) and to identify
fibrin, clots, and gas bubbles. Tubing can be clamped and, when changing out a
circuit due to oxygenator failure for instance, cut and reconnected to reinstitute
circuit flow.

Centrifugal Pumps

There are two types of pumps that have been employed in ECLS circuits: roller/
occlusive and centrifugal pumps. Practitioners of modern ECLS have settled on the
centrifugal pump design as the safer of the two, and this is the type that will be dis-
cussed here.
These devices increasingly employ a magnetically driven impeller within a spiral
housing. While many impellers are completely levitated by magnets, some have a
spindle to which fibrin can deposit, and this can contribute to hemolysis. The impel-
ler imparts mechanical energy to the blood, raising velocity and pressure as it moves
from the center of the vortex to the periphery. The housing constrains and directs the
blood flow toward the circumference where it exits the pump. This principle differs
entirely from that employed in roller head pumps, producing several advantages and
compromises. Where roller head pump flow is independent of afterload (to the point
of causing tubing rupture!), a centrifugal pump is unable to overcome excessive
afterload. Instead, flow will drop as afterload increases, despite an unchanged pump
rotation speed. In a similar vein, a centrifugal device is unlikely to cause cavitation
at the outflow cannula, as it is unable to generate sufficiently negative pressure. Of
course, this means that hypovolemia tends to threaten the adequacy of circuit flow
when a centrifugal pump is used. This lack of absolute relationship between pump
revolutions and blood flow necessitates a flow meter. An increasing discrepancy
between pump speed and measured flow is a strong signal of trouble with regard to
function of the circuit.
70 B. H. Rosen

In the case of power loss, many of these pumps can be hand-cranked. Internal
batteries can provide up to a couple hours of support without an external power
source, depending on the system and pump speeds needed. Additional discussion of
ECLS crises can be found in Chap. 13.
Advantages of centrifugal pumps over roller pumps include reduced tubing
length, safety benefits due to absent issues of spallation (i.e., liberation of micro-
scopic particles of tubing into the circulation), microemboli, raceway rupture, and
greater flexibility in circuit design because the pump does not need to be at the base
of the unit. Disadvantages are fewer, including direct priming volume required for
the pump (although tubing length is reduced) and blood flow that depends on pre-
load and afterload (requiring a blood flow probe). Centrifugal pumps were initially
reported to produce unacceptable rates of hemolysis due to heat generation, but
engineering improvements have solved this problem [5]. Designs have evolved to
reduce hemolysis and the risk of gaseous microembolism [6, 7], changing the land-
scape of ECLS circuits and leading to worldwide adoption of the technology [8, 9].

Membrane Oxygenators

The ideal membrane lung would be highly permeable to relevant gases (O2 and
CO2) while resisting fluid transudation from the blood to the gas phase (termed
“plasma leak”). Blood should flow through the device with minimal resistance,
allowing high flows with little pressure and without trauma to blood elements.
Surfaces exposed to blood would only minimally activate the host coagulation and
immune systems. These properties would be complemented by reliability, durabil-
ity, and a small priming volume. The human lung juxtaposes blood and gas over a
tremendous surface area, with incredibly thin diffusion distances, yet maintains a
clear separation between blood and gas phases. Scientists and engineers have strug-
gled to mimic these attributes: the history of ECLS is a remarkable story of inspira-
tion, invention, and persistence.
Both PMP and polypropylene hollow-fiber membranes employ large numbers of
fine capillary tubes to carry the sweep gas while being bathed by flowing blood.
This extra-capillary blood flows countercurrent to the direction of gas movement,
increasing the efficiency of gas exchange. It is important to realize the stark contrast
in surface area when comparing an oxygenator with the human lung it attempts to
replace: Most PMP devices provide at most two square meters of gas exchange
surface, while the lung exposes upward of 70 square meters to blood flow.
Additionally, in the best membrane lungs, oxygen must diffuse 150 μm from sweep
gas to blood, whereas the comparable distance within the human lung is a mere
0.5 μm. These disadvantages are offset by increasing the effective dwell time of the
blood within the artificial lung. In addition, so-called “secondary flows,” which
describe the mixing of blood around the gas-fluid interface due to purposefully cre-
ated turbulence, further enhance gas transfer. Blood cells are regularly being brought
into close approximation with the gas-filled capillaries, effectively reducing
2 Circuits, Membranes, and Pumps 71

Fig. 2.4 Oxygen transfer O2 Transfer

in mL/min as a function of
500
blood flow in L/min

400

O2 Transfer (mL/min)
300

200

100

0
0 2 4 6 8
Blood Flow (L/min)

diffusing distance [10, 11]. This layout allows for up to a two and a half-fold reduc-
tion in surface area necessary for gas exchange [12]. Typical gas-exchanging capac-
ities for membrane lungs are shown in Figs. 2.4 (oxygen) and 2.5 (carbon dioxide).
Providing sufficient oxygen transfer to meet the entire metabolic demand (roughly
250 mL O2/min) requires blood flow of roughly 4 L/min. In contrast, carbon dioxide
transfer is relatively advantaged so that much less blood flow is required, especially
at very high sweep gas flow rates. For example, nearly all of the metabolically pro-
duced carbon dioxide can be eliminated with only 1 L/min of blood flow, especially
at high gas flows (see Chap. 10) [13].
The most common external appearance of PMP oxygenators is an extruded
square evenly balanced on one corner (see Fig. 2.6). At the lower corner, blood
enters the device from the pump under pressure denoted as “pre-membrane pres-
sure” (PPRE); typical pressures are in the range of 225–275 mmHg, certainly less
than 400 mmHg; (see Table 2.1). The blood ascends to the opposite corner at the
highest elevation of the device and then flows down to the exit connector directly
opposite the inlet, by which point it is oxygenated and carbon dioxide has been
removed. At that point, the pressure (“post-membrane pressure,” PPOST) will be less
than PPRE. The difference between these is called the “ΔP”, representing the resis-
tive pressure drop across the membrane at the current flow. With the current genera-
tion of PMP devices, ΔP should range from the teens to low thirties at typical circuit
blood flow rates (see also Chap. 9).
With regard to the pressures and the information that may be gleaned from their
trends, an elevated PPRE has variable implications depending on whether PPOST is
also elevated. In the case where both are elevated (ΔP is preserved), the circuit
should be examined, not the oxygenator: the distal tubing and circuit may be
obstructed by kinks or thrombosis. Similar findings are seen transiently when
patients cough, Valsalva, or are suctioned. When PPRE rises along with an increase in
ΔP, resistance within the membrane is excessive, raising concerns for thrombosis,
heparin-induced thrombocytopenia, or accumulation of fibrin or cellular elements
72 B. H. Rosen

CO2 Transfer
500
CO2 2:1
400 CO2 1:1
CO2 Transfer (mL/min)

CO2 0.5:1
300

200

100

0
0 2 4 6 8
Blood Flow (L/min)

Fig. 2.5 Carbon dioxide transfer in mL/min as functions of both blood flow in L/min and sweep
gas flow

Fig. 2.6 Photograph of a

membrane oxygenator
(small adult model).
Identified on this model are
the blood inlet and outlet
ports, sweep gas, as well as
the two connections for the
water heater. A US quarter
dollar is shown for scale
2 Circuits, Membranes, and Pumps 73

Table 2.1 Circuit pressures

Location in the circuit Normal operating pressure

Proximal to the pump: PINLET −50 to −200 mmHg
Oxygenator inlet: PPRE 225 to 275 mmHg
Oxygenator outlet: PPOST 190 to 260 mmHg
ΔP (PPRE – PPOST) 10 to 35 mmHg

on the membrane as viewed through the transparent sides of the device. An increas-
ing ΔP along with evidence of impaired gas exchange, which would show as
reduced PO2 and elevated PCO2 from a post-oxygenator blood sample, at a given
flow and sweep portends failure of the membrane (see Chap. 9).
Another major advance in modern PMP oxygenators is their low resistance to
blood flow, producing several important advantages. First, this property ushered in
the era of lower pressure, afterload-sensitive centrifugal pumps, affording the safety
features discussed above. Secondly, lower pressures translate to safer, longer-lived
circuits, conferring additional safeguards against catastrophic rupture or circuit fail-
ure. Additionally, such low resistance to blood flow permits novel applications of
ECLS such as pumpless arteriovenous extracorporeal CO2 removal (AV ECCO2R)
that rely solely on the difference between arterial and venous blood pressures to
drive flow [14] (see Chap. 3).
In the modern era of ECLS, the centerpiece of the circuit is a PMP oxygenator,
and these devices have eclipsed prior generations of artificial lung. Nevertheless,
other oxygenator designs are seen occasionally and include both silicone mem-
branes first developed by Kolobow [15] (most recently marketed as the Medtronic
1–4500-A2) as well as the polypropylene microporous hollow-fiber membrane.
Silicone membrane oxygenators employed sheets enclosing a plastic polymer
screen and wrapped around a polycarbonate core. They remain the only gas
exchange device that is FDA-approved for long-term use (defined as use for more
than 6 h). However, they require priming with considerable volume (665 mL each
[16]), exhibit a large pressure drop across the membrane that limits the use of cen-
trifugal pumps, and are relatively inefficient in gas exchange so that at least two
large surface area units are needed per patient. Polypropylene hollow-fiber mem-
branes are highly efficient with respect to gas exchange but tend to develop plasma
leak (described further below). They also present a low resistance to blood flow,
need only a small priming volume, and remain in use for cardiopulmonary bypass
where they provide excellent short-term support.

Additional Limitations of Membrane Oxygenators

Rated Flow: Blood exiting the membrane is normally fully oxygenated, typically
with a PO2 in excess of 300 mmHg. As blood flow is increased, greater demands are
placed on the capacity for gas diffusion across the hollow fiber barrier. In part, this
74 B. H. Rosen

relates to the simple volume of oxygen that must diffuse as more deoxygenated
blood is pushed through the membrane but also to the increasing blood velocity
(thus reduced dwell time) produced at higher flows. At sufficiently high flows, the
membrane fails to fully saturate the blood. The flow threshold for full oxygenation
is termed the “rated flow.”
Plasma Leak: Hollow fiber membranes should be sufficiently permeable to allow
rapid gas diffusion while remaining impermeable to fluid movement. Plasma leak is
the phenomenon whereby plasma phospholipids leak from the circulating whole
blood to the gas compartment of the oxygenator and then serve to propagate further
plasma leakage in a positive feedback cycle [17]. Small to moderate amounts of
clear fluid may normally traverse the membrane, particularly with lower sweep gas
flows, and a drainage port is provided for egress of this condensation. Excessive
condensation will reduce oxygenator efficiency, which will lower post-oxygenator
PO2, and can be corrected when the clinician increases the sweep (>10 L/min) and
jostles the membrane by knocking on the housing to “cough” or “burp” the mem-
brane. Greater volumes of heme-tinged or proteinaceous-appearing fluid emanating
from the drainage port signal a failing membrane, severely impairing the efficiency
of gas exchange, accompanied by a rise in ΔP, and eventually requiring exchange
of the device. Clinically significant plasma leak can be confirmed by analysis of
liquid from the gas outlet for proteins [18]. Additional implications are a significant
loss of proteins and immunoglobulins, potentially significant from immunologic
and nutritional standpoints. PMP membranes are much less susceptible to plasma
leak than prior artificial lungs, yet they are not fully immune to this complica-
tion [19].

Renal Replacement Therapy

Acute kidney injury (AKI) is highly prevalent in those undergoing ECLS, affecting
up to 70% of patients in some case series [20, 21]. The resulting fluid overload is a
frequent indication for renal replacement therapy (RRT) as it will impede native
lung function and hinder efforts at animation of the patient during ECLS. The initial
management of non-oliguric renal failure can be accomplished by loop diuretics—
continuous infusions are preferred to avoid rapid changes in volume status—but
more precise management of volume status and metabolic derangements can be
achieved with continuous renal replacement therapy (CRRT). In addition to volume
management, CRRT may be used to avoid both profound metabolic acidosis, which
may increase respiratory drive such that it defeats the goal of ECLS to reduce pul-
monary injury, and reduced renal clearance of medications or uremia that can con-
tribute to acute encephalopathy. Accordingly, the decision regarding timing of
CRRT initiation should be individualized as with any critically ill patient and is
discussed in Chap. 12.
If already in place at the time of ECLS, standard dialysis catheters may be used
for CRRT as per usual protocols; however, given that many patients develop AKI
2 Circuits, Membranes, and Pumps 75

during their ECLS run, integration of CRRT into the circuit is common. There is
anecdotal evidence that use of the ECLS circuit for CRRT access can extend the
life of dialysis membranes and by reducing patient lines can facilitate ambulation
while on ECLS support. Disadvantages of integrating RRT systems into the ECLS
circuit include the need for additional connections and thus sites of air entrain-
ment, blood loss, and infection, along with risk of losing both circuits if one is
compromised. While there are numerous ways to integrate the CRRT, the return
should enter before the oxygenator to allow the membrane to eliminate any
entrained air.
The most common method of CRRT integration (see Fig. 2.1) is to have the
intake enter the CRRT post-pump, either from the arterial side post-membrane or
from between the pump and oxygenator (dotted line in Fig. 2.1). While it is possible
to have the positive pressure from the ECLS pump drive flow through the dialysis
filter on its own, such passive configurations are uncommon as the clinician loses
independent control of dialysis when circuit flows are adjusted for respiratory sup-
port. Instead, it is more common that standard CRRT systems are implemented in a
parallel circuit that will introduce a small shunt (generally <10% of flow runs
through CRRT), although pressure alarms frequently need to be adjusted to allow
for a much greater positive pressure than expected in normal CRRT runs.
Additionally, the outflow from the CRRT circuit will increase the PINLET, sometimes
enough that the pressure becomes positive. So long as the clinical staff are aware of
the small shunt fraction and adjust pressure alarms, this tends to be well-tolerated
by the other circuit components.
Anticoagulation used for ECLS is sufficient for CRRT; although if specific rea-
sons preclude anticoagulant use (e.g., peri-procedural hemostasis or pulmonary-­
renal syndromes, etc.), citrate can be run within the dialysis circuit alone. The
greater flows and degree of anticoagulation used in ECLS tend to allow for pro-
longed CRRT circuit longevity, with some centers being able to run up to 7–10 days
between changing dialysis filters.

Circuit Surface Coatings

The blood-circuit interface activates neutrophils, produces inflammation, and trig-

gers coagulation [22]. Changing from the silicone rubber membranes to PMP oxy-
genators allowed manufacturers to bond the surface of membrane oxygenators with
heparin or protein coatings similar to those found on other circuit components.
Various manufacturers have implemented this in diverse ways, altering the process
by which the molecule is attached to the surface and the binding sites that are acces-
sible to blood components. There is experimental evidence that heparin bonding of
cardiopulmonary bypass circuits for coronary artery bypass grafting attenuates the
inflammatory response as measured by terminal complement complex formation
and neutrophil elastase concentrations [23]. Complement and platelet activation is
limited as well [23, 24]. The effect in patients seems limited to a reduction in
76 B. H. Rosen

anticoagulation requirements [25], reduced plasma-free hemoglobin, and lower

TNF-α (tumor necrosis factor alpha) concentrations [26].
Although it is presumed that there is benefit to the practice of heparin bonding,
the data do not consistently support benefit when patient-centered outcomes are the
primary outcome. Coating could be beneficial in those who have an elevated pretest
probability or risk of hemorrhage, but overall there is controversy regarding the
value of heparin coatings [27, 28].
While heparin-induced thrombocytopenia (HIT) is not a common occurrence in
critical care with an estimated incidence of 0.3–0.6% [29], there are instances
reported in patients being supported by ECLS [30]. Moreover, the diagnosis of HIT
is fraught because thrombocytopenia is so common during ECLS for myriad rea-
sons (see Chap. 8). When HIT is confirmed (such as through serotonin-release
assay), infused heparin should be discontinued, while anticoagulation is maintained
using direct thrombin inhibitors such as argatroban [31] or bivalirudin [27].
However, the clinician should be mindful that circuit components are often heparin
bonded and that there is anecdotal evidence that this can be clinically significant
[32]. Manufacturers generally do have devices available without these biologic
coatings for patients with confirmed HIT.

Heat Exchanger and Heater-Cooler

While on ECLS, the patient’s blood is exposed to the environment of the outside
world, risking substantial heat loss and even hypothermia. Left unchecked, this
could produce coagulopathy, circulatory failure, and other organ dysfunction. The
heat exchanger consists of an external electric heater-cooler which pumps
temperature-­controlled water through tubing to the heat transfer unit. The heat
transfer unit separates the water phase from the circulating blood by a temperature-­
conductive material comprised of stainless steel, aluminum, or polypropylene. This
material is often coated with polymers or other surface coatings to abrogate blood
component activation that might stimulate inflammation or clotting. The blood and
water phases are directed in a countercurrent fashion to maximize efficiency of heat
transfer. While the trend is toward smaller units, larger heat exchangers are more
effective in regulating temperature but require larger priming volumes that contrib-
ute to hemodilution.
Heat exchangers may be integrated into the membrane lung or added as a stand-­
alone device. After blood is fully saturated with oxygen, warming carries the poten-
tial to release microbubbles. Thus, the heat exchanger is often placed proximal to
the oxygenator, which can then trap any oxygen released.
When circuit blood flow is sufficiently high, the heat exchanger can be effective
in stabilizing body temperature. Accordingly, the clinician should be aware that
fever can be masked and should thus pay close attention to other indicators of infec-
tion. This function of the ECLS circuit can be used to prevent hyperthermia in
patients following cardiac arrest who are undergoing targeted temperature
2 Circuits, Membranes, and Pumps 77

management. However, with smaller devices and lower flows, the efficiency can be
reduced considerably, such that patients can not only mount a fever, but even raise
the water bath above the set temperature.
Complications related to the heat exchanger are few. Leakage of water into the
blood phase will result in hemolysis, so this possibility should be included in the
differential diagnosis whenever an ECLS patient has unexplained hemolysis.
Finally, the water bath of the heat exchanger can serve as a reservoir for microbes.
Water should be exchanged on a regular basis, and some recommend surveillance
cultures to detect potentially pathogenic organisms.

Conclusions

The ECLS circuit can be an intimidating arrangement of sophisticated devices that

propel more than half a patient’s blood volume through a series of unfamiliar com-
ponents. With a functional understanding of how the components operate and work
together to provide gas exchange, the clinician can customize a circuit to best serve
the patient’s needs and anticipate problems before they escalate toward crises.

Acknowledgements The author wishes to thank the following content consultants:

Kristina L. Rudolph, B.S.N.; Jennifer Crumley, M.S.N.; Tom Rath, C.C.P.; and Elizabeth
Moore, M.S.N.

References

1. Khoshbin E, Dux AEW, Killer H, Sosnowski AW, Firmin RK, Peek GJ. A comparison of
radiographic signs of pulmonary inflammation during ECMO between silicon and poly-­
methyl pentene oxygenators. Perfusion. 2007;22:15–22.
2. Peek GJ, Firmin RK. The inflammatory and coagulative response to prolonged extracorporeal
membrane oxygenation. ASAIO J. 1999;45(4):250–63.
3. Lehle K, Philipp A, Gleich O, et al. Efficiency in extracorporeal membrane oxygenation-­
cellular deposits on polymethylpentene membranes increase resistance to blood flow
and reduce gas exchange capacity. ASAIO J. 2008;54(6):612–7. https://doi.org/10.1097/
MAT.0b013e318186a807.
4. Wildschut ED, Ahsman MJ, Allegaert K, Mathot RAA, Tibboel D. Determinants of drug
absorption in different ECMO circuits. Intensive Care Med. 2010;36:2109–16.
5. Lawson DS, Ing R, Cheifetz IM, et al. Hemolytic characteristics of three commercially
available centrifugal blood pumps. Pediatr Crit Care Med. 2005;6(5):573–7. https://doi.
org/10.1097/01.pcc.0000163282.63992.13.
6. Burnside J, Gomez D, Preston TJ, Olshove VF, Phillips A. In-vitro quantification of gas-
eous microemboli in two extracorporeal life support circuits. J Extra Corpor Technol.
2011;43:123–9.
7. Yee S, Qiu F, Su X, et al. Evaluation of HL-20 roller pump and Rotaflow centrifugal pump
on perfusion quality and gaseous microemboli delivery. Artif Organs. 2010;34(11):937–43.
https://doi.org/10.1111/j.1525-­1594.2010.01079.x.
78 B. H. Rosen

8. Lawson DS, Lawson AF, Walczak R, et al. North American neonatal extracorporeal mem-
brane oxygenation (ECMO) devices and team roles: 2008 survey results of Extracorporeal Life
Support Organization (ELSO) centers. J Extra Corpor Technol. 2008;40(3):166–74.
9. Lawson DS, Walczak R, Lawson AF, et al. North American neonatal extracorporeal membrane
oxygenation (ECMO) devices: 2002 survey results. J Extra Corpor Technol. 2004;36(1):16–21.
10. Drinker PA, Bartlett RH, Rishon MB, Noyes BS. Augmentation of membrane gas transfer by
induced secondary flows. Surgery. 1969;66(4):775–81.
11. Diller TE, Mikic BB, Drinker PA. Shear-induced augmentation of oxygen transfer in blood. J
Biomech Eng. 1980;102:67–72.
12. Gaylor JDS. Membrane oxygenators: current developments in design and application. J
Biomed Eng. 1988;10:541–7.
13. Mueller T, Lubnow M, Philipp A, et al. Extracorporeal pumpless interventional lung assist in
clinical practice: determinants of efficacy. Eur Respir J. 2009;33:551–9.
14. Johnson P, Frohlich S, Westbrook A. Use of extracorporeal membrane lung assist device
(Novalung) in H1N1 patients. J Card Surg. 2011;26(4):449–52. https://doi.org/10.1111/j.1540-­
8191.2011.01261.x.
15. Kolobow T, Bowman RL. Construction and evaluation of an alveolar membrane artificial
heart-lung. ASAIO J. 1963;9:238–43.
16. Khoshbin E, Roberts N, Harvey C, et al. Poly-methyl pentene oxygenators have improved
gas exchange capability and reduced transfusion requirements in adult extracorpo-
real membrane oxygenation. ASAIO J. 2005;51(3):281–7. https://doi.org/10.1097/01.
mat.0000159741.33681.f1.
17. Montoya JP, Shanley CJ, Merz SI, Bartlett RH. Plasma leakage through microporous mem-
branes: role of phospholipids. ASAIO J. 1992;38(M399-M405):M399–405.
18. Eash HJ, Jones HM, Hattler BG, Federspiel WJ. Evaluation of plasma resistant hollow fiber
membranes for artificial lungs. ASAIO J. 2004;50(5):491–7. https://doi.org/10.1097/01.
mat.0000138078.04558.fe.
19. Puis L, Ampe L, Hertleer R. Case report: plasma leakage in a polymethylpentene oxy-
genator during extracorporeal life support. Perfusion. 2009;24(1):51–2. https://doi.
org/10.1177/0267659109106294.
20. Antonucci E, Lamanna I, Fagnoul D, Vincent JL, De Backer D, Silvio TF. The impact of renal
failure and renal replacement therapy on outcome during extracorporeal membrane oxygen-
ation therapy. Artif Organs. 2016;40(8):746–54. https://doi.org/10.1111/aor.12695.
21. Ostermann M, Connor M Jr, Kashani K. Continuous renal replacement therapy during extracor-
poreal membrane oxygenation: why, when and how? Curr Opin Crit Care. 2018;24(6):493–503.
https://doi.org/10.1097/MCC.0000000000000559.
22. Fosse E, Moen O, Johnson E, et al. Reduced complement and granulocyte activation with
heparin-coated cardiopulmonary bypass. Ann Thorac Surg. 1994;58:472–7.
23. Jansen PGM, Velthius H, Huybregts RAJM, et al. Reduced complement activation and
improved postoperative performance after CPB with heparin-coated circuits. J Thorac
Cardiovasc Surg. 1995;110:829–34.
24. Pekna M, Hagman L, Halden E, Nilsson UR, Nilsson B, Thelin S. Complement activa-
tion during cardiopulmonary bypass: effects of immobilized heparin. Ann Thorac Surg.
1994;58:421–4.
25. Gravlee GP. Heparin-coated cardiopulmonary bypass circuits. J Cardiothorac Vasc Anesth.
1994;8(2):213–22.
26. Gu YJ, van Oeveren W, Akkerman C, Boonstra PW, Huyzen RJ, Wildevuur CRH. Heparin-­
coated circuits reduce the inflammatory response to cardiopulmonary bypass. Ann Thorac
Surg. 1993;55:917–22.
27. Ranucci M, Ballotta A, Kandil H, et al. Bivalrudin-based versus conventional heparin anticoag-
ulation for postcardiotomy extracorporeal membrane oxygenation. Crit Care. 2011;15:R275.
2 Circuits, Membranes, and Pumps 79

28. Silvetti S. Do we need heparin coating for extracorporeal membrane oxygenation? New
concepts and controversial positions about coating surfaces of extracorporeal circuits. Artif
Organs. 2015;39:176–9.
29. Cuker A. Clinical and laboratory diagnosis of heparin-induced thrombocytopenia: an inte-
grated approach. Semin Thromb Hemost. 2014;40:106–14.
30. Pollak U, Yacobobich J, Tamary H, Dagan O, Manor-Shulman O. Heparin-induced thrombo-
cytopenia and extracorporeal membrane oxygenation: a case report and review of the litera-
ture. J Extra Corpor Technol. 2011;43(1):5–12.
31. Beiderlinden M, Treschan T, Gorlinger K, Peters J. Argatroban in extracorporeal membrane
oxygenation. Artif Organs. 2007;31(6):461–5.
32. Pappalardo F, Maj G, Scandroglio A, Sampietro F, Zangrillo A, Koster A. Bioline heparin-­
coated ECMO with bivalirudin anticoagulation in a patient with acute heparin-induced throm-
bocytopenia: the immune reaction appeared to continue unabated. Perfusion. 2009;24:135–7.
Chapter 3
Modes of ECMO

Jonathan Eaton, Christopher Trosclair, and L. Keith Scott

Chapter 3 Modes

The ability to understand and manipulate extracorporeal therapies has become piv-
otal in practicing critical care medicine. Extracorporeal life support (ECLS) is a
collective term for therapies supporting a range of causes of cardiac or pulmonary
failure, emphasizing oxygenation, carbon dioxide removal, circulatory support, or a
combination. Extracorporeal membrane oxygenation (ECMO) is the provision of
oxygen and carbon dioxide exchange using an extracorporeal circuit. These terms
explicitly exclude cardiopulmonary bypass for surgical procedures. ECLS has rap-
idly evolved from a rescue therapy to an early interventional therapeutic modality.
Much of this change can be attributed to our familiarity with the technology, its
technological advances, and experience gained during the influenza and COVID-19
pandemics.
This text emphasizes adult ECMO, especially veno-venous ECMO (VV ECMO)
related to support for the acute respiratory distress syndrome (ARDS), since this has
been the major area of growth in the field. Nevertheless, practitioners must recog-
nize when veno-arterial (VA) ECMO is appropriate, as well as the role for

J. Eaton · L. K. Scott (*)

ECLS Program, Critical Care Medicine, Louisiana State University Health Shreveport,
Shreveport, LA, USA
Critical Care Medicine Division, LSU Health Shreveport, Shreveport, LA, USA
e-mail: [email protected]; [email protected]
C. Trosclair
ECLS Program, Critical Care Medicine, Louisiana State University Health Shreveport,
Shreveport, LA, USA
Critical Care Medicine, LSU Health Shreveport, Shreveport, LA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 81

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_3
82 J. Eaton et al.

extracorporeal cardiopulmonary resuscitation (ECPR) and extracorporeal carbon

dioxide removal (ECCO2R), discussed in Chaps. 10 and 16. We will not address
configurations involving central cannulation, such as those requiring median ster-
notomy, instead focusing only on peripheral cannulation via the carotid or femoral
arteries, or the femoral, jugular, or subclavian veins.

Nomenclature

This chapter will incorporate the recommendations of an international consensus to

standardize nomenclature surrounding ECLS [1]. Access is via a systemic artery
(A), systemic vein (V), or pulmonary artery (P), using uppercase letters to signify
primary (major) sites of access. Secondary arterial access used to limit limb isch-
emia is abbreviated as “a.” Cannulation sites include carotid artery (c), femoral
vessels (f), jugular vein (j), and subclavian vessels (s) (see Table 3.1). When a dual-
lumen cannula is used, the prefix “dl” can be added. For example, VV ECMO using
a drainage cannula in the right femoral vein and a return cannula in the right jugular
vein would be signified by “VV ECMO, Vf-Vj configuration.”
Multiple options have been devised both for blood removal and blood return,
each with advantages, disadvantages, and physiological consequences. Most adults
with acute respiratory distress syndrome (ARDS) are managed using VV ECMO,
meaning that blood is withdrawn from large central veins or the right atrium and
returned also to the venous side of the circulation. In VA ECMO, used when circula-
tory support is required, blood is returned instead to the aorta (or other large artery).
Occasionally a hybrid mode is used in which blood is removed from a central vein,
passed through the membrane, and returned both to the central venous and arterial
circulations, termed VVA ECMO. When extracorporeal carbon dioxide removal
(ECCO2R) is the primary goal, the patient’s native blood pressure can be used to
drive flow from artery to vein through a low-resistance membrane (AV ECCO2R).
Alternatively, VV ECCO2R is the provision of carbon dioxide exchange by remov-
ing blood from a large vein, pumping it through a membrane, and returning it to a

Table 3.1 This table shows the correct method to document ECLS cannulation configuration that
should be adopted by all ECMO centers to standardize reporting
Abbreviation
Access (primary, uppercase; secondary, A or a Systemic artery
lowercase)
V Systemic vein
P Pulmonary artery
Cannula entry site c Carotid artery
f Femoral vessel
j Jugular vein
s Subclavian vessel
3 Modes of ECMO 83

Table 3.2 Selecting the mode: veno-venous (VV), veno-arterial (VA), or arteriovenous (AV)
pumpless

Mode Patient criteria

VV ARDS
Severe pneumonia
COPD exacerbation
Status asthmaticus
End-stage chronic lung disease as a bridge to transplant
VA ARDS with acute cor pulmonale
Massive pulmonary embolism
Post-cardiac surgery
eCPR
AV pumpless COPD exacerbation
Status asthmaticus
Ultra-lung-protective ventilation for ARDS
ARDS acute respiratory distress syndrome, COPD chronic obstructive pulmonary disease, eCPR
extracorporeal cardiopulmonary resuscitation

Table 3.3 Venovenous (VV) and venoarterial (VA) ECMO—advantages and disadvantages
Factor VV VA
Cannulas One may suffice Two are needed
Gas exchange efficiency Lower Higher
PaO2 Lower Higher
Recirculation Moderate None
Circulatory support None Full is possible
RV loading Small effect Reduced
LV loading No effect Increased
Systemic pulsatility Normal Reduced
Risk of systemic embolism Low Higher
Circuit pressure Lower Higher
RV right ventricle, LV left ventricle, PaO2 partial pressure of oxygen

large vein. The rest of this chapter describes the circuit design for each mode, their
physiological impact, factors leading to mode selection for individual patients, and
the advantages of each mode (Tables 3.2 and 3.3).

Veno-venous ECMO (VV-ECMO)

The VV Circuit

Veno-venous ECMO is used primarily for gas exchange in patients with isolated,
refractory respiratory failure. Because blood is drained as well as returned into the
venous system, adequate native cardiac function is essential since this form of
ECMO provides no direct circulatory support [2].
84 J. Eaton et al.

Cannulation for VV ECLS can be accomplished either by a two-cannula configu-

ration or a single vessel, dual-lumen catheter. The two-cannula approach is often
selected for emergency cannulation or when fluoroscopic or transesophageal guid-
ance is not available. Typically, cannulas are placed in the inferior (IVC) and supe-
rior (SVC) vena cavae, usually via the femoral and internal jugular veins (Fig. 3.1).
Dual-lumen cannulas have channels for both removing and returning blood and
require only a single insertion site (Fig. 3.2). These are placed so that outflow ports
are located in both the SVC and IVC (traversing the right atrium; RA), while the
inflow port resides within the RA and is directed toward the tricuspid valve. With
both of these circuit configurations, blood flows from the central veins through
external tubing to pump, membrane, warmer, and other circuit components (see
Chap. 2) and then is returned back to the central veins or RA. Since blood is removed
from and returned to the central venous circulation in the VV mode, some blood
returned and fully oxygenated may be immediately withdrawn again in a process
called recirculation. Recirculation creates a partially closed circulatory loop that
decreases the efficiency of oxygenation. The degree of recirculation depends
strongly on cannula location but can be influenced by the intravascular volume
state, circuit blood flow rates, and patient positioning.

Fig. 3.1 Illustration of the

dual-cannula setup for VV
ECLS. Blood is withdrawn
from the venous cannula
and returned into the
superior vena cava.
Configuration is
Vf-­Vj ECMO
3 Modes of ECMO 85

Fig. 3.2 Illustration of the standard ECMO configuration for the single cannula, dual-lumen
setup. Blood is withdrawn through a distal (inferior vena cava) and proximal (superior vena cava)
port and returned through a port positioned near and directed toward the tricuspid valve.
Configuration is (dl) Vj-V ECMO

Single-site, dual-lumen cannulation for VV ECMO has some theoretical ben-

efits, such as less recirculation when compared with dual-site cannulation [3].
Downsides include increased cost, larger cannula diameter, higher risk, and need
for continuous visualization during insertion (Chap. 4). Dual-lumen cannulas
are more expensive to purchase than two single-lumen cannulas, but these costs
may decrease as technology advances and more vendors enter the marketplace.
Dual-lumen cannulas for adult VV ECLS range in size from 23 to 32 French.
Achievable circuit flow depends strongly on cannula diameter, so the largest size
86 J. Eaton et al.

possible is preferred. A general rule of thumb is that the vessel diameter (in mil-
limeters; judged by bedside ultrasound) should be at least one-third the cannula
size (in French). This can be estimated prior to cannulation by bedside vascular
ultrasound, which allows an additional opportunity to verify that there is no
major thrombus or stenosis present that would complicate cannulation at that
site. Cannulation is more technically challenging and riskier than for the two-
site approach. Placement of the guidewire into the IVC should be verified and
the cannula visualized continuously (transthoracic or transesophageal echocar-
diography or fluoroscopy) during insertion to avoid catastrophic vascular or car-
diac injuries, requiring additional personnel, space, and equipment. See also
Chap. 4.

Physiological Implications of VV ECMO

The physiology of VV ECMO is fully described in Chap. 1, but several points

deserve emphasis here. First, since oxygenated blood returns to the venous circula-
tion, it mixes with venous blood that was not captured by the outflow cannula. The
oxygen saturation (SO2) of the resulting mixture is often in the mid-80s, even though
the oxygen partial pressure (PO2) of post-membrane blood may be 400 mmHg.
Since this mixture passes through failed lungs, the consequence is often a level of
systemic PO2 that is lower than many clinicians are comfortable with and, as dis-
cussed below, lower than is typically achieved during VA ECMO. However, as long
as total cardiac output is adequate, saturations in the 80s (and even high 70s) are
usually well tolerated. Most importantly, SaO2 is raised by increasing the fraction of
native cardiac output captured by the outflow cannulas and directed to the mem-
brane. Thus full support of oxygenation typically requires circuit blood flow rates of
3–5 L/min, values that sometimes produce hemolysis, recirculation, or circuit “chat-
tering” (when the great veins collapse around the cannula). If the goal of ECLS is
largely carbon dioxide removal (ECCO2R; see Chap. 10), much lower flow rates are
possible, and smaller cannulas can be used. ECCO2R using a VV circuit may be
effective for patients with conditions largely characterized by hypercapnia, such as
chronic obstructive pulmonary disease (COPD) exacerbations of status asthmaticus,
or to support an ultra-lung-protective mechanical ventilation strategy for patients
with severe ARDS.
A second physiological consequence of VV ECMO is that it has virtually no
direct circulatory impact, a result of the fact that equal flows of blood are simultane-
ously withdrawn and returned. Because of this, right and left heart function, as well
as the pulmonary circulation, must be reasonably intact for VV ECMO to be an
appropriate choice. Thus this mode is used primarily for patients with isolated,
refractory respiratory failure. Roughly a quarter of patients with severe ARDS have
right ventricular (RV) dysfunction related to the pulmonary vascular effects of lung
injury, hypoxia, thrombosis, and mechanical ventilation. When cor pulmonale is
severe, VA ECMO may be more appropriate. On the other hand, the VV mode has
3 Modes of ECMO 87

some second-order effects that could unload the right heart. For example, to the
extent that the membrane supplants the need for mechanical ventilation, lower tidal
volumes and PEEP are usually used, and these changed ventilator settings could
unload the RV. Similarly, once the pulmonary circulation is perfused with blood
having a higher PO2 and lower PCO2, vascular resistance falls acutely. In the CESAR
Trial of ECMO for severe ARDS, all patients were managed with the VV mode, and
no patient had to transition to a VA circuit during the trial. In contrast to the VV
mode, VA ECMO can fully support the circulation, so is more appropriate for
patients with massive pulmonary thromboembolism, ARDS complicated by severe
cor pulmonale, extra-corporeal cardiopulmonary resuscitation (eCPR), or following
cardiac surgery (Table 3.2). In hemodynamically unstable patients for whom ECLS
is being contemplated, it is important to delineate the basis for circulatory failure
(using echocardiographic imaging, for example) as this may influence the choice of
VV or VA ECMO.

Advantages of VV ECMO

Several advantages are conferred because the VV mode returns blood to the central
veins, rather than the arteries (Table 3.3). These include a reduced risk of systemic
embolism, including catastrophic cerebral or coronary embolism; no injury to sys-
temic arteries that could cause hemorrhage or distal ischemia; lower circuit pres-
sures with consequently lower chance of catastrophic circuit failure; and the
potential for a single cannula to suffice, possibly facilitating mobilization (see Chap.
14). VV ECMO also maintains the pulsatility of the systemic circulation, which is
lost to some degree during VA support, as described below. Finally, decannulation
can be performed at the bedside with no need for surgical repair or ligation of
vessels.

Veno-arterial ECMO

Veno-arterial (VA) ECMO can provide both cardiac and respiratory support [4]. For
adults with respiratory failure but preserved cardiac function, VV-ECMO is prefer-
able since it avoids the risks associated with large bore-arterial cannulation.

The VA Circuit

With VA ECLS, blood is withdrawn from the venous circulation either by direct
surgical cannulation of the RA or by a cannula placed in a vein with the tip posi-
tioned in the RA, SVC, or IVC. Blood is returned through the carotid artery in
88 J. Eaton et al.

Fig. 3.3 Illustration of VA cannulation for neonates and younger children. The outflow cannula is
inserted through the internal jugular vein with return through the common carotid artery.
Configuration is Vj-Ac ECMO

neonates and infants (Fig. 3.3) and the descending aorta (via the femoral artery)
in older children and adults (Fig. 3.4). Where to place the return cannula depends
partly on the blood flow needed. If a need for full support is anticipated (e.g.,
100–125 cc/kg/min in a neonate), the excessive resistance of peripheral arteries
may not allow sufficient flow. Since blood flow is largely dependent on cannula
diameter, the largest possible cannula is usually inserted. Vessel size is judged by
ultrasound to select a cannula of adequate size to provide flow without obstructing
distal arterial flow beyond the cannula. If the arterial cannula compromises or
obstructs nutrient blood flow, as is occasionally the case when a femoral cannula
threatens perfusion to the leg, a small catheter can be aimed distally to provide
additional blood flow (Fig. 3.5). Central cannulation of the right atrium and the
ascending aorta allows larger cannulas to be used, providing higher flows and
3 Modes of ECMO 89

Fig. 3.4 Illustration of VA

cannulation for older
children and adults. The
outflow cannula is inserted
through the right common
femoral vein, with return
through the left common
femoral artery.
Configuration is
Vf-Af ECMO

reliable coronary and cerebral perfusion, but at a cost of problematic bleeding

from the surgical wound and sternum, and the risk and expense of two surgeries.

Physiological Implications of VA ECMO

The gas exchange effects of VA ECMO are similar to those of VV ECMO, but there
are important differences. First, PaO2 values are typically higher on the VA mode
because most of the blood has passed through the ECMO circuit, with less provided
by the (failing) native circulation. In contrast, with the VV configuration much
venous (desaturated) blood is not captured by the drainage cannula(s) and passes
90 J. Eaton et al.

Fig. 3.5 Illustration of a hybrid system that started as a VA (Vf-Af) configuration and was con-
verted to veno-venoarterial (VVA) because of upper body hypoxemia. This configuration (Vf-Vj-
Af) ECMO also arises following initial VV support when circulatory support becomes necessary,
requiring the addition of a veno-arterial component

through the diseased lungs, producing venous admixture and causing lower PaO2
values. Second, cannula position in VA ECMO precludes the possibility of recircu-
lation, so gas exchange is more efficient. Finally, there is often a gradient of arterial
PaO2 related to the competition from fully oxygenated blood moving retrograde
from the descending aorta and de-oxygenated blood coming from the native cardiac
output and moving antegrade through the aorta.
The higher the native cardiac output and the more diseased the lungs, the lower
will be the proximal aortic oxygen partial pressure and the more distally this effect
will be seen. Commonly, the right arm PaO2 (and similarly the pulse oxygen satura-
tion) is lower than that in the left arm or in the legs. Since the coronary and carotid
arteries are supplied from the proximal aorta, this peculiarity of varying arterial
oxygen values may further threaten the circulation or neurological status.
While the gas exchange differences between VV and VA ECMO are modest,
circulatory physiology is completely different. VV ECMO has essentially no
3 Modes of ECMO 91

circulatory effect, yet the basic function of VA ECMO is to provide circulatory sup-
port, returning oxygenated blood to the arterial circulation at physiological perfu-
sion pressures. Since blood is drawn from the vena cava or RA, VA ECMO unloads
the right ventricle, and this mode has been used for patients with massive pulmo-
nary embolism, for example. However, the left ventricle does not benefit similarly.
Even though the LV receives less blood from the lungs, a failing LV may only eject
little stroke volume, especially once extracorporeal support raises blood pressure to
normal levels. Related to the low native cardiac output, aortic root flow may be slug-
gish with areas of stasis. This has the potential to promote thrombosis, raising the
risk of stroke or other manifestation of systemic embolization. Moreover, blood
from Thebesian veins and other sources drains into the LV, so this LV distension can
lead to complications such as pulmonary edema and RV overload. Because of this,
efforts are necessary to augment LV systolic function, monitor LA and LV size, and
possibly decompress the LV. Such measures should be included in the VA ECLS
care pathway. One method to reduce LV distention is to vent the LA. With central
cannulation, the technique is surgically straightforward. However, venting the LA is
more challenging when patients are cannulated peripherally. Options include trans-
cutaneous septoplasty, insertion of a vent cannula through a mini thoracotomy, or
placement of an intra-aortic balloon pump (IABP). Concerns with the use of an
IABP is that it could obstruct ECLS flow returning through the descending aorta or
flow from the femoral arterial cannula could compete with the inflated state of the
IABP [3]. However, several studies have shown a benefit of the IABP in off-loading
the LV and possibly enhancing cerebral and coronary blood flow.
Another difference between VV and VA ECMO relates to systemic arterial pul-
satility. Since pump flow is non-pulsatile, the systemic blood pressure tends also to
be lacking in variability, although there is often some component related to the
native circulation. Even when mean blood pressures are in the normal range, the
systolic pressure on VA ECMO tends to be significantly lower (and diastolic higher)
compared to normal physiology. Concerns have been expressed that non-pulsatile
systemic blood pressures could lead to renal dysfunction, but this has not borne out
in practice. Mean blood pressure relates to total circuit flow, native cardiac output,
and systemic vascular resistance. The adequacy of perfusion can be judged by
assessing mean blood pressure, central venous oxyhemoglobin saturation, lactic
acid levels, and end-organ function, but metrics related to pulse contour analysis
(including many modern minimally invasive cardiac output technologies) will not
be valid. Still, the return of a pulsatile waveform on the peripheral arterial trans-
ducer can be useful as a gauge of cardiac recovery.

Advantages of VA ECMO

In comparison with VV ECMO, the VA mode leads to more efficient gas exchange
and no possibility of recirculation. Systemic PO2 is typically much higher than in
VV configuration, although this is sometimes not true of the cerebral, coronary, and
92 J. Eaton et al.

right upper extremity circulations as described above. Most importantly, the hemo-
dynamic effects discussed previously allow full support of the circulation, including
unloading of the RV, whereas VV does not. Thus, “eCPR” is feasible even in patients
suffering full cardiac arrest.

Hybrid ECMO Configurations

The traditional ECMO configurations each have limitations. For example,

VA-ECMO, particularly with peripheral cannulation, may not provide adequate car-
diac and cerebral oxygenation. Likewise, a patient on VV-ECMO who suffers
declining heart function may require some form of hemodynamic support. That is
where hybrid systems come in. Over the past few years, many different configura-
tions have been reported. We will focus on the systems that seem to have the best
data and are illustrated in Fig. 3.5.

VVA-ECMO

In this configuration, a venous outflow cannula leads to the pump and artificial
membrane, after which blood is returned both to the right atrium (or vena cava) and
to the systemic arteries (Fig. 3.5). Often this arises when a patient is started on VA
ECMO but has persistent upper extremity, cardiac, or cerebral hypoxia. This usually
evolves in the face of worsening lung function where the blood entering the LA
becomes progressively hypoxemic. Placement of a second venous cannula (inflow,
connected by a “Y” to the arterial inflow) in the IVC, SVC, or RA allows a stream
of fully oxygenated blood to traverse the lungs, raising the saturation of blood pass-
ing through the left heart and into the proximal aorta. A downside lies in the fact that
flow is diverted from the arterial system, possibly reducing oxygen delivery.
Increasing the circuit flow rate may compensate for this but may be limited by can-
nula size or the adequacy of venous drainage. Furthermore, after a venous inflow
cannula is spliced into the arterial limb of the circuit, much of the flow may prefer-
entially go toward the venous system, being of lower pressure. This may negatively
impact mean arterial pressure support but may be remedied by applying a partially
occlusive clamp to the venous branch of the inflow limb of the circuit (Fig. 3.6).
Alternatively, this hybrid configuration may be reached when a patient on VV
ECMO develops cardiac dysfunction, for example, due to progressive cor pulmo-
nale. Now the addition of an arterial inflow cannula unloads the RV and directly
supports the circulation. This configuration is often accompanied by significant ino-
tropic support of the circulation.
Other hybrid systems have been described, such as combining VV ECMO with
IABP support or performing an atrial septoplasty to unload the RV in a patient with
severe veno-occlusive disease. There are few data on these hybrid modes and their
impact on ECMO duration, complications, or survival. In one small series, ARDS
3 Modes of ECMO 93

Fig. 3.6 VVA cannulation

with partially occlusive
line clamp on venous
inflow

patients started on or transitioned to VVA ECMO appeared to have a higher sur-

vival. However, the rationale for transitioning from VV to VVA was not well defined
and appeared to be related to mechanical issues (poor venous drainage), not to an
augmented cardiac output.

Extracorporeal CO2 Removal Modes

Extracorporeal CO2 removal (ECCO2R) as a primary goal can be performed using

three main modalities: (1) veno-venous cannulation (VV ECCO2R) using low flows
and a standard ECMO circuit, modified to emphasize CO2 removal or employing a
94 J. Eaton et al.

membrane specialized for CO2 removal; (2) arterio-venous pumpless system (AV
ECCO2R); and (3) adapting a continuous renal replacement therapy (CRRT) circuit
to integrate a CO2 removal membrane into the circuit. We will discuss the limits and
merits of each system in more detail.

VV ECCO2R

Gattinoni and colleagues were one of the first groups to describe effective carbon
dioxide removal using an extracorporeal device [5]. This approach uses a pump to
drain blood from the venous system, pass it through a gas exchanger, and return it
to the venous system. Since CO2 diffuses through the membrane more readily than
oxygen, much lower circuit blood flow is required, meaning smaller cannula size,
narrower tubing, and lower pump speeds. In theory, this should reduce hemolysis
and reduce complications of cannula insertion, while providing sufficient support to
reduce ventilator settings [6]. The same goals can be achieved through much larger
ECMO circuits, but without some of the potential advantages.
VV ECCO2R can be used for patients with predominantly obstructive physiol-
ogy, as in COPD exacerbations or status asthmaticus (to prevent intubation or to
liberate more rapidly), as discussed in Chap. 10. Recently, VV ECCO2R has been
used to facilitate “ultra-lung-protective” ventilation for patients with severe ARDS,
allowing tidal volumes of only 2–4 mL/kg.
VV ECCO2R has been used successfully in dealing with COVID-19-related
hypercapnic respiratory failure. In one case series, 29 patients with COVID-19 were
treated with an extracorporeal CO2 removal device with adequate control of CO2
and pH levels [7]. Complications were mostly cannulation issues, and there was no
significant bleeding, hemolysis, or device failure. Others have reported similar
experience [8]. One such VV-ECCO2R device that is currently undergoing FDA
review is shown in Fig. 3.7.

Fig. 3.7 Illustration of AV

ECCO2R with an
Af-Vf configuration. This
simple pumpless system
connects arterial and
venous cannulas to a
hollow fiber oxygenator
with oxygen serving as the
sweep gas
3 Modes of ECMO 95

AV ECCO2R

AV-ECCO2R is a pumpless arteriovenous carbon dioxide removal device.

Extracorporeal carbon dioxide is removed utilizing a low-resistance hollow fiber
oxygenator and a simplified extracorporeal circuit, driven by the patient’s own
blood pressure. This method of extracorporeal support has been demonstrated to be
both safe and effective in adult patients with acute lung injury exhibiting hypercap-
nia and respiratory acidosis [9].
Mathematical simulation suggested that it is possible to achieve total CO2
removal using an extracorporeal shunt fraction of only 10–15% of cardiac output
and a sweep gas to blood flow ratio of 5 or greater [10]. One clinical trial demon-
strated a mean CO2 removal of 112 mL/min with blood flow rates above 500 mL/
min [11]. Blood flow through the circuit is determined by the size of the arterial
cannula (the highest resistance component of the circuit) and the pressure gradient
between the arterial and venous systems.
The circuit can be very simple. A 12-14F arterial catheter is inserted using the
Seldinger technique. A venous cannula is subsequently inserted using a large F
catheter to assure low resistance and to maximize blood flow through the oxygen-
ator. Oxygen is used as the sweep gas: varying its flow changes the amount of CO2
removed. This system has shown efficacy in life-threatening asthma, COPD exacer-
bation, and ARDS [12].

CVVRRT

Patients on renal replacement therapy who develop respiratory failure could be can-
didates for CVVRRT. This approach inserts a gas exchange membrane in the return
limb of the RRT circuit. By and large these systems can process a blood flow up to
450 mL/min, but most are limited to 200–300 mL/min due to cannula and circuit
pressure limitations [13]. The use of this mode is in its infancy and has numerous
limitations. Further investigation is needed before this is adopted widely.

References

1. Conrad SA, Broman LM, Taccone FS, Lorusso R, Malfertheiner MV, Pappalardo F, et al. The
extracorporeal life support organization maastricht treaty for nomenclature in extracorporeal
life support. a position paper of the extracorporeal life support organization. Am J Respir Crit
Care Med. 2018;198(4):447–51 https://doi.org/10.1164/rccm.201710-2130CP.
2. Brodie D, Bacchetta M. Extracorporeal membrane oxygenation for ARDS in adults. N Engl J
Med. 2011;365(20):1905–14.
4. Schroeter T, Vollroth M, Höbartner M, Dhein S, Sahlisch S, Borger MA, Mohr
FW. Extracorporeal Membrane Oxygenation and Intra-Aortic Ballon Pump – an appropriate
combination or useless battle of materials? Thorac Cardiovasc Surg. 2013;61(S 01):OP19.
96 J. Eaton et al.

4. Nguyen DQ, Kulick DM, Bolman RM III, Dunitz JM, Hertz MI, Park SJ. Temporary
ECMO support following lung and heart-lung transplantation. J Heart Lung Transplant.
2000;19(3):313–6.
5. Low-Frequency Positive Pressure Ventilation with Extracorpor...: Anesthesia & Analgesia.
Accessed May 20, 2014. http://journals.lww.com/anesthesia-­analgesia/Fulltext/1978/07000/
Low_Frequency_Positive_Pressure_Ventilation_with.18.aspx.
6. Fanelli V, Costamagna A, Terragni PP, Ranieri VM. Low-flow ECMO and CO2 removal.
In: Sangalli F, Patroniti N, Pesenti A, editors. ECMO-extracorporeal life support in adults.
Springer Milan; 2014. p. 303–15.
7. Akkanti B, Jagpal S, Darwish R, Romero RS, Scott LK, Dinh K, Hussain S, Radbel J, Saad
MA, Enfield KB, Conrad SA. Physiologic improvement in respiratory acidosis using extra-
corporeal Co2 removal with hemolung respiratory assist system in the management of severe
respiratory failure from coronavirus disease 2019. Crit Care Explor. 2021;3(3):e0372.
8. Boparai S, Keith Scott L, Conrad S, Motayar N. Extra-corporeal carbon dioxide removal in
COVID-19 ARDS. Chest. 2020;158(4):A1029.
9. Brunston RL Jr, Tao W, Bidani A, Alpard SK, Traber DL, Zwischenberger JB. Prolonged
hemodynamic stability during arteriovenous carbon dioxide removal for severe respiratory
failure. J Thorac Cardiovasc Surg. 1997;114(6):1107–14.
10. Arteriovenous Extracorporeal Carbon Dioxide Removal A Mathem... : ASAIO
J. Accessed May 16, 2014. http://journals.lww.com/asaiojournal/Fulltext/1998/07000/
Arteriovenous_Extracorporeal_Carbon_Dioxide.7.aspx.
11. Conrad SA, et al. Total extracorporeal arteriovenous carbon dioxide removal in acute respira-
tory failure: a phase I clinical study. I. Intensive Care Med. 2001;27:1340–51.
12. Mauri T, Zanella A, Pesenti A. Extracorporeal gas exchange: present and future. In: Jean-Louis
VP, editor. Annual update in intensive care and emergency medicine 2013. Berlin Heidelberg:
Springer; 2013. p. 609–19.
13. Grant AA, Hart VJ, Lineen EB, et al. Rescue therapy for hypercapnia due to high
PEEP mechanical ventilation in patients with ARDS and renal failure. Artif Organs.
2019;43:599–604.
Chapter 4
Vascular Access

Steven A. Conrad

Introduction

Extracorporeal membrane oxygenation (ECMO) requires high extracorporeal flow

rates on the order of that of cardiac output. Access to the central circulation to pro-
vide and maintain blood flow necessary for adequate gas exchange is one of the
most essential aspects of successful extracorporeal support. Inadequate extracorpo-
real flow can lead to failure to deliver sufficient support and limit any potential
benefit of ECMO. Cannulas and cannula insertion techniques are quite variable, and
the choice of each depends on the goals and mode of support, cannulation configu-
ration, size of the patient, size of the vessels, as well as institutional and logistical
concerns.
An understanding of cannula design, hemodynamics, and techniques of cannula
insertion is the foundation of safe and effective application of extracorporeal sup-
port. The nomenclature and abbreviations in this chapter conform to the Maastricht
Treaty for ECLS Nomenclature developed by the Extracorporeal Life Support
Organization [1, 2].

S. A. Conrad (*)
Departments of Medicine, Emergency Medicine, Pediatrics and Surgery, Louisiana State
University Health Sciences Center, Shreveport, LA, USA
Ochsner LSU Health Academic Medical Center, Shreveport, LA, USA
LSU Health Shreveport, Shreveport, LA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 97

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_4
98 S. A. Conrad

Cannulas for Extracorporeal Support

Historically adopted from short-term cannulas used in cardiopulmonary bypass,

contemporary cannulas are now specifically designed and marketed for long-term
support with ECMO. A variety of cannulas for peripheral vascular access are com-
mercially available. These cannulas differ with respect to the technique of insertion
(percutaneous or surgical), blood flow direction (drainage or reinfusion), and wall
reinforcement, as well as being available in various lengths and diameters to accom-
modate the choice of vessel used for access. Cannulas designed for percutaneous
peripheral insertion have some minor feature differences from those intended for
surgical placement. The loading dilator that accompanies a percutaneous cannula
has a long taper and a central lumen to accommodate a guidewire, whereas the sur-
gical cannula has a blunt dilator with a short tip and no central lumen. The tip of a
percutaneous cannula is designed to fit snugly against the loading dilator and is
tapered to facilitate insertion through tissue, whereas this feature is optional in sur-
gical cannulas.
Wire-reinforced cannulas contain a layer of spiral-wound metal wire imbedded
in the wall of the cannula. This reinforcement allows the cannula to flex without
kinking, resist flattening from external compression, and prevent collapse when
negative pressures are applied to the lumen. Since these complications can result in
loss of extracorporeal support, reinforced cannulas are the preferred design.

Single-Lumen Design

Cannulas with a single lumen are required for venoarterial and arteriovenous vascu-
lar access and are an optional approach for venovenous access. Two fundamental
designs are manufactured, intended for either drainage or return (often referred to as
venous and arterial cannulas, respectively). The drainage cannula design is charac-
terized by a greater length (up to approximately 50 cm), greater available diameter
(up to 29 Fr), and a longer distal segment with multiple side holes to facilitate drain-
age. The greater length allows insertion into more central veins such as the superior
vena cava (SVC) or inferior vena cava (IVC) (Fig. 4.1a). Drainage cannulas are
designed with different placement and configurations of side holes. The return can-
nula design is characterized by a shorter length and a shorter distal segment with a
limited number of side holes, since deeper insertion is not required, and flow is not
dependent on side holes as in the drainage design (Fig. 4.1b). An excessive number
of side holes can increase the risk of hemolysis in return cannulas.
Recently introduced expandable, wire-reinforced cannulas that incorporate a dis-
tal segment of wall free wire mesh (Smartcannula LLC, Switzerland) are available
in some markets. These cannulas expand to a larger diameter within the vessel to
minimize flow resistance, and the distal mesh maintains vessel patency for improved
drainage [3].
4 Vascular Access 99

Fig. 4.1 Single-lumen

cannulas for percutaneous
a b
cannulation. (a) Drainage
cannulas exhibiting greater
length and multiple side
holes distributed along
much of the distal segment.
(b) Return cannulas
exhibiting shorter length
and limited side holes
adjacent to the tip of the
cannula. (Reprinted with
permission from Gettinge)

Dual-Lumen Design

Cannulas incorporating two lumens are a more recent design that have facilitated
the application of venovenous support for respiratory failure. Although designed for
percutaneous insertion, they can be placed surgically as well. Two fundamental
designs are available that have features to support different needs.
The bicaval design requires insertion via the internal jugular vein with the cannula
traversing the right atrium and the tip positioned in the IVC (Avalon Elite, Maquet
Cardiopulmonary, Rastatt, Germany and Crescent, Medtronic, Bloomington, MN,
USA) [4, 5] (Fig. 4.2). The drainage lumen extends the length of the cannula with
two drainage ports, one in each the IVC and SVC. The return lumen is shorter, ter-
minating in the right atrium with the return port directed toward the tricuspid valve.
This bicaval design effectively separates upper and lower body venous drainage and
results in lower recirculation with more effective systemic oxygen delivery.
Recirculation fractions on the order of 2–3% are typical, unlike 20–30% recircula-
tion fractions characteristic of two-site single-lumen cannulation [4]. These cannulas
are available in sizes suitable for venovenous support ranging from neonates to adults.
The second design is similar to a hemodialysis catheter with a single proximal
drainage port and a distal reinfusion port. Two variations of this design exist. A
longer and larger diameter version (Protek Duo, LivaNova PLC) is intended for
insertion via the right internal jugular vein into the main pulmonary artery, with the
proximal drainage port in the right atrium and the return port in the pulmonary
artery. It is intended for high-flow venovenous support, with the added advantage
that it can support right heart function. Recirculation is minimal to absent since the
pulmonary valve prevents oxygenated blood from reaching the drainage port. A
shorter and smaller diameter version that can be inserted into the internal jugular or
femoral vein (as for hemodialysis) is available in 15.5 Fr diameter and intended for
low flow (500–800 mL/min) extracorporeal circuits used for extracorporeal carbon
dioxide removal (ECCO2R) [6, 7] (see Chap. 10). The cannula flow must not exceed
100 S. A. Conrad

Fig. 4.2 General design of

a dual-lumen bicaval
venous cannula, designed
for combined drainage
from the IVC and SVC
with reinfusion of the
blood into the right atrium
for venovenous
extracorporeal support

the insertion vessel flow, or recirculation will limit effective blood flow. Placement
in the SVC via an internal jugular vein or in an iliac vein via the femoral vein usu-
ally assures adequate vessel blood flow.

Determinants of Cannula Blood Flow

Blood flow through vascular cannulas is driven by the difference between the pres-
sure at the hub of the cannula and the intravascular pressure at the tip of the cannula.
Although an ECMO cannula is cylindrical in shape, in which the relationship
between flow and pressure is expected to be linear in the presence of laminar flow,
the relationship is only linear at the low end of the flow range where laminar flow
predominates. The laminar flow relationship between flow ( Q ) and pressure gradi-
ent (ΔP) can be described by the Hagen-Poiseuille equation:

∆P ⋅ r 4
Q =
µ⋅L

where r is the radius, L is the length, and μ is the blood viscosity. Maximizing can-
nula blood flow involves the insertion of the largest diameter cannula that can be
safely inserted and keeping the length as short as possible.
4 Vascular Access 101

Fig. 4.3 Representative pressure-flow relationships for various size single-lumen return cannulas.
The graph depicts the nonlinear relationship between flow and pressure due to a combination of
laminar, transitional, and turbulent flow resulting from the complex geometry of the cannula

The actual pressure-flow relationship over the full range of blood flow is nonlin-
ear, most likely due to the development of transitional and turbulent flow in the
lumen and at the side holes as flow increases. Turbulence increases the pressure
gradient required to maintain a given level of flow. This nonlinear relationship is
demonstrated in a typical pressure-flow diagram shown in Fig. 4.3. Note that manu-
facturers provide pressure-flow data for their cannulas using water as the fluid, so
that the actual pressures required to generate flow will be higher with blood.

Patient Preparation

Determination of Vessel Size

During surgical cannulation the vessel is exposed, and cannula size selection can be
made visually at the time of cannulation. Determination of cannula diameter prior
to percutaneous cannulation, however, requires imaging to minimize complications.
Without vessel sizing the use of too large a cannula can result in venous obstruction,
102 S. A. Conrad

Fig. 4.4 Ultrasound image

of femoral vessels. The
double arrow indicates the
diameter of the common
femoral artery (CFA), the
ellipse the circumference
of the common femoral
vein (CFV) at the insertion
of the saphenous vein
(SV). Vessel size is
calculated as the diameter
(mm) × 3, or the
circumference in mm

failure to cannulate, or other complications such as vessel laceration or transection.

Too small a cannula can result in suboptimal blood flow and ineffective support.
Bedside ultrasound with a vascular transducer can provide high-quality images
of the cervical and femoral vessels. Vessel size can be obtained by using the built-in
measurement tools (Fig. 4.4) and converting to the French gauge system as described
by Joseph-Frédéric-Benoît Charrière [8] used for sizing cannulas. In the case of ves-
sels with a circular shape, conversion of vessel diameter in millimeters to French
size is accomplished with the following simple formula:

Fr = D( mm ) ⋅ 3

For noncircular vessels the French size is approximately equal to the circumfer-
ence in millimeters. Both diameter and circumference measurements are easily per-
formed with bedside ultrasound machines. The chosen cannula should be 10–20%
smaller than the measured vessel to help assure successful placement and prevent
complete obstruction of blood flow.

Infection Control

Infection is not an uncommon risk during prolonged extracorporeal support [9].

Since extracorporeal support may be required for periods of weeks, steps to prevent
infection are warranted, and strict attention to skin asepsis is mandatory during can-
nulation. Full surgical skin preparation can be accomplished with both aqueous- and
alcohol-based chlorhexidine solutions and should be applied according to the manu-
facturer’s recommendations. For example, with aqueous-based 4% chlorhexidine, a
2-min scrub allowing the skin to dry then repeating the scrub is the recommended
technique.
4 Vascular Access 103

Peri-procedural prophylaxis with intravenous antibiotics can be considered for

patients who are not receiving antibiotics and with choice of antibiotic and schedule
provided according to the institution’s guidelines. Continuation of prophylactic
antibiotics for the duration of ECLS support, other than required for treatment of
underlying infection, is not recommended [10]. Following insertion, strict aseptic
technique for prevention of cannula-associated infection is mandatory. Since
patients on ECLS may be fully dependent on support for weeks, simple redressing
and observation for development of infection should be replaced with an active
approach, such as repeating a 2-min surgical scrub of the site with aqueous 4%
chlorhexidine every 24–48 h.

Insertion Technique

Three techniques for cannula insertion are commonplace. Historically, all cannula-
tions were performed by surgeons using an open surgical technique. While some
vessels still require an open approach, surgical cannulation in most cases has been
replaced with percutaneous cannulation and performed by surgeons, intensivists,
interventionalists, and emergency physicians.

Percutaneous

Percutaneous cannulation has been used successfully for both venovenous and
peripheral venoarterial support [11]. It is the preferred technique since it is associ-
ated with a lower incidence of cannulation site bleeding and infection [12, 13]. It
can also be nonobstructive, allowing blood flow around the cannula after placement,
avoiding ischemic and congestive complications. It can be used for any peripheral
arterial access as well as venous access, although experience with carotid cannula-
tion is limited.
The same Seldinger technique used for smaller vascular access catheters is used
for ECLS cannulation, but with multiple dilators of no more than 4 Fr increments in
size (typically 8, 12, 16, 20, 24, 28, and 30–32 Fr), with the largest size approxi-
mately equal to the size of the cannula to be inserted. Vessel size is determined prior
to insertion with ultrasound, and an appropriately sized cannula is chosen. If the
cannula is being placed into the thoracic cavity, adequate sedation and a neuromus-
cular blocker are administered to prevent inspiratory effort and air embolism. Under
aseptic conditions and following infiltration of a local anesthetic, the vessel is iden-
tified with ultrasound, and an approach is chosen to avoid injury to neighboring
vessels, since vessels may overlie each other. The access needle is inserted using
ultrasound to guide it through the center of the vessel, and a 0.35″ or 0.38″ guide-
wire is advanced. Fluoroscopy is invaluable for preventing guidewire misadventures
during advancement and recommended for the bicaval dual-lumen cannula to assure
placement of the wire directly across the atrium.
104 S. A. Conrad

Following placement of the guidewire, a skin incision is made just large

enough to admit the cannula such that the skin is taut when the cannula is
placed. Larger incisions will more likely bleed when anticoagulation is initi-
ated. The tract is then dilated sequentially to the cannula size or 1–2 Fr smaller.
The cannula is placed over its tapered loading dilator and advanced into posi-
tion over the guidewire. Using a tubing clamp to control back-bleeding, the
guidewire and dilator are removed, and the cannula is flushed with heparinized
saline (2 units/mL) to maintain patency until attached to the ECLS circuit, and
the extracorporeal circulation has begun. The cannula is secured with sutures to
prevent decannulation, taking care to avoid crimping the cannula or providing a
pivot point for cannula kinking.
Entry into the femoral vein should be made above the insertion of the saphenous
vein, since the diameter of the vessel is larger at this level. For femoral arterial can-
nulation, the common femoral (above the takeoff of the deep and superficial femoral
arteries) is targeted.
Percutaneous cannulation of the common femoral artery may result in inade-
quate distal perfusion and subsequent development of lower limb ischemia. This
can be mitigated by percutaneous placement of an antegrade arterial cannula (6–8
Fr) into the superficial femoral artery, or surgical cut down with retrograde cathe-
terization of the posterior tibial artery to assure adequate perfusion of the limb.
This cannula is connected to the return limb of the circuit through a Luer lock or
¼ inch connector as required. If an antegrade cannula is not placed at the time of
primary cannulation, close monitoring for signs of limb ischemia is mandatory. An
elevated difference in near-infrared spectroscopy between ipsilateral and contralat-
eral calf muscles [14] and a blood pressure in the dorsalis pedis artery under
50 mmHg obtained with Doppler ultrasound are indications for secondary cannu-
lation [15].

Semi-Open

A variation of percutaneous cannulation preferred by some is a technique which

combines percutaneous skin and vessel insertion under direct visualization through
an incision over the vessel entry point. Following sedation and neuromuscular
blockade, skin preparation, and anesthetic infiltration, an incision is made over the
expected vessel entry point, and dissection is carried out visually to expose the
superficial wall of the vessel. A needle puncture is made through the skin distally to
the incision, and a tract is created with a guidewire and dilator as for percutaneous
insertion. The vessel is then punctured and the guidewire advanced. Entry of the
cannula over the guidewire into the vessel is performed under direct visual inspec-
tion. The subcutaneous tissue and skin are then closed. Vessel ligation and incision
are avoided, and the skin can be closed without the cannula exiting through the inci-
sion, reducing bleeding and infectious complications.
4 Vascular Access 105

Open Surgical

The preferred technique for cervical cannulation when carotid or subclavian arterial
access is required, or other peripheral artery following failure of percutaneous can-
nulation, is the open surgical technique. Following sedation, neuromuscular block-
ade, and skin preparation, an incision is made perpendicular to the axis of the vessel
with dissection carried down to expose the vessels. The cannula can be sized by
visual comparison with the vessel diameter. The vessels are freed from surrounding
tissue, and ligatures are placed proximal and distal to control bleeding. An arteriot-
omy (and venotomy) is made, and the cannula with its blunt-tipped loading dilator
is inserted into the vessel while loosening the proximal ligature to admit the can-
nula. Following insertion to the proper depth, the ligatures are secured, typically
with pledgets to prevent vessel injury, as vessel repair may be performed after
decannulation. The subcutaneous tissue and skin are closed, taking care to securely
close the skin around the cannula. The above description is generic, and variations
are numerous, subject to the surgeon’s preferences and experience.
If open cannulation is performed on the common femoral artery, which has no
collateral circulation, then a smaller antegrade cannula is placed distal to the pri-
mary cannula in the superficial femoral artery to maintain distal perfusion. A 6–8 Fr
cannula is sufficient and attached to the circuit as for percutaneous cannulation.
An alternative approach to insertion of the cannula into the artery is to suture a
vascular graft to the artery (end to side), with placement of the cannula into the
graft. The artery then remains fully open, and distal ischemia is avoided, and a
larger cannula can be used. This approach is often chosen when long-duration sup-
port is anticipated, such as bridge to transplant.

Cannulation Configuration

The foremost decision regarding vascular access is the mode of support (respiratory,
cardiac, or hybrid) which dictates the configuration of cannula placement.
Extracorporeal life support for both respiratory and cardiac failure was historically
performed using a venoarterial (VA) configuration. While still preferred for cardiac
failure, other configurations have been developed that are more suitable for other
types of support. As these configurations are presented in Chap. 2, the following
will focus on cannula sizes and vascular access associated with each configuration.

Venoarterial (VA) Configuration

The venoarterial configuration drains blood from the central venous circulation and
returns it to the arterial circulation. The cervical approach is used in neonates,
infants, and small children, since the femoral vessels are small prior to the age of
106 S. A. Conrad

walking. Cervical vessels have collateral circulation from the contralateral side,
allowing ligation of the carotid artery and internal jugular vein. The arterial return
cannula, typically 8–10 Fr, is placed into the common carotid artery and advanced
into the proximal innominate artery. The venous drainage cannula, typically 10–14
Fr, is placed into the internal jugular vein and advanced into the right atrium. This
configuration supplies oxygenated blood to the proximal aorta, but coronary and
right upper extremity blood may be poorly saturated if pulmonary dysfunction is
present, and the left ventricle is ejecting (see also Chap. 16).
Femoral cannulation can be used for venoarterial cardiac support in adults and
older children. A long drainage cannula, typically 27–29 Fr, is inserted into the
common femoral vein and advanced to the right atrium, and a short return cannula,
typically 17–19 Fr, is placed into the common femoral artery. Alternatively, a short
drainage cannula, typically 23–24 Fr, can be placed into the right atrium via the
internal jugular in place of the long femoral venous cannula. These approaches are
suitable if the native lungs can provide adequate saturation of blood since, in the
presence of cardiac ejection, the upper half of the body is supplied by the native
heart and lungs and the lower half by the extracorporeal circuit.

Venovenous (VV) Configuration

Venovenous cannulation was introduced later than venoarterial and is suitable for
respiratory failure with adequate cardiovascular function. It provides oxygenated
blood into the venous system while relying on the native heart for oxygen delivery,
making oxygenated blood available to all tissues, including the myocardium.
Percutaneous cannulation for venovenous support has replaced surgical cannulation
as the preferred approach.
The venovenous configuration was introduced to extracorporeal support using
two single-lumen cannulas, one placed into the proximal IVC via the femora vein
for drainage, typically 27–29 Fr, and the second return cannula, typically 23–24 Fr,
placed into the SVC via the right or left internal jugular veins for return. This direc-
tion of flow (femoral-jugular) has been shown to result in less recirculation than the
jugular-femoral direction [16, 17]. An alternative configuration that can result in
higher flows and even lower recirculation is the placement of three cannulas, one in
the SVC via an internal jugular and one in the proximal IVC via a femoral vein, with
the return cannula placed into the right atrium via the femoral or jugular approach
[18]. This separation of venous return reduces recirculation to levels lower than pos-
sible with the two-cannula approach.
A significant advance in venovenous support was the introduction of the dual-­
lumen, bicaval venovenous cannula for respiratory support (see dual-lumen design,
above). Developed initially for neonates [19], cannulas are also available for pediat-
ric and adult patients. This cannula design allows a single vascular access via the
right internal jugular vein. The cannula is placed by accessing the right internal
jugular, advancing the guidewire across the atrium into the IVC under fluoroscopic
imaging, then inserting the cannula with the distal drainage port in the proximal
4 Vascular Access 107

IVC, return port in the mid-right atrium, and second drainage port in the SVC. The
catheter can be placed via the left internal jugular in selected patients, but other
access points are not supported.

Venopulmonary (VP) Configuration

An alternative to dual single-lumen and bicaval dual-lumen cannulation for respira-

tory support, especially useful in the presence of right ventricular failure, is the
atrio-pulmonary cannula. Insertion involves access of the right internal jugular vein
followed by placement of a sheath. A balloon-tipped catheter is advanced through
the sheath and floated into the pulmonary circulation under fluoroscopic guidance.
A guidewire is then advanced through this catheter into the pulmonary artery. The
guide catheter is removed under visualization, leaving the guidewire in place. The
internal jugular access site is then dilated to accept the cannula, which is then
advanced with its loading dilator into position with the tip in the pulmonary artery
and the drainage port in the right atrium.

