@anesthesia Books 2016 Extracorporeal

Respiratory Medicine
Series Editor: Sharon I.S. Rounds
Gregory A. Schmidt Editor
Extracorporeal
Life Support for
Adults
Series Editor :
Sharon I.S. Rounds
More information about this series at http://www.springer.com/series/7665

Gregory A. Schmidt
Editor
Extracorporeal Life Support

for Adults
Editor
Gregory A. Schmidt, MD
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)

ISBN 978-1-4939-3004-3 ISBN 978-1-4939-3005-0 (eBook)
DOI 10.1007/978-1-4939-3005-0
Library of Congress Control Number: 2015950466
Springer New York Heidelberg Dordrecht London

Springer Science+Business Media New York 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microlms 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 specic 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, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.
Printed on acid-free paper
Humana Press is a brand of Springer

Springer Science+Business Media LLC New York is part of Springer Science+Business Media
(www.springer.com)
To William R. Lynch, MD, who built and
nurtured an outstanding program, exhibited
remarkable vision in how to advance ECLS
care, and opened my eyes to its new
possibilities.
Preface
Extracorporeal life support (ECLS) consists of using an external gas-exchanging

membrane to support oxygenation or carbon dioxide removal (or both), at times
including circulatory assistance. ECLS has been used in severe hypoxemic respira-
tory failure (ARDS, pneumonia); 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 trans-
plant. Historically, ECLS has been used sparingly, often as a last resort, and in few
centers with the requisite expertise. Three factors have combined to change this.
First, technological improvements in membranes, pumps, circuits, and cannulas
have led to more efcient and safer ECLS. Second, the CESAR trial has shown that,
for adults with severe ARDS, referral to an ECLS center improves outcomes.
Finally, the adverse consequences of conventional management of lung failure,
including ventilator-induced lung injury, ICU-acquired weakness, and nosocomial
infection, have become abundantly clear. Some of these may be ameliorated by
using ECLS in preference to conventional care. As perceptions of the role of ECLS
have evolved, more practitioners and more centers are developing ECLS capability
or positioning themselves to offer ECLS.
The aim of this book is to deliver a concise, evidence-based review of ECLS for
adult disease. Adult medicine (rather than neonatal and pediatric disease, where
ECLS has an established but limited role) represents the growth area for
ECLS. 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 ECLS. The text concludes with a fascinating historical review, showing
just how far weve come.
This text has been written for practicing physicians, nurses, perfusion specialists,
therapists, and critical care trainees who are considering whether to refer their
patients for ECLS, debating whether to offer ECLS capability to their patients, or
are already providing ECLS but seek a practical reference to best practices and
updated information. It could never have been completed without the inspiration
vii
viii Preface
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 eld; and our publisher at Springer-Link who pushed
for this important book. Finally, I recognize all those who do the hard work: the
nurses, perfusionists, and therapists who dedicate their lives to the critically ill. This
is an exciting time, ripe with change and opportunity. We seek a path forward for the
benet of all our patients.
Iowa City, IA, USA Gregory A. Schmidt, MD

Contents
1 Physiology of Extracorporeal Life Support (ECLS) ........................... 1

Matthew J. Brain, Warwick W. Butt, and Graeme MacLaren
2 Hypoxemic Respiratory Failure: Evidence, Indications,
and Exclusions ......................................................................................... 61
Darryl Abrams, Matthew Bacchetta, and Daniel Brodie
3 Cardiogenic Shock: Evidence, Indications, and Exclusions................ 73
Nicolas Brchot and Alain Combes
4 ECCO2R in Obstructive Diseases: Evidence, Indications,
and Exclusions ......................................................................................... 87
Lorenzo Del Sorbo and V. Marco Ranieri
5 ECLS as a Bridge to Lung Transplantation ......................................... 105
Christian Kuehn
6 Modes of ECLS ....................................................................................... 117
L. Keith Scott and Benjamin Schmidt
7 Vascular Access for ECLS ...................................................................... 133
Steven A. Conrad
8 Circuits, Membranes, and Pumps ......................................................... 147
Bradley H. Rosen
9 Ventilator Management During ECLS ................................................. 163
Antonio Pesenti, Giacomo Bellani, Giacomo Grasselli,
and Tommaso Mauri
10 Daily Care on ECLS ............................................................................... 181
Giles J. Peek
11 Crises During ECLS ............................................................................... 193
Cara L. Agerstrand, Linda B. Mongero, Darryl Abrams,
Matthew Bacchetta, and Daniel Brodie
ix
x Contents
12 Mobilization During ECLS .................................................................... 211

Gregory A. Schmidt
13 ECMO Weaning and Decannulation ..................................................... 223
Sundar Krishnan and Gregory A. Schmidt
14 The Story of ECLS: History and Future .............................................. 233
J. Ann Morris, Robert Pollock, Brittany A. Zwischenberger,
Cherry Ballard-Croft, and Joseph B. Zwischenberger
Index ................................................................................................................. 261

Contributors
Darryl Abrams, MD Division of Pulmonary, Allergy and Critical Care, New York-
Presbyterian Hospital/Columbia University Medical Center, New York, NY, USA
Cara L. Agerstrand, MD Division of Pulmonary, Allergy, and Critical Care
Medicine, Department of Medicine, Columbia University College of Physicians
and Surgeons/New York-Presbyterian Hospital, New York, NY, USA
Matthew Bacchetta, MD, MBA, MA Division of Thoracic Surgery, New York-
Cherry Ballard-Croft, PhD Division of Cardiothoracic Surgery, Department of
Surgery, University of Kentucky College of Medicine, Lexington, KY, USA
Giacomo Bellani, MD, PhD Department of Health Sciences, University of
Milano-Bicocca, Monza, Italy
Department of Anesthesia and Critical Care, San Gerardo Hospital and Milano-
Bicocca University, Monza, Italy
Matthew J. Brain, MBBS (Hons), FRACP, FCICM, DDU School of Public
Health and Preventive Medicine, Monash University, Malvern East, VIC, Australia
The Alfred Intensive Care Unit, Melbourne, VIC, Australia
Department of Medicine, Launceston General Hospital, Launceston, TAS, Australia
Nicolas Brchot, MD, PhD Service de Ranimation Mdicale, Hospital
PitiSalptrire, Paris, France
Daniel Brodie, MD Division of Pulmonary, Allergy and Critical Care, New York-
Warwick W. Butt, FRACP, FCICM ICU RCH, Department of Paediatrics UoM,
Clinical Sciences Theme MCRI, Royal Childrens Hospital, Melbourne, VIC, Australia
Paediatric Intensive Care Unit, Parkville, VIC, Australia
xi
xii Contributors
Alain Combes, MD, PhD Service de Ranimation Mdicale, Institut de

Cardiologie, Groupe Hospitalier Piti-Salptrire, iCAN, Institute of
Cardiometabolism and Nutrition, Paris Cedex, France
Steven A. Conrad, MD, PhD, MCCM, FCCP Department of Medicine,
Emergency Medicine and Pediatrics, Louisiana State University Health Sciences
Center, Shreveport, LA, USA
Lorenzo Del Sorbo, MD Dipartimento di Anestesiologia e Rianimazione,
Azienda Ospedaliera Citt della Salute e della Scienza di Torino, Universit di Torino,
Torino, Italy
Inter-departmental Division of Critical Care Medicine, University Health Network,
University of Toronto, Toronto, ON, Canada
Giacomo Grasselli, MD Department of Anesthesia and Critical Care, San Gerardo
Hospital and Milano-Bicocca University, Monza, Italy
Sundar Krishnan, MBBS Department of Anesthesia, University of Iowa Hospitals
and Clinics, Iowa City, IA, USA
Christian Kuehn, MD Department of Cardiac, Thoracic, Transplantation and
Vascular Surgery, Privatdozent Dr. med., Hannover Medical School, Hannover,
Germany
Graeme MacLaren, MBBS, FRACP, FCICM, FRCP, FCCP, DipEcho ICU
RCH, Department of Paediatrics UoM, Clinical Sciences Theme MCRI, Royal
Childrens Hospital, Melbourne, VIC, Australia
Paediatric Intensive Care Unit, Parkville, VIC, Australia
Cardiothoracic ICU, National University Hospital, Singapore, Singapore
Tommaso Mauri, MD Department of Anesthesia and Critical Care, Fondazione
IRCCS Ca Granda Ospedale Maggiore Policlinico, Milan, Italy
Linda B. Mongero, CCP, BS Department of Clinical Perfusion, New York
Presbyterian-Columbia University Medical Center, Locust Valley, NY, USA
J. Ann Morris, BS Division of Cardiothoracic Surgery, Department of Surgery,
University of Kentucky College of Medicine, Lexington, KY, USA
Giles J. Peek, MD, FRCS CTh, FFICM Heartlink ECMO Centre, Gleneld
Hospital, Leicester, UK
Antonio Pesenti, MD Department of Health Sciences, University of Milano-
Bicocca, Monza, Italy
Department of Anesthesia and Critical Care, San Gerardo Hospital and Milano-
Bicocca University, Monza, Italy
Robert Pollock, BS Division of Cardiothoracic Surgery, Department of Surgery,
University of Kentucky College of Medicine, Lexington, KY, USA
Contributors xiii
V. Marco Ranieri, MD Dipartimento di Anestesiologia e Medicina degli Stati

Critici, Ospedale S. Giovanni Battista-Molinette, Universit di Torino, Torino, Italy
Dipartimento di Anestesiologia e Rianimazione, Azienda Ospedaliera Citt della
Salute e della Scienza di Torino, Universit di Torino, Torino, Italy
Bradley H. Rosen, DO Division of Pulmonary, Critical Care, and Occupational
Medicine, Department of Internal Medicine, Carver College of Medicine, University
of Iowa Hospitals and Clinics, Iowa City, IA, USA
Benjamin Schmidt, MD Department of Surgery, Wake Forest University, Medical
Center Boulevard, Winston-Salem, NC, 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 Department of Anesthesiology, Wake Forest University,
Medical Center Boulevard, Winston-Salem, NC, USA
Brittany A. Zwischenberger, MD Division of Cardiothoracic Surgery, Department
of Surgery, University of Kentucky College of Medicine, Lexington, KY, USA
Joseph B. Zwischenberger, MD Division of Cardiothoracic Surgery, Department
of Surgery, University of Kentucky College of Medicine, Lexington, KY, USA
Chapter 1
Physiology ofExtracorporeal
Life Support (ECLS)
MatthewJ.Brain, WarwickW.Butt, andGraemeMacLaren
Introduction
Extracorporeal life support (ECLS) and related implantable circulatory assistance

devices describe several advancing technologies with broadening scope that are
being increasingly incorporated into 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 to maintain systemic oxygen delivery by oxygenating
venous blood returning to the right heart. In veno-arterial mode (VA-ECMO),
M.J. Brain, MBBS (Hons), FRACP, FCICM, DDU (*)

School of Public Health and Preventive Medicine, Monash University,
Malvern East, VIC, Australia
The Alfred Intensive Care Unit, Melbourne, VIC, Australia
Department of Medicine, Launceston General Hospital,
274-280 Charles St, Launceston, TAS 7250, Australia
e-mail: m.brain@iinet.net.au
W.W. Butt, FRACP, FCICM
ICU RCH, Department of Paediatrics UoM, Clinical Sciences Theme MCRI,
Royal Childrens Hospital, Melbourne, VIC, Australia
Paediatric Intensive Care Unit, 50 Flemington Road, Parkville, VIC 3052, Australia
e-mail: Warwick.butt@rch.org.au
G. MacLaren, MBBS, FRACP, FCICM, FRCP, FCCP, DipEcho
ICU RCH, Department of Paediatrics UoM, Clinical Sciences Theme MCRI,
Royal Childrens Hospital, Melbourne, VIC, Australia
Paediatric Intensive Care Unit, 50 Flemington Road, Parkville, VIC 3052, Australia
Cardiothoracic ICU, National University Hospital,
5 Lower Kent Ridge Rd, Singapore 119074, Singapore
e-mail: graeme_maclaren@nuhs.edu.sg
Springer Science+Business Media New York 2016 1

G.A. Schmidt (ed.), Extracorporeal Life Support for Adults,
Respiratory Medicine 16, DOI10.1007/978-1-4939-3005-0_1
2 M.J. Brain et al.
systemic blood flow is augmented by the extracorporeal blood pump, while