Venovenoarterial (VVA) Configuration

A hybrid cannulation configuration that provides both venovenous and venoarterial

support is the venovenoarterial configuration. The configuration drains blood from
a venous drainage cannula and returns blood to both the venous system and the arte-
rial system. Cannulation for this approach is usually performed with a long single-­
lumen drainage cannula (27–29 Fr) placed into the proximal IVC through a femoral
vein, and short single-lumen return cannulas placed in the SVC via the internal
jugular vein and in the common femoral artery. Cannulation for VVA can be per-
formed at the initiation of support if both cardiac and respiratory support are
required or can be staged by starting with VV support for respiratory failure and
adding VA support if cardiac failure develops, or by starting with VA support for
cardiac failure and adding VV support if respiratory failure ensues.
It is feasible to convert VV cannulation in a patient with respiratory failure sup-
ported with a dual-lumen, bicaval cannula to VVA by adding a single-lumen can-
nula for arterial cannulation. Recirculation in the VV cannula is not affected.

Low-Flow Venovenous Configuration

Venovenous support can target carbon dioxide removal (extracorporeal carbon diox-
ide removal, VV ECCO2R) to support lung-protective ventilation in patients for
whom oxygenation can be adequately provided through mechanical ventilation (see
Chaps. 6 and 10). Since ECCO2R only requires 0.5–1 L/min blood flow, cannulation
108 S. A. Conrad

for VV ECCO2R can be achieved by two smaller single-lumen cannulas (e.g., 14–16
Fr) placed into the femoral and jugular veins, or a dual-lumen cannula (15–16 Fr)
placed into either a femoral or jugular vein, similar to a hemodialysis catheter.

Arteriovenous (AV) Configuration

Another approach to extracorporeal carbon dioxide removal is pumpless arteriove-

nous extracorporeal carbon dioxide removal (AV ECCO2R), previously termed
interventional lung assist (ILA) or arteriovenous CO2 removal (AVCO2R). This con-
figuration involves cannulation of the femoral artery and vein with small cannulas
that provide blood flow on the order of 800–1000 mL/min. A 12 Fr arterial cannula
and a 16 Fr venous cannula can provide this blood flow. Cannulation technique is
the same as for any other single-lumen cannula insertion. The patient’s arterial
blood pressure provides the pressure gradient for flow. The major risk is the need for
arterial cannulation, but the cannula sizes described above have very low risk of
limb ischemia. As commercial VV ECCO2R systems enter the market, they will
likely replace AV ECCO2R much as continuous venovenous hemofiltration (CVVH)
has replaced continuous arteriovenous hemofiltration (CAVH).

Transthoracic (Central) Cannulation

Although much more invasive, direct cannulation of the right atrium and aortic root
through a median sternotomy remains an important approach to vascular access for
extracorporeal support. The most common use is support following failure to wean
from cardiopulmonary bypass (CPB), in which the cardiopulmonary bypass circuit
is replaced with the ECLS circuit, connecting directly to the right atrial drain can-
nula and the aortic root [20]. Typically, the sternum is left open and draped with an
occlusive dressing. CBP arterial and venous cannulas are large and support more
flow than can be achieved using peripheral access.
The transthoracic approach is associated with more bleeding and greater infec-
tion risk, so it is generally used for patients expected to recover cardiac function
quickly. If prolonged support is required, the patient may be transitioned to periph-
eral cannulation or to a ventricular assist device. This approach has also been used
to provide high-flow support in patients with severe sepsis [21].

Decannulation

When extracorporeal support is no longer required and weaning trials have been
successful (see Chap. 15), the patient is prepared for decannulation by discontinu-
ing anticoagulation to allow coagulation studies to return to near normal. In the case
4 Vascular Access 109

of percutaneous venous cannulation, continuing anticoagulation is an option that

may decrease the risk of venous thrombus formation. The cannulation sites are
exposed and prepared with antimicrobial solution. The circuit flow is stopped, and
the patient is removed from support by clamping the circuit near the cannulas. If
decannulation is delayed beyond discontinuation for any reason, the circuit tubing
is clamped and cut away from the cannulas, and the cannulas are flushed with hepa-
rinized saline to prevent thrombus formation.
Percutaneous venous cannulas are removed by first placing a horizontal mattress
suture in the incision surrounding the cannula, withdrawing the cannula, and secur-
ing the suture. No compression of the site is required since venous pressure is low,
and compression may predispose to venous thrombus formation. Full-dose antico-
agulation for 72–96 h and antiplatelet therapy with aspirin can be considered to
further reduce the risk of thrombus formation. Venous cannulas placed by the semi-­
open technique can be removed as if placed percutaneously.
Removal of percutaneous arterial cannulas depends on the size of the cannula
and duration of placement. Sizes up to 14–16 Fr can be removed by withdrawing the
cannula and applying pressure with fingertips until hemostasis is achieved, with
monitoring of the distal pulse so as not to fully compress the artery. Pressure is held
for 30 min using manual pressure directly over the arterial entry site. Mechanical
compression devices are discouraged due to difficulty in proper placement and the
wider area of applied pressure that may fail to prevent extravasation. Arterial punc-
ture closure devices may be used for cannulas about 16 Fr and above, and two
devices may be required for larger cannulas. Open surgical decannulation is an
option for decannulation of percutaneously placed cannulas and may be preferable
following long-duration support since the artery may have lost some of its integrity
and elasticity.
Surgically placed cannulas are removed with an open technique. The skin inci-
sion is reopened, temporary ligatures placed, and the cannula removed. The vessel
is either repaired or ligated, and the incision closed.

Complications of Cannulation

Vascular Injury

Injury to the target or adjacent vessel during cannulation can result in inability to
achieve vascular access, hemorrhage into areas such as the retroperitoneal space,
transection of a vessel, exsanguination, and death. Immediate attempts at surgical
exploration to complete cannulation or to repair injured vessels may be attempted
but may not be successful. The risk is higher with percutaneous cannulation since
the vessels are not visible. The use of ultrasound before and during percutaneous
cannulation can mitigate these risks by allowing for selection of appropriate cannula
size, identification of adjacent vessels, selection of an approach to avoid overlying
vessels, and guidance of vessel puncture in the center to ensure proper entry into the
vessel and reduce risk of injury.
110 S. A. Conrad

Inadequate Flow

The inability to achieve the expected circuit flow can result in the inability to achieve
adequate cardiac support during venoarterial ECMO, or persistent hypoxemia with
inability to achieve lung-protective settings during venovenous support, thereby
decreasing the chance of survival. Three conditions that commonly lead to inade-
quate flow are placement of a cannula that is smaller than required, improper place-
ment resulting in impaired venous drainage, and hypovolemia.
Choice of cannula size should be driven by the flow needed for adequate sup-
port, typically 50–80 mL/kg/min for an adult and higher for pediatric patients.
Drainage can almost always be achieved using a single, appropriately sized
drainage cannula. Uncommonly and in special circumstances, two drainage can-
nulas maybe required. The reinfusion cannula is typically smaller than the
drainage cannula since flow is driven with a much greater pressure gradient
generated by the blood pump and is rarely the cause of inadequate flow. Improper
placement can be identified by radiography or echocardiography and subse-
quently be corrected.
Hypovolemia is the most common transient cause of inadequate flow. It can
result in “chattering” or “chugging” of the venous line in which the vascular struc-
tures cyclically collapse around the cannula resulting in intermittent flow. The nega-
tive pressure generated at the pump during this sudden cessation of flow contributes
to hemolysis and must be corrected. Initially reducing the pump speed until volume
expansion with colloid or blood can be achieved, then restoration of flow, resolves
the problem.

Limb Ischemia

Ischemia of the lower limb due to complete or near-complete obstruction of the

femoral artery is one of the major risks associated with cannulation of the femoral
artery. Surgical cannulation requires the placement of a distal antegrade perfusion
catheter since the femoral artery is ligated as part of the procedure. An antegrade
cannula (6–8 Fr) placed into the superficial femoral artery or in the posterior tibial
artery can provide the necessary 200 mL of blood flow per minute to maintain limb
viability. During percutaneous cannulation the common femoral artery should be
targeted for the primary retrograde cannula. An alternative percutaneous strategy
is to place two smaller arterial cannulas, one in each femoral artery, together pro-
viding the total flow of a single larger cannula but reducing the risk of ischemia in
both limbs. The downside of this approach is that it doubles the risk of vascu-
lar injury.
If limb ischemia is not detected in time, sufficient muscle necrosis can occur that
may require fasciotomy or even amputation. Careful clinical examination, Doppler
4 Vascular Access 111

monitoring of distal blood pressure, and assessment with near-infrared spectros-

copy as described above can help identify this condition early. Many cannulating
physicians will routinely place a retrograde cannula at the time of cannulation in all
patients to minimize this risk.

Insertion Site Bleeding

Bleeding from cannulation sites is the most common bleeding complication. In

most instances it is minimal, but it can require intervention. Initial approach to man-
agement is to verify appropriate levels of anticoagulation, adequate platelet count,
normal prothrombin time, and adequate fibrinogen levels. The use of thromboelas-
tography (TEG) or thromboelastometry (ROTEM) provides insight into functional
abnormalities of coagulation that may not be detected by measurement of tradi-
tional lab values. Reduction of anticoagulation target and the application of topical
hemostatic agents may be helpful. Bleeding can be minimized by limiting the size
of the skin incision to allow the cannula to fit snugly when it is inserted. Bleeding
may be amenable to electrocautery if more conservative measures fail. In the case
of surgical cannulation, however, failure of more conservative measures may require
re-exploration of the cannulation site.

Recirculation

Recirculation during venovenous support occurs when some of the reinfused blood
is aspirated directly into the venous drainage cannula rather than being delivered to
the patient’s circulation, thereby reducing the effective extracorporeal flow. It is
unavoidable with the use of single-lumen cannulas, where recirculation fraction can
reach 30% or higher. Recirculation manifests as a decrease in oxygen delivery and
drop in systemic arterial saturation, with an increase in drainage saturation along
with a decrease in the saturation difference between the drainage and return limbs.
Increases in recirculation can occur with displacement of the cannulas and may
require radiography to detect. It also increases with increasing flow such that higher
flows may reduce oxygen delivery. In single-lumen cannulation, placement of the
drainage cannula in the proximal IVC rather than in the right atrium and placement
of the return cannula in the SVC above the cavo-atrial junction results in the lowest
recirculation [22].
Recirculation is less extensive with dual-lumen cannulas. The bicaval design
is associated with the lowest degree of recirculation, often under 3%. The veno-
pulmonary design also has minimal to absent recirculation. Recirculation is
absent in venoarterial cannulation since the circuit blood is returned into the arte-
rial system.
112 S. A. Conrad

Infectious Complications

Infection of the cannula insertion sites is a challenging problem since the patient
may be totally dependent on extracorporeal support, and recannulation may be risky
or impossible. Prevention is by assuring good skin asepsis at the time of cannulation
as well as throughout the duration of extracorporeal support. If infection does
develop and appears to be localized to the skin, then the use of appropriate antibiot-
ics may be successful in eradicating the infection.
If bacteremia develops, then consideration should be given to replacing the extra-
corporeal circuit after an initial treatment period with antibiotics, since seeding of
the large surface area circuit can result in persistent bacteremia. If the cannula site
infection is not responsive to antibiotic therapy alone, then recannulation at an alter-
native site may be required.

References

1. Conrad SA, Broman LM, Taccone FS, Lorusso R, Malfertheiner MV, Pappalardo F, et al. The
extracorporeal life support organization Maastricht treaty for nomenclature in extracorporeal
life support. A position paper of the extracorporeal life support organization. Am J Respir Crit
Care Med. 2018;198(4):447–51.
2. Broman LM, Taccone FS, Lorusso R, Malfertheiner MV, Pappalardo F, Di Nardo M, et al. The
ELSO Maastricht treaty for ECLS nomenclature: abbreviations for cannulation configuration
in extracorporeal life support - a position paper of the extracorporeal life support organization.
Crit Care. 2019;23(1):36.
3. von Segesser LK, Berdajs D, Abdel-Sayed S, Ferrari E, Halbe M, Wilhelm M, et al. New,
optimized, dual-lumen cannula for veno-venous ECMO. Perfusion. 2018;33(1_suppl):18–23.
4. Wang D, Zhou X, Liu X, Sidor B, Lynch J, Zwischenberger JB. Wang-Zwische double
lumen cannula-toward a percutaneous and ambulatory paracorporeal artificial lung. ASAIO
J. 2008;54(6):606–11.
5. Bermudez CA, Rocha RV, Sappington PL, Toyoda Y, Murray HN, Boujoukos AJ. Initial expe-
rience with single cannulation for venovenous extracorporeal oxygenation in adults. Ann
Thorac Surg. 2010;90(3):991–5.
6. Batchinsky AI, Jordan BS, Regn D, Necsoiu C, Federspiel WJ, Morris MJ, et al. Respiratory
dialysis: reduction in dependence on mechanical ventilation by venovenous extracorporeal
CO2 removal. Crit Care Med. 2011;39(6):1382–7.
7. Morimont P, Batchinsky A, Lambermont B. Update on the role of extracorporeal CO(2)
removal as an adjunct to mechanical ventilation in ARDS. Crit Care. 2015;19:117.
8. Iserson KV. J.-F.-B. Charriere: the man behind the “French” gauge. J Emerg Med.
1987;5(6):545–8.
9. Bizzarro MJ, Conrad SA, Kaufman DA, Rycus P, Extracorporeal Life Support Organization
Task Force on Infections EMO. Infections acquired during extracorporeal membrane oxygen-
ation in neonates, children, and adults. Pediatr Crit Care Med. 2011;12(3):277–81.
10. Extracorporeal Life Support Organization Task Force on Infections. Infection control and extra-
corporeal life support. 2010. Available from: http://elso.org/downloads/resources/committees/
infectious-­disease-­and-­antibiotic/Infection-­Control-­and-­Extracorporeal-­Life-­Support.pdf.
11. Pranikoff T, Hirschl RB, Remenapp R, Swaniker F, Bartlett RH. Venovenous extracorporeal
life support via percutaneous cannulation in 94 patients. Chest. 1999;115(3):818–22.
4 Vascular Access 113

12. Conrad SA, Grier LR, Scott LK, Green R, Jordan M. Percutaneous cannulation for extracorpo-
real membrane oxygenation by intensivists: a retrospective single-institution case series. Crit
Care Med. 2015;43(5):1010–5.
13. Burrell AJ, Pellegrino VA, Sheldrake J, Pilcher DV. Percutaneous cannulation in predomi-
nantly venoarterial extracorporeal membrane oxygenation by intensivists. Crit Care Med.
2015;43(12):e595.
14. Kim DJ, Cho YJ, Park SH, Lim C, Park KH, Jheon S, et al. Near-infrared spectroscopy moni-
toring for early detection of limb ischemia in patients on veno-arterial extracorporeal mem-
brane oxygenation. ASAIO J. 2017;63(5):613–7.
15. Breeding J, Hamp T, Grealy R, Nair P, Iyer A, Kawanishi Y. Effects of extracorporeal membrane
oxygenation pump flow, backflow cannulae, mean arterial blood pressure, and pulse pressure
on Doppler-derived flow velocities of the lower limbs in patients on peripheral veno-arterial
extracorporeal membrane oxygenation: a pilot study. Aust Crit Care. 2019;32(3):206–12.
16. Lee JH, Won JY, Han JU, Son HS, Jung JS. Differences in recirculation: differences according
to different methods of cannulation in veno-venous extracorporeal membrane oxygenation.
Perfusion. 2018;33(1 Suppl):1420143.
17. Conrad SA, Wang D. Evaluation of recirculation during venovenous extracorporeal membrane
oxygenation using computational fluid dynamics incorporating fluid-structure interaction.
ASAIO J. 2021;67(8):943–53.
18. Ichiba S, Peek GJ, Sosnowski AW, Brennan KJ, Firmin RK. Modifying a venovenous
extracorporeal membrane oxygenation circuit to reduce recirculation. Ann Thorac Surg.
2000;69(1):298–9.
19. Anderson HL 3rd, Otsu T, Chapman RA, Barlett RH. Venovenous extracorporeal life support
in neonates using a double lumen catheter. ASAIO Trans. 1989;35(3):650–3.
20. Field ML, Al-Alao B, Mediratta N, Sosnowski A. Open and closed chest extrathoracic cannu-
lation for cardiopulmonary bypass and extracorporeal life support: methods, indications, and
outcomes. Postgrad Med J. 2006;82(967):323–31.
21. Maclaren G, Butt W, Best D, Donath S, Taylor A. Extracorporeal membrane oxygenation
for refractory septic shock in children: one institution’s experience. Pediatr Crit Care Med.
2007;8(5):447–51.
22. Conrad SA. Computational simulation of recirculation during venovenous extracorporeal
membrane oxygenation. Perfusion. 2018;33(1 Suppl):142–3.
Chapter 5
Hypoxemic Respiratory Failure: Evidence,
Indications, and Exclusions

Kathleen E. Melville, Cara Agerstrand, Daniel Brodie, and Darryl Abrams

Abbreviations

ARDS Acute respiratory distress syndrome

ECMO Extracorporeal membrane oxygenation
PaO2 Partial pressure of oxygen in arterial blood
PaCO2 Partial pressure of carbon dioxide in arterial blood
FiO2 Fraction of inspired oxygen
PEEP Positive end-expiratory pressure
VILI Ventilator-induced lung injury

K. E. Melville · D. Brodie · D. Abrams (*)

Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine,
Columbia University College of Physicians & Surgeons, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]; [email protected];
[email protected]
C. Agerstrand
Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine,
Columbia University College of Physicians and Surgeons/NewYork-Presbyterian Hospital,
New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 115

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_5
116 K. E. Melville et al.

Evidence

ECMO is increasingly being used for patients with the acute respiratory distress
syndrome (ARDS), particularly in cases of severe ARDS in which life-threatening
hypoxemia or hypercapnia persists despite maximal conventional mechanical ven-
tilatory support [1, 2]. In addition, ECMO is used in some patients in whom life-­
threatening gas exchange abnormalities are sufficiently improved with the use of
positive-pressure ventilation in order to sustain life, yet only at the expense of gen-
erating excessively high inspiratory airway pressures. ECMO in this setting facili-
tates lung-protective ventilation and minimizes ventilator-induced lung injury. For
decades, there was an increase in the use of ECMO in ARDS despite high level
evidence supporting its benefit. In the last several years, however, there has been
compelling data from a large randomized controlled trial as well as subsequent
analyses that have helped better inform the use of ECMO for this indication [3–7].
The first successful use of ECMO for ARDS dates back to the early 1970s [8]. In
the following decades, there was enthusiasm for the use of this modality, but small
case series had notably high mortality, and two randomized controlled trials, one
with ECMO and the other with extracorporeal carbon dioxide removal, did not
show benefit [9, 10]. The early technology being used at the time, combined with
the inexperience of the centers, led to serious complications and seemed to offset
any potential benefit that the patients may have realized.
Interest in ECMO for ARDS was reignited around 2009 due in part to increased
use in patients with respiratory failure during the influenza A(H1N1) pandemic, as
well as the publication of a large randomized controlled trial comparing ECMO to
conventional management [11]. An observational study of 68 patients with con-
firmed or suspected H1N1-associated ARDS treated with ECMO showed a survival
of 75% [12, 13]. However, a study with a similar patient population managed with-
out the use of ECMO demonstrated nearly identical outcomes, which raised ques-
tions about whether ECMO provides any survival advantage over conventional
medical management in this population [14]. In an attempt to reconcile this issue, a
comparison of outcomes between ECMO-referred and non-ECMO-referred sub-
jects with confirmed or suspected H1N1-associated ARDS was performed [15].
After matching for appropriate variables, in this intention-to-treat analysis, ECMO-­
referred subjects consistently had a mortality approximately half that of the non-­
ECMO-­referred subjects (24% vs. 47% by propensity score matching, RR 0.51,
95% CI 0.31–0.84, p = 0.008).
Survival rates of subjects receiving extracorporeal support in these observational
studies were higher than those reported for subjects with and without extracorporeal
support in the earlier randomized trials. However, there are inherent flaws in com-
paring non-randomized studies, and changes in clinical management over time con-
found the comparison of survival rates from different eras. Likewise, ECMO
technology had evolved significantly since the early randomized trials, with more
efficient membranes for gas exchange, the advent of centrifugal pumps,
heparin-­coated circuits that tolerate lower levels of anticoagulation resulting in
5 Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions 117

lower bleeding risk, and cannulae that permit single-vessel access with minimal
recirculation.
In an attempt to estimate the effect of ECMO in ARDS using mostly modern
ECMO technology and coinciding with increasing usage within the critical care
community in the setting of the 2009 pandemic, the Conventional Ventilation or
ECMO for Severe Adult Respiratory Failure (CESAR) trial was performed [11]. In
this prospective, randomized, controlled trial, 180 subjects, age 18–65, with severe
but potentially reversible respiratory failure and a Lung Injury Score (a.k.a. Murray
Score, a composite score based on the ratio of partial pressure of oxygen in arterial
blood to the fraction of inspired oxygen (PaO2:FIO2), positive end-expiratory pres-
sure (PEEP), respiratory system compliance, and radiographic findings) of ≥3.0 or
uncompensated hypercapnia (pH <7.2) despite “optimal conventional management”
were randomly assigned to receive ongoing conventional mechanical ventilation at
designated treatment centers or be transferred to a single ECMO center for consid-
eration of treatment with venovenous ECMO. Hemodynamically stable subjects
randomized to the ECMO referral arm were initially managed on transfer with a
standardized management protocol that included a pressure-restricted ventilation
strategy, diuresis, and prone positioning. Those who were hemodynamically unsta-
ble or failed to respond to this strategy within 12 h were placed on ECMO. Only
76% of the subjects referred for ECMO actually received ECMO; however all sub-
jects who received ECMO were managed with a lung-protective ventilation strat-
egy. In total, 93% of subjects in the ECMO referral arm received treatment with a
low-volume, low-pressure strategy at some point in their care. By comparison,
because there was no mandate for a lung-protective ventilation strategy in the con-
ventional management group and perhaps because many of these subjects were dif-
ficult to ventilate, only 70% of those subjects were managed with such a strategy at
any time during the study. The primary outcome—death or severe disability at
6 months after randomization—occurred in 37% of the subjects referred for ECMO,
as compared with 53% of those in the conventional management group, with a rela-
tive risk of 0.69 (95% confidence interval 0.05–0.97, p = 0.03).
The results from CESAR suggested that it is reasonable to transfer patients with
severe ARDS to a center capable of performing ECMO as part of a standardized
management protocol. However, this trial was not a randomized trial of ECMO as
compared with standard-of-care mechanical ventilation. The higher survival in the
ECMO-referred group could also be accounted for by differences in the care
between study groups, most importantly, the discrepancy in the use of a low-­volume,
low-pressure mechanical ventilation strategy.
In 2018, the ECMO to Rescue Lung Injury in Severe ARDS (EOLIA) trial was
published in an attempt to reconcile questions left unanswered by CESAR [3]. In
this international, prospective, randomized controlled trial, 259 patients with very
severe ARDS were randomized to immediately receive venovenous ECMO or to
continue conventional treatment. Eligible patients met the American-European
Consensus Conference definition of ARDS, had been receiving invasive mechanical
ventilation for fewer than 7 days, and met certain disease-severity criteria, all while
118 K. E. Melville et al.

adhering to a low tidal volume strategy and receiving an FiO2 of greater or equal to
0.8 with a PEEP of at least 10 cm H2O. The specific disease severity criteria included
a PaO2:FiO2 of less than 50 mmHg for more than 3 h, a PaO2:FiO2 of less than
80 mmHg for more than 6 h, or an arterial blood pH of less than 7.25 with a partial
pressure of arterial carbon dioxide (PaCO2) of at least 60 mmHg while maintaining
a high respiratory rate of 35 breaths per minute, in an attempt to achieve a plateau
pressure below 33 cm H2O (see Table 5.1).
The trial was stopped early for projected futility in meeting the primary end-
point, a prespecified 20% absolute mortality reduction in the group randomized to
receive ECMO.
There was, however, a large, but not statistically significant, difference in mortal-
ity at 60 days. Forty-four of 124 (35%) of patients in the ECMO group and 57 of
125 (46%) in the control group died, yielding a relative risk of death of 0.76 (95%
CI 0.55–1.04, P = 0.09).
Crossover to ECMO for patients with severe refractory hypoxemia was permit-
ted under the trial design based on strict criteria. A total of 35 patients (28%) in the
control group crossed over and were cannulated to venovenous ECMO. These
patients, on average, had more severe ARDS, and worse multiorgan dysfunction,
including cardiovascular failure, compared to the cohort overall. The 60-day mor-
tality in this subset was 57% (20 of 35 patients), but a reasonable argument can be
made that significantly more of them would have died without receiving
ECMO. Therefore, in the intention-to-treat analysis, the mortality rates in the

Table 5.1 Inclusion criteria for CESAR and EOLIA

CESAR, 2009 [11]
Age 18–65
Severe but potentially reversible respiratory failure
 Lung injury (Murray) scorea ≥ 3
 Uncompensated hypercapnia (pH <7.2) despite optimal ventilator management
EOLIA, 2018 [3]
Age 18 and above
Very severe ARDSb
 PaO2:FiO2 < 50 mmHg for more than 3 h
 PaO2:FiO2 < 80 mmHg for more than 6 h
 pH <7.25 with PaCO2 ≥ 60c
Mechanical ventilation <7 days
EOLIA extracorporeal membrane oxygenation for severe acute respiratory distress syndrome,
CESAR efficacy and economic assessment of conventional ventilatory support versus extracorpo-
real membrane oxygenation for severe adult respiratory failure, ARDS acute respiratory distress
syndrome, ECMO extracorporeal membrane oxygenation, PaO2 partial pressure of oxygen in arte-
rial blood, PaCO2 partial pressure of carbon dioxide in arterial blood, FiO2 fraction of inspired
oxygen, PEEP positive end-expiratory pressure
a
Composite score incorporating PaO2:FIO2, PEEP, respiratory system compliance, and radio-
graphic findings
b
With low tidal volume ventilation, FiO2 ≥ 0.8, and PEEP ≥10
c
While maintaining respiratory rate of 35 in attempt to achieve plateau pressure ≤ 32 cm H2O
5 Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions 119

control group were likely lower than they would have been without the allowance of
crossover. Similarly, an important secondary endpoint of “treatment failure,” which
was defined as death in the ECMO group and death or crossover to ECMO in the
control group, was significantly different between the two groups (35% in the
ECMO group and 58% in the treatment group).
In contrast to CESAR and other earlier trials, the patients in the control group
were held to strict standards of lung protective mechanical ventilation, consistent
with current standards of care. Ninety percent of patients in the control arm were
placed in the prone position, and 56% of patients in the intervention arm were
placed prone prior to ECMO initiation (an additional 10% were placed prone during
ECMO). Ventilator parameters, including tidal volumes, plateau airway pressures,
and driving pressures, were able to be decreased to a greater extent in the group
treated with ECMO, with post hoc analysis suggesting that the greatest benefit was
seen in those patients entering the trial under the third criteria related to poorly
compliant lungs. It is therefore reasonable to assume that much of the benefit
observed is due, in part, to minimization of ventilator-induced lung injury (VILI).
Importantly, overall differences in complications and adverse events were not
statistically significant between the two groups. However, there were more bleeding
events necessitating blood transfusion as well as cases of severe thrombocytopenia
in the ECMO group. There were fewer cases of ischemic stroke in the ECMO group
compared to the control group.
Using a standard frequentist approach to statistical analysis, the EOLIA trial was
negative, having not demonstrated a statistically significant difference in mortality
between the ECMO and control groups. However, given the difference in mortality
between groups, the data was reevaluated using a Bayesian analysis to add to study
interpretation [5]. In clinical trials evaluated using a traditional frequentist analysis,
the study is considered positive if the evidence is enough to confidently reject the
null hypothesis. In contrast, the Bayesian method allows a probability to be assigned
to the hypothesis, incorporating prior beliefs and prior data into the study interpreta-
tion. A range of subjective reference priors, from “strongly skeptical” to “strongly
enthusiastic” were used in modeling, along with data-derived prior distributions,
which were based on previous relevant studies of ECMO. After incorporation of
these “prior” probabilities, the study reported posterior probabilities of a mortality
benefit of ECMO. They found that the probability of a relative risk of less than one
(e.g., probability of a mortality benefit of early ECMO) was 88–99% across the
range of prior assumptions. For an absolute risk reduction (ARR) of mortality of 2%
or more, the posterior probabilities ranged from 78% to 98%, whereas for an ARR
of at least 20% (i.e., the prespecified anticipated mortality reduction), the posterior
probabilities were much lower, ranging from 0% to 2%. This elegant reanalysis
provides a wider context for interpreting the data from the EOLIA trial that may be
more clinically intuitive than the frequentist approach. It illustrates the high proba-
bility that ECMO lowers mortality in this context, and as the accompanying edito-
rial stated, the key questions left to be answered are “by how much does ECMO
work, in whom, and at what cost?” [16].
120 K. E. Melville et al.

Given the difficulty in enrolling patients for EOLIA, attributable, in part, to the
lack of clinical equipoise, which has only grown more among those performing
ECMO since the results of EOLIA, there will likely not be another large random-
ized trial of ECMO for ARDS. Therefore, following the publication of EOLIA,
composite studies have attempted to combine and better interpret all the existing
data we do have to quantify the overall benefit of ECMO.
A meta-analysis by Munshi et al. included five studies, with a total of 773
patients, and was the first to incorporate data from the two modern randomized
controlled trials of ECMO in severe acute respiratory failure (CESAR and EOLIA)
[6]. The other studies included were observational in nature with matching tech-
niques. Using the pooled data from the randomized controlled trials (429 patients),
the primary outcome of 60-day mortality was significantly lower for those receiving
ECMO (RR 0.73; 95% CI 0.58–0.92) than for those that did not. When data from all
five of the studies were analyzed together, there was also a significantly lower
30-day mortality in the ECMO group (RR 0.69; 95% CI 0.5–0.95). The meta-­
analysis did not pool adverse events given inconsistency in the reporting across
studies.
As discussed previously, over the time that these trials have been performed,
ventilator management has also improved in an attempt to limit VILI, and strategies
using low tidal volumes, low airway pressures, high PEEP, and prone positioning
have become standard to varying extents. Given that many of these studies, and
other randomized controlled trials of therapies in ARDS, do not directly compare all
possible modalities commonly used in the treatment of ARDS, network meta-­
analyses are useful to quantify the relative association of different management
strategies on mortality. Sud et al. compared the effects of low tidal volume ventila-
tion, prone positioning, high PEEP, and venovenous ECMO, among others, on in-­
hospital mortality in ARDS [7]. Analyzing 34 randomized controlled trials, with
9085 patients, they found that prone positioning combined with low tidal volume
ventilation yielded the best outcomes (RR 0.74, 95% CI 0.6–0.92), but with veno-
venous ECMO also rating highly (RR 0.78, 95% CI 0.58–1.05) when compared to
low tidal volume ventilation. The data used in the latter comparison comes from
EOLIA and, therefore, includes only those patients with very severe ARDS [3].
Data from the 429 patients in EOLIA and CESAR were then combined in an
individual patient data meta-analysis, looking for differences in 90-day mortality
between conventional treatment and VV ECMO [4]. This revealed a relative risk of
death of 0.75 in the ECMO group, (95% CI 0.6–0.94, p = 0.013) as 36% of those in
the ECMO group died, compared with 48% in the control group. The relative risk
of treatment failure (defined as death in the ECMO group and death or crossover to
ECMO in the control group) was 0.65 (95% CI 0.52–0.8). Additionally, those ran-
domized to ECMO had more ventilator-free days, more days out of the intensive
care unit, and more days without vasopressors, renal replacement therapy, and neu-
rologic failure than those in the control group.
Taken together, there is compelling evidence to suggest that early initiation of
venovenous ECMO for severe ARDS leads to reduction in mortality compared with
conventional management. The use of “ultra-lung-protective” ventilation with
5 Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions 121

ECMO and the minimization of VILI likely drive a considerable proportion of the
improved outcomes seen with ECMO [17].
The role of ECMO in ARDS related to emerging infectious diseases is evolving.
As discussed previously, a relative surge in very severe ARDS case volume during
the influenza A(H1N1) pandemic in 2009 greatly increased the use and interest in
ECMO for respiratory failure. ECMO was also used for severe cases of ARDS dur-
ing the outbreak of Middle East Respiratory Syndrome (MERS) in 2012, but that
data is more limited, especially given comparatively fewer cases. ECMO centers
began using ECMO for severe ARDS due to coronavirus disease 2019 (COVID-19)
near the onset of the pandemic in 2020. In 2021, a systematic review and meta-­
analysis was published, using observational studies of ECMO in adults with
COVID-19 ARDS over the first year of the pandemic [18]. Twenty-two studies,
with a total of 1896 patients, were included, and almost all cases (98.6%) used
venovenous ECMO. The primary outcome of interest was in-hospital mortality, and
secondary outcomes included duration of ECMO and mechanical ventilation, abil-
ity to wean from ECMO, and complications during therapy. In the studies that
reported pre-initiation PaO2:FiO2 ratios (1344 patients), the mean was 67.8. Prior to
ECMO initiation, incidence of prone positioning was 85.3%, and 96.3% of the
patients had received neuromuscular blocking agents. The in-hospital mortality of
those patients receiving venovenous ECMO was 35.7% (95% CI 30.7–40.7). Meta-­
regression found that both age and ECMO duration were associated with increased
mortality, whereas increasing BMI seemed protective. In the full cohort, mean dura-
tion of ECMO (18 studies, 1844 patients) was 15.1 days, mean intensive care unit
length of stay was 29 days, and liberation from ECMO therapy was accomplished
in 67.6% of the cases. Two of the studies included in this analysis compared mortal-
ity rates between patients receiving ECMO with those receiving conventional ther-
apy with mechanical ventilation. Mortality rates of those placed on ECMO were
46.2% and 57.1% compared with 47.8% and 63.2%, respectively. As experience
increases for this indication, appropriate patient selection criteria may also be
refined. The use of ECMO in a potentially strained and resource-limited environ-
ment, such as that encountered during a pandemic, also raises ethical and logistical
issues that need to be navigated thoughtfully [19–23].

Indications

The decision to initiate ECMO is a complicated one, ideally based on a risk-benefit

analysis that incorporates the risk of mortality with or without extracorporeal life
support while factoring in the risk of complications as a result of its use. The most
robust data for ECMO in ARDS is from the EOLIA trial, and thus, its inclusion
criteria may be used to help guide patient selection [3]. In cases where, despite opti-
mization of conventional therapy, patients meet these criteria, ECMO should be
considered [24]. Specifically, those with PaO2:FiO2 < 80 mmHg for 6 h, <50 mmHg
122 K. E. Melville et al.

for 3 h, or by an arterial blood pH of <7.25 with PaCO2 of at least 60 mmHg for 6 h

while receiving greater or equal to 80% FiO2 and 10 cm H2O of PEEP may benefit
from ECMO initiation.

Exclusions

Venovenous ECMO for hypoxemic respiratory failure should be used either as a

bridge to recovery or as a bridge to lung transplantation (discussed in detail else-
where). Therefore, the only absolute contraindication to VV ECMO for hypoxemic
respiratory failure is advanced, irreversible lung disease without the option of trans-
plantation. There are numerous relative contraindications to VV ECMO, and the
decision to initiate ECMO is always a balance of risk and benefit based on the
individual patient. Inability to tolerate anticoagulation, preexisting severe coagu-
lopathies, and inability to accept blood products should all be carefully considered
prior to initiation of ECMO. Systemic anticoagulation is ideally administered while
a patient is on circuit (although it is not necessary in all cases), and these issues may
significantly complicate management or lead to adverse events that decrease the
potential benefit of the intervention. The benefit of ECMO for severe ARDS may
also be attenuated in those who have been exposed to high airway pressures and
high FiO2 for prolonged periods of time. Generally, mechanical ventilation for over
7 days under these conditions is considered a relative contraindication to ECMO
initiation, although this contraindication in particular should be applied with cau-
tion. Other relative contraindications include limitations to vascular access that
would preclude cannula placements and any conditions in which ECMO would be
unlikely to alter the patient’s overall prognosis, including but not limited to,
advanced malignancy or severe and irreversible brain injury (see Table 5.2).

Table 5.2 Contraindications for ECMO for ARDS

Absolute contraindication
Severe, irreversible lung disease without the option for lung transplantation
Relative contraindications
Inability to tolerate anticoagulation
Invasive mechanical ventilation with high FiO2 and airway pressures for >7 days
Limitations to vascular access
Concurrent severe irreversible organ failures or comorbidities
ARDS acute respiratory distress syndrome, ECMO extracorporeal membrane oxygenation, FiO2
fraction of inspired oxygen
5 Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions 123

Conclusion

The use of venovenous ECMO as a therapy for severe ARDS is now supported by
high-level evidence [3–7]. Increased experience and improvements in technology,
coupled with appropriate patient selection and a better understanding of the mecha-
nism of benefit, have helped to improve outcomes.

References

1. Brodie D, Slutsky AS, Combes A. Extracorporeal life support for adults with respiratory fail-
ure and related indications: a review. JAMA. 2019;322(6):557–68.
2. Combes A, Schmidt M, Hodgson CL, Fan E, Ferguson ND, Fraser JF, et al. Extracorporeal
life support for adults with acute respiratory distress syndrome. Intensive Care Med.
2020;46(12):2464–76.
3. Combes A, Hajage D, Capellier G, Demoule A, Lavoue S, Guervilly C, et al. Extracorporeal
membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med.
2018;378(21):1965–75.
4. Combes A, Peek GJ, Hajage D, Hardy P, Abrams D, Schmidt M, et al. ECMO for severe
ARDS: systematic review and individual patient data meta-analysis. Intensive Care Med.
2020;46(11):2048–57.
5. Goligher EC, Tomlinson G, Hajage D, Wijeysundera DN, Fan E, Juni P, et al. Extracorporeal
membrane oxygenation for severe acute respiratory distress syndrome and posterior prob-
ability of mortality benefit in a post hoc Bayesian analysis of a randomized clinical trial.
JAMA. 2018;320(21):2251–9.
6. Munshi L, Walkey A, Goligher E, Pham T, Uleryk EM, Fan E. Venovenous extracorporeal
membrane oxygenation for acute respiratory distress syndrome: a systematic review and meta-­
analysis. Lancet Respir Med. 2019;7(2):163–72.
7. Sud S, Friedrich JO, Adhikari NKJ, Fan E, Ferguson ND, Guyatt G, et al. Comparative effec-
tiveness of protective ventilation strategies for moderate and severe acute respiratory distress
syndrome. A network meta-analysis. Am J Respir Crit Care Med. 2021;203(11):1366–77.
8. Hill JD, O’Brien TG, Murray JJ, Dontigny L, Bramson ML, Osborn JJ, et al. Prolonged extra-
corporeal oxygenation for acute post-traumatic respiratory failure (shock-lung syndrome). Use
of the Bramson membrane lung. N Engl J Med. 1972;286(12):629–34.
9. Zapol WM, Snider MT, Hill JD, Fallat RJ, Bartlett RH, Edmunds LH, et al. Extracorporeal
membrane oxygenation in severe acute respiratory failure. A randomized prospective study.
JAMA. 1979;242(20):2193–6.
10. Morris AH, Wallace CJ, Menlove RL, Clemmer TP, Orme JF Jr, Weaver LK, et al. Randomized
clinical trial of pressure-controlled inverse ratio ventilation and extracorporeal CO2 removal
for adult respiratory distress syndrome. Am J Respir Crit Care Med. 1994;149(2 Pt 1):295–305.
11. Peek GJ, Mugford M, Tiruvoipati R, Wilson A, Allen E, Thalanany MM, et al. Efficacy and
economic assessment of conventional ventilatory support versus extracorporeal membrane
oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised con-
trolled trial. Lancet. 2009;374(9698):1351–63.
12. Davies A, Jones D, Bailey M, Beca J, Bellomo R, Blackwell N, et al. Extracorporeal
membrane oxygenation for 2009 influenza a(H1N1) acute respiratory distress syndrome.
JAMA. 2009;302(17):1888–95.
13. Davies A, Jones D, Gattas D. Extracorporeal membrane oxygenation for ARDS due to 2009
influenza a(H1N1)—reply. JAMA. 2010;303(10):941–2.
124 K. E. Melville et al.

14. Miller RR 3rd, Markewitz BA, Rolfs RT, Brown SM, Dascomb KK, Grissom CK, et al.
Clinical findings and demographic factors associated with ICU admission in Utah due to novel
2009 influenza A(H1N1) infection. Chest. 2010;137(4):752–8.
15. Noah MA, Peek GJ, Finney SJ, Griffiths MJ, Harrison DA, Grieve R, et al. Referral to an
extracorporeal membrane oxygenation center and mortality among patients with severe 2009
influenza A(H1N1). JAMA. 2011;306(15):1659–68.
16. Lewis RJ, Angus DC. Time for clinicians to embrace their inner Bayesian?: reanalysis of results
of a clinical trial of extracorporeal membrane oxygenation. JAMA. 2018;320(21):2208–10.
17. Abrams D, Schmidt M, Pham T, Beitler JR, Fan E, Goligher EC, et al. Mechanical ventilation
for ARDS during extracorporeal life support: research and practice. Am J Respir Crit Care
Med. 2019;201(5):514–25.
18. Ramanathan K, Shekar K, Ling RR, Barbaro RP, Wong SN, Tan CS, et al. Extracorporeal
membrane oxygenation for COVID-19: a systematic review and meta-analysis. Crit Care.
2021;25(1):211.
19. Supady A, Badulak J, Evans L, Curtis JR, Brodie D. Should we ration extracorporeal mem-
brane oxygenation during the COVID-19 pandemic? Lancet Respir Med. 2021;9(4):326–8.
20. Supady A, Brodie D, Curtis JR. Ten things to consider when implementing rationing guide-
lines during a pandemic. Intensive Care Med. 2021;47(5):605–8.
21. Supady A, Curtis JR, Abrams D, Lorusso R, Bein T, Boldt J, et al. Allocating scarce inten-
sive care resources during the COVID-19 pandemic: practical challenges to theoretical frame-
works. Lancet Respir Med. 2021;9(4):430–4.
22. Abrams D, Lorusso R, Vincent JL, Brodie D. ECMO during the COVID-19 pandemic: when
is it unjustified? Crit Care. 2020;24(1):507.
23. Badulak J, Antonini MV, Stead CM, Shekerdemian L, Raman L, Paden ML, et al. Extracorporeal
membrane oxygenation for COVID-19: updated 2021 guidelines from the extracorporeal life
support organization. ASAIO J. 2021;67(5):485–95.
24. Abrams D, Ferguson ND, Brochard L, Fan E, Mercat A, Combes A, et al. ECMO for ARDS:
from salvage to standard of care? Lancet Respir Med. 2019;7(2):108–10.
Chapter 6
Ventilator Management During ECLS

Antonio Pesenti, Giacomo Bellani, Giacomo Grasselli, and Tommaso Mauri

“The wisdom of old men. They do not grow wise. They grow careful.”
From Farewell to arms
by Ernest Hemingway

Introduction

In this chapter we will discuss the ventilatory management during ECLS. We will
focus almost exclusively on acute hypoxemic respiratory failure, but the principles
here exposed are by and large applicable to hypercapnic respiratory failure too.
Conversely, we will sometimes introduce concepts that, though derived from the
experience of ECLS in chronic respiratory failure patients, we presume applicable
to the acute patient as well.
The reader will forgive us for taking first a historical approach. Extracorporeal
membrane oxygenation (ECMO) got to the stage in clinical medicine when in 1974
the National Institutes of Health (NIH) funded the first prospective controlled ran-
domized clinical study on ARDS [1]. It aimed at comparing the outcome of severe
acute respiratory failure patients treated by veno-arterial (VA) ECMO with the out-
come of patients treated conventionally by mechanical ventilation. The results of
the study were discouraging, showing the same high mortality (exceeding 90%) in
both arms [1]. In discussing their results, Zapol and colleagues identified, among

A. Pesenti (*) · G. Bellani · G. Grasselli · T. Mauri

Department of Pathophysiology and Transplant, University of Milan, Milan, Italy
Department of Anestesia and Critical Care, Fondazione IRCCS Ca’ Granda Ospedale
Maggiore Policlinico, Milan, Italy
e-mail: [email protected]; [email protected];
[email protected]; [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature 125

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_6
126 A. Pesenti et al.

others, two possible reasons for the study failure: (a) the VA mode, that could have
increased the incidence of pulmonary vascular thrombosis, and (b) the decrease in
tidal volume and respiratory rate applied in the ECMO group which might have
caused a worsening of respiratory compliance. However, following the institution of
ECLS, tidal volume was decreased only to a little extent, as it went from approxi-
mately 800 mL to an average of approximately 600 mL. Moreover, this decrease
was associated with a constantly high plateau pressure, in a range between 40 and
50 cmH2O [2]. At that time, indeed, the recommended tidal volume was 10–15 mL/
kg or more, as it was one of the major determinants of oxygenation [3, 4]. Taking an
opposite position, in 1981, Ted Kolobow wrote a provocative editorial: “Why did
ECMO fail?” [5] in which he suggested that “severely diseased lungs have a chance
to heal only if the environment remains conducive to the healing of the lungs. This
environment does not consist of high airway pressures, high tidal volumes, high
PEEP, high FiO2, or a severe pulmonary hypoperfusion with severe and lethal lung
tissue alkalosis” [5]. He had developed in his laboratory at the NIH a prototype that
became the first commercially available membrane lung, and he was one of the most
experienced scientists in this field [6]. He pointed out that providing viable blood
gases by extracorporeal means was not enough to foster lung healing. The ventila-
tory treatment should be reset to take into account that the native lung management
was no longer dictated by the need to maximize gas exchange. High airway pres-
sures and high tidal volume, a must to “optimize” gas exchange through the native
lung, had to be seen as the cause of what later became known as ventilator induced
lung injury (VILI) [7].
The goal of ECMO was redefined from “buying time for the lung to heal” [8] to
“rest the lung” [9]. In the same years, Gattinoni et al. introduced in clinical use a
novel ECLS technique called LFPPV ECCO2R (low-frequency positive pressure
ventilation with extracorporeal CO2 removal) [10]. The stage was set to allow a new
approach to the ventilator management of the patients undergoing ECLS.
ECLS evolved from pure hypoxia rescue therapy to a strategy that, by exploiting
the physiological advantages provided by extracorporeal gas exchange, fosters lung
healing and VILI prevention while avoiding systemic biotrauma and multiple organ
failure [9].
Thus, over the last 30 years, the focus of ECLS treatment of ARDS shifted
toward optimal ventilation management during ECLS. Still, data on how to best
ventilate ARDS patients undergoing ECLS are scanty and mainly come from obser-
vational or physiological studies.

Pathophysiology of Gas Exchange During ECLS

We will discuss almost exclusively the pathophysiology of extracorporeal gas

exchange during veno-venous bypass, though we will also mention some important
aspects of veno-arterial or arterio-venous bypasses.
6 Ventilator Management During ECLS 127

Oxygenation

ECLS was born as ECMO [11], and oxygenation is still the major focus of the tech-
nique. Oxygen transfer (to the extracorporeal blood) for any given membrane lung
will depend on blood flow, on hemoglobin (Hb) concentration, and on input Hb
oxygen saturation. For any given (fixed) intrapulmonary shunt (proportion of the
cardiac output that crosses the lung and leaves it with an unchanged oxygen content
equal to the mixed venous one), veno-venous ECMO will increase the arterial oxy-
gen content according to the change it is able to operate in the mixed venous oxygen
content. If a patient has a 50% intrapulmonary shunt fraction and a mixed venous
oxygen saturation (SvO2) of 50%, then the arterial oxygen saturation (SaO2) will
roughly be 75% (given an arteriovenous difference of 25%). If by oxygenating a
substantial (say 50%) proportion of venous blood ECMO can raise SvO2 to 75%,
then SaO2 will now be the average between 75% and 100%, that is, approximately
87.5% (approximation is due to not taking into account the dissolved O2). It becomes
obvious that the higher the ratio (extracorporeal blood flow/cardiac output), the
higher the effect on the mixed venous and therefore on arterial oxygenation (this is
one of the reasons to advocate β-blocking in certain VV bypasses) [12].
Real life is much more complicated than that, mainly because of the negative
effect of bypass recirculation (which increases with increasing blood flow) and
because intrapulmonary shunt is not fixed, but tends to increase substantially with
the increases in SvO2, in part because of the release of any residual hypoxic
vasoconstriction.
Our ability to provide oxygen to the patient is moreover limited by the fact that
we can add only a limited amount of oxygen to blood, due the concentration of
hemoglobin and to the fact that patients normally have, and probably require, a
mixed venous oxygen saturation around 70%.
A basic principle, described fully in Chap. 1, is that oxygenation is mainly
dependent on blood flow and requires extracorporeal flows very similar to the nor-
mal cardiac output.

Carbon Dioxide

In contrast, CO2 removal requires much lower blood flows [13]. This is due to the
very high carbon dioxide content of blood (mostly carried as bicarbonate ion). We
can expect the carbon dioxide content of normal venous blood to be approximately
600 mL/L and some 50% more in compensated chronic hypercapnia. This means
that one liter of blood contains two to three times the total CO2 produced by the
entire body metabolism in a minute. Therefore, we can assume that total CO2
removal (that is the removal of the minute CO2 production) can be achieved at a
blood flow in the range of 0.5 L/min, as opposed to the 5 L/min required to provide
the 250 mL/min physiological oxygen consumption.
128 A. Pesenti et al.

Fig. 6.1 In spontaneously FiO2

breathing subjects, FiO2 1
has to be increased to PACO2= 35 mmHg
maintain a constant FaO2
when the alveolar gas
equation is faced with RQs
much lower than 1, as it is
the case when the amount
of CO2 removed by the PAO2 300 mmHg
natural lung is decreased 0.5
by the amount removed by PAO2 200 mmHg
the membrane lung.
(Modified from Ref. [13]) PAO2 100 mmHg
AIR

0
0 0.5 1
R

In the second half of the 1970s, the pioneering work of Kolobow and Gattinoni
provided the basic pathophysiological foundations to suggest a possible clinical role
for ECCO2R.
This extraordinary pair of scientist-inventor and scientist-clinician realized that the
membrane oxygenator was indeed a membrane lung, exchanging both oxygen and
CO2. In a little more than a year, they described first a membrane lung optimized for
CO2 removal (the so-called CDML: the carbon dioxide membrane lung) [14], and
then they showed that by removing CO2 at incremental rates, it was possible to control
the ventilation of an awake sheep [15], down to complete apnea [16], and developed a
mode of ventilation that would maintain lung volume while using very low respiratory
rates [17]. They clarified that in apneic oxygenation alveolar PO2 is a function of
alveolar PN2, in turn at equilibrium with the PN2 of the gas ventilating the membrane
lung [16]. Another observation, particularly important in the spontaneously breathing
subject, defines the role of the ratio between the amount of CO2 eliminated and the
amount of oxygen consumed by the natural lung (respiratory quotient, RQ) in deter-
mining alveolar PO2. They stressed the need of increasing FiO2 to maintain a constant
FaO2 when the alveolar gas equation is faced with RQs much lower than 1, as is the
case when the amount of CO2 removed by the natural lung is decreased by the amount
removed by the membrane lung (Fig. 6.1). When no CO2 is exchanged through the
natural lung (apneic oxygenation), then FaO2 is maintained constant only by FiO2 1,
and the role of the membrane lung PN2 becomes predominant [18].

The Acute ARDS Phase: Controlled Mechanical Ventilation

While ECLS has been mainly considered as a rescue treatment for refractory hypox-
emia [11], the considerable amount of CO2 removed by the artificial lung allows
decreasing dramatically the load mechanical ventilation imposes on the natural lung
6 Ventilator Management During ECLS 129

Table 6.1 Guidelines proposed by different ECMO centers to guide ventilation settings in the
early ECMO phase
Tidal volume RR PEEP
Guidelines Ventilation mode (mL/kg) (1/min) (cmH2O) FiO2
ELSO [20] PCV To reach Ppeak of 20 4–5 10 Not
cmH2O reported
CESAR trial PCV To reach Ppeak of 10 10–15 0.3
[21] 20–25 cmH2O
Karolinska PCV or PSV To reach Ppeak of Not 5–10 0.4
[22] 20–25 cmH2O reported
EOLIA trial PCV or VCV assist To reach Pplat <20 15–30 ≥10 0.3–0.6
[23] control cmH2O
RR respiratory rate, PEEP positive end-expiratory pressure, FiO2 inspired O2 fraction, PCV pres-
sure controlled ventilation, PSV pressure support ventilation, VCV volume controlled ventilation,
Pplat airway plateau pressure

and implementing a truly protective ventilatory strategy (based upon decreasing

pressures, tidal volume, and respiratory rate) [19]. These adjustments will lead, in
most cases, to a reduction of mean airway pressure with the potential consequence
of alveolar derecruitment: if deemed necessary, this can be avoided by an appropri-
ate increase in positive end-expiratory pressure (PEEP).
No consensus or formal evidence exists on how to set the ventilator during the
early oxygenation rescue phase of ECMO treatment (Table 6.1) [20–23].

 alking the Tightrope Between Lung Recruitment

W
and Lung Rest

Three major goals are important in supporting the severely hypoxemic ARDS
patient: rescue from hypoxia, lung recruitment (open lung strategy), and VILI pre-
vention (lung rest). During conventional mechanical ventilation, a high price is paid
in terms of airway pressures (PEEP, plateau, and mean airway pressure) to maintain
viable arterial oxygenation, while sufficient alveolar ventilation has to be main-
tained to eliminate CO2. Veno-venous ECMO is capable of providing the entire
oxygen consumption, even when the native lung oxygen transfer is nil (100% intra-
pulmonary shunt); under these conditions, though, blood flow must be maximized,
and relatively low arterial oxygen saturation levels are accepted, sometimes around
80%. While these levels of oxygenation might not be better than the one the patient
could achieve before bypass, the good side of ECMO is that oxygenation can be
maintained while the lung is being managed under “protective” ventilatory settings.
However, a compromise is often recommended (Table 6.1), mostly because some
uncertainties exist as to whether a recruited lung will really heal faster than a col-
lapsed one. Airway pressures are set at compromise values, with the aim of main-
taining a reasonable recruitment status and favoring the oxygenation function of the
native lung, hence maximizing arterial oxygen saturation. A rather different
130 A. Pesenti et al.

approach is taken by some, as exemplified by the guidelines applied by the

Karolinska ECMO center [22].
We suspect a major aspect of this alternative strategy resides in the arterial oxy-
gen saturation one is prepared to accept. In a series of severe H1N1 ARDS patients
undergoing ECMO support, the authors accepted arterial oxygenation values lower
than those observed before ECMO, in exchange for very low airway pressure and
minimally invasive ventilation (pressure support ventilation rather than controlled
mechanical ventilation was used very early in this series) [22].
It is obvious that there is no general consensus on how the ventilator should be
set during ECLS. We have been advocating the use of low-frequency ventilation
(4–6 bpm) with limited plateau pressures since the early 1980s [9], but we have
been unable to set rules or recommendations besides these general indications.
Lately, we have been reporting tidal volumes as low as 1.9 ml/kg [24], while Bein
et al. applied ultra-protective ventilation at 3 mL/kg tidal volume in a series of
ARDS patients [25]. Others have reported the occasional use of high-frequency
ventilation combined with ECMO. Unable to distinguish between cause and effect,
Pham et al. [26] reported that in the first day of ECMO, the highest plateau pressures
were observed in the more severe patients, as judged by their highest mortality rate.
If we come to PEEP settings, no uniform guidelines can be found either. As a
personal rule, at the start of ECMO, we decrease plateau pressures and respiratory
rate, but we adjust PEEP to avoid a sudden drop in mean airway pressure. Many
severe ARDS patients are referred to ECMO when their mean airway pressure is
between 22 and 26 cmH2O and even higher. A sudden decrease to 15 cmH2O will
most probably cause a situation of relative sudden hypervolemia with the risk of
lung flooding. Certainly, when deciding the PEEP level, we select between two
opposites: either to privilege lung volume, avoiding atelectasis, or to minimize alve-
olar strain. Most often, a compromise is reached, this time rather independently
from oxygenation needs, now satisfied by the extracorporeal blood flow.
The lack of consensus on how to set ventilation during ECMO is reflected again
in Table 6.2, which reports ventilator management data during the early ECMO
phase from three different groups [27–29].

Rescue Treatments for Persistent Hypoxemia During ECLS

The term “rescue therapies” encompasses a group of nonconventional treatments

that can be used in ARDS patients with persistent impairment of gas exchange
despite optimization of the mechanical ventilation settings [30–31]. The most com-
monly used rescue therapies are recruitment maneuvers, prone positioning, inhaled
nitric oxide, and high-frequency oscillatory ventilation (HFOV). The possibility of
combining ECMO with one or more of the other rescue treatments has some poten-
tial interest [32]. In the next paragraphs, we will briefly review the mechanism of
action of the different rescue therapies and, when available, the clinical data on their
application in ECMO patients.
6 Ventilator Management During ECLS 131

Table 6.2 Main ventilator settings from three different published case series: consensus during
the early ECMO phase is lacking
Tidal volume RR PEEP
(mL/kg) (1/min) (cmH2O) FiO2
Brogan et al. [27] Before ECLS Not Reported 20 [15;25] 12 [10;17] 1 [1;1]
24 h of ECLS Not Reported 10 [10;15] 10[8;14] 0.5 [0.4;0.5]
Zimmermann et al. [28] Before ECLS 6.6 [5.3;7.2] 25 [22;27] 17 [14;20] 1 [0.8;1]
24 h of ECLS 4.4 [3.4;5.4] 21 [18;26] 17 [14;20] 0.7 [0.6;0.9]
Patroniti et al. [29] Before ECLS 6.2 [4.7;7.7] 28 [20;33] 16 [14;19] 1 [1;1]
24 h of ECLS 4.6 [3.0;6.3] 10 [8;12] 16 [14;19] 0.6 [0.4;0.8]
RR respiratory rate, PEEP positive end-expiratory pressure, FiO2 inspired O2 fraction, ECLS extra-
corporeal lung support

Prone Positioning

Prone positioning was first proposed in 1974 as a form of respiratory physiokinetic

therapy for pediatric patients [33]. Since then, its use has progressively increased,
and it is now recommended as a complement to conventional ventilation in severely
hypoxemic patients [34]. An increase in oxygenation is observed in 60–70% of
patients [35], probably obtained through the combination of different pathophysio-
logic mechanisms: more homogeneous distribution of ventilation, recruitment of
dorsal lung segments, and improved ventilation-perfusion matching [36]. This may
ultimately lead to an optimal recruitment at a given level of PEEP with a reduced
risk of VILI [37].
Practically, prone positioning can be achieved using specialized rotating beds or
simply using the bed sheets and extra nursing personnel. In both cases adequate
padding and skin protection are mandatory to prevent pressure skin breakdown.
Before the PROSEVA trial was published [34], the sustained increase in oxygen-
ation obtained with pronation could not be translated in a significant reduction of
mortality. However, most of these trials were underpowered, included patients of
different severity, and differed in the length of pronation and ventilator settings. A
number of meta-analyses supported prone positioning as an effective treatment in
decreasing mortality in the most severe ARDS patients, while a higher rate of com-
plications is observed in the subgroup of less severe patients [38]. The meta-­analysis
published by Abroug et al. included 1675 patients from seven randomized trials and
investigated the effect of subgroup severity (ALI vs ARDS patients) and of duration
of pronation: the authors showed that prone positioning significantly decreased
mortality only in ARDS patients and suggested that a longer duration of ventilation
in the prone position should be used [39].
The PROSEVA trial finally demonstrated that when applied in severe ARDS
patients for a sufficiently high number of hours per day, prone position effectively
decreases mortality [34].
Prone positioning is associated with well-known complications, such as pressure
sores, loss or obstruction of the endotracheal tube, dislodgement of tubes or
132 A. Pesenti et al.

vascular catheters, and hemodynamic instability. It is easy to understand that these

complications may be particularly important in ECMO patients, in whom compres-
sion or inadvertent removal of the vascular cannulas may lead to reduction or inter-
ruption of the extracorporeal support.
Very few studies provide information on the application of prone positioning
during ECMO. In 2004, Hemmila et al. reported the extensive experience of the
University of Michigan, describing a case series of 255 patients placed on extracor-
poreal support for severe ARDS [40]. Prone positioning for 12–18 h/day was rou-
tinely applied in these patients, but no information is provided on the quality and
number of complications attributable to position changes.
Haefner et al. described a population of 63 pediatric patients (median age
12 months, median weight 9.8 kg) receiving intermittent prone positioning while on
ECMO for respiratory failure [41]. The only complications attributable to pronation
were bleeding from a cannulation site in 18% of patients and chest tube dislodge-
ment in two cases; no unplanned extubations, appliance displacements, cutaneous
ulcerations, or corneal abrasions were observed. It is clear, however, that positional
changes are easier to perform in pediatric patients, thanks to the lower body weight.
Goettler et al. performed a retrospective analysis on ten patients supported with
ECMO and 42 patients on continuous renal replacement therapy (CRRT), to assess
specifically the risk of cannula-related complications [42]: No patients experienced
inadvertent cannula removal during positioning, and only in two cases was extracor-
poreal blood flow reduced; the site of cannula insertion (jugular or femoral) did not
affect the risk of malfunction.
Finally, Litmathe et al. in 2011 reported two cases of morbidly obese ARDS
patients (body mass index of 61 and 51 kg/m2) in whom prone positioning was
applied during ECMO: No major complications were observed, and both patients
had a significant clinical improvement after pronation and were successfully dis-
charged from the ICU [43].
In summary, the rationale of performing a trial of prone positioning during
ECMO is to optimize alveolar recruitment and ventilation-perfusion matching with-
out further increasing airway pressure and possibly reducing the risk of VILI. It
must be remembered that the effect of pronation upon oxygenation is hardly pre-
dictable in the individual patient; moreover, the improvement in oxygenation may
take several hours, so a long duration of prone position seems to be advisable [34].
Based on the available data, prone positioning in ECMO patients seems feasible and
safe, but due to the risk of potentially disastrous complications, this procedure
should probably be performed only in centers with extensive experience in the field.

Inhaled Pulmonary Vasodilators

Intravenous administration of pulmonary vasodilators can lead to systemic hypoten-

sion and increased intrapulmonary shunt due to loss of hypoxic pulmonary vaso-
contriction; on the contrary, inhaled delivery of short-acting pulmonary vasodilators
6 Ventilator Management During ECLS 133

such as nitric oxide (NO) or prostacyclin selectively increases pulmonary blood

flow only to ventilated lung units, thus improving ventilation-perfusion matching
[44]. In patients with a high shunt fraction, this can be associated with a significant
improvement of arterial oxygenation and reduction of pulmonary artery pres-
sure [45].
Inhaled NO (iNO) is the best studied selective pulmonary vasodilator, and sev-
eral trials have demonstrated its efficacy in improving oxygenation and pulmonary
vascular resistances in ARDS patients. These effects are short-lived and do not
improve outcome: a meta-analysis of 12 RCTs including a total of 1237 patients
confirmed that iNO improves oxygenation at 24 h (13% increase in PaO2/FiO2 ratio)
at the cost of an increased risk of renal dysfunction (relative risk 1.5) and no benefit
in survival [46].
To date, no studies have been published focusing on the use of iNO during
ECMO: Ullrich et al. described the use of iNO and ECMO as part of an integrated
approach to ARDS [32], but no information is provided on the combination of these
rescue treatments. However, since iNO can raise PO2 rapidly with negligible acute
toxic effects, a trial of NO inhalation (dose range 5–40 ppm) can be attempted in
patients who remain severely hypoxemic despite ECMO support, especially in the
presence of a concomitant pulmonary artery hypertension.
Inhaled aerosolized prostacyclin (iAP) has effects similar to iNO on the pulmo-
nary vasculature [47], but few studies have investigated its use in ARDS patients
[48–50]. To our knowledge, no data on the use of iAP in ECMO patients have been
reported.

High-Frequency Oscillatory Ventilation (HFOV)

HFOV combines a bias flow and an oscillating piston to deliver tidal volumes lower
than the anatomical dead space (around 1–2 mL/kg) at very high respiratory fre-
quency (usually 3–6 Hz in adults) [51]. The rapid oscillations of gas are delivered
above and below a constant mean airway pressure (Paw), usually set 5 cmH2O
higher than the level achieved during conventional ventilation. Oxygenation during
HFOV depends on the set level of Paw and FiO2, while CO2 removal depends on the
pressure amplitude and frequency of oscillation [51]. Compared to conventional
ventilation, the delivery of small tidal volumes should limit alveolar overdistension,
and the use of higher Paw should promote alveolar recruitment while avoiding
cyclical alveolar collapse, thus improving gas exchange and maintaining the goal of
lung protection.
HFOV is extensively used in neonates with respiratory distress syndrome [52],
while the experience in adult ARDS patients is more limited and more
controversial.
Recently, the results of two large, multicenter, randomized trials comparing
HFOV to conventional ventilation as a first-line ventilator strategy in adult ARDS
patients have been published. The Oscillation for Acute Respiratory Distress
134 A. Pesenti et al.

Syndrome Treated Early (OSCILLATE) trial was stopped after inclusion of 548 of
a planned 1200 patients because the mortality of patients treated with HFOV was
significantly higher than the control group (47% vs 35%); in addition, patients
assigned to HFOV received higher doses of sedatives, neuromuscular blocking
agents, and vasoactive drugs [53]. The Oscillation for ARDS (OSCAR) trial, which
included a total of 795 patients, failed to show any difference in 30-day mortality
between HFOV and conventional ventilation (41.7% vs 41.1%) [54]. Despite some
important methodological differences, the results of these trials argue against a
widespread application of HFOV in ARDS patients. These two studies however
enrolled patients in whom oxygenation was far from the level that would have quali-
fied them for an oxygenation rescue maneuver, and it is doubtful that we should
disqualify at this time HFOV as a possible oxygenation rescue procedure.
Literature on the application of HFOV during ECMO are scarce and limited to
few case reports. Banach et al. described the case of a 35-year-old man with lobar
pneumonia undergoing ECMO for refractory hypoxemia who required discontinu-
ation of the extracorporeal support because of hemorrhagic complications: ECMO
was then substituted with HFOV coupled with pumpless arteriovenous extracorpo-
real lung assist (PECLA) for CO2 removal [55].
Finally, Muellenbach et al. showed in an animal model [56] and in a human case
of post-traumatic ARDS [57] that the combination of an arteriovenous extracorpo-
real lung assist (AV-ECLA) with HFOV allowed the use of oscillatory frequencies
higher than those usually applied (up to 10–15 Hz), thus minimizing the risk of
barotrauma and volutrauma.

Management of Pneumothoraces

Finally, we would like to discuss ventilation during ECMO in special circumstances,

namely, in patients with established barotrauma (pneumothorax or pneumomedias-
tinum, subcutaneous emphysema). In such cases we have learned [23], like others,
that the best way to manage pneumothoraces in ECMO patients is probably to
decrease ventilation further, decrease airway pressure as low as possible, and wait
for the gas to be reabsorbed. When, however, a pneumothorax causes a hemody-
namic impairment, then chest drainage becomes mandatory.

Assisted Ventilation During ECLS

Following the phase of initial ECMO application and when the patient’s conditions
are more stable, mechanical ventilation is switched from controlled to assisted
mode. The aim is then a progressive decrease in support, ECLS discontinuation,
and, finally, patient’s extubation. In particular, early switch to protective-assisted
MV may improve respiratory muscle function and gas exchange, decrease the need
6 Ventilator Management During ECLS 135

for sedation, and aid weaning from ventilator [58]. These changes, though dictated
mainly by the individual patient evolution (some patients stay days on bypass, but a
few stay months), are however sought for at different times by different groups. In
recent years, important technological advances contributed to significantly improve
the safety of ECLS circuits’ use. This led many investigators to a change in strategy,
sometimes liberating the patient first from the ventilator and later from bypass. It has
then become possible to report patients being extubated while still on bypass (awake
ECMO) [59]. We have not adopted this approach but feel safe to recommend initiat-
ing weaning (i.e., assisted breathing rather than controlled) as soon as possible while
assuring patients’ safety and comfort. While most groups delay the application of
assisted ventilation until the native lung performance improves substantially, others
almost immediately switch the ventilator to pressure support mode [60]. Namely,
Karolinska group reported their experience in 17 adult ARDS patients treated by
ECLS between 1995 and 1999. They discontinued muscle relaxation and set MV to
PSV right after starting ECLS. Sedation targets were raised, and within a few hours,
patients were only mildly sedated. They report that patients remained awake during
the day, able to interact with staff and family, and participate in quiet activities like
watching TV. Spontaneous assisted breathing was enhanced and maintained, when
needed, by adding 5% CO2 to the gas ventilating the membrane lung. Assisted
mechanical ventilation mode was pressure support ventilation (PSV) for all patients;
no data are reported to indicate how PSV was titrated during ECLS. We know that
tidal volume was reduced from a range of 450–947 mL before ECLS start to
≈122–662 mL during extracorporeal treatment. Nonetheless, patients’ mortality
(24%) was surprisingly low given their baseline severity. Thus, from this study we
can conclude that in severe ARDS patients ECLS treatment coupled with minimal
sedation and PSV with low tidal volumes may be associated with high survival.
Our group previously showed that PSV may be difficult to implement in ARDS
patients with very low respiratory system compliance (Crs), likely because peak
inspiratory flow is reached rapidly and the flow-based expiratory phase of PSV
starts while patient is still inspiring (premature expiratory cycling) [61]. Thus, these
patients are at high risk of patient-ventilator asynchrony. Asynchrony is a serious
threat, as higher asynchrony is associated with iatrogenic injury and delayed wean-
ing from the ventilator. To this end, we compared PSV with neurally adjusted ven-
tilator assist (NAVA) [62] in a group of severe ARDS patients with Crs ranging
between 7 and 31 mL/cmH2O undergoing veno-venous ECMO and low tidal vol-
ume (i.e., 3–4 mL/kg) PSV. We could show that, in this selected group of patients,
the low Crs values coupled to short inspiratory time led to premature expiratory
cycling and high patient-ventilator asynchrony. NAVA was associated with improved
patient-ventilator interaction compared to PSV, the more so in patients with the low-
est Crs values. More recently, we also evaluated the respiratory pattern of severe
ARDS patients switched to assisted ventilation as per clinical decision when
increasing gas flows were applied (i.e., at increasing rate of extracorporeal CO2
removal) [63]. The respiratory drive and effort were effectively controlled at higher
gas flow rates, suggesting that after switch to assisted ventilation, careful titration of
the extracorporeal CO2 removal rate may be critical for success.
136 A. Pesenti et al.

Other modes of assisted ventilation have been applied during ECMO. Besides
CPAP, we believe we should mention airway pressure release ventilation since this
particular mode of ventilation, alternating two pressure levels at fixed inspiratory/
expiratory time, allows spontaneous breathing at all times during the respiratory
cycle, hence avoiding respiratory muscle atrophy and favoring optimal distribution
of the inspired tidal volume.

Avoiding Intubation by ECMO: Is It Possible?

Institution of ECMO to avoid intubation represents the most advanced frontier for
assisted/spontaneous breathing during ECLS. Reports on this approach in bridge to
lung transplant patients (a few of whom have been affected by severe ARDS) are so
many that it may be considered as standard of care. For example, Fuehner et al.
recently published a retrospective analysis of 26 patients treated with ECLS before
intubation at the onset of end-stage respiratory failure awaiting lung transplant [64].
Of these, 16 patients did not require intubation and were transplanted after a median
time of 11 days on ECMO support. Two of these patients died from septic multior-
gan failure after transplantation. Another patient died from lung cancer 60 days after
transplantation. The remaining 13 patients were discharged from the hospital
20–87 days after transplantation, and all of these patients remained alive during the
follow-up period (7–39 months). These authors also selected a group of patients
who underwent intubation and MV and no ECMO while waiting transplant: of
them, those who survived until hospital discharge required more ventilator days,
spent a longer time in the ICU, and were discharged later from hospital [64]. The
overall survival of the intention-to-treat populations at 6 months after transplant was
62% in the awake ECMO group and 35% in the MV group (P = 0.05). Considering
only patients who reached transplantation, the 6-month post-transplant survival
rates were 80% in the awake ECMO group and 50% in the MV group, respectively
(P = 0.02). Thus, ECMO should be considered before intubation in patients with
end-stage respiratory failure awaiting lung transplant but only in highly specialized
centers with prepared and dedicated staff.
Is awake ECMO approach feasible in severe ARDS patients? Dr. Hoeper and
colleagues reported the results of a single-center, uncontrolled pilot trial designed to
assess the feasibility of veno-venous ECMO in awake, non-intubated, spontane-
ously breathing patients with ARDS [65]. They enrolled six patients, four of whom
were immunocompromised. Three patients were weaned from ECMO without
being intubated and were subsequently discharged from hospital alive, while the
other three required intubation and invasive MV (and two of them died in hospital).
The authors concluded that an “awake ECMO” strategy appears feasible in selected
patients with ARDS and deserves further evaluation as a potential alternative to
intubation and mechanical ventilation. More recently, another clinical study [66]
showed that a strategy of early extubation during ECMO in severe ARDS was asso-
ciated with 50% success (i.e., ECMO weaning without re-intubation). Interestingly,
6 Ventilator Management During ECLS 137

the patients successfully treated by awake ECMO had higher lung weight measured
by CT scan. Subsequent physiological study confirmed the correlation between dif-
ficult control of respiratory drive by ECMO and higher lung weight [67].
Avoiding intubation particularly in specific situations, like ARDS in immuno-
compromised patients, might become a major goal of ECMO. Careful selection of
candidates is necessary to individualize the bypass technique to be applied and the
optimal ventilatory management. The field is very promising, but systematic clini-
cal experience is still lacking.

 onitoring Respiratory Effort in Spontaneously Breathing

M
Patients on ECLS

The physiological response to extracorporeal CO2 removal is a decrease in minute

ventilation, to maintain a constant PaCO2 in spite of a decreased need of CO2
elimination through the native lung [66]. In patients with severe ARDS undergoing
ECMO support, this physiological mechanism to control respiratory drive and
effort appears to be lost, the more so the more severe the condition of the patient.
Failure to control respiratory effort may be a specific mechanism leading to a
failed attempt to switch from controlled to assisted ventilation in patients sup-
ported by ECLS. Higher effort despite significant extra-corporeal CO2 removal
may be a marker of the underlying severity [67] and the main determinant of addi-
tional patient self-inflicted lung injury (P-SILI), either potentially causing failure
of the spontaneous breathing strategy. Respiratory rate is not a reliable index of
effort [68], but is often used as a surrogate because it’s easy to assess: surely,
patients with respiratory rate >30 bpm despite maximal ECMO support should be
carefully evaluated for switch back to sedation and controlled ventilation. Expired
tidal volume >6 ml/kg PBW despite maximal ECMO support and minimal airway
pressure support may indicate lack of relief of respiratory effort and may suggest
ineffective decrease of the risk of P-SILI leading to failure [69]. Airway occlusion
pressure at 0.1 s (P0.1) represents a simple and more accurate indicator of exces-
sive respiratory drive when it exceeds value 4 cmH2O [70]. Most commercially
available ventilators can measure it noninvasively. Esophageal pressure monitor-
ing represents the gold standard for assessing respiratory drive and effort. Negative
deflection during inspiration (ΔPes) of more than 10 cmH2O may increase the risk
of both lung (through high transpulmonary pressure) and diaphragm (through
lesional edema) injury [69]. Finally, the airway pressure negative swing during an
expiratory hold (ΔPocc) and the difference between plateau and peak pressure
measured during an inspiratory pause (PMI) represent validated alternatives to
estimate ΔPes from the standard ventilator waveforms [71, 72]. Assessing respira-
tory drive and effort to fine-tune the interaction between the ECLS and ventilator
support is critical to fully exploit benefits of assisted ventilation (reduced sedation,
increased venous return, muscle activity) and limit risks (lung and diaphragm
injury).
138 A. Pesenti et al.

Disconnection from ECLS

No universally accepted rules exist as to when to disconnect a patient from ECMO

as discussed more fully in Chap. 15. Before disconnecting ECMO, gas exchange
and work of breathing should be acceptable on non-injurious ventilator settings. We
consider very useful the guideline proposed by Palmer: a test of at least 4 h is con-
ducted with the membrane gas flow shut off; ventilator is set on PSV mode with
FiO2 0.5 or less, mean airway pressure between 10 and 15 cmH2O, peak inspiratory
pressure not more than 25 cmH2O, and PEEP between 5 and 10 cmH2O. The patient
passing the test is decannulated under local anesthesia in combination with mild
sedation [60].

 artial Extracorporeal CO2 Removal (PECOR)

P
and Ventilator Management

Modern extracorporeal CO2 removal techniques are characterized by very low

(<500 mL/min), or intermediate blood flow (<1.5 L/min). Clinical experience and
data are very scanty and, at time of writing, totally insufficient to provide any guide-
line or recommendation. The field of extracorporeal CO2 removal still remains an
experimental one. It is important however to review the principles upon which the
physiology of extracorporeal CO2 removal is based, to understand its side effects
and evaluate potential advantages and intrinsic limitations.
As anticipated in the introduction, clinical application of ECCO2R started in
Milan in 1979 [10], following the experimental evidence provided by the work con-
ducted in Kolobow’s laboratory at National Institutes of Health in the previous years
[73]. The technique was in fact a hybrid between ECMO and pure ECCO2R. It was
based on a V-V bypass at 1.5–2.5 l/min, with oxygenation achieved mostly through
the native lung, kept inflated by PEEP with a continuous oxygen flow and a very low
mechanical respiratory rate at 4–6 breaths/min with limited plateau pressure. At that
time [9] we claimed that the “technique seems to overcome the ventilation/perfu-
sion mismatching due to ventilation maldistribution in stiff, non-homogeneous
lungs and prevents the pulmonary barotrauma and extrapulmonary derangements
caused by conventional mechanical ventilation.” Technology was still in its infancy,
but the targets were already well defined.
The first real low flow partial CO2 removal technique was applied in a patient
with high flow bilateral airleaks following bilateral pleurectomy and undergoing
controlled mechanical ventilation [74]. Thanks to the decreased ventilator needs
afforded by ECCO2R, in a few days the patient could be switched to CPAP and
finally extubated with complete resolution of the air leak. In this case cannulation
was percutaneous, blood flow ranged between 0.4 and 0.6 l/min, and CO2 removal
between 22% and 40% of the total patient VCO2 (Fig. 6.2).
6 Ventilator Management During ECLS 139

Ventilatory
PS CPAP T-tube CPAP
Mode (+22 cm H2O)(3 cm H2O) (3 cm H2O)

FIO2 100
(%)
0

100

PaO2 80
(mmHg)
60

PaCO2 40
(mmHg)
20

PECOR 0
(%) 0 1 2 3 4 5 6 7 PECOR days

Fig. 6.2 Data from the first partial extracorporeal low-flow partial CO2 removal (PECOR) clinical
case: a patient with high flow bilateral airleaks, after institution of PECOR, could promptly be
switched from high-level PSV to CPAP and finally extubated with complete resolution of the air
leak. In this case cannulation was percutaneous, blood flow ranged between 0.4 and 0.6 l/min, and
CO2 removal between 22% and 40% and the total patient VCO2. (Modified from Ref. [74])

As already stated, Kolobow and Gattinoni showed how ECCO2R could control
spontaneous breathing and the need for mechanical ventilation in sheep [15, 18].
When we faced the first ARDS patients recovering from severe ARDS on ECLS, we
could show that ARDS patients undergoing CPAP ventilation while on bypass could
decrease their ventilation according to the amount of CO2 removed by the mem-
brane lung [75]. The same effect has been recently observed in ARDS patients being
ventilated using NAVA while on ECMO [76].
These data however, while supporting the possibility of targeting ventilatory
needs and respiratory drive in ARDS patients, should be interpreted with caution.
Experience shows that in the early acute interstitial edema phase, ARDS patients
have a very high respiratory drive and are severely dyspneic and tachypneic. In the
early phase of severe ARDS, blood gases represent a minor portion of the respira-
tory drive. In the clinical setting of ECMO in severe ARDS patients, it is not uncom-
mon to verify that even when PaCO2 is brought down to very low levels
(25–28 mmHg), the respiratory rate is still very high, and the inspiratory effort may
still generate very deep negative intrathoracic pressures and very high transpulmo-
nary pressure, carrying a high risk of barotrauma and VILI in spite of ECCO2R
(Fig. 6.3). These observations contrast with what has been reported in acutely
decompensated COPD patients, in whom respiratory drive (and the need for intuba-
tion) can be effectively controlled by ECCO2R [77–78].
Which is then the possible role of ECCO2R in ARDS patients? Most important at
this time is the ultra-protective ventilator approach proposed by Terragni et al. in
2009 [79]. This group of investigators showed that 25–30% of ARDS patients,
140 A. Pesenti et al.

Fig. 6.3 Airway pressure (Paw), esophageal pressure (Pes), and electrical activity of the dia-
phragm (EAdi) tracings from one representative patient undergoing ECMO and pressure support
ventilation: ECMO sweep gas flow (GF) was gradually decreased from 4 L/min (100%) to 2 L/min
to 0 L/min (0%). Note the increasing esophageal pressure swings at GF 0%. (Bellani, Grasselli,
Mauri, Pesenti: unpublished data)

though ventilated according to the NIH lung-protective recommendations, showed

evidence of severe hyperinflation on CT scan [80]. They reasoned that this unex-
pected VILI risk could be lowered by decreasing the tidal volume to 4 mL/kg from
6 mL/kg while maintaining an almost normal arterial pH and PCO2 by a low flow
(250–400 mL/min) bypass applied through a modified hemofiltration machine com-
prising a membrane lung. Their study was not designed to provide outcome data,
but it provided enough physiological and biochemical evidence to support the fea-
sibility of such an approach. Two separate case reports have been published to date
suggesting the possibility of using ultraprotective ventilation (down to 2 ml/kg) to
achieve survival in ARDS patients undergoing extracorporeal gas exchange ([23,
81] case report 2). The Xtravent study reported the effect of applying 3 mL/kg tidal
volume with ECCO2R in a group of severe ARDS patients [25]. The study was ran-
domized and controlled, with a primary outcome of mortality. The study was
stopped because of slow enrollment and failed to demonstrate improved mortality
but showed feasibility, reduced sedative use, and lower serum inflammatory mark-
ers. A post hoc analysis suggested earlier weaning from MV in the most hypoxemic
subgroup. The largest trial was recently published by the REST investigators,
enrolling 412 adults with hypoxemic respiratory failure who were randomized to
conventional lung-protective ventilation (6 mL/kg PBW was recommended) or
ECCO2R and reduced tidal volume (aiming for less than or equal to 3 mL/kg PBW)
[82]. The primary outcome, all-cause mortality at 90 days, was not different.
Moreover, there were significantly fewer ventilator-free days in the ECCO2R group
and more serious adverse events.
We could foresee the use of minimally invasive, low-flow ECCO2R devices to be
applied in the early phases of ARDS, either to avoid or delay intubation or to extend
the range of application of NIV or CPAP. It is entirely possible that, even if muscle
paralysis and controlled mechanical ventilation will certainly maintain a role in
ARDS management, a judicious application of ECCO2R may lead to early assisted
breathing, early weaning, and extubation, perhaps with an extended use of NIV. The
possibility of less sedation, shorter or no intubation, and improved contact and com-
munication with the surroundings is certainly very appealing. Decreased prevalence
6 Ventilator Management During ECLS 141

and duration of intubation might also lower the incidence of ventilator-associated

pneumonia. The most promising applications of low-flow ECCO2R appear to be
related however to the concept of ultraprotective ventilation in ARDS [75] and to
the treatment of COPD patients [83].

Conclusions

The concept of VILI and the need for protective ventilation have gained widespread
acceptance, having been clearly validated in the clinical field. The practical applica-
tion of these concepts in the patient undergoing ECLS, though based on solid patho-
physiological grounds, lacks evidence to allow the drawing of guidelines or
recommendations. Clinical experience suggests two alternative approaches, possi-
bly applicable at different times in the same patient: One looks for optimized lung
recruitment by a higher PEEP level, and the other privileges lung rest in spite of the
risk of substantial lung collapse. Both approaches require a third essential element:
time and with it the patience necessary for the lung to recover.

References

1. Zapol WM, Snider MT, Hill JD, Fallat RJ, Bartlett RH, Edmunds LH, Morris AH, Peirce
EC 2nd, Thomas AN, Proctor HJ, Drinker PA, Pratt PC, Bagniewski A, Miller RG Jr.
Extracorporeal membrane oxygenation in severe acute respiratory failure. A randomized pro-
spective study. JAMA. 1979;242(20):2193–6.
2. National Heart, Lung, and Blood Institute: Division of Lung Disease. Section four: hemo-
dynamics, gas exchange and mechanics. In: Extracorporeal support for respiratory insuffi-
ciency. A collaborative study in response to RFP-NHLI-73-20. Ed. US Department of Health,
Education and Welfare. Washington, DC, 1979.
3. Pontoppidan H, Geffin B, Lowenstein E. Acute respiratory failure in the adult. 3. N Engl J
Med. 1972;287(16):799–806.
4. Suter PM, Fairley HB, Isenberg MD. Effect of tidal volume and positive end-expiratory pres-
sure on compliance during mechanical ventilation. Chest. 1978;73(2):158–62.
5. Kolobow T, Solca M, Gattinoni L, Pesenti A. Adult respiratory distress syndrome (ARDS):
why did ECMO fail? Int J Artif Organs. 1981;4(2):58–9.
6. Kolobow T, Bowman RL. Construction and evaluation of an alveolar membrane artificial
heart-lung. Trans Am Soc Artif Intern Organs. 1963;9:238–43.
7. Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies. Am
J Respir Crit Care Med. 1998;157(1):294–323.
8. Zapol WM, Kitz RJ. Buying time with artificial lungs. N Engl J Med. 1972;286(12):657–8.
9. Gattinoni L, Agostoni A, Pesenti A, Pelizzola A, Rossi GP, Langer M, Vesconi S, Uziel L, Fox
U, Longoni F, Kolobow T, Damia G. Treatment of acute respiratory failure with low-frequency
positive-pressure ventilation and extracorporeal removal of CO2. Lancet. 1980;2(8189):292–4.
10. Gattinoni L, Kolobow T, Agostoni A, Damia G, Pelizzola A, Rossi GP, Langer M, Solca M,
Citterio R, Pesenti A, Fox U, Uziel L. Clinical application of low frequency positive pressure
ventilation with extracorporeal CO2 removal (LFPPV-ECCO2R) in treatment of adult respira-
tory distress syndrome (ARDS). Int J Artif Organs. 1979;2(6):282–3.
142 A. Pesenti et al.

11. Hill JD, O'Brien TG, Murray JJ, Dontigny L, Bramson ML, Osborn JJ, Gerbode F. Prolonged
extracorporeal oxygenation for acute post-traumatic respiratory failure (shock-lung syn-
drome). Use of the Bramson membrane lung. N Engl J Med. 1972;286(12):629–34.
12. Guarracino F, Zangrillo A, Ruggeri L, Pieri M, Calabrò MG, Landoni G, Stefani M, Doroni L,
Pappalardo F. β-Blockers to optimize peripheral oxygenation during extracorporeal membrane
oxygenation: a case series. J Cardiothorac Vasc Anesth. 2012;26(1):58–63.
13. Gattinoni L, Pesenti A, Kolobow T, Damia G. A new look at therapy of the adult respiratory
distress syndrome: motionless lungs. Int Anesthesiol Clin. 1983;21(2):97–117.
14. Kolobow T, Gattinoni L, Tomlinson T, White D, Pierce J, Iapichino G. The carbon dioxide
membrane lung (CDML): a new concept. Trans Am Soc Artif Intern Organs. 1977;23:17–21.
15. Kolobow T, Gattinoni L, Tomlinson TA, Pierce JE. Control of breathing using an extracorpo-
real membrane lung. Anesthesiology. 1977;46(2):138–41.
16. Kolobow T, Gattinoni L, Tomlinson T, Pierce JE. An alternative to breathing. J Thorac
Cardiovasc Surg. 1978;75(2):261–6.
17. Gattinoni L, Kolobow T, Tomlinson T, Iapichino G, Samaja M, White D, Pierce J. Low-­
frequency positive pressure ventilation with extracorporeal carbon dioxide removal (LFPPV-­
ECCO2R): an experimental study. Anesth Analg. 1978;57(4):470–7.
18. Gattinoni L, Kolobow T, Tomlinson T, White D, Pierce J. Control of intermittent posi-
tive pressure breathing (IPPB) by extracorporeal removal of carbon dioxide. Br J Anaesth.
1978;50(8):753–8.
19. Brodie D, Bacchetta M. Extracorporeal membrane oxygenation for ARDS in adults. N Engl J
Med. 2011;365(20):1905–14.
20. ELSO Patient Specific guidelines. Available at http://www.elsonet.org/index.
php?option=com_phocadownload&view=category&id=4&Itemid=627.
21. Peek GJ, Mugford M, Tiruvoipati R, Wilson A, Allen E, Thalanany MM, Hibbert CL, Truesdale
A, Clemens F, Cooper N, Firmin RK, Elbourne D, CESAR trial collaboration. Efficacy and
economic assessment of conventional ventilator support versus extracorporeal membrane oxy-
genation for severe adult respiratory failure (CESAR): a multicentre randomised controlled
trial. Lancet. 2009;374(9698):1351–63.
22. Holzgraefe B, Broomé M, Kalzén H, Konrad D, Palmér K, Frenckner B. Extracorporeal
membrane oxygenation for pandemic H1N1 2009 respiratory failure. Minerva Anestesiol.
2010;76(12):1043–51.
23. Combes A, Hajage D, Capellier G, Demoule A, Lavoué S, Guervilly C, Da Silva D, Zafrani L,
Tirot P, Veber B, Maury E, Levy B, Cohen Y, Richard C, Kalfon P, Bouadma L, Mehdaoui H,
Beduneau G, Lebreton G, Brochard L, Ferguson ND, Fan E, Slutsky AS, Brodie D, Mercat A,
EOLIA Trial Group, REVA, and ECMONet. Extracorporeal membrane oxygenation for severe
acute respiratory distress syndrome. N Engl J Med. 2018;378(21):1965–75.
24. Mauri T, Bellani G, Foti G, Grasselli G, Pesenti A. Successful use of neurally adjusted ven-
tilatory assist in a patient with extremely low respiratory system compliance undergoing
ECMO. Intensive Care Med. 2011;37(1):166–7.
25. Bein T, Weber-Carstens S, Goldmann A, Müller T, Staudinger T, Brederlau J, Muellenbach R,
Dembinski R, Graf BM, Wewalka M, Philipp A, Wernecke KD, Lubnow M, Slutsky AS. Lower
tidal volume strategy (≈3 ml/kg) combined with extracorporeal CO2 removal versus ‘conven-
tional’ protective ventilation (6 ml/kg) in severe ARDS: the prospective randomized Xtravent-­
study. Intensive Care Med. 2013;39(5):847–56.
26. Pham T, Combes A, Roze H, Chevret S, Mercat A, Roch A, Mourvillier B, Ara-Somohano C,
Bastien O, Zogheib E, et al. Extracorporeal membrane oxygenation for pandemic influenza
A(H1N1)-induced acute respiratory distress syndrome: a cohort study and propensity matched
analysis. Am J Respir Crit Care Med. 2013;187(3):276–85.
27. Brogan TV, Thiagarajan RR, Rycus PT, Bartlett RH, Bratton SL. Extracorporeal membrane
oxygenation in adults with severe respiratory failure: a multi-center database. Intensive Care
Med. 2009;35(12):2105–14.
6 Ventilator Management During ECLS 143

28. Zimmermann M, Bein T, Arlt M, Philipp A, Rupprecht L, Mueller T, Lubnow M, Graf BM,
Schlitt HJ. Pumpless extracorporeal interventional lung assist in patients with acute respiratory
distress syndrome: a prospective pilot study. Crit Care. 2009;13(1):R10.
29. Patroniti N, Zangrillo A, Pappalardo F, Peris A, Cianchi G, Braschi A, Iotti GA, Arcadipane
A, Panarello G, Ranieri VM, Terragni P, Antonelli M, Gattinoni L, Oleari F, Pesenti A. The
Italian ECMO network experience during the 2009 influenza A(H1N1) pandemic: preparation
for severe respiratory emergency outbreaks. Intensive Care Med. 2011;37(9):1447–57.
30. Pipeling MR, Fan E. Therapies for refractory hypoxemia in acute respiratory distress syn-
drome. JAMA. 2010;304(22):2521–7.
31. Patroniti N, Isgrò S, Zanella A. Clinical management of severely hypoxemic patients. Curr
Opin Crit Care. 2011;17(1):50–6.
32. Ullrich R, Lorber C, Röder G, Urak G, Faryniak B, Sladen RN, Germann P. Controlled air-
way pressure therapy, nitric oxide inhalation, prone position, and extracorporeal membrane
oxygenation (ECMO) as components of an integrated approach to ARDS. Anesthesiology.
1999;91(6):1577–86.
33. Bryan AC. Conference on the scientific basis of respiratory therapy. Pulmonary physiotherapy
in the pediatric age group. Comments of a devil’s advocate. Am Rev Respir Dis. 1974;110(6
Pt 2):143–4.
34. Guérin C, Reignier J, Richard JC, Beuret P, Gacouin A, Boulain T, Mercier E, Badet M, Mercat
A, Baudin O, Clavel M, Chatellier D, Jaber S, Rosselli S, Mancebo J, Sirodot M, Hilbert G,
Bengler C, Richecoeur J, Gainnier M, Bayle F, Bourdin G, Leray V, Girard R, Baboi L, Ayzac
L, PROSEVA Study Group. Prone positioning in severe acute respiratory distress syndrome. N
Engl J Med. 2013;368(23):2159–68.
35. Hemmila MR, Napolitano LM. Severe respiratory failure: advanced treatment options. Crit
Care Med. 2006;34(9 Suppl):S278–90.
36. Pelosi P, Brazzi L, Gattinoni L. Prone position in acute respiratory distress syndrome. Eur
Respir J. 2002;20(4):1017–28.
37. Galiatsou E, Kostanti E, Svarna E, Kitsakos A, Koulouras V, Efremidis SC, Nakos G. Prone
position augments recruitment and prevents alveolar overinflation in acute lung injury. Am J
Respir Crit Care Med. 2006;174(2):187–97.
38. Sud S, Friedrich JO, Taccone P, Polli F, Adhikari NK, Latini R, Pesenti A, Guérin C, Mancebo
J, Curley MA, Fernandez R, Chan MC, Beuret P, Voggenreiter G, Sud M, Tognoni G, Gattinoni
L. Prone ventilation reduces mortality in patients with acute respiratory failure and severe
hypoxemia: systematic review and meta-analysis. Intensive Care Med. 2010;36(4):585–99.
39. Abroug F, Ouanes-Besbes L, Dachraoui F, Ouanes I, Brochard L. An updated study-level
meta-analysis of randomized controlled trials on proning in ARDS and acute lung injury. Crit
Care. 2011;15(1):R6.
40. Hemmila MR, Rowe SA, Boules TN, Miskulin J, McGillicuddy JW, Schuerer DJ, Haft JW,
Swaniker F, Arbabi S, Hirschl RB, Bartlett RH. Extracorporeal life support for severe acute
respiratory distress syndrome in adults. Ann Surg. 2004;240(4):595–607.
41. Haefner SM, Bratton SL, Annich GM, Bartlett RH, Custer JR. Complications of intermittent
prone positioning in pediatric patients receiving extracorporeal membrane oxygenation for
respiratory failure. Chest. 2003;123(5):1589–94.
42. Goettler CE, Pryor JP, Hoey BA, Phillips JK, Balas MC, Shapiro MB. Prone positioning does
not affect cannula function during extracorporeal membrane oxygenation or continuous renal
replacement therapy. Crit Care. 2002;6(5):452–5.
43. Litmathe J, Sucker C, Easo J, Wigger L, Dapunt O. Prone and ECMO – a contradiction per se?
Perfusion. 2012;27(1):78–82.
44. Griffiths MJ, Evans TW. Inhaled nitric oxide therapy in adults. N Engl J Med.
2005;353(25):2683–95.
45. Rossaint R, Falke KJ, López F, Slama K, Pison U, Zapol WM. Inhaled nitric oxide for the adult
respiratory distress syndrome. N Engl J Med. 1993;328(6):399–405.
144 A. Pesenti et al.

46. Adhikari NK, Burns KE, Friedrich JO, Granton JT, Cook DJ, Meade MO. Effect of nitric
oxide on oxygenation and mortality in acute lung injury: systematic review and meta-analysis.
BMJ. 2007;334(7597):779.
47. Haraldsson A, Kieler-Jensen N, Nathorst-Westfelt U, Bergh CH, Ricksten SE. Comparison of
inhaled nitric oxide and inhaled aerosolized prostacyclin in the evaluation of heart transplant
candidates with elevated pulmonary vascular resistance. Chest. 1998;114(3):780–6.
48. Walmrath D, Schneider T, Schermuly R, Olschewski H, Grimminger F, Seeger W. Direct com-
parison of inhaled nitric oxide and aerosolized prostacyclin in acute respiratory distress syn-
drome. Am J Respir Crit Care Med. 1996;153(3):991–6.
49. Walmrath D, Schneider T, Pilch J, Grimminger F, Seeger W. Aerosolised prostacyclin in adult
respiratory distress syndrome. Lancet. 1993;342(8877):961–2.
50. van Heerden PV, Barden A, Michalopoulos N, Bulsara MK, Roberts BL. Dose-response to
inhaled aerosolized prostacyclin for hypoxemia due to ARDS. Chest. 2000;117(3):819–27.
51. Hager DN. High-frequency oscillatory ventilation in adults with acute respiratory distress syn-
drome. Curr Opin Anaesthesiol. 2012;25(1):17–23.
52. Courtney SE, Durand DJ, Asselin JM, Hudak ML, Aschner JL, Shoemaker CT, Neonatal
Ventilation Study Group. High-frequency oscillatory ventilation versus conventional mechani-
cal ventilation for very-low-birth-weight infants. N Engl J Med. 2002;347(9):643–52.
53. Ferguson ND, Cook DJ, Guyatt GH, Mehta S, Hand L, Austin P, Zhou Q, Matte A, Walter
SD, Lamontagne F, Granton JT, Arabi YM, Arroliga AC, Stewart TE, Slutsky AS, Meade MO,
OSCILLATE Trial Investigators; Canadian Critical Care Trials Group. High-frequency oscil-
lation in early acute respiratory distress syndrome. N Engl J Med. 2013;368(9):795–805.
54. Young D, Lamb SE, Shah S, MacKenzie I, Tunnicliffe W, Lall R, Rowan K, Cuthbertson BH,
OSCAR Study Group. High-frequency oscillation for acute respiratory distress syndrome. N
Engl J Med. 2013;368(9):806–13.
55. Banach M, Soukup J, Bucher M, Andres J. High frequency oscillation, extracorporeal mem-
brane oxygenation and pumpless arteriovenous lung assist in the management of severe
ARDS. Anestezjol Intens Ter. 2010;42(4):201–5.
56. Muellenbach RM, Kuestermann J, Kredel M, Johannes A, Wolfsteiner U, Schuster F, Wunder
C, Kranke P, Roewer N, Brederlau J. Arteriovenous extracorporeal lung assist allows for
maximization of oscillatory frequencies: a large-animal model of respiratory distress. BMC
Anesthesiol. 2008;8:7.
57. Muellenbach RM, Wunder C, Nuechter DC, Smul T, Trautner H, Kredel M, Roewer N,
Brederlau J. Early treatment with arteriovenous extracorporeal lung assist and high-­frequency
oscillatory ventilation in a case of severe acute respiratory distress syndrome. Acta Anaesthesiol
Scand. 2007;51(6):766–9.
58. Putensen C, Muders T, Varelmann D, Wrigge H. The impact of spontaneous breathing during
mechanical ventilation. Curr Opin Crit Care. 2006;12(1):13–8.
59. Xia J, Gu S, Li M, Liu D, Huang X, Yi L, Wu L, et al. Spontaneous breathing in patients with
severe acute respiratory distress syndrome receiving prolonged extracorporeal membrane oxy-
genation. BMC Pulm Med. 2019;19(1):237.
60. Lindén V, Palmér K, Reinhard J, Westman R, Ehrén H, Granholm T, Frenckner B. High sur-
vival in adult patients with acute respiratory distress syndrome treated by extracorporeal mem-
brane oxygenation, minimal sedation, and pressure supported ventilation. Intensive Care Med.
2000;26(11):1630–7.
61. Cereda M, Foti G, Marcora B, Gili M, Giacomini M, Sparacino ME, Pesenti A. Pressure sup-
port ventilation in patients with acute lung injury. Crit Care Med. 2000;28(5):1269–75.
62. Mauri T, Bellani G, Grasselli G, Confalonieri A, Rona R, Patroniti N, Pesenti A. Patient-­
ventilator interaction in ARDS patients with extremely low compliance undergoing ECMO: a
novel approach based on diaphragm electrical activity. Intensive Care Med. 2013;39(2):282–91.
63. Mauri T, Grasselli G, Suriano G, Eronia N, Spadaro S, Turrini C, Patroniti N, Bellani G,
Pesenti A. Control of respiratory drive and effort in extracorporeal membrane oxygen-
ation patients recovering from severe acute respiratory distress syndrome. Anesthesiology.
2016;125(1):159–67.
6 Ventilator Management During ECLS 145

64. Fuehner T, Kuehn C, Hadem J, Wiesner O, Gottlieb J, Tudorache I, Olsson KM, Greer M,
Sommer W, Welte T, Haverich A, Hoeper MM, Warnecke G. Extracorporeal membrane oxy-
genation in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med.
2012;185(7):763–8.
65. Hoeper MM, Wiesner O, Hadem J, Wahl O, Suhling H, Duesberg C, Sommer W, Warnecke
G, Greer M, Boenisch O, Busch M, Kielstein JT, Schneider A, Haverich A, Welte T, Kühn
C. Extracorporeal membrane oxygenation instead of invasive mechanical ventilation in
patients with acute respiratory distress syndrome. Intensive Care Med. 2013;39(11):2056–7.
66. Crotti S, Bottino N, Ruggeri GM, Spinelli E, Tubiolo D, Lissoni A, Protti A, Gattinoni
L. Spontaneous breathing during extracorporeal membrane oxygenation in acute respiratory
failure. Anesthesiology. 2017;126(4):678–87.
67. Spinelli E, Mauri T, Lissoni A, Crotti S, Langer T, Albanese M, Volta CA, Fornari C, Tagliabue
P, Grasselli G, Pesenti A. Spontaneous Breathing Patterns During Maximum Extracorporeal
CO2 Removal in Subjects With Early Severe ARDS. Respir Care. 2020;65(7):911–9.
68. Akoumianaki E, Vaporidi K, Georgopoulos D. The Injurious Effects of Elevated or
Nonelevated Respiratory Rate during Mechanical Ventilation. Am J Respir Crit Care Med.
2019;199:149–57.
69. Goligher EC, Jonkman AH, Dianti J, Vaporidi K, Beitler JR, Patel BK, Yoshida T, Jaber S,
Dres M, Mauri T, Bellani G, Demoule A, Brochard L, Heunks L. Clinical strategies for imple-
menting lung and diaphragm-protective ventilation: avoiding insufficient and excessive effort.
Intensive Care Med. 2020;46(12):2314–26. https://doi.org/10.1007/s00134-­020-­06288-­9.
70. Telias I, Junhasavasdikul D, Rittayamai N, Piquilloud L, Chen L, Ferguson ND, Goligher EC,
Brochard L. Airway Occlusion Pressure As an Estimate of Respiratory Drive and Inspiratory
Effort during Assisted Ventilation. Am J Respir Crit Care Med. 2020;201(9):1086–98.
71. Bertoni M, Telias I, Urner M, Long M, Del Sorbo L, Fan E, Sinderby C, Beck J, Liu L, Qiu H,
Wong J, Slutsky AS, Ferguson ND, Brochard LJ, Goligher EC. A novel non-invasive method
to detect excessively high respiratory effort and dynamic transpulmonary driving pressure dur-
ing mechanical ventilation. Crit Care. 2019;23(1):346.
72. Foti G, Cereda M, Banfi G, Pelosi P, Fumagalli R, Pesenti A. End-inspiratory airway occlusion:
a method to assess the pressure developed by inspiratory muscles in patients with acute lung
injury undergoing pressure support. Am J Respir Crit Care Med. 1997;156(4 Pt 1):1210–6.
73. Gattinoni L, Kolobow T, Damia G, Agostoni A, Pesenti A. Extracorporeal carbon dioxide
removal (ECCO2R): a new form of respiratory assistance. Int J Artif Organs. 1979;2(4):183–5.
74. Pesenti A, Rossi GP, Pelosi P, Brazzi L, Gattinoni L. Percutaneous extracorporeal CO2 removal
in a patient with bullous emphysema with recurrent bilateral pneumothoraces and respiratory
failure. Anesthesiology. 1990;72(3):571–3.
75. Marcolin R, Mascheroni D, Pesenti A, Bombino M, Gattinoni L. Ventilatory impact of partial
extracorporeal CO2 removal (PECOR) in ARF patients. ASAIO Trans. 1986;32(1):508–10.
76. Karagiannidis C, Lubnow M, Philipp A, Riegger GA, Schmid C, Pfeifer M, Mueller
T. Autoregulation of ventilation with neurally adjusted ventilatory assist on extracorporeal
lung support. Intensive Care Med. 2010;36(12):2038–44.
77. Burki NK, Mani RK, Herth FJ, Schmidt W, Teschler H, Bonin F, Becker H, Randerath WJ,
Stieglitz S, Hagmeyer L, Priegnitz C, Pfeifer M, Blaas SH, Putensen C, Theuerkauf N, Quintel
M, Moerer O. A novel extracorporeal CO2 removal system: results of a pilot study of hyper-
capnic respiratory failure in patients with COPD. Chest. 2013;143(3):678–86.
78. Cardenas VJ Jr, Lynch JE, Ates R, Miller L, Zwischenberger JB. Venovenous carbon diox-
ide removal in chronic obstructive pulmonary disease: experience in one patient. ASAIO
J. 2009;55(4):420–2.
79. Terragni PP, Del Sorbo L, Mascia L, Urbino R, Martin EL, Birocco A, Faggiano C, Quintel M,
Gattinoni L, Ranieri VM. Tidal volume lower than 6 ml/kg enhances lung protection: role of
extracorporeal carbon dioxide removal. Anesthesiology. 2009;111(4):826–35.
80. Terragni PP, Rosboch G, Tealdi A, Corno E, Menaldo E, Davini O, Gandini G, Herrmann P,
Mascia L, Quintel M, Slutsky AS, Gattinoni L, Ranieri VM. Tidal hyperinflation during low
146 A. Pesenti et al.

tidal volume ventilation in acute respiratory distress syndrome. Am J Respir Crit Care Med.
2007;175(2):160–6.
81. Bein T, Osborn E, Hofmann HS, Zimmermann M, Philipp A, Schlitt HJ, Graf BM. Successful
treatment of a severely injured soldier from Afghanistan with pumpless extracorporeal lung
assist and neurally adjusted ventilatory support. Int J Emerg Med. 2010;3(3):177–9.
82. McNamee JJ, Gillies MA, Barrett NA, Perkins GD, Tunnicliffe W, Young D, et al. Effect of
lower tidal volume ventilation facilitated by extracorporeal carbon dioxide removal vs stan-
dard care ventilation on 90-day mortality in patients with acute hypoxemic respiratory failure:
the REST randomized clinical trial. JAMA. 2021;326(11):1013–23.
83. Kluge S, Braune SA, Engel M, Nierhaus A, Frings D, Ebelt H, Uhrig A, Metschke M,
Wegscheider K, Suttorp N, Rousseau S. Avoiding invasive mechanical ventilation by extracor-
poreal carbon dioxide removal in patients failing noninvasive ventilation. Intensive Care Med.
2012;38(10):1632–9.
Chapter 7
Managing the Systemic Circulation:
Volume Status and RV Function

Sundar Krishnan and Gregory A. Schmidt

In patients with acute, severe respiratory failure, the pulmonary parenchymal dis-
ease process is often associated with severe pulmonary arterial hypertension and
right ventricular (RV) dysfunction, also called acute cor pulmonale. Intensive care
therapies such as fluid management, mechanical ventilation, and vasoactive infu-
sions can also exacerbate RV dysfunction and shock. In such patients, monitoring
and management of hemodynamic parameters can then be further complicated
when extracorporeal support is needed. It is hence essential for the intensivist to
develop familiarity with the unique considerations of fluid management and RV
dysfunction in patients on extracorporeal membrane oxygenation (ECMO) [1].
In this chapter, we will review the validity of various hemodynamic parameters
in assessing volume status and RV function in patients with severe respiratory fail-
ure on veno-venous extracorporeal membrane oxygenation (V-V ECMO) support.
We then discuss the nuances of fluid management in these patients. The next section
describes the challenges with monitoring RV function and cardiac output in patients
on extracorporeal support and the principles for RV support in this patient popula-
tion. Throughout the chapter, we lay a strong emphasis on the role of point-of-care
echocardiography in guiding bedside management. The chapter focuses primarily
on the care of patients on V-V ECMO. Where relevant, some additional points are
made regarding veno-arterial extracorporeal membrane oxygenation (V-A ECMO)
patients.

S. Krishnan (*)
Department of Anesthesiology, Duke University School of Medicine, Duke University
Medical Center, Durham, NC, USA
e-mail: [email protected]
G. A. Schmidt
Division of Pulmonary Diseases, Critical Care, and Occupational Medicine, Department of
Internal Medicine, University of Iowa, Iowa City, IA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 147

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_7
148 S. Krishnan and G. A. Schmidt

Volume Status

As with all critically ill patients, adequate intravascular volume is required for
maintaining native cardiac output for organ perfusion. Maintaining adequate intra-
vascular volume is essential for one additional reason in patients on extracorporeal
support. Hypovolemia would lead to reduced extracorporeal support and crucially
in patients on V-V ECMO, reduced extracorporeal gas exchange, and hence, hypox-
emia. At the same time, positive fluid balance has been associated with worse out-
comes in critical illness. Most patients at this stage of disease would benefit from
fluid removal for relief of pulmonary and systemic edema.
Hence, the goal for volume status is dual – an adequate preload is required to
optimize cardiac output as well as ECMO flow, while also avoiding fluid overload.
Often, the goal of maintaining ECMO flow can be the driver for volume
resuscitation.

Monitoring

Circuit Negative Pressure and Cannula Chatter An early sign of reduced venous
return is a worsening negative pressure on the venous drainage cannula, as moni-
tored by the ECMO circuit. At typical flows (2–4 lpm) with typical sized cannulae
in normal sized adults, negative pressure readings lower than −80 to −100 mm Hg
usually reflect hypovolemia. In patients with severe hypovolemia, the ECMO drain-
age cannulae will often start vibrating (colloquially called “chatter”) as the pump
attempts to drain volume and the inferior vena cava (IVC) collapses around the can-
nula. As with negative pressure values, chatter is dependent on the pump settings
and is more common when high ECMO flows are attempted in conditions with
reduced venous return. At lower pump settings, the cannula chatter might be less
obvious.

CVP In critically ill patients, it has been well established that static parameters of
cardiac filling, including central venous pressure and pulmonary wedge pressure,
are inadequate markers of preload. In patients with severe ARDS on mechanical
support, despite intravascular hypovolemia, CVP can be elevated due to RV dys-
function or elevated pleural pressure. Additionally, the position of the central venous
catheter vis-a-vis the ECMO cannulae can further impact the CVP value in either
direction.

PPV Pulse pressure variation (PPV) is an accurate predictor of fluid responsive-

ness under controlled circumstances. However, in patients with ARDS on ECMO,
the accuracy of PPV in predicting cardiac output response to fluid administration is
affected by the low to ultra-low tidal volumes and the presence of RV dysfunction.
Similarly, ventilation-induced variations in stroke volume measured by pulse
7 Managing the Systemic Circulation: Volume Status and RV Function 149

c­ ontour analysis or bioreactance are invalid during ECMO and should not be used
for this purpose. In patients on V-A ECMO, the reduction of native cardiac pulsatil-
ity in the arterial waveform further invalidates PPV.

Passive Leg Raising On the other hand, passive leg raising (PLR) does not rely on
changes in pleural pressure and has been shown useful during V-V ECMO, with an
increment in PLR-induced stroke volume greater than 10% (using transthoracic
echocardiography) predicting fluid response. PLR-related changes in end-tidal CO2
are likely to be erroneous as most CO2 is removed via the extracorporeal mem-
brane. Prone ventilation precludes using PLR to predict fluid responsiveness. An
alternative is to infuse a mini-fluid bolus (100 cc crystalloid) over 1 min, using pulse
contour-based methods to judge the effect [2].

POCUS The presence of a venous cannula can make IVC measurements difficult
to interpret. However, collapse of the IVC around the ECMO cannula on POCUS
examination suggests hypovolemia. Additionally, the very low tidal volumes that
are often employed in ventilatory management of V-V ECMO patients would
decrease the accuracy of respiratory variation in IVC diameter as a marker of fluid
responsiveness. In patients who are breathing spontaneously, high respiratory drive
and greatly reduced lung compliance may facilitate inspiratory collapse of the IVC,
mimicking hypovolemia.

Abdominal Ultrasound More recently, POCUS evaluation of abdominal organs

has been used to evaluate venous congestion. The VEXUS score (venous excess on
ultrasound) uses the IVC, hepatic venous waveform, portal venous waveform, and
intrarenal venous Doppler waveforms to grade the severity of venous congestion
[3]. The combination of an IVC size >2 cm along with severe abnormalities in 2 out
of 3 of hepatic, portal, and renal veins was associated with the development of acute
kidney injury (AKI) in postoperative cardiac surgical patients. While this score had
a hazard ratio of 3.69 and high specificity (96% and 95% confidence interval
89–99%) for predicting AKI, it carried a low sensitivity (27%). This is likely due to
the multifactorial nature of postoperative AKI. Importantly though, evaluation of
volume excess on ultrasound outperformed CVP in predicting renal injury.

Management

Fluid management in ARDS patients on extracorporeal support is guided by two

conflicting goals. On the one hand, pulmonary edema is alleviated by a restrictive
fluid therapy regimen [4]. On the other hand, limiting intravascular volume could
limit ECMO flow, hence affecting the gas exchange provided by the ECMO circuit.
Fluid therapy should hence be titrated by frequent monitoring of both end points.
Table 7.1 lists commonly used markers for hypovolemia and for volume overload in
patients on ECMO.
150 S. Krishnan and G. A. Schmidt

Table 7.1 Markers for volume overload and hypovolemia in patients on ECMO
Markers that suggest volume overload Markers that suggest hypovolemia
Elevated CVPa High negative pressures on the ECMO inflowb
Severe RV dilation (RVEDD/LVEDD >1.0) Chattering in the ECMO tubingb
Interventricular septal shift, D-shaped LV IVC collapse around ECMO cannula
Systolic blunting of hepatic venous flow Increasing vasopressor requirement
Pulsatile portal venous flow (>50%)
Systolic blunting of renal venous flow
Increased renal arterial resistive index
a
In addition to the other, well-recognized caveats related to the usefulness of CVP as a marker of
fluid responsiveness in critically ill patients, CVP can also be falsely elevated or reduced in ECMO
patients, depending on cannula position
b
High negative pressures on ECMO inflow and ECMO cannula chatter can also occur due to can-
nula malpositioning or tamponade, if the cannula is unable to drain venous return

Resuscitation Patients with ARDS are usually managed in a volume-restricted

manner to limit pulmonary congestion. The initiation of ECMO can worsen hypo-
volemia due to release of cytokines causing vasoplegia, as well as any bleeding
resulting from the cannulation procedure. Hence, patients often need volume resus-
citation initially.

Diuresis Once adequate flow has been established, volume removal should be initi-
ated, with the goal of relieving pulmonary and systemic edema. As intravascular
volume is decreased, signs of low preload might begin to appear. Potential responses
to cannula chatter and inadequate ECMO circuit flow include volume challenge,
reducing circuit flow, or calming inspiratory effort by raising sweep gas flow or
through pharmacologic means. It is important to note, however, that conditions such
as cannula malposition or tamponade might limit venous return to the specific site
of drainage from the cannula, despite there being adequate intravascular volume.
Echocardiography would play an important role in ruling out these confounders for
cannula chatter and high negative pressures. The presence of clots within the venous
drainage cannula might also lead to these extrinsic signs of hypovolemia and can be
more difficult to rule out.

Right Ventricular Function

Patients with ARDS develop pulmonary hypertension related to atelectasis, ventila-

tory hyperinflation, hypoxic pulmonary vasoconstriction, hypercapnia, and the
impact of inflammatory and thrombotic processes on the pulmonary vasculature.
The right ventricle is a thin-walled structure that is very sensitive to increases in
afterload. Further, sepsis and metabolic derangements can affect RV function
directly. RV dysfunction causes venous congestion and low cardiac output, leading
to poor tissue perfusion and organ failure.
7 Managing the Systemic Circulation: Volume Status and RV Function 151

Incidence Severe ARDS is associated with a 20–50% incidence of acute RV dys-

function. Important risk factors include infectious processes, worse gas exchange
(ratio of arterial partial pressure of oxygen to fractional inspired oxygen <150),
hypercarbia (PaCO2 > 48 mm Hg), and high ventilatory driving pressures (>18 mm
Hg). Similarly, among patients with COVID-19 pneumonia, RV dysfunction was
more common in patients with worse clinical conditions.

Outcomes The presence of RV dysfunction is associated with longer hospital stays

and decreased survival in patients with ARDS. Among ARDS patients who require
V-V ECMO, RV failure is again associated with decreased survival to decannulation
and hospital discharge [5]. As described further below, while some patients with
mild RV dysfunction can be managed with V-V ECMO alone, those with severe RV
dysfunction should be managed with V-A ECMO for cardiopulmonary support.
Among patients with pulmonary embolism and acute myocardial infarction, RV
dysfunction is seen in significant percentage (30–50%) of patients and is also asso-
ciated with worse clinical outcomes. Among patients undergoing cardiac surgery,
the incidence of RV dysfunction is very low (<1%) for routine cardiotomy but much
higher for patients undergoing heart transplantation and left ventricular assist device
placement (up to 50%).

Monitoring

The cannulation strategy, circulatory mechanics, and minimized ventilatory settings

can all affect the validity of routine hemodynamic monitoring parameters in patients
on ECMO. It is hence important to incorporate multiple parameters into clinical
decision-making while acknowledging their individual nuances.
Echocardiography Echocardiography plays an important role in the detection,
monitoring, and management of acute RV dysfunction [6, 7]. Table 7.2 lists various
echocardiographic parameters for evaluating RV dysfunction in patients on VV
ECMO. Usual metrics include RV:LV end-diastolic area, LV eccentricity index, tri-
cuspid annular plane systolic excursion, tissue Doppler peak systolic velocity at the
lateral annulus, and longitudinal strain of the RV free wall. RV fractional area
change, measured on the apical four-chamber view, most accurately reflects cardiac
MRI measurements of RV ejection fraction. The presence of a biphasic pulmonary
arterial flow waveform and pulmonary artery acceleration time less than 100 ms on
pulse wave Doppler indicates pulmonary arterial hypertension.
After ECMO support is initiated, monitoring for RV dysfunction remains impor-
tant. We suggest routine evaluation of the RV throughout the period of critical ill-
ness (see Chap. 12 Daily Management). Most echocardiographic parameters of RV
size and function remain valid monitoring markers despite the patient being sup-
ported on V-V ECMO. Of note, the tricuspid regurgitant jet velocity can no longer
be used to estimate pulmonary artery pressures because of confounding by return of
152 S. Krishnan and G. A. Schmidt

Table 7.2 Monitoring RV dysfunction by echocardiography, while on ECMO

Parameter Abnormal value Comments
Preload
IVC size and >2 cm diameter with minimal IVC collapse around ECMO cannula
variation respiratory variation suggests hypovolemia
Ultra-low tidal volume ventilation on ECMO
limits respiratory variation
RVEDA/LVEDA >0.6 Measured in the apical four-chamber or
mid-esophageal four-chamber view
Contractility
RV FAC <35% Clear endocardial border definition is
required
TAPSE <17 mm Optimal Doppler beam alignment is required
RV S′ <10 cm/s Optimal Doppler beam alignment is required
RV free wall <20%
strain
Afterload
RVSP >40 mm Hg ECMO return flow directed at the tricuspid
valve confounds TR jet assessment
PA acceleration <100 ms
time
PA waveform Biphasic

blood flow from the ECMO cannula. On the other hand, all echocardiographic
parameters of RV function become less accurate in patients on V-A ECMO because
cardiac filling is unloaded by the pump.
Arterial Line-Based Cardiac Output Monitoring Minimally invasive, noncali-
brated hemodynamic monitors that utilize the arterial waveform are likely not
affected by V-V ECMO [8]. However, the quality of the arterial waveform and
inherent inconsistencies in this technology might make the data unreliable. In
patients on V-A ECMO, the addition of continuous pump flow to the reduced LV
pulsatility invalidates analysis of arterial waveform-based cardiac output assessment.
Monitors that rely on indicator dilution (thermal or lithium) would suffer indica-
tor loss into the extracorporeal circuit. The amount of flow in the ECMO circuit and
native cardiac output would impact the validity of this assessment. In patients on
low flow V-V ECMO for hypercarbic respiratory failure, transpulmonary thermodi-
lution cardiac output measurements are similar to those obtained by pulse contour
analysis [9].
Thermodilution Cardiac Output While pulmonary artery catheters (PACs) have
fallen out of favor in most critical care settings, they are still recommended in the
management of patients with severe cardiogenic shock. PACs offer important infor-
mation about left- and right-sided filling pressures, cardiac output, and the ability to
follow trends. However, these values should be interpreted with caution in patients
with severe pulmonary disease and on extracorporeal circulatory support. Various
7 Managing the Systemic Circulation: Volume Status and RV Function 153

cardiac chambers might be partially unloaded, depending upon circuit configura-

tion. Thermodilution cardiac output is rendered inaccurate because of diversion of
part of the injectate to the extracorporeal circuit.

Cardiac Output by Echo: Monitoring cardiac output by measuring left ventricular

outflow tract velocity time integral (LVOT VTI) remains accurate in patients on VV
ECMO. In patients on V-A ECMO with preserved pulsatility, monitoring changes in
LVOT VTI can still be useful. For example, an increase in LVOT VTI with
Trendelenburg maneuver can help predict volume responsiveness [10].

SvO2 Venous oxygen saturation measured in the pulmonary artery (SvO2) or the
superior vena cava (ScvO2) is often used as a surrogate marker of cardiac output. In
patients on V-V ECMO, ScvO2 might or might not be reflective of the adequacy of
tissue perfusion, depending on the site of blood sampling and the position of the
reinfusion cannula. Similarly, Fick cardiac output measurements using ScvO2 will
be inaccurate in patients on V-V ECMO. Further downstream, oxygenated blood
returning from the V-V ECMO circuit mixes with the native cardiac output, making
the SvO2 an invalid measure of tissue oxygenation.
In patients on peripheral cannulation for V-A ECMO, pulmonary arterial flow
would consist mostly of blood from the superior vena cava and coronary sinus,
while blood from the inferior vena cava is drawn into the ECMO circuit. Measuring
SvO2 on the pre-membrane side of the oxygenator provides another method of eval-
uating tissue perfusion.
Lactate Serum lactate concentrations remain a reliable parameter for organ perfu-
sion in patients on ECMO. Elevated lactate levels correlate with mortality outcomes
in patients with ARDS with respiratory failure. Monitoring lactate trends can further
assist in evaluating the efficacy of circulatory support [11].

NIRS Near-infrared spectroscopy (NIRS) evaluates the tissue absorption of near-­

infrared wavelengths to provide information on regional oxygen saturation (rSO2).
A decrease in cerebral rSO2 value reflects a decrease in regional or global perfusion
and should be treated with an increase in cardiac output, mean arterial pressure,
hemoglobin level, or arterial oxygen saturation. Paradoxically, an increase in cere-
bral rSO2 in patients on V-A ECMO might reflect a decrease in native cardiac output
and should be investigated further.

Abdominal Ultrasound As discussed above, assessment of abdominal organs

(hepatic vein, portal vein, renal vein) on POCUS has been shown to correlate with
right ventricular dysfunction and venous congestion. Evidence of severe venous
congestion in multiple organs on POCUS can be used to guide volume removal.
These parameters have not yet been validated in the setting of extracorporeal
support.
154 S. Krishnan and G. A. Schmidt

Management

Both intravascular volume depletion and volume overload are detrimental to organ
perfusion. Specifically, volume overload increases RV wall tension, with reduced
coronary perfusion pressure as well as reduced left ventricular diastolic filling due
to septal shift toward the left. This could lead to a worsening spiral of reduced car-
diac output, organ hypoperfusion, and venous congestion.
Vasoactives Vasoactive management for RV failure follows the same principles as
prior to initiation of V-V ECMO. Norepinephrine improves systemic pressure and
RV perfusion. At very high doses, norepinephrine and other alpha-agonists might
worsen pulmonary hypertension. On the other hand, vasopressin has been shown to
have minimal effects on pulmonary vascular resistance. Inotropic support with epi-
nephrine or dopamine might often still be required. Inodilators like milrinone and
dobutamine should be used cautiously because of the potential decrease in systemic
afterload. Inhaled vasodilators provide the benefit of decrease in pulmonary vascu-
lar resistance, without affecting systemic pressures. These inhaled therapies likely
still remain physiologically effective while the patient is on ECMO, despite low
tidal volume ventilation.

Ventilation Both atelectasis and hyperinflation can cause an increase in pulmonary

vascular resistance. In addition, the exacerbation of hypoxic pulmonary vasocon-
striction with poor respiratory gas exchange impacts right ventricular afterload. On
V-V ECMO, high positive end-expiratory pressure is usually maintained to prevent
decruitment and alveolar atelectasis [12]. The extracorporeal oxygenator is very
effective in clearing CO2. Hence, tidal volumes can be decreased significantly on
VV ECMO, often to as low as 2–3 ml/kg. Further, prone positioning decreases RV
afterload and improves oxygenation and might be used as additional therapy in
patients not supported adequately by V-V ECMO.
In contrast, in patients on peripheral V-A ECMO support, ventilation through
the native lungs is responsible for oxygenation of the upper half of the body.
Hence, ventilatory parameters might need to be kept close to normal. The ven-
tilatory strategy should still incorporate lung-protective and RV-protective
strategies.
Impact of V-V ECMO on RV dysfunction: The institution of extracorporeal gas
exchange with V-V ECMO augments SvO2 and lowers PaCO2. These in turn ame-
liorate pulmonary vasoconstriction, hence improving RV afterload. This is evi-
denced by a decrease in pulmonary artery and central venous pressure and an
increase in cardiac index with the improvement in pre-pulmonary capillary oxygen-
ation [13, 14]. Further, the relief of overdistension from mechanical ventilatory air-
way pressures might also improve RV function. It is hence important to recognize
that for appropriately selected patients requiring inotropic therapy prior to initiation
of extracorporeal support, RV dysfunction could be managed with V-V ECMO
7 Managing the Systemic Circulation: Volume Status and RV Function 155

Table 7.3 Clinical features that suggest the use of VV vs VA ECMO in patients with
respiratory failure
VV ECMO is appropriate VA ECMO is appropriate
None or mild requirement for inotropic High dose requirement for inotropic support
support
Echocardiographic measures do not show Echocardiographic measures show severe RV
severe RV dysfunction dysfunction
Continued evidence of improving tissue Continued deterioration of circulatory parameters
perfusion while on VV ECMO and organ function, despite VV ECMO support
Septic shock with adequate cardiac Bi-ventricular failure or severe LV failure
output

alone. Table 7.3 lists criteria that suggest appropriateness of V-V or V-A ECMO in
patients with respiratory failure.

V-V ECMO as Initial Support In a review of 717 ARDS patients with shock in the
ELSO registry, only 18% of patients were started on V-A ECMO [15]. V-V ECMO
was associated with less gastrointestinal bleeding and hemolysis, but overall rates
of bleeding, stroke, and renal failure were similar. Survival to discharge was 58%
for V-V ECMO in contrast to 43% for V-A ECMO (p = 0.002). Multivariable regres-
sion analysis revealed V-V ECMO to be an independent predictor of survival to
discharge relative to V-A ECMO. Further, the conversion from V-V to V-A ECMO
in the study was only 4%.

V-A ECMO Progression of RV dysfunction during V-V ECMO is an ominous sign,

pointing to lack of recovery. There has been some evidence of improvement in RV
parameters with institution of intra-aortic balloon pump (IABP) counterpulsation
for V-V ECMO patients with cardiogenic shock. More commonly, for patients with
severe hemodynamic compromise prior to ECMO initiation, or those who remain
underperfused despite extracorporeal gas exchange on V-V ECMO, cardiopulmo-
nary support with V-A ECMO would be required.

Harlequin Syndrome Conversion to peripheral V-A ECMO in a patient with

severe pulmonary disease could lead to Harlequin syndrome, due to inadequate gas
exchange for the blood passing through the lungs. Hence, more central arterial (e.g.,
subclavian) cannulation might be required. Alternatively, maintaining partial venous
return from the ECMO circuit through veno-veno arterial hybrid support (V-VA)
would allow for oxygenation of pulmonary arterial blood.

V-PA ECMO A different cannulation strategy, using a right ventricular assist

device with ECMO has also been described for COVID-19 ARDS [16]. This is
done via a large (29 or 31 French) dual lumen cannula that is inserted percutane-
ously into the right internal jugular vein and passed into the main pulmonary artery,
allowing for mechanical unloading of the RV in addition to gas exchange and
oxygenation.
156 S. Krishnan and G. A. Schmidt

Conclusion

Acute RV dysfunction is common in patients with severe ARDS. Patients on extra-

corporeal support for pulmonary or cardiopulmonary failure require uniquely tai-
lored therapy. Regular attention should be paid to patient-specific goals. Point-of-care
echocardiography plays an important role in guiding therapy.
Funding Statement Support was provided solely from institutional and/or depart-
mental sources.

Conflicts of Interest G.A.S. receives royalties from McGraw-Hill, Springer- Link,

and UpToDate.

References

1. Krishnan S, Schmidt GA. Hemodynamic monitoring in the extracorporeal membrane oxygen-

ation patient. Curr Opin Crit Care. 2019;25:285–91.
2. Mallat J, Meddour M, Durville E, Lemyze M, Pepy F, Temime J, Vangrunderbeeck N, Tronchon
L, Thevenin D, Tavernier B. Decrease in pulse pressure and stroke volume variations after mini-­
fluid challenge accurately predicts fluid responsiveness. Br J Anaesth. 2015;115(3):449–56.
3. Beaubien-Souligny W, Rola P, Haycock K, et al. Quantifying systemic congestion with point-­
of-­care ultrasound: development of the venous excess ultrasound grading system. Ultrasound
J. 2020;12:16.
4. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS)
Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D,
deBoisblanc B, Connors AF Jr, Hite RD, Harabin AL. Comparison of two fluid-management
strategies in acute lung injury. N Engl J Med. 2006;354(24):2564–75.
5. Ortiz F, Brunsvold ME, Bartos JA. Right ventricular dysfunction and mortality after cannula-
tion for venovenous extracorporeal membrane oxygenation. Crit Care Explor. 2020;2:e0268.
6. Dandel M. Heart–lung interactions in COVID-19: prognostic impact and usefulness of bedside
echocardiography for monitoring of the right ventricle involvement. Heart Fail Rev. Published
online April 17, 2021. 1–15.
7. Krishnan S, Schmidt GA. Acute right ventricular dysfunction: real-time management with
echocardiography. Chest. 2015147:835–46.
8. Bond O, Pozzebon S, Franchi F, Zama Cavicchi F, Creteur J, Vincent JL, et al. Comparison of
estimation of cardiac output using an uncalibrated pulse contour method and echocardiography
during veno-venous extracorporeal membrane oxygenation. Perfusion. 2019;35(5):397–401.
9. Minini A, Raes M, Taccone FS, Malbrain MLNG. Transpulmonary thermodilution during
extracorporeal organ support (ECOS): is it worth it? A brief commentary on the effects of the
extracorporeal circuit on TPTD-derived parameters. J Clin Monit Comput. 2021;35(4):681–7.
10. Luo JC, Su Y, Dong LL, Hou JY, Li X, Zhang Y, Ma GG, Zheng JL, Hao GW, Wang H, Zhang
YJ, Luo Z, Tu GW. Trendelenburg maneuver predicts fluid responsiveness in patients on veno-­
arterial extracorporeal membrane oxygenation. Ann Intensive Care. 2021;11(1):16.
11. Slottosch I, Liakopoulos O, Kuhn E, et al. Lactate and lactate clearance as valuable tool to
evaluate ECMO therapy in cardio- genic shock. J Crit Care. 2017;42:35–41.
12. Schmidt M, Stewart C, Bailey M, et al. Mechanical ventilation management during extracor-
poreal membrane oxygenation for acute respiratory distress syndrome: a retrospective interna-
tional multicenter study. Crit Care Med. 2015;43:654–64.
7 Managing the Systemic Circulation: Volume Status and RV Function 157

13. Bunge JJH, Caliskan K, Gommers D, Reis MD. Right ventricular dysfunction during acute
respiratory distress syndrome and veno-venous extracorporeal membrane oxygenation. J
Thorac Dis. 2018;10(Suppl 5):S674–82.
14. Reis Miranda D, van Thiel R, Brodie D, Bakker J. Right ventricular unloading after ini-
tiation of venovenous extracorporeal membrane oxygenation. Am J Respir Crit Care Med.
2015;191(3):346–8.
15. Kon ZN, Bittle GJ, Pasrija C, et al. Venovenous versus venoarterial extracorporeal membrane
oxygenation for adult patients with acute respiratory distress syndrome requiring precannula-
tion hemodynamic support: a review of the ELSO registry. Ann Thorac Surg. 2017;104:645–9.
16. Cain MT, Smith NJ, Barash M, Simpson P, Durham LA 3rd, Makker H, Roberts C, Falcucci
O, Wang D, Walker R, Ahmed G, Brown SA, Nanchal RS, Joyce DL. Extracorporeal mem-
brane oxygenation with right ventricular assist device for COVID-19 ARDS. J Surg Res.
2021;264:81–9.
Chapter 8
Antithrombotic Therapy for ECMO

Usha S. Perepu

Introduction

Extra corporeal membrane oxygenation (ECMO) has been shown to improve sur-
vival in patients with acute respiratory failure when conventional ventilatory strate-
gies are inadequate [1]. Blood flow through these artificial devices disrupts normal
hemostasis and promotes coagulation by activation of the coagulation pathway.
Coagulation is induced by contact of the blood with the surface of the circuit and
membranes of the ECMO device. Therefore, anticoagulation is necessary to prevent
thrombus formation and maintain device function. However, bleeding risk is high
due to consumption of coagulation factors along with platelet dysfunction. Hence a
precarious balance between thrombosis and bleeding exists. Studies have shown
that thrombosis of the circuit and systemic thromboembolism is around 53% and is
partly dependent on the anticoagulation regimen used [4]. Severe hemorrhage of
around 40% with intracranial bleeding in 16–21% has been reported. It is also
known that hemostatic complications are associated with increased mortality.

Hemostatic Alterations During ECMO

Extracorporeal surfaces while designed to withstand high pressures in the devices

alter hemostatic function significantly. The hemostatic changes can be attributed to
the Virchow’s triad which include endothelial injury, altered blood flow, and

U. S. Perepu (*)
Department of Internal Medicine, University of Iowa, Iowa City, IA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 159

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_8
160 U. S. Perepu

a ECMO Circuit

Hy
ce

pe
rfa
• Procoagulant hydrophobic • Free hemoglobin from hemolysis

rc
su

oag
artificial surface VV-ECMO • Circulating microparticles

lial
• Early fibrinogen deposition

ula
the
• Lack of endothelium

bil
do

y it
En

Abnormal blood flow

• Shear stress through pump

• Large surface area of
oxygenator membrane
• Low flow at connector sites

b Patient
Hy
ce

pe
rfa

• Vessel damage at • Disseminated intrvascular coagulation

rco
su

cannula sites • Elevated Factor VIII

ag
lial

• Surgical sites • Tissue factor release from monocytes

ula
the

bil
do

ity
En

Abnormal blood flow

• Low flow at venous cannula

insertion site
• Immobility

Fig. 8.1 The prothrombotic changes described by Virchow’s triad in respect to (a) the ECMO
circuit and (b) patient factors (Adapted from Doyle and Hunt [22])

hypercoagulability. The etiology of the hemostatic changes can be divided into two
categories: (1) interactions between the blood and the extracorporeal circuit and (2)
patient-related factors (Fig. 8.1).

Hematologic Consequences with ECMO

The contact system activation does not play a significant role in cell-based coagula-
tion; this is thought to be the primary mode of thrombin generation in blood flow
associated with extracorporeal circulation. The contact or intrinsic system is
8 Antithrombotic Therapy for ECMO 161

ECMO Biomaterial En
Endothelium
with absorbed proteins

C3
C3a
C3 Anaphylatoxins
C3b
C
C33 C5a
Contact Activation
b
C3bBb C3b
C3b3
C 3b3bBb
3b3bBb
bB
C3b3bBb
fXIIf fXII C5
C5
+ C6,
C C7,
C6 C C8,
C7 C CC9
Kallikrien
SC5b-9 or C5b-9 (MAC)
fXIIa Prekallikrein
HMWK Complement Activation

Bradykinin
fXII fXIa
fXIa
Ia Coagulation Cascade

fIX
X fiXa
fiX
fiXa
Xaa
Neutrophil Extravasation
ffX
X fXa

Fibrinogen hro
ombin
bin
iin
Thrombinn Prot
Pro
Pr
Prrot
o
otthrom
hrom
ro
ombin
ombin
Prothrombinbin

Fibrin

Integrans

Platelet Degranulation
IL-1β

TNF-α
Pro-inflammatory cytokines
Neutrophil Activation

Complement Products P-selectin and other adhesion molecules

Thrombin and fXlla

Endothelial Activation

Platelet Activation

Fig. 8.2 The inflammatory response to extracorporeal membrane oxygenation (ECMO). During
ECMO, the complement and contact systems are activated as a result of blood-biomaterial interac-
tion. The alternative complement pathway (AP) is primarily responsible for producing the ana-
phylatoxins C3a and C5a and the membrane attack complex (MAC). This occurs as the result of
increased hydrolysis of C3 on the biomaterial surface. The contact system is responsible for pro-
ducing activated factor XII (FXIIa), which induces the intrinsic coagulation pathway, leading to
thrombin formation. Products produced by each of these systems promote the production of proin-
flammatory cytokines and have direct effects on leukocytes, platelets, and the vascular endothe-
lium. In particular, neutrophils are activated, leading to increased neutrophilic infiltration of tissue
and eventual organ damage (adapted from Millar et al. [36])

composed of factor XII, factor XI, high-molecular-weight kininogen and prekalli-

krein (Fig. 8.2). Factor XII is activated by binding to nonphysiologic surfaces in the
extra corporeal system. This results in activation of factor XII which cleaves pre-­
kallikrein to release kallikrein and high-molecular-weight kallikrein to bradykinin.
This process has been shown to occur rapidly within 10 min of initiation of ECMO
[26]. Factor XIIa formed as a result of contact activation leads to activation of factor
162 U. S. Perepu

XI to XIa which in turn converts factor IX to IXa thus leading to activation of factor
X and thrombin generation. Although tissue factor-induced coagulation is thought
to be less prominent in extracorporeal circulation due to the absence of endothelial
injury, disseminated intravascular coagulation due to sepsis leads to tissue factor
expression by monocytes and macrophages which results in thrombin generation.
Also, TNF alpha and IL-6 have been described to induce endothelial cell expression
of soluble tissue factor.
Platelets not only play a role in hemostasis but are also a key mediator of inflam-
mation during ECMO by releasing proinflammatory cytokines, adhesion molecules,
and growth factors (Fig. 8.2). Complement activation, thrombin generation, and
adhesion of fibrinogen to the circuit surface all result in platelet activation. On the
other hand, continuous activation of the coagulation system can result in consump-
tion of coagulation factors and platelets leading to a bleeding due to thrombocyto-
penia. Severe thrombocytopenia of <50,000 cells/mm3 is seen in 22% of patients on
ECMO making anticoagulation challenging and increasing transfusion require-
ment [30].
Fibrinolysis is an essential next step in the hemostatic regulation once endothe-
lial injury is resolved. Activation of fibrinolysis is initiated by release of tissue plas-
minogen activator which activates plasminogen to plasmin causing fibrin breakdown.
Hyperfibrinolysis is excessive fibrin breakdown which can be associated with very
high D dimers resulting in bleeding. This can be commonly seen in patients on
ECMO [29].
Extracorporeal circulation results in red blood cell breakdown leading to
increased levels of free circulating hemoglobin (Hb) [22]. Free Hb is known to
deplete nitric oxide by binding which in turn leads to vasoconstriction and increas-
ing platelet activation thereby creating a prothrombotic state. Free plasma hemoglo-
bin levels >50 mg/dL after 24 h have been shown to be an independent predictor of
mortality in ECMO [27]. In addition, erythrocyte-derived microparticles resulting
from hemolysis or from transfused blood cells generate thrombin via contact path-
way [28].

Inflammatory Consequences with ECMO

The complement system is activated by three pathways: the classical pathway (CP),
the alternate pathway (AP), and the lectin pathway (LP), all of which are triggered
by distinct mechanisms with a common goal of C3 activation leading to the forma-
tion of the lytic membrane attack complex (MAC). The AP is activated in patients
on ECMO because of contact of the blood with the foreign material of the circuit.
Due to the absence of regulatory proteins to suppress the complement system by the
artificial surface, there is propagation of the complement cascade leading to exces-
sive inflammatory response and capillary leak syndrome. Studies have shown that
there is rapid activation of complement and formation of MAC within the first hour
of initiation of ECMO [20]. These levels tend to normalize within 2 days.
8 Antithrombotic Therapy for ECMO 163

Heparin-­coated circuits have shown reduction in AP activation as heparin has an

anticomplement effect [21]. The CP and LP pathways are activated by antibody
antigen complexes and carbohydrates in patients with sepsis receiving ECMO.
Critical illness triggers systemic inflammatory response syndrome (SIRS) due to
initiation of the adaptive and innate immune systems which is a marker of poor
outcome [31]. This can be aggravated when extracorporeal life support is initiated.
Interactions between the extracorporeal circuit and blood further enhance SIRS. With
this inflammation there is monocyte and neutrophil activation which express tissue
factor on their cell surfaces resulting in activation of the coagulation pathway. SIRS
also induces widespread activation of endothelium which secretes proinflammatory
cytokines such as TNF alpha, this being a potent activator of neutrophils. It also
stimulates transmigration of neutrophils along with macrophage phagocytosis.
Activated neutrophils release cytotoxic enzymes and reactive oxygen species which
has been thought to be responsible for end-organ damage associated with ECMO in
experimental models [32]. Interleukin-6 and interleukin-8 are potent activators of
neutrophils and T cells and promote differentiation of B cells. High interleukin lev-
els are associated with poor survival.

Anticoagulation

Various modalities have been adopted to counteract the prothrombotic state with
ECMO. Nowadays almost all ECMO circuits are coated with heparin to reduce
activation of the coagulation pathway. Anticoagulants and antiplatelet agents have
been considered to overcome the risk for thrombosis. However, there is no consen-
sus on the ideal anticoagulation strategy [1, 4]. Large surveys showed that 96–100%
of centers reported using unfractionated heparin (UFH) as their main anticoagulant
of choice [3]. This is likely as UFH is the anticoagulant with the most experience, is
inexpensive, has short half-life with protamine as a reversal agent, and has relatively
low side effect profile. UFH is a complex glycosaminoglycan that binds to anti-
thrombin (AT); once bound, the UFH-AT complex inactivates coagulation factors
such as thrombin and factor Xa. It inhibits thrombin after it is formed but does not
prevent thrombin generation nor does it inhibit thrombin already bound to fibrin.
The onset of action is instantaneous after an intravenous infusion with a half-life of
around 45 min. UFH is metabolized in the reticuloendothelial system and the liver
and excreted in the urine. No dose modifications are needed for renal impairment
making it ideal for patients on ECMO with multiorgan failure. ELSO recommends
starting a 50–100 unit/kg bolus dose of UFH at the time of cannulation followed by
the initiation of a UFH infusion at 7.5–20 units/kg/h to achieve an ACT goal of
180–200s although there is significant institutional variability [2].
Argatroban and bivalirudin are synthetic parenteral short-acting direct thrombin
inhibitors (DTIs) which bind to both circulating and clot-bound thrombin inhibiting
thrombin-mediated cleavage of fibrinogen to fibrin; activation of coagulation factor
V, VIII, and XIII; and platelet aggregation. They have a more predictable
164 U. S. Perepu

pharmacokinetics and result in greater reduction in thrombin generation making

them more efficacious compared to UFH. DTIs work independently of antithrombin
which makes them more reliable in patients with low or fluctuating antithrombin
activity. DTIs do not interact with heparin-induced antibodies making them the
choice of anticoagulation in patients with heparin-induced thrombocytopenia (HIT).
Finally, they do not bind to plasma proteins or other cells leading to fewer fluctua-
tions in serum electrolytes and other cell counts. Although DTIs do not have a spe-
cific antidote, they have a relatively short half-life with rapid tapering anticoagulant
effect despite end-organ damage in ECMO patients. Moreover, ECMO patients
rarely require complete reversal of anticoagulation due to the high risk for
thrombosis.
Argatroban is the anticoagulation of choice in patients with confirmed or sus-
pected HIT. Case series and several case reports have shown successful use of arg-
atroban in patients with HIT on ECMO [16]. A tenfold lower dose than
manufacturer-recommended dose was adequate to achieve appropriate anticoagula-
tion [14, 15]. There are no reported cases with argatroban as a first-line anticoagu-
lant in ECMO.
Argatroban is administered by intravenous infusion, with an onset of action
within 30 min and a half-life of 45 min. It is primarily metabolized by enzyme
breakdown in the liver. Infusion is started at 0.2–1 mcg/kg/min and adjusted to
maintain aPTT 1.5–2.5 times the baseline values.
Bivalirudin is the primary anticoagulant in patients with HIT needing percutane-
ous coronary interventions. Several case series and retrospective studies in patients
on ECMO have showed no difference in thromboembolic or bleeding complications
between UFH and bivalirudin. aPTT was maintained in therapeutic range with
fewer fluctuations with bivalirudin compared to UFH [10–13].
Bivalirudin is administered by intravenous infusion, onset of action is within
minutes, and half-life is 25 min in patients with normal renal function. It is started
with an initial bolus dose of 0.05–0.5 mg/kg followed by an infusion rate of
0.03–0.1 mg/kg/h to maintain aPTT 1.5–2.5 times baseline values [2]. Dose adjust-
ments may be needed based on the duration on ECMO.
Low-molecular-weight heparin (LMWH) is derived by enzymatic depolymeriza-
tion of UFH to result in a lower-molecular-weight heparin. Similar to UFH, it binds
to antithrombin resulting in inactivation of factor Xa. LMWHs are metabolized in
the liver and excreted by the kidney; patients with renal impairment have reduced
clearance of LMWH. Predictable pharmacokinetics and absent need for drug moni-
toring make it a more appealing alternative. Data on the use of LMWH as antico-
agulation in ECMO is sparse. Prophylactic subcutaneous anticoagulation with
LMWH in patients on ECMO has been shown to be safe and feasible in one obser-
vational study with no increased thrombosis. The transfusion requirement was
lower compared to historical controls [5]. Another study compared UFH to LMWH
in split doses of 0.5 mg/kg body weight per day. Thromboembolic events were
higher in the UFH arm compared to LMWH with no difference in bleeding out-
comes between the groups [25]. Although LMWH seems to be promising, more
studies are needed prior to replacing UFH with LMWH.
8 Antithrombotic Therapy for ECMO 165

Anticoagulation Sparing Strategies While anticoagulation is the standard of

care to prevent thrombosis in patients on ECMO, the risk of serious bleeding also
remains a concern. Patients with polytrauma on ECMO are at a high risk for bleed-
ing where therapeutic anticoagulation may be detrimental. Several small case
series have evaluated novel anticoagulation sparing strategies such as low dose of
heparin and citrate anticoagulation as options. Carter et al. demonstrated that hepa-
rin sparing strategy with low-dose heparin compared to standard heparin dosing on
patients with venovenous ECMO showed no statistical differences in mortality,
bleeding risk, or thrombotic complications [6]. A recent meta-analysis of seven
studies including 553 patients compared low-dose heparin versus full-dose heparin
in patients supported with ECMO and showed similar rates of thrombosis, bleed-
ing, and mortality between the two groups [7]. This suggests that low dose of hepa-
rin is a feasible and a safe option in select group of patients at high risk for bleeding.
The target ACT for low-dose heparin used in the studies was 140–160 s or
aPTT <45 s.
Regional citrate anticoagulation has long been used for circuit anticoagulation in
renal replacement therapy and therapeutic apheresis. Ionized Ca plays a major role
as a cofactor for activation of the several coagulation factors and platelet aggrega-
tion. Citrate chelates ionized Ca which in turn makes it unavailable for coagulation.
Due to the rapid metabolism of citrate, its effects are largely localized. Possible
adverse effects of citrate can include symptomatic hypocalcemia and acid base dis-
turbances which can be prevented by Ca supplementation if necessary. There are
some case reports of the use of regional citrate anticoagulation in patients needing
CRRT on ECMO with successful outcomes [8]. Early phase I studies are ongoing in
infants needing ECMO using citrate as the sole anticoagulant [9].
Theoretically platelet inhibition could help reduce activation of coagulation and
thrombosis of the circuit. Although this has proven to be effective in reducing
thrombosis in the laboratory [33], there is little published data on the usefulness of
antiplatelet agents in patients receiving ECMO. It is common practice to add anti-
platelet agents to anticoagulation protocols for the management of ventricular assist
devices and VA ECMO in patients with cardiogenic shock, and there are some insti-
tutions who implement the same rationale for the management of patients on
ECMO. While this could be an option, it is important to assess risk-benefit of adding
these therapies.