VPA-ECMO describes augmentation of pulmonary arterial flow. Both of the latter
configurations can also incorporate support of oxygenation.
Configurations can also be classified by the site of vascular access, with cannulas
being either peripherally placed via the great vessels or centrally placed via thoracot-
omy. Rotary pumps (without oxygenators) have been miniaturised, allowing develop-
ment of left and right ventricular assist devices (LVAD and RVAD) respectively.
A basic ECMO circuit consists of a blood pump and oxygenator connected by con-
duits (Fig.1.1). Other components may be added to this basic configuration, in particu-
lar other extracorporeal circuits such as renal replacement devices. However,
maintaining simplicity is important for safety, infection control and troubleshooting.
Each configuration creates a unique interaction with the cardiorespiratory
system. Sound understanding of the physiology and limitations of each mode is
Fig. 1.1 Schematic of ECMO configurations, circles represent pumps, 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 (ECCO2R) is commonly performed with a VV-ECMO configuration,
usually with a single dual-lumen catheter. Intravascular membrane oxygenators have also been
developed [1] but are not currently in clinical use
1 Physiology ofExtracorporeal Life Support (ECLS) 3
required to prescribe, manage and wean this support and recognise evolving
complications of the therapy. Although designed primarily to replace cardiorespira-
tory function, the interaction of ECLS with several other physiologic systems must
be 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. The systemic inflammatory response syndrome
(SIRS) may arise from the underlying disease or as a reaction to the non-biological
material of the ECLS circuit. The metabolic response to critical illness has direct
implications for oxygenation and CO2 removal, as well as nutritional supplementation
to facilitate later weaning.
In order to comprehensively understand ECLS and its effects on human physiology,
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 oxidative
metabolism can maintain proper cellular and organ function.
Cardiorespiratory physiology and any mechanical support must provide an
adequate hydrostatic pressure gradient across capillary beds to support blood flow,
as well as maintain concentration gradients by which substrates, including oxygen,
diffuse 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 andAerobic andAnaerobic Metabolism
Glucose and other simple carbohydrates enter cells down a concentration gradient
through glucose transporters that allow for tissue-specific behaviour such as prefer-
ential basal uptake by the brain, concentration-dependent uptake by the liver,
concentration-sensing by the insulin-secreting pancreatic -cells and insulin-
dependent uptake in skeletal muscle and fat [2].
Intracellular glucose is rapidly phosphorylated in the cytosol by hexokinases,
after which it becomes the primary substrate for energy production or biosynthetic
reactions including glycogen storage (Fig.1.2). Utilisable intracellular energy is
Fig. 1.2 Key intermediates in intracellular metabolism: After entering cells, glucose is phosphor-
ylated (-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 tricarboxylic acid (TCA) cycle and oxidative phosphorylation only occur in mito-
chondria 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 [3]. Different amino
acids can enter or be synthesised from the pathway 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, ketone body production when insufficient oxaloac-
etate exists for acetyl groups to enter the TCA cycle. Ketone bodies are produced predominantly
in the liver from fatty acid breakdown and constitute a glucose-sparing fuel for the brain and heart.
Humoral promoters and inhibitors of reactions are shown
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 net 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. This yields 38 ATP molecules, significantly
more than glycolysis, but generates NADH in such quantities that a powerful elec-
tron 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 normal
cellular processes.
When oxygen is not available in sufficient quantities, ATP generation from ADP
can only continue in the cytosol by 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
or global hypoxemia of the organism.
Carbon Dioxide Production andtheRespiratory Quotient
The respiratory quotient (RQ) describes the ratio of the amount of carbon dioxide
( VCO 2 ) produced per unit time to the amount of oxygen consumed ( VO 2 ).
VCO2
RQ = (1.1)
VO
2

The respiratory quotient depends on the sources of fuel being used. For glucose
metabolism, the six carbon atoms result in production of six molecules of carbon
dioxide while consuming six molecules of oxygen; it thus has a respiratory quotient
of 1. The reactions for oxidation of some amino acids and fatty acids (lipolysis)
produce less CO2 (by not including the pyruvate to acetyl-CoA reaction, Fig.1.2)
and hence have respiratory quotients of less than 1.
In contrast, each acetyl-CoA molecule utilised for fatty acid synthesis (lipogenesis)
results in production of a molecule of CO2 (from pyruvate to acetyl-CoA, Fig.1.2)
6 M.J. Brain et al.
without increasing mitochondrial NADH.As the rate of oxygen consumption

depends on the mitochondrial concentration of NADH, lipogenesis 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 [4, 5]:
Glucose Oxidation : C6 H12 O6 + 6O2 6CO2 + H 2 O RQ = 1

Lipolysis of glycerol triestearate : C57 H110 O6 + 80.25O2 + 57CO2
+ 55.5H 2 O RQ = 0.667

Amino Acid ( Glycine ) Oxidation : NC2 H 5 O2 + 2.25O2 2CO2
+ 2.5H 2 O RQ = 0..88

Lipogenesis from Glucose* : 4C6 H12 O6 + O2 C16 H 32 O2
+ 8CO2 + 8H 2 O RQ = 8
Lipogenesis* : 13.83C6 H12 O6 + 5O2 C55 H104 O6 + 28CO2 + 31H 2 O RQ = 5.6

*Note that the RQ of lipogenesis depends on the fatty acid being produced and
the carbohydrate that is utilised. C16H32O2 palmitic acid. C55H104O6 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 metabolism is coupled to TCA intermediate availability and, when carbo-
hydrate loads are excessive (such as with glucose supplementation exceeding
4mgkg1min1 [5.8gkg1day1]), lipogenesis occurs with a respiratory quotient as
high as 8 [57] resulting in a high CO2 burden.
Metabolism intheStressed State
Key hormones coordinate the response to nutrition supply and stress. Insulin marks
the fed state, promoting hepatic glucose uptake, glycogen and amino acid synthesis
and conversion of acetyl-CoA to free-fatty acid production while in the peripheral
tissues stimulating myocyte synthesis of contractile elements and adipocyte
triglyceride deposition.
Glucagon is secreted by pancreatic -cells in response to low blood glucose
levels 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, 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, 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 massprimarily
amino acids from muscle catabolism [3]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 weaning, a requirement for tracheostomy and the risk of
secondary infection.
Erythrocyte Metabolism
Being a specialised organ for oxygen transport, erythrocytes are nearly 90%
haemoglobin-by-weight, with very few other organelles. Nevertheless, they require
an ongoing energy source to maintain membrane integrity, cytoskeleton structure,
intracellular electrolyte and osmotic equilibrium and to keep the iron 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. However,
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 [8].
8 M.J. Brain et al.
The RapoportLuebering shunt (Fig.1.2) describes the pathway for 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 [9]. Oxygen depletion (resulting in fewer
haemoglobin-binding sites for 2,3-DPG), acidotic conditions that inhibit 2,3-DPG
synthesis and the accumulation of inorganic phosphate (Pi) which increases 2,3-DPG
breakdown [8], result in decreased intracellular 2,3-DPG concentrations. This is most
relevant under conditions of red cell storage where lower glycolysis rates and accu-
mulation 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 receiv-
ing ECLS, this effect may have significant implications for oxygen carriage. The role
of 2,3-DPG will be further discussed below when considering oxygen carriage.
Biophysics ofMembrane Gas Exchange
Mitochondria can couple ATP production to NADH oxidation only if sufficient oxy-
gen exists in the environment of cells. Similarly, carbon dioxide diffuses 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 oxygen and carbon dioxide in blood
When considering pulmonary gas exchange a third component must be considered:
the convective transport of the gas to the alveolar epithelium. However, exposure of
the extracorporeal membrane to fresh gas flow is somewhat simpler in ECLS and
will be considered later in the context of 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
untoward biological responses, reduce priming volumes, avoid plasma leakage and
increase their simplicity and integration as systems. Materials and construction of
gas exchange membranes will be discussed in later chapters; however, a brief introduc-
tion 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
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. (1.9) are demonstrated with being the porositythe
area of membrane occupied by gas, 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 the constric-
tivitythe resistance to passage
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 [10]. Most systems direct fresh gas
through the lumen of the hollow fibres, while blood flows between the tubules
(termed extra-capillary flow). The reverse configuration is also sometimes utilised,
however, and overall characteristics such as total surface area for gas exchange,
resistance to flow and trauma to formed blood components will be determined by
factors such as membrane material, fibre diameter and length, fibre density and the
velocity of the blood [11].
10 M.J. Brain et al.
Heat loss over the extracorporeal circuit into the environment can be substantial
and heat exchangers are commonly incorporated into the oxygenator. Figure1.3
demonstrates one such design where the microporous membrane fibres are laid
perpendicularly to impermeable capillaries that circulate heated water, allowing
heat to be regulated.
Diffusion ofGas Molecules into aLiquid Phase
Concentration ofGases inSolutions
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 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 con-
tainer in litres. R is the ideal gas constant which in SI units is 8.314JK1mol1 or
in conventional units: 62.36mmHgK1mol1
At a constant temperature, Eq. (1.2) simplifies to P n / V . Concentration is
defined as moles/unit volume, i.e. n/V; hence pressure is proportional to gas concen-
tration, i.e. the greater the number of gaseous molecules in a given volume, the
more force those gas molecules will exert on the walls of the container. The physical
reaction of dissolving in solution is also proportional to the partial pressure of the
gas above the solution, so that for oxygen dissolution [12]:
K Forward
O2 ( gas )

O2 ( dissolved ) (1.3)
K Reverse

[O2 ](gas) K Forward = [O2 ](dissolved ) K Reverse (1.4)

[O2 ](dissolved ) = SC [O2 ](gas) (1.5)

The rate constant KForward in Eq. (1.4) describes the proportion of oxygen gas that
dissolves per unit time, while KReverse describes the proportion of the dissolved con-
centration that leaves the solution to the gas phase. When the system described
in Eq. (1.4) is at thermodynamic equilibrium, the concentrations in the gas and
liquid phases are stable and the constants may then be combined, resulting in
Eq.(1.5), also known as Henrys Law. SC is the Bunsen solubility coefficient where
SC=KForward/KReverse and is gas- and solvent-specific. SC is affected by other dissolved
solutes and falls with increasing temperature (i.e. KForward becomes smaller and
KReverse larger).
Due to difficulties in measuring the molar concentration of oxygen compared to
measuring a volume of 100% oxygen at standard conditions (STPD: 0C, 760mmHg,
dry gas), it is customary to report oxygen content in mLdL1. Under these conditions,
oxygen approximates an ideal gas such that 6.021023 gas molecules (i.e. 1mol)
occupy 22.414L at 0C.Quantification of human oxygen consumption is performed
using STPD rather than BTPS (body temperature and pressure, saturated: Eq. (1.7))
[13] as water vapour in the latter partially condenses with increasing pressure. This
vapour results in a significant deviation from an ideal gas and invalidates the relation-
ship between the number of molecules and volume defined in Eq. (1.2).
The Solubility ofRespiratory Gases inSolution
The Bunsen solubility coefficient of oxygen is 0.003082mLdL1mmHg1 and

describes the measured solubility corrected to STPD.Utilising this conversion, the
solubility coefficient of oxygen in normal plasma is 1.38103mmolL1mmHg1
while carbon dioxide is nearly 22 times more soluble at 3.08102mmolL1mmHg1
[14]. Thus, using Henrys Law (Eq.1.5) 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 mmolL-1 Or 0.0278 mLdL-1
(1.6)
[CO2 ] = 3.08 10-2 40 mmHg
= 1.232 mmolL-1 Or 2.7 mLdL-1
It should be noted that this does not include oxygen and CO2 in chemical equilibrium
with the dissolved gas such as that combined with haemoglobin or in reaction with
water. The significantly higher plasma concentration of dissolved carbon dioxide
(Eq.1.6) resulting from its greater solubility allows for more rapid elimination by gas
exchange membranes when compared to oxygen under the same flow conditions.
In the gaseous phase, the partial pressures of individual gases combine to equal the
total ambient pressure that the gas mixture exerts on its container, allowing each gas
to be reported as a fraction of the total. For example, the partial pressure of oxygen in
inhaled 37C air that is fully saturated with water vapour at 1atm. (i.e. BTPS) is:
PIO2 = ( 760 mmHg - 47 mmHg ) FiO2
(1.7)
= 149.7 mmHg for an FiO2 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 dissolved gas
in a given quantity of solution that is in contact with a gas phase has no such equiva-
lent summation, i.e. the summation of the partial pressures in solution will not equal
atmospheric pressure.
Since the partial pressure of a gas is proportional to the concentration in solution

(at equilibrium), 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 pressuresapart from
measurement techniques, it allows easy quantification of the concentration gradient
from the site of gas exposure to the site of usage. Its inconvenience comes
when considering stoichiometric relationships such as described under respiratory
quotient above.
Biophysics ofMembrane 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 materials, a gasblood interface is usually prevented by forma-
tion of a biofilm comprising adsorbed blood proteins that forms after a short period
of operation. These factors impose a time constraint in 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
Ficks Law of Diffusion [15]:
DC
J = -D A (1.8)
Dl
This expression says the flux (J, mmols1) of a solute over the thickness of a mem-
brane (l, cm) is proportional to the diffusivity coefficient, D, the concentration
gradient across the membrane (C, mmolmL1 or mmolcm3) 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 [15]. This universal statement of
mass transport is applicable not only to the extracorporeal oxygenator, but also of
oxygen moving from plasma to interstitial fluid and into cells.
The diffusivity coefficient, D, is expressed as area over time (cm2s1) and describes
a unique constant of the 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 tem-
peratures provide gas molecules with greater kinetic energy, increasing diffusion
rates [11]. This is in contrast to the solubility of the gas 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 mediasuch as membranes in hollow fibre oxy-
genatorsrequires more parameters to be incorporated into the constant, resulting
in effective diffusivity, DEff, that for an isolated membrane has the following
parameters:
Dd
DEff = e (1.9)
t
D is the diffusion coefficient described above for the gasliquid interface within the
pores; is the porositythe fraction of the interface occupied by gas; is the tortu-
ositya geometric description accounting for the increased length of diffusion
within the membrane; and is constrictivity, which describes resistance to mole-
cules traversing pores due to their size relative to the pore diameter [16, 17].
Figure1.3 illustrates a section of membrane and the diffusion path within.
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, 16]. 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 termed
an enhancement factor. After incorporating haemoglobin, the DEff for oxygen
depends not only on the membrane characteristics discussed, but is also a function
of haemoglobin concentration (the haematocrit, %Hct). Equation (1.10) demon-
strates an example of a term for the effective diffusivity of a membrane exposed to
a turbulent bovine blood stream [18, 19]:

DEff = ( 2.13 - 0.0092 %Hct ) 10 -5 ( cm 2 -1
s ) (1.10)
The Driving Force forDiffusive Transport ofGases
By combining Eq. (1.5) (Henrys Law) for solute concentration and Ficks
Law (Eq.1.8) an equation for diffusive membrane flux can be derived where k
is the permeability constantthe product of the solubility coefficient (SC,
mmolL1mmHg1) already defined and effective diffusivity (DEff, cm2s1) having
the units molcm1s1mmHg1 [20]:
PO2 ( Gas) - PO2 ( Plasma )
J = - k O2 A (1.11)
Dl
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 membranes operational life due to macroscopically visible
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 -kO into a single constant, resulting
2
in a quantitative statement that the flux of gas is proportional to the driving force
and is opposed by certain resistances [15] and should be familiar as equivalent to
statements of resistance:
Driving Force
Mass transfer per unit area =
Resistance to Transport
J (
PO2 ( Gas) - PO2 ( Plasma ) )
( mmols -1
cm -2 ) A
=
RTotal
(1.12)
Dl
where RTotal =
- kO 2

From Eq. (1.12) it is apparent that maintaining the partial pressure gradient between
the gas phase and the blood phase is important to maximise oxygen flux. Maintenance
of the local pressure gradient at any point along a hollow fibre is influenced by three
design factors and one operational factor: the time the blood is exposed to the mem-
brane, relative flow direction, local turbulence and haematocrit.
Membrane Exposure Time
For a static liquid below a static gas, the rate of diffusion will decrease exponentially
until equilibrium when the conditions of Eq. (1.5) are met, i.e. the rate of gas dis-
solving into the liquid is equal to the rate of molecules leaving the liquid (Fig.1.4a).
It is clear from this figure that too short an exposure time will result in submaximal
oxygenation.
The more complicated relationship when haemoglobin is present is displayed
in Fig.1.4b. Here the oxygen content of the blood displays a plateau due to the
oxygenhaemoglobin 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
likely to be encountered is an important design parameter. With current hollow fibre
oxygenators this length is around 4cm for adequate oxygenation, with shorter
lengths needed for carbon dioxide removal (see ECCO2R below).
a b
14.00
Partial Pressure vs Time for a static gas/liquid O2 Content vs Membrane Exposure Time
12.00
Partial Pressure of Gas Phase

10.00
Partial Pressure Gas
Partial Pressure Liquid Phase
O2 Content mL/dL
8.00
Oxygen Content (ml/dL) for Hb 10 g/dL pH
6.00 7.44 Temp 37.5 & BE 0
Dissolved O2 in solution
4.00
2.00
0.00
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: (a) Depicts partial pressures
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. (b) 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 oxygenators. Data
derived from Katoh [18]
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 blood stream.
ffect ofTurbulence andHaematocrit onLocal

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

haemoglobin maintains diffusion in plasma toward red cells [14] and augmenting
haematocrit increases the flux of oxygen into the blood (Fig.1.4b). Creation of
turbulent flow vortices more effectively purges tubule contents than laminar flow
(discussed below) and brings erythrocytes into closer proximity to gas exchange
membranes, increasing the local concentration gradient.
Resistances toDiffusive Transport
The total resistance to mass solute movement in Eq. (1.12), RTotal, is the sum of
component resistances, which can be divided into gas phase resistance (RG), mem-
brane resistance RM, and blood side resistance, RB: RTotal=RG+RM+RB. Of these RG
is negligible and the factors influencing RM have been extensively discussed. The most
variable component is RB due to the formation of a stationary film on the blood
side of the membrane (Fig.1.3 boundary layer).
Area oftheGas Exchange Membrane
It is not practical for manufacturers to specify a membrane surface area (A, Eq.
(1.12)) that encompasses the complex microscopic geometry of a membrane and its
pores, the latter being incorporated into DEff as discussed [16]. Furthermore, it can-
not be guaranteed that any oxygenator design utilises the entire membrane area
evenly. Thus, while flux per unit area (J/A) is a useful description of isolated mem-
brane performance, oxygenators are better characterised by their total flux by incor-
porating area into the equation for resistance:
(P - PO2 ( Plasma ) )
( mL min ) J
-1 O2 ( Gas )
Oxygenator =
RTotal
(1.13)
Dl
where RTotal =
- kO 2 A

Although lacking the precision of Eq. (1.12) in defining properties of the membrane,
Eq. (1.13) can be incorporated into monitoring gas exchange efficiency of an indi-
vidual oxygenator over time and will be discussed after oxygen carriage is consid-
ered. The value of JOxygenator is usually reported in oxygenator product specification
sheets at varying blood flows.
Ultrafiltration ofPlasma Water Over theOxygenator Membrane [15]
Analogous to the membrane flux of oxygen down a concentration gradient is the

movement of water from the plasma to the gas phase. The driving force here is
hydrostatic pressure, rather than concentration gradient, and is generally defined as
a conductance (the inverse of resistance) termed the coefficient of ultrafiltration
(KUF, mLmin1mmHg1):
water flux
K UF =
transmembrane pressure
(1.14)
QF
=
PBlood - PGas

QF (mLmin ) is termed the ultrafiltration rate and describes the appearance of fluid
1
within the gas containing hollow fibres, while the terms for pressure (PBlood and Pgas)
describe the heights of a fluid column relative to atmospheric pressure in each com-
partment. Many of the factors already described for resistance to diffusion will be
contained in the ultrafiltration coefficient and will not be discussed further. The ultra-
filtrate represents a homogenous fluid that will contain dissolved solutes from plasma
proportional to the size of membrane pores, which are typically smaller than 1m.
A major drawback of early microporous membranes was significant plasma leak-
age, as the open porous structure allowed significant ultrafiltration. This has been
significantly alleviated by newer closed-fibre membranes; however, some water
flux still occurs under normal operating conditions. The water then 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 [10].
Oxygen Transport
As outlined above, oxygen has a limited solubility in plasma of 1.39103mmolmmHg1

or 0.0031mLdL1mmHg1 at 37C.For a normal arterial oxygen partial pressure
of 100mmHg this equates to about 3mL of dissolved oxygen per litre of blood.
At that oxygen content, maintaining a nominal body oxygen consumption of
250mLmin1 would require a cardiac output of 80120Lmin1 and even breathing
100% oxygen at normal atmospheric pressure would not sustain aerobic cellular
metabolism, providing only 20mL of oxygen per litre of blood [14].
Oxygen Carriage
Oxygen carriage in blood is massively increased 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. This rapid
structural alteration changes the affinity of haemoglobin for oxygen, resulting in
higher affinity in areas where oxygen is abundant and lower affinity in areas of
oxygen consumption. A normal adult has a haemoglobin concentration of between
12 and 16gdL1. However, it is quite common for anaemia to be present in criti-
cally ill patients, and those receiving ECMO are often transfused to maintain a haemo-
globin concentration around 10gdL1, equating to 1.55mmolL1 (assuming a
molecular weight of haemoglobin of 64,458gmol1) [21].
Haemoglobin is a spherical molecule consisting of four globin subunits (two

and two chains) with each globin containing a haem group in a peripheral molecu-
lar crevice. Each haem molecule consists of a central iron atom in the ferrous (Fe2+)
state between two histidine amino-acids. This structure allows the iron atom to bind
oxygen without being oxidised to Fe3+, a change that would prevent further oxygen
binding. Haemoglobin demonstrates cooperative binding whereby the binding of
oxygen to the iron moieties is enhanced if another binding site on the same molecule
is already occupied by oxygen. As each haemoglobin molecule has four binding
sites, it can exclusively be 0, 25, 50, 75 or 100% oxygenated. Haemoglobin satura-
tion refers to the fractional occupancy of all the oxygen binding sites in a solution,
and due to cooperative binding results in the sigmoid haemoglobin dissociation
curve (Fig.1.5) [3].
Apart from the conformational change induced by oxygen itself, four other major
factors influence the conformational state of haemoglobincarbon 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 h aemoglobin
in a given quantity of blood is with oxygen at any partial pressure, and thus the oxy-
gen content of that blood. If all haemoglobin binding sites are occupied by oxygen
Oxygen Hemoglobin Dissociation and Oxygen Content

100%
20.00
90%
80%
70% 15.00
Blood Oxygen Content ml/dL
Hemoglobin Saturation %
60%
50%
10.00
40%
SaO2 at pH 7.44 Temp 37.5 & BE 0

30%
SaO2 at pH 7.32 Temp 37.5 & BE 0
5.00
Oxygen Content (ml/dL) for Hb 10g/dL
20% pH 7.44 Temp 37.5 & BE 0
Oxygen Content (ml/dL) for Hb 10g/dL
pH 7.32 Temp 37.5 & BE 0
10%
0% 0.00
0 20 40 60 80 100 120
Partial Pressure Oxygen (mmHg)
Fig. 1.5 The oxygenhaemoglobin dissociation curve as calculated by the Thomas modification
of the Kelman Eq. (1.9). Also shown is the oxygen content for haemoglobin concentration of
10gdL1 after applying Eq. (1.15)
(100% saturation) the maximum oxygen carrying capacity is 1.39mL per gram of
haemoglobin in adults [13] or 1.312mL per gram of foetal haemoglobin [14]. Thus,
the total oxygen content of adult arterial blood (CaO2) can be described by:

(
CaO2 = Hb SaO2 1.39 + 0.0031 PaO2 mLdL-1 ) (1.15)
The total oxygen content can then be multiplied by blood flow (cardiac output) to
give oxygen delivery, DO 2 which has the units of flux: mLmin1. Note the scaling
factor of 10 to convert the units of oxygen content to mLL1.
DO2 = Q B CaO2 10 (1.16)

Modulation ofHaemoglobins Affinity forOxygen
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 ferrous elements in deeper crypts and facilitating release of oxy-
gen from haemoglobin. Similarly, increasing temperature and increasing hydrogen
ion concentrations stabilise haemoglobin in the deoxygenated state. Similar to oxy-
gen, carbon dioxide binding is reversible and in compartments with a low CO2 con-
centration 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 will be discussed later in the chapter.
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 [9]. This has the most significance in transfused blood where, after 2 weeks
of storage, 2,3-DPG levels become negligible. After transfusion of this blood DPG
levels do not return to normal until nearly 48h [22]. In the absence of 2,3-DPG the
affinity for oxygen is increased resulting in less oxygen delivery at the same periph-
eral PO2. As will be seen, this is only likely to be a factor in the most severe oxygen-
ation problems and the benefit of increased oxygen binding sites from transfusion
will outweigh the transiently lower delivery (see Figs.1.8 and 1.9).
The OxygenHaemoglobin Dissociation Curve
The oxygenhaemoglobin dissociation curve is characterised by an upper plateau at

higher partial pressures of oxygen where haemoglobin is between 90 and 100%
saturated. Below this is a steep shoulder where the saturation of haemoglobin rap-
idly decays as the partial pressure of oxygen falls. Physiologically the factors that
shift the curve to the right (i.e. to a lower affinity state) are local tissue metabolism
and hence local oxygen consumption (Fig.1.5).
Several equations exist to model the normal oxygenhaemoglobin dissociation
curve. One of the most informative is the Thomas modification [23] of the Kelman
[24] equation and its inverse [25], which calculate the haemoglobin saturation for
any partial pressure of oxygen and allow for shifts of the curve due to temperature,
[H+] and carbon dioxide.
For convenience, the oxygenhaemoglobin dissociation curve is frequently
described by the partial pressure at which 50% of haemoglobin is saturated. A nor-
mal P50 for arterial blood is 26.3mmHg. Values higher than this describe a right-
shifted curve, i.e. a haemoglobin with lower affinity for oxygen. Two curves of
haemoglobin saturation are demonstrated in Fig.1.7, 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 differ-
ence in haemoglobin saturation is minimal. In contrast, there is a significant differ-
ence in the saturation of Hb at a partial pressure of 40mmHg commonly found in
venous blood.
The implications of the oxygen Hb dissociation curve become clearer when
oxygen content is also plotted on the same chart (Fig.1.5). At haemoglobin of
10gdL1 the oxygen content at a partial pressure of 100mmHg is 13mLdL1 and
minimally affected by the arteriovenous pH difference. However, at an oxygen
partial pressure of 40mmHg, a 1mLdL1 difference becomes apparent between
the two content curves, being 10mLdL1 at pH of 7.44 and 9mLdL1 at pH of 7.32.
In the tissues a drop in content can only occur if the oxygen is utilised by metabo-
lism, so this right-shifting of the curve as the products of metabolism acidify
capillary blood serves to bolster the partial pressure gradient for diffusion from
capillary to cell.
It should be noted that right-shifted oxygenhaemoglobin dissociation curves,
while advantageous for unloading oxygen in acidotic tissues, may be counterpro-
ductive at sites of oxygen uptake if abnormally low alveolar partial pressures exist.
Inspection of Fig.1.5 reveals that if oxygen uptake were to occur at a partial pressure
of 60mmHg, a right shifted curve (pH7.32) will carry 0.5mLdL1 less oxygen.
This gap widens if uptake occurs at even lower partial pressures.
Mixing Blood ofDiffering Oxygen Partial Pressures
The oxygenhaemoglobin dissociation curve is of importance when considering

mixing blood streams with differing oxygen concentrations. If the volumetric 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 tension and satu-
ration 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 (Table1.1).
Table 1.1 Mixing blood streams with differing oxygen concentrations