Role of Hemostatic Adjuncts

Blood transfusion protocols are not evidence-based but rather based on clinical
experience and transfusion thresholds which vary largely between centers.
Platelet transfusions are considered to aim for a platelet count around
>100,000 cells/mm3. A lower threshold may be adequate in adults at low risk for
bleeding. Fresh frozen plasma may be considered if the prothrombin time is
166 U. S. Perepu

prolonged in the setting of bleeding. Cryoprecipitate can be considered for fibrin-

ogen levels <100–150 mg/dL.
Antifibrinolytics such as aminocaproic acid and tranexamic acid have been
shown to reduce surgical site bleeding. Data on the usefulness of these agents in
ECMO patients is lacking. However, they may have a role if there is evidence of
excessive fibrinolysis as evidenced by TEG/ROTEM.
Recombinant activated factor VII or prothrombin complex concentrates are an
option for refractory bleeding despite other interventions. Cautious use of these
agents is recommended as there is a significant risk for thrombosis.

Monitoring of Anticoagulation

Monitoring anticoagulation in critically ill patients is challenging in itself; addition

of an extracorporeal circuit and anticoagulation increases the complexity even fur-
ther. Stringent monitoring of hemostasis is necessary to quantify the effects of anti-
coagulation and stratify the bleeding and thrombotic risk. Currently, there are a
number of whole blood and plasma-based tests to assess coagulation in vitro; how-
ever, each has its limitations, and they are not well standardized. Although using one
method of monitoring UFH activity is no longer acceptable, multiple tests several
times a day can increase confusion for monitoring. ELSO recommends each ECMO
program has individual protocols that work best for their institution based on the
available tests at their center.
Activated clotting time (ACT): The ACT measures the time it takes for whole
blood to clot when exposed to substances that activate the contact pathway. This is
performed by mixing whole blood with an activator (celite or kaolin) and measuring
the time to clot formation (Table 8.2). Its efficacy is reduced at lower concentrations
of heparin that are typically administered during ECMO. ACT results are affected
by factors other than heparin including thrombocytopenia, platelet dysfunction,
antiplatelet medications, hypofibrinogenemia, and coagulation factor deficiencies.
Hypothermia and hemodilution can also affect ACT results and this may also vary
based on the ACT device. Despite these shortcomings ACT is a widely available
point-of-care test and is useful when it can be complemented with other measure-
ments to monitor UFH. A target of 180–220 s is usually recommended.
Activated partial thromboplastin time (aPTT): The aPTT is a plasma-based test
that measures the time to fibrin formation by using an activator (silica, ellagic acid),
calcium, and phospholipids (Table 8.2). This remains the mainstay of laboratory-­
based heparin monitoring in ECMO centers. ELSO recommends that each center
establish a therapeutic range for aPTT to compensate for the variable response of
aPTT reagents to UFH. This is thought to be a more useful test compared to ACT in
adult patients on ECMO. aPTT may be prolonged due to coagulation factor defi-
ciencies, von Willebrand disease, fibrinogen deficiency, alteration in liver function,
and systemic inflammation. Therapeutic target range is generally 1.5–2.5 times the
baseline aPTT.
8 Antithrombotic Therapy for ECMO 167

Thromboelastography: Thromboelastography (TEG) or rotational thromboelas-

tography (ROTEM) has the ability to reflect on all aspects of hemostasis compared
to standard tests and is used for monitoring at some ECMO programs. ELSO rec-
ommends the use of these tests if available. This test assesses the viscoelastic prop-
erties of clot formation in fresh or citrated whole blood measuring the integrity of
the coagulation cascade from the time of fibrin formation to clot lysis and impor-
tantly includes the contribution of platelets (Tables 8.1 and 8.2). By adding inhibi-
tors of heparin such as heparinase, it can also confirm the effect of the drug on
hemostasis. As a result, UFH responsiveness can be evaluated by TEG/ROTEM by
examining the difference in R or CT-times between tests with and without hepari-
nase, which may be beneficial when there is concern for heparin resistance. Some
centers utilize TEG/ROTEM in addition to AT activity in the replacement of anti-
thrombin. Platelet inhibition can also be evaluated by using arachidonic acid and
adenosine diphosphate. Primary and secondary fibrinolysis can also be measured.
Anti-factor Xa activity (anti-Xa): This assay is distinct from ACT and aPTT in
that it is a measure of heparin effect rather than the heparin concentration by mea-
suring the inhibition of activated factor X by the ability of heparin to catalyze AT
(Table 8.2). Although it is the most reliable test as it is not influenced by

Table 8.1 Parameters and components of hemostasis measured by TEG/ROTEM

Parameters on TEG/ROTEM Description
Reaction time ®/clotting time (CT) sec Time to initial fibrin
Formation. Measures coagulation factor activity
and heparin effect
α angle/α angle degrees Rate of clot formation. Dependent on fibrinogen
and platelet function
Maximum amplitude (MA)/maximum clot Maximal strength of the clot. Dependent on
firmness (MCF) mm platelet and fibrinogen activity
Clot lysis (Ly30)/lysis index (LI) Rate of fibrinolysis

Table 8.2 Anticoagulation monitoring

Monitoring
test Advantages Disadvantages
ACT (sec) Rapid Insensitive to low doses of UFH
Low cost Affected by anemia, thrombocytopenia
Bed-side method
Whole blood analysis
aPTT (sec) Well known Time-consuming
Gold standard to monitor UFH Plasma-based test that does not reflect
therapy cellular contribution to hemostasis
TEG/ Whole blood Time-consuming
ROTEM Reflects cellular contribution to Poor specificity/sensitivity
hemostasis
Anti Xa (IU/ Sensitive to UFH Expensive
mL) Not influenced by coagulopathy, Time-consuming
thrombocytopenia, or dilution Influenced by AT levels
168 U. S. Perepu

coagulopathy, thrombocytopenia, or hemodilution, it is time-consuming and needs

separate calibration for UFH and LMWH. It is important to note that some labora-
tories add exogenous AT to the sample which may lead to inaccurate results. Since
anti-Xa assays require AT to determine the result, it is important to consider AT
deficiency when anti-Xa concentrations are not increasing with increasing doses of
heparin. Moreover, most assays are affected by hyperlipidemia, hyperbilirubinemia,
and high plasma free hemoglobin from hemolysis which are not uncommon in criti-
cally ill patients on ECMO. The typical target ranges are between 0.3 and 0.7 IU/mL.
In summary, we have several modalities for monitoring anticoagulation and pre-
venting bleeding. However, the challenge arises when there is discrepancy between
the various tests. Viscoelastic testing may be the most reliable when such a situation
occurs. Furthermore, TEG-based protocols have been shown to be safe and effective
for guiding anticoagulation compared to PTT-based protocols by improved blood
product utilization and lower heparin dosing without an increase in thrombosis or
bleeding complications [34, 35]. Hence, it is important to establish institutional
standardization of anticoagulation management based on the resources available at
each center.

Specific Situations

Heparin-induced thrombocytopenia (HIT): This is a life-threatening prothrombotic

immune-mediated complication of UFH and LMWH therapy. It is a result of auto-
antibody directed against endogenous platelet factor 4 (PF4) in complex with hepa-
rin. Thrombocytopenia is a result of removal of IG-coated platelets by macrophages
and the reticuloendothelial system. This antibody activates platelets and causes arte-
rial and venous thrombosis. This is specifically relevant in patients on ECMO who
are naturally at risk for thrombosis and HIT increases that risk even further. The true
incidence of HIT in patients receiving ECMO is not well established with ranges
around 0.5–5%. HIT is suspected when there is a decline in platelet count with or
without thrombosis after exposure to heparin analogues. This is particularly chal-
lenging in patients on ECMO where thrombocytopenia and thrombosis due to plate-
let activation, consumption is not uncommon. Moreover, mechanical devices can
result in platelet activation with ongoing release of PF4 which may influence devel-
opment of HIT. The diagnosis of HIT is based on clinical and laboratory confirma-
tion. Standardized assessment of the clinical probability of HIT is crucial with an
assessment tool “the 4Ts score” which is calculated based on the degree of throm-
bocytopenia, timing of thrombocytopenia, presence of thrombosis, and absence of
other causes of thrombocytopenia. A high pretest probability is suspicious for a true
HIT. Laboratory testing includes immunologic assays such as PF4 ELISA. Although
this test is widely available in most institutions, it has a poor specificity of 74–86%
leading to false-positive results. Functional assays such as serotonin release assay
are considered gold standard and measure platelet-activating effects of anti-PF4/
heparin antibodies with high sensitivity and specificity. However, this test is not
8 Antithrombotic Therapy for ECMO 169

widely available and is complex to perform. Management of HIT includes immedi-

ate cessation of all heparin products, including line flushes and heparin-bound cir-
cuit components. As there is a significant risk of a thrombosis without anticoagulation,
prompt initiation of alternative anticoagulation is essential. The options for antico-
agulation in patients on ECMO include direct thrombin inhibitors such as argatro-
ban and bivalirudin. Intravenous immunoglobulins and plasmapheresis have been
suggested as adjunct therapies in severe cases with persistent thrombocytopenia
due to HIT.
Heparin resistance: This is a phenomenon whereby therapeutic levels are unable
to be maintained despite high doses of heparin. It is typically defined as a require-
ment of 35,000 units or more of UFH over a 24 h period to maintain an anti-Xa level
or aPTT within the desired therapeutic range. The causes for heparin resistance
include congenital or acquired antithrombin III deficiency, increased heparin clear-
ance, and heparin-binding acute-phase reactants. Patients on ECMO are at much
higher risk of heparin resistance due to their systemic inflammatory state leading to
antithrombin-independent heparin sequestration [17]. Moreover, activated neutro-
phils in sepsis can secrete heparin-binding protein leading to resistance [18].
Heparin resistance can be treated by increasing the dose of heparin to overcome
resistance and antithrombin concentrates or by switching to an alternative
anticoagulant.
Antithrombin deficiency: Antithrombin is natural anticoagulant which is pro-
duced in the liver. It inhibits coagulation by lysing thrombin and factor Xa. AT
activity is markedly potentiated by heparin; potentiation of its activity is the princi-
pal mechanism by which both UFH and LMWH result in anticoagulation. AT defi-
ciency is common upon initiation of ECMO due to a combination of accelerated
consumption and reduced synthesis [19]. However, the optimal AT activity for adult
patients on ECMO receiving UFH is unknown. Studies in neonates and infants on
ECMO showed no increased bleeding risk when AT supplementation is used.
Practice varies among programs regarding routine monitoring and replacement of
AT. Some programs aim to keep the AT level>50% whereas others treat low AT
activity only if there is evidence of heparin resistance based on low aPTT or anti Xa
levels [2]. AT concentrates are available for replacement if indicated.
Acquired von Willebrand disease: von Willebrand factor (VWF) is a large
plasma glycoprotein synthesized by endothelial cells and megakaryocytes and
undergoes multimerization. VWF contributes to primary hemostasis by platelet
adhesion and aggregation at the site of endothelial injury. In areas of high shear
stress such as ECMO circuits, the multimers become unfolded and are susceptible
to cleavage by ADAMTS13. The loss of high-molecular weight multimers leads to
platelet dysfunction, together resulting in bleeding typically from the mucosal sur-
faces, respiratory tract, and puncture sites. Studies have shown rapid development
of acquired VWD and platelet dysfunction within hours of implantation of
ECMO. The effect is reversible within hours after explantation [23, 24]. Serological
tests will show reduced VWF activity with reduction in high-molecular-weight
multimers. Treatment with desmopressin and VWF-containing factor VIII concen-
trates can be considered if indicated.
170 U. S. Perepu

Conclusion

Patients requiring ECMO are critically ill with various underlying conditions which
impact the coagulation and immune system. Balancing bleeding and thrombosis
remains a challenge. Robust guidelines with the help of quality research are needed
for the management of anticoagulation, treatment of bleeding, and improvement of
survival.

References

1. Esper SA, Welsby IJ, Subramaniam KJ, et al. Adult extracorporeal membrane oxygen-
ation: an international survey of transfusion and anticoagulation techniques. Vox Sang.
2017;112(5):443–52.
2. ELSO anticoagulation guidelines 2014.
3. Bembea MM, Annich G, Rycus P, et al. Variability in anticoagulation management of patients
on extracorporeal membrane oxygenation: an international survey. Pediatr Crit Care Med.
2013;14:e77–84.
4. Sklar MC, Sy E, Lequier L, Fan E, Kanji HD. Anticoagulation practices during venovenous
extracorporeal membrane oxygenation for respiratory failure. A systematic review. Ann Am
Thorac Soc. 2016;13(12):2242–50. https://doi.org/10.1513/AnnalsATS.201605-­364SR. PMID:
27690525.
5. Krueger K, Schmutz A, Zieger B, Kalbhenn J. Venovenous extracorporeal membrane oxygen-
ation with prophylactic subcutaneous anticoagulation only: an observational study in more
than 60 patients. Artif Organs. 2017;41(2):186–92.
6. Carter KT, Kutcher ME, Shake JG, Panos AL, Cochran RP, Creswell LL, et al. Heparin-sparing
anticoagulation strategies are viable options for patients on veno-venous ECMO. J Surg Res.
2019;243:399–409. https://doi.org/10.1016/j.jss.2019.05.050. PMID: 31277018.
7. Lv X, Deng M, Wang L, Dong Y, Chen L, Dai X. Low vs standardized dose anticoagula-
tion regimens for extracorporeal membrane oxygenation: a meta-analysis. PLoS One.
2021;16(4):e0249854. https://doi.org/10.1371/journal.pone.0249854. eCollection 2021.
8. Study using citrate to replace heparin in babies requiring extracorporeal membrane oxygen-
ation (ECMO)
9. Shum HP, Kwan AM, Chan KC, Yan WW. The use of regional citrate anticoagulation con-
tinuous venovenous hemofiltration in extracorporeal membrane oxygenation. ASAIO
J. 2014;60(4):413–8.
10. Kaseer H, Soto-Arenall M, Sanghavi D, Moss J, Ratzlaff R, Pham S, Guru P. Heparin vs
bivalirudin anticoagulation for extracorporeal membrane oxygenation. J Card Surg.
2020;35(4):779–86. https://doi.org/10.1111/jocs.14458. PMID: 32048330.
11. Pieri M, Agracheva N, Bonaveglio E, Greco T, De Bonis M, Covello RD, Zangrillo A,
Pappalardo F. Bivalirudin versus heparin as an anticoagulant during extracorporeal membrane
oxygenation: a case-control study. J Cardiothorac Vasc Anesth. 2013;27(1):30–4.
12. Ranucci M, Ballotta A, Kandil H, et al. Bivalirudin-based versus conventional hepa-
rin anticoagulation for postcardiotomy extracorporeal membrane oxygenation. Crit Care.
2011;15(6):R275.
13. Bohman JK, Yi L, Seelhammer T. Observational case-control comparison of bivalirudin versus
heparin anticoagulation for adult extracorporeal membrane oxygenation (ECMO). Am J Res
Crit Care Med. 2018;197:A5114.
8 Antithrombotic Therapy for ECMO 171

14. Beiderlinden M, Treschan T, Görlinger K, Peters J. Argatroban in extracorporeal membrane oxy-

genation. Artif Organs. 2007;31(6):461–5. https://doi.org/10.1111/j.1525-­1594.2007.00388.x.
PMID: 17537058.
15. Menk M, Briem P, Weiss B, Gassner M, Schwaiberger D, Goldmann A, Pille C, Weber-­
Carstens S. Efficacy and safety of argatroban in patients with acute respiratory distress syn-
drome and extracorporeal lung support. Ann Intensive Care. 2017;7:82.
16. Rougé A, Pelen F, Durand M, Schwebel C. Argatroban for an alternative anticoagulant in HIT
during ECMO. J Intensive Care. 2017;5:39.
17. Raghunathan V, Liu P, Kohs TC, Amirsoltani R, Oakes M, OJ MC, Olson SR, Zonies D, Shatzel
JJ. Resistance is common in patients undergoing extracorporeal membrane oxygenation but
is not associated with worse clinical outcomes. ASAIO J. 2021;67(8):899–906. https://doi.
org/10.1097/MAT.0000000000001334.
18. Linder A, Christensson B, Herwald H, Björck L, Akesson P. Heparin-binding protein: an early
marker of circulatory failure in sepsis. Clin Infect Dis. 2009;49:1044–50.
19. Hirose H, Sarik J, Pitcher H, et al. Antithrombin III deficiency in patients requiring mechanical
circulatory support. J Cardiol Clin Res. 2014;2:1017.
20. Vallhonrat H, Swinford RD, Ingelfinger JR, Williams WW, Ryan DP, Tolkoff-Rubin N,
et al. Rapid activation of the alternative pathway of complement by ECMO. ASAIO
J. 1999;45:113–4.
21. Kopp R, Mottaghy K, Kirschfink M. Mechanism of complement activation during extracor-
poreal blood-biomaterial interaction: effects of heparin coated and uncoated surfaces. ASAIO
J. 2002;48:598–605. https://doi.org/10.1097/00002480-­200211000-­00005.
22. Doyle AJ, Hunt BJ. Current understanding of how extracorporeal membrane oxygenators acti-
vate haemostasis and other blood components. Front Med (Lausanne). 2018;5:352. https://doi.
org/10.3389/fmed.2018.00352. PMID: 30619862.
23. Tauber H, Ott H, Streif W, Weigel G, Loacker L, Fritz J, Heinz A, Velik-Salchner
C. Extracorporeal membrane oxygenation induces short-term loss of high-molecular-weight
von Willebrand factor multimers. Anesth Analg. 2015;120(4):730–6.
24. Kalbhenn J, Schlagenhauf A, Rosenfelder S, Schmutz A, Zieger B. Acquired von Willebrand
syndrome and impaired platelet function during venovenous extracorporeal membrane oxy-
genation: rapid onset and fast recovery. J Heart Lung Transplant. 2018;37(8):985–91. https://
doi.org/10.1016/j.healun.2018.03.013. PMID: 29650295.
25. Cooper E, Burns J, Retter A, Salt G, Camporota L, Meadows CI, Langrish CC, Wyncoll D,
Glover G, Ioannou N, Daly K, Barrett NA. Prevalence of venous thrombosis following veno-
venous extracorporeal membrane oxygenation in patients with severe respiratory failure. Crit
Care Med. 2015;43(12):e581–4. https://doi.org/10.1097/CCM.0000000000001277. PMID:
26308437.
26. Wendel HP, Scheule AM, Eckstein FS, Ziemer G. Haemocompatibility of paediatric mem-
brane oxygenators with heparin-coated surfaces. Perfusion. 1999;14(1):21–8.
27. Omar HR, Mirsaeidi M, Socias S, Sprenker C, Caldeira C, Camporesi EM, Mangar D. Plasma
free hemoglobin is an independent predictor of mortality among patients on extracorporeal
membrane oxygenation support. PLoS One. 2015;10(4):e0124034. https://doi.org/10.1371/
journal.pone.0124034. PMID: 25902047
28. Van Der Meijden PE, Van Schilfgaarde M, Van Oerle R, Renné T, ten Cate H, Spronk
HM. Platelet- and erythrocyte-derived microparticles trigger thrombin generation via factor
XIIa. J Thromb Haemost. 2012;10:1355–62. https://doi.org/10.1111/j.1538-­7836.2012.04758.
29. Hunt BJ, Parratt MS, Segal HC, Sheikh S, Kallis P, Yacoub M. Activation of coagulation and
fibrinolysis during cardiothoracic operations. Ann Thorac Surg. 1998;65:712–8.
30. Abrams D, Baldwin MR, Champion M, Agerstrand C, Eisenberger A, Bacchetta M, et al.
Thrombocytopenia and extracorporeal membrane oxygenation in adults with acute respira-
tory failure: a cohort study. Intensive Care Med. 2016;42:844–52. https://doi.org/10.1007/
s00134-­016-­4312-­9.
172 U. S. Perepu

31. Duffy MJ, Mullan BA, Craig TR, Shyamsundar M, MacSweeney RE, Thompson G, Stevenson
M, McAuley DF. Impaired endothelium-dependent vasodilatation is a novel predictor of mor-
tality in intensive care. Crit Care Med. 2011;39(4):629–35.
32. McILwain RB, Timpa JG, Kurundkar AR, Holt DW, Kelly DR, Hartman YE, Neel ML,
Karnatak RK, Schelonka RL, Anantharamaiah GM, et al. Plasma concentrations of inflamma-
tory cytokines rise rapidly during ECMO-related SIRS due to the release of preformed stores
in the intestine. Lab Investig. 2010;90(1):128–39.
33. Reynolds MM, Annich GM. The artificial endothelium. Organogenesis. 2011;7(1):42–9.
https://doi.org/10.4161/org.7.1.14029. PMID: 21289481; PMCID: PMC3082033.
34. Phillips RC, Shahi N, Leopold D, Levek C, Shirek G, Hilton S, Hyslop R, Gien J, Kinsella JP,
Buckvold S, Liechty KW, Kim JS, Marwan AI. Thromboelastography-guided management
of coagulopathy in neonates with congenital diaphragmatic hernia supported by extracorpo-
real membrane oxygenation. Pediatr Surg Int. 2020;36(9):1027–33. https://doi.org/10.1007/
s00383-­020-­04694-­0. PMID: 32607833.
35. Panigada M, Iapichino G E, Brioni M, et al. Thromboelastography-based anticoagulation man-
agement during extracorporeal membrane oxygenation: a safety and feasibility pilot study.
Ann. Intensive Care. 2018;8:7. https://doi.org/10.1186/s13613-­017-­0352-­8.
36. Millar JE, Fanning JP, McDonald CI, McAuley DF, Fraser JF. The inflammatory response to
extracorporeal membrane oxygenation (ECMO): a review of the pathophysiology. Crit Care.
2016;20(1):387. https://doi.org/10.1186/s13054-­016-­1570-­4. PMID: 27890016; PMCID:
PMC5125043
Chapter 9
Membrane Dysfunction

B. D. Warren, M. J. Sobieszczyk, and P. E. Mason

Introduction

Expanded implementation of extracorporeal membrane oxygenation (ECMO) sup-

port for respiratory and cardiac failure has continued over the last fifteen years [1].
Progression of this support modality is due to both improved circuit technology
(centrifugal blood pumps, decreased priming volume with minimization of external
foreign surfaces, specialized gas-permeable, hollow, microfibers, and extensive
thromboresistant coatings) and expanded expertise in the setting of the H1N1 influ-
enza and SARS-CoV2 pandemics starting in 2009 and 2020, respectively [2].
Despite advances in technology and provider proficiency, mortality per the ELSO
registry remains significant with a range between 35% and 60% depending on initial
patient indication as well as mode of mechanical support required [3]. While the
major determinants of overall survival are often independent of the ECMO appara-
tus, the main drivers of morbidity and mortality in these patients are related to
bleeding and thrombosis [1, 4–6]. Conceptually, these can be separated into patient-
related physiologic complications (hemolysis, hemorrhage, thromboembolism) vs
mechanical support dysfunction (cannula issues, pump/membrane oxygenator

B. D. Warren (*)
Department of Internal Medicine, Brooke Army Medical Center, San Antonio, TX, USA
M. J. Sobieszczyk
Department of Pulmonary and Critical Care Medicine, Brooke Army Medical Center,
San Antonio, TX, USA
e-mail: [email protected]
P. E. Mason
Department of Surgery, Brooke Army Medical Center, San Antonio, TX, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 173

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_9
174 B. D. Warren et al.

malfunction, circuit thrombosis). Here we review pertinent key literature as well as

outline our approach for evaluation of membrane lung dysfunction in patients
requiring ECMO support.

Mechanisms of Membrane Lung Dysfunction

In the typical adult, miniaturized ECMO circuit consists of an electrically pow-

ered blood pump, heparin, or phosphorylcholine (PCC)-coated tubing and an arti-
ficial membrane lung (ML). While there is some inter-manufacturer variability
with regard to the size, surface area, and layout of the ML fibrils, the current
industry standard consists of multiple polymethylpentene (PMP) hollow microfi-
bers [7, 8] which facilitate efficient oxygen uptake and carbon dioxide removal
without a direct blood-gas interface. Despite the advent of tip-to-tip antithrom-
botic coatings within the extracorporeal circuitry, the nonbiologic surfaces experi-
ence near instantaneous adsorption of plasma proteins (especially albumin and
fibrinogen) [9–11]. This initial protein scaffold activates inflammatory, comple-
ment, and contact coagulation pathways resulting in thrombus formation, fibrino-
lysis, leukocyte mobilization, cytokine production, and endothelial cell activation
both within the extracorporeal circuit and systemically throughout the patient’s
endogenous cardiovascular system [5, 12]. At the level of the ML, cellular and
fibrinous deposition can produce pseudomembranous structures along the fiber
walls that increase blood flow resistance and reduce gas transfer efficiency result-
ing in shunt physiology [13–16]. Simultaneously, these depositions are responsi-
ble for thromboembolic complications, dysregulated coagulation, and increased
hemolysis creating regional dead space physiology as a result of widespread clot
developing with in the ML [13, 14]. Coagulation abnormalities, obstruction to
blood flow, or impaired gas transfer is responsible for the majority of ML
exchanges.

Membrane Lung Monitoring

Hematologic Profile

The normal physiologic balance between thrombosis and fibrinolysis, largely regu-
lated at the level of the endothelium, is exceedingly complex. Stability of the coagu-
lation system in patients requiring ECMO support is further complicated by the
disease state as well as the prothrombotic nature of continuous blood to nonbiologic
surface interaction within the extracorporeal circuit [12, 17]. Details of the complex
cross talk between the coagulation system, complementation activation, and sys-
temic inflammatory response as well as the influence of flow dynamics, pressure
changes, and sheer stress within the ECMO circuit have been expertly reviewed
9 Membrane Dysfunction 175

elsewhere [5, 12, 17, 18]. Despite technological advances in pump and ML design,
combined with systemic anticoagulation and local antithrombotic surface modifica-
tions within the circuit, there remains considerable hematologic dysfunction.
Additionally, there is marked variability between ECMO centers regarding the
choice of anticoagulant (heparin vs direct thrombin inhibitor), use of biopassive
(phosphorylcholine)- vs biomimetic (heparin)-coated circuits, as well as methodol-
ogy for monitoring hematologic changes [11, 19–21] (see also Chap. 8). Frequent
assessment of the hematologic profile is integral to the evaluation of ML dysfunc-
tion and possible preemptive replacement as discussed below. While various hema-
tologic markers have been evaluated, our institution utilizes a battery of INR,
activated partial thromboplastin time (aPTT), anti-Xa, D-dimer, fibrinogen, lactate
dehydrogenase (LDH), plasma-free hemoglobin (PfHb), hemolysis index, platelet
count, and thromboelastography (TEG®) in our integrated assessment of ML
performance.

Circuit Pressure Monitoring

The first of three essential functional measures of oxygenator performance is the

transmembrane pressure gradient (ΔPML). Thrombus formation within the ML
increases resistance (RML) to blood flow as reflected by an increase in ΔPML [15, 16,
22, 23]. The pressure change across the ML is shown in Eq. (9.1):

∆PML = PPRE , ML − PPOST , ML (9.1)

There is no consensus standard as to an absolute ΔPML value that necessitates

circuit exchange. However, sustained values above 50 mmHg, a progressive increase
in ΔPML over serial measures, or decreasing blood flow at a sustained pump speed
should raise clinical concern for ML dysfunction. Finally, we recommend relating
ΔPML to blood flow rate (QB) as displayed in Eq. (9.2), as this allows for normaliza-
tion of the pressure change across serial measurements and more accurately repre-
sents the RML.

∅PML / QB (9.2)

ML Gas Transfer

Oxygenation support during ECMO is influenced by the amount of oxygen trans-

ferred across the ML [13, 24, 25]. Adapting the Fick principle, oxygen transfer ( V
O2,ML) may be calculated as in Eq. (9.3):

VO 2, ML = QB ( CPOST O 2 − CPRE O 2 ) × 10 (9.3)

176 B. D. Warren et al.

where V O2,ML = O2 transfer across the ML (ml/min), QB = blood flow rate (L/min),
and CxO2 = O2 content of (pre-/post-ML) blood (ml/dL). The oxygen content of
pre-/post-membrane (CPRE,O2 and CPOST,O2) blood is defined by Eq. (9.4) [26].

Cx O 2 = 1.34 × Hb × S x O 2 + 0.003 × Px O 2 (9.4)

where Hb = hemoglobin (g/dL), SxO2 = O2 saturation of (pre-/post-ML) blood,

PxO2 = O2 partial pressure of (pre-/post-ML) blood (mmHg), and 0.003 is a solubil-
ity coefficient for oxygen.
Overall oxygenator performance is influenced by cannula recirculation [27],
inherent properties of the ML (hollow fiber thickness, total surface area for gas
exchange, and hollow fiber composition), age of the oxygenator, the oxygen content
of the blood entering the circuit, as well as the ratio of ventilation to perfusion (ML
shunt fraction; discussed below). Calculating V O2,ML is the only way to quantify
oxygen transfer and establish the presence of ML dysfunction.

ML Shunt Fraction

A final technical assessment of oxygenator performance involves evaluation of the

ML shunt fraction (Qs/QB). Efficient gas transfer requires matching of ML sweep
gas flow (ventilation) and blood flow (perfusion) across the hollow fiber interface.
While the intrinsic shunt fraction of ML is low (reportedly <10% [28]), accumula-
tion of fibrin and cellular debris creates regions of mismatch (shunt physiology). By
analogy with native lung physiology, one can calculate an ideal “capillary” oxygen
content (CCAPO2) to represent the content of blood in equilibrium with hollow fiber
oxygen (Eqs. (9.5) and (9.6)):

PCAP O 2 = FS O 2 × ( PATM − PH 2 O ) − PPOST CO 2 / RQ (9.5)

where FSO2 is the sweep gas inlet oxygen fraction, PATM is atmospheric pressure,
PH2O is water vapor pressure, PPOSTCO2 is the partial pressure of CO2 in the post-
membrane blood (mmHg), and RQ is the respiratory quotient (VO2/VCO2) [13,
14, 29].

CCAP O 2 = PCAP O 2 × 0.003 mL / mmHg + Hb × 1.34 × SCAP O 2 (9.6)

Then a shunt fraction, QS/QB, can be estimated:

QS / QB = ( CCAP O 2 − CPOST O 2 ) / ( CCAP O 2 − CPRE O 2 ) (9.7)

Frequent evaluation of QS/QB can be cumbersome based on the necessary moni-

toring equipment and is less utilized compared to a calculation of the V O2,ML. An
alternative simplified approach involves calculating the ML P/F ratio (mmHg) by
9 Membrane Dysfunction 177

dividing the PPOSTO2 by FSO2 as this has been shown to be inversely proportional to
QS/QB [13, 30, 31]. However, a progressive increase in the ML shunt fraction as well
as sustained ratio >30% has been shown to correlate with overall ML aging and gas
transfer efficiency [32].

ML Dysfunction

Optimal timing as well as specific criteria for an ECMO circuit change is a nonstan-
dardized practice between various centers. While multiple technical or patient-­
specific factors can contribute to the decision, one essential element is the evaluation
of ML efficiency. Prompt recognition of ML dysfunction allows elective replace-
ment, as described below, and reduces the likelihood of emergent exchange with its
inherent risks. However, premature replacement of an adequately functioning cir-
cuit also generates risk [16]. Membrane lung exchange may be indicated by one or
more of the following factors: (a) circuit-associated hematologic abnormalities, (b)
progressive blood flow obstruction, or (c) inadequate gas exchange (Fig. 9.1).

Coagulopathy Hemolysis
↑ aPTT
↑ anti-Xa ↑ LDH
1 ↓ Plt
D-Dimer > 25-30 mg/dL
↑PfHb
Hemolysis Index
Fibrinogen < 200 mg/dL Hemoglobinuria
Hematologic ↑ TEG® R –Time
Profile

Pressure Monitoring Obstructed Blood Flow

∆PML= PPREML – PPOSTML Progressive increase in RML ML Dysfunction;

2 Consideration for
RML= ∆PML/QB Increased pump Circuit Exchange
speed to maintain a
Mechanical stable blood flow rate
Obstruction Membrane Lung

Impaired O2Transfer Reduced CO2Clearance

. .
VO2, ML= QB (CPOSTO2– VCO2, ML= QG x FOCO2
CPREO2) x 10

3 PPOSTO2 < 200 mmHg &

.
a) PPOSTCO2 > 40 mmHg
VO2, ML< 100 –150 b) PPRECO2 – PPOSTCO2<
ml/min after optimization 10 mmHg with QG≥
Impaired Gas
Exchange of QB, CPREO2, and FSO2 10L/min

Fig. 9.1 ML dysfunction is recognized through changes in hematologic profile, pressure across
the ML, or impaired gas exchange. aPTT activated partial thromboplastin time, anti-Xa anti-factor
Xa, Plt platelet count, TEG R-time thromboelastography reaction time to initial thrombin forma-
tion, LDH lactate dehydrogenase, PfHb plasma-free hemoglobin, ΔPML pressure difference across
the ML, PPREML inlet pressure, PPOSTML outlet pressure, RML resistance across the ML, QB circuit
blood flow, V O2, ML oxygen transfer across the ML, CPOSTO2 outlet oxygen content, CPREO2 inlet
oxygen content, PPOSTO2 outlet oxygen partial pressure, FSO2 oxygen fraction in the sweep gas, V
CO2, ML carbon dioxide transfer across the membrane, QG sweep gas flow, FOCO2 gas outlet fraction
of carbon dioxide, PPOSTCO2 outlet partial pressure of carbon dioxide, PPRECO2 inlet partial pres-
sure of carbon dioxide
178 B. D. Warren et al.

Hematologic Abnormalities

Hematologic derangements can serve as a clinical indicator of impending circuit

dysfunction. Given the predilection for concurrent bleeding, hemolytic, and throm-
botic complications, it is important to evaluate all aspects of ECMO-related hema-
tologic abnormities when assessing the health of the ML. The literature offers
variable definitions for ECMO-induced coagulopathy; however, the following are
integral laboratory markers that should be assessed with the goal of early identifica-
tion of ML dysfunction. Management and monitoring of anticoagulation during
ECMO have been previously well-reviewed [5, 19, 33]. There is considerable varia-
tion among ECMO centers with regard to choice of anticoagulant, monitoring strat-
egies, and therapeutic targets [34]. The majority of centers use unfractionated
heparin (UFH) based primarily on its familiarity. However, UFH does carry
increased risk for heparin-induced thrombocytopenia [35] and heparin resistance
via depletion of antithrombin [19, 36–38]. Additional concerns with UFH use
include interlaboratory aPTT monitoring variability [39], discordance between
aPTT vs anti-Xa values [40], and impaired accuracy secondary to antiphospholipid
antibodies or hyperfibrinolysis.
Low-molecular-weight heparin (LMWH) has increased factor Xa inhibition
compared to UFH [41] but is less commonly used in ECMO [42] secondary to poor
renal clearance, longer half-life, and the requirement of anti-Xa level monitoring.
However, LMWH is easier to dose and administer and less likely to cause
HIT. Finally, direct thrombin inhibitors (DTIs; argatroban and bivalirudin) function
independent of antithrombin, do not induce the production of anti-platelet antibod-
ies, and have a short half-life (30 min bivalirudin; 40 min for argatroban) [43].
Accurate monitoring of DTIs is accomplished with a combination of anti-Xa levels
and thromboelastography (TEG®) [44]. Anti-Xa concentrations reflect the magni-
tude of activated factor X (FXa) inhibition and likely better correlate with the
degree of anticoagulation than aPTT [45]. Notable limitations include false reduc-
tion in the presence of hyperbilirubinemia and elevated plasma free-hemoglobin
[46]. TEG® data represents total hemostasis from initial platelet aggregation
through fibrinolysis. In the context of ECMO, the TEG® R-time (latency from start
of test to initial fibrin formation) has the strongest correlation with aPTT values
[47] and has been proven safe with fewer thrombotic complications and a lower
heparin dose requirement when compared to aPTT monitoring [48, 49].
Conversion of fibrinogen to insoluble fibrin is the concluding event of the coagu-
lation cascade. Through various mechanisms, ECLS predisposes patients to both a
consumptive coagulopathy and hyperfibrinolysis with increased incidence of hypo-
fibrinogenemia [50]. Fibrinogen monitoring signals both coagulation factor con-
sumption and hemorrhage risk when levels fall below 200 mg/dL. Importantly,
values should be assessed serially as fibrinogen is an acute phase reactant that is
elevated in the setting of inflammation, rendering relative changes as important as
absolute values.
9 Membrane Dysfunction 179

The D-dimer (DD) is a degradation product of cross-linked fibrin that signi-

fies fibrinolytic activity. DD values greater than 25–30 mg/dL in the absence of
an alternative source suggest clot accumulation within the membrane and cor-
relate with decremental oxygenator efficiency that improves after ML exchange
[51, 52].
Evaluation of platelets in the setting of ECMO is complex as they have dual
functionality as mediators of hemostasis and inflammation (via release of granular
content) [53, 54]. Further, patients on ECMO are uniquely prone to both consump-
tive thrombocytopenia and a qualitative platelet dysfunction (acquired von
Willebrand factor syndrome [55]). Extracorporeal support, in isolation, does not
correlate with progressive thrombocytopenia or severe thrombocytopenia (≤50,000/
μL). Rather, severity of illness and platelet count at the time of cannulation are
strong predictors of developing severe thrombocytopenia [56]. Ultimately there is
no defined threshold for thrombocytopenia that would trigger an oxygenator
exchange in isolation, but acute onset thrombocytopenia without an alternative eti-
ology can suggest pump head thrombosis or increase thrombus burden within the
oxygenator [57].
Red blood cell destruction during ECMO is a common occurrence with nearly
70% of patients having elevated levels of plasma-free hemoglobin [18, 58] and
approximately 18% having clinically significant hemolysis [59]. Primary drivers of
hemolysis include underlying disease, high negative pressure at the pump intake
[60], thrombotic deposition within the membrane oxygenator [61], and high circuit
blood flow rates through small diameter cannulas [23]. Traditional hemolytic mark-
ers include lactate dehydrogenation (LDH) and plasma-free hemoglobin (PfHb).
Importantly, PfHb has cytotoxic properties resulting in tissue hypoxia [49] as well
as an ability to scavenge nitric oxide causing vasoconstriction, platelet aggregation,
and endothelial dysfunction [62, 63]. Clinically, elevated PfHb correlates with
increased overall mortality and ECMO duration. Values greater 1000 mg/L are asso-
ciated with increased need for renal replacement therapy and acute thrombosis of
the ECMO circuit pump head [18]. The predominant limitation with PfHb monitor-
ing is the time delay for the results as this is a send-out study for many institutions.
Alternative, though less elegant, screening tools for hemolysis include the hemoly-
sis index by spectrophotometry [64] and a bedside assessment for hemoglobinuria.
We consider rising PfHb values above 50 mg/L without alternative etiology con-
cerning for circuit-derived hemolysis which should prompt consideration of
exchange [16].
Consumptive coagulopathy (prolonged aPTT, hypofibrinogenemia, thrombocy-
topenia, and elevated D-dimer) and hemolysis (elevated PfHb and LDH) without
alternate etiology raise concern for a circuit-related hematologic disorder.
Differentiating ECMO-specific findings from those related to the patient’s primary
disease is only confirmed after the circuit is replaced and the values normalize [30].
We monitor D-dimer, platelets, fibrinogen, and PfHb to diagnose circuit-related
coagulopathy.
180 B. D. Warren et al.

Obstruction to Blood Flow

Increasing ΔPML/QB suggests obstruction to blood flow secondary to thrombus

development within the ML. Intrinsic blood flow resistance (RML), generally mini-
mal, is variable between the different manufactures based on size, design, and lay-
out of the microfiber [28]. Consequently, there is no standardized cut-off value for
ΔP that defines ML dysfunction. Membrane exchange criteria based on ΔP largely
ranges between 30 and 50 mmHg; however acute changes and overall pressure trend
profiles are likely more clinically relevant [15, 16]. A rapid increase in ΔPML/QB is
always concerning and membrane exchange should be considered even in the
absence of other indicators of ML dysfunction. In the absence of pre- and postmem-
brane lung pressure measurements, increasing pump speed to maintain a stable
blood flow rate should be interpreted as an increasing ΔP [22].

Inadequate Oxygen Uptake

Persistent hypoxemia in the presence of a post-oxygenator PO2 < 200 mmHg should
prompt calculation of V O2,ML [30] (see Eqs. (9.3) and (9.4) above).
Before calculating V O2,ML, conditions must be optimized to maximize oxygen
transfer with potential targets including circuit blood flow, blood oxygen content
entering the circuit (CPREO2), and the sweep gas oxygen concentration [16]. Elevated
CPREO2 due to recirculation or impaired tissue extraction will limit oxygen transfer.
Oxygen fraction of the sweep gas (FSO2) provides the gradient for transfer such that
values below 1.0 can limit maximal transfer. If V O2,ML is <100–150 ml/min after
optimization in a patient with sustained hypoxemia, ML exchange is indicated [16,
23, 26].

Inadequate Carbon Dioxide Clearance

ECMO facilitates CO2 clearance by generating a CO2 pressure gradient between the
transiting blood and the gas phase within the ML. Carbon dioxide removal ( V
CO2,ML) is proportional to the sweep gas flow, largely independent of the blood flow
rate through the circuit [65], and adversely affected by regions of increased clot
burden that allow for ventilation without adequate perfusion (ML dead space) [26].
The calculation for V CO2,ML is shown in Eq. (9.8) [13, 66].

VCO 2, ML = QG × FO CO 2 × 1000 (9.8)

where QG is sweep gas flow, FOCO2 is sweep gas outlet fraction of CO2, and 1000
is mL/L.
9 Membrane Dysfunction 181

While ML dysfunction can present as inadequate or declining CO2 clearance, we

do not routinely calculate V CO2,ML since this requires sampling ML outlet CO2. An
alternative clue to inadequate CO2 clearance is persistent PPOSTCO2 greater than
40 mmHg despite sweep gas flow rates above 10 L/min. Substantially impaired CO2
clearance should prompt consideration of ML exchange [16].

Exchanging the ML

Ideally, ML exchange can be planned semi-electively based on recognizing the

hematologic, pressure, or gas exchange abnormalities described above. Replacing
the ML entails risk even in the best of circumstances, more so in an emergency.
Exchange requires that the patient be temporarily removed from ECMO, so the time
required must be kept to a minimum. Simulation training provides a foundation for
the coordination necessary for a safe and expeditious procedure. An experienced
team can replace the ML in well under 1 min.
When time allows, the team should anticipate the consequences of a brief period
of ineffective gas exchange by preparing to adjust the ventilator, infuse vasoactive
drugs, give additional doses of sedative or narcotics, and reassure the awake patient.
When the replacement ML is ready and the team is in position, the circuit is clamped
near the ML (pre and post), pump turned off, tubing cut using sterile scissors, and
the new ML inserted. Clamps are removed, gas and heater lines are connected to the
new ML, and the pump is gradually restored to appropriate speed. Pump speed and
sweep gas flow settings may need adjustment from pre-exchange values since the
new ML is expected to function more efficiently. This is a good opportunity to
assess the entire circuit thoroughly, make sure connectors are tight and ty-bands are
intact, and that circuit clamps are returned to their proper location so that they can
be located quickly in any emergency (see Chap. 12).

References

1. Brodie D, Slutsky AS, Combes A. Extracorporeal life support for adults with respiratory fail-
ure and related indications: a review. JAMA. 2019;322(6):557–68.
2. Davies A, Jones D, Bailey M, Beca J, Bellomo R, Blackwell N, et al. Extracorporeal mem-
brane oxygenation for 2009 influenza A (H1N1) acute respiratory distress syndrome.
JAMA. 2009;302:1888–95.
3. Extracorporeal life support registry report. Available online: https://www.elso.org/Registry/
Statistics/InternationalSummary.aspx. Accessed on 15 Aug 2021.
4. Aubron C, Cheng AC, Pilcher D, Leong T, Magrin G, Cooper DJ, et al. Factors associated with
outcomes of patients on extracorporeal membrane oxygenation support: a 5-year cohort study.
Crit Care. 2013;17:R73.
5. Mulder MM, Fawzy I, Lancé MD. ECMO and anticoagulation: a comprehensive review. Neth
J Crit Care. 2018;26:6–13.
182 B. D. Warren et al.

6. Rastan AJ, Lachmann N, Walther T, Doll N, Gradistanac T, Gommert JF, et al. Autopsy find-
ings in patients on postcardiotomy extracorporeal membrane oxygenation (ECMO). Int J Artif
Organs. 2006;29:1121–31.
7. Betit P. Technical advances in the field of ecmo. Respir Care. 2018;63(9):1162–73. https://doi.
org/10.4187/respcare.06320.
8. Ündar A, Wang S, Palanzo DA. Impact of Polymethylpentene OXYGENATORS on out-
comes of all extracorporeal life support patients in the United States. Artif Organs.
2013;37(12):1080–1. https://doi.org/10.1111/aor.12242.
9. Besser MW, Klein AA. The coagulopathy of cardiopulmonary bypass. Crit Rev Clin Lab Sci.
2010;47(5–6):197–212. https://doi.org/10.3109/10408363.2010.549291.
10. Gorbet MB, Sefton MV. Biomaterial-associated thrombosis: roles of coagulation factors,
complement, platelets and leukocytes. Biomaterials. 2004;25(26):5681–703. https://doi.
org/10.1016/j.biomaterials.2004.01.023.
11. Oliver WC. Anticoagulation and coagulation management for ECMO. Semin Cardiothorac
Vasc Anesth. 2009;13(3):154–75. https://doi.org/10.1177/1089253209347384.
12. Millar JE, Fanning JP, McDonald CI, McAuley DF, Fraser JF. The inflammatory response to
extracorporeal membrane oxygenation (ecmo): a review of the pathophysiology. Crit Care.
2016;20(1):1–10. https://doi.org/10.1186/s13054-­016-­1570-­4.
13. Epis F, Belliato M. Oxygenator performance and artificial-native lung interaction. J Thorac
Dis. 2018;10(S5):S596–605. https://doi.org/10.21037/jtd.2017.10.05.
14. Ficial B, Vasques F, Zhang J, Whebell S, Slattery M, Lamas T, Daly K, Agnew N, Camporota
L. Physiological basis of extracorporeal membrane oxygenation AND extracorporeal carbon
dioxide removal in respiratory failure. Membranes. 2021;11(3):225. https://doi.org/10.3390/
membranes11030225.
15. Lehle K, Philipp A, Müller T, Schettler F, Bein T, Schmid C, Lubnow M. Flow dynamics
of different ADULT Ecmo systems: a clinical evaluation. Artif Organs. 2013;38(5):391–8.
https://doi.org/10.1111/aor.12180.
16. Zakhary B, Vercaemst L, Mason P, Antonini MV, Lorusso R, Brodie D. How I approach mem-
brane lung dysfunction in patients receiving ECMO. Crit Care. 2020;24(1):1–4. https://doi.
org/10.1186/s13054-­020-­03388-­2.
17. Doyle AJ, Hunt BJ. Current understanding of how extracorporeal membrane oxygenators acti-
vate haemostasis and other blood components. Front Med. 2018;5:352. https://doi.org/10.3389/
fmed.2018.00352.
18. Pan KC, McKenzie DP, Pellegrino V, Murphy D, Butt W. The meaning of a high PLASMA
free haemoglobin: retrospective review of the prevalence of haemolysis and circuit throm-
bosis in an ADULT ECMO Centre over 5 years. Perfusion. 2015;31(3):223–31. https://doi.
org/10.1177/0267659115595282.
19. Chlebowski MM, Baltagi S, Carlson M, Levy JH, Spinella PC. Clinical controversies in antico-
agulation monitoring and antithrombin supplementation for ecmo. Crit Care. 2020;24(1):1–2.
https://doi.org/10.1186/s13054-­020-­2726-­9.
20. Despotis GJ, Avidan MS, Hogue CW. Mechanisms and attenuation of hemostatic activation
during extracorporeal circulation. Ann Thorac Surg. 2001;72:S1821–31.
21. Esper SA, Levy JH, Waters JH, Welsby IJ. Extracorporeal membrane oxygenation in the adult:
a review of anticoagulation monitoring and transfusion. Anesth Analg. 2014;118:731–43.
22. Lehle K, Philipp A, Gleich O, Holzamer A, Müller T, Bein T, Schmid C. Efficiency in
extracorporeal MEMBRANE oxygenation—cellular deposits on polymethylpentene
MEMBRANES increase resistance to blood flow and reduce gas exchange capacity. ASAIO
J. 2008;54(6):612–7. https://doi.org/10.1097/mat.0b013e318186a807.
23. Lehle K, Philipp A, Hiller KA, et al. Efficiency of gas transfer in venovenous extracorporeal
membrane oxygenation: analysis of 317 cases with four different ECMO systems. Intensive
Care Med. 2014;40:1870–7. https://doi.org/10.1007/s00134-­014-­3489-­z.
24. Park M, Costa EL, Maciel AT, Silva DP, Friedrich N, Barbosa EV, Hirota AS, Schettino G,
Azevedo LC. Determinants of oxygen and carbon DIOXIDE transfer during extracorporeal
9 Membrane Dysfunction 183

membrane oxygenation in an experimental model of multiple organ dysfunction syndrome.

PLoS One. 2013;8(1):e54954. https://doi.org/10.1371/journal.pone.0054954.
25. Schmidt M, Tachon G, Devilliers C, Muller G, Hekimian G, Bréchot N, Merceron S, Luyt
CE, Trouillet J-L, Chastre J, Leprince P, Combes A. Blood oxygenation and decarboxylation
determinants during venovenous ecmo for respiratory failure in adults. Intensive Care Med.
2013;39(5):838–46. https://doi.org/10.1007/s00134-­012-­2785-­8.
26. Scaravilli V, Zanella A, Sangalli F, Patroniti N. Basic aspects of physiology during ECMO
support. In: ECMO-extracorporeal life support in adults. Milano: Springer; 2014. p. 19–36.
https://doi.org/10.1007/978-­88-­470-­5427-­1_3.
27. Xie A, Yan TD, Forrest P. Recirculation in venovenous extracorporeal membrane oxygenation.
J Crit Care. 2016;36:107–10. https://doi.org/10.1016/j.jcrc.2016.05.027.
28. Jegger D, Tevaearai HT, Mallabiabarrena I, Horisberger J, Seigneul I, von Segesser
LK. Comparing oxygen transfer performance between three membrane oxygenators: effect
of temperature changes during cardiopulmonary bypass. Artif Organs. 2007;31(4):290–300.
https://doi.org/10.1111/j.1525-­1594.2007.00379.x.
29. Isgrò S, Mojoli F, Avalli L. Monitoring the ECMO patient: the extracorporeal circuit. In:
ECMO-extracorporeal life support in adults. Milano: Springer; 2014. p. 401–11. https://doi.
org/10.1007/978-­88-­470-­5427-­1_35.
30. Lubnow M, Philipp A, Foltan M, Bull Enger T, Lunz D, Bein T, Haneya A, Schmid C, Riegger
G, Müller T, Lehle K. Technical complications during veno-venous extracorporeal membrane
oxygenation and their relevance predicting a system-exchange–retrospective analysis of 265
cases. PLoS One. 2014;9(12):e112316. https://doi.org/10.1371/journal.pone.0112316. PMID:
25464516; PMCID: PMC4251903.
31. Philipp A, De Somer F, Foltan M, Bredthauer A, Krenkel L, Zeman F, Lehle K. Life span of
different extracorporeal membrane systems for severe respiratory failure in the clinical prac-
tice. PLoS One. 2018;13(6):e0198392. https://doi.org/10.1371/journal.pone.0198392.
32. Panigada M, L’Acqua C, Passamonti SM, Mietto C, Protti A, Riva R, Gattinoni L. Comparison
between clinical indicators of transmembrane oxygenator thrombosis and multidetec-
tor computed tomographic analysis. J Crit Care. 2015;30(2):441. https://doi.org/10.1016/j.
jcrc.2014.12.005.
33. ELSO Anticoagulation Guidelines 2014. Extracorporeal life support organization. Available
online: http://www.elso.org/Portals/0/Files/elsoanticoagulationguideline8-­2014-­table-­
contents.pdf. Accessed 15 Aug 2021.
34. Bembea MM, Annich G, Rycus P, Oldenburg G, Berkowitz I, Pronovost P. Variability in anti-
coagulation management of patients on extracorporeal membrane oxygenation. Pediatr Crit
Care Med. 2013;14(2):e77. https://doi.org/10.1097/pcc.0b013e31827127e4.
35. Glick D, Dzierba AL, Abrams D, Muir J, Eisenberger A, Diuguid D, Abel E, Agerstrand C,
Bacchetta M, Brodie D. Clinically suspected heparin-induced thrombocytopenia during extra-
corporeal membrane oxygenation. J Crit Care. 2015;30(6):1190–4. https://doi.org/10.1016/j.
jcrc.2015.07.030.
36. Iapichino GE, Protti A, Andreis DT, Panigada M, Artoni A, Novembrino C, Pesenti A,
Gattinoni L. Antithrombin during extracorporeal membrane oxygenation in adults: national
survey and retrospective analysis. ASAIO J. 2019;65(3):257–63. https://doi.org/10.1097/
mat.0000000000000806.
37. Levy JH, Connors JM. Heparin resistance — CLINICAL perspectives and management strate-
gies. N Engl J Med. 2021;385(9):826–32. https://doi.org/10.1056/nejmra2104091.
38. Protti A, Iapichino GE, Di Nardo M, Panigada M, Gattinoni L. Anticoagulation management
and antithrombin supplementation practice during veno-venous extracorporeal membrane oxy-
genation: a worldwide survey. Anesthesiology. 2020;132(3):562–70. https://doi.org/10.1097/
ALN.0000000000003044. PMID: 31764152.
39. Olson JD, Arkin CF, Brandt JT, Cunningham MT, Giles A, Koepke JA, Witte DL. College
of American Pathologists Conference XXXI on laboratory monitoring of anticoagulant
184 B. D. Warren et al.

therapy: laboratory monitoring of unfractionated heparin therapy. Arch Pathol Lab Med.
1998;122(9):782–98. PMID: 9740136.
40. Liveris A, Bello RA, Friedmann P, Duffy MA, Manwani D, Killinger JS, Rodriquez D,
Weinstein S. Anti-factor Xa assay is a superior correlate of heparin dose than activated
partial thromboplastin time or activated clotting time in pediatric extracorporeal mem-
brane oxygenation. Pediatr Crit Care Med. 2014;15(2):e72–9. https://doi.org/10.1097/
PCC.0000000000000028. PMID: 24335992.
41. Alban S. Pharmacological strategies for inhibition of thrombin activity. Curr Pharm Des.
2008;14(12):1152–75. https://doi.org/10.2174/138161208784246135. PMID: 18473863.
42. Krueger K, Schmutz A, Zieger B, Kalbhenn J. Venovenous extracorporeal membrane oxygen-
ation with prophylactic subcutaneous anticoagulation only: an observational study in more
than 60 patients. Artif Organs. 2017;41(2):186–92. https://doi.org/10.1111/aor.12737. PMID:
27256966.
43. Di Nisio M, Middeldorp S, Büller HR. Direct thrombin inhibitors. N Engl J Med.
2005;353(10):1028–40. https://doi.org/10.1056/NEJMra044440. Erratum in: N Engl J Med.
2005 Dec 29;353(26):2827. PMID: 16148288.
44. Colman E, Yin EB, Laine G, Chatterjee S, Saatee S, Herlihy JP, Reyes MA, Bracey
AW. Evaluation of a heparin monitoring protocol for extracorporeal membrane oxygenation
and review of the literature. J Thorac Dis. 2019;11(8):3325–35. https://doi.org/10.21037/
jtd.2019.08.44. PMID: 31559035; PMCID: PMC6753426.
45. Delmas C, Jacquemin A, Vardon-Bounes F, Georges B, Guerrero F, Hernandez N, Marcheix
B, Seguin T, Minville V, Conil JM, Silva S. Anticoagulation monitoring under ECMO support:
a comparative study between the activated coagulation time and the anti-Xa activity assay. J
Intensive Care Med. 2020;35(7):679–86. https://doi.org/10.1177/0885066618776937. PMID:
29768983.
46. Kostousov V, Nguyen K, Hundalani SG, Teruya J. The influence of free hemoglobin and bili-
rubin on heparin monitoring by activated partial thromboplastin time and anti-Xa assay. Arch
Pathol Lab Med. 2014;138(11):1503–6. https://doi.org/10.5858/arpa.2013-­0572-­OA. PMID:
25357112.
47. Alexander DC, Butt WW, Best JD, Donath SM, Monagle PT, Shekerdemian LS. Correlation
of thromboelastography with standard tests of anticoagulation in paediatric patients receiv-
ing extracorporeal life support. Thromb Res. 2010;125(5):387–92. https://doi.org/10.1016/j.
thromres.2009.07.001. PMID: 19674773.
48. Panigada M, E Iapichino G, Brioni M, Panarello G, Protti A, Grasselli G, Occhipinti G,
Novembrino C, Consonni D, Arcadipane A, Gattinoni L. Thromboelastography-based anti-
coagulation management during extracorporeal membrane oxygenation: a safety and feasibil-
ity pilot study. Ann Intensive Care. 2018;8(1):7. https://doi.org/10.1186/s13613-­017-­0352-­8.
PMID: 29340875; PMCID: PMC5770349.
49. Rother RP, Bell L, Hillmen P, Gladwin MT. The clinical sequelae of intravascular hemo-
lysis and extracellular plasma hemoglobin: a novel mechanism of human disease.
JAMA. 2005;293(13):1653–62. https://doi.org/10.1001/jama.293.13.1653. PMID: 15811985.
50. Boggio LN, Simpson ML. Fibrinogen replacement underutilization in ECMO. Blood.
2020;136(Supplement 1):4–5. https://doi.org/10.1182/blood-­2020-­142812.
51. Dornia C, Philipp A, Bauer S, Stroszczynski C, Schreyer AG, Müller T, Koehl GE, Lehle
K. D-dimers are a predictor of clot volume inside membrane oxygenators during extracor-
poreal membrane oxygenation. Artif Organs. 2015;39(9):782–7. https://doi.org/10.1111/
aor.12460. PMID: 25845704.
52. Lubnow M, Philipp A, Dornia C, Schroll S, Bein T, Creutzenberg M, Diez C, Schmid C, Pfeifer
M, Riegger G, Müller T, Lehle K. D-dimers as an early marker for oxygenator exchange in
extracorporeal membrane oxygenation. J Crit Care. 2014;29(3):473. https://doi.org/10.1016/j.
jcrc.2013.12.008.
53. Kraft F, Schmidt C, Van Aken H, Zarbock A. Inflammatory response and extracorporeal cir-
culation. Best Pract Res Clin Anaesthesiol. 2015;29(2):113–23. https://doi.org/10.1016/j.
bpa.2015.03.001. PMID: 26060024.
9 Membrane Dysfunction 185

54. Whiteheart SW. Platelet granules: surprise packages. Blood. 2011;118(5):1190–1. https://doi.
org/10.1182/blood-­2011-­06-­359836. PMID: 21816838.
55. Kalbhenn J, Schmidt R, Nakamura L, Schelling J, Rosenfelder S, Zieger B. Early diagno-
sis of acquired von Willebrand Syndrome (AVWS) is elementary for clinical practice in
patients treated with ECMO therapy. J Atheroscler Thromb. 2015;22(3):265–71. https://doi.
org/10.5551/jat.27268. PMID: 25186021.
56. Abrams D, Baldwin MR, Champion M, Agerstrand C, Eisenberger A, Bacchetta M, Brodie
D. Thrombocytopenia and extracorporeal membrane oxygenation in adults with acute respira-
tory failure: a cohort study. Intensive Care Med. 2016;42(5):844–52. https://doi.org/10.1007/
s00134-­016-­4312-­9. PMID: 27007099; PMCID: PMC5407307.
57. Fisser C, Winkler M, Malfertheiner MV, Philipp A, Foltan M, Lunz D, Zeman F, Maier LS,
Lubnow M, Müller T. Argatroban versus heparin in patients without heparin-induced throm-
bocytopenia during venovenous extracorporeal membrane oxygenation: a propensity-score
matched study. Crit Care. 2021;25(1):160. https://doi.org/10.1186/s13054-­021-­03581-­x.
58. Dufour N, Radjou A, Thuong M. Hemolysis and plasma free hemoglobin during extracorpo-
real membrane oxygenation support: from clinical implications to laboratory details. ASAIO
J. 2020;66(3):239–46. https://doi.org/10.1097/MAT.0000000000000974. PMID: 30985331.
59. Zangrillo A, Landoni G, Biondi-Zoccai G, Greco M, Greco T, Frati G, Patroniti N, Antonelli
M, Pesenti A, Pappalardo F. A meta-analysis of complications and mortality of extracorporeal
membrane oxygenation. Crit Care Resusc. 2013;15(3):172–8. PMID: 23944202.
60. Horton AM, Butt W. Pump-induced haemolysis: is the constrained vortex pump
better or worse than the roller pump? Perfusion. 1992;7(2):103–8. https://doi.
org/10.1177/026765919200700204.
61. Kawahito S, Maeda T, Motomura T, Ishitoya H, Takano T, Nonaka K, Linneweber J, Ichikawa
S, Kawamura M, Hanazaki K, Glueck J, Nosé Y. Hemolytic characteristics of oxygenators
during clinical extracorporeal membrane oxygenation. ASAIO J. 2002;48(6):636–9. https://
doi.org/10.1097/00002480-­200211000-­00010. PMID: 12455774.
62. Jeney V, Balla J, Yachie A, Varga Z, Vercellotti GM, Eaton JW, Balla G. Pro-oxidant and cyto-
toxic effects of circulating heme. Blood. 2002;100(3):879–87. https://doi.org/10.1182/blood.
v100.3.879. PMID: 12130498.
63. Schaer DJ, Buehler PW, Alayash AI, Belcher JD, Vercellotti GM. Hemolysis and free hemo-
globin revisited: exploring hemoglobin and hemin scavengers as a novel class of therapeu-
tic proteins. Blood. 2013;121(8):1276–84. https://doi.org/10.1182/blood-­2012-­11-­451229.
PMID: 23264591; PMCID: PMC3578950.
64. Marques-Garcia F. Methods for hemolysis interference study in laboratory medicine – a criti-
cal review. EJIFCC. 2020;31(1):85–97. PMID: 32256292; PMCID: PMC7109502.
65. Kolobow T, Gattinoni L, Tomlinson TA, Pierce JE. Control of breathing using an
extracorporeal membrane lung. Anesthesiology. 1977;46(2):138–41. https://doi.
org/10.1097/00000542-­197702000-­00012. PMID: 13683.
66. Zanella A, Mangili P, Giani M, Redaelli S, Scaravilli V, Castagna L, Sosio S, Pirrone F,
Albertini M, Patroniti N, Pesenti A. Extracorporeal carbon dioxide removal through
ventilation of acidified dialysate: an experimental study. J Heart Lung Transplant.
2014;33(5):536–41. https://doi.org/10.1016/j.healun.2013.12.006. PMID: 24439968.
Chapter 10
ECCO2R in Obstructive Diseases:
Evidence, Indications, and Exclusions

Lorenzo Del Sorbo, V. Marco Ranieri, and Vito Fanelli

Introduction

In the last decade, there has been an impressive technological and clinical evolution
of extracorporeal life support (ECLS) strategies [1–4]. To date, a variety of modali-
ties of ECLS have been developed, characterized by different configurations, differ-
ent vascular accesses, different rates of blood flow, and different clinical goals [5].
In particular, remarkable interest has been focused on extracorporeal carbon dioxide
removal (ECCO2R) [6, 7], due to the relative ease and efficiency in blood CO2 clear-
ance granted by extracorporeal gas exchangers as compared to oxygen delivery.
CO2 present in blood, mainly dissociated into bicarbonates and hydrogen ions and
bound to hemoglobin, is characterized by a steep, linear dissociation kinetic and
high solubility, which provide an elevated diffusion rate across the membrane lung
(ML). Therefore, ventilation (sweep gas flow) rather than perfusion (extracorporeal
blood flow rate) of the ML and its total surface area is the main determinant of the
CO2 clearance from the blood. Since CO2 is, on average, metabolically produced at
the rate of 200–250 ml/min, ECCO2R systems running blood flow through the ML
even as low as 0.5–1 L/min may theoretically remove the entire CO2 production
(without providing significant oxygenation) [8–11].

L. Del Sorbo (*)

Interdepartmental Division of Critical Care Medicine, University Health Network, University
of Toronto, Toronto, ON, Canada
e-mail: [email protected]
V. M. Ranieri
Department of Medical and Surgical Sciences (DIMEC), Alma Mater Studiorum, University
of Bologna, Bologna, Italy
V. Fanelli
Dipartimento di Anestesiologia e Rianimazione, Azienda Ospedaliera Città della Salute e
della Scienza di Torino, Università di Torino, Torino, Italy

© The Author(s), under exclusive license to Springer Nature 187

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_10
188 L. Del Sorbo et al.

These features have been exploited by the industry and technologically trans-
lated in clinical strategies of ECLS providing ECCO2R that are remarkably less
invasive as compared to the full extracorporeal membrane oxygenation (ECMO),
easier to manage, and as efficient in CO2 clearance [6]. Several ECCO2R devices
have been recently approved, at least in Europe, and are available for clinical use.
In most of these systems, the blood is drawn from a central vein by a draining
cannula, driven with a centrifugal or roller pump through the ML, where CO2 is
removed by diffusion sustained by a sweep gas flow, and returned into the venous
circulation usually through a second lumen of the draining cannula (Fig. 10.1a).
Hence, central venous access with a dual-lumen cannula is commonly required by
most ECCO2R systems. It is nonetheless possible to set an ECCO2R circuit so that
two central veins, one for drainage and the other for reinfusion, are accessed
(Fig. 10.1b). However, reducing the dimensions and number of central venous

a Veno-venous ECCO2-R
with one endovascular dual-lumen catheter

L/min

Sweep O2
gas flow

Gas
exchanger

Blood
flow
CO2

Pump Double
lumen
cannula
0.3-1 L/min
Console

Fig. 10.1 Possible configurations of extracorporeal carbon dioxide removal (ECCO2R) circuits.
(a) Minimally invasive veno-venous ECCO2R circuit with a single venous vascular access through
a dual-lumen cannula that can be typically inserted in the internal jugular vein or in the femoral
vein. (b) Veno-venous ECCO2R circuit with two separate venous vascular accesses, with the flow
usually running from femoral vein to the jugular vein or from femoral vein to femoral vein. (c)
Pumpless arteriovenous ECCO2R device with the typical placement of the ML connecting the
femoral artery with the contralateral femoral vein
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 189

Veno-venous ECCO2-R
b with double vascular access

L/min

Sweep O2
gas flow

Gas
exchanger

Blood
flow
CO2

Pump Double
lumen
cannula
0.3-1 L/min
Console
Arterio-venous ECCO2-R
c

Blood
flow
Gas
exchanger
L/min Sweep
gas flow

O2 CO2

Fig. 10.1 (continued)

190 L. Del Sorbo et al.

access sites, as well as the degree of anticoagulation, but maintaining or increasing

the efficiency of CO2 removal, has been the key factor in progressively minimizing
the invasiveness of these systems [6].
An alternative configuration of ECCO2R involves inserting an arteriovenous
(AV) bypass, usually by connecting the femoral artery to the contralateral femoral
vein, with an interposed ML (Fig. 10.1c). In this system, blood flows through the
ML based on the AV pressure gradient, not on a pump, thus requiring significantly
less machinery. Although this approach is efficient, it is more invasive and prone to
related complications [6].
Rapid technological progress and effective marketing have seen ECCO2R
applied to several different patient populations. In patients with acute respiratory
distress syndrome (ARDS) [2, 12], ECCO2R has been used to limit the hypercapnic
acidosis resulting from very low tidal volumes during ultra-protective mechanical
ventilation (see Chap. 6). Zimmermann and colleagues demonstrated that pumpless
AV-ECCO2R cleared sufficient CO2 from 51 patients with ARDS as to enable pro-
tective mechanical ventilation with tidal volumes lower than 6 cc/kg of predicted
body weight. It is worth noting that 6% of these patients suffered from ECCO2R-­
related complications, such as ischemia of the limb due to arterial cannulation [13].
In a randomized, controlled trial in patients with ARDS, low tidal volume ventila-
tion (6 mL/kg) was compared with ultra-protective mechanical ventilation (3 mL/
kg) combined with pumpless AV-ECCO2R. Although the study was not powered to
find any difference in terms of mortality or ventilator-free days, a reduced inflam-
matory response was detected in the group treated with ECCO2R [14]. Another
study in severe ARDS patients used a minimally invasive venovenous ECCO2R
device to reduce tidal volume from 6.3 ± 0.2 to 4.2 ± 0.2 mL/kg of predicted body
weight: [15] pulmonary inflammatory mediators were reduced significantly, as was
lung hyperinflation on CT.
At the same time that ECCO2R is being used more widely in ARDS, the
evolution toward progressively less invasive devices has prompted investiga-
tions into a broader spectrum of disease [1, 7]. In particular, the idea of apply-
ing ECCO2R in patients with respiratory acidosis secondary to acute
exacerbations of obstructive lung disease has gained great interest [1, 7, 16,
17]. Exacerbations of chronic obstructive pulmonary disease (COPD) and
asthma are characterized by hypercapnia, without necessarily compromising
oxygenation [18, 19]. This pathophysiological feature represents the optimal
setting for ECCO2R, potentially providing an alternative, not just an adjunct, to
mechanical ventilation as a way to support respiration [17] until the lung recov-
ers [20–22]. This chapter focuses on the main evidence for the innovative use
of ECCO2R in patients with obstructive lung diseases, along with clinical indi-
cations and exclusion criteria.
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 191

Evidence

Although the use of ECCO2R has been described in several case series of patients
with obstructive lung disease, we lack randomized trials to clearly demonstrate effi-
cacy. However, numerous clinical reports demonstrate the efficiency of this techno-
logically advanced strategy, which is based on a strong pathophysiological rationale
[17, 23]. The main feature of exacerbations of obstructive diseases is acute worsen-
ing of expiratory flow-limitation caused by increased resistance in small airways
[19, 24]. Increased resistive load is associated with a longer expiratory time con-
stant (the product of airway resistance and lung compliance), which defines the time
for exponential fall in lung volume during passive exhalation [25]. The direct con-
sequence of a longer time constant is the development of dynamic alveolar hyperin-
flation, which is also generally associated with an increase in the end-expiratory
elastic recoil of the respiratory system, also called intrinsic positive end-expiratory
pressure (Fig. 10.2) [25].
Dynamic alveolar hyperinflation and intrinsic-PEEP have detrimental effects on
the respiratory and cardiovascular system, remarkably increasing the work of
breathing, reducing venous return, and compromising cardiac output and lung per-
fusion [19, 26]. The persistence of these pathophysiological features results in
respiratory muscle fatigue and rapid shallow breathing, with the consequent reduc-
tion of alveolar ventilation (Fig. 10.2) [24, 26–28]. Since alveolar ventilation is
inversely proportional to PaCO2 (according to the equation PaCO2= V CO2/ V A,
where V CO2 is the CO2 production and V A is alveolar ventilation), acute exacerba-
tions of obstructive lung diseases produce respiratory acidosis. The rise in PaCO2,
in turn, increases ventilatory demand, which may worsen alveolar hyperinflation
and muscle fatigue, creating a vicious loop [18, 29, 30].
In this situation, noninvasive ventilation (NIV) can reduce the work of breathing
and increase alveolar ventilation, but its efficiency may not be sufficient especially
in the most severe cases [31]. Invasive mechanical ventilation (MV) also reduces the
work of breathing and is effective in correcting gas exchange but may produce
severe side effects. Moreover, with improper settings, invasive ventilation may
worsen alveolar hyperinflation, causing barotrauma, lung rupture, and hemody-
namic decompensation [19, 32–34].
ECCO2R, on the other hand, represents an innovative and physiologically attrac-
tive option, at least in theory. In contrast to NIV and invasive ventilation, whose goal
is to reduce PaCO2 by mechanically augmenting alveolar ventilation in already
hyperinflated lungs, ECCO2R reduces the alveolar ventilation required to eliminate
the CO2 production. Reduced tidal volume and respiratory rate extends expiratory
time, suiting better the high expiratory time constant characteristic of diseases with
expiratory flow limitation. By these physiological mechanisms ECCO2R can inter-
rupt the vicious circle of dynamic hyperinflation, thereby reestablishing a more
favorable balance between the work of breathing and the respiratory load in patients
with acute obstructive lung diseases exacerbations [17, 23].
192 L. Del Sorbo et al.

COPD exacerbation
a

worsening of
1. Airflow resistance
and
2. Expiratory flow limitation

Dynamic alveolar hyperinflation

PEEPi

 threshold loading of
 CDYN  VD/VT
inspiratory muscles

 pulmonary
Neuro-mechanical artery pressure
uncoupling
with dyspnea  RV preload

 LV afterload

Muscles fatigue Rapid shallow

Ventilatory pump failure breathing
1. low VT
2. high RR

Hypercapnia
and
respiratory acidosis

 ventilatory drive
and  VA demand

Fig. 10.2 (a) Schematic representation of the pathophysiological mechanisms underlying exacer-
bations of chronic obstructive pulmonary disease (COPD). (b) Schematic representation of the
pathophysiological rationale of ECCO2R application in COPD. PEEPi intrinsic positive end expi-
ratory pressure, CDYN dynamic compliance, VD dead space, VT tidal volume, VD/VT fraction of dead
space, RV right ventricle, LV left ventricle, RR respiratory rate, VA alveolar ventilation
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 193

COPD exacerbation
b

worsening of
1. Airflow resistance
and
2. Expiratory flow limitation

Dynamic alveolar hyperinflation

PEEPi

Muscles fatigue Rapid shallow

Ventilatory pump failure breathing
1. low VT
2. high RR

Hypercapnia and
respiratory acidosis

ECCO2R

 VAdemand: lower RR and  expiratory time

Resolution of the dynamic alveolar hyperinflation

and  PEEPi

 threshold loading of  pulmonary

 VD/VT inspiratory muscles artery pressure

 RV preload
 Neuro-mechanical
 CDYN uncoupling
 LV afterload

Fig. 10.2 (continued)

194 L. Del Sorbo et al.

Indications

The strong pathophysiological rationale for using ECCO2R in acute exacerbations

of obstructive lung diseases has led to a growing number of studies to verify its
clinical efficiency. These studies represent the first necessary step to translate the
physiologically sound rationale into evidence-based clinical practice.
Most of the clinical applications of ECCO2R for the treatment of obstructive lung
diseases have been reported in patients with COPD exacerbations. Nonetheless,
several case reports describe ECCO2R in patients with near-fatal asthma.

 xacerbation of Chronic Obstructive Pulmonary

E
Disease (COPD)

COPD is one of the most significant social and economic burdens in western coun-
tries, being the fourth leading cause of death [35–37]. Its natural history is charac-
terized by progressive deterioration of lung function, punctuated by a variable
number of acute exacerbations [38, 39]. COPD exacerbations are associated with
considerable morbidity and mortality, which ranges between 8% and 26% for
patients admitted to the intensive care unit [40–42].
During acute exacerbations of COPD, patients develop alveolar hyperinflation,
which in the most severe cases may lead to muscle fatigue and respiratory acidosis,
creating a vicious loop refractory to medical treatment [18, 29, 30]. Advanced strat-
egies are available to break this cycle, improve respiratory function, and gain time
for treatment of the exacerbation [43–45].
Firstly, NIV has become the standard of care for exacerbations refractory to med-
ical treatment [31, 44–46]. NIV effectively supports the respiratory muscles,
increases alveolar ventilation, reduces PaCO2, and improves oxygenation. Moreover,
externally applied PEEP counteracts intrinsic-PEEP, reducing the inspiratory
threshold load, further lowering the work of breathing and helping to reverse muscle
fatigue. Together these effects are instrumental in decreasing the need for endotra-
cheal intubation and improving survival [47–49].
A second option is invasive mechanical ventilation (MV), used when NIV fails
or altered consciousness precludes effective NIV [32, 34]. Failure of NIV occurs in
a significant percentage of cases (30%) and is associated with increased mortality
[50]. Moreover, MV is associated with a number of undesired side effects [32, 34,
51], which may produce morbidity or mortality [52]. There is good reason to believe
that avoiding endotracheal intubation and MV in patients with severe COPD exac-
erbation would substantially improve outcome.
ECCO2R is a promising, innovative strategy to improve the respiratory function
of patients with severe COPD exacerbation and obviate the need for MV [16, 17].
ECCO2R effectively lowers the alveolar ventilation required to eliminate CO2
(Fig. 10.3), thereby reestablishing a compensated balance between metabolic
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 195

PaCO2 pH Respiratory rate

140 7,6 40

* 35
120 7,5

30
100 7,4 *

Respirations / min
mmHg

pH
25
80 7,3
* 20

60 7,2
15

40 7,1 10

20 7,0 5

ICU BL 5-8 13-16 21-24 ICU BL 5-8 13-16 21-24 ICU BL 5-8 13-16 21-24
NIV 0-4 0-12 17-20 NIV 0-4 9-12 17-20 NIV 0-4 9-12 17-20
Time intervals

Fig. 10.3 Changes over time of partial pressure of arterial carbon dioxide (PaCO2), pH, and respi-
ratory rate from admission in the intensive care unit (ICU), during noninvasive ventilation (NIV),
at baseline (BL), and after initiation of extracorporeal carbon dioxide removal (ECCO2R) in
patients with hypercapnic respiratory failure treated with the pumpless arteriovenous ECCO2R
device with the goal of avoiding endotracheal intubation. The data demonstrate the significant
decrease in respiratory rate secondary to the reduction of PaCO2 facilitated by the ECCO2R device.
Reproduced with permission from reference [81]

production and the impaired respiratory system. Several reports verify this hypoth-
esis, showing the successful use of ECCO2R in chronic obstructive respiratory dis-
eases [16, 17, 53, 54].
One of the first reports on the application of ECCO2R to support respiratory
function of a COPD patient was published in 1990 by Pesenti and colleagues [55].
A 19-year-old woman with bullous emphysema, recurrent pneumothoraxes, bilat-
eral air leaks, and lung infection had failed attempts at weaning after 28 days of
MV. A CT scan revealed the presence of pneumomediastinum, subcutaneous
emphysema, and hyperinflation of multiple lung bullae, likely sustained by pro-
longed MV and the high minute volumes required for CO2 clearance. This vicious
cycle was broken by low-blood-flow (0.4–0.6 L/min) venovenous ECCO2R, which
allowed the patient to transition to a spontaneous ventilatory mode and then to full
liberation from MV. A subsequent CT showed a decrease in the size of bullae with
expansion of the healthier lung parenchyma, suggesting that ECCO2R relieved
dynamic hyperinflation.
Since this first description, a number of observational studies have been reported
on the clinical application of ECCO2R in patients with COPD exacerbations to
facilitate early liberation from MV or to avoid MV after NIV failure [53, 56].
Devices characterized by venovenous configurations with a dual-lumen cannula
similar in size to a dialysis catheter are used in most of these clinical applications.
Abrams and colleagues [57] reported five older patients (age 73 ± 8.7 years) with
acute COPD exacerbation who failed NIV, requiring MV. After an average time of
16.5 ± 5.9 h of MV, ECCO2R was initiated. In all five patients, ECCO2R facilitated
extubation within 24 h of treatment (median duration of MV post ECCO2R = 4 h,
196 L. Del Sorbo et al.

range 1.5–21.5 h). The ECCO2R flow required to accomplish the goal of extubation
ranged between 1 and 1.7 L/min resulting in pH of 7.34–7.48. Once extubated,
patients were intensively rehabilitated while on ECCO2R, with a mean time to
ambulation of 29.4 ± 12.6 h after ECCO2R.
In a larger case series [58], ECCO2R was used to treat 30 patients who failed NIV
but refused endotracheal intubation. The mortality rate of this group of patients was
significantly decreased compared to that of 30 historical controls who received MV.
In another particularly interesting case report, Diehl and colleagues reported two
intubated patients with COPD exacerbation in whom ECCO2R lowered PaCO2 and
work of breathing [59]. Interestingly, during ECCO2R, the total CO2 production
decreased likely due to the reduced work of breathing, suggesting a direct and indi-
rect effect of ECCO2R on CO2 clearance and metabolic homeostasis.
It is worth highlighting that, in these case reports, ECCO2R was applied as an
adjunct to mechanical ventilation, subjecting patients to the potential complications
of two invasive strategies of respiratory support. Excitingly, a new approach pro-
poses that ECCO2R may be effective in preventing endotracheal intubation alto-
gether in patients failing NIV. This innovative strategy could represent an important
step toward ECCO2R as alternative to MV, with the important result of sidestepping
dynamic hyperinflation, barotrauma, hemodynamic impairment, ventilator-­
associated pneumonia, immobility, critical illness polyneuropathy, and difficult
weaning secondary to ventilator-induced diaphragmatic dysfunction [20, 21, 60].
Avoiding these adverse effects could improve prognosis, especially considering the
severity of underlying lung disease and frequent presence of comorbidities [52, 61].
Along this line, in a multicenter study 25 patients supported with ECCO2R due
to the high risk of NIV failure after COPD exacerbation were compared with 21
historical controls, identified with a rigorous matching method (GenMatch), who
received only NIV [62]. ECCO2R applied in addition to NIV for an average of
29 ± 5 h with blood flow of 255 ± 78 ml/min significantly reduced the risk of intuba-
tion (hazard ratio 0.27, 95% CI, 0.07–0.98). Moreover, in the cohort of patients
treated with ECCO2R, hospital mortality was significantly lower compared with
matched controls (8% vs 33%, p = 0.035). ECCO2R-related complications, includ-
ing bleeding, clots in the circuit, pump malfunction, and membrane lung failure,
were reported in 13 patients (52%).
Another multicenter study [63] showed that, in 25 patients with COPD exacerba-
tion failing NIV and treated with ECCO2R, intubation was avoided in 56%. There
were 11 patients who received MV despite ECCO2R due to either progressive
hypoxemia or worsening hypoventilation. Duration of mechanical ventilation in
these patients was significantly lower (8 vs. 14 days, p = 0.02) than in 25 matched
controls selected by age, SAPS II, and pH at the time of NIV failure. However, ICU
and hospital length of stay, as well as mortality, did not differ between the two
groups. ECCO2R-related major adverse events were observed in 11 patients (44%).
The effects of ECCO2R have been investigated also in stable COPD patients with
chronic hypercapnia despite domiciliary NIV [64]. Ten COPD patients enrolled in
the home care NIV program for at least 6 months with PaCO2 >50 mmHg and pH
>7.35 were treated with a 24 h trial of ECCO2R. The treatment was prematurely
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 197

interrupted in four patients due to the onset of ECCO2R-related mechanical adverse

events. In the remaining six patients, ECCO2R resulted in a PaCO2 drop of 23–47%.
After termination of ECCO2R, PaCO2 returned to baseline within 48 to 96 h, sug-
gesting that it is possible to reach “a hypercapnia-free window” and thus “CO2
dialysis” is potentially feasible.
Taken together, these results suggest that ECCO2R is effective in improving gas
exchange, reducing dynamic lung hyperinflation, and freeing some patients from
MV. However, given the incidence of complications, further study is warranted to
determine whether the risk-to-benefit balance of ECCO2R is superior to that of
MV. Moreover, a large clinical trial could also define more precisely the clinical
indications for ECCO2R in COPD exacerbation, which now are derived only from
small case series (Table 10.1) and subject to rapid technological progress. Indeed,
the development of progressively smaller and less invasive devices spurs innovative
clinical strategies, such as physiotherapy and early mobilization, as described in
Chap. 14 [57]. Smaller devices become potentially more portable and facilitate
patients walking, instead of being bedridden, despite the severity of disease [65].
One can even imagine further evolution along these lines, as occurred for ventricu-
lar assist devices [66], so that small and implantable ECCO2R circuits could provide
destination treatment for patients discharged to home.

Acute Severe Asthma

Severe acute asthma is a major health burden, as its prevalence is growing world-
wide with increasing admissions to hospital and intensive care unit [19, 67]. The
in-hospital mortality rate due to acute exacerbation of asthma in the United States is
0.5%, yet this accounts for about one third of all asthma deaths. Moreover, mortality
reaches 8% for those requiring admission to the intensive care unit for intubation
and MV [19, 68].
The treatment of severe acute exacerbation of asthma consists of measures to
reverse the airflow obstruction. Ventilatory assistance is provided to patients refrac-
tory to medical treatment and developing hypoxemia or hypoventilation [19, 67].
However, while hypercapnia is observed in about 10% of asthma patients in the
acute setting, only a small percentage of these require invasive MV [69]. Interestingly,
from the combined data in two reports of 2655 asthma admissions over 20 years,
only 135 patients were diagnosed with life-threatening episodes, and a mere 48
were intubated [70, 71].
The accepted indication for MV is progressive hypercapnia with impending
hemodynamic decompensation, secondary to severe lung hyperinflation. The goal
of MV is to provide adequate gas exchange while waiting for airflow obstruction to
respond to bronchodilator therapy. However, MV may aggravate alveolar hyperin-
flation, thereby causing worsening hypercapnia, barotrauma, pneumothorax, and
further hemodynamic deterioration, with attendant risk of morbidity or mortality
[19, 33, 67].
198 L. Del Sorbo et al.

Table 10.1 Indications, exclusions, and goals of ECCO2R in COPD

Indications Invasive mechanical
ventilation for COPD
exacerbation with two or
more failed weaning
attempts
Failing treatment with NIV for at least 1–2 h with signs of respiratory distress
NIV for COPD (respiratory rate ≥30 breaths/min and use of accessory
exacerbation muscles or paradoxical abdominal movements) and
PaCO2 >55 mmHg and pH <7.25 or pH <7.30 and
PaCO2 >55 mm Hg, with PaCO2 decrease <20% from
baseline
Exclusions No absolute exclusion Anatomical abnormalities or vascular diseases
criteria, but thorough preventing the correct insertion of the ECCO2R cannula
evaluation of risks in Body mass index >31.1 kg/m2 for men or >32.2 kg/m2
each patient for women
PaO2 to FiO2 ratio consistently <250
Known or suspected pregnancy
Hemodynamic instability
Uncontrolled arrhythmia
Decompensated congestive heart failure
Recent major surgery
Hemorrhagic diathesis
Contraindications to the administration of
anticoagulants (i.e., hemorrhagic disease,
thrombocytopenia, or heparin-induced
thrombocytopenia)
History of intracranial bleeding
Inability to receive blood products
History of complications from extracorporeal support
Relative exclusion Terminal diseases
criteria Severe malnutrition or cachexia
Severely debilitated state
Advanced malignancy
Immunocompromised condition
Goals for Physiological goals Reduction of respiratory rate with consequent increased
ECCO2R expiratory time, facilitating resolution of dynamic
alveolar hyperinflation and hence improving the balance
between work of breathing and respiratory load
Bridge to recovery Prevention of endotracheal intubation
Facilitation of weaning from invasive mechanical
ventilation
Early mobilization and physiotherapy
Potential future goals Destination treatment

To prevent these potentially fatal consequences, ECCO2R has been applied as an

adjunct to MV for patients with life-threatening asthma refractory to conventional
therapy [72–74]. In theory, ECCO2R should correct respiratory acidosis, allow a
lower respiratory rate and tidal volume, reduce alveolar hyperinflation, and restore
a more normal hemodynamic status. However, when hypoxemia or hemodynamic
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 199

failure is profound, ECCO2R may not be sufficient and full extracorporeal mem-
brane oxygenation (ECMO) may provide a more adequate means of support
[75–77].
Extracorporeal membrane oxygenation was first applied in near-fatal asthma in
1981, followed by a progressively increasing number of reports [78]. In the interna-
tional extracorporeal life support organization (ELSO) registry [77], asthma was the
basis for ECMO in 24 out of 1257 adult patients between 1986 and 2006. Before
ECMO, the average pH was 7.17 ± 0.16, PaCO2 119.7 ± 58.1 mmHg, and PaO2/
FiO2 ratio 244 ± 180, despite invasive MV. Extracorporeal support was provided for
111.9 ± 71.2 h and 83.3% of patients survived. Complications were described in 19
of the 24 asthmatic subjects (79.2%). Between 1986 and 2007, 64 pediatric asthmat-
ics were treated with ECMO for life-threatening asthma, and 60 of those survived
(94%) [76]. Prior to initiation of ECMO, median pH was 6.96 (range 6.78–7.28),
median PaCO2 was 123 mmHg (range 70–237 mmHg), and PaO2 was 126 mmHg
(range 59–636 mmHg), despite MV. The median time of ECMO support was 94 h,
during which a remarkable number of cardiovascular, hemorrhagic, and mechanical
complications were reported. These results show that in both pediatric and adult
patients treated with ECMO for near-fatal asthma, hypercapnia, rather than hypox-
emia, was the dominant gas exchange problem, suggesting that less invasive
ECCO2R might have sufficed, perhaps with fewer complications.
In this vein, several case reports describe using pumpless AV-ECCO2R in adult
and pediatric patients with acute severe asthma [72–74]. More recently, a minimally
invasive venovenous ECCO2R device was used as an adjunct to MV in two adults
with asthma refractory to conventional therapies [79]. The first had a pH of 6.94,
PaCO2 of 147 mmHg, and PaO2 of 416 mmHg despite optimal MV. ECCO2R was
started with a blood flow at 1.3 L/min and sweep gas flow of 2.3 L/min, markedly
improving the respiratory acidosis to a pH of 7.24, PaCO2 of 56 mmHg, and PaO2
of 310 mmHg. After 24 h of ECCO2R support, the patient was liberated from MV
and progressively weaned from the extracorporeal device, before being successfully
discharged. The second patient not only had severe respiratory acidosis despite MV
but required vasopressors for hemodynamic support. ECCO2R contributed to nor-
malization of pH and PaCO2, while respiratory rate was lowered from 8 to 4 breaths/
min and tidal volume from 6 to 4 mL/kg. In one day, intrinsic PEEP resolved, shock
was reversed, and neuromuscular blocking agents were discontinued.
In summary, minimally invasive ECCO2R appears very promising for patients
with near-fatal asthma refractory to conventional treatment, but systematic evalua-
tion is needed to prove its efficacy and determine the true risks.

Exclusions

It is important to emphasize that the current lack of large, systematic clinical trials
providing definitive evidence of the efficacy of ECCO2R in improving clinically
significant outcomes makes the decision of including or excluding patients from the
200 L. Del Sorbo et al.

treatment with ECCO2R particularly complex. Since ECCO2R has become a mini-
mally invasive treatment, well-tolerated, and easy to manage, requiring similar
resources as renal dialysis, every patient with adequate indications should be con-
sidered as a potential candidate for its application, without absolute exclusions [4].
The balance between risks and benefits should be evaluated in each clinical case
before exposing a patient to this innovative respiratory support strategy. Despite
remarkable improvements in technology, there are still a number of clinically rele-
vant, unwanted side effects related to the application of ECCO2R, which may limit
its clinical consideration in patients with specific conditions [4, 6, 16]. The most
frequent complications caused by ECCO2R are mechanical, secondary to the poten-
tial vascular trauma during the intravascular insertion of the circuit cannula, and
bleeding due to the need of therapeutic anticoagulation to avoid thrombosis of the
circuit. Therefore, anatomical abnormalities or vascular diseases preventing the cor-
rect insertion of the ECCO2R cannula, preexisting risk of bleeding, contraindica-
tions to the administration of anticoagulants, or actual hemorrhagic diseases,
represent major conditions that may preclude the application of ECCO2R
(Table 10.1).
Moreover, additional exclusion criteria are applied in clinical trials investigating
the efficacy ECCO2R in improving the recovery from respiratory failure [57, 80,
81]. In this setting, conditions that limit recovery, such as terminal diseases, severe
malnutrition or cachexia, severely debilitated state, advanced malignancy, and
immunocompromised condition, are exclusionary (Table 10.1). An exception would
be a patient with end-stage, obstructive lung disease in whom ECCO2R might serve
as a bridge to lung transplant [82].

References

1. Abrams D, Brodie D. Emerging indications for extracorporeal membrane oxygenation in

adults with respiratory failure. Ann Am Thorac Soc. 2013;10:371–7.
2. Del Sorbo L, Cypel M, Fan E. Extracorporeal life support for adults with severe acute respira-
tory failure. Lancet Respir Med. 2014;2:154–64.
3. Brodie D, Bacchetta M. Extracorporeal membrane oxygenation for ARDS in adults. N Engl J
Med. 2011;365:1905–14.
4. MacLaren G, Combes A, Bartlett RH. Contemporary extracorporeal membrane oxygenation
for adult respiratory failure: life support in the new era. Intensive Care Med. 2012;38:210–20.
5. Sidebotham D, McGeorge A, McGuinness S, Edwards M, Willcox T, Beca J. Extracorporeal
membrane oxygenation for treating severe cardiac and respiratory failure in adults: part
2-­technical considerations. J Cardiothorac Vasc Anesth. 2010;24:164–72.
6. Cove ME, Maclaren G, Federspiel WJ, Kellum JA. Bench to bedside review: extracorporeal
carbon dioxide removal, past present and future. Crit Care. 2012;16:232.
7. Terragni P, Maiolo G, Ranieri VM. Role and potentials of low-flow CO(2) removal system in
mechanical ventilation. Curr Opin Crit Care. 2012;18:93–8.
8. Bartlett RH. Physiology of extracorporeal life support. ECMO, extracorporeal cardiopulmo-
nary support in critical care. In: Annich GM, Lynch WR, MacLaren G, Wilson JM, Barlett RH,
editors. Extracorporeal life support organization. 4th ed. Ann Arbor: Springer; 2012. p. 11–31.
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 201

9. Gattinoni L, Pesenti A, Mascheroni D, Marcolin R, Fumagalli R, Rossi F, Iapichino G,

Romagnoli G, Uziel L, Agostoni A, et al. Low-frequency positive-pressure ventilation with
extracorporeal CO2 removal in severe acute respiratory failure. JAMA. 1986;256:881–6.
10. Gattinoni L, Carlesso E, Langer T. Clinical review: extracorporeal membrane oxygenation.
Crit Care. 2011;15:243.
11. Gattinoni L, Kolobow T, Damia G, Agostoni A, Pesenti A. Extracorporeal carbon dioxide
removal (ECCO2R): a new form of respiratory assistance. Int J Artif Organs. 1979;2:183–5.
12. Marcolin R, Mascheroni D, Pesenti A, Bombino M, Gattinoni L. Ventilatory impact of partial
extracorporeal CO2 removal (PECOR) in ARF patients. ASAIO Trans. 1986;32:508–10.
13. Zimmermann M, Bein T, Arlt M, Philipp A, Rupprecht L, Mueller T, Lubnow M, Graf BM,
Schlitt HJ. Pumpless extracorporeal interventional lung assist in patients with acute respiratory
distress syndrome: a prospective pilot study. Crit Care. 2009;13:R10.
14. Bein T, Weber-Carstens S, Goldmann A, Muller T, Staudinger T, Brederlau J, Muellenbach R,
Dembinski R, Graf BM, Wewalka M, Philipp A, Wernecke KD, Lubnow M, Slutsky AS. Lower
tidal volume strategy (approximately 3 ml/kg) combined with extracorporeal CO(2) removal
versus ‘conventional’ protective ventilation (6 ml/kg) in severe ARDS : the prospective ran-
domized Xtravent-study. Intensive Care Med. 2013;39(5):847–56.
15. Terragni PP, Del Sorbo L, Mascia L, Urbino R, Martin EL, Birocco A, Faggiano C, Quintel M,
Gattinoni L, Ranieri VM. Tidal volume lower than 6 ml/kg enhances lung protection: role of
extracorporeal carbon dioxide removal. Anesthesiology. 2009;111:826–35.
16. Braune SA, Kluge S. Extracorporeal lung support in patients with chronic obstructive pulmo-
nary disease. Minerva Anestesiol. 2013;79:934–43.
17. Lund LW, Federspiel WJ. Removing extra CO2 in COPD patients. Curr Respir Care Rep.
2013;2:131–8.
18. Barbera JA, Roca J, Ferrer A, Felez MA, Diaz O, Roger N, Rodriguez-Roisin R. Mechanisms
of worsening gas exchange during acute exacerbations of chronic obstructive pulmonary dis-
ease. Eur Respir J. 1997;10:1285–91.
19. McFadden ER Jr. Acute severe asthma. Am J Respir Crit Care Med. 2003;168:740–59.
20. Del Sorbo L, Ranieri VM. We do not need mechanical ventilation any more. Crit Care Med.
2010;38:S555–8.
21. Malagon I, Greenhalgh D. Extracorporeal membrane oxygenation as an alternative to ventila-
tion. Curr Opin Anaesthesiol. 2013;26:47–52.
22. Hoeper MM, Wiesner O, Hadem J, Wahl O, Suhling H, Duesberg C, Sommer W, Warnecke
G, Greer M, Boenisch O, Busch M, Kielstein JT, Schneider A, Haverich A, Welte T, Kuhn
C. Extracorporeal membrane oxygenation instead of invasive mechanical ventilation in
patients with acute respiratory distress syndrome. Intensive Care Med. 2013;39(11):2056–7.
23. Crotti S, Lissoni A, Tubiolo D, Azzari S, Tarsia P, Caspani L, Gattinoni L. Artificial lung as an
alternative to mechanical ventilation in COPD exacerbation. Eur Respir J. 2012;39:212–5.
24. Hogg JC. Pathophysiology of airflow limitation in chronic obstructive pulmonary disease.
Lancet. 2004;364:709–21.
25. Laghi F, Goyal A. Auto-PEEP in respiratory failure. Minerva Anestesiol. 2012;78:201–21.
26. Loring SH, Garcia-Jacques M, Malhotra A. Pulmonary characteristics in COPD and mecha-
nisms of increased work of breathing. J Appl Physiol. 2009;107:309–14.
27. Stevenson NJ, Walker PP, Costello RW, Calverley PM. Lung mechanics and dyspnea dur-
ing exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med.
2005;172:1510–6.
28. O’Donnell DE, Parker CM. COPD exacerbations 3: Pathophysiology. Thorax. 2006;61:354–61.
29. Roberts CM, Stone RA, Buckingham RJ, Pursey NA, Lowe D. Acidosis, non-invasive ventila-
tion and mortality in hospitalised COPD exacerbations. Thorax. 2011;66:43–8.
30. Oliven A, Kelsen SG, Deal EC, Cherniack NS. Mechanisms underlying CO2 retention during
flow-resistive loading in patients with chronic obstructive pulmonary disease. J Clin Invest.
1983;71:1442–9.
31. Nava S, Hill N. Non-invasive ventilation in acute respiratory failure. Lancet. 2009;374:250–9.
202 L. Del Sorbo et al.

32. Reddy RM, Guntupalli KK. Review of ventilatory techniques to optimize mechanical venti-
lation in acute exacerbation of chronic obstructive pulmonary disease. Int J Chron Obstruct
Pulmon Dis. 2007;2:441–52.
33. Kao CC, Jain S, Guntupalli KK, Bandi V. Mechanical ventilation for asthma: a 10-year experi-
ence. J Asthma. 2008;45:552–6.
34. Ward NS, Dushay KM. Clinical concise review: mechanical ventilation of patients with
chronic obstructive pulmonary disease. Crit Care Med. 2008;36:1614–9.
35. Calverley PM, Walker P. Chronic obstructive pulmonary disease. Lancet. 2003;362:1053–61.
36. Chapman KR, Mannino DM, Soriano JB, Vermeire PA, Buist AS, Thun MJ, Connell C, Jemal
A, Lee TA, Miravitlles M, Aldington S, Beasley R. Epidemiology and costs of chronic obstruc-
tive pulmonary disease. Eur Respir J. 2006;27:188–207.
37. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to
2030. PLoS Med. 2006;3:e442.
38. Burge S, Wedzicha JA. COPD exacerbations: definitions and classifications. Eur Respir J
Suppl. 2003;41:46s–53s.
39. Hurst JR, Vestbo J, Anzueto A, Locantore N, Mullerova H, Tal-Singer R, Miller B, Lomas
DA, Agusti A, Macnee W, Calverley P, Rennard S, Wouters EF, Wedzicha JA. Susceptibility to
exacerbation in chronic obstructive pulmonary disease. N Engl J Med. 2010;363:1128–38.
40. Patil SP, Krishnan JA, Lechtzin N, Diette GB. In-hospital mortality following acute exacerba-
tions of chronic obstructive pulmonary disease. Arch Intern Med. 2003;163:1180–6.
41. Donaldson GC, Wedzicha JA. COPD exacerbations 1: epidemiology. Thorax. 2006;61:164–8.
42. Soler-Cataluna JJ, Martinez-Garcia MA, Roman Sanchez P, Salcedo E, Navarro M, Ochando
R. Severe acute exacerbations and mortality in patients with chronic obstructive pulmonary
disease. Thorax. 2005;60:925–31.
43. Quon BS, Gan WQ, Sin DD. Contemporary management of acute exacerbations of COPD: a
systematic review and meta-analysis. Chest. 2008;133:756–66.
44. Chronic Obstructive Pulmonary Disease. National clinical guideline on management of chronic
obstructive pulmonary disease in adults in primary and secondary care. Thorax. 2004;59(Suppl
1):1–232.
45. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS. Global strategy for the diagnosis,
management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global
Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir
Crit Care Med. 2001;163:1256–76.
46. Organized jointly by the American Thoracic Society, the European Respiratory Society, the
European Society of Intensive Care Medicine, and the Société de Réanimation de Langue
Française, and approved by ATS Board of Directors, December 2000. International Consensus
Conferences in Intensive Care Medicine: noninvasive positive pressure ventilation in acute
Respiratory failure. Am J Respir Crit Care Med. 2001;163:283–91.
47. Brochard L, Mancebo J, Wysocki M, Lofaso F, Conti G, Rauss A, Simonneau G, Benito S,
Gasparetto A, Lemaire F, et al. Noninvasive ventilation for acute exacerbations of chronic
obstructive pulmonary disease. N Engl J Med. 1995;333:817–22.
48. Ram FS, Picot J, Lightowler J, Wedzicha JA. Non-invasive positive pressure ventilation for
treatment of respiratory failure due to exacerbations of chronic obstructive pulmonary disease.
Cochrane Database Syst Rev. 2004;(1):CD004104.
49. Scala R, Naldi M, Archinucci I, Coniglio G, Nava S. Noninvasive positive pressure ventila-
tion in patients with acute exacerbations of COPD and varying levels of consciousness. Chest.
2005;128:1657–66.
50. Chandra D, Stamm JA, Taylor B, Ramos RM, Satterwhite L, Krishnan JA, Mannino D,
Sciurba FC, Holguin F. Outcomes of noninvasive ventilation for acute exacerbations of chronic
obstructive pulmonary disease in the United States, 1998–2008. Am J Respir Crit Care Med.
2012;185:152–9.
51. Rello J, Ollendorf DA, Oster G, Vera-Llonch M, Bellm L, Redman R, Kollef MH, Group
VAPOSA. Epidemiology and outcomes of ventilator-associated pneumonia in a large US data-
base. Chest. 2002;122:2115–21.
10 ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions 203

52. Ai-Ping C, Lee KH, Lim TK. In-hospital and 5-year mortality of patients treated in the ICU for
acute exacerbation of COPD: a retrospective study. Chest. 2005;128:518–24.
53. Sklar MC, Beloncle F, Katsios CM, Brochard L, Friedrich JO. Extracorporeal carbon dioxide
removal in patients with chronic obstructive pulmonary disease: a systematic review. Intensive
Care Med. 2015;41:1752–62.
54. d’Andrea A, Banfi C, Bendjelid K, Giraud R. The use of extracorporeal carbon dioxide
removal in acute chronic obstructive pulmonary disease exacerbation: a narrative review. Can
J Anaesth. 2020;67:462–74.
55. Pesenti A, Rossi GP, Pelosi P, Brazzi L, Gattinoni L. Percutaneous extracorporeal CO2 removal
in a patient with bullous emphysema with recurrent bilateral pneumothoraces and respiratory
failure. Anesthesiology. 1990;72:571–3.
56. Taccone FS, Malfertheiner MV, Ferrari F, Di Nardo M, Swol J, Broman LM, Vercaemst L,
Barrett N, Pappalardo F, Belohlavek J, Mueller T, Lorusso R, Belliato M, et al. Extracorporeal
CO2 removal in critically ill patients: a systematic review. Minerva Anestesiol. 2017;83:762–72.
57. Abrams DC, Brenner K, Burkart KM, Agerstrand CL, Thomashow BM, Bacchetta M, Brodie
D. Pilot study of extracorporeal carbon dioxide removal to facilitate extubation and ambu-
lation in exacerbations of chronic obstructive pulmonary disease. Ann Am Thorac Soc.
2013;10:307–14.
58. Morelli A, D’Egidio A, Orecchioni A, Alessandri F, Mascia L, Ranieri VM. Extracorporeal
CO2 removal in hypercapnic patients who fail noninvasive ventilation and refuse endotracheal
intubation: a case series. Intensive Care Med Exp. 2015;3:824.
59. Diehl JL, Piquilloud L, Richard JM, Mancebo J, Mercat A. Effects of extracorporeal carbon
dioxide removal on work of breathing in patients with chronic obstructive pulmonary disease.
Intensive Care Med. 2016;42:951–2.
60. Bonin F, Sommerwerck U, Lund LW, Teschler H. Avoidance of intubation during acute
exacerbation of chronic obstructive pulmonary disease for a lung transplant candidate using
extracorporeal carbon dioxide removal with the Hemolung. J Thorac Cardiovasc Surg.
2013;145:e43–4.
61. Menzies R, Gibbons W, Goldberg P. Determinants of weaning and survival among patients
with COPD who require mechanical ventilation for acute respiratory failure. Chest.
1989;95:398–405.
62. Del Sorbo L, Pisani L, Filippini C, Fanelli V, Fasano L, Terragni P, Dell’Amore A, Urbino R,
Mascia L, Evangelista A, Antro C, D’Amato R, Sucre MJ, Simonetti U, Persico P, Nava S,
Ranieri VM. Extracorporeal Co2 removal in hypercapnic patients at risk of noninvasive venti-
lation failure: a matched cohort study with historical control. Crit Care Med. 2015;43:120–7.
63. Braune S, Sieweke A, Brettner F, Staudinger T, Joannidis M, Verbrugge S, Frings D, Nierhaus
A, Wegscheider K, Kluge S. The feasibility and safety of extracorporeal carbon dioxide
removal to avoid intubation in patients with COPD unresponsive to noninvasive ventilation
for acute hypercapnic respiratory failure (ECLAIR study): multicentre case-control study.
Intensive Care Med. 2016;42:1437–44.
64. Pisani L, Nava S, Desiderio E, Polverino M, Tonetti T, Ranieri VM. Extracorporeal CO2
removal (ECCO2R) in patients with stable COPD with chronic hypercapnia: a proof-of-­
concept study. Thorax. 2020;75:897–900.
65. Garcia JP, Kon ZN, Evans C, Wu Z, Iacono AT, McCormick B, Griffith BP. Ambulatory veno-­
venous extracorporeal membrane oxygenation: innovation and pitfalls. J Thorac Cardiovasc
Surg. 2011;142:755–61.
66. Moazami N, Sun B, Feldman D. Stable patients on left ventricular assist device support have
a disproportionate advantage: time to re-evaluate the current UNOS policy. J Heart Lung
Transplant. 2011;30:971–4.
67. Phipps P, Garrard CS. The pulmonary physician in critical care. 12: Acute severe asthma in the
intensive care unit. Thorax. 2003;58:81–8.
68. Krishnan V, Diette GB, Rand CS, Bilderback AL, Merriman B, Hansel NN, Krishnan
JA. Mortality in patients hospitalized for asthma exacerbations in the United States. Am J
Respir Crit Care Med. 2006;174:633–8.
204 L. Del Sorbo et al.

69. McFadden ER Jr, Lyons HA. Arterial-blood gas tension in asthma. N Engl J Med.
1968;278:1027–32.
70. Braman SS, Kaemmerlen JT. Intensive care of status asthmaticus. A 10-year experience.
JAMA. 1990;264:366–8.
71. Trawick DR, Holm C, Wirth J. Influence of gender on rates of hospitalization, hospital course,
and hypercapnea in high-risk patients admitted for asthma : a 10-year retrospective study at
Yale-New Haven Hospital. Chest. 2001;119:115–9.
72. Jung C, Lauten A, Pfeifer R, Bahrmann P, Figulla HR, Ferrari M. Pumpless extracorporeal
lung assist for the treatment of severe, refractory status asthmaticus. J Asthma. 2011;48:111–3.
73. Conrad SA, Green R, Scott LK. Near-fatal pediatric asthma managed with pumpless arteriove-
nous carbon dioxide removal. Crit Care Med. 2007;35:2624–9.
74. Elliot SC, Paramasivam K, Oram J, Bodenham AR, Howell SJ, Mallick A. Pumpless extracor-
poreal carbon dioxide removal for life-threatening asthma. Crit Care Med. 2007;35:945–8.
75. Shapiro MB, Kleaveland AC, Bartlett RH. Extracorporeal life support for status asthmaticus.
Chest. 1993;103:1651–4.
76. Hebbar KB, Petrillo-Albarano T, Coto-Puckett W, Heard M, Rycus PT, Fortenberry
JD. Experience with use of extracorporeal life support for severe refractory status asthmaticus
in children. Crit Care. 2009;13:R29.
77. Mikkelsen ME, Woo YJ, Sager JS, Fuchs BD, Christie JD. Outcomes using extracorporeal life
support for adult respiratory failure due to status asthmaticus. ASAIO J. 2009;55:47–52.
78. MacDonnell KF, Moon HS, Sekar TS, Ahluwalia MP. Extracorporeal membrane oxygenator
support in a case of severe status asthmaticus. Ann Thorac Surg. 1981;31:171–5.
79. Brenner K, Abrams DC, Agerstrand CL, Brodie D. Extracorporeal carbon dioxide removal
for refractory status asthmaticus: experience in distinct exacerbation phenotypes. Perfusion.
2014;29:26–8.
80. Burki NK, Mani RK, Herth FJ, Schmidt W, Teschler H, Bonin F, Becker H, Randerath WJ,
Stieglitz S, Hagmeyer L, Priegnitz C, Pfeifer M, Blaas SH, Putensen C, Theuerkauf N, Quintel
M, Moerer O. A novel extracorporeal CO(2) removal system: results of a pilot study of hyper-
capnic respiratory failure in patients with COPD. Chest. 2013;143:678–86.
81. Kluge S, Braune SA, Engel M, Nierhaus A, Frings D, Ebelt H, Uhrig A, Metschke M,
Wegscheider K, Suttorp N, Rousseau S. Avoiding invasive mechanical ventilation by extracor-
poreal carbon dioxide removal in patients failing noninvasive ventilation. Intensive Care Med.
2012;38:1632–9.
82. Del Sorbo L, Boffini M, Rinaldi M, Ranieri VM. Bridging to lung transplantation by extracor-
poreal support. Minerva Anestesiol. 2012;78:243–50.
Chapter 11
ECMO as a Bridge to Lung
Transplantation

Christian Kuehn and Ruslan Natanov

Introduction

Extracorporeal life support (ECLS) has become an established therapy in patients

with life-threatening cardiac, respiratory, or cardiopulmonary failure over the last
two decades. Depending on the pathology, the use of veno-venous, venoarterial, or
a combination of both may be needed for patient support. In patients with end-stage
disease, this type of support can either be used as a bridge to recovery or as a bridge
to transplantation (BTT). As ECLS has become more widely available over time, its
use in the daily clinical setting has increased exponentially. In this chapter, we dis-
cuss various support types and their application in lung transplantation patients.

Historical Aspects

Over the last four decades, lung transplantation has evolved from an experimental
technique to an established option for patients with terminal pulmonary disease and
disease of the pulmonary circulation. Technical advances in isolated lung transplan-
tation were driven especially by Cooper and colleagues in Toronto. In the early era
of lung transplantation, ECMO was seen as a contraindication for transplantation as
mechanical support was marked by complications and poor outcome. In particular,
bleeding, hemolysis, infection, and technical complications marked the postopera-
tive course in the early days of mechanical support in lung transplantation patients.

C. Kuehn · R. Natanov (*)

Department of Cardiothoracic, Transplantation, and Vascular Surgery, Hannover Medical
School, Hannover, Germany
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature 205

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_11
206 C. Kuehn and R. Natanov

Despite these complications, the number of transplantations rose and transplanta-

tion techniques evolved allowing for more complex patients to be transplanted. This
led to the first successful use of ECMO as bridging to retransplantation in patients
suffering from early graft failure [1]. As for most centers, graft deterioration to the
point of mechanical ventilation or the necessity of ECMO marked the end of trans-
plant candidacy. The notion of using ECMO for bridging to retransplantation pro-
voked a scientific discussion.
At the time, restricted use of ECMO was understandable due to the existing
complications while on ECMO and the inability to support patients for more than
a few days. For this reason, only sporadic case reports on the use of ECMO in
patients as a bridge to transplant from this early time exist. Over the ensuing
decade, lung transplantation became even more established and the number of
transplants continued to rise. Improving success rates meant that lung transplanta-
tion became a realistic option for end-stage lung disease so that the number of
patients on the waiting list has surpassed the number of available and suitable
donor organs. A longer waiting time on the transplantation list meant an increasing
number of patients that decompensated while awaiting a donor. Treatment options
involved intubation and mechanical ventilation despite the fact this was still
viewed as a significant risk factor for mortality and the fact that many patients
remained hypercapnic despite all attempts to ventilate adequately [2, 3].
Improvement in ECLS techniques and the introduction of a pumpless support sys-
tem offered a novel therapeutic option: extracorporeal carbon dioxide removal
(ECCO2R) as a bridge to lung transplantation. Successful bridging with a pump-
less (arteriovenous) device showed a notable 1-year survival rate after transplanta-
tion of 80%. Although these patients could be bridged over 2 weeks, pumpless
AV-ECLS support only allows a relatively low blood flow, thus only CO2 removal
(and no oxygenation) was possible. For patients with greatly impaired gas exchange
function, low-flow, pumpless circuits are therefore not suitable. The integration of
a pump and the use of low-resistance membranes allowed for greater blood flow
and made decarboxylation and oxygenation via a veno-venous (VV-) ECMO pos-
sible. Since the introduction of ECLS, technical innovations have led to a signifi-
cant reduction of complications and lowered the threshold for using ECLS early,
rather than as a last resort.
Several technological advances can be noted. The most significant technological
change was the introduction of the coated, polymethylpentene, hollow-fiber mem-
branes in the oxygenator, enabling excellent gas transfer [4]. The second improve-
ment was the refinement of the system surface coating to enhance blood compatibility.
This, in combination with novel, powerful, centrifugal pumps, made long-lasting
support possible. Based on the technical improvements and increased experience,
the Hannover group published the first series of patients bridged to transplantation
under the support of awake-ECMO [4]. Patients could be supported without the
need for sedation and mechanical ventilation. This led to the awareness in more
centers of the possibility of ECLS and awake-ECMO as an option for carrying
patients successfully to transplantation.
11 ECMO as a Bridge to Lung Transplantation 207

Modes of Support

In general, patients on the transplant waiting list with a deteriorating clinical status
may be eligible for ECLS. In patients with deteriorating pulmonary function, the
choice of support is either the pumpless interventional lung assist (iLA) or the more
common VV-ECMO. In the setting of the iLA, systemic pressure drives blood from
the femoral artery via a membrane oxygenator to the femoral vein. In the VV-ECMO
setting, oxygen-poor blood is drawn from the femoral vein and, after oxygenation,
returned in either the jugular or the contralateral femoral vein. In VV-ECMO, blood
is actively pumped using a centrifugal pump. In case of hemodynamic instability, or
patients with strained right ventricular function, the use of venoarterial (VA-)
ECMO should be considered. In this type of support, venous blood is drawn from
the right heart via the inferior vena cava, and oxygenated blood is returned in either
the femoral or subclavian artery. In rare cases, the arterial cannula can be placed in
the ascending aorta. In extreme cases where both cardiac and pulmonary failures are
present, the choice for a combined veno-venoarterial (VVA-) ECMO may be con-
sidered. This constellation however is reserved for extremely ill patients and requires
a high level of ECMO expertise.

Interventional Lung Assist

iLA has proven useful in patients awaiting lung transplantation with refractory
hypercapnia [5]. Although significant reduction of carbon dioxide has been reported
by several authors, to date no randomized clinical trials comparing iLA and
VV-ECMO are provided in the literature. Several limitations of this support system
have been discussed. Due to the lack of a pump, the flow rate is limited to 2.5 L/min
which allows for adequate removal of CO2 but is not sufficient for oxygenation.
Furthermore, cannulation of the femoral artery can occlude the vessel leading to
distal leg ischemia. In the worst case, amputation of the limb can follow [6, 7]. Due
to the significant limitations, this system has mainly been replaced by VV-ECMO.

Veno-venous ECMO

In the case of VV-ECMO, support is mainly indicated for patients with severe respi-
ratory failure. Depending on the type of respiratory failure, either the classical
VV-ECMO with two cannulas or a dual-lumen, single cannula VV-ECMO is pre-
ferred. In patients with predominantly hypercapnic respiratory failure, relatively
low flow rates may be sufficient to provide normalization of the CO2 concentrations.
To achieve this, a dual-lumen cannula can be inserted into the jugular vein (usually
on the right side). Echocardiographic or fluoroscopic guidance for optimal
208 C. Kuehn and R. Natanov

positioning is necessary for this type of cannulation (see Chap. 4). Although the
flow rate is usually too low to achieve oxygenation, it is effective in the removal of
excess carbon dioxide. This usually leads to the normalization of respiratory acido-
sis and may reduce the need for vasoactive agents. Furthermore, as a dual-lumen
cannula may be inserted under local anesthesia, sedation and intubation of the
patient may be avoided. Physical therapy and patient mobilization are very much
possible under this type of ECLS (see Chap. 14). A significant limitation is the rela-
tively low blood flow rate (usually up to 3.5 L/min), therefore limiting the rate of
blood oxygenation. Patients with severe, end-stage respiratory disease due to com-
bined hypoxic and hypercapnic respiratory failure may not be sufficiently treated
with a single cannula only. For these patients, the classic VV-ECMO with cannula-
tion of both the jugular and femoral vein may be a preferred option. In this setting,
high blood flow (up to 6 L/min) may be achieved, providing near-complete replace-
ment of native lung function. Classical, two-cannula VV-ECMO may also be initi-
ated in the awake patient, provided the patient is cooperative. Large diameter
cannulas are placed in the jugular and femoral vein (both usually on the right side).
Blood is drained from the inferior vena cava, oxygenated, and returned to the right
atrium. Optimal positioning of the cannulas provides maximal blood flow without
significant recirculation of oxygenated blood. VV-ECMO for BTT may be suitable
in patients with end-stage pulmonary fibrosis, sarcoidosis, or bronchiolitis obliter-
ans. As COVID-19 has become a serious source of pulmonary disease, our center
has recently bridged several patients to transplantation with severe COVID-19-
associated ARDS. To prevent loss of muscle mass and deterioration of the general
patient’s condition, early implantation of VV-ECMO is advised. However, to avoid
the situation of a bridge to nowhere, the chances for listing for lung transplantation
must be critically evaluated before initiation of VV-ECMO (see Chap. 17).

Venoarterial ECMO

VA-ECMO is typically indicated in patients with hemodynamic instability. However,

should this instability originate from myocardial hypoxia due to respiratory disease,
then patients may not benefit from VA-ECMO alone. In these patients, poorly oxy-
genated blood is ejected from the left ventricle into the systemic circulation (and
thus the coronary arteries). In time, myocardial hypoxia may lead to forward failure
and hemodynamic instability. Caution must be taken in this situation, as a watershed
between the blood of low and high oxygen concentration can be formed, typically
in the aortic arch. Poorly oxygenated blood will continue to be ejected from the left
ventricle into the coronary arteries thus enabling further hemodynamic deteriora-
tion. In contrast, when hemodynamic instability is due to pulmonary hypertension,
oxygenation of the blood is adequate and does not pose an imminent hypoxic threat
to the myocardium. Instead, increased workload of the right ventricle may be the
culprit causing cardiogenic shock. Peripheral implantation of VA-ECMO support
alleviates the strain on the right ventricle immediately, stabilizing the
11 ECMO as a Bridge to Lung Transplantation 209

hemodynamics. Cannulas are typically implanted in the right groin using ultrasound
guidance. Local anesthesia may be used to avoid the need for sedation and intuba-
tion. In the ideal setting, VA-ECMO is implanted electively. Intraoperative continu-
ation of VA-ECMO may be necessary after successful BTT.

Complex Forms of Support

In both previously described modes of support, patients have either pulmonary or

circulatory failure. However, the difficult question is often what mode of support is
optimal in patients with combined dysfunction. Right ventricular failure with dete-
riorating pulmonary function generally does not profit from VV-ECMO alone.
Furthermore, the development of a watershed can lead to hypoxemia of the upper
body [8]. For these patients, it is of utmost importance that sedation and intubation
are avoided, thus central cannulation is not an option. In general, therapy is started
with VV-ECMO and, when needed due to hemodynamic instability, is expanded
with an additional arterial branch in the femoral artery. Distal leg perfusion is rou-
tinely placed to avoid leg ischemia. The arterial branch is connected to the afferent
venous line via a Y-connector. In this mode of support, blood is being drained from
the right atrium via the femoral cannula and is being returned to the patient in the
upper vena cava and the femoral artery. Blood flow in either the venous or arterial
branch can be regulated using a throttle screw. Measurement of individual blood
flow is recommended and an additional flow sensor is advised. Patients with this
kind of support are extremely fragile and require a highly experienced staff. Special
consideration should be taken in these patients concerning the wisdom of BTT.

ECLS as a Bridge to Transplantation: General Considerations

In our clinic, examination and evaluation of patients for transplantation are done by
a multidisciplinary team. Patient education concerning the waiting period, technical
aspects of the procedure, and long-term prognosis are pivotal in the preoperative
patient preparation. Patients are furthermore educated in the possibility of deteriora-
tion while on the waiting list and the available invasive therapeutic options. As each
patient is a unique individual, there is no simple algorithm to address these possi-
bilities. In the evaluation of the possibility of ECLS as BTT, patient age and comor-
bidities are significant variables. Patients with single organ failure are usually
considered as potential candidates for ECLS as BTT. Usually, there are two sce-
narios where ECLS can be considered as a potential BTT for lung transplantation.
In the first type of patient, there is an isolated gas exchange failure, thus prompting
VV-ECMO. The second type of patient presents with right-sided heart failure and is
suitable for VA-ECMO. In ideal circumstances, patients and physicians recognize
deterioration and end-stage failure. In this case, actions can be taken before the need
210 C. Kuehn and R. Natanov

for intubation, sedation, and potential complications of ICU admission arise. In

some cases, noninvasive ventilation can provide the time needed to bridge to can-
nulation. Once stabilized with ECLS, patients can resume their participation in
everyday life as much as possible. Mobilization and physical therapy while on
ECLS provide for optimal physical conditions and a chance of success after trans-
plantation. Should there be the possibility for elective cannulation, utmost care
should be taken to initiate ECLS in the awake patient. This provides the possibility
for the patient to eat and drink independently and participate in everyday interaction
between medical staff [9]. While on ECLS, intermittent noninvasive ventilation as
support during the night may be helpful.
Early evaluation for ECLS in these patients is essential to avoid critical situa-
tions. Often this ideal situation is not realized: patients may deteriorate unexpect-
edly and urgent intubation and mechanical ventilation may be unavoidable.
Alternatively, in some patients, ECLS is not sufficient alone and intubation may be
needed. Independent of the cause for intubation, once ECLS is established weaning
from ventilation should be pursued. If complete weaning is not feasible, sedation
should be withheld and spontaneous breathing facilitated. Neurological evaluation
is essential to determine whether the patient is still a suitable candidate for trans-
plantation. Although a successful postoperative outcome depends on whether new
complications arise during the waiting period, trying to reduce the waiting time by
accepting marginal organs must be avoided. Once a suitable organ is accepted,
transplantation is usually done via the minimally invasive technique, and the deci-
sion of ECLS removal after transplantation lies with the surgeon. Care must be
taken in patients left on ECLS after transplantation due to the increased risk of
postoperative bleeding. Optimization of the coagulation factors, platelets, and fre-
quent measurement of heparin activity minimizes the risk of postoperative bleeding.
Should the patient deteriorate while on ECLS, care must be taken to reevaluate
whether transplantation remains a realistic option. In some cases, absolute contrain-
dications for transplantation develop during the waiting period and ECLS may
become a bridge to nowhere. This situation should be recognized early and ECLS
terminally withdrawn. Additional organ failure (kidney, liver), fulminant infections,
and rising need for vasoactive infusions usually preclude successful transplantation.
In these cases, patients may be taken off the active waiting list pending recovery.

Pulmonary Failure

Treatment of patients with progressive pulmonary failure traditionally involves

intubation and mechanical ventilation. There is, however, compelling evidence that
early implantation of VV-ECMO in awake patients provides a survival benefit [10].
When patients are put on ECMO electively, proper education on the procedure and
possible complications contributes to the success of awake ECMO. Regrettably,
often patients deteriorate at a rapid pace or are not in the ECMO center at the time
of the onset of respiratory failure. In this case, primary stabilization with
11 ECMO as a Bridge to Lung Transplantation 211

mechanical ventilation must be prioritized. Subsequently, initiation of VV-ECMO

and weaning from ventilation are primary goals. Should the patient be admitted to
an ICU other than the ECMO center, early interdisciplinary evaluation of the patient
is essential to expedite interhospital transfer on or for ECMO.

Pulmonary Hypertension and Right Ventricular Failure

Increased strain on the right ventricle due to high pulmonary pressures may lead to
right-sided decompensation. Clinical evaluation of the patient involves echocar-
diography, hemodynamic monitoring, and clinical chemistry (see Chap. 7).
Echocardiographic signs include enlargement of the right ventricle and flattening of
the interventricular septum, the so-called D-sign. Invasive measurements show high
central venous pressure, low central venous saturation, and low cardiac output.
Acute kidney failure and acute liver dysfunction are common end-organ failures in
right heart failure. In these patients, implantation of VA-ECMO usually stabilizes
the hemodynamic situation immediately and restores end-organ function. Continued
use of VA-ECMO during and after transplantation has been associated with an
improved postoperative outcome when compared to patients with mechanical ven-
tilation alone. Midterm results show a benefit for intra- and postoperative use of
VA-ECMO in terms of reduced mortality and improvement of left ventricular func-
tion 1 year after transplantation [11]. This effect was also seen in patients trans-
planted due to pulmonary fibrosis combined with pulmonary hypertension [12]. In
contrast to the early practice of VA-ECMO, where the arterial cannula was placed
directly into the pulmonary artery, peripheral cannulation in the femoral artery and
vein is currently the standard. When done percutaneously, easy removal at the bed-
side avoids the need for a second operation and thus provides the possibility for
early postoperative weaning from ventilation.

Patient Transport While on ECLS

In most cases, listed patients are at the transplantation center at the time of decom-
pensation so that ECMO may be initiated electively. Sometimes, however, patients
are in an external center or are not yet listed at the time of decompensation. In these
patients, interdisciplinary evaluation should be done early, preferably prior to intu-
bation. When eligible for transplantation, patients must be evaluated for the type of
ECLS support and a retrieval team is sent to the destination. After careful evaluation
of the patient, cannulation is done in typical manner. Interhospital patient transport
on ECLS is quite possible, even in awake patients [13]. Transport can be by air or
land and enables the possibility of listing once the patient has arrived at the trans-
plantation center. In patients with pulmonary hypertension, it is preferable to arrange
transport while the patient is awake.
212 C. Kuehn and R. Natanov

Preventing Mechanical Ventilation with ECLS

As previously stated, classical therapy options for patients with end-stage pulmo-
nary failure involved sedation and mechanical ventilation. Follow-up of these
patients however showed an extremely poor outcome; survival rates 1 year after
transplantation were less than 50%. With such poor prognosis, the question arises
whether ELCS has a place. Due to the scarcity of available donor organs, all efforts
should be taken to maximize postoperative success rates in high-risk patients. Early
patient stabilization with ECLS could provide the tool needed for this. In our experi-
ence, we have learned that sedation, intubation, and mechanical ventilation are
associated with extremely poor postoperative survival. Therefore, our practice has
evolved toward avoiding mechanical ventilation. Should the patient already be on
mechanical ventilation, then weaning should be prioritized enabling them to eat,
drink, exercise, and communicate with medical staff. Our strategy to keep patients
awake on ECLS has been shown to be beneficial when compared to patients on
mechanical ventilation.

Other Indications for ECLS in Lung Transplantation

Intraoperative VA-ECMO

The development of the minimally invasive, bilateral lung transplantation technique

has had several postoperative benefits for the patients. As for the intraoperative use
of mechanical circulatory support, this meant that central cannulation was no longer
an option, thus the use of cardiopulmonary bypass (CPB) was reconsidered. Major
disadvantages to CPB include the need for full anticoagulation (leading to increased
risk of bleeding) and activation of nonspecific inflammation due to the large artifi-
cial surface area and blood-air contact. Our experience showed improved postopera-
tive outcome in patients treated with intraoperative ECLS when compared to CPB
[14]. Therefore, we have changed our strategy toward peripheral VA-ECMO
implantation via the groin. Percutaneous cannulation enables relatively easy can-
nulation access and removal of the cannulas when ECMO is no longer needed.

Postoperative ELCS

In selected patients, prolonged postoperative VA-ECMO may be beneficial after

lung transplantation. Severe impairment of pulmonary function leading to swelling
of the pulmonary parenchyma and thus impairment of gas exchange is often seen in
patients with acute ischemia-reperfusion damage. Impaired gas exchange generally
11 ECMO as a Bridge to Lung Transplantation 213

leads to elevation of CO2 concentrations and falling O2 concentrations despite

potentially damaging levels of mechanical ventilation. As a result, pulmonary vas-
cular resistance increases with a decrease in cardiac output leading to circulatory
failure and acute kidney failure. Immediate stabilization can be achieved using
VA-ECMO support by unloading the heart and restoring cardiac output. After stabi-
lization, lung-protective mechanical ventilation is often possible and recovery from
ischemia-reperfusion damage is promoted.
The second indication for postoperative VA-ECMO is circulatory support in
patients posttransplant due to long-lasting pulmonary hypertension. Early results in
lung transplantation patients due to pulmonary hypertension were very discourag-
ing due to high rates of pulmonary failure after transplant. In pulmonary hyperten-
sion, the left ventricle is usually small and stiff. After transplantation, blood flow via
the lungs is normalized and an increased volume is pumped from right to left. As the
left ventricle is relatively weak, it is unable to cope with normal left ventricular
blood flow. Increased left ventricular volume load usually leads to rising left-sided
filling pressures and development of acute pulmonary edema. When not corrected,
irreversible graft failure occurs. Therefore, VA-ECMO is implanted at the beginning
of transplantation to reduce the preload of both ventricles. After transplantation, the
patient is transferred to the ICU with ongoing VA-ECMO support, and weaning
from mechanical ventilation is attempted as soon as possible. After successful extu-
bation, VA-ECMO may be weaned over 5 days, under continuous monitoring of the
left ventricular function and filling pressures.

ECLS for Organ Preservation

To maximize organ donor acceptance, novel organ perfusion systems have been
developed in recent years. In ex vivo lung perfusion, donor lungs are connected to a
system of tubes and perfused to evaluate marginal organs. Parameters such as oxy-
gen uptake, lung elasticity, perfusion, and ventilation pressures determine the suit-
ability for transplantation. Organs that seem to be marginal at first glance may still
be accepted once an adequate function is confirmed using the organ perfusion sys-
tem. Furthermore, due to continuous perfusion of the organ with a normothermic
solution containing nutrients, cold ischemic time can be significantly reduced and
donor organ damage may be ameliorated. To date, extensive research in the normo-
thermic perfusion system has not shown a benefit for medium-term organ function
[15]. Further development of this system into a portable system, the organ care
system (OCS), has led to a pilot study between the Hannover group in cooperation
with colleagues from Madrid [16]. In this study, 12 donor lungs were transplanted
after initial cold perfusion followed by warm perfusion and ventilation using the
OCS. Follow-up data has not shown a benefit in short-term survival [17]. However,
further research must be done to determine whether the use of the OCS has a benefit
in organ function in the long term.
214 C. Kuehn and R. Natanov

Conclusions

Continued increase in the number of lung transplantations and technological

advancements has transformed the use of ECLS from a short-term stabilization to a
support system that can be maintained for weeks and even months. Improvement of
technology and the development of ECLS experience have further reduced the com-
plication rate and have enabled the use of ECLS in a wide variety of patients as a
BTT. The concept of awake ECMO is a recent milestone in the treatment of end-­
stage pulmonary disease. Interhospital patient transfer on ECLS has furthermore
enabled the possibility of transplantation in patients otherwise physically not able to
make the transfer into the transplantation center. The use of periprocedural ECLS is
feasible and improves outcome in patients with postoperative pulmonary edema and
pulmonary hypertension. Regular use of ECLS in modern transplantation centers
should therefore be established as part of the treatment spectrum.

References

1. Jurmanm MJ, Haverich A, Demertzis S, Schaefers HJ, Wagner TO, Borst HG. Extracorporeal
membrane oxygenation as a bridge to lung transplantation. Eur J Cardiothorac Surg.
1991;5(2):94–7.
2. Fischer S, Simon AR, Welte T, Hoeper MM, Meyer A, Tessmann R, et al. Bridge to lung
transplantation with the novel pumpless interventional lung assist device NovaLung. J Thorac
Cardiovasc Surg. 2006;131(3):719–23.
3. Smits JMA, Mertens BJA, Van Houwelingen HC, Haverich A, Persijn GG, Laufer G. Predictors
of lung transplant survival in eurotransplant. Am J Transplant. 2003;3(11):1400–6.
4. Agati S, Ciccarello G, Fachile N, Scappatura RM, Grasso D, Salvo D, et al. DIDECMO: a new
polymethylpentene oxygenator for pediatric extracorporeal membrane oxygenation. ASAIO
J. 1992;52(5):509–12.
5. Fischer S, Hoeper MM, Bein T, Simon AR, Gottlieb J, Wisser W, et al. Interventional lung
assist: a new concept of protective ventilation in bridge to lung transplantation. ASAIO
J. 2008;54(1):3–10.
6. Kinaschuk K, Bozso SJ, Halloran K, Kapasi A, Jackson K, Nagendran J. Mechanical circula-
tory support as a bridge to lung transplantation: a single Canadian institution review. Can
Respir J. 2017;2017:5947978.
7. Secretariat MA. Extracorporeal lung support technologies – bridge to recovery and bridge to
lung transplantation in adult patients. Ont Health Technol Assess Ser. 2010;10:1–47.
8. Hoeper MM, Tudorache I, Kühn C, Marsch G, Hartung D, Wiesner O, et al. Extracorporeal
membrane oxygenation watershed. Circulation. 2014;130:864–5.
9. Fuehner T, Kuehn C, Hadem J, Wiesner O, Gottlieb J, Tudorache I, et al. Extracorporeal mem-
brane oxygenation in awake patients as bridge to lung transplantation. Am J Respir Crit Care
Med. 2012;185(7):763–8.
10. Ius F, Natanov R, Salman J, Kuehn C, Sommer W, Avsar M, et al. Extracorporeal membrane
oxygenation as a bridge to lung transplantation may not impact overall mortality risk after
transplantation: results from a 7-year single-centre experience. Eur J Cardiothorac Surg.
2018;54(2):334–40.
11. Salman J, Ius F, Sommer W, Siemeni T, Kuehn C, Avsar M, et al. Mid-term results of bilateral
lung transplant with postoperatively extended intraoperative extracorporeal membrane oxy-
genation for severe pulmonary hypertension. Eur J Cardiothorac Surg. 2017;52(1):163–70.
11 ECMO as a Bridge to Lung Transplantation 215

12. Salman J, Bernhard BA, Ius F, Poyanmehr R, Sommer W, Aburahma K, et al. Intraoperative
extracorporeal circulatory support in lung transplantation for pulmonary fibrosis. Ann Thorac
Surg. 2021;111(4):1316–24.
13. Fleissner F, Mogaldea A, Martens A, Natanov R, Rümke S, Salman J, et al. ECLS supported
transport of ICU patients: does out-of -house implantation impact survival? J Cardiothorac
Surg. 2021;16(1):1–6.
14. Ius F, Kuehn C, Tudorache I, Sommer W, Avsar M, Boethig D, et al. Lung transplantation on
cardiopulmonary support: venoarterial extracorporeal membrane oxygenation outperformed
cardiopulmonary bypass. J Thorac Cardiovasc Surg. 2012;144(6):1510–6.
15. Nilsson T, Wallinder A, Henriksen I, Nilsson JC, Ricksten SE, Møller-Sørensen H, et al. Lung
transplantation after ex vivo lung perfusion in two Scandinavian centres. Eur J Cardiothorac
Surg. 2019;55(4):766–72.
16. Warnecke G, Moradiellos J, Tudorache I, Kühn C, Avsar M, Wiegmann B, et al. Normothermic
perfusion of donor lungs for preservation and assessment with the organ care system lung
before bilateral transplantation: a pilot study of 12 patients. Lancet. 2012;380(9856):1851–8.
17. Warnecke G, Van Raemdonck D, Smith MA, Massard G, Kukreja J, Rea F, et al. Normothermic
ex-vivo preservation with the portable Organ Care System Lung device for bilateral lung trans-
plantation (INSPIRE): a randomised, open-label, non-inferiority, phase 3 study. Lancet Respir
Med. 2018;6(5):357–67.
Chapter 12
Daily Management of Patients on VV
ECMO

Charles Rappaport and Kristina Rappaport

The Daily Survey

Daily management consists of judging gas exchange, reviewing organ system func-
tion, identifying and preventing complications, and ensuring communication
between the ECMO managing LIP, ECMO specialists, bedside nurses, respiratory
and physical therapists, and the patient or surrogate decision-makers (see Table 12.1).

Gas Exchange

The fundamental purpose of ECMO is to accomplish gas exchange when the native
lung cannot. The factors underlying adequate oxygen and carbon dioxide exchange
are described in Chap. 1 and these should be reviewed systematically on (at least) a
daily basis.
Oxygenation Assuming that the patient’s native cardiovascular system is intact, an
acceptable arterial oxygen saturation (SaO2) assures adequate systemic oxygen-
ation. Arterial saturation is measured using pulse oximetry (SpO2) or arterial blood
gas analysis. In patients on VV ECMO, values are often lower than typically toler-
ated in mechanically ventilated patients. For example, an SpO2 > 75% may be toler-

C. Rappaport (*)
Division of Pulmonary, Critical Care and Occupational Medicine, University of Iowa Hospital
and Clinics, Iowa City, IA, USA
e-mail: [email protected]
K. Rappaport
University of Iowa Hospital and Clinics, Iowa City, IA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 217

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_12
218 C. Rappaport and K. Rappaport

Table 12.1 Daily survey checklist

1. Review ECMO orders and parameters to make sure they are accurate and up to date
2. Review and interpret blood gas samples
 Pre-oxygenator and post-oxygenator blood gases; ensure adequate CO2 clearance and
appropriate post-oxygenator PO2
 Patient ABG; venous oxygen saturation trend (developing shock, recirculation)
3. Review sweep gas flow and circuit blood flow trends over the past 24 hours
4. Review coagulation profile and anticoagulation trend
5. Review hemoglobin and lactate trends
6. Review any new and/or relevant imaging
7. Examine the patient; POCUS for cardiac assessment and cannula position
8. Discuss concerns with the ECMO specialist, nurse, and all other team members on rounds.
This should include a daily discussion about trialing off ECMO support
9. Include family discussion on rounds of possible
10. Review the intended bridge for the patient and if that bridge is still realistic

ated well, especially when there is no evidence of organ dysfunction or lactic

acidosis. Nevertheless, values in the high 80s or low 90s can be achieved in most
patients.
When systemic oxygenation is adequate, it may be appropriate to consider
whether weaning from ECMO should be attempted, as described further in Chap.
15, or whether the pump speed (and circuit blood flow) can be reduced somewhat.
In general, circuit flow should always be maintained above 1.5 L/min to reduce the
risk of thrombosis and membrane dysfunction. If systemic oxygenation is not
acceptable, attention should turn to circuit blood flow, membrane lung function,
recirculation, and the venous oxyhemoglobin saturation (which may signal circula-
tory failure).
Circuit blood flow: Since oxygenation relies critically on capturing a substantial
fraction of the cardiac output, inadequate circuit blood flow must be addressed.
Suboptimal flow is usually due to hypovolemia or cannula malposition, although
device malfunction, thrombosis, excessive pump speed, or cannula kinking contrib-
utes occasionally.
The outflow cannula lowers the local pressure within the floppy vena cavae,
tending to collapse the vessel and block further outflow. This often manifests as
“chugging” of the external cannulas or tubing, increasingly negative pump inlet
pressures, or alarms signaling low flow. Lowering the pump speed may occasionally
restore flow but, more often, additional steps are needed. Typically, this involves a
fluid bolus, which raises the transmural IVC pressure, limiting its tendency to col-
lapse. Judging overall intravascular volume is complex in the ECMO patient, as
reviewed in Chap. 7. Point-of-care-ultrasound (POCUS) is often informative as part
of a comprehensive assessment (see Table 12.2). Since ARDS is usually managed
with a restricted fluid approach [1], indiscriminate fluid loading should be avoided.
The propensity of the IVC to collapse is affected also by changing pleural pressure,
so that inspiratory efforts and PEEP play a role at times. Especially when circuit
12 Daily Management of Patients on VV ECMO 219

Table 12.2 POCUS exam for the ECMO patient

1. Cannula position, including color Doppler to judge the return jet
2. IVC and tendency to collapse (assuming active inspiration) as one component of judging
volemia
3. RV function, since this is often impaired and may worsen when the lung lesion is progressive
 (a) TAPSE
 (b) RV:LV ED area
 (c) S′
4. LV function
5. Lung B lines, especially during PEEP-titration or in response to clinical change
6. Seeking sites of bleeding or pleural disease, pleural space, abdomen, groin, or thigh

flow compromise follows PEEP reduction or lowered sweep gas flow (and accord-
ingly greater inspiratory efforts), attention to the ventilator and patient’s work of
breathing is necessary. Abolishing inspiratory effort by raising sweep gas flow or
infusing narcotics may be simpler and more effective than giving repeated fluid
boluses.
The other common basis for poor circuit blood flow is cannula malposition. One
common example is migration of the distal aspect of a dual-lumen internal jugular
(dl Vj-V) into the hepatic vein. This is readily identified with POCUS, using a sub-
xiphoid or transhepatic window, and aided by color Doppler flow imaging.
Inadvertent withdrawal of a cannula, as can occur during turning, mobilization, pro-
ning, or transport of the EMCO patient may also be to blame and can be detected
with POCUS or radiographic imaging. The daily survey for patients with dl Vj-V
cannulation includes using color Doppler to identify the return jet [directed toward
the tricuspid valve (TV)]. The operator should ensure that the jet has not rotated
away from the TV, migrated distally (e.g., hepatic vein), or shifted proximally (in
the superior vena cava). For those with Vf-Vf cannulation, color Doppler can iden-
tify the return jet near or in the right atrium, rather than distally (superior vena cava)
or proximally (toward the hepatic vein). If POCUS windows do not allow adequate
visualization, chest (for Vj) or abdominal (for Vf) plain radiographs can detect gross
malposition but not the orientation of the return jet.
Membrane lung function: With a fully functional oxygenator, the outlet PO2 is
usually >250 mmHg and outlet saturation (SPOSTO2) is 1.0. Additional key indicators
of membrane lung function include the pre-oxygenator-to-post-oxygenator PCO2
difference (normally >10 mmHg), which signals clearance of CO2, and the mem-
brane pressure drop, normally <50 mmHg (see also Chap. 9). Most often deteriora-
tion in membrane lung function develops over many hours, even days, but
occasionally can evolve over a much shorter period, producing a crisis (see
Chap. 13).
Recirculation: Recirculation occurs when highly oxygenated blood, having just
returned to the body, is withdrawn again via the drainage cannula, reducing the
efficiency of ECMO support. At times, recirculation can be so dramatic as to pro-
duce critical systemic hypoxemia. More often, the ECMO specialist notices brief
220 C. Rappaport and K. Rappaport

flashes of bright red blood in the withdrawal cannula. Recirculation is confirmed

when the inlet (SPREO2) and outlet saturations (SPOSTO2) are high and approximate
each other. The recirculation fraction (Rf) can be calculated as (SPREO2 – ­SvO2)/
(SPOSTO2 – SvO2), where SvO2 is the mixed venous saturation. Recirculation is usu-
ally minor when dual-lumen cannulas are used (hepatic vein malposition is an
exception) but can be a major problem when two cannulas are employed. Clinically
significant recirculation may require repositioning of cannulas, although sometimes
simply lowering the pump speed is effective.
Venous oxyhemoglobin saturation: When the patient’s overall circulation is
impaired, SvO2 falls as described by the Fick principle. Since some of this low oxy-
gen blood is not captured by the withdrawal cannula and traverses the lung (where
it does not become well-saturated), low cardiac output can amplify arterial hypox-
emia. In this situation, identifying the basis for circulatory compromise (hypovole-
mia, right ventricular failure, or other) may be necessary to restore adequate
systemic oxygenation.
Carbon dioxide elimination ECMO is more efficient for CO2 exchange than for
oxygenation, so that achieving CO2 elimination is generally much more readily
achieved. Adjusting sweep gas flow has a prompt and usually dramatic impact on
CO2 diffusion across the membrane lung. Nevertheless, there are highly significant
interactions between sweep gas flow and patient-ventilator interactions (assuming
the patient is not therapeutically paralyzed) that must be kept in mind during the
daily survey, following changes in sweep gas flow, after modifications to sedatives
and analgesics, and when mobilizing the patient.
While the patient is paralyzed, sweep gas flow is typically targeted to the arterial
partial pressure of CO2 (PaCO2). For patients beginning ECMO with very elevated
values for PaCO2, a gradual reduction toward normal is advised to reduce the risk of
neurological complications, as discussed below. Once paralytic drugs are stopped,
sweep gas flow should be set in accord with the desired work of breathing, not arbi-
trary PaCO2 levels. Early in the course of illness, the patient’s respiratory effort
should generally be minimized, raising sweep gas flow to lower central respiratory
drive. This allows for maximal lung healing and minimizing ventilator-induced lung
injury (VILI), as spontaneous breathing efforts can amplify lung injury [2]. Increases
in sedation are often ineffective and may worsen delirium and hinder early mobili-
zation. The daily survey should seek to judge work of breathing by examining respi-
ratory muscle use and respiratory rate, examining ventilator graphical displays of
pressure and flow, and measuring the airway occlusion pressure 100 msec following
the onset of inspiration (P0.1). A P0.1 value of less than 1.1 cmH2O implies low respi-
ratory drive. Conversely, a P0.1 value of greater than 4.0 cm H2O can accurately
identify excessive (and likely injurious) respiratory effort [3].
A common error when setting the sweep gas flow is to respond to a low PaCO2
of, say, 30 mmHg in an awake, actively breathing ECMO patient by lowering
sweep gas flow. This has the effect of shifting the burden of CO2 excretion to the
native lung and respiratory muscles and away from the ECMO circuit. Of course,
if the goal is to determine readiness for weaning, lowering sweep gas flow is
12 Daily Management of Patients on VV ECMO 221

appropriate. If the plan, however, is to support the patient as a bridge to lung heal-
ing, lowering sweep gas flow is counterproductive. A better approach is to ignore
the PaCO2 (and pH), leave the sweep gas flow at a level that accommodates a low
work of breathing, and wait until native lung function has improved. Some patients
exhibit high work of breathing despite raising the sweep gas flow, even once
PaCO2 is dramatically lowered. Some conditions to include in the differential
diagnosis of such patients are agitation, anxiety, severe pain, developing shock,
fever, or systemic inflammation. Without identifying and treating these underlying
conditions, high respiratory drive may persist despite increased extracorporeal
CO2 removal [4].
As the patient’s trajectory stabilizes, greater respiratory work may be acceptable,
or even desired, in accord with the principles of weaning described in Chap. 15.
Spontaneous breathing while on VV ECMO has been shown to reduce ventilation-­
perfusion (V/Q) mismatch, preserve respiratory muscle and diaphragm tone, and
preserve functional residual capacity [5]. The managing physician should assess
daily the status of native lung function and the role for continuing ECMO support.
If native lung function is improving, then support can be shifted from ECMO to the
patient and ventilator. Reduced sweep gas flow should be monitored by clinical
exam, ventilator graphics, and P0.1 as described above. If the inevitable increase in
respiratory effort is well-tolerated, this suggests that further reductions in sweep gas
flow may be achieved.