O2 Tension O2 Content
Equal flows Volume (mL) Hb (g) (mmHg) Saturation (%) (mL)
Venous blood 100 10 40 76 10.07
Oxygenated blood 100 10 100 98 13.09
Mean (for comparison) (70) (86.8)
Mixed blood 200 20 23.17
(=2dL) Final O2 content (mLdL1) 11.58
Final saturation (%) 87
Final partial pressure (mmHg) 51.7
O2 Tension O2 Content
Unequal flows Volume (mL) Hb (g) (mmHg) Saturation (%) (mL)
Venous blood 3000 300 30 58 236
Oxygenated blood 2000 200 300 100 287
Mean (for comparison) (165) (79.0)
Mixed blood 5000 500 523
(=50dL) Final O2 content (mLdL1) 10.5
Final saturation (%) 77
Final partial pressure (mmHg) 41
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 blood stream 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 10gdL1 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 1dLmin1 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
As demonstrated in Table1.1, 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
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 may
result in suboptimal oxygen delivery.
Veno-venous ECMO andOxygen Transport
Having covered oxygen carriage and mixing blood with differing oxygen content,
veno-venous (VV) extracorporeal membrane oxygenation can now be considered.
As blood is accessed and returned to the venous system/right atrium (Fig.1.6),
Fig. 1.6 Schematic of a VV-ECMO circuit. It will be assumed that both femoral and internal jugu-
lar access exists so all venous return is modeled with one vessel. A mixture of fully oxygenated and
venous blood enters the right heart and perfuses the pulmonary circulation before the left heart
distributes it systemically (hence end-tidal CO2 will not reflect mixed venous PCO2). Varying with
the cannula position and venous return (cardiac output), some amount of recirculation is very com-
mon, reducing the amount of oxygen delivered. Depending on recirculation, it may not be possible
to sample true mixed venous blood
systemic oxygen delivery remains dependent on cardiac output, making this system
simpler to quantitatively analyse.
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
along the inferior vena cava. This access cannula typically ends around 10cm below
the cavo-atrial junction, while the return cannula has a single terminal orifice in the
right atrium. If higher extracorporeal circuit flows are attempted, the inferior vena
cava may collapse around the multistage cannula, intermittently obstructing flow
and causing the external circuit to shudder, with consequent hypoxia. If these
higher circuit flows are necessary to achieve adequate oxygenation and this negative
access pressure cannot be resolved by giving fluid, a second access cannula may
need to be placed in the internal jugular vein.
Figure 1.6 demonstrates a basic circuit. It can be appreciated 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 supported 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.
The patient has the following parameters: cardiac output of 4Lmin1, ECMO
flow of 4Lmin1, 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 7Lmin1 or fall to 2.5Lmin1?
Apart from the non-contribution of the lungs, other assumptions include an
oxyhaemoglobin 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 param-
eters are required: the haemoglobin concentration and the patients total oxygen con-
sumption (VO2). Initially, this hypothetical patient has a haemoglobin concentration
of 10gdL1 and is at steady state consuming 250mLmin1 of oxygen with no mark-
ers of tissue hypoxia. Thus, there are four independent variables: the cardiac output,
the ECMO flow rate, the haemoglobin concentration and the target VO2.
While providing circulating oxygen 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 said to be 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, oliguria and ris-
ing lactate as anaerobic metabolism ensues.
Figure1.7a suggests that under these conditions, increasing the cardiac output
from 4 to 7Lmin1 will cause a drop in the arterial saturations from 99 to 84%.
In contrast dropping the cardiac output to 3Lmin1 will not affect the arterial satu-
ration but will cause the central mixed venous saturation to fall from 52 to 25%.
The equations required to generate this model are the content and delivery equa-
tions (Eqs.1.15 and 1.16) 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). It should be noted that
in VV-ECMO the pulmonary artery is not the correct sampling site for mixed venous
saturation. Instead, the pre-oxygenator blood is the best approximation (but depend-
ing on the vessels cannulated may not include superior vena cava flow).
The methods for mixing blood streams (Table1.1) are used to determine the pulmo-
nary artery oxygen content and saturations. In the absence of any lung function this is
modeled as the systemic arterial oxygen delivery. 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 extractionthis determines the venous oxygen
flux to the right heart. Example values from Fig.1.7a are shown in Table1.2.
rterial Saturations Are Dependent ontheFraction

A
ofCardiac Output Captured
To explain the fall in arterial saturations with increasing cardiac output it should be
appreciated that in this case the required VO2 did not change. What did change was
the total venous return, which increased by 75% with the increased cardiac output.
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 ECMO flow conditions and haemoglo-
bin. See text for discussion. (a) VV-ECMO blood flow 5L.min1, body oxygen consumption 250ml.
min1 and hemoglobin 10g.dl1. (b) VV-ECMO blood flow 3L.min1, body oxygen consumption
250ml.min1 and hemoglobin 10g.dl1. (c) VV-ECMO blood flow 4L.min1, body oxygen consump-
tion increasing with cardiac output, hemoglobin 10g.dl1. (d) VV-ECMO blood flow 4L.min1, body
oxygen consumption 250ml.min1 and hemoglobin 7g.dl1
The ECMO flow remained at 4Lmin1 so the ECMO shunt, i.e. the fraction of
deoxygenated venous blood in the right ventricle increased from 3 to 43%. The
shunt is calculated as (Cardiac outputOxygenated Blood flow)/Cardiac output
where oxygenated blood flow equals the cardiac output if the cardiac output is less
than ECMO blood flow. The oxygenated blood volumetric flow will equal ECMO
set flow minus any recirculation when cardiac output is greater than ECMO flow.
In summary, the resulting arterial saturation is analogous to the example of mixing
blood streams 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 reality, increases
in cardiac output are likely to be accompanied by an increased VO2 and in this
setting the mixed venous saturation will fall (Fig.1.7c).
Table 1.2 Solutions for Fig.1.7a
Flux of
fully
Oxygenated oxygenated Cardiac Fraction of
blood flow blood output/ Venous O2 venous
ECMO (factors delivered venous flux return (deoxygenated) Total O2 flux CaO2-
set rate recirculation) by ECMO return to heart CvO2 SvO2 blood leaving leaving RV CaO2 SaO2 CvO2 VO2
(Lmin1) (Lmin1) (mLmin1) (Lmin1) (mLmin1) (mLL1) (%) RV (%) (mLmin1) (mLL1) (%) (mLL1) (mLmin1)
4.0 1.5 201.0 1.5 0.0 0.0 0 0 201.0 134.0 100 134.0 201.0
1 Physiology ofExtracorporeal Life Support (ECLS)
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 10gdL1. Oxygenated blood flow refers to the volumetric flow of oxygenated blood into the right heart. For
calculation purposes oxygen content is expressed in mLL1 rather than mLdL1
25
ixed Venous Saturations Are Determined by VO2

M
andCardiac Output
To appreciate why the mixed venous saturations fall to the left of Fig.1.7a (where VO2
was held constant) when the cardiac output falls, recirculation between the access
and return cannula must be considered. Recirculation is an additional variable to the
independent variables identified above.
Several factors influence recirculation; however, this model incorporates only
cardiac output: if the heart stops, there will be no venous return and recirculation
might approach 100% (if the great veins do not collapse), i.e. blood will enter the
ECMO circuit from the IVC and return via the right atrium, flowing in a retrograde
fashion back down the IVC, resulting in no systemic oxygenation (amongst other
adverse effects!). Similarly at low cardiac outputs, the blood not ejected will recir-
culate; however, it becomes saturated with oxygen on the first pass and will be
unable to carry more. This leads to an important point: when the ECMO flow falls
below cardiac output, the amount of oxygenated blood ejected by the right heart is
cardiac output-limited. Conversely, at high cardiac outputs and hence high venous
return, recirculation will be minimal as the oxygenated blood will be washed
through the right atrium into the right ventricle.
Under conditions where the cardiac output is less than the ECMO flow, the blood
entering the arterial circulation is fully saturated with oxygen. However, to supply
the required VO2, oxygen extraction from the low output has to increase, and hence,
mixed central venous oxygen content and venous saturation fall.
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 ECMO fraction of cardiac output mecha-
nism described above AND increased extraction to achieve the higher VO2
requirement.
The Effect ofOxygen Carrying Capacity
The parameters in Fig.1.7d are identical to Fig.1.7a except for a lower haemoglo-
bin, so these graphs illustrate the effect of haemorrhage and transfusion. The lower
oxygen-carrying capacity means that at any cardiac output, the oxygen 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 extraction can occur
and the achieved VO2 falls lower than the target VO2under 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 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. This is in contrast to venous saturations (Fig.1.9)
which 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.
VV-ECMO Arterial Saturations at ECMO Flow 4 L/min and VO2 250 ml/min
100.00%
90.00%
80.00%
90.00%-100.00%
Arterial Sat
70.00% 80.00%-90.00%
70.00%-80.00%
60.00%
60.00%-70.00%
50.00% 50.00%-60.00%
40.00%-50.00%
40.00%
1.50
2.50
3.50
4.50
5.50
6.50 Hb 13
Hb 12
Cardiac Output L/min Hb 11
7.50 Hb 10
Hb 9
Hb 8
Hb 7 /dL
8.50 Hb 6 bin g
oglo
Hb 5 Hem
Fig. 1.8 Arterial oxygen saturation vs. haemoglobin and cardiac output in a patient receiving
VV-ECMO.VO2 is held constant
VV-ECMO Venous Saturations at ECMO Flow 4 L/min and VO2 250 ml/min
100%
90%-100%
90%
80%-90%
80%
70%-80%
70%
60%-70%
60% 50%-60%
Venous Sat
50% 40%-50%
40% 30%-40%
30% 20%-30%
20% 10%-20%
10% 0%-10%
Hb 11
0%
dL
1.50 2.00 2.50 Hb 8 g/
3.00 3.50 4.00
4.50 5.00 5.50 b in
6.00 6.50 7.00 Hb 5 lo
7.50 8.00 8.50 og
Cardiac Output L/min e m
H
Fig. 1.9 Central mixed-venous oxygen saturation vs. haemoglobin and cardiac output in a patient
receiving VV-ECMO.VO2 is held constant
Further Consideration ofRecirculation
In the above analysis, recirculation is primarily considered in terms of the cardiac

output and is considered as 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 catheters being in too close proximity to each other or suboptimal posi-
tioning (e.g. jugular 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 certainly result in some retrograde flow in the great veins through part of the
cardiac cycle.
More importantly is the effect of a patent tricuspid valve, as in general the
tricuspid valve will be closed for most of ventricular systole. Although it is con-
ceivable that high ECMO flows directed at the valve may cause early tricuspid
opening this is unlikely in patients where right ventricular hypertrophy or systemic
inotropes are augmenting right heart function. In the period of tricuspid closure the
continuous ECMO flow can only recirculate through the right atrium and vena cava
(Fig.1.10).
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
The Importance ofMixed-Venous Saturation
The ability to conveniently measure the arterial oxygen saturation non-invasively in

real time via finger oximetry is of considerable clinical importance. However, in the
setting of ECLS it is important to understand what the number represents.
It is common practice to target arterial oxygen saturations of ~90%, this being
the shoulder of the oxygenhaemoglobin dissociation curve below which the
partial pressure of oxygen falls more steeply. In the general ward setting this is a
robust target, as achieving it generally implies adequate oxygen delivery, while
exceeding it may have adverse consequences for selected subgroups of patients. In that
context, failing to achieve a saturation of 90% necessitates a review regarding esca-
lation of support, with the likelihood that only a short time remains to remedy the
situation.
What is being measured 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 requiring ECLS the peripheral
circulation may lack detectable pulsatility and these monitors will struggle to give
an accurate reading. In these situations, blood gas monitoring (which usually
provides a calculated saturation based on the partial pressure, temperature, pH and
CO2) is essential.
In the ECLS setting, the primary concern is oxygen delivery and optimising this
within the limits of the available system. Equations1.16 and 1.17 describe oxygen
content and delivery. From these equations it can be appreciated that increasing the
haemoglobin concentration can accommodate lower arterial saturations yet still
provide the same oxygen content and thus delivery. This is important to recognise
when other limitations such as ECMO flow rates or problems with recirculation
prevent higher saturation.
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 sampling
blood from one of these sites, the central venous oxygen content (CvO2) can be
calculated, again using Eq. (1.15). This concept has led to classic descriptions of
supply-dependent and independent oxygen uptake and it is usually informative
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
(even from an optimal sampling site) is that it does not represent the lowest satura-
tion (see mixing blood of differing content above). Different organs (or 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 pressure 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 lactate 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 then 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.
Limitations Imposed onVV-ECMO Oxygenation
In patients with severely reduced lung function, high cardiac output or poor
VV-ECMO circuit flows may not be able to oxygenate an adequate amount of the
cardiac output to achieve normal arterial oxygen saturations. This may be
improved by adding a second inflow cannula, thus allowing increased circuit flows
without venous collapse. However, if this is not technically possible then targeting
a lower arterial saturation may be necessary.
Carbon Dioxide Physiology
An adult consuming 300mLmin1 of oxygen with a respiratory quotient of 0.833

will produce 250mLmin1 or 360Lday1 of carbon dioxide via aerobic metabo-
lism and this volume must be removed from the body. The properties of carbon
dioxide make its clearance more favourable than oxygen absorption. In ECMO,
steps are occasionally required to avoid excessive hypocapnia.
Carbon Dioxide Transport inPhysical Solution
The solubility of carbon dioxide in plasma is described by Henrys Law (Eq.1.5)

with a value for SC at 37C of 0.0308mmolL1mmHg1 [14]. At a PCO2 of
40mmHg this equates to 1.232mmolL1 or 2.7mLdL1 of dissolved carbon
dioxide. The solubility decreases to 2.88102mmolL1mmHg1 at 40C.
Like oxygen, only a small proportion of carbon dioxide is transported as dis-
solved gas; the remainder being in chemical equilibrium with this phase through
various reactions with water for a total blood content of nearly 500mLL1. In
blood, these reactions largely occur in erythrocytes where the enzyme carbonic
anhydrase (C.A.) is abundant [26]; however, C.A. is distributed in many other tis-
sues and in pulmonary and renal capillaries [14].
Reactions ofCarbon Dioxide inSolution
The following reactions describe the hydration of dissolved carbon dioxide

with arrows, indicating where the majority of reactants are at equilibrium. It
should be noted that 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 (largely on haemoglobin) is also
listed here:

CO2 ( gas ) + H 2 O
CO2 ( dissolved ) + H 2 O a. Dissolving in water
CO2 ( dissolved ) + H 2 O
H 2 CO3
C. A .
C. A .
b. Formation of carbonic acid
CO2 ( dissolved ) + OH -

C. A . -

C. A .
HCO3 c. Reaction with hydroxyl ion
H 2 CO3
H + HCO3
+ -
d. Dissociation to bicarbonate
CO2 ( dissolved ) + R - NH 2 e. Formation of carbamino
- +

R - NH - COO + H groups
(1.17)
The equilibrium constant for the dissociation of carbonic acid to bicarbonate is so

small (61011molL1) that the carbonic acid concentration is near undetectable. 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 (Eqs.1.17bd) in the body can be summarised
into a single equation and the equilibrium constants combined (K) such that [12]:
CO2 ( dissolved ) + H 2 O

+ -
H + HCO3 a. Summary Equation
K
[CO2 ] = HCO3- b. Equilibrium Reaction
H
+
(1.18)
HCO3-
pH = pK + log c. Logarithmic Form
[CO2 ]
Where [CO2 ] = 0.0308 PCO2 at 37C. ( Henry s Law )

K now describes the compound equilibrium of multiple reactions and is derived
experimentally. [H2O] is notably excluded from Eq. (1.18b) because it is incorpo-
rated into K. This is permissible because its 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.1
Although less quoted than the logarithmic form (the HendersonHasselbalch
equation), Eq. (1.18b) is most useful in understanding carbon dioxide in body
fluids. For plasma with a [H+] of 3.98108molL1 (pH of 7.4) at 37C, the
apparent K is 8.13107EqL1mmHg1 (pK 6.09) and thus K/[H+] is around
20, i.e. 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 and so the total CO2 contained in blood is
largely stored as bicarbonate ions in plasma and erythrocyte water (Fig.1.12).
Effect oftheHydrogen 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 Eq. (1.18a), CO2 is the only independent variable with the
amount in the body being determined by the balance of production and flux out of
the body [12]. The [H+] depends not only on the reactions of carbon dioxide with
water but also on the concentrations of other strong and weak electrolytes with the
final balance being determined by the need to maintain electrical neutrality and the
1
Occasionally the term PCO2 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 Henrys Law without alteration of the
solubility constant (Eq.1.5).
Fig. 1.11 Dissolved carbon dioxide diffuses across all tissue planes and dissociates in each com-
partment to HCO3 with the ratio of HCO3 to CO2 depending on the [H+]
ionisation constant of water [12, 27]. The simplest summary of this complex
interaction is that bicarbonate behaves as an electrical spacer in the following
equation for electrical neutrality2:
2
Being electrical neutrality, the units are available charge (milli-equivalents per litre mEqL1);
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.
Na + + K + + 2 Ca 2 + + 2 Mg + + z +
= Cl - + HCO3 - + 1.8 PO 4 2 - + 0.28 Albumin - + lactate - + x -

(1.19)
This occurs because of the large supply of dissolved CO2 throughout the body and
the ready reversibility of its reaction with water. In contrast, the other charged
species in Eq. (1.19) are all tightly regulated by other cellular processes and can-
not change quickly (Fig.1.11). Thus if lactic acid production were to rise, the
immediate effect is a shift of [HCO3] to CO2 to maintain electrical neutrality;
over time renal chloride loss is increased to restore balance. These mechanisms
contribute to understanding the alkalotic effects of hypoalbuminemia and acidify-
ing effects of hypernatremia (dilutional acidosis) commonly encountered in criti-
cally ill patients. The reader is referred to other sources for a more detailed
discussion [12, 27].
Carbonic Anhydrase
Carbonic anhydrase is present in many tissues, including erythrocytes and

pulmonary capillaries, but not in blood plasma [14]. As a consequence most of
the above reactions occur at a faster rate within erythrocytes and a membrane-
bound Cl/HCO3 exchange protein allows erythrocyte bicarbonate to diffuse
through plasma water.
C.A. catalyses the reaction of carbon dioxide with water which otherwise has a
half-time for achieving equilibrium of around 15s [12, 28]. Inhibition of carbonic
anhydrase does not affect the final equilibrium concentrations of the Henderson
Hasselbalch equation; however, it may affect whether equilibrium is actually
reached in a body fluid compartment before diffusion of dissolved carbon dioxide
into surrounding compartments (Fig.1.9) occurs (recall HCO3 cannot easily cross
most lipid membranes). This is reflected in higher tissue partial pressures of carbon
dioxide if a total blockade of carbonic anhydrase is imposed [14].
The Arteriovenous-CO2 Difference andCarbamino Carriage
Figure1.12 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 produced after the addition of a strong
acid to a blood sample which shifts Eq. (1.18a) fully to the left [29]. In contrast the
bicarbonate reported on an arterial blood gas sample is calculated from the measured
pH and PCO2 using Eq. (1.18c) and is typically 24mmolL1 lower than total CO2.
30 35
Oxygenated Blood [CO2]Total (mmol.L-1)
AV Difference (mmol.L-1)
Components of Total CO2 by compartment (mmol)

Deoxygenated Blood [CO2]Total (mmol.L-1) 30
25
[CO2]Total (mmol.L-1 whole blood)

25
20
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 37C and haematocrit of 45%. Upper
lines represent dissociation of total CO2 in mmolL1 in deoxygenated and oxygenated whole
blood; the line AV Difference demonstrates increased carriage by carbamino groups as the oxygen
saturation falls. To convert mmolL1 to mLdL1 multiply by 2.226. Shaded lower areas demon-
strate the amount of CO2 (mmol) in erythrocyte water (325mL per L blood at PCO2 40mmHg) and
plasma water (509mL per L blood at PCO2 40mmHg). Dissolved CO2 is represented across all
water in 1L whole blood (836mL). Carbamino CO2 demonstrates the maximum amount (mmol)
that fully deoxygenated haemoglobin can carry. Summation of the amount of CO2 in mmol (shaded
areas) at any PCO2 divided by whole blood water (836mL) yields the concentration of CO2 per
litre. After [3034]
This difference is due to carbon dioxide carriage on amino acids and particularly
carbamino formation on haemoglobin (Eq.1.17e).
Inspection of Fig.1.12 is the dissociation curve for CO2 at a haematocrit of 45%
and demonstrates the PCO2 increases from 40 to 47mmHg as tissue metabolism
consumes oxygen. In this example the total CO2 rises correspondingly from
19.6mmolL1 (43.5mLdL1) to 22.4mmolL1 (49.8mLdL1) the difference
being due to tissue CO2 production and is proportional to the respiratory quotient
[30]. The proportion of CO2 carried as carbamino groups increases in this process
as oxygen unloading increases the affinity of haemoglobin for carbon dioxide (hae-
moglobin is more basic when deoxygenated, shifting Eq. (1.18b) toward bicarbon-
ate [14]). It should be emphasised that the blood bicarbonate is carried by blood
water which consists of intracellular and extracellular water [31].
Determinants ofCarbon 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 250mLmin1 in each system. The resistance to diffusive transport (RTotal) is
identical for the two membranes so that from Eq. (1.12):
(
J RTotal = Area PCO2 ( Plasma ) - PCO2 ( Gas) ) where RTotal =
Dl
- k O2
= Area1 D P1 Oxygenator 1 (1.20)
Area 2
= ( x D 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 (x=2) will double the required trans-
membrane 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. This often results in
hypocapnia and is avoided by reducing the partial pressure gradient of CO2 by either
adding carbon dioxide to the sweep gas or by reducing the sweep gas flow so that
the convection of carbon dioxide away from the membrane is slowed.
Extracorporeal Carbon Dioxide Removal: ECCO2R
ECCO2R refers to the support of hypercapnic respiratory failure, usually in the setting
of acutely decompensated pulmonary disease. Modern systems utilise a scaled
down VV-ECMO circuit utilising significantly smaller vascular access catheters.
While clearing carbon dioxide, the limited membrane area and lower blood flow
significantly reduce the oxygenation that can be achieved from these systems to
clinically negligible effects.
Blood Flow Requirements
Lower blood flows are made possible by the higher solubility of carbon dioxide and
its more linear dissociation curve across the physiologic range compared to the
sigmoid curve for oxygen saturation of haemoglobin (Fig.1.13). Complete satura-
tion of haemoglobin at a concentration of 10gdL1 with oxygen results in a
Fig. 1.13 Comparison of Oxygen and CO2 Content vs Partial Pressure

oxygen and carbon dioxide kPa
content vs. partial pressure 0.00 2.00 4.00 6.00 8.00 10.00 12.00
60.00
25.00
50.00
20.00
Amount in blood mmol/L

Content in Blood ml/dL
40.00
15.00
30.00
Oxygen Content for Hb 10g/dL
pH 7.44 Temp 37.5 & BE 0
Total CO2 ml/dL

10.00
20.00
5.00
10.00
0.00 0.00
0 20 40 60 80 100
Partial Pressure Oxygen (mmHg)
maximum oxygen carriage of approximately 13mLdL1 while venous blood with

a saturation of 70% has a content of 9.3mLdL1 (Eq.1.15, Fig.1.5). Using these
figures, 37mL of oxygen can be added to each litre of venous blood and thus an
adult consuming oxygen at 300mLmin1 would require an ECMO flow and cardiac
output over 7Lmin1. As discussed, lower cardiac outputs result in a fall in mixed
venous saturations that allows more oxygen uptake (Fig.1.7).
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
60mmHg equates to 52mLdL1. Hence, from 5dL of blood at this PCO2, 250mL
of CO2 could be removed by an efficient gas exchange system allowing an extracor-
poreal circuit running at 500mLmin1 to clear the CO2 load of adult patients.
Recirculation withSmaller Access Catheters
The example above assumes no recirculation, however, this is an increased risk with
the integrated vascular access devices used for ECCO2R.Significant recirculation of
blood that had already been completely cleared of CO2 will limit overall removal, and in
this setting increasing the ECMO flow rate may not influence the membranes expo-
sure 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 inECCO2R: Integration withPulmonary Oxygenation
Despite limitations imposed by haemoglobin saturation with oxygen, some improve-

ment in systemic oxygenation may occur with ECCO2R which can be explained by
three mechanisms [35]:
1. Flux of oxygen to the venous blood across the membrane still occurs as outlined
in the section on oxygen transfer.
2. Facilitation of mechanical ventilation strategies that maximise oxygen exchange
over carbon dioxide clearance.
3. Improving the partial pressure gradient for pulmonary oxygen exchange.
Ventilation Strategies withECCO2R
A significant mismatch between pulmonary ventilation and perfusion characterises

many forms of respiratory failure that are significant enough to require ECLS, with
a true pathologic shunt developing in more severe conditions such as acute respira-
tory 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 pulmonary blood flow through the
ventilated areas.
Applying relatively high positive end-expiratory pressure (PEEP) recruits
pulmonary surface area, particularly in dependent lung areas that receive more
blood flow [14], while avoiding barotrauma to the pathologically non-compliant
lung necessitates reduced tidal volumes (6mLkg1) targeted to achieve safe inspiratory
plateau pressures (<30cm H2O). This strategy improves oxygenation and decreases
mortality compared to high tidal volume ventilation [36, 35].
The high PEEP and consequent high mean airway pressure and coexistent
pulmonary hypertension resulting from hypoxic pulmonary vasoconstriction place
a significant load on the right ventricle which often already has impaired contractile
function from hypoxia and the acidosis imposed by hypercarbia. Right ventricular
dilation occurs and impairs left ventricular filling by shifting the interventricular
septum, and right heart output falls (Fig.1.14). The overall reduction in cardiac
output reduces systemic oxygen delivery with a high extraction ratio and low mixed
venous saturation (Eq.1.16). Inhaled pulmonary vasodilators can be added to
improve matching of pulmonary blood flow to ventilated areas and reduce the load
imposed on the right heart.
Fig. 1.14 Acute right ventricular (RV) enlargement with left ventricular (LV) compression 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
Ongoing oxygen consumption by the body maintains a pressure gradient between

oxygenated alveoli and venous blood in pulmonary capillaries allowing passive
diffusion to occur. This removal of gas from the alveolus permits convection of
further fresh gas into the alveolusso-called passive oxygenation. 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
ventilation 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 strategy;
however, in some patients the resulting arterial PCO2 is so high that significant
acidosis ensues. Institution of ECCO2R in these settings improves carbon dioxide
clearance and alleviates the hypercapnic acidosis. In other settings such as head
trauma, evidence suggests that maintaining normal PCO2 is optimal but this may not
be achievable with coexistent lung injury.
ECCO2R andAlveolar Oxygen Concentrations
Oxygenation utilising the low tidal volumes outlined above is further augmented by
maximising the oxygen partial pressure gradient between perfused alveoli and
pulmonary capillary blood to increase 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.1.7). The partial pres-
sure 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):
In Eq. (1.21), the Alveolar Gas Equation. PA denotes alveolar partial pressure,
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, Equation (1.21c) is
the most precise expression of alveolar gas and makes no assumptions about the
respiratory quotient [14].
PaCO2 1 - R.Q.
PAO2 = PIO2 - + PaCO2 FIO2 a.
R.Q. R.Q.
PaCO2
PAO2 PIO2 - b. (1.21)
R.Q.
PIO - PEO2
PAO2 = PIO2 - PaCO2 2 c.
PEO2