Ancillary Therapies

Antithrombotic Therapy

The managing physician should review the coagulation profile, target measures for
antithrombotic therapy, and any evidence of bleeding that might prompt a lower
antithrombotic goal or cessation of anticoagulation altogether. The antithrombotic
regimen (heparin, bivalirudin, argatroban) and the measures of antithrombotic effect
(aPTT, thromboelastogram, anti-factor Xa levels, activated clotting time) are center-­
specific as further described in Chap. 8. The anticoagulation titration algorithm
should be reviewed with the ECMO specialist on daily rounds to ensure the level of
anticoagulation is appropriate. Signs of active, external bleeding should prompt the
team to either pause anticoagulation or switch to a lower-dose algorithm. Internal
bleeding is harder to appreciate, but a significant hemorrhage can cause a reduction
in ECMO circuit flow, reduced central venous saturation, lactic acidosis, fall in
hemoglobin concentration, or signs of shock.
Low platelet counts often prompt concern for heparin-induced thrombocytope-
nia (HIT). However, thrombocytopenia is quite common in patients on ECMO,
whereas proven HIT is quite uncommon, with a reported incidence of <1% [6]. If
HIT is diagnosed or strongly suspected, heparin must be stopped and treatment with
argatroban or bivalirudin initiated. Furthermore, many circuits are heparin-coated,
222 C. Rappaport and K. Rappaport

thus a circuit change is necessary to eliminate all contact with heparin. Given the
nontrivial implications of treating for HIT, testing should be done only if the clinical
suspicion is quite high (see Chap. 8).

Renal Replacement Therapy

Acute kidney injury (AKI) is quite common in patients requiring VV ECMO,

exceeding 70% in some reports [7]. The mechanism is multifactorial and includes
critical hypoxemia, hypotension, low cardiac output, and septic shock. A high per-
centage of patients who develop AKI while on VV ECMO will eventually require
renal replacement therapy (RRT), and these patients have an increased risk of mor-
tality compared to patients who do not require RRT [8]. Interestingly, the timing of
RRT in critically ill patients has not been shown to affect mortality. Early initiation
of RRT compared to standard timing did not reduce 90-day mortality in critically ill
patients with AKI [9, 10]. If the patient does indeed require RRT for either electro-
lyte abnormalities or intractable volume overload refractory to diuretics, CRRT is
usually the preferred method as it limits large volume shifts. Access can be achieved
by adding a hemofilter in line with the ECMO circuit, combining the CRRT machine
with the ECMO circuit (see Chap. 2), or by obtaining separate vascular access
(sometimes undesirable due to anticoagulation) [11, 12].

Prone Position and Mobilization

Prone positioning reduces mortality in severe acute respiratory distress syndrome

(ARDS) [13]. Physiologically, prone ventilation improves static lung compliance,
facilitates more homogenous distribution of ventilation-induced lung stress, and
unloads the right ventricle [14]. Although SaO2 tends to rise when ARDS patients
are ventilated in the prone position, such improvement probably does not contribute
the mortality benefit [15] so that prone ventilation might be beneficial in ECMO
patients even once SaO2 is normalized. Potentially beneficial physiologic effects
have been demonstrated in patients supported on VV ECMO [16] and observational
and retrospective studies support a mortality benefit [17, 18].
Proning requires a highly experienced team. Special consideration should be
paid to the circuit location, cannula stability, and ventilator circuit so that positional
changes do not threaten these vital support devices. Proper skin interventions for
cannula sites and circuitry help to prevent pressure injuries and skin breakdown.
Any procedures (line insertion) or diagnostic procedures (POCUS, ECG) should be
completed prior to turning the patient. It is prudent to conduct a POCUS exam fol-
lowing proning to reevaluate cannula position, volume status, and biventricular
function (an apical four-chamber view is generally feasible while prone).
In critically ill patients, early sedative interruption and physical therapy reduces
the duration of delirium, mechanical ventilation, and ICU stay, while enhancing
functional status at hospital discharge [19, 20]. Physical therapy in critically ill
12 Daily Management of Patients on VV ECMO 223

patients has also been shown to reduce muscle wasting [21]. These findings likely
apply also to ECMO patients. Mobilization is feasible and safe when conducted in
experienced centers, even for patients with femoral cannulas (see Chap. 14). Given
the complexity of lines and devices, a multidisciplinary team of nurse, respiratory
therapist, ECMO specialist, physical therapist, occupational therapist, and intensiv-
ist is advisable [22].

Necessity of Lines and Tubes

Managing physicians and specialists should engage in a daily assessment of the

need for central venous catheters, arterial lines, urinary catheters, and the endotra-
cheal tube (ETT). Prolonged, stable, supported patients might benefit from removal
of hardware to prevent infections. When deciding to remove any lines, the team
must be cognizant of the fact that the patient will be on anticoagulation, thus replac-
ing lines carries an above average bleeding risk.
Extubation is feasible in some ECMO patients [23]. When ECMO is successful
in fully replacing lung function, the ventilator and ETT may no longer be essential.
This is an attractive option for carefully selected patients and could limit VILI,
reduce infection risk, and aid in patient comfort. Extubation also limits the amount
of sedation a patient receives which can help reduce delirium and maximize physi-
cal therapy. Despite these potential benefits, a EuroELSO survey showed that only
32% of providers surveyed responded that their center extubates patients prior to
weaning ECMO support [24]. The decision to extubate the patient on ECMO is usu-
ally not a straightforward one. It requires judging whether the work of breathing
will be acceptable and is likely to stay acceptable; this cannot be assumed even if
gas exchange is successfully carried out by the ECMO circuit. Moreover, PEEP
itself may limit atelectrauma, a potential benefit that is lost with extubation.
Nevertheless, for many awake ECMO patients who tolerate spontaneous breathing,
a trial of extubation is reasonable. On the other hand, for those with neurological
compromise or who demonstrate distress or excessive breathing effort, tracheos-
tomy may be indicated.

Assessing Other Systems

Hemodynamics

Acceptable cardiovascular function is a prerequisite when considering candidacy

for VV ECMO. Cardiac assessment prior to cannulation is very important, as
patients with severe RV or LV dysfunction might need circulatory support in addi-
tion to respiratory support (VA ECMO). Isolated RV dysfunction is not a contrain-
dication to VV ECMO per se, especially in cases where it is judged to be a
consequence of severe lung disease. In fact, RV function can improve following VV
224 C. Rappaport and K. Rappaport

ECMO initiation due to decreased hypoxic vasoconstriction, improved pulmonary

artery pressure, and correction of severe hypoxemia [25, 26]. Once VV ECMO is
initiated, the patient’s native circulation should be assessed regularly, relying on
vital signs, central venous oxygen saturation (and trends), vasoactive drug infu-
sions, and POCUS of RV and LV.
Every attempt should be made to optimize volume status while on VV ECMO as
described in Chap. 7. The percentage of venous capture determines the ability to
oxygenate patients on VV ECMO, so that hypovolemia can make oxygenation quite
challenging. ECMO flow rates have been shown to inversely correlate with the IVC
distensibility index [27]. Most patients placed on VV ECMO for respiratory failure
have received diuresis prior to cannulation. Initial cannula placement causes an
inflammatory response which further reduces intravascular volume. In the initial
phase of VV ECMO support, patients might need volume augmentation rather than
diuresis to maintain adequate circuit flow. Diuresis can resume once a steady state
has been reached and ECMO support is acceptable. Dynamic, rather than static,
fluid-response predictors are preferable, such as by combining passive leg raising
(PLR) with transthoracic echocardiography (LV outflow tract velocity-time integral
from an apical five-chamber view) [28]. Given the very low tidal volumes typically
used in ECMO patients, respiration-related metrics (such as pulse pressure or stroke
volume variation) will not be valid. Similarly, end-tidal CO2 values should not be
used to judge the impact of PLR since most CO2 is eliminated through the ECMO
circuit.
Changes in vital signs, increased vasoactive drug requirements, or increased
recirculation should prompt a comprehensive reassessment of volume status, ven-
tricular function, and vascular tone. VV ECMO itself has no significant hemody-
namic effect: therefore, hemodynamic crises in ECMO patients should be evaluated
similarly to other ICU patients. Of course, some causes of hemodynamic instability
(hemorrhagic hypovolemia, RV failure) are particularly common in ECMO patients
and should lead the differential diagnosis. If cardiac function declines, inotropic
support, ventilator adjustments, or pulmonary vasodilators may be indicated as dis-
cussed in Chap. 7 but, if medical management fails, mechanical support or conver-
sion to VA ECMO might be necessary.

Neurological Function

New neurological events are common on VV ECMO and portend a poor prognosis.
Intracranial bleeding occurs in approximately 5% of patients [29]. Multiple mecha-
nisms are likely in play, but AKI and rapid reductions in PaCO2 after ECMO initia-
tion (with subsequent alterations in cerebral blood flow) have been independently
associated with intracranial bleeding [30]. Interestingly, intracranial bleeding has
not been associated with anticoagulation profile, age, or comorbid conditions. The
daily survey should include a detailed neurological assessment, preferably after
sedative interruption.
12 Daily Management of Patients on VV ECMO 225

Circuit Evaluation and Safety

Every shift change, emergency interventions should be identified and properly

prepped and ready for utilization should the need arise. These include, but are not
limited to, emergency ventilator settings, backup-circuit location, ensuring that the
circuit is plugged into emergency power outlets, tubing, and clamp location in case of
a circuit emergency that would require clamping off. A thorough circuit assessment
should be carried out at each shift change as well or more often based on circuit health
or concerns (increasing clot formation, decreasing blood flow through circuit, sub-
therapeutic anticoagulation, etc.). Circuit assessments should involve the following:
(a) Using a light source to look for any clot or fibrin formation
(b) Checking the gas source to ensure it is properly connected and flowing with
no leaks
(c) Assuring adequate membrane lung function based on color change pre- to post-­
oxygenator or more completely as described in Chap. 9
(d) Making sure connectors are tight and ty-bands are intact
(e) Utilizing skin pressure interventions at any sites that the circuit is touching the
patient’s skin to prevent any pressure injuries
The bedside specialist is responsible for the cannula assessment as well. This
usually occurs during circuit assessment and involves checking cannula measure-
ments to monitor for cannula migration, assessing sites for bleeding or signs of
infection, ensuring that the cannula dressing is intact and noting when the next
dressing change is due, and making sure that suture sites are intact. If a patient is
placed prone, securement devices may need to be moved more posteriorly to keep
little of the circuit underneath the patient. If there is a concern with the cannula
assessment, the managing physician and cannulating surgeon should be notified
promptly.

Goals of Therapy

VV ECMO is a bridge to recovery or lung transplant, rather than a destination ther-

apy. This should be clearly outlined prior to cannulation and reassessed daily to
ensure that the goal is realistic. Prior to cannulation, the managing physician should
obtain informed consent from the patient or surrogate decision-maker, clearly out-
line what ECMO is a bridge to, explain the limitations of VV ECMO, and review
expectations in case complications arise or the prognosis worsens. Multidisciplinary
conversations are essential daily (occasionally more than once per day) to make sure
each team member understands and agrees with the plan. Family members should
be updated every day, ideally as part of multidisciplinary rounds. This allows for
transparency and a chance for family to air their concerns and ask questions. If
patient prognosis worsens or complications arise due to continuing ECMO (bleed-
ing, infection, oxygenator failure, etc.), the conversation then shifts to discontinuing
ECMO as discussed more fully in Chap. 17.
226 C. Rappaport and K. Rappaport

References

1. The National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS)
Clinical Trials Network. Comparison of two fluid-management strategies in acute lung injury.
N Engl J Med. 2006;354:2564–75.
2. Mauri T, Grasselli G, Suriano G, Eronia N, et al. Control of respiratory drive and effort in
extracorporeal membrane oxygenation patients recovering from severe acute respiratory dis-
tress syndrome. Anesthesiology. 2016;125:159–67.
3. Telias I, Junhasavasdikul D, Rittayamai N, Piquilloud L, et al. Airway occlusion pressure as
an estimate of respiratory drive and inspiratory effort during assisted ventilation. Am J Respir
Crit Care Med. 2020;201(9):1086–98.
4. Crotti S, Bottino N, Ruggeri GM, Spinelli E, et al. Spontaneous breathing during extracorpo-
real membrane oxygenation in acute respiratory failure. Anesthesiology. 2017;126:678–87.
10.1097.
5. Langer T, Vecchi V, Belenkiy SM, Cannon JW, et al. Extracorporeal gas exchange and spon-
taneous breathing for the treatment of acute respiratory distress syndrome: an alternative to
mechanical ventilation? Crit Care Med. 2014;42(3):e211–20.
6. Natt B, Hypes C, Basken R, Malo J, et al. Suspected heparin-induced thrombocytope-
nia in patients receiving extracorporeal membrane oxygenation. J Extra Corpor Technol.
2017;49:54–8.
7. Chen Y, Tsai F, Fang J, Yang C. Acute kidney injury in adults receiving extracorporeal mem-
brane oxygenation. J Formos Med Assoc. 2014;113:778–85.
8. Kielstein J, Heiden A, Beutel G, Gottlieb J, et al. Renal function and survival in 200 patients
undergoing ECMO therapy. Nephrol Dial Transplant. 2013;28:86–90.
9. Timing of Initiation of Renal-Replacement Therapy in Acute Kidney Injury The STARRT-AKI
Investigators for the Canadian Critical Care Trials Group, the Australian and New Zealand
Intensive Care Society Clinical Trials Group, the United Kingdom Critical Care Research
Group, the Canadian Nephrology Trials Network, and the Irish Critical Care Trials Group. N
Engl J Med. 2020;383:240–51.
10. Cove ME, MacLaren G, Brodie D, Kellum J. Optimising the timing of renal replacement
therapy in acute kidney injury. Crit Care. 2021;25:184.
11. Askenazi DJ, Selewski DT, Paden ML, et al. Renal replacement therapy in critically ill patients
receiving extracorporeal membrane oxygenation. Clin J Am Soc Nephrol. 2012;7(8):1328–36.
12. Dyk M. The use of CRRT in ECMO patients the Egyptian. J. Crit. Care Med. 2018;6(3):95–100.
13. Guérin C, Reignier J, Richard JC, Beuret P, et al. PROSEVA Study Group. Prone positioning
in severe acute respiratory distress syndrome. N Engl J Med. 2013;368(23):2159–68.
14. Jozwiak M, Teboul JL, Anguel N, Persichini R, et al. Beneficial hemodynamic effects of prone
positioning in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med.
2013;188(12):1428–33.
15. Albert RK, Keniston A, Baboi L, Ayzac L, et al. Prone position-induced improvement in gas
exchange does not predict improved survival in the acute respiratory distress syndrome. Am J
Respir Crit Care Med. 2014;189:494–6.
16. Franchineau G, Bréchot N, Hekimian G, et al. Prone positioning monitored by electrical
impedance tomography in patients with severe acute respiratory distress syndrome on veno-­
venous ECMO. Ann Intensive Care. 2020;10(1):12.
17. Guervilly C, Prud'homme E, Pauly V, Bourenne J. Prone positioning and extracorporeal mem-
brane oxygenation for severe acute respiratory distress syndrome: time for a randomized trial?
Intensive Care Med. 2019;45(7):1040–2.
18. Giani M, Martucci G, Madotto F, Belliato M, et al. Prone positioning during venovenous extra-
corporeal membrane oxygenation in acute respiratory distress syndrome: a multicenter cohort
study and propensity-matched analysis. Ann Am Thorac Soc. 2021;18:495–50.
12 Daily Management of Patients on VV ECMO 227

19. Schweickert WD, Pohlman MC, Pohlman AS, Nigos C, et al. Early physical and occupational
therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet.
2009;373(9678):1874–82.
20. Parker A, Sricharoenchai T, Needham DM. Early rehabilitation in the intensive care
unit: preventing physical and mental health impairments. Curr Phys Med Rehabil Rep.
2013;1(4):307–14.
21. Puthucheary ZA, Rawal J, McPhail M, Connolly B, et al. Acute skeletal muscle wasting in
critical illness. HEJAMA. 2013;310(15):1591–600.
22. Haji JY, Mehra S, Doraiswamy P. Awake ECMO and mobilizing patients on ECMO. Indian J
Thorac Cardiovasc Surg. 2021;37:309–18.
23. Ellouze O, Lamirel J, Perrot J, Missaoui A, et al. Extubation of patients undergoing extracor-
poreal life support. A retrospective study. Perfusion. 2019;34(1):50–7.
24. Justyna S, Shekar K, Protti A, Tukacs M, et al. Extubate before venovenous extracorporeal
membranous oxygenation decannulation or decannulate while remaining on the ventilator?
The EuroELSO 2019 weaning survey. ASAIO J. 2021;67(4):e86–9.
25. Srivastava MC, Ramani GV, Garcia JP, Griffith BP, et al. Veno-venous extracorporeal mem-
brane oxygenation bridging to pharmacotherapy in pulmonary arterial hypertensive crisis. J
Heart Lung Transplant. 2010;29(7):811–3.
26. Grant C, Richards JB, Frakes M, Cohen J, et al. ECMO and right ventricular failure: review of
the literature. J Intensive Care Med. 2021;36(3):352–60.
27. Patel K, Baran D, Nazir H, Anandarangam T. Effect of V-V ECMO flow rates on IVC diameter
variation on trans-thoracic echocardiography. Crit Care Med. 2014;42(12):1406–7.
28. Guinot PG, Zogheib E, Detave M, Moubarak M, et al. Passive leg raising can predict fluid
responsiveness in patients placed on venovenous extracorporeal membrane oxygenation. Crit
Care. 2011;15(5):R216.
29. Luyt CE, Bréchot N, Demondion P, Jovanovic T, et al. Brain injury during venovenous extra-
corporeal membrane oxygenation. Intensive Care Med. 2016;42(5):897–907.
30. Cavayas YA, Munshi L, Del Sorbo L, Fan E. The early change in PaCO2 after extracorporeal
membrane oxygenation initiation is associated with neurological complications. Am J Respir
Crit Care Med. 2020;201(12):1525–35.
Chapter 13
Crises During ECLS

Purnema Madahar, Dana A. Mullin, Meaghan Flatley, Darryl Abrams,

Phillipe H. Lemaitre, Daniel Brodie, and Cara Agerstrand

Introduction

The use and success of extracorporeal life support (ECLS) has historically been
limited by frequent and potentially severe complications. Advancements in circuit
technology and increased experience with ECLS management have reduced, but

P. Madahar · D. Brodie · C. Agerstrand (*)

Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine,
Columbia University College of Physicians and Surgeons/NewYork-Presbyterian Hospital,
New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]; [email protected];
[email protected]
D. Abrams
Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine,
Columbia University College of Physicians & Surgeons, New York, NY, USA
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]
D. A. Mullin
Department of Clinical Perfusion and Anesthesia Support Services, NewYork-Presbyterian
Hospital, New York, NY, USA
e-mail: [email protected]
M. Flatley
Department of Surgery, Columbia University College of Physicians and Surgeons/NewYork-­
Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]
P. H. Lemaitre
Center for Acute Respiratory Failure, New York-Presbyterian Hospital, New York, NY, USA
Division of Cardiothoracic Surgery, Department of Surgery, Columbia University College of
Physicians and Surgeons/NewYork-Presbyterian Hospital, New York, NY, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 229

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_13
230 P. Madahar et al.

certainly not eliminated, these complications in the modern era of ECLS. Accurate
estimates of current complication rates are difficult to enumerate as published rates,
and those found in registries, typically reflect the use of both older and newer tech-
nology and reporting of complications may be underestimated. The Extracorporeal
Life Support Organization (ELSO) registry, which compiles the most extensive
database on complications related to ECLS, includes ECLS cases with both modern
and outdated technology, as well as centers with greater and lesser degrees of expe-
rience with ECLS using a variety of ECLS configurations. Determining the expected
rate of complications in any given circumstance should be attempted with caution,
although the registry allows for some degree of risk adjustment.
Crises in ECLS can be partitioned into those primarily originating in the circuit
and those originating in the patient. Patient outcomes are inextricably linked to the
timely recognition and management of both. Recurring provider training is essential
to maintain these skills and facilitate appropriate emergency responses.

Circuit Crises

Emergency Management During Circuit Clamping

During any ECLS crisis requiring circuit clamping, all limbs of the circuit should be
clamped in the most proximal position to the patient, in order to isolate the patient
from the circuit (Fig. 13.1). Clamps should be placed directly on the cannula and are
best positioned between the stainless steel wires and the hard connector on the mal-
leable tubing. Depending on the configuration of the ECLS circuit, there should
always be at least one tubing clamp per cannula present at the bedside and easily
accessible during an emergency. In cases of massive hemorrhage, the tubing line
associated with the source of the bleed should be clamped first. If it is unclear where
the source of bleeding is occurring, then the reinfusion cannulae should be clamped

Fig. 13.1 Appropriate Blender

Oxygenator
clamp position when Console

isolating patient from

circuit. During an ECLS
crises in which the
integrity of the circuit is at Pump
risk or in question, the
patient may be separated
from the circuit by
clamping the drainage and
reinfusion limbs. (Courtesy
of Coach. Reprinted with
permission from
CollectedMed.com)
13 Crises During ECLS 231

first, followed by the drainage cannulae, and lastly the distal perfusion cannula, if
present. Clamping of the ECLS circuit requires simultaneous placement of the
patient on emergency ventilator settings or maximal supplemental oxygen or nonin-
vasive ventilatory support in non-intubated patients. Vasopressors and inotropes
should be readily available for hemodynamic support. Notify the ECMO specialist
and/or perfusionist immediately to help troubleshoot and assist with management.

Circuit Air

Any loss of circuit integrity may result in the entrainment of air into the circuit.
Circuit air is an emergency situation that must be addressed immediately; however,
since the oxygenator may serve as an air trap, risk to the patient is highest when air
is present on the post-oxygenator side of the circuit. In venovenous ECLS, circuit
air risks propagation of a venous air embolism. In venoarterial ECLS, or venove-
nous ECLS in the presence of an intracardiac or intrapulmonary shunt, there is a risk
of arterial air embolism.
Pre-oxygenator air can originate from peripheral or central venous lines that are
damaged, inadvertently left open, or accessed for intravenous infusions. Because
circuit pressure is negative between the drainage cannula and the pump, cracked
stopcocks, connectors, tubing, or loose connections can entrain air, as can partial
dislodgement of the drainage cannula, which results in exposure of a side port. Air
can arise within the oxygenator itself from improper priming, oxygenator mem-
brane rupture, or occlusion of the gas exhaust outlet. Post-oxygenator air is seen if
air passes through the oxygenator from the pre-oxygenator side or is directly
entrained through defects in the circuit after the oxygenator.
Another mechanism through which air may enter the circuit is known as cavita-
tion, which occurs when a high negative pressure within the drainage cannula or
tubing draws a large amount of gas out of solution. Extreme changes in pressure
induce hydrodynamic phenomena, including bubble nucleation, growth, and col-
lapse. This process can be minimized by avoiding extremes in negative pressure on
the venous side of the circuit. Cavitation can also occur in the setting of a high
sweep gas flow rate and a low blood flow rate.
In order to minimize introduction of air into the circuit, medication and blood
products should be administered to the patient through separate venous access, and
independent vascular access should be placed for continuous renal replacement
therapy, if possible. If the circuit must be used, the pre-oxygenator port/limb should
be accessed so that the oxygenator can trap any air inadvertently introduced (see
also Chap. 2). A bubble detector may also be placed on the post-oxygenator tubing,
which can be set to stop the pump immediately if air is detected.
Small amounts of air on the venous side of the circuit can often be removed
without interrupting blood flow by carefully withdrawing the air into a syringe. If a
large quantity of air is present or the patient is at particularly high risk of systemic
air embolization, ECLS should be suspended temporarily by clamping all of the
232 P. Madahar et al.

drainage and reinfusion limbs at the most proximal position to the patient, in order
to isolate the circuit from the patient. When air appears on the post-oxygenator side
of the circuit, the drainage and reinfusion lines should be clamped immediately,
thereby halting blood flow (Fig. 13.1). The patient should be placed in the
Trendelenburg position so that any embolization of air is directed toward the lower
extremities and emergency circuit clamping measures should be followed [1, 2].
To remove air from the circuit tubing, attach a syringe to an access port and posi-
tion the tubing that contains the air below the circuit access point. If the pump head
is affected, remove the pump head from its casing, hold it below the circuit access
point, and tap it until the air flows into the circuit tubing. The air will rise to the most
superior position, which should be toward the access point. Aspirate with the syringe
until all the air is removed. With a clamped circuit, removing volume creates even
more negative circuit pressure if fluid is not administered through a second access
point. When large volumes are withdrawn, excessive negative pressure risks hemo-
lysis and further entrainment of air through cavitation. If air is present within the
oxygenator itself, the oxygenator is removed from its holder and tapped gently until
air rises toward a port on its upper portion where it can be aspirated. If a significant
amount of air is present within the oxygenator, the circuit may have to be reprimed.
A cardiotomy reservoir spliced into the circuit can allow the circuit to be reprimed
more efficiently. In the event the circuit cannot be de-aired quickly and easily, the
oxygenator, or possibly the entire circuit, may need to be replaced.
If air reaches the patient, management is mostly supportive, although venous air
may be aspirated under some circumstances [1]. (Table 13.1 highlights the diagno-
sis and management of circuit-related crises.)

Thrombosis

The interaction between blood and the ECLS circuit may affect both procoagulant
and anticoagulant pathways, potentially resulting in thrombosis occurring in the
circuit or the patient. Thrombosis in the ECLS circuit appears as dark red clot or
white strands of fibrin (Fig. 13.2). Inadequate anticoagulation, disseminated intra-
vascular coagulation, heparin-induced thrombocytopenia with thrombosis (HITT),
antithrombin III deficiency, and other hypercoagulable states may further increase
the risk of circuit thrombosis, especially at sites of low or turbulent flow. A simpli-
fied circuit with only the essential components may reduce this risk.
Whereas older ECLS circuits required high levels of anticoagulation to prevent
thrombosis, newer biocompatible or heparin-coated circuit components can be
managed with lower levels of anticoagulation [3, 4]. Although there are no univer-
sally accepted anticoagulation guidelines in ECLS, the administration of heparin
with a goal activated partial thromboplastin time 1.5 times the normal reference
range or an anti-Xa level of 0.1–0.3 IU/mL has been suggested as sufficient to main-
tain circuit patency in many patients [5–8] (see also Chap. 8).
13 Crises During ECLS 233

Table 13.1 Circuit-related emergencies and management

Circuit crises Diagnostics Management
Circuit air
Pre-oxygenator Inspect for defects in pre-oxygenator Isolate air within tubing at closest
tubing and access points access point
Check drainage cannula for exposed Aspirate air through syringe
drainage ports Restore circuit volumea
Ensure no air in tubing or uncapped Repair or replace circuit defects
patient lines/disconnected devices
(CRRT, plasmapheresis, etc.)
Oxygenator Inspect for defects in pre-oxygenator Remove oxygenator from its
tubing and access points holder
Check drainage cannula for exposed Tap oxygenator to isolate air
drainage ports superiorly
Ensure no air in tubing or uncapped Aspirate air through syringe
patient lines/disconnected devices Restore circuit volumea
(CRRT, plasmapheresis, etc.) If large amount of air, the
oxygenator or circuit may need
to be reprimed or replaced
Post-oxygenator Inspect for defects in pre-oxygenator, Clamp arterial and venous limbs
oxygenator, and post-oxygenator of the circuit
tubing and access points Reduce RPMs
Check drainage cannula for exposed Follow emergency circuit
drainage ports clamping measures
Isolate air within tubing at closest
access point
Aspirate air through syringe
Replace blood volume removed
during aspiration
Pump/oxygenator failure
Thrombosis Assess oxygenator function; follow Ensure adequate systemic
pre- and post-oxygenator PO2 and anticoagulation
pressures Investigate causes of
Assess for patient-related diagnoses hypercoagulability
as precipitants of hypercoagulability If HITT suspected based on the
4 T’s score:
 Stop heparin
 Start direct thrombin inhibitor
 Measure platelet factor 4
antibody and confirm with
SRA
 Evaluate for patient
thrombosis
(continued)
234 P. Madahar et al.

Table 13.1 (continued)

Circuit crises Diagnostics Management
Oxygenator failure Assess oxygenator function; follow Prep for oxygenator changeout
pre- and post-oxygenator PO2 and Follow emergency circuit
pressures clamping measures
Prep ECLS tubing
Double clamp the drainage and
reinfusion limbs of the circuit
Sterile cut tubing and reconnect
to new oxygenator
Pump malfunction/ Check power supply Initiate hand-cranking
failure Check pump head integrity (assess Replace malfunctioning
for pump head clot) component or entire circuit
Assess for inadequate preload and/or
excessive afterload
Assess for hemolysis
Tubing disconnect/cannula fracture
Pre-pump Identify site of disconnect or fracture Reconnect tubing in cases of
disconnect
Clamp all limbs of the circuit to
isolate the patient
Reduce RPMs
Follow emergency circuit
clamping measures
Replace tubing at the site of the
defect in cases of fracture
Transfuse for clinically relevant
blood loss
Post-pump Identify site of disconnect or fracture Reconnect tubing in cases of
disconnect
Clamp all limbs of the circuit to
isolate the patient
Reduce RPMs
Follow emergency circuit
clamping measures
Replace tubing at the site of the
defect in cases of fracture
Transfuse for clinically relevant
blood loss
Tubing rupture
Pre-pump Identify site of rupture Clamp all limbs of the circuit to
isolate the patient
Reduce RPMs
Follow emergency circuit
clamping measures
Replace tubing at the site of the
defect
Transfuse for clinically relevant
blood loss
13 Crises During ECLS 235

Table 13.1 (continued)

Circuit crises Diagnostics Management
Post-pump Identify site of rupture Clamp all limbs of the circuit to
isolate the patient
Reduce RPMs
Follow emergency circuit
clamping measures
Replace tubing at the site of the
defect
Transfuse for clinically relevant
blood loss
Heat exchanger Assess for water in the oxygenator or Replace heat exchanger
malfunction blood in the water bath Replace oxygenator if evidence
Assess if blood temperature starts to of water or blood leakage
drop and patient starts to rapidly cool Turn off the heater/cooler or
unplug
Treat hypothermia if present
Inadvertent decannulation
Partial dislodgement Assess cannula for exposure of Reinforce cannula with sutures
with preserved blood drainage port If drainage port exposed, replace
flow cannula
Do not reintroduce exposed
portion of cannula
Assess ECLS blood flows and
pressures
Imaging to confirm cannula
placement after cannula is secure
Complete Clamp all limbs of the circuit to
dislodgement isolate the patient
or Reduce RPMs
Partial dislodgement Follow emergency circuit
with compromised clamping measures
blood flow Control cannulation site bleeding
Discontinue anticoagulation
Replace cannulae
Assess ECLS blood flows and
pressures
Imaging to confirm cannula
placement after cannula is secure
CRRT continuous renal replacement therapy, ECLS extracorporeal life support, HITT heparin-­
induced thrombocytopenia with thrombosis, PO2 partial pressure of oxygen, RPMs revolutions per
minute, SRA serotonin release assay
a
Blood aspirated from the circuit may be replaced by either transfusion or by leaving the venous
line open and using the patient as a reservoir
236 P. Madahar et al.

Fig. 13.2 Picture of blood

and fibrin clot in
oxygenator. Arrows depict
blood and fibrin clot
formation on the
oxygenator. (Courtesy of
Coach. Reprinted with
permission of
CollectedMed.com)

If thrombosis has developed in the setting of thrombocytopenia, a diagnosis of

heparin-induced thrombocytopenia with thrombosis (HITT) should be considered
[9–11]. If there is a reasonable suspicion for HITT, heparin should be stopped
immediately and anticoagulation continued with a direct thrombin inhibitor [9, 12–
15]. Measurement of platelet factor 4 antibody levels is the initial serologic screen-
ing test, which, if positive, should be followed by a serotonin release assay
(SRA) [16].

Oxygenator Failure

Modern polymethylpentene hollow fiber oxygenators are more efficient and durable
than older polypropylene or silicone oxygenators, though over time their effective-
ness may wane and oxygen transfer across the membrane should be assessed if there
is concern for an impairment in oxygenator function. A decreasing post-oxygenator
partial pressure of oxygen (PO2), an increasing transmembrane pressure gradient,
blood flow obstruction, development of circuit-related coagulopathy or hemolysis,
or a need to steadily increase sweep gas flow rate to manage the partial pressure of
arterial carbon dioxide (PaCO2) may be indications of a failing oxygenator [17] (see
also Chap. 9).
Thrombosis within the oxygenator is a common cause of an increase in the trans-
membrane pressure gradient and impaired oxygenator function. This phenomenon
can be partially assessed by visual inspection of the oxygenator during a circuit
check, though visualization of clot is limited to only the most superficial layers of
13 Crises During ECLS 237

the oxygenator membrane and may not be possible for the inlet side of the oxygen-
ator in some pump–oxygenator combinations. Therefore, monitoring the pressure
drop across the oxygenator (delta pressure) is essential. If the pressure gradient
increases for a given blood flow over time, a significant amount of thrombus may be
developing within the oxygenator. Increasing the level of anticoagulation may sta-
bilize these occurrences, though the oxygenator may ultimately need replacement.
Membrane leakage, with blood passing across the membrane and into the hollow
fibers, as seen in the gas exhaust line, is another cause of oxygenator failure. If
blood is found in the gas exhaust line, the oxygenator must be replaced. The gas
exhaust outlet must remain vented to the atmosphere; otherwise, significant increase
in oxygenator gas pressure could place the patient at risk for air embolism across the
compromised membrane. Oxygenator replacement requires that the patient tempo-
rarily be removed from ECLS. Experienced teams can change an oxygenator in less
than 30 seconds. However, this takes careful coordination and planning. Simulation
training is recommended for such scenarios. Emergency circuit clamping measures
should be followed and the pump turned off or to a very low flow rate. Vasopressors
and inotropic agents should be immediately available to manage any hemodynamic
compromise. The circuit is clamped near the inflow and outflow ports of the oxy-
genator, which is then removed and replaced; in the case of integrated circuits, the
full disposable pump and oxygenator will need to be exchanged. Gas and heater
lines are connected to the new oxygenator and ECLS support is resumed.

Pump Failure

Pump failure is a known but uncommon complication of ELCS [18]. If pump fail-
ure occurs for any reason, emergency circuit clamping measures should be fol-
lowed immediately. During VA ECLS, it is imperative to clamp the circuit to
prevent reversal of blood flow from the patient’s arterial system, under the pressure
of native antegrade blood flow through the ECLS circuit into the venous system
(leading to creation of a large arteriovenous shunt) that can lead to cardiopulmo-
nary collapse. A hand crank, which should be kept alongside ECLS circuits, is used
to generate blood flow while the source of the problem is identified and corrected.
If disruption of the electrical power supply is the cause of pump failure, then bat-
tery power or a backup power supply should be obtained. If the ECLS circuit was
inadvertently powered off, it should be restarted immediately and alarm functional-
ity assessed.
If the circuit has adequate power supply and the pump is rotating, but there is no
flow, there may be inadequate occlusion in the case of a roller pump or inadequate
preload or excess afterload with a centrifugal pump in the circuit. Centrifugal pumps
may decouple from the motor, in which case the pump should be stopped and
removed from its seat on the motor. Pump integrity should be confirmed by thor-
ough visual inspection, assessing for cracks or thromboses. If the pump head is
238 P. Madahar et al.

intact, it may be placed back onto the motor and the pump restarted. In the event
there is still no flow, the motor should be replaced.

Tubing Rupture/Disconnect/Cannula Fracture

Tubing rupture, disconnect, or cannula fracture is an ECLS emergency and can

result in exsanguination or entrainment of air into the circuit and patient. Risk of
exsanguination is greatest if the rupture, disconnect, or fracture occurs post-pump,
where the circuit is under positive pressure. The risk of air entrainment is greatest
when tubing rupture, disconnect, or cannula fracture occurs pre-pump, under nega-
tive pressure conditions. For post-pump rupture or fracture, the circuit should be
immediately clamped, emergency circuit clamping measures followed, and the
defective tubing or cannula isolated and replaced. In cases of tubing disconnect, the
first step should be to reconnect the tubing. For small cracks in the tubing or cannula
that occur pre-pump, it may be possible to continue ECLS support while repairing
the problem, particularly if the patient is supported on venovenous ECLS and no
intracardiac or intrapulmonary shunt exists, although in most cases the circuit is
clamped while the defect is repaired or replaced.

Heat Exchanger

ECLS circuits contain a heat exchanger connected to a water source that enables
temperature regulation of the patient. Before the patient is placed on ECLS, the heat
exchanger should be tested by connecting to the water source to ensure that there is
no water leakage between the heat exchanger and the membrane fiber bundle within
the oxygenator. If a water leak develops while the patient is on ECLS, the oxygen-
ator must be changed immediately, as hemolysis, blood volume changes, or infec-
tion may ensue from water transfer to the patient or blood loss into the water bath.

Inadvertent Decannulation

Inadvertent decannulation is a rare but potentially fatal complication of ECLS [19].

Increased tension on the cannula or sutures may occur during patient movement, or
when bed position is adjusted. Additionally, an agitated, awake patient may pull at
the tubing or cannula.
If a cannula becomes partially dislodged, additional sutures may be used to rein-
force its position. This procedure should be done under sterile conditions to mini-
mize the risk of infection. A partially dislodged cannula should not be reinserted, if
at all possible, into the patient as this could result in vascular injury or infection.
13 Crises During ECLS 239

Complete decannulation or partial decannulation to the point of compromised

blood flow is a surgical emergency. If complete decannulation occurs, the remaining
cannula should be clamped immediately to prevent exsanguination and the patient
should come off ECLS support and emergency circuit clamping measures should be
followed. Direct, firm pressure should be applied to the vascular site and anticoagu-
lation should be withheld.
Education and in-servicing of staff not familiar with ECLS may help reduce the
risk of inadvertent decannulation. Furthermore, the cannula positions, and associ-
ated securing sutures, should be checked on a routine basis to ensure their integrity
(see also Chap. 12). An experienced provider should monitor the patient and circuit
during radiologic or other bedside procedures and should ensure that both the ECLS
cart and bed are returned to a locked position after any movement.

Patient Crises

Bleeding and Transfusions

Bleeding is the most common reported complication of ECLS support and has been
described in nearly 25% of cases [20, 21]. The etiology and severity of ECLS-­
related bleeding varies from minor oozing at surgical sites to fatal intracerebral
hemorrhage; however, the risk of severe hemorrhage appears to have decreased in
recent years. Bleeding and the associated high transfusion requirements necessary
to maintain an acceptable level of hemoglobin were nearly ubiquitous in early
ECLS experience, with reports of intracerebral hemorrhage in greater than 50% of
patients and transfusion requirements of greater than 6 units of packed red blood
cells (pRBCs) per day in select series [22, 23].
Modern ECLS circuitry and evolving approaches to patient management have
reduced both the risk of hemorrhage and the need for transfusion of pRBCs in
patients receiving extracorporeal support. Two modern randomized controlled trials
of ECLS in severe acute respiratory distress syndrome (ARDS), known as the
Efficacy and economic assessment of conventional ventilator support versus extra-
corporeal membrane oxygenation for severe adult respiratory failure (CESAR) trial
and the ECMO to Rescue Lung Injury in Severe ARDS (EOLIA) trial, reported on
serious bleeding events. In the CESAR trial, only one death related to hemorrhage
was reported. In the EOLIA trial, three subjects (2% of those in the intervention
arm) were reported to have massive hemorrhage [6, 24]. However, in the EOLIA
trial, there was a significant difference in bleeding events leading to transfusion in
those who received ECLS (48%) compared to those who did not receive ECLS
(28%) [6].
Bleeding during ECLS may be difficult to manage due to both the potential coag-
ulopathy induced by the ECLS circuit and the need to provide systemic anticoagula-
tion to prevent thrombus formation within the circuit [24]. Modern circuits are more
240 P. Madahar et al.

biocompatible and require a lesser degree of anticoagulation than older ECLS tech-
nology, which may also reduce the risk and severity of bleeding.
Contemporary trends in critical care have reduced the hemoglobin threshold at
which transfusions are thought to be necessary, including in patients supported with
ECLS [6, 25–30]. Therefore, while transfusions remain common during ECLS sup-
port, and severe or hemodynamically unstable bleeding requires transfusion of
pRBCs, transfusion for the sake of maintaining a normal hemoglobin level may not
be necessary [31, 32].
The optimal hemoglobin for patients receiving ECLS is not known and practice
varies widely by center. Abundant evidence has demonstrated that liberal transfu-
sion strategies targeting higher hemoglobin levels result in increased morbidity
and mortality in critically ill medical and surgical patients when compared with
more conservative approaches [27–29, 33–35]. This includes higher rates of
transfusion-­related acute lung injury, worsened ARDS, volume overload, infec-
tions, sepsis, poor wound healing, and intensive care unit, hospital, and 1-year
mortality [33, 34, 36–38]. Transfusion-associated morbidity extends to patients
receiving even small amounts of blood [33, 39]. Adverse events associated with
transfusions increase in a dose-dependent manner and provide evidence suggest-
ing that even as little as 1–2 units of blood may increase morbidity and mortality
considerably [40].
Currently, there is no randomized trial comparing liberal and conservative trans-
fusion protocols in patients receiving ECLS. Nevertheless, several studies have
shown similar, if not improved, outcomes in patients receiving ECLS who were
managed with a conservative transfusion approach compared to utilizing a liberal
transfusion threshold [26, 34, 41]. Given the potentially harmful consequences
associated with the transfusion of pRBCs and the questionable ability of transfused
blood to actually improve oxygen delivery to tissues, it may be beneficial to lower
the transfusion threshold in patients receiving ECLS and tolerate a moderate degree
of anemia [42–44]. Similar to our practice with other critically ill patients, our pro-
tocol is to transfuse non-bleeding patients without active cardiac disease when the
hemoglobin is below 7.0 g/dL, although this threshold may be even lower in patients
awaiting transplantation, where blood product transfusions may affect donor–recip-
ient cross-matching.
Thrombocytopenia is also a common complication of ECLS, with platelet activa-
tion and consumption potentially resulting in a notable amount of thrombocytope-
nia over the first 4–5 days of ECLS support [45–48]. As with hemoglobin, the
optimal platelet level in patients receiving ECLS is not known. One approach is to
maintain a count greater than 20,000 platelets per milliliter in non-bleeding patients
and greater than 50,000 platelets per milliliter in bleeding patients or those undergo-
ing an invasive procedure. One exception to this approach may be in patients with
severe pulmonary hypertension, in whom the risk of platelet sequestration in the
pulmonary vasculature, which can precipitate pulmonary hypertensive crises, may
outweigh the benefit of transfusion in the absence of severe bleeding.
It may be possible to continue anticoagulation in the setting of mild to moderate
bleeding. With modern extracorporeal circuit components that are more
13 Crises During ECLS 241

biocompatible and less prone to thrombosis, low levels of systemic anticoagulation

may generally be administered without a significant increase in the rate of circuit
thrombosis, therefore minimizing the likelihood of exacerbating any bleeding. In
the setting of life-threatening hemorrhage, anticoagulation may be withheld for
periods of time without clinically significant thrombosis, depending on the clinical
situation, although anticoagulation should be resumed as soon as possible [8, 49,
50]. Fresh frozen plasma can be administered as needed. Fibrinogen levels may be
maintained above 100 mg/dL with the transfusion of cryoprecipitate in the setting
of bleeding.
If bleeding from a surgical or cannulation site occurs, reinforcement of the site
with additional sutures, application of a pressure dressing, or the use of electrocau-
tery may reduce bleeding. In cases of pulmonary hemorrhage, bronchoscopy may
be used to localize the source of bleeding to potentially guide definitive
interventions.
Uncontrollable, life-threatening bleeding has been treated successfully with anti-
fibrinolytics such as aminocaproic acid (AMICAR) and tranexamic acid [49, 51–
56]. Recombinant factor VII, which complexes with tissue factor to increase the
production of thrombin and enhance clot formation, has also been used to manage
hemorrhage during ECLS [55, 57–60]. However, with any of these agents, the risk
of oxygenator and circuit thromboses, including fatal clots, may be increased [61,
62]. These agents should only be considered in life-threatening situations when
other measures have failed. Prior to use of any of these therapies, consultation with
a hematologist is recommended. (Table 13.2 highlights diagnosis and management
of patient-related crises.)

Hemolysis

Improvements in circuit technology, including the transition from roller pumps to

centrifugal pumps and the use of gas exchangers with decreased pressure differ-
ences across the membrane, have reduced the incidence of hemolysis in ECLS [63–
68]. Despite these advances, hemolysis can still occur from extremes in negative
pressure in the venous drainage cannula and tubing, typically when pressure exceeds
−100 mmHg, although it may also occur at less negative pressures. Hemolysis can
occur at sites of turbulent, non-laminar blood flow in the circuit or patient, such as
at points of kinked or intentionally split tubing, or at vascular access points. In situ
thromboses can also cause non-laminar flow and result in hemolysis, as can over-
heated water baths [69–71]. Hemolysis may also be due to medications, certain
infections, hematologic abnormalities, and other etiologies unrelated to the ECLS
circuit.
The components of any ECLS circuit depend on institutional preference; how-
ever, a simplified circuit (e.g., without a bridge) may reduce the rate of hemolysis.
Likewise, externally applied, noninvasive monitors can reduce the necessity for
access points, further minimizing the amount of turbulent flow. The use of a
242 P. Madahar et al.

Table 13.2 Patient-related emergencies and management

Patient crises Diagnostics Management
Bleeding
Mild to Monitor hemoglobin Local bleeding control
moderate Check platelet, PT, aPTT, and Maintain target hemoglobin (typically ≥7.0 g/
fibrinogen dL)
Maintain platelet count ≥50,000/mL
Transfuse cryoprecipitate if fibrinogen <100 mg/
dL
Consider FFP for elevated PT
Administer DDAVP if platelet dysfunction
suspected
Consider dose reduction or discontinuation of
systemic anticoagulation
Severe Monitor hemoglobin Local bleeding control
Check platelet, PT, aPTT, and Maintain target hemoglobin (typically ≥7.0 g/
fibrinogen dL)
Maintain platelet count ≥50,000/mL
Discontinue anticoagulation
Transfuse FFP for elevated PT
Transfuse cryoprecipitate if fibrinogen <100 mg/
dL
Administer DDAVP if concern for platelet
dysfunction
In catastrophic bleeding, consider protamine if
anticoagulated with heparin, consider factor
VIIa, aminocaproic acid, or tranexamic acid, if
necessary
Hemolysis Monitor hemoglobin, LDH, Avoid venous limb pressure more negative than
bilirubin, and free −100 mmHg
hemoglobin Minimize circuit access points
Check haptoglobin and Correct tubing kinks
preform urinalysis Replace heat exchanger if malfunctioning
Assess for tubing kinks Correct patient or treatment-related causes of
Assess for circuit thromboses hemolysis
Check heat exchanger
temperature
Assess for cause of hemolysis
in the patient
Refractory Assess for recirculation; If recirculation present, consider reducing pump
hypoxemia monitor pre-oxygenator SO2 RPMs or repositioning cannulae
Assess gas delivery line for Reconnect gas delivery line and/or fix cause of
disconnection or obstruction gas delivery line obstruction
Assess oxygenator function; If oxygenator failure, replace oxygenator
check pre- and post-­ Increase blood flow rate, augment maximal flow
oxygenator PO2 and pressures with additional venous drainage cannula
Assess circuit blood flow rate If femoral venoarterial configuration and dual
Assess for dual circulation by circulation present, consider addition of venous
presence of differential reinfusion cannula to upper body
oxygenation Utilize supportive measures with deep sedation,
neuromuscular blockade, targeted temperature
management, and/or proning, as clinically
indicated
13 Crises During ECLS 243

Table 13.2 (continued)

Patient crises Diagnostics Management
Shock
Low-output Check TTE Consider the use of inotropes
state CVP If venovenous ECLS, consider conversion to a
Central or mixed venous venoarterial or veno–venoarterial configuration
oxygen saturation If venoarterial ECLS, augment circuit blood
Lactate flow, and/or consider other mechanical cardiac
Other diagnostics as per support
routine care
High-output Check TTE Augment ECLS blood flow rate, as able (may
state CVP require additional venous drainage)
Central or mixed venous
oxygen saturation
Lactate
Other diagnostics as per
routine care
Limb ischemia
Non-DPC Physical exam Treat causes of shock and reduce vasopressors as
related NIRS oxygen saturation able
Doppler ultrasound External warming of limbs
Place DPC if ongoing signs of limb ischemia
DPC related Physical exam DPC flow should be ~8–10% of total arterial
NIRS oxygen saturation reinfusion flow
Doppler ultrasound For bleeding due to disconnected DPC,
Monitor for bleeding reconnect DPC immediately and transfuse for
clinically significant blood loss
For thrombosed DPC, remove thrombus via
sterile aspiration or remove DPC entirely
Consider therapeutic anticoagulation if
propagation of clot despite DPC removal
aPTT activated partial thromboplastin time, CVP central venous pressure, DDAVP desmopressin,
DPC distal perfusion cannula, ECLS extracorporeal life support, FFP fresh frozen plasma, LDH
lactate dehydrogenase, NIRS near-infrared spectroscopy, PO2 partial pressure of oxygen, PaO2
partial pressure of oxygen in arterial blood, PT prothrombin time, RPMs revolutions per minute,
TTE transthoracic echocardiogram

simplified circuit consisting of only the key components, including centrifugal

pump, membrane oxygenator, pre- and post-oxygenator pressure monitors, and heat
exchanger, is preferred. As compared with older or more complex ECLS circuits,
significant hemolysis does not occur as commonly with the use of a simplified cir-
cuit design.
If severe hemolysis is identified that cannot be attributed to a correctable circuit
problem (such as tubing kink or excessively high drainage pressures) or a systemic
process in the patient and cannot be corrected by reducing the temperature of the
water bath, then the circuit may need to be replaced. With any occurrence, a source
of non-circuit-related hemolysis should be excluded clinically [71, 72].
244 P. Madahar et al.

Refractory Hypoxemia

Refractory hypoxemia is considered an indication for the use of ECLS in severe

ARDS; however, refractory hypoxemia may persist even after the initiation of ECLS
support [6, 21, 24, 73]. Factors that determine the relative contribution of the circuit
to oxygenation include the severity of the gas exchange abnormality in the native
lungs, the contribution of the mechanical ventilator to gas exchange, the patient’s
native cardiac output, the degree of recirculation, the rate of extracorporeal blood
flow, the membrane’s gas transfer efficiency, oxygen extraction in the patient’s tis-
sues, and the metabolic demands of the patient.
Accurate estimation of blood flow and oxygen requirements is essential prior to
ECLS initiation in order to choose properly sized cannulae to meet the patient’s
metabolic demand. While this calculation can be estimated by body surface area
(BSA), patient-specific physiologic characteristics, such as pregnancy, vasodilatory
shock, or other high cardiac output states, may necessitate blood flows that consid-
erably exceed those calculated by BSA [74–76]. In venovenous ECLS, the addition
of a second venous drainage cannula may be used to augment blood flow if the ini-
tial configuration cannot achieve adequate blood flow rates.
In venoarterial support, most commonly via the femoral vein and artery, rein-
fused, oxygenated blood flows retrograde in the aorta, where it mixes with blood
ejected from the left ventricle at a mixing point. Depending on the relative contri-
bution of circuit blood flow and native cardiac output, regional perfusion and flow
differences can occur, especially when cardiac dysfunction is not severe [77]. In
these circumstances, there is a dual circulation (native and extracorporeal),
wherein the extracorporeal circulation provides oxygenated blood, while the
native circulation may provide deoxygenated blood, especially in the arch of the
aorta to the carotid and coronary arterial beds, resulting in differential oxygen-
ation. For this reason, systemic oxygen delivery in venoarterial ECLS is best
assessed by measurement of the PaO2 from a radial artery, particularly the right
radial artery, as an estimate of the oxygen content of blood perfusing the brain and
heart. An increase in the circuit blood flow rate may improve delivery of oxygen-
ated blood to the aortic arch. Alternatively, the addition of a second reinfusion
cannula into an internal jugular vein, thereby creating a veno-arterial configura-
tion, can improve oxygen delivery to the systemic circulation by utilizing the
patient’s native cardiac function to deliver reinfused, oxygenated blood into the
ascending aorta. A venoarterial ECLS configuration via the axillary, innominate,
or subclavian artery may or may not ameliorate this issue, whereas direct reinfu-
sion into the aorta would preclude the need for veno-arterial conversion, though
these approaches are more technically difficult and not suitable for emergent situ-
ations [78–81].
Hypoxemia can also result from circuit malfunction, disconnection or obstruc-
tion of the gas delivery line, recirculation, or a failing oxygenator. Recirculation, in
13 Crises During ECLS 245

which reinfused, oxygenated blood is returned to the circuit without having passed
through the patient, is common in the traditional two-site venovenous ECLS con-
figuration, especially when the drainage and reinfusion ports are in close proximity
to one another or the pump flows are too high and the venous cannula drains a dis-
proportionate amount of reinfused blood [82, 83]. Recirculation should be sus-
pected when a high pre-oxygenator SO2 is observed. However, clinical judgment
must be used to distinguish whether the high pre-oxygenator SO2 is due to recircula-
tion or whether it is a consequence of either pulmonary recovery, poor tissue oxygen
extraction in the setting of sepsis, or decreased cardiac output without correlating
change in blood flow rate in venovenous ECLS or venoarterial ECLS with central
configurations (right atrium to aorta or right atrium to left atrium). Repositioning of
the cannulae or use of a bicaval dual-lumen cannula may reduce recirculation [84–
86]. Oxygenator function may be assessed by monitoring the pre- and post-­
oxygenator blood gases, the pressure gradient across the oxygenator, the color
differential of the circulating blood in the drainage and re-infusion limbs by visual
inspection, and the SO2 of the reinfusion limb. Provided that the blood flow is less
than the rated flow of the oxygenator, the post-oxygenator oxygen saturation should
be greater than 99% in a functioning oxygenator.

Shock

Management of shock states during venovenous ECLS is the same as in patients

without ECLS support because venovenous configurations do not provide direct
circulatory support. If cardiogenic shock develops in the setting of venovenous
ECLS, a trial of inotropes and/or vasopressors should be attempted first. If shock
persists, conversion to venoarterial or veno–venoarterial ECLS should be consid-
ered, although differential oxygenation with femoral arterial reinfusion, as men-
tioned above, must be kept in mind [87, 88].
If a high-output shock state develops during venovenous or antegrade venoarte-
rial ECLS, the extracorporeal blood flow rate will account for a lower percentage
of the patient’s cardiac output. Consequently, the proportion of blood being oxy-
genated by the circuit will be lower, which will result in worsened hypoxemia.
While the use of ECLS in high-output shock has not been well studied, reports
have demonstrated the successful use of venoarterial ECLS in adult patients with
septic shock in order to augment cardiopulmonary function in the setting of vaso-
dilation, reduced cardiac output, and high metabolic demand. Judicious patient
selection is paramount to maximize the benefit of ECLS in the setting of shock
[89–93].
246 P. Madahar et al.

Limb Ischemia

Femoral arterial cannulation, in combination with the hemodynamic instability of

critically ill patients supported with ECLS, places the limb at risk for thromboem-
bolic complications. The reported incidence of limb complications associated with
VA ECLS is variable, but is generally recognized as being greater than 10% [94].
Limb ischemia after peripheral cannulation can be attributed to mechanical obstruc-
tion of arterial flow by arterial cannulas, venous stasis secondary to compression by
large venous cannulas, and/or systemic vasoconstriction resulting from shock or
pharmacologic vasoconstriction [95]. The risk is greater in patients with larger can-
nulas, smaller vessels, and peripheral vascular disease. Distal perfusion can be mon-
itored by physical exam, Doppler ultrasound, and/or near-infrared spectroscopy
(NIRS). NIRS oxygen saturations less than 40%, lack of distal pulse on exam, or
lack of Doppler signal on ultrasound warrants investigation, and possible interven-
tion, in order to prevent limb ischemia. Decreasing vasopressor infusion, treating
circulatory shock, and external limb warming may improve blood flow to extremi-
ties [96]. Nonetheless, some patients may require augmentation of blood flow
through the use of a distal perfusion cannula (DPC) or venous introducer sheaths in
order to ensure adequate limb perfusion.
There are several mechanisms commonly utilized to augment limb perfusion.
Anterograde cannulation of the superficial femoral artery with a 5- to 8-Fr intro-
ducer sheath with subsequent connection of the sheath to the side port of the main
arterial cannula is an easy and inexpensive way to perfuse the distal extremity.
Alternatively, 10- to 12-Fr pediatric cannulae can be used for distal artery perfusion
[96]. Data on the routine placement of DPC during initial VA cannulation is mixed
and superficial femoral arterial cannulation is not without risk. However, salvage
DPC placement resulting in resolution of limb ischemia has been described and
placement of DPC when suspicion for limb ischemia exists is generally a well-­
accepted practice [94].
Once a distal perfusion strategy is employed, monitoring is continued to ensure
early detection of potential complications. Most complications are either hemor-
rhagic or thrombotic in nature. Hemorrhage can occur from the DPC insertion site
itself or as a result of disconnection of the cannula from the arterial side port, with
potential for significant blood loss, especially if the site of bleeding or disconnect
is hidden by dressings or bed sheets. Prompt recognition and reconnection of tub-
ing if disconnected are imperative in order to avoid exsanguination. Bleeding
around the DPC may require additional suture placement or utilization of hemo-
static agents. If the bleeding cannot be adequately controlled by these means,
removal of the DPC may be necessary, and possible reconfiguration if removal of
the DPC threatens limb ischemia, in order to remove the arterial cannula that is
restricting blood flow.
By nature of their small size and lower flow state, distal perfusion catheters and
venous sheaths utilized in this manner are at higher risk of thrombosis. Ideally, flow
through the DPC should be monitored via a flow probe and should exceed 8%–10%
13 Crises During ECLS 247

of the total arterial reinfusion flow. If unable to monitor flow, pedal pulses should be
assessed frequently, along with visual inspection of the DPC tubing to look for
separation of blood as a sign of reduced flow. Thrombosis of the cannula results in
leg ischemia, and therefore, continuous NIRS oxygen saturation monitoring, as well
as routine Doppler signal monitoring, is important. Management of thrombosed
DPC necessitates immediate removal of the thrombus via sterile aspiration or
removal of the entire DPC. If the clot propagates and is not adequately removed
with catheter removal, therapeutic anticoagulation may be needed. Thrombosis of
the SFA necessitates reconfiguration if a DPC cannot be replaced.

Conclusion

Successful application of ECLS requires expeditious identification and intervention

during both patient and circuit-related crises. Through clinical instruction and simu-
lation, ECLS centers should train providers to recognize and respond to both com-
mon and uncommon complications in order to maximize patient safety and improve
outcomes.

Conflict of Interests Drs. Abrams and Agerstrand are members of the steering committee of the
Extracorporeal Life Support Organization (ELSO).
Dr. Brodie receives research support from ALung Technologies. He has been on the medical
advisory boards for Abiomed, Xenios, Medtronic, Cellenkos, and Hemovent. He is the president-­
elect of the Extracorporeal Life Support Organization (ELSO).
All other authors have no conflicts of interest to disclose.

References

1. Chan KM, Wan WTP, Ling L, So JMC, Wong CHL, Tam SBS. Management of circuit air in
extracorporeal membrane oxygenation: a single center experience. ASAIO J. 2021. https://doi.
org/10.1097/mat.0000000000001494.
2. Yan S, Lou S, Zhao Y, Liu G, Ji B. Air in extracorporeal membrane oxygenation: can never be
overemphasized. Perfusion. 2021;36:97–9.
3. Kopp R, Mottaghy K, Kirschfink M. Mechanism of complement activation during extracor-
poreal blood-biomaterial interaction: effects of heparin coated and uncoated surfaces. ASAIO
J. 2002;48:598–605.
4. Deptula J, Glogowski K, Merrigan K, Hanson K, Felix D, Hammel J, Duncan K. Evaluation
of biocompatible cardiopulmonary bypass circuit use during pediatric open heart surgery. J
Extra-corporeal Technol. 2006;38:22–6.
5. Bembea MM, Annich G, Rycus P, Oldenburg G, Berkowitz I, Pronovost P. Variability in anti-
coagulation management of patients on extracorporeal membrane oxygenation. Pediatr Crit
Care Me. 2013;14:e77–84.
6. Combes A, Hajage D, Capellier G, et al. Extracorporeal membrane oxygenation for severe
acute respiratory distress syndrome. N Engl J Med. 2018;378:1965–75.
248 P. Madahar et al.

7. Extracorporeal Life Support Organization ELSO Anticoagulation Guideline. 2014. https://

www.elso.org/Portals/0/Files/elsoanticoagulationguideline8-­2014-­table-­contents.pdf.
8. Extracorporeal Life Support Organization (ELSO) General Guidelines for all ECLS Cases.
2017. https://www.elso.org/Portals/0/ELSO%20Guidelines%20General%20All%20
ECLS%20Version%201_4.pdf.
9. Pollak U. Heparin-induced thrombocytopenia complicating extracorporeal membrane oxy-
genation support: review of the literature and alternative anticoagulants. J Thromb Haemost.
2019;17:1608–22.
10. Choi JH, Luc JGY, Weber MP, et al. Heparin-induced thrombocytopenia during extracorporeal
life support: incidence, management and outcomes. Ann Cardiothorac Surg. 2019;8:19–31.
11. Zaaqoq AM, Brammer RC, Chan CM, Shorr AF. Heparin-induced thrombocytopenia in extra-­
corporeal membrane oxygenation: epidemiology, outcomes, and diagnostic challenges. J
Thromb Thrombolysis. 2021:1–7.
12. Geli J, Capoccia M, Maybauer DM, Maybauer MO. Argatroban anticoagulation for
adult extracorporeal membrane oxygenation: a systematic review. J Intensive Care Med.
2021:088506662199373.
13. Sanfilippo F, Asmussen S, Maybauer DM, Santonocito C, Fraser JF, Erdoes G, Maybauer
MO. Bivalirudin for alternative anticoagulation in extracorporeal membrane oxygenation: a
systematic review. J Intensive Care Med. 2016;32:312–9.
14. Coughlin MA, Bartlett RH. Anticoagulation for extracorporeal life support. ASAIO
J. 2015;61:652–5.
15. Linkins L-A, Dans AL, Moores LK, Bona R, Davidson BL, Schulman S, Crowther M,
Physicians AC of C. Treatment and prevention of heparin-induced thrombocytopenia anti-
thrombotic therapy and prevention of thrombosis, 9th ed.: American College of Chest
Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141:e495S–530S.
16. Warkentin TE, Greinacher A, Gruel Y, Aster RH, Chong BH, haemostasis scientific and stan-
dardization committee of the international society on thrombosis and haemostasis. Laboratory
testing for heparin-induced thrombocytopenia: a conceptual framework and implications for
diagnosis. J Thromb Haemost. 2011;9:2498–500.
17. Zakhary B, Vercaemst L, Mason P, Antonini MV, Lorusso R, Brodie D. How I approach mem-
brane lung dysfunction in patients receiving ECMO. Crit Care. 2020;24:671.
18. Thiagarajan RR, Barbaro RP, Rycus PT, Mcmullan DM, Conrad SA, Fortenberry JD, Paden
ML, centers E member. Extracorporeal life support organization registry international report
2016. ASAIO J. 2017;63:60–7.
19. Kim DH, Cho WH, Son J, Lee SK, Yeo HJ. Catastrophic mechanical complications
of extracorporeal membrane oxygenation. ASAIO J. 2021. https://doi.org/10.1097/
mat.0000000000001354.
20. Organization ELS. ECLS registry report, international summary; 2019.
21. Brodie D, Slutsky AS, Combes A. Extracorporeal life support for adults with respiratory fail-
ure and related indications. JAMA. 2019;322:557–68.
22. Sell LL, Cullen ML, Whittlesey GC, Yedlin ST, Philippart AI, Bedard MP, Klein
MD. Hemorrhagic complications during extracorporeal membrane oxygenation: Prevention
and treatment. J Pediatr Surg. 1986;21:1087–91.
23. Gattinoni L, Pesenti A, Mascheroni D, et al. Low-frequency positive-pressure ventilation with
extracorporeal CO2 removal in severe acute respiratory failure. JAMA. 1986;256:881–6.
24. Peek GJ, Mugford M, Tiruvoipati R, et al. Efficacy and economic assessment of conventional
ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory
failure (CESAR): a multicentre randomised controlled trial. Lancet. 2009;374:1351–63.
25. Zhang W, Zheng Y, Yu K, Gu J. Liberal transfusion versus restrictive transfusion and outcomes
in critically ill adults: a meta-analysis. Transfus Med Hemoth. 2021;48:60–8.
26. Agerstrand CL, Burkart KM, Abrams DC, Bacchetta MD, Brodie D. Blood conservation in
extracorporeal membrane oxygenation for acute respiratory distress syndrome. Ann Thorac
Surg. 2015;99:590–5.
13 Crises During ECLS 249

27. Walsh TS, Boyd JA, Watson D, et al. Restrictive versus liberal transfusion strategies for older
mechanically ventilated critically ill patients. Crit Care Med. 2013;41:2354–63.
28. Vincent JL, Baron J-F, Reinhart K, Gattinoni L, Thijs L, Webb A, Meier-Hellmann A, Nollet
G, Peres-Bota D, Investigators for the A. Anemia and blood transfusion in critically ill patients.
JAMA. 2002;288:1499–507.
29. Hébert PC, Fergusson DA. Red blood cell transfusions in critically ill patients.
JAMA. 2002;288:1525–6.
30. Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, Tweeddale M,
Schweitzer I, Yetisir E. A multicenter, randomized, controlled clinical trial of transfusion
requirements in critical care. N Engl J Med. 1999;340:409–17.
31. Martucci G, Grasselli G, Tanaka K, Tuzzolino F, Panarello G, Schmidt M, Bellani G,
Arcadipane A. Hemoglobin trigger and approach to red blood cell transfusions during veno-­
venous extracorporeal membrane oxygenation: the international TRAIN-ECMO survey.
Perfusion. 2019;34:39–48.
32. Butch SH, Knafl P, Oberman HA, Bartlett RH. Blood utilization in adult patients undergoing
extracorporeal membrane oxygenated therapy. Transfusion. 1996;36:61–3.
33. Qin CX, Yesantharao LV, Merkel KR, Goswami DK, Garcia AV, Whitman GJR, Frank SM,
Bembea MM. Blood utilization and clinical outcomes in extracorporeal membrane oxygen-
ation patients. Anesth Analg. 2020;131:901–8.
34. Martucci G, Panarello G, Occhipinti G, et al. Anticoagulation and transfusions management in
veno-venous extracorporeal membrane oxygenation for acute respiratory distress syndrome:
assessment of factors associated with transfusion requirements and mortality. J Intensive Care
Med. 2017;34:630–9.
35. Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Abraham E, MacIntyre NR, Shabot
MM, Duh M-S, Shapiro MJ. The CRIT Study; anemia and blood transfusion in the critically
ill—current clinical practice in the United States. Crit Care Med. 2004;32:39–52.
36. Marik PE, Corwin HL. Acute lung injury following blood transfusion; Expanding the defini-
tion. Crit Care Med. 2008;36:3080–4.
37. Koch CG, Li L, Sessler DI, Figueroa P, Hoeltge GA, Mihaljevic T, Blackstone EH. Duration
of red-cell storage and complications after cardiac surgery. N Engl J Med. 2008;358:1229–39.
38. Shorr AF, Duh M-S, Kelly KM, Kollef MH, Group CS. Red blood cell transfusion and
ventilator-­associated pneumonia; a potential link? Crit Care Med. 2004;32:666–74.
39. Ferraris VA, Davenport DL, Saha SP, Bernard A, Austin PC, Zwischenberger JB. Intraoperative
transfusion of small amounts of blood heralds worse postoperative outcome in patients having
noncardiac thoracic operations. Ann Thorac Surg. 2011;91:1674–80.
40. Ferraris VA, Davenport DL, Saha SP, Austin PC, Zwischenberger JB. Surgical outcomes
and transfusion of minimal amounts of blood in the operating room. Arch Surg-Chicago.
2012;147:49–55.
41. Abbasciano RG, Yusuff H, Vlaar A, Lai F, Murphy GJ. Blood transfusion threshold in patients
receiving Extra Corporeal Membrane Oxygenation (ECMO) support for cardiac and respira-
tory failure – a systematic review and meta-analysis. J Cardiothor Vasc An. 2020;35:1192–202.
42. Rossi’s Principles of Transfusion Medicine, 4th ed.
43. Roback JD, Neuman RB, Quyyumi A, Sutliff R. Insufficient nitric oxide bioavailabil-
ity: a hypothesis to explain adverse effects of red blood cell transfusion. Transfusion.
2011;51:859–66.
44. Carson JL, Grossman BJ, Kleinman S, et al. Red blood cell transfusion: a clinical practice
guideline from the AABB*. Ann Intern Med. 2012;157:49.
45. Anderson HL, Cilley RE, Zwischenberger JB, Bartlett RH. Thrombocytopenia in neonates
after extracorporeal membrane oxygenation. ASAIO J. 1986;32:534–7.
46. Robinson TM, Kickler TS, Walker LK, Ness P, Bell W. Effect of extracorporeal membrane
oxygenation on platelets in newborns. Crit Care Med. 1993;21:1029–34.
47. Stallion A, Cofer BR, Rafferty JA, Ziegler MM, Ryckman FC. The significant relationship
between platelet count and haemorrhagic complications on ECMO. Perfusion. 1994;9:265–9.
250 P. Madahar et al.

48. Jiritano F, Serraino GF, Ten Cate H, Fina D, Matteucci M, Mastroroberto P, Lorusso R. Platelets
and extra-corporeal membrane oxygenation in adult patients: a systematic review and meta-­
analysis. Intensive Care Med. 2020;46:1154–69.
49. Lamb K, Cowan S, Evans N, Pitcher H, Moritz T, Lazar M, Hirose H, Cavarocchi N. Successful
management of bleeding complications in patients supported with extracorporeal membrane
oxygenation with primary respiratory failure. Perfusion. 2013;28:125–31.
50. von Stumm M, Subbotina I, Biermann D, Gottschalk U, Mueller G, Kozlik-Feldmann R,
Reichenspurner H, Riso A, Sachweh JS. Impact of delayed systemic heparinization on post-
operative bleeding and thromboembolism during post-cardiotomy extracorporeal membrane
oxygenation in neonates. Perfusion. 2020;35:626–32.
51. Wilson JM, Bower LK, Fackler JC, Beals DA, Bergus BO, Kevy SV. Aminocaproic acid
decreases the incidence of intracranial hemorrhage and other hemorrhagic complications of
ECMO. J Pediatr Surg. 1993;28:536–41.
52. van der Staak FHJ, de Haan AFJ, Geven WB, Festen C. Surgical repair of congenital diaphrag-
matic hernia during extracorporeal membrane oxygenation: hemorrhagic complications and
the effect of tranexamic acid. J Pediatr Surg. 1997;32:594–9.
53. Downard CD, Betit P, Chang RW, Garza JJ, Arnold JH, Wilson JM. Impact of Amicar on hem-
orrhagic complications of ECMO: a ten-year review. J Pediatr Surg. 2003;38:1212–6.
54. Buckley LF, Reardon DP, Camp PC, Weinhouse GL, Silver DA, Couper GS, Connors
JM. Aminocaproic acid for the management of bleeding in patients on extracorporeal mem-
brane oxygenation: four adult case reports and a review of the literature. Hear Lung J Acute
Critical Care. 2016;45:232–6.
55. Lotz C, Streiber N, Roewer N, Lepper PM, Muellenbach RM, Kredel M. Therapeutic inter-
ventions and risk factors of bleeding during extracorporeal membrane oxygenation. ASAIO
J. 2017;63:624–30.
56. Coleman M, Davis J, Maher KO, Deshpande SR. Clinical and hematological outcomes of ami-
nocaproic acid use during pediatric cardiac ECMO. J Extra-corporeal Technol. 2021;53:40–5.
57. Dominguez TE, Mitchell M, Friess SH, Huh JW, Manno CS, Ravishankar C, Gaynor JW,
Tabbutt S. Use of recombinant factor VIIa for refractory hemorrhage during extracorporeal
membrane oxygenation&ast. Pediatr Crit Care Me. 2005;6:348–51.
58. Wittenstein B, Ng C, Ravn H, Goldman A. Recombinant factor VII for severe bleeding during
extracorporeal membrane oxygenation following open heart surgery. Pediatr Crit Care Me.
2005;6:473–6.
59. Repessé X, Au SM, Bréchot N, Trouillet J-L, Leprince P, Chastre J, Combes A, Luyt
C-E. Recombinant factor VIIa for uncontrollable bleeding in patients with extracorporeal
membrane oxygenation: report on 15 cases and literature review. Crit Care. 2013;17:R55.
60. Anselmi A, Guinet P, Ruggieri VG, et al. Safety of recombinant factor VIIa in patients under
extracorporeal membrane oxygenation. Eur J Cardio Thorac. 2016;49:78–84.
61. Bui JD, Despotis GD, Trulock EP, Patterson GA, Goodnough LT. Fatal thrombosis after
administration of activated prothrombin complex concentrates in a patient supported by extra-
corporeal membrane oxygenation who had received activated recombinant factor VII. J Thorac
Cardiovasc Surg. 2002;124:852–4.
62. Syburra T, Lachat M, Genoni M, Wilhelm MJ. Fatal outcome of recombinant factor VIIa
in heart transplantation with extracorporeal membrane oxygenation. Ann Thorac Surg.
2010;89:1643–5.
63. Morgan IS, Codispoti M, Sanger K, Mankad PS. Superiority of centrifugal pump over roller
pump in paediatric cardiac surgery: prospective randomised trial. Eur J Cardio Thorac.
1998;13:526–32.
64. Klein M, Dauben H, Schulte HD, Gam E. Centrifugal pumping during routine open heart
surgery improves clinical outcome. Artif Organs. 1998;22:326–36.
65. Parolari A, Alamanni F, Naliato M, Spirito R, Franzè V, Pompilio G, Agrifoglio M, Biglioli
P. Adult cardiac surgery outcomes: role of the pump type. Eur J Cardio Thorac. 2000;18:575–82.
13 Crises During ECLS 251

66. Khoshbin E, Roberts N, Harvey C, Machin D, Killer H, Peek GJ, Sosnowski AW, Firmin
RK. Poly-methyl pentene oxygenators have improved gas exchange capability and
reduced transfusion requirements in adult extracorporeal membrane oxygenation. ASAIO
J. 2005;51:281–7.
67. Dogal N, Mathis R, Lin J, Qiu F, Kunselman A, Ündar A. Evaluation of three hollow-fiber
membrane oxygenators without integrated arterial filters for neonatal cardiopulmonary bypass.
Perfusion. 2012;27:132–40.
68. Gu YJ, Boonstra PW, Graaff R, Rijnsburger AA, Mungroop H, van Oeveren W. Pressure
drop, shear stress, and activation of leukocytes during cardiopulmonary bypass: a comparison
between hollow fiber and flat sheet membrane oxygenators. Artif Organs. 2000;24:43–8.
69. Lehle K, Philipp A, Zeman F, Lunz D, Lubnow M, Wendel H-P, Göbölös L, Schmid C, Müller
T. Technical-induced hemolysis in patients with respiratory failure supported with veno-­
venous ECMO – prevalence and risk factors. Plos One. 2015;10:e0143527.
70. Dufour N, Radjou A, Thuong M. Hemolysis and plasma free hemoglobin during extracorpo-
real membrane oxygenation support. Asaio J. 2019. Publish Ahead of Print:NA.
71. Appelt H, Philipp A, Mueller T, Foltan M, Lubnow M, Lunz D, Zeman F, Lehle K. Factors asso-
ciated with hemolysis during extracorporeal membrane oxygenation (ECMO)—Comparison
of VA- versus VV ECMO. Plos One. 2020;15:e0227793.
72. Nagler B, Hermann A, Robak O, et al. Incidence and etiology of system exchanges in
patients receiving extracorporeal membrane oxygenation. Asaio J. 2021. Publish Ahead of
Print:776–84.
73. Goligher EC, Tomlinson G, Hajage D, Wijeysundera DN, Fan E, Jüni P, Brodie D, Slutsky
AS, Combes A. Extracorporeal membrane oxygenation for severe acute respiratory distress
syndrome and posterior probability of mortality benefit in a post hoc bayesian analysis of a
randomized clinical trial. JAMA. 2018;320:2251.
74. Zakhary B, Vercaemst L, Mason P, Lorusso R, Brodie D. How I manage drainage insufficiency
on extracorporeal membrane oxygenation. Crit Care. 2020;24:151.
75. Broman LM, Wittberg LP, Westlund CJ, et al. Pressure and flow properties of cannulae for extra-
corporeal membrane oxygenation II: drainage (venous) cannulae. Perfusion. 2019;34:65–73.
76. On bypass, advanced perfusion techniques. 2008. https://doi.org/10.1007/978-­1-­59745-­305-­9.
77. Combes A, Price S, Slutsky AS, Brodie D. Temporary circulatory support for cardiogenic
shock. Lancet. 2020;396:199–212.
78. Javidfar J, Brodie D, Costa J, Miller J, Jurrado J, LaVelle M, Newmark A, Takayama H, Sonett
JR, Bacchetta M. Subclavian artery cannulation for venoarterial extracorporeal membrane
oxygenation. ASAIO J. 2012;58:494–8.
79. Biscotti M, Bacchetta M. The “Sport Model”: extracorporeal membrane oxygenation using the
subclavian artery. Ann Thorac Surg. 2014;98:1487–9.
80. Chicotka S, Rosenzweig EB, Brodie D, Bacchetta M. The “Central Sport Model”. ASAIO
J. 2017;63:e39–44.
81. Ranney DN, Benrashid E, Meza JM, et al. Central cannulation as a viable alternative to periph-
eral cannulation in extracorporeal membrane oxygenation. Semin Thorac Cardiovasc Surg.
2017;29:188–95.
82. Abrams D, Bacchetta M, Brodie D. Recirculation in venovenous extracorporeal membrane
oxygenation. ASAIO J. 2015;61:115–21.
83. Palmér O, Palmér K, Hultman J, Broman M. Cannula design and recirculation during venove-
nous extracorporeal membrane oxygenation. ASAIO J. 2016;62:737–42.
84. Frenckner B, Broman M, Broomé M. Position of draining venous cannula in extracorporeal
membrane oxygenation for respiratory and respiratory/circulatory support in adult patients.
Crit Care. 2018;22:163.
85. Javidfar J, Brodie D, Wang D, Ibrahimiye AN, Yang J, Zwischenberger JB, Sonett J, Bacchetta
M. Use of bicaval dual-lumen catheter for adult venovenous extracorporeal membrane oxy-
genation. Ann Thorac Surg. 2011;91:1763–9.
252 P. Madahar et al.

86. Tipograf Y, Gannon WD, Foley NM, Hozain A, Ukita R, Warhoover M, McMaster W, Nesbitt
JC, Shah AS, Bacchetta M. A dual-lumen bicaval cannula for venovenous extracorporeal
membrane oxygenation. Ann Thorac Surg. 2020;109:1047–53.
87. Brasseur A, Scolletta S, Lorusso R, Taccone FS. Hybrid extracorporeal membrane oxygen-
ation. J Thorac Dis. 2018;10:S707–15.
88. Coco VL, Swol J, Piero MED, Massimi G, Chiarini G, Broman LM, Lorusso R. Dynamic
extracorporeal life support: a novel management modality in temporary cardio-circulatory
assistance. Artif Organs. 2021;45:427–34.
89. Bréchot N, Hajage D, Kimmoun A, et al. Venoarterial extracorporeal membrane oxygenation
to rescue sepsis-induced cardiogenic shock: a retrospective, multicentre, international cohort
study. Lancet. 2020;396:545–52.
90. Ling RR, Ramanathan K, Poon WH, Tan CS, Brechot N, Brodie D, Combes A, MacLaren
G. Venoarterial extracorporeal membrane oxygenation as mechanical circulatory support in
adult septic shock: a systematic review and meta-analysis with individual participant data
meta-regression analysis. Crit Care. 2021;25:246.
91. Myers LC, Lee C, Thompson BT, Cudemus G, Raz Y, Roy N. Outcomes of adult patients with
septic shock undergoing extracorporeal membrane oxygenation therapy. Ann Thorac Surg.
2020;110:871–7.
92. Huang C-T, Tsai Y-J, Tsai P-R, Ko W-J. Extracorporeal membrane oxygenation resuscitation
in adult patients with refractory septic shock. J Thorac Cardiovasc Surg. 2013;146:1041–6.
93. MacLaren G, Pellegrino V, Butt W, Preovolos A, Salamonsen R. Successful use of ECMO in
adults with life-threatening infections. Anaesth Intensive Care. 2004;32:707–10.
94. Elmously A, Bobka T, Khin S, et al. Distal perfusion cannulation and limb complica-
tions in venoarterial extracorporeal membrane oxygenation. J Extra-corporeal Technol.
2018;50:155–60.
95. Juo Y, Skancke M, Sanaiha Y, Mantha A, Jimenez JC, Benharash P. Efficacy of distal perfu-
sion cannulae in preventing limb ischemia during extracorporeal membrane oxygenation: a
systematic review and meta-analysis. Artif Organs. 2017;41:E263–73.
96. Mohite PN, Fatullayev J, Maunz O, et al. Distal limb perfusion in peripheral VA-ECMO. Artif
Organs. 2014;38:940–4.
Chapter 14
Mobilization During ECLS

Gregory A. Schmidt

Until recently, patients managed on ECMO were deeply sedated, often paralyzed,
always mechanically ventilated, and moved as gingerly as possible. Now, techno-
logical innovations in gas-exchanging membranes, pumps, and cannulas, combined
with the recognition that life-sustaining gas exchange can be achieved through the
veno-venous (VV) mode, have spurred enthusiasm to mobilize ECMO patients, as
we do other critically ill patients [1, 2]. These changes have also forced us to recon-
sider the role of ECMO in severe ARDS. Is it merely to rescue the desperately ill
once mechanical ventilation, paralysis, prone positioning, and inhaled vasodilators
have failed? Or, instead, can it serve as a means to avoid noxious interventions and
their sequelae, earning an earlier and more prominent place in the intensivist’s
armamentarium for the treatment of the acute respiratory distress syndrome (ARDS)?

Technological Advances

Patients with severe ARDS can generally be treated with the VV, rather than the
veno-arterial (VA), mode. For example, in the Conventional Ventilation or ECMO
for Severe Acute Respiratory Failure (CESAR) trial, all patients were treated with
the VV mode and none required conversion to VA ECMO [3]. VV ECMO offers
three major advantages. First, returning blood from the gas exchange membrane to
the veins, rather than systemic arteries, requires much lower pressures and corre-
spondingly simpler, safer circuits (see Chap. 2). Second, since only venous access
is required, mobilization-related risks of large vessel arterial cannulation

G. A. Schmidt (*)
Division of Pulmonary Diseases, Critical Care, and Occupational Medicine, Department of
Internal Medicine, University of Iowa, Iowa City, IA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 253

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_14
254 G. A. Schmidt

(hemorrhage, limb ischemia, dislodgement) are avoided. Finally, by its very nature,
VA ECMO requires at least two cannulas, adding to the complexity of movement,
whereas VV support typically needs only a single cannula (see Chap. 4).
ECMO devices, too, have evolved in a direction to make mobilization more
feasible. Modern polymethylpentene (PMP) gas-exchanging membranes present
a much lower resistance to blood flow than earlier technologies (see Chap. 2).
This has two key advantages: (a) higher circuit blood flow is possible, capturing a
greater fraction of the native cardiac output, making it easier to attain acceptable
oxygenation, as described in Chap. 1 and (b) the lower resistance requires less
pressure to drive blood flow, producing safer circuits and smaller, more reliable
pumps. Safer circuits and pumps demand less in the way of monitors and other
accoutrements, reducing the clutter that formerly surrounded ECMO patients. It is
also hard to overstate the importance of dual-lumen cannulas. Only one access
site is needed, usually freeing the groin and making it easier to have patients sit,
stand, and even walk. Even as recently as the CESAR trial, all patients had two
(some three) cannulas, including one in the femoral vein. The presence of only
one critical cannula also reduces the burden on the mobilization team. As described
below, when we mobilize ECMO patients, we dedicate one member of the team to
focus entirely on the cannula(s), since the patient’s life depends on its reliable
function.

The Health Costs of Immobility

Mobilization depends not only on technological innovation, but also on a changing

approach to critical illness in general. Until recently, critically ill patients were typi-
cally sedated and moved only occasionally, emphasizing comfort, physiological
stability, and (presumed) safety. Less than 20 years ago, the idea of interrupting
sedative infusions for mechanically ventilated patients was novel and controversial
[4]. It is now apparent that sedatives contribute to delirium [5], prolonged ventila-
tion [4], posttraumatic stress disorder [6], and poor long-term outcomes [7].
Conversely, daily sedative interruption shortens the duration of mechanical ventila-
tion and speeds discharge from the ICU [4]. Coordinating this spontaneous awaken-
ing trial with a spontaneous breathing trial reduces time on the ventilator and in the
ICU and lowers the risk of dying [8].
There is ample evidence that critical illness begets systemic and respiratory mus-
cle weakness, termed “ICU-acquired weakness (ICU-AW)” and “ventilator-induced
diaphragm dysfunction (VIDD),” respectively [9, 10]. Both accelerated muscle pro-
teolysis and decreased protein synthesis contribute to sepsis-induced myopathy
[11]. Two aspects of the time course of ICU-AW are remarkable: (1) muscle is lost
as early as the first day of critical illness and (2) weakness may persist for years.
These findings suggest that early approaches to treatment and prevention might
yield the most gains.
14 Mobilization During ECLS 255

Like skeletal muscles, the diaphragm suffers atrophy and contractile dysfunction
during critical illness and mechanical ventilation. This occurs acutely, worsens pro-
gressively, and is associated with prolonged ventilation and risk of death. Muscle
protein synthesis is inhibited and multiple pathways of self-destruction are upregu-
lated. Also like in peripheral muscle, active contraction (that is, active breathing)
can effectively modify the degree of catabolism, helping to maintain contractile
function [12]. This has potentially important implications for the role of ECMO. By
freeing the diseased lungs from the burden of gas exchange, extracorporeal support
could allow less ventilation (even extubation; see Chap. 12) and, therefore, less
sedation, delirium, immobility, and VIDD.
Probably the greatest causal factor for ICU-AW is muscular silence, a hypothesis
that may explain the contributory role of sedatives, neuromuscular blocking drugs,
controlled mechanical ventilation, and bed rest. Reducing these influences requires
changing our view of the critically ill from fragile beings in need of (over)protection
to subjects at risk of complication if not liberated from noxious treatments. Of
course, some patients are incredibly delicate and unable to sustain any provocation,
but their number is probably fewer than we commonly believe. As discussed below,
there is accumulating evidence that patients can be safely liberated while on ECMO.

Early Mobilization Promotes Recovery

Immobility leads to myopathy and weakness; thus, mobility seems a reasonable

defense against ICU-AW. Several clinical trials have shown active physical therapy
to be both safe and effective. In one study, 103 ventilated subjects were prospec-
tively enrolled for a twice-daily intervention by a team of physical therapist, respira-
tory therapist, nurse, and critical care technician [13]. Therapy was initiated once
subjects were awake and progressed from sitting on the bed to sitting in chair to
ambulation. There were 1449 activity events, nearly 600 of which included intu-
bated patients. In sessions that involved subjects with an endotracheal tube, nearly
half included ambulation. Importantly, no accidental extubations occurred, and
other adverse events were seen in fewer than 1% of sessions. The conclusions of this
study are limited by the lack of a control group. In a before-and-after study, an early
mobility protocol was associated with significantly lower ICU and hospital length
of stay [14].
The first randomized, controlled, multicentered study of early mobilization
enrolled subjects within 72 hours of intubation, allowing physical therapy to be
provided, on average, within 2 days of intubation [1]. Intervention subjects achieved
important milestones previously thought unattainable during mechanical ventila-
tion: 43% transferred to a chair and 24% were able to walk—some as far as 30
meters. Subjects who received the intervention had less delirium, had more
ventilator-­free days, and were more likely to be functionally independent at hospital
discharge. When combining Awakening, Breathing Coordination, Delirium moni-
toring and management, and Early exercise and mobility in a bundled approach
256 G. A. Schmidt

(ABCDE Bundle), subjects gained more ventilator-free days, experienced less

delirium, and were more likely to be mobilized during the ICU stay than usual care
subjects [15]. Findings like these have led quality organizations, such as the Institute
for Healthcare Improvement, to create action-focused programs to spur cultural
change in ICUs.

Mobilization During ECLS

The published experience regarding physical therapy, mobilization, and ambulation

during ECLS is modest but growing [16–21]. In one of the first cases reported, a
patient with chronic obstructive pulmonary disease (COPD) remained severely
hypercapnic and ventilator-bound despite 2 weeks of intensive therapy. In order to
bridge him to transplantation and encourage pretransplant rehabilitation, VV ECMO
was instituted through a dual-lumen cannula in the right internal jugular vein [22].
Mechanical ventilation was weaned by 24 hours, and using a wheeled intravenous
pole to carry the pump, oxygenator, and oxygen tank, the patient was able to ambu-
late throughout the hospital, use a treadmill, and ride an exercise bicycle. He was
successfully transplanted and discharged home. These same authors subsequently
described 10 subjects in whom ambulatory ECMO was attempted for severe respi-
ratory failure due to ARDS, idiopathic pulmonary fibrosis, COPD, or pulmonary
arterial hypertension [23]. Six were ultimately liberated from ventilatory support
and four of these ambulated while on ECMO.
Much of the early ambulatory ECMO experience was gained in patients awaiting
transplantation. Because pretransplant mechanical ventilation increases the risk of
complications and predicts worse outcomes following surgery, using ECMO to pre-
vent or limit the duration of mechanical ventilation appears attractive. In a small
series, three patients intubated for end-stage lung disease were cannulated while
awaiting transplant [24]. Active rehabilitation, physical therapy, and ambulation
were achieved pretransplant and, by 1 week following surgery, all were free of
mechanical ventilation, out of the ICU, and walking. A similar approach has been
used in cystic fibrosis patients awaiting transplant [25]. In a small, retrospective
series, patients bridged with ECMO who participated in rehabilitation had shorter
posttransplant mechanical ventilation, incidence of ICU-AW, and length of stay in
the ICU and in the hospital [26]. ECMO has also been used to avoid intubation
altogether. In a small cohort, five patients with cardiopulmonary failure due to pul-
monary hypertension were placed on ECMO using local anesthesia [27]. All stabi-
lized immediately and three survived to transplantation (18–35 days later) and were
breathing spontaneously, eating and drinking, and participating actively in physical
therapy in the interim. This series is notable in that VA ECMO was the chosen
mode, in light of the underlying illness. In a larger series, 31 patients were trans-
planted from ECMO (range 2–53 days) and 19 were ambulatory at the time of sur-
gery [28]. Roughly half of these patients were managed with VA (or VVA) ECMO.
14 Mobilization During ECLS 257

This early experience has led many centers, not only to consider and trial ambu-
latory ECMO, but to actively develop and nurture specialized teams to aid mobiliza-
tion. Following the creation of a mobilization program aimed at ECMO patients, 35
of 100 patients with refractory respiratory or cardiac failure could be actively reha-
bilitated [29]. Since bridge-to-recovery patients tend to be more ill and suffering
from more organ failures than bridge-to-transplant patients, it is interesting that 16
of the 35 patients undergoing physical therapy were not transplant candidates.
Further, while most were treated with VV ECMO through a dual-lumen cannula in
the internal jugular vein, 4 patients were on VA ECMO (internal jugular vein to
subclavian artery) and 8 had femoral access. Of the 16 bridge-to-recovery patients,
14 survived, most going directly home following discharge. It should be empha-
sized that these results followed an explicit plan to create and train a specialized
team including physical and occupational therapists, perfusionists, critical care
RNs, critical care nurse practitioners, respiratory therapists, and ECMO intensivists
and surgeons.
Increasingly, patients with arterial cannulas for VA ECMO or having femoral
cannulas for VV ECMO are considered excellent candidates for mobilization.
Indeed, in some trials the great majority of patients underwent active mobilization
while cannulated via the femoral route [18, 21]. Rare instances of accidental decan-
nulation [18], cannula site bleeding [18], and iliopsoas hematoma [30] are reported.
Nevertheless, overall, the evidence supports the safety of ambulation with femoral
cannulation, even for VA ECMO [31, 32].

Barriers to Mobilization During ECLS

It is no small task to safely mobilize patients during ECMO. Barriers include sever-
ity of illness, resource and staffing limitations, obesity, equipment needs, risks of
catastrophic complications, and cultural barriers [20]. There is broad agreement that
mobilization is safe for critically ill patients, including those with endotracheal
tubes requiring mechanical ventilation [33], but still very limited confidence that we
can safely do this with the very sickest. Before mobilization, patients should be
screened for contraindications, such as those described in the trial by Schweickert
and colleagues as modified for ECMO in Table 14.1 [34]. Many ECMO patients
will have neurological failure as a consequence of the underlying illness or its treat-
ment, precluding extubation, ambulation, or any meaningful therapy. Severe circu-
latory failure may also be limiting, although many patients in shock and on
vasoactive infusion can sit, stand, and even walk [1]. In others, mobilization threat-
ens circuit function because movement changes cannula position, causes catheter
kinking, or provokes transient caval collapse. Morbid obesity, extensive surgical
wounds or dressings, hardware, and other patient-specific features may limit
mobility.
Fear of complications when ambulating is understandable since circuit failure, or
even a simple stumble, can be life-threatening. Accumulating data, however, are
258 G. A. Schmidt

Table 14.1 Safety screen: contraindications to mobilization

Mean arterial blood pressure <60 mmHg
Heart rate <40 or >130/min
Respiratory rate >40/min
Critical hypoxemia despite ECMO (SaO2 < 78%)
Increased intracranial pressure
Active hemorrhage
Active myocardial ischemia
Undergoing a procedure
Agitation requiring increased sedative in the last 30 min
Delirium precluding meaningful cooperation
Insecure device, especially the ECMO cannula
Malfunction of cannula or circuit with movement
Modified from Ref. [34]

reassuring. For example, in the small series cited above [17, 18], serious complica-
tions were not seen. In the larger series of bridge-to-transplant and bridge-to-­
recovery, there were no complications related to physical therapy [24]. The
physiological stress related to mobilization appears to be quite modest [35]. Even in
the broader experience gained in generally critically ill patients, physical therapy
and mobilization have been remarkably complication-free [14, 36]. In a prospec-
tive, randomized trial of early, aggressive physical therapy, only 4% of sessions
were terminated for safety concerns [1]. The most commonly reported were periods
of patient-ventilator dyssynchrony. Falls to knees without injury did not occur in
two studies [1, 14], but occurred rarely in another [13]. Accidental extubations were
not seen in these early studies. In a large, prospective evaluation of safety in a single
center, fewer than 1% of sessions had a potential safety event (most often arrhyth-
mia) and only 4 of 1110 required minimal additional treatment or cost [27].
A final barrier relates to each ICU’s culture of liberation in general and mobiliza-
tion in particular. For example, physical rehabilitation is built on a foundation of
sedative interruption, regular spontaneous breathing trials, delirium prevention, and
a multidisciplinary, team-based approach to patient care. A quality improvement
project aimed at reducing heavy sedation and bolstering staffing to include full-time
physical and occupational therapists succeeded in improving delirium, functional
mobility, and ICU length of stay [37]. Leadership is also essential. When imple-
menting a new mobility culture, champions should anticipate some resistance from
clinicians unconvinced of safety. At times, such resistance arises from physicians,
primary care clinicians, bedside nurses, and even family members. Physical thera-
pists appear ready to mobilize patients to higher levels than nurses, suggesting that
added therapist staffing is an effective means to change practice [38]. Time invested
in identifying supporters is valuable, but skeptics may respond to education or
active listening regarding their concerns.
The risks of early mobilization, although rare, are immediate, obvious, and
potentially catastrophic. In contrast, the benefits are subtle and delayed, leading
14 Mobilization During ECLS 259

some clinicians to urge “first do no harm.” When an adverse event complicates

mobilization (and there is little doubt that an eventual patient will suffer a complica-
tion), there is a risk that an emotional, short-term-oriented response could threaten
an entire mobility program. The ECMO team must emphasize the data supporting
mobilization, acknowledge that rare patients may be harmed, and be prepared to
defend the mobility team if their role is questioned. In the event of a complication,
we encourage a “just culture” approach, rather than finger-pointing.

Mobilization: Team and Methods

Once a patient passes the safety screen described above, the mobility team should
be assembled. Safely ushering an ECMO patient out of bed, and especially out of
the ICU, requires monitoring, attention to the cannula, and a perfusionist to roll and
attend to the circuit, in addition to the physical therapist: three caregivers at a mini-
mum and often several more (Fig. 14.1). It can be expensive to provide such an
experienced team on a regular basis, especially since physical therapy resources are
currently in high demand in the ICU. Specialized equipment can make this easier,

Fig. 14.1 Patient with

severe ARDS making a trip
outdoors, accompanied by
caregivers and family
260 G. A. Schmidt

but adds an additional layer of expense. The potential for effective therapy depends
on circuit design that is conducive to transport, as described in Chap. 2. Centers
beginning a program of mobilization are advised to consult with or visit experi-
enced ICUs.
The team should be experienced in the care and mobilization of the critically ill.
The complexity of mobilizing the ECMO patient, combined with the ever-present
risk of circuit disruption, demands a coordinated effort. The therapist should be the
single, clear team leader to state each step of the plan, communicate with the patient,
and judge whether to continue or abort the session, akin to the role of team leader
during cardiopulmonary resuscitation [39]. When the team leader clearly states
what will happen next, the perfusionist can anticipate how to maneuver and monitor
the circuit; the assistant can stabilize the cannula and guide its movement along with
the patient; and the patient can muster his resources to make a good effort. Because
patients are often anxious, a calm but firm approach is essential. Responsibilities of
the nurse include vital sign monitoring, observation for distress, and attention to the
stability of tubes and lines. The perfusionist attends to the cannula and circuit,
including adequacy of circuit flow, and anticipates and communicates to the team
leader any threat to its continued function. Additional staff may be needed to deal
with a mechanical ventilator (transport ventilator), provide hand-bag ventilation,
assist especially debilitated or obese patients, carry backup supplies, push wheel-
chairs and carts, or provide an additional level of monitoring for very fragile patients.
Before embarking on ambulation, necessary devices, such as endotracheal tubes,
should be examined for integrity and securement. Any inessential devices, cathe-
ters, or infusions should be discontinued temporarily. Sweep gas flow should be
disconnected from the wall oxygen, using a transport cylinder. Sweep gas flow rates
higher than maintenance may be needed to adjust for the increased metabolic
demand of exercise, and this should be factored in when calculating whether the
cylinder will support the entire session. Extra oxygen and a spare battery provide an
added level of safety. When leaving the bedside, it is useful to bring along a
wheelchair in case the patient tires or needs additional assistance.
Physical activity in the critically ill raises total body oxygen consumption mean-
ingfully and probably even more than in healthy volunteers performing a similar
degree of work [40]. This added burden could threaten a limited cardiopulmonary
reserve, so that careful monitoring is essential for safety. At a minimum, this
includes continuous assessment of rhythm, heart rate, oximetry, and subjective dis-
tress, as well as the ability to measure blood pressure. Any deterioration, including
those listed in Table 14.2, should prompt an immediate halt to mobilization, reas-
sessment of safety, and a decision whether to terminate the session. Returning the
patient to the ICU bed requires careful planning and may benefit from a checklist so
that essential steps (e.g., reconnecting the sweep gas to wall oxygen) are not omitted.
Cycle ergometry, with passive or increasingly active motion, increases muscle
force and functional status at the time of ICU discharge, providing another option
for strengthening [41]. Electrical stimulation of muscles has been hypothesized to
improve strength and limit wasting, but efficacy remains unproved [42]. As the
14 Mobilization During ECLS 261

Table 14.2 Terminating a therapy session

Mean arterial blood pressure <60 mmHg
Heart rate <40 or >130/min
Respiratory rate >40 that is sustained
Critical hypoxemia (SaO2 falling more than 5% from baseline)
New arrhythmia
Significant ventilator dyssynchrony
Significant new symptoms, such as chest pain
Inability to maintain adequate circuit flow
Circuit malfunction or alarming
Device dislodgement or instability
Subjective distress: verbal or facial cues, overt agitation
Unsteadiness sufficient to risk safely continuing
Fall

acceptance of ICU mobilization increases, equipment is becoming widely available

to aid mobilization of critically ill patients and make it safer.

Conclusion

Critically ill patients are at very high risk for ICU-AW and VIDD, potentially
increasing morbidity, complications, and length of stay. Physical therapy is clearly
effective in raising functional capacity and appears remarkably safe. Even during
ECMO, mobilization is possible with an experienced, dedicated team, bringing into
focus a new vision: the awake, extubated, ambulating patient.

References

1. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physical and occupational ther-
apy in mechanically ventilated, critically ill patients: a randomized controlled trial. Lancet.
2009;373:1874–82.
2. Engel HJ, Needham DM, Morris PE, et al. ICU early mobilization: from recommendation to
implementation at three medical centers. Crit Care Med. 2013;41:S69–80.
3. Peek GJ, Mugford M, Tiruvoipati R, et al. Efficacy and economic assessment of conventional
ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory
failure (CESAR): a multicentre randomised controlled trial. Lancet. 2009;374:1351–63.
4. Kress JP, Pohlman AS, O’Connor MF, et al. Daily interruption of sedative infusions in criti-
cally ill patients undergoing mechanical ventilation. N Engl J Med. 2000;342:1471–7.
5. Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine vs midazolam for sedation of criti-
cally ill patients: a randomized trial. JAMA. 2009;301:489–99.
6. Kress JP, Gehlbach B, Lacy M, et al. The long-term psychological effects of daily sedative
interruption on critically ill patients. Am J Respir Crit Care Med. 2003;168:1457–61.
262 G. A. Schmidt

7. Shehabi Y, Bellomo R, Reade MC, et al. Early intensive care sedation predicts long-term mor-
tality in ventilated critically ill patients. Am J Respir Crit Care Med. 2012;186:724–31.
8. Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safety of a paired sedation and ventilator
weaning protocol for mechanically ventilated patients in intensive care (awakening and breath-
ing controlled trial): a randomised controlled trial. Lancet. 2008;371:126–34.
9. Fan E, Cheek F, Chlan L, et al. An official American Thoracic Society clinical practice guide-
line: the diagnosis of intensive care unit-acquired weakness in adults. Am J Respir Crit Care
Med. 2014;190:1437–46.
10. Demoule A, Jung B, Prodanovic H, et al. Diaphragm dysfunction on admission to the intensive
care unit. Prevalence, risk factors, and prognostic impact – a prospective study. Am J Respir
Crit Care Med. 2013;188:213–9.
11. Wollersheim T, Woehlecke J, Krebs M, et al. Dynamics of myosin degradation in intensive care
unit-acquired weakness during severe critical illness. Intensive Care Med. 2014;40:528–38.
12. Sassoon CSH, Zhu E, Caiozzo VJ. Assist-control mechanical ventilation attenuates ventilator-­
induced diaphragm dysfunction. Am J Respir Crit Care Med. 2004;170:626–32.
13. Bailey P, Thomsen GE, Spuhler VJ, et al. Early activity is feasible and safe in respiratory fail-
ure patients. Crit Care Med. 2007;35:139–45.
14. Morris PE, Goad A, Thompson C, et al. Early intensive care unit mobility therapy in the treat-
ment of acute respiratory failure. Crit Care Med. 2008;36:2238–43.
15. Balas MC, Vasilevskis EE, Olsen KM, et al. Effectiveness and safety of the awakening and
breathing coordination, delirium monitoring/management, and early exercise/mobility bundle.
Crit Care Med. 2014;42:1024–36.
16. Thiagarajan RR, Teele SA, Teele KP, et al. Physical therapy and rehabilitation issues for
patients supported with extracorporeal membrane oxygenation. J Pediatr Rehabil Med.
2012;5:47–52.
17. Bonizzoli M, Lazzeri C, Drago A, Boninsegni LT, Donati M, Di Valvasone S, et al. Effects of a
physiotherapic program in patients on venovenous extracorporeal membrane oxygenation: an
8-year single-center experience. Minerva Anestesiol. 2019;85:989–94.
18. Braune S, Bojes P, Mecklenburg A, Angriman F, Soeffker G, Warnke K, et al. Feasibility,
safety, and resource utilization of active mobilization of patients on extracorporeal life support:
a prospective observational study. Ann Intensive Care. 2020;10:161.
19. The ECMO-PT Study Investigators and International ECMO Network. Early mobilization
during extracorporeal membrane oxygenation was safe and feasible: a pilot randomized con-
trolled trial. Intensive Care Med. 2020;46:1057–9.
20. Marhong JD, DeBacker J, Viau-Lapointe J, Munshi L, Del Sorbo L, Burry L, et al. Sedation
and mobilization during venovenous extracorporeal membrane oxygenation for acute respira-
tory failure: an international survey. Crit Care Med. 2017;45:1893–9.
21. Wells CL, Forrester J, Vogel J, Rector R, Tabatabai A, Herr D. Safety and feasibility of early
physical therapy for patients on extracorporeal membrane oxygenator: University of Maryland
Medical Center experience. Crit Care Med. 2018;46:53–9.
22. Garcia JP, Iacono A, Kon ZN, et al. Ambulatory extracorporeal membrane oxygenation: a new
approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg. 2010;139:e137–9.
23. Garcia JP, Kon ZN, Evans C, et al. Ambulatory veno-venous extracorporeal membrane oxy-
genation: innovation and pitfalls. J Thorac Cardiovasc Surg. 2011;142:755–61.
24. Turner DA, Cheifetz IM, Rehder KJ, et al. Active rehabilitation and physical therapy dur-
ing extracorporeal membrane oxygenation while awaiting lung transplantation: a practical
approach. Crit Care Med. 2011;39:2593–8.
25. Hayes D Jr, Kukreja J, Tobias JD, et al. Ambulatory venovenous extracorporeal respiratory
support as a bridge for cystic fibrosis patients to emergent lung transplantation. J Cyst Fibros.
2012;11:40–5.
26. Rehder KJ, Turner DA, Hartwig MG, et al. Active rehabilitation during extracorporeal mem-
brane oxygenation as a bridge to lung transplantation. Respir Care. 2013;58:1291–8.
14 Mobilization During ECLS 263

27. Olsson KM, Simon A, Strueber M, et al. Extracorporeal membrane oxygenation in nonintu-
bated patients as bridge to lung transplantation. Am J Transplant. 2010;10:2173–8.
28. Hoopes CW, Kukreja J, Golden J, et al. Extracorporeal membrane oxygenation as a bridge to
pulmonary transplantation. J Thorac Cardiovasc Surg. 2013;145:862–8.
29. Abrams D, Javidfar J, Farrand E, et al. Early mobilization of patients receiving extracorporeal
membrane oxygenation: a retrospective cohort study. Crit Care. 2014;18:R38.
30. Taniguchi H, Ikeda T, Takeuchi I, Ichiba S. Iliopsoas hematoma in patients undergoing veno-
venous ECMO. Am J Crit Care. 2021;30:55–63.
31. Pasrija C, Mackowick KM, Raithel M, Tran D, Boulos FM, Deatrick KB, et al. Ambulation
with femoral arterial cannulation can be safely performed on venoarterial extracorporeal mem-
brane oxygenation. Ann Thorac Surg. 2019;107:1389–94.
32. Da Cunha FD, Marcolino MAZ, Macagnan FE, Della Mea Plentz R, Kessler A. Safety and
potential benefits of physical therapy in adult patients on extracorporeal membrane oxygen-
ation support: a systematic review. Rev Bras Ter Intensiva. 2019;31:227–39.
33. Hodgson CL, Stiller K, Needham DM, et al. Expert consensus and recommendations on
safety criteria for active mobilization of mechanically ventilated critically ill adults. Crit Care.
2014;18:658. https://doi.org/10.1186/s13054-­014-­0658-­y.
34. Pohlman MC, Schweickert WD, Pohlman AS, et al. Feasibility of physical and occupational
therapy beginning from initiation of mechanical ventilation. Crit Care Med. 2010;38:2089–94.
35. Hayes K, Holland AE, Pellegrino VA, Young M, Paul E, Hodgson CL. Early rehabilitation dur-
ing extracorporeal membrane oxygenation has minimal impact on physiological parameters: a
pilot randomized controlled trial. Aust Crit Care. 2021;34:217–25.
36. Sricharoenchai T, Parker AM, Zanni JM, et al. Safety of physical therapy interventions in criti-
cally ill patients: a single-center prospective evaluation of 1110 intensive care unit admissions.
J Crit Care. 2014;29:395–400.
37. Needham DM, Korupolu R, Zanni JM, et al. Early physical medicine and rehabilitation for
patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil.
2010;91:536–42.
38. Garzon-Serrano J, Ryan C, Waak K, et al. Early mobilization in critically ill patients:
patients’ mobilization level depends on health care provider’s profession. Phys Med Rehabil.
2011;3:307–13.
39. Yeung JH, Ong GJ, Davies RP, et al. Factors affecting team leadership skill and their relation-
ship with quality of cardiopulmonary resuscitation. Crit Care Med. 2012;40:2617–21.
40. Hickmann CE, Roeseler J, Castanares-Zapatero D, et al. Energy expenditure in the critically ill
performing early physical therapy. Intensive Care Med. 2014;40:548–55.
41. Burtin C, Clerckx B, Robbeets C, et al. Early exercise in critically ill patients enhances short-­
term functional recovery. Crit Care Med. 2009;37:2499–505.
42. Kho ME, Truong AD, Zanni JM, et al. Neuromuscular electrical stimulation in mechanically
ventilated patients: a randomized, sham-controlled pilot trial with blinded outcome assess-
ment. J Crit Care. 2015;30:32–9.
Chapter 15
ECMO Weaning and Decannulation

Sharon L. McCartney and Sundar Krishnan

Introduction

As native lung function improves in patients with acute respiratory distress syn-
drome (ARDS) on extracorporeal life support, therapy is directed toward weaning
from extracorporeal membrane oxygenation (ECMO) and decannulation. The speed
with which patients can be weaned off support depends on the improvement in their
cardiopulmonary function and needs to be titrated to individual response. In this
chapter, we will discuss the signs of readiness; management of the ventilator, oxy-
genator sweep gas, and ECMO flow during the weaning process; and cardiopulmo-
nary assessment during weaning. Further, we will discuss the procedure of
decannulation and the follow-up of patients after decannulation. The process of
weaning and trialing off extracorporeal support is radically different for venovenous
(VV) and venoarterial (VA) ECMO, because of the difference in physiological sup-
port provided by the two modalities. Hence, we will present a short, separate section
on weaning from VA ECMO. Finally, we review the termination of ECMO in
patients where further therapy is deemed unrecoverable.

S. L. McCartney · S. Krishnan (*)

Divisions of Cardiothoracic Anesthesia and Critical Care, Department of Anesthesiology,
Duke University Medical Center, Durham, NC, USA
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature 265

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_15
266 S. L. McCartney and S. Krishnan

Weaning from VV ECMO

Signs of Readiness

Therapy directed toward improving pulmonary function should start prior to patients
being placed on ECMO and continue after ECMO support has been initiated.
Therapies should be targeted to the individual pathology that led to the need for
respiratory support, but in general may include diuresis, antibiotic therapy, broncho-
dilators, and bronchoscopy with airway clearance. Ventilator management while on
ECMO involves limiting FiO2 to the lowest possible (21–50%), maintaining PEEP
to preserve recruitment, and limiting tidal volumes to 4–8 mL/kg.
For ventilated ARDS patients supported with mechanical ventilation, improve-
ment in pulmonary function can be assessed initially while the patient is still on
full ECMO support. Even with ventilator settings unchanged from “rest” settings
typically used during extracorporeal support, an increase in PaO2 or a decrease in
sweep flow to maintain normocarbia might occur, indicating improving native
lung function. Additionally, improved pulmonary compliance will lead to
increased tidal volumes on unchanged pressure-control ventilator settings.
Usually, respiratory system compliance >20 mL/cm H2O suggests adequate recov-
ery to proceed with weaning. An improvement in pulmonary aeration on chest
radiograph can accompany improving alveolar recruitment, but at times the radio-
graph may lag behind clinical improvement. Lung ultrasound can be used to eval-
uate the subpleural areas of the lung. Signs of improving lung function are listed
in Table 15.1.
Once significant aeration is achieved, as assessed by improving tidal volumes
(>150 mL) on rest settings and improvements in chest radiograph, the Cilley test
should be performed daily. The patient should be hemodynamically stable and
comfortable on mechanical ventilation. To perform the Cilley test, also called the
oxygen challenge test, the FiO2 is increased to 100%, with no other changes in
ventilator settings. If the patient has improved lung aeration, there will be a rapid
increase in oxygen saturation within a couple of minutes and an arterial blood gas
should show a PaO2 of >225 mmHg at 15 min after the increase in FiO2 [1]. This

Table 15.1 Signs of improving lung function

Improved pulmonary Improved lung compliance, as observed by increased tidal volumes
mechanics on stable inspiratory pressure
Improved gas exchange Increase in PaO2 or decrease in PaCO2 on rest settings on the
ventilator
Increased CO2 levels in capnometry on rest settings on the ventilator
Decrease in sweep gas flow needed to maintain normocarbia
Rapid improvement in SpO2 or PaO2 with increase in FiO2 to 100%
(Cilley test)
Signs on imaging studies Improved lung aeration on chest X-ray
Resolving B-lines on lung ultrasound
15 ECMO Weaning and Decannulation 267

demonstrates the patient is ready for advancement of ventilator settings and a

sweep trial. However, one recent single-center study of 253 patients concluded that
the oxygen challenge test was a poor predictor of readiness to decannulate from
VV ECMO [2].

ECMO Weaning and Liberation

ARDS patients with improving pulmonary function can be weaned and liberated
from extracorporeal support. When the lowest ECMO flow to provide adequate sup-
port at low ventilator settings is used, weaning from VV ECMO becomes a more
gradual process. Additionally, with improving lung function, sweep gas flow rate
will decrease to maintain normocarbia, indicating improved alveolar gas exchange.
As such, weaning of VV ECMO should be a routine progression of ECMO manage-
ment in the setting of improving pulmonary function. On occasion, the decision to
wean off extracorporeal support might be hastened by bleeding complications
related to anticoagulation. With moderate ventilatory support, a trial off during VV
ECMO includes assessing native lung function while discontinuing all sweep gas
flow to the ECMO circuit. Importantly, blood flow continues through the VV circuit,
without extracorporeal gas exchange. Finally, liberation from ECMO involves
decannulation and discontinuation of anticoagulation. It is generally accepted that
when extracorporeal support is contributing to less than 30% of native organ func-
tion, it is possible to trial off ECMO [3, 4]. An algorithm for VV ECMO decannula-
tion used at the authors’ institution is presented in Fig. 15.1.

Ventilator Management

When patients are supported by VV ECMO, it is common practice to place the

patient on “rest” ventilator settings with reduced tidal volumes, peak pressures,
FiO2, and PEEP [5, 6]. When the decision is made to advance from “rest” settings,
recruitment maneuvers might be required before increased ventilator settings are
employed. Further weaning is only possible when the patient can achieve adequate
tidal volumes (approximately 6 mL/kg), with acceptable airway pressure (plateau
airway pressure <30 cm H2O). Patients can be on assisted or spontaneous modes of
ventilation during weaning and trials off VV ECMO. Adequate minute ventilation
will have to be established before patients can be weaned or trialed off extracorpo-
real support. Ventilator settings during weaning and trials off must be both lung-­
protective and at a level at which the team would feel comfortable decannulating.
Ideally, the patient should be supportable with minimal ventilator settings, with the
ability to increase ventilatory support if needed, once extracorporeal support is
removed.Respiratory rate, tidal volume, gas exchange, and respiratory distress are
vital clues to pulmonary function during weaning and trialing off from ECMO,
268 S. L. McCartney and S. Krishnan

Improved pulmonary function

Recruitment Maneuvers

Ventilator settings:
FiO2 <50% TV 6 ml/kg,
Pplat<30 mm Hg, RR <25

Maintain ECMO flow. Decrease sweep

gas flow in steps to 1Ipm.

NO Are respiratory and

hemodynamic
parameters stable?

YES

Turn off sweep gas flow.

Maintain ECMO blood flow.

Patient is ready for

Patient is not ready to be
NO Over the next 4-24 YES discontinuation of
weaned off ECMO. Continue
hours, are ECMO support.
VV ECMO support. Maintain
adequate oxygenator sweep respiratory and
hemodynamic Plan for
gas flow and ECMO blood
parameters stable? decannulation.
flow.

Fig. 15.1 Algorithm for VV ECMO weaning and trial off

much as one would judge the success of a spontaneous breathing trial. Additionally,
P0.1 (negative pressure generated during the first 100 msec after the initiation of an
inspiratory effort against a closed respiratory circuit) will increase in patients with
increased work of breathing during the ECMO wean and trial off. While values
>2–3 cm H2O have been shown to be predictive of failed extubation, there is cur-
rently no validated P0.1 cutoff to assess readiness for liberation from ECMO
support.
15 ECMO Weaning and Decannulation 269

Oxygenator Sweep Gas Flow Management

There are multiple ways to wean ECMO and different institutions have different
preferences. Various methods will be described subsequently.

Separate Deoxy and Carboxy Challenge

In this weaning strategy, the patient’s oxygenation is trialed prior to their ability to
decarboxylate [1]. Before commencing the oxygenation challenge, the FiO2 on the
ventilator is increased to 60%, since 60% or less would be an adequate FiO2 on the
ventilator for decannulation. Additionally, ventilator settings are increased to non-­
rest settings, but targeting tidal volumes of 6–8 mL/kg ideal body weight. After a
few minutes at these settings, the sweep gas FiO2 is decreased sequentially from
100% to 60%, 30%, and then 21% [1]. During the FiO2 wean, a peripheral satura-
tion of >88% should be maintained. If the patient’s saturation falls below this value,
the patient does not meet criteria for continuing the wean and the patient is placed
back on full ECMO support and rest settings on the ventilator can be resumed. If the
patient continues to meet weaning criteria and the FiO2 of the sweep gas is decreased
to 21%, the carboxy challenge can be performed. In the carboxy challenge, the
sweep gas flow rate is reduced by 30% every 5–10 min while measuring the patient’s
response (end-tidal CO2, CO2 by blood gases, in addition to physical exam findings
such as tachypnea) [1]. Once the sweep gas flow is turned off, the weaning test is
successful and a trial off ECMO can ensue. Blood gases can be drawn to continue
to ensure adequate oxygenation and CO2 removal. ECMO circuit flow is never
reduced and additional anticoagulation is not necessary during trial off ECMO. Trials
off ECMO can last between 4 and 24 h prior to decannulation.

Simultaneous Deoxy and Carboxy Challenge

In this weaning strategy, oxygenation and ventilation are trialed simultaneously.

Before commencing the trial, the ventilator is set to achieve adequate tidal volumes
and minute ventilation on lung-protective ventilation. The FiO2 on the ventilator is
increased to 60% or less, since this would be an appropriate FiO2 for decannulation.
For this trial, the sweep gas flow should be low while on ECMO (typically ≤2 L/
min). Once ready, the sweep flow gas is discontinued, while the ECMO flow remains
stable. Gas exchange, work of breathing, and hemodynamic status should be
assessed during this trial. Importantly, there needs to be no interruption in blood
flow in the extracorporeal circuit during a trial off VV ECMO. The return tubing
from the oxygenator will change in color from oxygenated blood to deoxygenated
blood, similar to the tubing draining blood from the patient. Blood gases should be
checked frequently at the beginning of the trial and show adequate oxygenation and
270 S. L. McCartney and S. Krishnan

Table 15.2 Signs of trial off ECMO failure

Ventilator parameters Tachypnea
FiO2 requirement >60%
Inability to maintain oxygenation and ventilation on lung-protective
ventilation
Very high P0.1
Blood gas parameters pO2 < 60 mmHg
Hypercarbia
Acidosis
Hemodynamic Hypotension
parameters Tachycardia

ventilation. As the trial off ECMO continues, blood gases can be less frequent if the
oxygenation and ventilation have been sufficient. Signs of failure of trial off are
detailed in Table 15.2. Decannulation can follow if trials are successful for 2–24 h.
At some institutions, or in select patients, a trial off can proceed for a longer time
before decannulation is deemed appropriate. If the patient shows signs of inade-
quate native pulmonary function, the sweep gas flow is simply turned back on and
extracorporeal support is resumed. Ventilatory settings and sedation can be man-
aged to the patient’s comfort, until the next trial off period.

ECMO Flow Management

Circuit blood flow does not need to be changed while attempting weaning from VV
ECMO. However, as the patient’s intravascular volume is decreased with diuresis, it
might be difficult to maintain a high rate of blood flow through the extracorporeal
circuit. As the patient’s pulmonary gas exchange improves, their demands from
ECMO may lessen and it may be possible to decrease circuit blood flow. A mini-
mum flow rate of 500–1000 mL/min needs to be maintained to prevent stasis of
blood in the cannulae, circuit, pump, and oxygenator. Low blood flow rates through
the ECMO circuit should also prompt higher anticoagulation goals to prevent
thrombosis.

Sedation Management

While patients are on VV ECMO, sedation should be maintained at the lowest level
possible to maintain patient comfort and hemodynamic stability. In patients with a
tracheostomy, it may be possible to maintain complete wakefulness without seda-
tion during ECMO weaning or trial off ECMO, while intubated patients may require
some degree of sedation to maintain comfort. In patients on mechanical ventilation,
sometimes sedation may need to be increased during ECMO weaning or trial off to
prevent ventilator dyssynchrony. Each patient will have different needs, but the low-
est level of sedation to maintain patient comfort, hemodynamic stability, and venti-
lator synchrony should be used.
15 ECMO Weaning and Decannulation 271

Hemodynamic Management

Cardiopulmonary function should be closely monitored during the weaning and

trial off VV ECMO. Increased work of breathing, altered gas exchange, and
increased airway pressures might lead to cardiac decompensation during weaning
and the trial off period. In addition to standard hemodynamic monitoring and assess-
ment of the need for inotropic therapy, echocardiography can be a valuable tool in
assessing cardiac function. Right ventricular dilation when oxygenator sweep gas
flow is decreased or turned off can be a clue toward hemodynamic intolerance.

Anticoagulation Management

Since there is no interruption of blood through the venous cannulae, extracorporeal

tubing, pump, or oxygenator during a trial off VV ECMO, there is no increased risk
of thrombosis. If the patient appears to have improved lung function to the point that
the trial off ECMO is expected to be successful, it would be reasonable to turn the
heparin infusion off in anticipation of decannulation at the end of the trial off period.
More commonly, anticoagulation is continued through the trial off period. If the
patient passes the trial off, ECMO flow is continued while the sweep gas remains
off, and anticoagulation is turned off in coordination with the plan for decannula-
tion. If the patient then fails their trial off VV ECMO, anticoagulation continues as
before and the extracorporeal support is resumed.

Decannulation

Decannulation is performed after a trial off proves that recovery is adequate to

maintain life support through more conventional means. Decannulation should be a
progression from trial off ECMO and changes to level of ventilatory support com-
pared to trial off should not be necessary. Decannulation is usually performed as a
planned bedside procedure. A checklist for VV ECMO decannulation is provided in
Fig. 15.2.

Fig. 15.2 Checklist for

VV ECMO decannulation
272 S. L. McCartney and S. Krishnan

Before decannulation, cardiopulmonary stability since the trial off needs to be

confirmed. Heparin is discontinued at least 30–60 min before the procedure.
Extracorporeal circuit flow is turned off and the drainage and return lines are
clamped. Following cannula removal, it should be reasonable to administer prot-
amine to correct residual coagulopathy if necessary. Removal of large venous can-
nulae with side holes might be associated with venous air embolism. Patients should
hence be positioned such that the cannula site is dependent. Additionally, a Valsalva
maneuver should prevent a significant negative pressure gradient into the venous
system. After removal of percutaneously placed cannulae, nonocclusive pressure
needs to be held at the venous puncture sites for at least 30 min. For cannulae placed
via surgical cutdown, a surgical approach and repair might be necessary.

Following Decannulation

Since patients have demonstrated adequate gas exchange (and hemodynamic sta-
bility) prior to stopping ECLS, decannulation is usually well-tolerated. Continued
recovery can be anticipated and attention turned to further healing, rehabilitation,
and prevention of complications. Additionally, if appropriate, further weaning and
liberation from mechanical ventilator may be appropriate. Procedures that were
withheld because of the anticoagulation necessary for extracorporeal support (e.g.,
tracheostomy, chest tube placement, invasive line placement) can now be
performed.
Some patients will have setbacks, for example, with worsening dyspnea. It is
helpful to return to basic principles, considering the differential diagnosis for the
deterioration (fluid overload, thromboembolism), performing appropriate diagnos-
tic studies, and using the ventilator to support respiratory function (or, for the extu-
bated patient, noninvasive ventilation or re-intubation). In the 24–48 h after
decannulation, patients also need to be assessed for venous thrombosis at the can-
nula site.

Extubation Prior to Decannulation

In an international survey, 90% of ECMO centers keep patients mechanically venti-

lated while weaning VV ECMO [5]. The decision to remove ventilator support first
versus weaning off ECMO first depends on multiple factors. As mentioned previ-
ously, hemorrhagic complications might lead to discontinuing ECMO sooner than
planned. On the other hand, for patients who require significant sedation to tolerate
the endotracheal tube, extubation with continued ECMO support might assist in
mobilization and enhanced recovery. The institution’s ability to provide prolonged
ECMO support must be taken into account in patients expected to require respira-
tory support for a long duration. In addition, the condition of the ECMO circuit
(thrombus in cannulas, thrombus in oxygenator, function of oxygenator) is a
15 ECMO Weaning and Decannulation 273

paramount consideration before withdrawing ventilator support. Finally, the inten-

sive care unit should be able to provide emergency ventilatory support in case of
respiratory failure in an extubated patient recently decannulated from ECMO.

Weaning from VA ECMO

Weaning from VA ECMO is different from VV ECMO for three major reasons: first,
both the cardiac and pulmonary systems will be gradually removed from extracor-
poreal support; second, the ECMO flow is decreased but sweep gas flow is
unchanged; and third, ECMO support can never truly be interrupted for long periods
of time because of concerns for stasis.
Assessment of readiness for weaning and discontinuation should be performed
daily. Signs of improving organ function include improved pulsatility in the arterial
waveform, reduction in cardiac filling pressures, improved organ function, and an
improvement in radiographic and sonographic signs [7].
For weaning, ECMO flow is reduced slowly by 0.5–1 L/min every 4–6 h.
Ventilatory and hemodynamic support might have to be uptitrated as ECMO wean
progresses. At a flow of about 1 L/min, if the patient can maintain a mean arterial
pressure >65 mmHg, mixed venous oxygen saturation >65%, and an arterial satura-
tion > 90% with low ventilator settings and low vasoactive support, they might be
ready for a trial off from ECMO support. After adequate information has been
acquired, the patient is placed back on 1.5–2 LPM flow while awaiting
decannulation.
Often, the trial off from ECMO occurs in the operating room, with hemodynamic
and echocardiographic monitoring. With clamps on the cannulae, the patient is sep-
arated from the pump. A bridge between the two limbs of the circuit keeps blood
flowing across the extracorporeal circuit. If a prolonged trial is desired, occasional
flashing (release of clamps on the cannulae) can be used to keep clots from forming.
However, since this maneuver introduces oxygenated blood into the patient, any
further assessments of gas exchange should be made prior to flashing the cannulae.
Decannulation is performed in the operating room, to allow for good vascular
control of the arterial cannulation site. For patients placed on VA or VV ECMO as
a bridge to transplantation or durable mechanical support, decannulation occurs in
the operating room along with the surgical procedure.

Termination of ECLS in Those Who Do Not Recover

Most patients treated with ECLS will be successfully weaned or bridged to trans-
plant. Others, however, simply never recover. The primary disease may fail to
resolve; contraindications to transplant arise; ECLS complications (especially intra-
cranial hemorrhage) change the trajectory of illness; or the consequences of critical
274 S. L. McCartney and S. Krishnan

Table 15.3 Common reasons for failure of ECLS in ARDS

Category of failure Reason for failure
Primary illness progresses or fails to remit Proliferative ARDS, lung fibrosis
Cor pulmonale
Multiorgan failure
Contraindication prevents transplant Treatment-resistant infection
Severe heparin-induced thrombocytopenia
Highly HLA-sensitized status
Extensive pleural disease
ECLS complication Intracranial hemorrhage
Intractable hemorrhage
Systemic embolism, especially neurological
Access failure
Critical illness complication Septic shock
Extensive myocardial infarction

illness, such as nosocomial pneumonia, septic shock, or acute renal failure turn a
desperate situation into an unrecoverable downward spiral (Table 15.3). It is impor-
tant to remember that some patients will recover despite many weeks, even months,
of ECLS, so decisions to terminate support should not be made casually.
Determining that continued critical care is futile is controversial and exceedingly
complex [8, 9], perhaps more so in ECLS patients [10, 11]. Difficult decisions are
made easier when there is clarity of goals even before ECLS is begun and when
these goals are revisited regularly. When initiating ECLS, its purpose should be
stated and these goals should be shared amongst the healthcare team and the patient
(or surrogates). If ECLS is begun as a bridge to transplant, it is imperative to assess
candidacy realistically and to be certain that there are no contraindications to trans-
plant that could strand the patient on a “bridge to nowhere” [11]. Laying out goals
for patients who are being bridged to recovery is more challenging, because recov-
ery can be protracted and hard to predict accurately, and the course may be punctu-
ated with periodic downturns that make survival seem unlikely. The ECLS team
should meet regularly with the patient to address status, progress toward the goals
(or regression), and prognosis. For those unlikely to survive, the concept of a time-
limited trial (in which explicit metrics and a clear schedule are enunciated) [12] can
be helpful for patients and families. While the ethics of prolonging or terminating
extracorporeal support are discussed in future chapters, we will discuss the practical
aspects of ECLS termination here.
When the caregivers and patient concur that continued ECLS cannot achieve its
goals and should be withdrawn, there is no ethical dilemma. In this circumstance,
we advocate changing the treatment goals to focus on comfort and withdrawing all
unnecessary ICU interventions. Since ECLS is never instituted as destination ther-
apy, most consider its termination to be similar to withdrawing other life-support
treatments. Nevertheless, since withdrawal generally leads to immediate death,
some discomfort is understandable, especially when caring for conscious patients.
15 ECMO Weaning and Decannulation 275

Once a decision is made to cease ECLS in a patient expected to die, this change
in plan should be communicated to primary caregivers, consultants, referring physi-
cians, ICU nurses, perfusionists, therapists, and other stakeholders. Involvement of
palliative care specialists and spiritual care services may be helpful [13]. Patient
comfort should be assured, using analgesics and sedatives as needed, and anticipat-
ing that withdrawal of VV ECLS may provoke or heighten dyspnea. We then turn
off the sweep gas, preventing further gas exchange in the membrane. Assuming VV
ECMO was life-sustaining, the termination of support will lead to death. For those
on VA ECLS, circuit blood flow can be reduced as vasoactive drug infusions are
stopped. The patient should be kept comfortable from worsening dyspnea or other
perceived discomfort.

Funding Statement Support was provided solely from institutional and/or departmental sources.

References

1. Vasques F, Romitti F, Gattinoni L, et al. How I wean patients from veno-venous extra-­corporeal
membrane oxygenation. Crit Care. 2019;23:316.
2. Hartley EL, Sanderson B, Vasques F, Daly KJ, Lozinski M, Barrett NA, Camporota
L. Prediction of readiness to decannulation from venovenous extracorporeal membrane oxy-
genation. Perfusion. 2020;35(1_suppl):57–64.
3. Lynch WR. Weaning, trialing and futility. In: ECMO: extracorporeal cardiopulmonary sup-
port in critical care. 4th ed. ELSO; 2012.
4. ELSO guidelines for adult respiratory failure; August 2017.
5. Marhong JD, Munshi L, Detsky M, et al. Mechanical ventilation during extracorporeal life
support (ECLS): a systematic review. Intensive Care Med. 2015;41(6):994–1003.
6. Marhong JD, Telesnicki T, Munshi L, et al. Mechanical ventilation during extracorporeal
membrane oxygenation. An international survey. Ann Am Thorac Soc. 2014;11(6):956–61.
7. Fried JA, Masoumi A, Takeda K, Brodie D. How I approach weaning from venoarterial
ECMO. Crit Care. 2020;24(1):307.
8. Luce JM. A history of resolving conflicts over end-of-life care in intensive care units in the
United States. Crit Care Med. 2010;38:1623–9.
9. Siegel MD. End-of-life decision making in the ICU. Clin Chest Med. 2009;30:181–94.
10. Rosenberg AA, Haft JW, Bartlett R, Iwashyna TJ, Huang SK, Lynch WR, Napolitano
LM. Prolonged duration ECMO for ARDS: futility, native lung recovery, or transplantation?
ASAIO J. 2013;59:642–50.
11. Abrams DC, Prager K, Blinderman CD, Burkart KM, Brodie D. Ethical dilemmas encountered
with the use of extracorporeal membrane oxygenation in adults. Chest. 2014;145:876–82.
12. Schenker Y, Tiver GA, Hong SY, White DB. Discussion of treatment trials in intensive care. J
Crit Care. 2013;28:862–9.
13. Cook D, Rocker G. Dying with dignity in the intensive care unit. N Engl J Med.
2014;370:2506–14.
Chapter 16
Venoarterial ECMO in Respiratory Failure

Avery Tung and Tae H. Song

Introduction

When used to treat respiratory failure, the primary benefit of extracorporeal circula-
tion (ECMO) is to facilitate gas exchange. With a membrane lung (ML) in the extra-
corporeal circuit, oxygen can be added, and carbon dioxide can be removed from
venous blood to preserve oxygenation and carbon dioxide homeostasis. In that way,
patients with damaged or diseased lungs incapable of adequate gas exchange can be
supported until their lungs recover.
Although venoarterial (VA) ECMO is more commonly used in adults overall [1],
venovenous (VV) ECMO is more common in patients with respiratory failure due
to ARDS [2]. However, VA ECMO may also be used to provide cardiovascular sup-
port to patients with both hemodynamic instability and respiratory failure. When
configured in this manner, the ECMO circuit returns oxygenated blood not to the
venous system but to the arterial circulation. This approach can widen the range of
patients with respiratory failure who can be supported by ECMO technology.
VA ECMO may be indicated in two broad categories of patients with respiratory
failure. The first involves disease states that, when combined with respiratory fail-
ure, may require VA support. Examples include cardiogenic shock due to failure of
the right or left ventricle (or both), septic/vasoplegic shock requiring augmentation
of cardiac output for blood pressure support, and other conditions that prevent ade-
quate cardiac output such as pulmonary emboli, myocarditis, intracardiac shunts, or
persistent arrhythmias.

A. Tung (*)
Department of Anesthesia and Critical Care, University of Chicago, Chicago, IL, USA
e-mail: [email protected]
T. H. Song
Section of Cardiac Surgery, Department of Surgery, University of Chicago, Chicago, IL, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature 277

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_16
278 A. Tung and T. H. Song

In addition, VA ECMO may be used when oxygenation via VV ECMO alone

is inadequate. Because both drainage and return cannulas for VV ECMO reside
in the same vascular tree, recirculation (where oxygenated blood from the
ECMO return cannula is drawn back into the ECMO circuit rather than passing
through end organs) may be present [3]. When enough ECMO output is “wasted”
in this manner, the ability of VV ECMO to support gas exchange is diminished.
Under such conditions adjusting cannula position may improve systemic
oxygenation.
In some cases, a patient with hypoxemic respiratory failure may have low arte-
rial oxygen saturations despite maximal use of VV ECMO and optimal delivery
(minimal or no recirculation). Targeting higher hemoglobin levels to increase
oxygen delivery is one strategy. Another is conversion to VA ECMO or addition
of an arterial cannula to the ECMO return flow to reroute oxygenated blood to the
aortic arch. This problem typically occurs when the native cardiac output is high,
and VV ECMO is not able to generate a sufficiently high arterial O2 saturation. In
such cases, and particularly when the patient is agitated due to low O2 satura-
tions, VA ECMO may improve delivery of oxygenated blood to the cerebral cir-
culation and allow the patient to be awake and interactive despite severe
respiratory failure.
This chapter will review basic principles of VA ECMO, discuss relevant manage-
ment paradigms for VA ECMO and how they differ from VV ECMO, and suggest
how clinicians might determine when VV ECMO is inappropriate or ineffective,
what additional tests or clinical findings might lead to choosing VA ECMO over VV,
and what cues might drive conversion to VA ECMO for a patient already on
VV ECMO.
Specific cases of respiratory failure where VA ECMO may offer a clinical advan-
tage to VV ECMO alone will be discussed (Table 16.1).

Table 16.1 Potential indications for VA ECMO in respiratory failure

1. Concurrent LV dysfunction
 (a) Ischemic heart disease
 (b) Takotsubo syndrome
 (c) Stunned myocardium
 (d) Myocarditis
2. Severe RV dysfunction
 (a) Acute cor pulmonale due to ARDS
 (b) Pulmonary embolism
3. High cardiac output states
 (a) Septic shock
4. Intractable arrhythmias
16 Venoarterial ECMO in Respiratory Failure 279

Fundamentals of VA ECMO

Anatomy

As the name suggests, VA ECMO differs from VV ECMO in locating the ECMO
return cannula in the arterial side instead of the venous side of the circulation.
Venous blood is drained from the right atrium or superior/inferior vena cava into the
pump, routed through an ML, and returned to the body via the arterial system.
Locations for the venous (drainage) cannula are described in Chaps. 3 and 4 and
include central, femoral, internal jugular, and (less commonly) subclavian sites. For
VA ECMO, the size and location of the return cannula is limited by the size of the
target artery and the ability to preserve flow distal to the artery after partial or total
occlusion due to the cannula. Common locations for the return cannula thus include
central, femoral, and axillary arteries. Although cannula locations are similar to
those of a standard cardiopulmonary bypass (CPB) circuit, usually a CPB circuit
involves passive drainage of venous blood into a reservoir which then feeds the
centrifugal pump. In contrast, VA ECMO has no reservoir, so adequacy of venous
drainage is crucial to arterial return.

Equipment

Cannulation techniques for the drainage cannula are similar to those used for VV
ECMO. Because VA ECMO is often used to support cardiac output and blood pres-
sure in hemodynamically unstable patients, adequacy of venous drainage is critical
to hemodynamic support and locating the tip of the venous cannula in the right
atrium and/or vena cavae is thus a necessary component of VA ECMO. Venous can-
nula sizes for VA ECMO in adults typically range from 21 to 27F.
As with VV ECMO, a centrifugal pump is used to draw blood from the venous
drainage line, actively pump it through a ML, and return it to the circulation. Unlike
VV ECMO however, VA ECMO creates a parallel circuit where blood bypasses the
heart and lungs and is pumped from the venous to arterial circulation. A ML is thus
mandatory for VA ECMO as deoxygenated venous blood would otherwise be
shunted into the arterial system and produce severe arterial oxygen desaturation.
The sweep gas therefore cannot be turned off in VA ECMO. With large, well-­
positioned cannulas, it is possible to generate blood flows as high as 7 LPM, but 3–5
LPM flows are more typical.
Return cannulas are usually 15–21F in diameter. Although larger cannulas may
permit higher flow, they are also associated with a higher incidence of cannulation-­
related adverse events such as bleeding and vascular injury. In addition, for a return
cannula that is inserted into the femoral artery, ischemia of the ipsilateral leg may
occur unless a distal perfusion catheter is placed. This catheter is often a 6- to 8-F
introducer placed distal to the insertion site that diverts blood flow from the return
280 A. Tung and T. H. Song

cannula to the artery distal to the cannula insertion site (usually the superficial fem-
oral artery). The incidence of lower extremity ischemia with femoral cannulation
historically ranges from 12% to 22% [4] and fasciotomy for compartment syndrome
or amputation occurs with a measurable frequency. Distal perfusion catheters reduce
the incidence of such complications [5].
As with VV ECMO, anticoagulation is required for VA ECMO to prevent throm-
bosis which may result in thromboembolic events, ML failure, or pump malfunc-
tion. Bleeding occurs in approximately 25% of all VA ECMO patients [6] and is
seen even in the absence of anticoagulation. Both heparin and direct thrombin inhib-
itors such as bivalirudin and argatroban have been used for systemic anticoagulation
[7]. Few data exist to identify an optimal anticoagulation strategy and it is likely that
no single approach is best (see Chap. 8).
Adequacy of anticoagulation may be monitored via a variety of strategies. For
patients on heparin, the 2014 ELSO guidelines suggest titrating to an activated clot-
ting time (ACT) = 180–220 s [8]. Anti-Xa monitoring has also been used with an
ELSO-suggested goal of 0.3–0.7 iU/mL [8]. Prothrombin time (PTT) ranges
between 50 and 80 s are also used in some centers. Divergence between measures
of coagulation is common, and clinical experience suggests that, after prolonged
ECMO runs, target Anti-Xa levels may be associated with supratherapeutic PTT
levels [9]. Although thromboelastography (TEG) may permit more precise titration
of the degree of anticoagulation, existing data suggest a poor correlation between
point-of-care TEG testing and bleeding and thrombotic complications [10]. As with
any therapeutic titration, evidence of bleeding or clotting should prompt a review of
anticoagulation management.

Monitoring

VA ECMO differs from VV ECMO in affecting cardiac output and end-organ perfu-
sion in addition to gas exchange. As a result, monitoring goals for VA ECMO
include adequacy of limb perfusion distal to the (usually femoral) return cannula,
function of the native heart and lungs, adequacy of upper body oxygenation (for
femoral cannulation), and effectiveness of overall cardiorespiratory support in addi-
tion to circuit, cannula, and ML function and integrity. In addition, weaning from
VA ECMO differs from weaning from VV ECMO because both hemodynamic and
respiratory support are being weaned.

Adequacy of Oxygenation via the ECMO Circuit

Because gas exchange in patients on VA ECMO occurs via the ECMO circuit,
working in parallel with the patient’s native heart and lungs, monitoring adequacy
of oxygenation via the ECMO circuit may be needed to troubleshoot inadequate
systemic O2 saturation. Factors affecting the PO2 at the ML outlet include blender
16 Venoarterial ECMO in Respiratory Failure 281

concentration, circuit blood flow rate, PO2 at the oxygenator intake, and amount of
microthrombi in the ML. In general, if blood flow through the ML is below the rated
flow for the device and the inflow saturation is 70% or higher, the saturation at the
ML outlet should be >95% [11]. Because many factors affect post-membrane oxy-
genation, a reasonable clinical strategy for monitoring ML function is to obtain a
baseline post-ML PO2 5–10 min after initiation and monitor trends over time (see
Chap. 9).

Adequacy of Overall Gas Exchange (ECMO and Native Circulation)

Unlike VV ECMO, VA ECMO affects oxygen delivery not just by improving arte-
rial oxygen saturation but also by augmenting systemic cardiac output to deliver
oxygenated blood to peripheral organs. Because systemic blood flow in patients on
VA ECMO involves the output of both the native circulation and the ECMO system
itself, directly monitoring systemic cardiac output (and thus oxygen delivery)
requires monitoring both native cardiac output and ECMO flow. As a rough rule of
thumb, a normal blood pressure and systemic SvO2 (>70%) suggest adequate oxy-
gen delivery to peripheral organs.
Because overall oxygenation represents a combination of ECMO and native lung
function, diagnosing the cause of changes in systemic oxygenation can be challeng-
ing. Native lung function, ML function, or differences in flow delivery between the
ML and the native circulation may all play a role. Patients with respiratory failure
who require VA ECMO may also have septic or cardiogenic shock, and as a result,
the amount of native cardiac output passing through the lungs and heart (instead of
the ECMO device) may be higher or lower.
Estimating the cardiac output passing through the native heart and lungs is dif-
ficult. Arterial pulsatility is an imperfect monitor for left ventricle (LV) ejection but
suggests at least some intrinsic cardiac output. Transthoracic echocardiography
allows more accurate estimates of stroke volume by measuring the left ventricular
outflow tract velocity-time integral. Perhaps the most straightforward monitoring
approach is pulmonary artery (PA) catheterization, which permits direct measure-
ment of intracardiac flow.
The interaction between native and ECMO flows in a VA ECMO circuit can also
have consequences on cardiac function. In principle, increasing ECMO flow would
augment venous drainage from the right atrium and decrease the amount of blood
flow passing through the native heart/lungs. In practice, however, the increased LV
afterload due to increased ECMO pump output may also limit LV ejection and con-
tribute to LV distention. In patients with cardiogenic shock and LV dysfunction,
such distention may slow cardiac recovery [12]. In addition, LV distention can
increase left atrial pressure, potentially worsening pulmonary edema (and oxygen-
ation via the native lungs).
Several strategies for avoiding, detecting, and monitoring LV distention have
been suggested. These include echocardiography to assess for aortic valve opening,
LV size, and stasis/clot formation; chest radiography to detect worsening
282 A. Tung and T. H. Song

pulmonary edema; and pulmonary artery catheterization to measure the pulmonary

artery diastolic pressure (or occlusion pressure). Of these, PA catheterization is
increasingly considered the most accurate approach and emerging guidelines for the
management of cardiogenic shock suggest benefit from doing so [13, 14]. By moni-
toring pulmonary artery occlusion pressure and native cardiac output, PA catheter-
ization allows the greatest clarity with respect to the distribution of blood flow
between native and ECMO circuits.
PA catheterization also facilitates fluid management of patients on VA
ECMO. Patients who require initiation of VA ECMO support are often hypotensive
and receiving vasopressor support. Appropriately titrating intravascular volume and
vasopressor dose to achieve adequate blood pressure without LV distention is best
achieved using hemodynamic monitoring with a PA catheter.
A complete discussion of LV venting strategies for LV distention is beyond the
scope of this chapter and a 2018 review provides a referenced summary [15].
Inotropes and vasodilators may permit greater ejection from the LV but at potential
cost to blood pressure and LV oxygen supply-demand balance. ECMO flow can be
lowered to reduce afterload. Catheters may be introduced via the LV apex, retro-
grade via the aortic valve, and from the right heart via the interatrial septum to
actively drain blood from the LV.

Adequacy of Upper Body Oxygenation

When lung failure is present and the VA ECMO return cannula is in the femoral
artery, deoxygenated blood from the native circulation may enter subclavian and
carotid arteries before mixing with oxygenated blood traveling retrograde from the
distal aorta. The extent to which desaturated blood from the native circulation enters
the cerebral vessels depends on the balance between ECMO and native flows. Due
to dynamic changes in the ratio of native to ECMO blood flow, the “watershed” area
may change as cardiac function improves.
If oxygenated blood from the femoral ECMO return cannula does not reach the
upper body, the syndrome is variously termed “differential hypoxia,” “North-­
South,” or “Harlequin” syndrome. Relative or absolute cerebral hypoxemia may
result, so patients on VA ECMO with a femoral return cannula require monitoring
[16]. Pulse oximetry, arterial waveform, and blood gas monitoring at the right radial
location will allow continuous assessment of oxygen saturation at the right brachio-
cephalic artery and hence the right common carotid artery. A large right radial pulse
pressure suggests potentially significant native ejection (of deoxygenated blood in
patients with lung failure). Blood gas monitoring from the right radial artery will
enable more precise monitoring of oxygen saturation but can only be performed
intermittently.
Strategies to ameliorate the risk of upper body desaturation due to Harlequin
syndrome include optimization of native lung function and increasing ECMO flow
to push the watershed zone proximal to the aortic arch [15]. Increasing ECMO flow
may also worsen LV distention as described above, potentially worsening
16 Venoarterial ECMO in Respiratory Failure 283

pulmonary edema. Reducing native cardiac ejection, either by targeting a lower

preload or less ejection, may also allow oxygenated ECMO blood to flow retrograde
to the proximal aorta. An alternative approach is to divert some oxygenated ECMO
blood to the superior vena cava to increase oxygenation in the native circulation.
This strategy, described as veno-venoarterial (V-VA) ECMO, involves placing a
return cannula in the internal jugular vein to provide oxygenated blood to the heart
and lungs. The balance between flow through the femoral and internal jugular can-
nulas can be adjusted via clamps on the cannula tubing.

Distal Limb Ischemia

Unlike VV ECMO, VA ECMO requires that a return cannula be placed in an artery.