It is clear from Eq. (1.21b) that the alveolar PAO2 and thus the pressure gradient for
gas exchange falls with increasing PaCO2. Table1.3 demonstrates this effect at differ-
ent FIO2 levels. At a low FIO2, removal of CO2 with ECCO2R significantly increases
the partial pressure gradient for oxygen diffusion from the alveolus into plasma.
However, the effect becomes proportionally lower with increasing PIO2 and its
relevance in someone with a high FIO2 requirement is less clear:
Table 1.3 The proportional Atmospheric pressure (mmHg) 760 760 760
effect on alveolar oxygen
Water vapour (mmHg) 47 47 47
tension of CO2 removal at
increasing FIO2 Calculated inspired oxygen partial pressure (Eq.1.21)
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 ECCO2R
PaCO2 (mmHg) 40 40 40
PAO2 (mmHg) 106 390 536
Proportional effect of CO2 removal on PAO2
% Change 125 18 12
Biophysics oftheExtracorporeal 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
(venous to venous or venous to arterial) that determine whether the mode of support
is primarily oxygenation 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. Several of the parameters that describe the
extracorporeal blood path also apply to flow within the systemic circulation and
interactions with the native circulation will be considered where appropriate.
Various coatings (discussed later) may be applied to the circuitry to enhance bio-
compatibility, 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-
tion (LVAD in Fig.1.1) before also discussing the oxygenator.
Types ofBlood 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 the tubing is compressed by a roller. Two to three rollers are mounted on a rotat-
ing disc while the tubing is encased around this disc with the geometry 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 dictates the volume displaced by each roller pass while revo-
lution 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 hemolysis
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 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.
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 80150rpm
range these devices operate at 20004000rpm and considerable engineering 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 perpendicularly against the
periphery of the pump along which the outlet is contained. These devices are non-
occlusive and will rotate freely whether the pump is de-primed by entrainment of air
or occlusion of the blood path. The output is thus significantly influenced by loading
conditions and the physical properties of blood.
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 9000rpm). Currently they are only
implemented in ventricular assist devices and, comparable 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 ECLS circuits this dis-
sociation between pump speed and flow has resulted in Doppler monitoring of the
velocity of blood in the tubing. In implanted devices (ventricular assist devices), it
is not currently possible to provide a long term accurate flow measurement system
so flow is calculated from the rotation speed. Accurate haematocrit measurements
are important in determining the actual flow, as will be discussed. However, in prac-
tice, exact determination of flow may be less important than clinical evidence of
adequate forward output (perfusion, exercise tolerance) and ventricular unloading
(pulmonary congestion), while echocardiography can demonstrate ventricular
collapse (Fig.1.20).
Physical Properties ofBlood Relevant toFlow
The basic hydraulic principles outlined below describe the flow of a uniform incom-
pressible fluid in solid tubing. The conduits utilised in ECLS can be considered
solid tubes; however, blood can only approximate a uniform fluid due to the pres-
ence of suspended cells and chemical interactions of suspended proteins with the
conduit walls.
Two fluid characteristics are relevant to flow: density and viscosity. As a descrip-
tion of mass per unit volume, density () particularly affects the momentum of mov-
ing fluid and resistance to acceleration and for blood is slightly greater than water at
1.06gmL1 at 37C.
Viscoelasticity
Viscosity () 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.
Viscosity increases in a non-linear fashion relative to haematocrit and is shown
graphically in Fig.1.15.
The effect of haematocrit 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:
1. Implanted ventricular assist devices where blood flow is estimated from the rotor
rpmthe software of these devices often requires a current haematocrit.
2. In ECLS 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).
Figure1.15 also displays decreasing viscosity with increasing shear rate. Shear
rate is a measure of the rate at which adjacent fluid layers move with respect to each
other (see laminar flow below). At low blood velocities, aggregates of erythrocytes
form with plasma proteins contributing to adhesive forces. As blood velocity
increases, aggregates become smaller as erythrocytes move with respect to each
other. At high velocities, layers of erythrocytes and plasma move with significantly
less intracellular adhesive forces [39]. Vessel size contributes to the shear rate in
combination with velocity, arterioles and capillaries having the highest values and
the great veins the lowest (Fig.1.15) [38, 40]. Erythrocytes also centralise in small
capillaries, decreasing the effective haematocrit and contributing to lower viscosi-
ties [38, 41].
Fig. 1.15 Relative Viscosity of Blood compared to water vs. haematocrit and shear rate. Viscosity
rises exponentially with increasing haematocrit but falls with increasing shear rates, as aggregating
forces between red cells become less prominent (back panel). Viscosity data plotted from equa-
tions in [37] and shear rates of vessels from [38, 39]
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 [37]. These
components of force will be lost as heat (viscous friction). Some of the accelerating
force will also deform erythrocytes and in that respect impart potential energy that
is released as the erythrocytes relax back to their normal shape (elasticity).
Flow ofIdeal Fluids inSolid Conduits
Analogous to the concentration gradient driving molecular flux across membranes,

flow through a tube will occur from a high pressure area at one end to a low pressure
area at the other, while certain resistances will oppose this flow:
P - POutlet
Q = Inlet (1.22)
R
Fig. 1.16 Graphic description of the Bernoulli equation for laminar flow in a rigid conduit from a
high pressure compartment to a low pressure compartment. 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) across the system is constant
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 limit 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, i.e. 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 ECLS cannulas in vessels using colour flow Doppler techniques.
Figure1.16 also displays the total internal energy in the system, i.e. sum of poten-
tial and kinetic energy. This is nearly constant, falling only 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 and energy of momentum raises the pressure (potential energy) in the
receiving container. This 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,
constrained pressure and any kinetic energy:
r velocity2
r gh + + P = Constant where g = 9.807ms2
2
h = fluid column height ( m )
r velocity2 (1.23)
+ P = Constant For neglible gravity
2
Kinetic Potential
+ = Constant
Energy Energy
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 pres-
sure. For tubes of varying cross section the velocity will increase as diameter
decreases, leading to drops in pressure; however, the total internal energy will still
remain constant such that:
r V12 r V22
+ P1 = + P2
2 2 (1.24)
r (V22 - V12 )
DP =
2
Inspection of Fig.1.16 and the former consideration of viscoelasticity reveals why
the Bernoulli equation only strictly applies to an incompressible Newtonian fluid in
a solid conduitenergy lost to viscous forces or stored as elastic potential energy is
not included in Eq. (1.24). Nevertheless the concepts of conservation of energy from
potential (pressure) to kinetic (velocity) provide the foundation for understanding
conduit flow and Eq. (1.24) is used clinically in Doppler calculations.
Resistance toFlow
Equation (1.22) implies that any resistance simply alters the ratio between the pres-
sure 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, the 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 pressure
drop from altering velocity.
Resistance toLaminar 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 stationary blood film in contact with
the conduit walls from 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 h
Resistance to laminar flow =
p radius4
(
mmHgL-1 min -1 ) (1.25)

Combining Eqs. (1.22) and (1.25) and converting radius to diameter, d, yields
Poiseuilles equation:
( P - POutlet ) p d 4
Q = Inlet (1.26)
128 length h

The effect of viscosity () 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, which can be limited by vessel size and tech-
nical factors relating to insertion.
Turbulent Flow
In contrast to laminar flow in straight tubules, at circuit constrictions, rotors and

around the complex geometry of the oxygenator, flow is turbulent. In turbulent flow,
more of the impulse energy goes to random kinetic motion, eddy currents and fric-
tion between the fluid, suspended cells and the walls of the tube. In comparison to
Eq. (1.26), the pressure gradient over a conduit demonstrating turbulence increases
exponentially with increasing blood flow [14].
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 benefits 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 [42] where, for a conduit of diameter d and a fluid with a mean velocity v :
r dv
Re = (1.27)
h
At Re numbers less than 100, pure laminar flow is apparent; from 100 to 1000 lami-
nar flow occurs but an increasingly wide boundary layer appears against the station-
ary conduit; above 1000 there is a transition to turbulent flow streams and eddy
currents; Re numbers >10,000 are purely turbulent flow [42]. The linking of Re to
velocity is critical in determining the maximal flow through extracorporeal circuits
and will be discussed below. The distance blood must flow before laminar flow can
be established is also related to Re and is known as the entrance length:
Re0.03tube diameter (Fig. 1.16).
A Reynolds number can also be calculated for the resistance (drag) that a particu-
late experiences in a viscous fluid. For erythrocytes suspended in plasma the cell
diameter (dRBC) and a coefficient, , for cell deformity Replasma is defined as [37]:
r plasma dRBC ( vplasma - vRBC ) f

Re plasma = (1.28)
h plasma

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 forGas 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, mLs1) through any two points of the extracorporeal circuit must be
constant and flow can be described as the product of blood velocity, V (cms1) and
cross sectional area, CSA (cm2):
Q1 = Q2
(1.29)
CSA1 V1 = CSA 2 V2

The cross section of the conduit is calculated as radius2 and an ECLS flow of
5Lmin1 in 1cm diameter tubing yields a velocity of 1.06ms1. Oxygenator cross-
sectional areas vary but an approximation can be made by dividing the oxygenator
priming volume by the oxygenator width in cm, yielding a volume per cm. This can
be divided by the volume in 1cm 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 1.3cms1 or around 100 times slower than the conduit velocity, 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
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 ECLS circuit with a centrifugal vortex pump, a membrane
oxygenator and conduits between the patient access cannulae. Systems consisting of centrifugal
vortex pumps (as opposed to roller pumps) are typically valveless, allowing 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
hollow fibre oxygenators, the blood must spread across the opening face of the
tubules. More clot deposition over time will gradually decrease the number of paths
that blood can take.
The Oxygenator asaResistor
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 or membrane of the oxygenator equate to an extracorporeal capillary system
and considerably impede the flow of blood, which in the case of adult ECLS is in
the order of 36Lmin1. 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 is life-threatening and early warning of circuit clotting is
important. It is common practice to continuously monitor the oxygenator inlet and
outlet pressures by placing pressure transducers at these points (Fig.1.17). The flow
generated by the blood pump is usually known and, if kept constant, Eq. (1.22) can
be simplified to: pressure drop is proportional to resistance. 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.
Conservation ofEnergy Across theECLS Circuit
Figure1.18 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 lost to viscous fric-
tion (most notably across the oxygenator) and added by the pump are now depicted.
As we are now using blood, the full statement for stable energy transfer must now
replace the Bernoulli equation:
r v12 r v22
r gh1 + + P1 + hpump + hRBC = r gh2 + + P2 + hfriction + hRBC (1.30)
2 2
Fig. 1.18 Energy exchange across the extracorporeal circuit; after [43]. Refer to Fig.1.16 for
discussion of energy transfer from pressure to velocity. The total internal energy of the fluid is
depicted without inclusion of heat generated from viscous friction. Thus the total energy falls at all
points, most prominently in the low flow region of the oxygenator, being added only by the pump.
Narrow spacing between laminar flow lines represents high velocity at low pressure
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 erythrocyte.
Energy is added by the pump (pump header, 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 (pressure and velocity) of the
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 [42].
At startup, the high velocity veins of a 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. (1.30). This does not mean turbulent flow will exist throughout the
system; rather the energy lost from small areas of turbulence will summate to
oppose the applied energy of the rotor resulting in a steady terminal velocity.
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
(conservation of momentum). Instead, the kinetic energy is converted to pressure
against the pump housing, against the 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 to
rpm by dividing by 60).
rpm
P = torque 2p (1.31)
60
Torque represents force multiplied by angular distance and has SI units of Newton
metre (Nm); however, it is more understandable to clinicians as joules per radian.
It should be noted in Eq. (1.31) that this is the force required to move the rotors to
both propel the fluid column and overcome viscous friction at the rotor blades.
Occlusion oftheFlow Path
It is important to appreciate that the total system energy in a velocity pump as

described above depends on there being flow: 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. (1.30) will be zero at the pump inlet and no
work will done maintaining blood velocity against viscous friction at any point in
the circuit, other than viscous friction at the rotor blades themselves.
The only moving blood now is in circular motion around the rotors and the rotor
velocity will accelerate, raising Reynolds number until sufficient local turbulence
exists around them that the resistance from this viscous friction again opposes the
applied energy and a new steady rotor velocity results. In other words, by Eq. (1.31),
if the same power is applied to the rotors then rpm will increase, conversely if the
rotors are set to maintain a continuous rpm, power consumption must fall.
This effect of occlusion lowering the work done by 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 pressure in the distal conduit until mechanical failure occurs, i.e. the work the
roller pump performs on the blood increases with occlusion.
Flow Regurgitation
It should be apparent from Eq. (1.30) that a high opposing velocity or pressure in the
outflow region opposes forward flow and the energy imparted by the pump must
overcome this. This is of importance 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 be wholly spent
overcoming viscous friction at the blades from extremely turbulent flow.
As introduced above, turbulent flow refers not only to eddies of the blood stream
but also to local turbulence of plasma around erythrocytes (Eq.1.27) and the shear
stress on erythrocyte cell membranes in this setting will be high, often resulting in
haemolysis.
Cavitation
As introduced above, turbulent flow describes rapidly varying random inverse fluc-
tuations of pressure and velocity that sum to the total internal energy of the fluid
[42]. The amplitude of these fluctuations increases with the internal energy of the
fluid and, as discussed, high local velocities may induce erythrocyte haemolysis
from shear stress. A second mechanism of haemolysis is known as cavitation.
If the peak negative pressure 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 abso-
lute property of the fluid at a particular temperature. 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 ofVelocity Pumps withtheSystemic Circulation
Ventricular Assist Devices: VADs
The interaction of power consumption with axial velocity just described for velocity
pumps is of particular relevance to the operation and monitoring of VADs. The loca-
tion of the device between left ventricle and aorta (Fig.1.1) exposes it to pulsatile
flow (if enough ventricular function exists) while the short path length and lack of
an oxygenator results in minimal dampening of this pressure fluctuation. The effects
of a varying pressure gradient can be considered in an analogous fashion to the
discussion of occlusion (Fig.1.19).
In a normal heart, the aortic valve opens in ventricular systole and the slightly
higher pressure in the left ventricle results in antegrade flow and transfer of pressure
energy to the aortic blood stream and elastic aortic wall. In diastole, the pressure
falls in the relaxing ventricle and the pressure of blood in the aorta closes the aortic
valve, maintaining a pressure gradient between the ventricular cavity and aorta. The
stored potential energy of the aortic blood drives flow through the peripheral circu-
lation after ventricular contraction has ceased for as long as aortic pressure is greater
than right atrial pressure (Fig.1.19, lower left panel).
In the LVAD supported circulation, the total energy of blood in the LVAD outlet
exceeds that in the aorta, and in the zone of flow deceleration (Fig.1.16 and Eq.
(1.23)) the kinetic energy of the two streams becomes equal with the additional
energy augmenting aortic pressure. Any pressure increase in the left ventricle
(whether enough to open the aortic valve or not) will lower the pressure gradient
between the ventricle and aorta, requiring less force to be exerted on the blood by
the rotors. By Eq. (1.31), this results in a higher flow and motor rpm for the same
motor power output or a reduction in motor power to maintain the same rpm
(Fig.1.19). These devices are thus described as afterload-sensitive.
In reality, pulsatile fluctuations in the arterial trace (accentuated if the aortic valve
starts opening), motor rpm and motor power occur, with the proportions dependent
on the LVAD controller software, pump speed and conditions. LVAD computer-
controllers monitor the waveform of speed oscillations to provide numerical indica-
tors of the ventricleLVAD interaction. The averaged peak-trough speed difference is
known as the pulsatility index. Monitoring this value over time gives a guide to
changing ventricular conditions, with high values suggesting significant LV contrac-
tility (which may mean the LVAD flow can be increased to further unload the ven-
tricle) and falling values indicating loss of contractility (that acutely may mean the
LV is collapsing and beginning to occlude the LVAD inflow). Similarly, the shape of
the waveform (rate of speed change) compared to a sine wave (over pumping index)
can be monitored and a numeric index produced. As they depend on multiple factors
including LV function, RV function, aortic pressure, aortic valve opening and heart
rate, and may be confounded by dynamic factors such as thrombus formation and
position [44], such parameters are only useful when monitored for changes over
time, with alterations triggering other tests such as echocardiography.
Flow Access Cannula Pump (Axial Return Cannula Flow

Acceleration in LV Configuration) in Aorta Deceleration
Energy
Systole LVAD POut
Systole LVAD PIn

Pressure
Diastole LVAD POut
Diastole LVAD PIn

Velocity
Progression through LVAD
AoVClose
PAo AoVOpen
Systole LVAD POut
Diastole LVAD POut
Pressure
Systole P
Systole LVAD PIn
PLV
Diastole P
MVClose
MVOpen Diastole LVAD PIn
PLA
Time Time
Normal Circulation LVAD Supported Circulation
Fig. 1.19 Afterload sensitivity of a left ventricular assist device. Top panel depicts an axial flow
LVAD device flowing from the Left Ventricle (LV) to the Aorta (Ao). Middle panel (not to scale)
depicts combined total energy of the blood stream as it progresses through the LVAD and the
contribution of potential energy (pressure) and kinetic energy (velocity) in systole (black lines)
and diastole (red). Viscous friction results in net energy loss at all points except the pump where
external energy is added. Left lower panel depicts the pressure vs. time graph for aorta, left ven-
tricle and left atrium (LA) for a normal circulation; mitral (MV) and aortic (AoV) valve state is
indicated. Right lower panel depicts pressure vs. time graph for an LVAD supported circulation.
The aortic valve will not open unless the LV pressure exceeds the aortic pressure for part of car-
diac cycle. Even without valve opening, aortic pressure fluctuations result from instantaneous
changes in pressure between the LV (LVAD PIn) and Aorta (LVAD POut) that occur through the
cardiac cycle from any cardiac activity
VA-ECMO
Central VA-ECMO usually involves right atrial cannulation with blood returning to
the proximal aorta, while peripheral VA-ECMO drains the great veins or right
atrium and returns to the descending aorta (Fig.1.1). The same principles determine
circuit flow as for VAD configurations; however, a pulsatile pressure gradient across
the extracorporeal circuit will not occur unless enough ventricular pressure is gener-
ated to open the aortic valve. Furthermore, the pressure difference between the aorta
and ventricle in systole and diastole will be proportionally greater than that between
the aorta and right atrium due to the lower diastolic pressure in the ventricle. Hence,
VA-ECMO will not produce a pulsatile flow unless enough ventricular function
exists to open the aortic valve.
Flow Reversal
The absence of valves in the blood path of many ECLS configurations allows blood
to flow in reverse when the pump is not running. This is particularly prominent in
VA-ECMO configurations and may be damaging, as such flow creates a relatively
low resistance systemic shunt from the arterial limb to the venous limb and in the
failing circulation will rapidly worsen organ perfusion. Flow reversal is avoided by
clamping the ECLS circuit at any time where low or no flow is anticipated, includ-
ing at initiation of support, utilising the characteristic of velocity pumps to continue
rotation when occluded. Before unclamping, the rotor speed is increased to a level
that should oppose flow reversal and then increased further as required (see occluded
flow above).
Differential Cyanosis
In peripheral VA-ECMO, flow in the upper aorta is retrograde from the point of
maximal pressure generation at the site of the ECMO outlet. As cardiac function
improves and the aortic valve begins opening, the moving column of blood ejected
by the ventricle will meet the moving column from the extracorporeal circuit. At
this point the velocity of the two streams will become equal (and low) and a swirling
region of high pressure will exist. The level of this region of relative stasis will
depend on the energy contained in the fluid columns.
If the pathology that required peripheral VA-ECMO also resulted in poor oxy-
genation by the lungs then the ventricular component of the aortic pressure stream
will contain hypoxic blood and the great vessels supplying the head and neck will
be perfused with this. The first vessels originating from the aorta are the coronary
arteries followed by the brachiocephalic trunk supplying the right arm and right
carotid. It is thus common practice to monitor saturations in the right arm to detect
this phenomenon.
The management to restore oxygenated perfusion to the brain and coronary

arteries depends on the level of ventricular function. If the ventricle appears likely
to be able to support systemic perfusion then VA-ECMO can be withdrawn or con-
verted to VV-ECMO to provide oxygenation alone. If circulatory support is still
required then maximising oxygenation with appropriate ventilator settings may
alleviate the problem. Decreasing the inflow of blood to the left heart will also
decrease the left ventricle cardiac output and this can be achieved by placing
increased load on the right ventricle (by increasing PEEP) and increasing the ECMO
circuit flows to drain the right atrium thus removing right ventricular preload.
Valvular Incompetence
Aortic incompetence is a significant problem in VA-ECMO and, to a lesser degree,

LVAD support. Aside from structural valve disease, both mitral and aortic incompe-
tence commonly result from dilation of the valve annulus in the failing heart.
Unfortunately, it is often in this setting that VA-ECMO is instituted. In VA-ECMO,
persistent blood flow through the right heart and pulmonary vasculature often results
in further dilation, yet the inability to generate enough pressure to open the aortic
valve means blood cannot be ejected. In this situation a small amount of aortic
regurgitation may worsen with time, often with development of worsening mitral
regurgitation. This combination leads to elevated pulmonary vascular pressures and
in severe cases overt pulmonary edema. Management strategies are similar to those
for differential cyanosis, increasing pulmonary vascular resistance to decrease right
heart output, although decreasing circuit flows instead of increasing them may alle-
viate the situation through facilitating more ventricular ejection. Targeting lower
systemic arterial pressure may also help, as this lowers the gradient across the aortic
valve, minimising regurgitation and allowing antegrade LV ejection.
Ventricular Preload vs. Device Preload
Ventricular performance is classically described by the Frank-Starling mechanism

where, for a given contractility, an increase in myocardial wall tension results in
increased stroke volume, with the myocardial wall tension being termed preload. This
mechanism allows beat to beat variations in venous return to be ejected and contributes
to matching the output of the right ventricle with that of the left during respiration.
Measurable surrogates for preload include the left ventricular end diastolic
pressure (LVEDP) and failure of the stroke volume to adequately increase with
increments of LVEDP over the normal range is part of the pathophysiology of sys-
tolic heart failure. Failure to achieve an optimal wall tension over a given time
period (rate of relaxation) or at elevated LVEDP characterises the pathophysiology
of diastolic dysfunction.
As described above, velocity flow devices are afterload-sensitive. However,
unlike the native ventricle, continuous flow devices draining the ventricles require a
constant flow at their intake and do not increase their output relative to venous
return, making them more preload-insensitive. Current pump controller software

algorithms can attempt to automatically modify the pump speed in response to
pump speed waveform analysis (described above); however, no current device has
live feedback from a true measure of LVEDP or LV volume [44]. Instead of respond-
ing to physiologic demand, most of these feedback loops are tuned to drop the pump
speed if indicators of total LV collapse are evident. The requirement for continuous
flow has several implications for both VADs and ECLS.
Collapse Around theInflow Cannula
At the intake of the circuit depicted in Fig.1.18, pressure rapidly falls as the kinetic
energy of the blood increases. Low pressure vortices of turbulent flow will exist
around the zone of flow acceleration and these forces will tend to draw the compli-
ant venous vessel or heart chamber wall toward the intake if not opposed by the
distending pressure from the inflow of further blood. In ECLS circuits this may
manifest as transient circuit shuddering, increased haemolysis of blood and low
flow alarms. In VAD, ventricular arrhythmias may be precipitated, numeric indices
of pulsatility will decrease (ventricular contraction is ineffective due to lack of pre-
load) and poor peripheral perfusion may result (Fig.1.20).
Fig. 1.20 Left ventricular suck down due to hypovolemia and excessive VAD flows, all imaged in
diastole. The top left panel shows a long axis view with the VAD inflow cannula to the left of the
ventricular cavity. The anterior leaflet of the mitral valve is mid-cavity; however, the posterior
leaflet and left atrium are compressed with a narrow slit accommodating ventricular inflow. Top
right is a short axis view of the same valve demonstrating a small distorted mitral valve and an
unusual concavity to the ventricular septum. The bottom panels are the result of decreasing pump
speed and giving fluid; the left ventricular cavity has a more normal rounded appearance and the
left atrium and mitral inflow are no longer collapsed. Image M.Brain by permission of The Alfred
Intensive Care Unit, The Alfred Hospital, Melbourne
Mitral Inflow Restriction
Of particular importance to LVAD function is a low resistance path for blood to

enter the left ventricle. Significant mitral stenosis is an obvious contraindication but
even the flow resistance created by prosthetic mitral valves may not allow enough
flow to avoid ventricular collapse without requiring excessively elevated pulmonary
pressures that may impair oxygenation.
Right Ventricular (RV) Function andPulmonary Haemodynamics
The lack of a preload-sensitive mechanism in an LVAD means that physiologic

variation in right ventricular cardiac output is not coupled to that of the left.
Poor RV function or elevated pulmonary vascular resistance may impair blood flow
to the left heart, leading to restriction of maximal LVAD flows that can be achieved
without ventricular collapse.
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 resis-
tance can often be achieved by minimising excessive ventilator pressures and the
use of pulmonary vasodilators such as inhaled nitric oxide or epoprostenol. Right
ventricular 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).
Compared to LVAD therapy, VA-ECMO performance is relatively independent
of poor RV function.
Thrombus Formation
As well as thrombus formation in the low flow components of the oxygenator,