When the femoral site is used, the incidence of limb ischemia can range as high as
20%. With current ECMO practice, a 6- to 8-F distal perfusor is often used to divert
some ECMO flow to the distal limb. Although this approach dramatically reduces
the incidence of limb ischemia [5], lower limb ischemia requiring fasciotomy or
amputation may still occur, so monitoring of the distal limb is needed. In addition
to clinical examination, Doppler examination of the foot may also allow some
assessment of distal limb flow although pulses are often absent due to continuous
flow from the distal perfusor. Elevated systemic lactate levels may signal inadequate
limb perfusion. Adequacy of flow through the distal perfusor, particularly with low
levels of anticoagulation or low ECMO flows, is also a concern as low flows may
predispose to thrombosis of the distal perfusor.

Weaning

Unlike VV ECMO, which is initiated for hypoxia or hypercapnia and can be weaned
by reducing the sweep gas flow to zero, VA ECMO requires that the ML be kept on
to prevent returning deoxygenated blood to the arterial system. Weaning from VA
ECMO thus involves reducing the ECMO circuit flow while monitoring systemic
perfusion and pressure (see Chap. 15). During VA ECMO weaning, oxygen satura-
tion may not change, or not be helpful for guiding the weaning process. Because
most uses of VA ECMO are for cardiogenic shock with adequate lung function,
oxygenation is usually not as much of a concern. But when VA ECMO is used for
respiratory failure, a reduction in ECMO flow may produce desaturation. Since the
ECMO circuit cannot be completely stopped due to thrombosis concerns, VA
ECMO is generally weaned by reducing the flow to a lower limit (1–2 LPM) while
monitoring oxygen saturation and hemodynamic stability. Adequacy of native lung
function to support oxygenation unassisted should be verified before decannulation.
In most cases, by the time the patient is considered ready for decannulation, hemo-
dynamic instability has resolved, the native heart is ejecting, and ECMO flow is
contributing little oxygenated blood to the upper body where saturation is usually
monitored.
284 A. Tung and T. H. Song

Clinical Indications for VA ECMO for Respiratory Failure

In general, indications for VA ECMO in patients with respiratory failure can be

grouped into two categories. The first involves a separate indication for VA ECMO
in addition to VV support. Examples include respiratory failure with concomitant
vasoplegic shock; respiratory failure with cardiogenic shock due to failure of the
RV, LV, or both; or respiratory failure with other etiologies for insufficient cardiac
output such as intracardiac shunt, arrhythmias, pulmonary embolus, or elevated pul-
monary vascular resistance.
The second category includes patients with persistent (cerebral) hypoxemia
despite maximal VV ECMO support. This section will briefly describe the use of
VA ECMO for these conditions and suggest what clinical cues in patients with
ARDS may suggest the use of VA versus VV ECMO and what additional evaluation
may be indicated before choosing either VV or VA modalities.

Vasoplegic Shock

In principle, the ability of VA ECMO to augment cardiac output in addition to oxy-

genation suggests a role in respiratory failure with associated septic shock. However,
several factors may limit any survival advantage with VA ECMO. Increased diffi-
culty achieving source control with ECMO cannulas in place, inflammation from
exposure to the ECMO circuit, inability to generate sufficient additional flow to
overcome systemic vasoplegia, and progressive end-organ dysfunction unrelated to
hemodynamic instability may all prevent VA ECMO from improving survival in
critically ill patients who require both hemodynamic and respiratory support.
Indications for choosing VA over VV ECMO for septic or vasoplegic shock associ-
ated with respiratory failure are not well established. Existing data suggest that VV
ECMO may reduce vasopressor requirements in hemodynamically unstable patients
with ARDS and concurrent vasopressor use [17]. As a result, patients with normal
cardiac function and low vasopressor requirements may not always require
VA ECMO.
The benefit of VA versus VV ECMO for ARDS with septic shock is unclear.
Current literature consists mostly of case reports and small observational series
[18]. Outcome studies are mixed. One found much worse outcomes when compared
to a historical series of patients with cardiogenic shock [19]. On the other hand, a
2013 case series of 14 patients suggested “rescue” rates of 70% in a population with
normal baseline LV function, bacterial septic shock, and severe septic cardiomyopa-
thy [20]. Another case series reported 90% survival for VA ECMO in septic shock
with LV dysfunction and 65% survival for VV ECMO for distributive shock alone
[21]. Yet another trial compared 82 patients with septic shock to 130 matched con-
trols and found a mortality risk ratio of 0.54 for patients receiving VA ECMO versus
no ECMO [22]. Despite this trend towards benefit with VA ECMO for septic shock,
16 Venoarterial ECMO in Respiratory Failure 285

current Surviving Sepsis Guidelines make a weak recommendation to use VV

ECMO when mechanical ventilation fails, but do not make a recommendation
regarding use of VA ECMO [23]. Factors such as timing (early vs late initiation),
type and location of septic source, and baseline organ dysfunction likely play an
important role in moderating any survival benefit. In light of these data, VA ECMO
may be considered a rescue strategy for patients with septic shock and associated
cardiomyopathy.
Cues for instituting VA ECMO in ARDS patients with septic or vasoplegic shock
include a history of cardiac dysfunction (ischemic coronary disease, valvular dys-
function, low ejection fraction, arrhythmias), high vasopressor requirements, dete-
riorating clinical status, and refractory hypotension. In larger patients, VA ECMO
may not be able to generate sufficient additional flow to overcome the low systemic
vascular resistance.

Respiratory Failure with Cardiogenic Shock

Since VV ECMO does not support the circulation, respiratory failure complicated
by severe LV or RV dysfunction may require VA ECMO. When considering VV
ECMO for patients with hemodynamic instability, the circulation should be assessed
comprehensively, generally including echocardiography. Indications for VA instead
of VV ECMO include hemodynamic instability despite vasoactive infusion, preex-
isting right or left heart failure, elevated lactate or low SvO2 despite maximal medi-
cal therapy, valvular dysfunction preventing adequate cardiac output, or intractable
arrhythmias.
LV Dysfunction The most common indication for VA ECMO is severe LV failure.
In such cases and especially after cardiac arrest, respiratory failure may follow car-
diogenic shock due to elevations in LV filling pressure and consequent pulmonary
edema. For those patients, cardiac dysfunction is the primary reason for VA ECMO
support. Alternatively, respiratory failure precedes hemodynamic failure, as when
stress (takotsubo) cardiomyopathy complicates ARDS or when severe hypoxemia
produces a stunned myocardium.
By supporting both lung and LV function, VA ECMO may restore myocardial
oxygen supply-demand balance in patients with ischemic cardiomyopathy or buy
time for recovery when LV dysfunction is reversible. Other causes of cardiogenic
shock that respond well to VA ECMO but are beyond the scope of this chapter
include fulminant myocarditis [24] and refractory arrhythmias [25]. As noted above,
increased left ventricular loading with high ECMO flow rates may hamper myocar-
dial recovery [26] so attention should be paid to the degree of LV distention during
VA ECMO.
RV Dysfunction Severe hypoxemic respiratory failure may itself induce a suffi-
cient increase in pulmonary vascular resistance to cause RV dysfunction or failure.
Clinical clues include elevated right atrial pressure, tricuspid regurgitation, or echo-
286 A. Tung and T. H. Song

cardiographic signs (see Chap. 7). A decision to initiate VV ECMO in such patients
initially can be reasonable as existing data suggests that right ventricular function
may improve with VV support alone [17, 27]. However, if RV dysfunction is pro-
found, VA ECMO may be a better choice. Similarly, if RV function deteriorates on
VV ECMO, vasopressor requirements escalate, or hypoxemia persists after initia-
tion, then converting to VVA ECMO may stabilize cardiorespiratory function while
the lung and heart recover. Such support may allow time to optimize mechanical
ventilation, restore metabolic homeostasis, and permit the right ventricle to recover.
Evidence to support VA ECMO for RV support in patients with ARDS remains
at the level of case reports [28, 29]. A small trial compared VV, VA, and VVA ECMO
and found no clear difference in mortality. However, the incidence of RV dysfunc-
tion in ARDS is as high as 25% [30] and associated with greater mortality [31]. It
remains unclear whether the increased complexity of dual return cannulas and arte-
rial cannulation offsets the physiological benefit of improved oxygenation and sys-
temic blood flow. More recently, a specialized, dual-lumen cannula that returns flow
directly into the pulmonary artery has been used to support both lung and RV func-
tion. However, VA and VVA ECMO remain viable options in certain situations.
Indications for adding an arterial return cannula to VV ECMO or converting to
VA ECMO for cardiogenic shock due to associated right heart failure include dete-
riorating hemodynamic status while on VV ECMO, increasing lactate levels despite
maximal vasopressor therapy, worsening or persistently severe hypoxemia, and
echocardiographic evidence of severe RV dysfunction (RV dilation, decreased RV
contractility, new or worsening tricuspid regurgitation).
Acute pulmonary embolism is another cause of severe pulmonary and cardiovas-
cular dysfunction which may respond to VA ECMO. Such patients are often hypox-
emic due to acute lung failure and in cardiogenic shock due to acute increases in RV
afterload. While initial treatment with thrombolytic therapy lowers mortality in
hemodynamically unstable patients, VA ECMO may improve survival in those who
remain in cardiorespiratory failure after thrombolytic therapy or in whom throm-
bolysis is contraindicated. One case series described 95% survival-to-discharge in
patients receiving VA ECMO as a primary management strategy for PE with cardio-
genic shock and an 80% survival in patients presenting with cardiac arrest [27].
These results likely represent an upper bound on survival as other case series report
worse outcomes [32, 33]. As with ECMO for sepsis, survival rates vary widely
between studies and a clear understanding of what factors predict successful out-
come remains elusive. As an example, an elevated lactate level on presentation has
been reported as a risk factor for poor outcome in some cohorts [33] but not others
[27]. Regardless of predictors, an organized ECMO protocol is likely needed to
optimize outcomes with VA ECMO for hemodynamically unstable patients with PE
as they may rapidly evolve to cardiac arrest.
In general, PE sufficiently severe to cause hypoxemia usually also causes right
ventricular dysfunction. As a result and because the clinical course of PE is often
uncertain, VA ECMO is more commonly chosen when extracorporeal support is
needed. Indications for choosing VA ECMO for patients with PE include hemody-
namic instability; ready availability of ECMO support; failure of thrombolytic
16 Venoarterial ECMO in Respiratory Failure 287

therapy to reverse hemodynamic instability; high or increasing doses of vasoactive

drugs; preexisting right or left heart failure, valvular disease, or pulmonary hyper-
tension; and deteriorating clinical status, elevated lactate levels, or low SvO2 despite
maximal medical therapy.

 atients with Inadequate Oxygenation Despite Maximal VV

P
ECMO Support

The last category of patients with hypoxemic respiratory failure who may benefit
from VA ECMO are those who remain persistently hypoxemic despite maximal VV
support. Since correction of hypoxemia using VV ECMO relies on capturing a large
fraction of cardiac output (see Chap. 1), this scenario is usually seen in high output
states. Hypoxia and hypercarbia can contribute to agitation in patients on ECMO,
impeding their ability to participate in physical therapy. ELSO guidelines define
adequate oxygen support as an arterial oxygen saturation > 80% and adequate CO2
removal as ability to remove the CO2 produced by consuming oxygen at a 3 cc/kg/
min rate [11].
Reasons for persistent hypoxemia or inadequate CO2 clearance during VV
ECMO include ML malfunction, excessive recirculation, and inadequate circuit
blood flow. Of these, the former is easily diagnosed and treated (Chap. 9). Managing
recirculation may not be as straightforward. Some degree of recirculation is often
present with VV ECMO. Factors that increase the likelihood of recirculation include
close positioning of the SVC and IVC cannulas, single (vs multistage)-lumen drain-
age cannulas [34], and increased ECMO flow rates [35]. Circuit blood flow may rise
with fluid loading, cannula repositioning, or exchange for larger cannulas, but when
cardiac output is high, it may simply be impossible to achieve flows sufficient for
adequate oxygenation.
Inadequate oxygen saturation despite optimal VV ECMO support is difficult to
treat. Even if hemoglobin levels are increased to allow adequate oxygen delivery, an
arterial oxygen saturation < 90% can lead to persistent agitation requiring heavy
sedation. Although ELSO guidelines define adequate oxygen support as an arterial
oxygen saturation > 80% [11], in clinical practice such patients almost always
require heavy sedation to control persistent agitation.
Conversion to VA ECMO with femoral cannulation invites a severe harlequin
syndrome as such patients often have elevated cardiac outputs. Central cannulation
is extremely invasive. A third potential solution is to insert an arterial return cannula
in the right axillary artery. Oxygenated blood from the ECMO circuit then enters the
right common carotid artery, ensuring that blood entering the cerebral vessels is
adequately oxygenated. Generally, 15- to 19-F cannulas are used for this purpose
and can be placed percutaneously or surgically with a graft. Small case series [36,
37] suggest that full VA ECMO support is feasible with this approach although
complications include arm swelling, hematoma, and infection [37]. With an ambu-
latory VA configuration (drainage usually from the internal jugular vein and return
288 A. Tung and T. H. Song

via axillary artery cannulation), a venous drainage limb is not necessary although
may be useful to prevent reperfusion injury to the lung.
Indications for adding an arterial return cannula to existing VV ECMO or con-
verting to ambulatory VA ECMO include persistent hypoxemia/hypercarbia, agita-
tion due to inadequate arterial oxygen saturation, poor response to cannula
repositioning or increasing ECMO flows, and adequacy of vascular access.

Conclusion

Although the most common indication for VA ECMO is to maintain hemodynamic

stability in patients with cardiogenic shock, it is occasionally needed to support
patients with primary lung failure. The spectrum of lung failure states potentially
treatable by VA ECMO includes ARDS with septic shock, respiratory failure with
cardiogenic shock (either concurrently or as a result), pulmonary hypertension with
right ventricular failure due to pulmonary embolus or to ARDS, and in situations
when VV ECMO is unable to adequately oxygenate patients with severe hypox-
emia. The addition of an arterial cannula may introduce monitoring needs not
needed during VV ECMO alone, including for distal limb ischemia, LV distention,
and harlequin syndrome. Weaning from VA ECMO also adds considerations since
the ECMO circuit remains oxygenated while flow is reduced and the ability of
native lung function to support oxygenation may not be fully assessable prior to
decannulation.
The use of VA ECMO for respiratory failure remains a dynamic, evolving field
and little consensus regarding indications, management, conversion from VV to
VVA/VA ECMO, or weaning exists. Regardless, it is likely that a clinically relevant
role for VA ECMO in respiratory failure will exist and grow as the boundaries of
ECMO support expand.

References

1. ECLS Registry Report, October 2021. Extracorporeal Life Support Organization (ELSO).
Available at: https://www.elso.org/Registry/Statistics/InternationalSummary.aspx. Accessed
11/4/2021.
2. Kon ZN, Bittle GJ, Pasrija C, Pham SM, Mazzeffi MA, Herr DL, Sanchez PG, Griffith
BP. Venovenous versus venoarterial extracorporeal membrane oxygenation for adult patients
with acute respiratory distress syndrome requiring precannulation hemodynamic support: a
review of the ELSO registry. Ann Thorac Surg. 2017;104:645–9.
3. Abrams D, Bacchetta M, Brodie D. Recirculation in venovenous extracorporeal membrane
oxygenation. ASAIO J. 2015;61:115–21.
4. Cheng R, Hachamovitch R, Kittleson M, Patel J, Arabia F, Moriguchi J, Esmailian F, Azarbal
B. Complications of extracorporeal membrane oxygenation for treatment of cardiogenic shock
and cardiac arrest: a meta-analysis of 1,866 adult patients. Ann Thorac Surg. 2014;97:610–6.
16 Venoarterial ECMO in Respiratory Failure 289

5. Juo YY, Skancke M, Sanaiha Y, Mantha A, Jimenez JC, Benharash P. Efficacy of distal perfu-
sion cannulae in preventing limb ischemia during extracorporeal membrane oxygenation: a
systematic review and meta-analysis. Artif Organs. 2017;41:E263–73.
6. Sy E, Sklar MC, Lequier L, Fan E, Kanji HD. Anticoagulation practices and the prevalence
of major bleeding, thromboembolic events, and mortality in venoarterial extracorporeal mem-
brane oxygenation: A systematic review and meta-analysis. J Crit Care. 2017;39:87–96.
7. Kaseer H, Soto-Arenall M, Sanghavi D, Moss J, Ratzlaff R, Pham S, Guru P. Heparin vs
bivalirudin anticoagulation for extracorporeal membrane oxygenation. J Card Surg.
2020;35(4):779–86.
8. ELSO Anticoagulation Guideline. Extracorporeal Life Support Organization (ELSO).
Available at: http://www.elso.org/Portals/0/Files/elsoanticoagulationguideline8-­2014-­table-­
contents.pdf. Accessed 11/4/2021.
9. Adatya S, Sunny R, Fitzpatrick MJ, Colvin M, Thennapan T, John R, Dodge Zantek N, Pritzker
M, Eckman P, Uriel N. Coagulation factor abnormalities related to discordance between anti-­
factor Xa and activated partial thromboplastin time in patients supported with continuous-flow
left ventricular assist devices. J Heart Lung Transplant. 2016;35:1311–20.
10. Jiritano F, Fina D, Lorusso R, Ten Cate H, Kowalewski M, Matteucci M, Serra R, Mastroroberto
P, Serraino GF. Systematic review and meta-analysis of the clinical effectiveness of point-of-­
care testing for anticoagulation management during ECMO. J Clin Anesth. 2021;73:110330.
11. General guidelines for all ECLS cases, August 2017. Extracorporeal Life Support Organization
(ELSO). Available at: https://www.elso.org/Portals/0/ELSO%20Guidelines%20General%20
All%20ECLS%20Version%201_4.pdf. Accessed 11/4/2021.
12. Kurihara H, Kitamura M, Shibuya M, Tsuda Y, Endo M, Koyangi H. Effect of transaortic catheter
venting on left ventricular function during venoarterial bypass. ASAIO J. 1997;43:M838–41.
13. Rihal CS, Naidu SS, Givertz MM, Szeto WY, Burke JA, Kapur NK, Kern M, Garratt KN,
Goldstein JA, Dimas V, Tu T, Society for Cardiovascular Angiography and Interventions
(SCAI), Heart Failure Society of America (HFSA), Society of Thoracic Surgeons (STS),
American Heart Association (AHA), and American College of Cardiology (ACC). 2015
SCAI/ACC/HFSA/STS Clinical Expert Consensus Statement on the Use of Percutaneous
Mechanical Circulatory Support Devices in Cardiovascular Care: Endorsed by the American
Heart Association, the Cardiological Society of India, and Sociedad Latino Americana
de Cardiología Intervencionista; Affirmation of Value by the Canadian Association of
Interventional Cardiology-Association Canadienne de Cardiologie d’intervention. J Am Coll
Cardiol. 2015;65:2140–1.
14. Ranka S, Mastoris I, Kapur NK, Tedford RJ, Rali A, Acharya P, Weidling R, Goyal A, Sauer
AJ, Gupta B, Haglund N, Gupta K, Fang JC, Lindenfeld J, Shah Z. Right heart catheteriza-
tion in cardiogenic shock is associated with improved outcomes: insights from the nationwide
readmissions database. J Am Heart Assoc. 2021;10:e019843.
15. Rao P, Khalpey Z, Smith R, Burkhoff D, Kociol RD. Venoarterial extracorporeal membrane
oxygenation for cardiogenic shock and cardiac arrest. Circ Heart Fail. 2018;11:e004905.
16. Hoeper MM, Tudorache I, Kühn C, Marsch G, Hartung D, Wiesner O, Boenisch O, Haverich A,
Hinrichs J. Extracorporeal membrane oxygenation watershed. Circulation. 2014;130:864–5.
17. Gutsche JT, Mikkelsen ME, McCarthy FH, Miano TA, Vernick WJ, Ramakrishna H, Patel
PA, Augoustides Y, Szeto WY, Desai ND, Lane-Fall MB, Williams ML. Veno-venous extra-
corporeal life support in hemodynamically unstable patients with ARDS. Anesth Analg.
2017;124:846–8.
18. Buia A, Hopf HB, Herrmann E, Schmandra T, Hanisch E. Septic shock: ECMO beyond
ARDS? Introducing the Simon two-stage protocol when randomisation is considered unethi-
cal. Scand J Trauma Resusc Emerg Med. 2020;28:22.
19. Ro SK, Kim WK, Lim JY, Yoo JS, Hong SB, Kim JB. Extracorporeal life support for adults
with refractory septic shock. J Thorac Cardiovasc Surg. 2018;156:1104–9.
290 A. Tung and T. H. Song

20. Bréchot N, Luyt CE, Schmidt M, Leprince P, Trouillet JL, Léger P, Pavie A, Chastre J, Combes
A. Venoarterial extracorporeal membrane oxygenation support for refractory cardiovascular
dysfunction during severe bacterial septic shock. Crit Care Med. 2013;41:1616–26.
21. Falk L, Hultman J, Broman LM. Extracorporeal membrane oxygenation for septic shock. Crit
Care Med. 2019;47:1097–105.
22. Bréchot N, Hajage D, Kimmoun A, Demiselle J, Agerstrand C, Montero S, Schmidt M, Luyt
CE, Lebreton G, Hékimian G, Flecher E, Zogheib E, Levy B, Slutsky AS, Brodie D, Asfar P,
Combes A, International ECMO Network. Venoarterial extracorporeal membrane oxygenation
to rescue sepsis-induced cardiogenic shock: a retrospective, multicentre, international cohort
study. Lancet. 2020;396(10250):545–52.
23. Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et.al. Surviving
sepsis campaign: international guidelines for management of sepsis and septic shock 2021.
Intensive Care Med 2021; 47:1181–1247.
24. Diddle JW, Almodovar MC, Rajagopal SK, Rycus PT, Thiagarajan RR. Extracorporeal
membrane oxygenation for the support of adults with acute myocarditis. Crit Care Med.
2015;43:1016–25.
25. Bhandary SP, Joseph N, Hofmann JP, Saranteas T, Papadimos TJ. Extracorporeal life support
for refractory ventricular tachycardia. Ann Transl Med. 2017;5:73.
26. Schrage B, Becher PM, Bernhardt A, Bezerra H, Blankenberg S, Brunner S, Colson P, Cudemus
Deseda G, Dabboura S, Eckner D, Eden M, Eitel I, Frank D, Frey N, Funamoto M, Goßling
A, Graf T, Hagl C, Kirchhof P, Kupka D, Landmesser U, Lipinski J, Lopes M, Majunke N,
Maniuc O, McGrath D, Möbius-Winkler S, Morrow DA, Mourad M, Noel C, Nordbeck P,
Orban M, Pappalardo F, Patel SM, Pauschinger M, Pazzanese V, Reichenspurner H, Sandri
M, Schulze PC, Schwinger RHG, Sinning JM, Aksoy A, Skurk C, Szczanowicz L, Thiele H,
Tietz F, Varshney A, Wechsler L, Westermann D. Left ventricular unloading is associated with
lower mortality in patients with cardiogenic shock treated with venoarterial extracorporeal
membrane oxygenation: results from an International, Multicenter Cohort study. Circulation.
2020;142:2095–106.
27. Pasrija C, Kon Z. A protocolized approach to veno-arterial extracorporeal membrane oxygen-
ation for massive pulmonary embolism. Resuscitation. 2018;126:e12. Epub 2018 Jan 31
28. Sauza-Sosa JC, Rojas-Gomez CA, Montes de Oca-Sandoval MA, Hernandez-Rendon E,
Lima-Linares R, Camarena-Alejo G. Acute cor pulmonale in a critically ill patient with
COVID-19 resolved with venoarteriovenous extracorporeal membrane oxygenation. CASE
(Phila). 2021;5:138–42.
29. Bergman ZR, Prathibha S, Bauman BD, Yannopoulos D, Brunsvold ME. Venoarteriovenous
ECMO in concomitant acute respiratory distress syndrome and cardiomyopathy associated
with COVID-19 infection. Case Rep Crit Care. 2021;2021:8848013.
30. Mekontso Dessap A, Boissier F, Charron C, Bégot E, Repessé X, Legras A, Brun-Buisson
C, Vignon P, Vieillard-Baron A. Acute cor pulmonale during protective ventilation for acute
respiratory distress syndrome: prevalence, predictors, and clinical impact. Intensive Care Med.
2016;42:862–70.
31. Boissier F, Katsahian S, Razazi K, Thille AW, Roche-Campo F, Leon R, Vivier E, Brochard L,
Vieillard-Baron A, Brun-Buisson C, Mekontso DA. Prevalence and prognosis of cor pulmo-
nale during protective ventilation for acute respiratory distress syndrome. Intensive Care Med.
2013;39:1725–33.
32. Meneveau N, Guillon B, Planquette B, Piton G, Kimmoun A, Gaide-Chevronnay L, Aissaoui
N, Neuschwander A, Zogheib E, Dupont H, Pili-Floury S, Ecarnot F, Schiele F, Deye N, de
Prost N, Favory R, Girard P, Cristinar M, Ferré A, Meyer G, Capellier G, Sanchez O. Outcomes
after extracorporeal membrane oxygenation for the treatment of high-risk pulmonary embo-
lism: a multicentre series of 52 cases. Eur Heart J. 2018;39:4196–204.
33. George B, Parazino M, Omar HR, Davis G, Guglin M, Gurley J, Smyth S. A retrospective
comparison of survivors and non-survivors of massive pulmonary embolism receiving veno-­
arterial extracorporeal membrane oxygenation support. Resuscitation. 2018;122:1–5.
16 Venoarterial ECMO in Respiratory Failure 291

34. Palmér O, Palmér K, Hultman J, Broman M. Cannula design and recirculation during venove-
nous extracorporeal membrane oxygenation. ASAIO J. 2016;62:737–42.
35. van Heijst AF, van der Staak FH, de Haan AF, Liem KD, Festen C, Geven WB, van de Bor
M. Recirculation in double lumen catheter veno-venous extracorporeal membrane oxygen-
ation measured by an ultrasound dilution technique. ASAIO J. 2001;47:372–6.
36. Javidfar J, Brodie D, Costa J, Miller J, Jurrado J, LaVelle M, Newmark A, Takayama H, Sonett
JR, Bacchetta M. Subclavian artery cannulation for venoarterial extracorporeal membrane
oxygenation. ASAIO J. 2012;58:494–8.
37. Biscotti M, Bacchetta M. The “sport model”: extracorporeal membrane oxygenation using the
subclavian artery. Ann Thorac Surg. 2014;98:1487–9.
Chapter 17
Ethical Challenges in Extracorporeal
Membrane Oxygenation

Elizabeth Sonntag and Meera Pahuja

Introduction

The use of extracorporeal membrane oxygenation (ECMO) is fraught with ethical

challenges – most of which resemble those that critical care physicians have been
facing for generations. However, unlike other forms of life-sustaining therapy,
VV-ECMO’s possible trajectories are limited to recovery or lung transplantation
[1]. The cost of ECMO can be exorbitant and clinical outcomes remain largely
uncertain. Ethical practice compels providers to act in patient’s best interest; how-
ever, the uncertainty in outcomes with VV-ECMO makes applying best interest
standards ambiguous [2].
This results in significant center-to-center variation in how ECMO is applied,
and different centers may have varied criteria for who are considered ECMO candi-
dates [3]. Furthermore, given the emergent setting in which ECMO is initiated, time
constraints limit information-gathering, making anticipatory guidance extremely
difficult. Common ethical issues unique to ECMO care include determining ECMO
initiation candidacy given limits of prognostication; balancing benefits and burdens
during continuation of ECMO; decisions about when to withdraw ECMO therapy in
the face of “bridge to nowhere” scenarios, medical complications, or patient suffer-
ing; and, finally, resource allocation [4, 5]. The ethical concepts of beneficence,
nonmaleficence, autonomy, justice, shared decision-making, surrogate decision-­
making, decisional capacity, advance care planning, substituted judgment, time-­
limited trials, futile care, and moral distress are all inherent to the care of patients on
VV-ECMO and are further described in Table 17.1.

E. Sonntag (*) · M. Pahuja

Department of Internal Medicine, Virginia Commonwealth University School of Medicine,
Richmond, VA, USA
e-mail: [email protected]; [email protected]

© The Author(s), under exclusive license to Springer Nature 293

Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6_17
294 E. Sonntag and M. Pahuja

Table 17.1 This table defines and describes common ethical principles found throughout this
chapter [5, 6, 8, 21, 22]
Term Definition Comment
Beneficence Acting in the best interest of
the patient; doing good
Nonmaleficence Avoiding harm to the patient
Autonomy A patient’s right to choose A patient must have decision-making
what medical recommendations capacity to exercise their autonomy.
they will accept or reject. A patient cannot demand a medical
intervention that has not been offered, or is
inappropriate or non-indicated.
Justice Equal distribution of goods
amongst a society.
Fairness.
Shared When a physician and patient Proper shared decision-making involves a
decision-making work together to make health physician making recommendations based
care decisions. on clinical knowledge and evidence-based
medicine and a patient sharing their values
and preferences so that a decision can be
made together regarding what is in the
patient’s best interest.
This requires a physician to respect
diversity in patient values, attitudes,
cultures, religions, etc.
Surrogate When a physician and patient Typically, a surrogate decision-maker is
decision-making representative work together to asked to base decisions on the patient’s
make health care decisions for known values and preferences.
a patient who does not have
capacity.
Decisional A patient’s ability to Decisional capacity is determined by
capacity understand the medical provider judgment. In general, a provider
situation and reason through will evaluate whether a patient is thinking
available options in order to clearly, displaying consistency, and making
consent to or refuse a particular rational decisions.
medical treatment. Decisional capacity exists on a spectrum,
that is to say, a different level of capacity
may be required for a simple, low-risk
procedure (i.e., central line insertion), than
for a complex, higher-risk procedure (i.e.,
VV-ECMO).
Advance care A patient expressing their Advance care planning encompasses a wide
planning wishes for medical care should variety of conversations about medical care
they become seriously ill. which may range from a conversation with
a health care provider to a legal
documentation of wishes.
Substituted When a surrogate decision-­ The patient’s values and preferences may
judgement maker uses a patient’s values have been explicitly expressed; however,
and preferences rather than often they must be inferred by the surrogate
their own to make decisions based on known characteristics of the
about medical care. patient.
17 Ethical Challenges in Extracorporeal Membrane Oxygenation 295

Table 17.1 (continued)

Term Definition Comment
Time-limited A therapy is applied and then
trial subsequently reevaluated for
appropriateness at
predetermined time intervals.
Futile care A therapy that cannot achieve a Continuing VV-ECMO in a patient who is
desired goal. on a “bridge to nowhere” (is not a
transplant candidate and is unlikely to
recover) is often considered futile.
Moral distress The emotional response one Moral distress may occur in cases where
may experience when they feel health care team members are called to
they know the “right” thing to carry out therapy in VV-ECMO patients
do, but are compelled to do who they perceive to be suffering.
otherwise.

This chapter will specifically discuss VV-ECMO, though venoarterial ECMO

and ECMO as cardiopulmonary resuscitation (ECPR) do have their own set of ethi-
cal considerations.

Initiating ECMO

VV-ECMO is reserved for patients with refractory respiratory failure as either a

bridge to recovery or to lung transplantation. Predicting which patients will recover
is complicated. Although survival rates to discharge are well-documented by the
Extracorporeal Life Support Organization (ELSO) at 59%, there is a lack of studies
that document more detailed outcomes [2, 3]. As discussed earlier in this text, clini-
cians use a number of clinical predictors to determine candidacy for VV-ECMO. In
addition to these clinical factors, the provider must also be prepared to engage in
shared decision-making, which incorporates the patient’s values and preferences
into the decision for ECMO care. The goals of shared decision-making include
eliciting the values and preferences of the patient, guiding the patient through com-
plex medical decision-making, and understanding motivating factors that are unique
to each patient. This requires the provider to have insight into their own values and
to be able to set them aside to ensure they are not influencing recommendations.
Truly informed consent is often impossible in the setting of ECMO given the
urgency of the procedure. Therefore, initial consent should include reviewing indi-
cations for ECMO, outlining risks such as major complications and failure to
recover, and addressing criteria for ECMO termination [2]. In most cases of
VV-ECMO, the patient is incapacitated secondary to their critical illness prior to the
initiation of ECMO. Therefore, providers must rely on advance directives or surro-
gate decision-makers for consent. ECMO is not generally considered during advance
care planning so that the burden of decision-making then falls on surrogate decision-­
makers [4]. Surrogate decision-makers are tasked with relaying a patient’s values
296 E. Sonntag and M. Pahuja

and preferences and making decisions based upon them. This is called substituted
judgement [6, 7]. It is important to take into consideration that decision-making is
often being done in a clinically emergent situation, where the decision-maker may
be under a great deal of fear or stress. Clinicians must be prepared to guide decision-­
makers through the process of applying substituted judgement. For example, many
patients have not discussed with their loved ones what they would want in the many
scenarios that may come to light during an ICU stay, especially the idea of ECMO;
however, they likely have shared aspects of their life they find meaningful, valuable,
and important, such as mobility, cognitive function, spending time with family and
friends, work, and independence. Often a patient’s medical preferences can be
inferred based on the values they displayed or espoused in their day-to-day lives. A
point should be made to ensure that the quality of life that is expected with the treat-
ment plan will be acceptable to the patient. For example, being kept alive to be
discharged to a nursing home, fully reliant on others for care, may be acceptable to
one patient, but not to another. Because ECMO is often started in the acute setting,
often under a great deal of emotional stress, expectation-setting is extremely diffi-
cult. Most patients and surrogates are unfamiliar with ECMO and see it as a lifesav-
ing procedure. For this reason, once the patient is stabilized, it is extremely important
to continue conversations with the family to continue to build reasonable expecta-
tions, clarify the goals of ECMO, and discuss scenarios of recovery and
non-recovery.

Continuing ECMO

Often with ECMO initiation, knowledge of whether or not the patient will be a can-
didate for recovery or transplantation is not known until after cannulation and stabi-
lization [5]. By its inherent characteristics, ECMO is a time-limited therapy. That is
to say, if treatment goals are not met (transplant or recovery), the therapy becomes
a bridge to nowhere and, if ECMO is continued, will ultimately be met with com-
plications. This should be explained to family in detail as soon as possible. ELSO
recommends that the possibility of stopping for futility should be explained to the
family before ECMO is initiated. Furthermore, families should be prepared for the
goals of care to shift if treatment goals are not met, or in the face of complications
[1]. One important part of setting expectations with a complicated therapy like
ECMO is to meet families where they are, and to communicate big picture updates,
rather than getting lost in the details. This is best done by setting clear and specific
goals, like recovery of certain organ systems, wakefulness, and rehabilitation [1]. It
is also important to note that expectations may change throughout the course and
multiple meetings are prudent to reassess understanding and reset expectations.
There are many clinicians involved in the care of an ECMO patient, and providers
may change frequently, making consistent messaging difficult [1]. An effort should
be made to ensure that all team members, including providers, nurses, respiratory
therapists, and ECMO specialists, are on the same page about the plan of care. A
17 Ethical Challenges in Extracorporeal Membrane Oxygenation 297

useful tool to guide conversations surrounding the initiation of life-sustaining ther-

apy with ECMO is a “time-limited trial.” This is an agreement between a clinical
team and a patient (or surrogate) to use ECMO for a predetermined time interval
and assess for clinical response [8]. Providers may then use this information to
guide decision-making about continued ECMO therapy versus a transition to care
focused on comfort [9]. In 2021, an article on time-limited trials amongst critically
ill patients was published in JAMA Internal Medicine. This study showed that when
physicians were trained in communication skills, including integrating time-limited
trials to their ICU care plan, both the process measure of quality of family meetings
and the clinical outcome measure of ICU length of stay were improved [10].

Terminating ECMO

Morbidity and mortality rates increase the longer a patient is on ECMO [2]. Ongoing
ECMO use may be considered inappropriate if a patient is no longer a transplant
candidate, or if recovery is improbable or impossible. For example, if ECMO is
being used as a bridge to lung transplantation and the patient faces severe decon-
ditioning, multiorgan failure, or a difficult to eradicate infection, he may no longer
be a suitable candidate. Similarly, a patient who was originally placed on ECMO as
a bridge to recovery may suffer complications such as severe brain damage, multi-
organ failure, or permanent lung damage [2]. Given the limitations in outcome data,
it is sometimes very difficult to know whether sufficient time has been given to
assess the potential for meaningful recovery [3]. ELSO does give some recommen-
dations regarding stopping ECMO for futility. For example, ECMO “should be dis-
continued promptly if there’s no hope for a healthy survival.” Additional guidance
includes that 2 weeks of no lung function in a patient who is not a transplant candi-
date is often considered futile. They note, however, that there have been cases of
lung recovery after 50 days of ECMO. Therefore, discontinuation of ECMO for
futility should be undertaken in a case-by-case basis [2, 3].
As discussed in the section on initiating ECMO, the team should have done con-
siderable work up-front to prepare a patient and surrogates for the possibility of
terminating ECMO. In one study, decisions to withdraw ECMO were influenced by
patient comorbidities, patient wishes, and the etiology of respiratory failure (such as
pulmonary fibrosis). Unfortunately, these investigators found that survey respon-
dents reported low rates of engaging in shared decision-making in patients receiving
ECMO [11]. Understanding and applying the principles of shared decision-making,
surrogate decision-making, and substituted judgment are imperative to respecting a
patient’s wishes when they cannot speak for themselves and can also help lessen the
burden that families feel in making difficult end-of-life decisions. Treatment plans
should align not only with patients’ preferences and values but also with prognosis
and appropriateness of the care [10]. There is no ethical obligation to provide an
intervention for which there is no reasonable chance of achieving its aim. Therefore,
rather than a shared decision-making approach of when to discontinue ECMO, the
298 E. Sonntag and M. Pahuja

providers should use their clinical judgment to determine when treatment goals are
unlikely to be met and recommend discontinuation [1]. This means the provider will
need to feel comfortable making recommendations in the space of uncertainty as is
common in the application of ECMO.
The best-case scenario is that the family assents to discontinuation of ECMO at
the health care team’s recommendation. Of course, even when best practices are
used for patient selection and expectation-setting, withdrawal of any life-sustaining
therapy is often challenging for the patient, family, and providers. In many cases
surrogates may request to continue ECMO, even after the team has determined it is
no longer a bridge to recovery or transplant. In fact, in a single-center study, one
ECMO program described disagreement about continuation of ECMO as the most
common ethical issue faced in ECMO care [4].
This ethical dilemma occurs when autonomy and beneficence/nonmaleficence
are in conflict. Autonomy is the basic ethical principle that supports a patient’s right
to choose and this should be upheld whenever possible. Autonomy, however, is not
absolute. It gives a patient a right to choose between offered medical therapies and
to refuse therapies, but not to demand therapy the health care team feels is inappro-
priate, harmful, or futile [2]. Furthermore, respect for patient autonomy does not
require burdening a patient or family with the entire decision-making process.
However, a fully paternalistic approach is not appropriate either. Some centers have
patients or their surrogates sign documents stating that ECMO withdrawal decisions
will be in the hands of the providers. This may help set expectations and avoid dis-
putes at the time of ECMO discontinuation; however, it is not a perfect solution. It
is a provider’s duty to build trust and understand a patient’s values and preferences,
so that end-of-life decisions can be framed by both clinical realism and the patient’s
best interest. These types of decisions can take an emotional toll on surrogates [2].
Surrogate decision-makers for ICU patients have a PTSD rate of 33%. That rises to
82% when making end-of-life decisions [7].
Additionally, disagreements about continuation of ECMO do not always occur
between surrogates and health care providers, but also amongst members of the
team [4]. For example, respiratory therapists, ECMO technicians, and nurses may
have different perceptions of quality of life or suffering based on their time at the
bedside. It is incredibly important that throughout ECMO initiation, continuation,
and withdrawal, all members of the health care team are communicating and on the
same page. This not only helps to create consistent messaging to patients and sur-
rogates, but also helps avoid moral distress amongst team members.
If conflicting views cannot be resolved to allow ECMO withdrawal, fallback
approaches include “do not escalate” (for complications or subsequent organ fail-
ures) or awaiting ECMO circuit or membrane failure [2]. Although it has been
established that withdrawal and withholding of life-sustaining therapy are ethically
equivalent, they can often feel emotionally different: some clinicians equate with-
drawal of support as ending a life. For this reason, it may be easier for some to await
complications, rather than withdraw the circuit. Ethics consultation may be helpful
in such circumstances.
17 Ethical Challenges in Extracorporeal Membrane Oxygenation 299

Ethics Consultation

It may be appropriate to consider ethics consultation for individual ethics cases, or

as a system-wide policy. Due to the ethical issues common to ECMO care, some
centers have integrated ethics consultation into the care of all ECMO patients [4].
Benefits of having ethics consultation in ECMO cases include the following: elicit-
ing goals and preferences, managing family expectations, mitigating ethical con-
flicts, reevaluating goals as treatment continues, and anticipating and addressing
moral distress felt by staff [1, 5].

Health Disparities

Health care disparities, as evidenced by differences in cardiovascular outcomes,

diabetes, end-stage renal disease rates, COVID-19, and life expectancy, are seen
throughout our health systems every day [12]. There is a growing body of evidence
showing similar racial and ethnic disparities in the utilization and outcomes of
ECMO. These disparities are complex and multifactorial but, in part, could be
driven by systemic, institutional, and interpersonal racism, which join to create the
health inequities we see in minority populations today across the spectrum of health
care [13].
Studies specifically detailing VV-ECMO are lacking; however, other studies
have shown significant racial and ethnic disparities in ECMO utilization and out-
comes. For example, one study of 130,860 congenital cardiac surgery patients
between 2004 and 2015 found that minorities had increased odds of mortality after
congenital heart surgery and suggested that these disparities could be related to
disparate ECMO utilization [14].
Further, in addition to decreased rates of utilization, studies have shown increased
mortality for Black and Latinx patients who do receive ECMO. Data from the
Extracorporeal Life Support International Registry between 1998 and 2012 was
analyzed for all 7106 North American pediatric patients. After adjusting for demo-
graphics, diagnosis, pre-extracorporeal life support care, extracorporeal life support
variables, and extracorporeal life support-related complications, they found that
Black race and Latinx ethnicity were independently associated with increased mor-
tality [15].
It is our ethical imperative to critically challenge the status quo in medicine and
make whole systems change. Equitable care or justice is a core ethical principle
underpinning all of medicine, yet we continue to study health disparities across
medicine without understanding the systemic and institutional policies in place
that allow the unjust disparities to continue. Further study is needed on why there
are disparate rates of both VV-ECMO utilization and mortality in minority
patients.
300 E. Sonntag and M. Pahuja

Scarce Resource Allocation

Scarcity in medicine is ubiquitous and requires a bioethics framework to provide

just allocation of resources. Though physicians struggle with the notion of ration-
ing, some view this as a professional obligation and part of their ethical duties [16].
ECMO is a resource-intensive therapy and requires multidisciplinary specialist care
in an ICU setting. Decisions to balance finite resources happen every day in medi-
cine, albeit on a smaller scale. For instance, we may discharge a patient home
despite understanding the potential benefit from one more night of rest or monitored
care, because the cost of the extra night outweighs the benefit. We do this through-
out medicine without thinking; however, when we apply these same principles to
advanced lifesaving technologies such as ECMO, things become more complicated
because the balance is literally life and death.
Prior to the COVID-19 pandemic, there have been few studies about the manage-
ment of ECMO allocation and certainly no widely accepted algorithms. A survey of
539 physicians from 39 countries in 2019 aimed to characterize physicians’ atti-
tudes towards ECMO initiation, limitation, and withdrawal, finding that resource
allocation was not one of the major influencing factors [11].
Though medicine operates on the notion that there is a finite resource of all social
goods, the pandemic magnified the possibility of large-scale rationing. However,
the pandemic-specific literature is heavily focused on critical care-level resources
such as ventilators and intensive care beds and not specifically ECMO. Most of
these guidelines follow an ethical framework focused on utilitarianism or maximiz-
ing medical benefit and minimizing harm. Often this takes into consideration
patient-specific factors such as age, comorbidities, and functional status to estimate
the likelihood of acceptable recovery. In addition, an influential New England
Journal of Medicine paper by Emanuel et al. prioritized health care workers, avoided
allocating scarce resources on a first-come first-serve basis, and highlighted apply-
ing the same principles to all COVID-19 and non-COVID-19 patients [17].
There has been limited guidance specifically addressing scarce allocation of
resources as it relates to ECMO, although ELSO prepared a consensus guideline on
when and how to use ECMO in COVID-19 patients. They underscored focusing on
existing ECMO centers rather than creating new ones because of the intensive hos-
pital resource utilization, substantial staff training, and multidisciplinary needs with
ECMO – this despite preliminary data that ECMO could be lifesaving for patients
with severe ARDS and refractory cardiocirculatory compromise. Given the inten-
sity of resources required for ECMO, the guidelines focus on hospital- or regional-­
directed case-by-case decisions regarding initiation. They advocate reserving
ECMO for younger patients with few comorbidities while also prioritizing health
care workers [18].
As discussed earlier in the chapter, health care disparities are widespread and
exist throughout medicine. Critics of crisis standards of care guidelines or rationing
ventilators, ICU beds, and possibly ECMO point out that these guidelines almost
certainly will exacerbate health care disparities. Despite many of these algorithms
17 Ethical Challenges in Extracorporeal Membrane Oxygenation 301

having the explicit intention to avoid racial discrimination, focusing on chronic

comorbid conditions such as end-stage renal disease, cardiovascular disease, or dia-
betes inadvertently increases the likelihood that minority populations will suffer.
These populations disproportionately have increased rates of many chronic condi-
tions including end-stage renal disease, cardiovascular disease, diabetes, and can-
cer. Therefore, if we emphasize selecting patients without comorbid conditions for
lifesaving critical care, we will systematically disadvantage minorities. Miller et al.
encourage deprioritizing allocation rules to decrease the risk of exacerbating these
disparities [19].

Integrating Ethics into Education

ECMO has changed the landscape of how we care for patients with refractory respi-
ratory failure and brought with it new ethical challenges. We have detailed in this
chapter how a thorough understanding of medical ethics, health care disparities, and
communication skills may help a physician navigate the ethical dilemmas that arise
in ECMO care [2, 10]. This education and understanding of ethics will continue to
be imperative as the number of ECMO programs and treated patients increase over
time [20]. However, residency and fellowship programs who train in ECMO use
currently lack a formal medical ethics curriculum. Integrating the four basic prin-
ciples of beneficence, nonmaleficence, autonomy, and justice and promulgating cur-
ricula on informed consent, shared decision-making, surrogate decision-making,
and end-of-life care could help physicians prepare for the common ethical chal-
lenges they are sure to encounter in practice [21].

References

1. Stephens AL, Bruce CR. Setting expectations for ECMO: improving communication between
clinical teams and decision makers. Methodist Debakey Cardiovasc J. 2018;14:120.
2. Ramanathan K, Cove ME, Caleb MG, Teoh KLK, Maclaren G. Ethical dilemmas of adult
ECMO: emerging conceptual challenges. J Cardiothorac Vasc Anesth. 2015;29:229–33.
3. Extracorporeal Life Support Organization – ECMO and ECLS Extracorporeal Life Support
Organization (ELSO). In: Extracorporeal Life Support Organization (ELSO) World’s Largest
Registry of ECMO Runs and ECLS Centers. https://www.elso.org/. Accessed 21 Nov 2021.
4. Courtwright AM, Robinson EM, Feins K, Carr-Loveland J, Donahue V, Roy N, McCannon
J. Ethics committee consultation and extracorporeal membrane oxygenation. Ann Am Thorac
Soc. 2016;13:1553–8.
5. Wirpsa MJ, Carabini LM, Neely KJ, Kroll C, Wocial LD. Mitigating ethical conflict and moral
distress in the care of patients on ECMO: impact of an automatic ethics consultation protocol.
J Med Ethics. 2021; https://doi.org/10.1136/medethics-­2020-­106881.
6. Chen DT, Shepherd LL, Mohrmann ME. Substituted judgment. J Gen Intern Med. 2008;24:145.
7. Azoulay E, Pochard F, Kentish-Barnes N, et al. Risk of post-traumatic stress symptoms in fam-
ily members of intensive care unit patients. Am J Respir Crit Care Med. 2005;171:987–94.
302 E. Sonntag and M. Pahuja

8. Quill TE, Holloway R. Time-limited trials near the end of life. JAMA. 2011;306:1483.
9. Graffagnino JP, Avant LC, Calkins BC, Swetz KM. Home therapies in advanced heart failure:
inotropes and diuretics. Curr Heart Fail Rep. 2020;17:314–23.
10. Chang DW, Neville TH, Parrish J, et al. Evaluation of time-limited trials among critically
ill patients with advanced medical illnesses and reduction of nonbeneficial ICU treatments.
JAMA Intern Med. 2021;181:786.
11. Abrams D, Pham T, Burns KE, et al. Practice patterns and ethical considerations in the man-
agement of venovenous extracorporeal membrane oxygenation patients. Crit Care Med.
2019;47:1346–55.
12. Lopez L, Hart LH, Katz MH. Racial and ethnic health disparities related to covid-19.
JAMA. 2021;325:719.
13. Havranek EP, Mujahid MS, Barr DA, et al; American Heart Association Council on quality of
care and outcomes research, council on epidemiology and prevention, council on cardiovas-
cular and stroke nursing, council on lifestyle and cardiometabolic health, and stroke council.
Social determinants of risk and outcomes for cardiovascular disease: a scientific statement
from the American Heart Association Circulation 2015;132(9):873–898.
14. Chan T, Barrett CS, Tjoeng YL, Wilkes J, Bratton SL, Thiagarajan RR. Racial variations
in extracorporeal membrane oxygenation use following congenital heart surgery. J Thorac
Cardiovasc Surg. 2018;156:306–15.
15. Chan T, Di Gennaro J, Farris RW, Radman M, McMullan DM. Racial and ethnic variation in
pediatric cardiac extracorporeal life support survival. Crit Care Med. 2017;45:670–8.
16. Scheunemann LP, White DB. The ethics and reality of rationing in medicine. Chest.
2011;140:1625–32.
17. Emanuel EJ, Persad G, Upshur R, Thome B, Parker M, Glickman A, Zhang C, Boyle C, Smith
M, Phillips JP. Fair allocation of scarce medical resources in the time of covid-19. N Engl J
Med. 2020;382:2049–55.
18. Bartlett RH, Ogino MT, Brodie D, et al. Initial elso guidance document: ECMO for COVID-19
patients with severe cardiopulmonary failure. ASAIO J. 2020;66:472–4.
19. Miller WD, Peek ME, Parker WF. Scarce resource allocation scores threaten to exacerbate
racial disparities in health care. Chest. 2020;158:1332–4.
20. Enumah ZO, Carrese J, Choi CW. The ethics of extracorporeal membrane oxygenation: revis-
iting the principles of clinical bioethics. Ann Thorac Surg. 2021;112:61–6.
21. Sonntag E, Shah K, Katz J. Educating resident and fellow physicians on the ethics of mechani-
cal circulatory support. AMA J Ethics. 2019; https://doi.org/10.1001/amajethics.2019.407.
22. Jonsen AR, Siegler M, Winslade WJ, Mishra R. Clinical ethics: a practical approach to ethical
decisions in clinical medicine. New York: McGraw Hill; 2022.
Index

A Arteriovenous extracorporeal CO2 removal

Abdominal ultrasound, 153 (AV-ECCO2R), 95
Acceptable cardiovascular function, 223 Assisted ventilation during ECLS,
Acquired von Willebrand disease, 169 134–136
Activated neutrophils, 163, 169 Atelectasis and hyperinflation, 154
Acute ARDS phase, controlled mechanical Avoiding intubation, 137
ventilation, 128–130 Awake ECMO, 214
Acute kidney injury (AKI), 222 Axial flow pumps, 43
Acute pulmonary embolism, 286
Airway pressure, 129, 140
Anterograde cannulation of the superficial B
femoral artery, 246 Bedside ultrasound with a vascular
Anticoagulation, 75, 163–165, 240 transducer, 102
management, 271 Bernoulli equation for laminar flow, 46
monitoring, 167, 168 Bicaval design, 99, 111
sparing strategies, 163, 165 Bilateral lung transplantation technique,
Antifibrinolytics, 166 212
Antithrombin deficiency, 169 Biophysics
Anti-thrombotic therapy, 221–222 of membrane gas exchange, 8, 10
for ECMO, 159–169 of membrane oxygenation, 12–13
Anti-Xa concentrations, 178 Bivalirudin, 163, 164
Argatroban, 163, 164 Bleeding, 239
Arterial (SaO2) and central mixed-venous Blood and fibrin clot in oxygenator,
oxygen saturations (SvO2) vs. 236
cardiac output, 21 Blood flow for gas transfer, 49
Arterial line based cardiac output Blood flow through vascular cannulas,
monitoring, 152 100, 101
Arterial oxygen saturation, 130 Blood pump
vs. haemoglobin and cardiac output, 28 positive displacement, 42
Arterial oxygenation, 129 velocity pumps, 42, 43
Arterial saturations are dependent on the Blood transfusion protocols, 165
fraction of cardiac output, 25 Blood-circuit interface, 75
Arteriovenous (AV) configuration, 108 Bohr effect, 20
Arteriovenous-CO2 difference and carbamino Bridge to transplantation (BTT), 205
carriage, 35–36 Bunsen solubility coefficient of oxygen, 11

© The Editor(s) (if applicable) and The Author(s), under exclusive license to 303
Springer Nature Switzerland AG 2022
G. A. Schmidt (ed.), Extracorporeal Membrane Oxygenation for Adults,
Respiratory Medicine, https://doi.org/10.1007/978-3-031-05299-6
304 Index

C Coagulation cascade and inhibition by

Cannulas for extracorporeal support, 98–101 common anticoagulants, 57
Cannulas with a single lumen, 98 Coagulation system, 174
Cannulation Coagulopathy, 3
complications Color doppler flow imaging, 219
inadequate flow, 110 Complement system, 162
infectious complications, 112 Conservation of energy, 51–53
insertion site bleeding, 111 Consumptive coagulopathy, 179
vascular injury, 109 Contact system activation, 160
configuration, 105–108 Continuous activation of coagulation
strategy, 151 system, 162
Carbaminohaemoglobin, 20 Continuous renal replacement therapy
Carbon dioxide (CO2) (CRRT), 74, 132
clearance, 180–181 Continuous venovenous renal replacement
dissociation curve, 34 therapy (CVVRRT), 95
elimination, 220 Contraindications for ECMO for ARDS, 122
partial pressure, 36 Conventional Ventilation or ECMO for Severe
physiology, 30–41 Adult Respiratory Failure (CESAR)
production, 4 trial, 117, 119
transfer, 72 Cycle ergometry, 260
transport in physical solution, 31–36
Carbonic anhydrase, 35
Cardiac and pulmonary failures, 207 D
Cardiac assessment prior to cannulation, 223 Daily management of patients on VV
Cardiac output, 54 ECMO, 217–225
Cardiac output by echo, 153 Daily survey, 217–225
Cardiopulmonary bypass circuits, 66 Daily survey checklist, 218
Cardiorespiratory physiology and mechanical D-dimer (DD), 179
support, 3 2,3-diphosphoglycerate (DPG), 20
Cardiotomy reservoir, 232 Decannulation, 108–109, 271–273
Cavitation, 53 Deoxy and carboxy-challenge, 269–270
Central venous pressure (CVP), 148 Diffusion of gas molecules into a liquid phase
Centrifugal constrained vortex pumps, 43 gas concentrations, 10, 11
Centrifugal pumps, 69–70 solubility coefficient of oxygen, 11, 12
Cerebral hypoxemia, 282 Diffusive transport of gasses, driving
Cervical cannulation, 105 force, 13–16
Circuit Direct thrombin inhibitors (DTIs), 163
air, 231–232 Disconnection from ECLS, 138
assessments, 225 Dissolved carbon dioxide, 11
blood flow, 218 Distal limb ischemia, 283
crises, 230–239 Distal perfusion strategy, 246
designs, 63, 64 Diuresis, 150
evaluation and safety, 225 Dual lumen bicaval venous cannula, 100
integrity, 231 Dual lumen design, 99–100
length and complexity, 65 Dual-lumen internal jugular (dl Vj-V), 219
negative pressure and cannula chatter, 148 Dynamic alveolar hyperinflation and
orientation, 66–69 intrinsic-PEEP, 191
pressure monitoring, 73, 175
priming, 66
single-and double-lumen cannulas, 69 E
surface coatings, 75–76 Early mobilization, 255–256
tubing, 69 ECCO2R-related major adverse events,
Circuit-related emergencies and 196
management, 233–235 Echocardiography, 151
Index 305

Electrical activity of the diaphragm (EAdi) intrinsic positive end-expiratory

tracings, 140 pressure, 191
Emergency interventions, 225 invasive mechanical ventilation, 194
Emergency management during circuit loop refractory to medical
clamping, 230–231 treatment, 194
Energy exchange across the extracorporeal mechanical ventilation, 196
circuit, 51 mechanical, secondary to the potential
Energy metabolism to stress, 6, 7 vascular trauma, 200
Enzymatic processing of glucose, 6 ML based on the AV pressure
Epinephrine, 7 gradient, 190
Erythrocyte derived microparticles, 162 modalities, 187
Erythrocyte metabolism, 7, 8 NIV failure after COPD
Esophageal pressure, 140 exacerbation, 196
Ethical challenges, ECMO, 293, 295–301 noninvasive ventilation, 194
Ethical principles, 294–295 partial pressure of arterial carbon
EuroELSO survey, 223 dioxide, 195
Extracorporeal carbon dioxide removal pathophysiological features, 191
(ECCO2R), 191 pathophysiological mechanisms, 192
and alveolar oxygen concentrations, pathophysiological rationale, 194
40–41 progressive hypoxemia/worsening
in ARDS, 139 hypoventilation, 196
blood flow requirements, 37 pulmonary inflammatory
circuits, 188 mediators, 190
circulation with smaller access pumpless AV-ECCO2R cleared
catheters, 37 sufficient CO2, 190
in obstructive diseases respiratory acidosis secondary to acute
acute exacerbations of COPD exacerbations of obstructive lung
patients, 194 disease, 190
acute worsening of expiratory respiratory function, 195
flow-limitation, 191 respiratory support strategy, 200
anatomical abnormalities/vascular risk-to-benefit balance, 197
diseases, 200 social and economic burdens, 194
arterial cannulation, 190 technological and clinical
arterio-venous by-pass, 190 evolution, 187
bicarbonates and hydrogen ions, 187 technological progress and effective
blood CO2 clearance, 187 marketing, 190
central venous access sites, 188–190 technologically advanced strategy, 191
central venous access with dual-lumen tidal volume and respiratory rate,
cannula, 188 191
clinical application of ECCO2R, 195 venovenous ECCO2R, 195
clinical applications, 194 with near-fatal asthma, 194
clinical indications, 190, 197 oxygenation in, 37–39
clinical strategies, 188 passive oxygenation and low tidal volume
clinical trials, 199 strategies, 40
CO2 clearance, 187 ventilation strategies with, 39–40
efficacy, 191 Extracorporeal circuits for ECMO, 41
endotracheal intubation, 194, 196 Extracorporeal circulation, 162
evidence-based clinical practice, 194 Extracorporeal life support (ECLS), 163
exclusion criteria, 190, 200 cannulation configuration, 82
gas exchange, 197 console, 68
hypercapnic acidosis, 190 failure of, in ARDS, 274
indications, exclusions, and goals, removal after transplantation, 210
198, 200 Extracorporeal Life Support Organization
inflammatory response, 190 (ELSO) Registry, 230
306 Index

Extracorporeal membrane oxygenation Heat exchanger, 76, 238

(ECMO), 199 Heater-cooler, 76–77
in ARDS, 121 Hematologic consequences with
centers to guide ventilation settings in the ECMO, 160–162
early ECMO phase, 129 Hemodynamic crises in ECMO patients, 224
circuit change, 177 Hemodynamic management, 271
circuit with a centrifugal vortex pump, 50 Hemolysis, 241–243
configuration, 2, 85 Hemostatic alterations during
flow management, 270 ECMO, 159–160
initiation knowledge, 296 Henderson-Hasselbalch equation, 32
referral arm, 117 Heparin induced thrombocytopenia
technology, 117 (HIT), 56, 164
therapy vs. a transition to care, 297 Heparin resistance, 169, 178
weaning and liberation, 267–271 Heparin-bonding, 76
Extracorporeal membrane oxygenation to Heparin-coated circuits, 116
Rescue Lung Injury in Severe Heparin-induced thrombocytopenia (HIT), 76,
ARDS (EOLIA) trial, 117, 119 168, 178
Extracorporeal support, 116, 179 HFOV coupled with pumpless arteriovenous
Extracorporeal therapies, 81 extracorporeal lung assist (PECLA)
Extubation is feasible in some ECMO for CO2 removal, 134
patients, 223 High frequency oscillatory ventilation
(HFOV), 133, 134
Hollow fibre oxygenator construction, 9
F Hybrid cannulation configuration, 107
Femoral arterial cannulation, 246 Hybrid ECMO Configurations, 92
Femoral cannulation, 106 Hydrogen ion concentration, 32–35
Femoral vessels, 102 Hyperfibrinolysis, 162
Fibrinolysis, 162 Hypofibrinogenemia, 178
Fick’s Law, 13 Hypovolemia, 110
Flow of ideal fluids in solid conduits, 45–47 Hypoxemia, 244
Flow regurgitation, 53
Fluid management in ARDS patients on
extracorporeal support, 149 I
Fluid-response predictors, 224 ICU’s culture of liberation, 258
Immobility, myopathy and weakness, 255
Inadvertent decannulation, 238, 239
G Inclusion criteria for CESAR and EOLIA, 118
Gas exchange, 217–221 Infection control, 102–103
Gas exchange membrane, 16 Inflammatory consequences with
Glucagon, 7 ECMO, 162–163
Glucose, 3 Inflammatory response to extracorporeal
Glucose metabolism, 4 membrane oxygenation
Glycolysis, 4 (ECMO), 161
Glycolysis activity in hypoxic environment, 4 Inhaled aerosolized prostacyclin (iAP), 133
Inhaled NO (iNO), 133
Inhaled pulmonary vasodilators, 132, 133
H Institution of ECMO to avoid intubation,
H1N1-associated ARDS, 116 136, 137
Haemoglobin, 18 Insulin, 6
Haemoglobin’s affinity for oxygen, 20–21 Intensive care bed, 300
Harlequin syndrome, 155 Interventional lung assist, 207
Health costs of immobility, 254–255 Intracellular glucose, 4
Healthcare disparities, 299 Intracellular metabolism, 5
Index 307

Intracerebral hemorrhage, 239 red blood cell destruction, 179

Intraoperative va-ECMO, 212 technology and provider
Intravascular volume depletion and proficiency, 173
overload, 154 thromboembolic complications, 174
Intubation and mechanical ventilation, exchange, 177, 181
206, 210 function, 219
Invasive mechanical ventilation (MV), 191 gas transfer, 175–176
shunt fraction, 176–177
Membrane oxygen flux matches demand, 17
L Membrane oxygenators, 49, 70–73
Lactate, 153 plasma leak, 74
Laminar flow, 47, 48 rated flow, 73
Left ventricle (LV) ejection, 281 Miniaturisation of rotary pumps (without
Life-threatening bleeding, 241 oxygenators), 1
Limb ischemia, 110–111, 246–247 Minimally invasive technique, 210
Limb perfusion, 246 Mitochondrial coupling of ATP production to
Lipogenesis, 6 NADH oxidation, 8
Low-flow venovenous configuration, 107–108 Mixed-venous saturation, 25–27, 29–30
Low-molecular-weight heparin (LMWH), Mixing blood streams with differing oxygen
164, 178 concentrations, 23
Lung function, 266 Mobilization
Lung oxygen transfer, 129 barriers, 257, 258
Lung transplantation, 205–214 during ECLS, 253–257, 259–261
Lung-protective ventilation strategy, 117 team and methods, 259, 260
Modelling VV-ECMO, 27
Modes of ECMO, physiological impact, 83
M Monitoring anticoagulation in critically ill
Mechanical support in lung patients, 166, 167
transplantation, 205 Monitoring respiratory effort in spontaneously
Medical ethics, 301 breathing patients on ECLS, 137
Membrane exchange criteria based on ΔP, 180 Monitoring RV dysfunction by
Membrane exposure time, 14–15 echocardiography, while on
Membrane gas exchange, biophysics of, 8, 10 ECMO, 152
Membrane leakage, 237
Membrane lung (ML)
dysfunction, 177 N
acute onset thrombocytopenia, 179 National Institutes of Health (NIH), 125
CO2 clearance, 180, 181 Near-infrared spectroscopy (NIRS), 153
complementation activation, and Neurological evaluation, 210
systemic inflammatory Neurological events, 224
response, 174 Nomenclature, ECLS, 82
continuous blood to non-biologic Noninvasive ventilation (NIV), 191
surface interaction, 174 Norepinephrine, 7
ECMO-related hematologic
abnormities, 178
ECMO-specific findings, 179 O
extracorporeal circuit, 174 Occlusion of the flow path, 52–53
evaluation of platelets, 179 Open surgical technique, 105
hematologic derangements, 178 Organ donor acceptance, 213
hematologic profile, 175 Oscillation for Acute Respiratory Distress
mechanisms of, 174 Syndrome Treated Early
obstruction to blood flow, 180 (OSCILLATE) trial, 133–134
oxygen transfer, 180 Oscillation for ARDS (OSCAR) trial, 134
308 Index

Oxidative phosphorylation, 4 Platelets, 162

Oxygen carriage in blood, 18 Pneumothoraces, 134
Oxygen carrying capacity, 27 Point-of-care echocardiography, 156
Oxygen haemoglobin dissociation Point-of-care-ultrasound (POCUS)
curve, 19, 22 evaluation of abdominal organs, 149
Oxygen transfer, 71 examination, 149, 219
Oxygen transport, 18–30 Polymethylpentene (PMP) gas-exchanging
Oxygenation, 17, 217, 218 membranes, 254
Oxygenator Sweep Gas Flow Polypropylene microporous membranes, 8–9
Management, 269–270 Positive displacement pumps, 42
Oxygenators, 49 Pre-oxygenator air, 231
failure, 236–237 Pressure-flow relationships for various size
performance, 175 single lumen return cannulas, 101
as a resistor, 50 Preventing mechanical ventilation with
Oxygen-haemoglobin dissociation ECLS, 212
curve, 21, 22 Priming, 66
Oxygen-poor blood, 207 Progressive pulmonary failure, 210
Prolonged postoperative va-ECMO, 212, 213
Prompt recognition, 246
P Prone positioning, 222
Partial extracorporeal CO2 removal (PECOR) Proportional effect on alveolar oxygen tension
and ventilator of CO2 removal, 41
Management, 138–141 PROSEVA trial, 131
Partial extracorporeal low-flow partial CO2 Pulmonary arterial mixed-venous oxygen
removal (PECOR) clinical case, 139 saturation, 30
Partial pressure vs. time for a gas dissolving in Pulmonary artery catheters (PACs), 152, 282
a liquid, 15 Pulmonary artery oxygen content and
Passive leg raising (PLR), 149 saturations, 25
Patient crises, 239–241, 244–247 Pulmonary congestion, 150
Patient transport on ECLS, 211 Pulmonary failure, 210, 211
Patient-related emergencies and Pulmonary gas exchange, 8
management, 242–243 Pulmonary haemodynamics, 54–56
Percutaneous cannulation, 103, 104 Pulmonary hypertension, 211
Percutaneous skin and vessel insertion, 104 Pulmonary or circulatory failure, modes of
Peri-procedural prophylaxis with intravenous support, 209
antibiotics, 103 Pulse pressure variation (PPV), 148
Persistent hypoxemia, 180 Pump failure, 237–238
Physical activity in critically ill, 260 Pumpless arteriovenous extracorporeal lung
Physical therapy assist (PECLA) for CO2
in critically ill patients, 222–223 removal, 134
and patient mobilization, 208 Pumpless AV-ECCO2R in adult and pediatric
Physiological response to extracorporeal CO2 patients, 199
removal, 137 Pumpless interventional lung assist (iLA), 207
Physiology, extracorporeal life support
blood pump and oxygenator, 2
and limitations, 2 R
metabolic responses to critical illness and Randomized controlled trials of ECMO,
chronic pathology, 3 116, 120
recirculation, 29 Rapoport Leubering shunt, 8
recovery of a failing cardio-respiratory Reactions of carbon dioxide in solution, 31–32
system or a destination therapy, 3 Readiness, 266–267
Plasma-free hemoglobin (PfHb) Recirculation, 219
monitoring, 179 due to tricuspid closure, 29
Plastic conduits, 42 during venovenous support, 111
Index 309

Recombinant activated factor VII or Shared decision-making, 301

prothrombin complex Shock, 245
concentrates, 166 Single lumen cannulas for percutaneous
Recombinant factor VII, 241 cannulation, 99
Red blood cell destruction during ECMO, 179 Spectrophotometric sensors, 67
Refractory hypoxemia, 244, 245 Surgical cannulation the vessel, 101
Regional citrate anticoagulation, 165 Surgically-placed cannulas, 109
Relative flow direction, 15 Systemic circulation management, 147, 148,
Relative viscosity of blood, 44 150, 151, 153–155
Renal dialysis, 200 Systemic inflammatory response syndrome
Renal replacement therapy (RRT), 74, 75, 222 (SIRS), 3, 163
Reparative properties of lung, 56–58
Rescue treatments for persistent hypoxemia
during ECLS T
clinical improvement after pronation, Termination
132 ECLS, 273–275
complications attributable to ECMO, 297–298
pronation, 132 Theoretically platelet inhibition, 165
dislodgement of tubes or vascular Therapy session, 261
catheters, 131–132 Thermodilution cardiac output, 152
hemodynamic instability, 132 Thrombocytopenia, 240
HFOV, 133, 134 Thromboelastography (TEG), 167
intravenous administration of pulmonary Thromboembolic events, 164
vasodilators, 132, 133 Thrombosis, 232–236
loss or obstruction of the endotracheal of the cannula, 247
tube, 131 within the oxygenator, 236
pathophysiologic mechanisms, 131 Thrombus formation, 55, 56, 175
pressure sores, 131 Time-limited trials, ECHO, 297
prone positioning, 131, 132 Tissue perfusion, 3
pulmonary vasodilators, 132, 133 Transmembrane pressure gradient, 175
Resistance to laminar flow, 47–48 Transthoracic (central) cannulation, 108
Resistances to diffusive transport, 16 Trial-off ECMO failure, 270
Respiration-related metrics, 224 Tubing rupture, disconnect, or cannula
Respiratory quotient (RQ), 4, 6 fracture, 238
Resuscitation, 150 Turbulence and hematocrit on local
Reynolds number, 48–49 concentration gradients, 16
Right ventricular (RV) failure, 211 Turbulent flow, 48
Right ventricular (RV) function, 54–56,
150–155, 223
Right-shifted oxygen haemoglobin U
dissociation curves, 22 Ultrafiltration of plasma water over the
Rotational thromboelastography oxygenator membrane, 17
(ROTEM), 167 Unfractionated heparin (UFH), 56, 163
Upper body oxygenation, 282–283

S
Saturable oxygen carrier (haemoglobin), 23 V
Scarce resource allocation, 300, 301 Vasoactive drug requirements, 224
Sedation and mechanical ventilation, 206 Vasoactive management for RV failure,
Sedation management, 270 154
Seldinger technique, 103 Velocity pumps, 42, 43
Severe acute asthma, 197, 199 Veno-arterial (VA) configuration, 105–106
Severe adult respiratory distress Veno-arterial (VA) ECLS, circuit blood
syndrome, 259 flow, 275
310 Index

Veno-arterial (VA) ECMO, 125, 155 advantages, 87

advantages, 83, 91 cannulation, 84
anatomy, 279 circuit configuration, 24
anticoagulation, 280 degree of recirculation, 84
cannulation for older children and dual-lumen cannulas, 84, 85
adults, 88, 89 for hypoxemic respiratory failure, 122
cannulation techniques, 279 as initial support, 155
VA circuit, 87–89 physiology of, 86, 87
disadvantages, 83 on RV dysfunction, 154, 155
dual-cannula setup, 84 for severe ARDS, 120
gas exchange effects, 89 Venovenous (VV) configuration, 106–107
gas exchange in patients, 280 Ventilation, 154
LV distention, 281 Ventilator induced lung injury (VILI), 126
monitoring, 280–283 Ventilator management, 120, 267–268
overall gas exchange, 281–282 Ventilator management during ECLS
overall oxygenation, 281 acute hypoxemic respiratory failure, 125
PA catheterization, 282 consensus during the early ECMO
physiological implications, 89–91 phase, 131
for respiratory failure, 208–209 extracorporeal blood flow, 130
cardiogenic shock, 277 high airway pressures, 126
gas exchange, 277 high tidal volume, 126
hypoxemic respiratory failure, 278 lung recruitment (open lung strategy),
inadequate oxygenation despite 129
maximal VV ECMO observational or physiological studies,
support, 287–288 126
LV failure, 285 pathophysiology of extracorporeal gas
oxygenation, 278 exchange
primary management strategy, 286 apneic oxygenation alveolar PO2, 128
respiratory failure, 285 bypass recirculation, 127
RV dysfunction or failure, 285 CO2 removal, 127, 128
septic or vasoplegic shock, 285 membrane oxygenator, 128
septic shock, 284 oxygenation, 127
vasoplegic shock, 284–285 residual hypoxic vasoconstriction, 127
return cannulas, 279 pure hypoxia rescue therapy, 126
upper body desaturation, 282 rescue from hypoxia, 129
upper body oxygenation, 282–283 severe pulmonary hypoperfusion, 126
Venopulmonary (VP) configuration, 107 tidal volume, 126
Venopulmonary design, 111 VILI prevention (lung rest), 129
Venous capture, 224 Viscosity (η), 44, 45
Venous oxygen saturation measured in the Volume overload and hypovolemia in patients
pulmonary artery (SvO2) or the on ECMO, 150
superior vena cava (ScvO2), 153 Volume status, 148–150
Venous oxygen saturation vs. haemoglobin von Willebrand factor (VWF), 169
and cardiac output, 20 VVA cannulation with partially occlusive line
Venous oxyhemoglobin saturation, 30, 220 clamp on venous inflow, 93
Venous pulmonary artey (V-PA) ECMO, 155
Venovenoarterial (VVA) configuration, 107
Veno-venous cannulation (VV ECCO2R), 94 W
Veno-venous extracorporeal membrane Weaning, 283
oxygenation (VV-ECMO), 1, 23, from VA ECMO, 273
83–87, 207–208, 225 from VV ECMO, 266, 267, 271, 272