thrombus may develop in or at the orifice of any of the ECLS components. This is
more likely to occur when lower flows have been utilised or anticoagulation ceased.
Anticoagulation for extracorporeal circuits will be discussed in a later chapter.
Conclusions
ECLS is a means of removing carbon dioxide and enhancing global oxygen delivery,
whether by increasing the oxygen content of blood or by augmenting flow. Although
seemingly complex, it can be comprehensively understood when framed in a
physiological and biomechanical context. The practical use of the technology 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 membrane
module for using implantable artificial lung. J Membr Sci. 2009;326(1):1306.
2. Scheepers A, Joost H-G, Schrmann A.The glucose transporter families SGLT and GLUT:
molecular basis of normal and aberrant function. JPEN J Parenter Enteral Nutr.
2004;28(5):36471.
3. Stryer L.Biochemistry [Internet]. W.H.Freeman; 1995. Available from: http://books.google.
com.au/books?id=5PdpAAAAMAAJ
4. Beals M, Gross L, Harrell S.Efficiency of ATP production 4th Edition. Metabolism for energy
and the respiratory quotient. 1999 [cited 2012 Aug 15]. Available from: http://www.tiem.utk.
edu/~gross/bioed/webmodules/respiratoryquotient.html
5. Elia M, Livesey G.Theory and validity of indirect calorimetry during net lipid synthesis.
Am J Clin Nutr. 1988;47(4):591607.
6. Wolfe RR, Allsop JR, Burke JF.Glucose metabolism in man: responses to intravenous glucose
infusion. Metabolism. 1979;28(3):21020.
7. Hellerstein MK etal. De novo lipogenesis in humans: metabolic and regulatory aspects. Eur J
Clin Nutr. 1999;53:5365.
8. Mulquiney PJ, Kuchel PW.Control analysis of 2,3-bisphosphoglycerate metabolism [Internet].
Biochem J.1999 [cited 2012 Oct 10];342:597604. Available from: http://www.biochemj.org/
bj/342/bj3420597.htm
9. Mulquiney PJ, Bubb WA, Kuchel PW.In situ kinetic characterization of 2,3-bisphosphoglycerate
synthase/phosphatase [Internet]. Biochem J.1999 [cited 2012 Oct 10];342:56780. Available
from: http://www.biochemj.org/bj/342/bj3420567.htm
10. Lawson DS, Holt D.Insensible water loss from the Jostra Quadrox D oxygenator: an invitro
study. Perfusion. 2007;22(6):40710.
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):4327.
12. Peter S.Strong ions plus carbon dioxide. In: Kellum JA, Elbers PW, editors. Stewarts text-
book of acid-base. 2nd ed. London, UK: Lulu Enterprises; 2009. p.11132.
13. Anaesthsia Correspondence Web Site [Internet]. [cited 2012 Sep 17]. Available from: http://
www.anaesthesiacorrespondence.net/Correspond3.asp?articleid=3598&archive=1
14. Lumb AB.Nunns applied respiratory physiology. 5th ed. Oxford: Butterworth-Heinemann; 2000.
15. Sargent, John A, Gotch, Frank.Principles and biophysics of dialysis. Replacement of Renal
function by Dialysis. Dordrecht: Springer; 1996. p.34145.
16. Cussler EL.Diffusion: mass transfer in fluid systems. Cambridge: Cambridge University
Press; 1997.
17. Grathwohl P.Diffusion in natural porous media: contaminant transport, sorption/desorption
and dissolution kinetics. Boston, MA: Kluwer Academic Publishers; 1998.
18. Katoh S, Yoshida F.Rates of absorption of oxygen into blood under turbulent conditions.
Chem Eng J.1972;3:27685.
19. Matsuda N, Sakai K.Blood flow and oxygen transfer rate of an outside blood flow membrane
oxygenator. J Membr Sci. 2000;170:1538.
20. Weibel ER.The pathway for oxygen: structure and function in the mammalian respiratory
system. Cambridge, MA: Harvard University Press; 1984.
21. Epstein FH, Hsia CCW.Respiratory function of hemoglobin. N Engl J Med. 1998;

338(4):23948.
22. 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):1316.
23. Thomas LJ.Algorithms for selected blood acid-base and blood gas calculations. J Appl Physiol.
1972;33(1):1548.
24. Kelman GR.Digital computer subroutine for the conversion of oxygen tension into saturation.
J Appl Physiol. 1966;21(4):13756.
25. 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
26. Arthurs GJ, Sudhakar M.Carbon dioxide transport. Contin Educ Anaesth Crit Care Pain.
2005;5(6):20710.
27. 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):224651.
28. Gutknecht J, Bisson MA, Tosteson FC.Diffusion of carbon dioxide through lipid bilayer mem-
branes. Effects of carbonic anhydrase, bicarbonate, and unstirred layers. J Gen Physiol.
1977;69(6):779.
29. Centor R.Serum total carbon dioxideclinical methodsNCBI 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/
30. Mochizuki M.The CO2 dissociation curve at steady state invivo. Yamagata Med J. 2004;
22(1):258.
31. Beilin LJ, Knight GJ, Munro-Faure AD, Anderson J.The sodium, potassium, and water contents
of red blood cells of healthy human adults. J Clin Invest. 1966;45(11):1817.
32. Mochizuki M.Analysis of bicarbonate concentration in human blood plasma at steady state
invivo. Yamagata Med J.2004;22(1):0924.
33. 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):57999.
34. Tazawa H, Mochizuki M, Tamura M, Kagawa T.Quantitative analyses of the CO2 dissociation
curve of oxygenated blood and the Haldane effect in human blood. Jpn J Physiol.
1983;33(4):60118.
35. Baker A, Richardson D, Craig G.Extracorporeal carbon dioxide removal (ECCO2R) in
respiratory failure: an overview, and where next? Journal of the Intensive Care Society
2012;13(3):232237
36. 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):13018.
37. Jung J, Lyczkowski RW, Panchal CB, Hassanein A.Multiphase hemodynamic simulation of
pulsatile flow in a coronary artery. J Biomech. 2006;39(11):206473.
38. Papaioannou TG, Stefanadis C. Vascular wall shear stress: basic principles and methods.
Hellenic J Cardiol. 2005;46(1):915.
39. Lipowsky H.Microvascular rheology and hemodynamics. Microcirculation. 2005;12(1):515.
40. Lipowsky HH, Kovalcheck S, Zweifach BW.The distribution of blood rheological parameters
in the microvasculature of cat mesentery. Circ Res. 1978;43(5):73849.
41. Ercan M, Koksal C.The relationship between shear rate and vessel diameter. Anesth Analg.
2003;96(1):3078.
42. Bahrami M.Introduction to fluid mechanics: ENSC283 2011 [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/
43. Bloomfield LA.How everything works: making physics out of the ordinary. Hoboken, NJ:
Wiley; 2007.
44. Salamonsen RF, Lim E, Gaddum N, AlOmari A-HH, Gregory SD, Stevens M, etal. Theoretical
foundations of a starling-like controller for rotary blood pumps. Artif Organs. 2012;
36(9):78796.
Chapter 2
Hypoxemic Respiratory Failure: Evidence,
Indications, and Exclusions
Darryl Abrams, Matthew Bacchetta, and Daniel Brodie
Abbreviations
ARDS Acute respiratory distress syndrome

ECMO Extracorporeal membrane oxygenation
PaO2 Partial pressure of oxygen in arterial blood
FiO2 Fraction of inspired oxygen
ECCO2R extracorporeal carbon dioxide removal
LFPPV Low-frequency positive-pressure ventilation
PCIRV Pressure-control inverse ratio ventilation
PEEP Positive end-expiratory pressure
APACHE Acute physiology and chronic health evaluation
SOFA Sequential organ failure assessment
BMI Body mass index
Qs/Qt Shunt fraction
D. Abrams, MD (*) D. Brodie, MD

Division of Pulmonary, Allergy and Critical Care, New York-Presbyterian Hospital/Columbia
University Medical Center, 622 W. 168th Street, PH 8 East, Room 101,
New York, NY 10032, USA
e-mail: da2256@cumc.columbia.edu
M. Bacchetta, MD, MBA, MA
Division of Thoracic Surgery, New York-Presbyterian Hospital/Columbia University Medical
Center, 161 Fort Washington Avenue, Room 336, New York, NY 10032, USA

Respiratory Medicine 16, DOI 10.1007/978-1-4939-3005-0_2
62 D. Abrams 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, but only at the expense of generating excessively high
inspiratory airway pressures. ECMO in this setting facilitates lung-protective venti-
lation and minimizes ventilator-associated lung injury. Despite the increasing use of
ECMO in ARDS, high-level evidence supporting its benefit for this indication is
lacking. The first successful use of ECMO for severe post-traumatic hypoxemic
respiratory failure was in 1971 by Dr. J.D. Hill and colleagues. The patient received
venoarterial extracorporeal support with a Bramson membrane lung for 75 h and
survived [3]. This success prompted others to attempt the institution of ECMO sup-
port for severe hypoxemic respiratory failure, with reports of approximately 150
cases performed from the original Hill report through 1974, though mortality was
8590 % [46]. In 1979, Warren Zapol and colleagues published the results of a
multicenter, randomized, controlled trial of mechanical ventilation supplemented
with venoarterial ECMO versus conventional mechanical ventilation alone, accord-
ing to the standard of care at that time, as a therapy for severe acute hypoxemic
respiratory failure [7]. Ninety subjects were randomized to conventional mechani-
cal ventilation alone or mechanical ventilation and ECMO, with slow and fast entry
criteria utilized for enrollment based on a combination of severity and duration of
hypoxemia, resulting in a maximum PaO2 to FiO2 ratio of 83. Subjects with pro-
longed mechanical ventilation (greater than 21 days), evidence of left atrial hyper-
tension (pulmonary capillary wedge pressure greater than 25), and chronic or
irreversible diseases were excluded. The trial demonstrated no significant difference
in survival (9.5 % in the ECMO group, 8.3 % in the control group, Table 2.1), with
high rates of infectious and bleeding complications. The low rate of survival in the
Table 2.1 Demographics and outcomes of prospective randomized trials of ECLS for ARDS
ECMO Non-ECMO
Study Year No. of pts. PaO2:FIO2 Modality survival (%) survival (%)
Zapol et al. [7] 1979 90 <83 ECCO2R 9.5a 8.3
Morris et al. [12] 1994 40 63 ECCO2R 33a 42
Peek et al. [33] 2009 180 75 ECMO 63b, c 47
ECLS extracorporeal life support, ECMO extracorporeal membrane oxygenation, ARDS acute
respiratory distress syndrome, ECCO2R extracorporeal carbon dioxide removal, PaO2 partial pres-
sure of arterial oxygen, FIO2 fraction of inspired oxygen
a
No statistically significant difference in survival between groups
b
22/90 patients (24 %) within ECMO referral group did not receive ECMO
c
Relative risk of death or severe disability 0.69 (95 % CI 0.050.97, p = 0.03) compared to non-
ECMO group
2 Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions 63
control group also highlights the high severity of illness in the study population and
the different standard of care for mechanical ventilation in ARDS at that time.
In the late 1970s and into the 1980s, Gattinoni and colleagues in Italy, among
others, began experimenting with the use of extracorporeal support in animals, and
then in humans with severe forms of ARDS. The concept involved using venove-
nous extracorporeal support at low blood flow rates principally to remove CO2, a
technique known as extracorporeal CO2 removal or ECCO2R. By removing CO2
directly from the blood, they could ventilate the subjects at lower airway pressures
and respiratory rateswhat they termed low-frequency positive-pressure ventila-
tion (LFPPV) while providing passive oxygenation through the endotracheal tube
[811]. The goal was to provide what Gattinoni referred to as pulmonary rest, the
precursor to low-volume, low-pressure ventilation strategies [11]. Results of a pro-
spective, uncontrolled study of 43 subjects treated with a LFPPV-ECCO2R strategy
for severe respiratory failure (average PaO2 to FIO2 ratio of 67) demonstrated a
survival of 49 %, despite an anticipated survival of less than 10 % based on similar
inclusion and exclusion criteria to those used in the trial by Zapol [9].
Prompted by a significantly higher survival than prior studies of subjects with
severe ARDS, a lack of randomization or control group in the Gattinoni study, and
the hypothesis that ventilator-associated lung injury plays a critical role in mortality
in ARDS, Morris and colleagues at the University of Utah evaluated the role of
LFPPV-ECCO2R in a prospective, randomized, controlled trial of subjects with
severe ARDS [12]. Forty subjects were randomized between 1987 and 1991 to
receive either conventional mechanical ventilation or a strategy of pressure-control
inverse ratio ventilation (PCIRV), followed by LFPPV-ECCO2R if PCIRV was
unable to maintain adequate PaO2 or pH. Entry criteria were similar to those used
by Zapol and Gattinoni. Oxygenation was markedly impaired, with an average PaO2
to FIO2 ratio of 63. The results demonstrated no difference in mortality between
groups, with a survival of 42 % in the control group and 33 % in the PCIRV plus
LFPPV-ECCO2R group. Hemorrhage with high rates of transfusion led to ECCO2R
discontinuation in seven (37 %) of the subjects in the experimental group, and four
subjects (21 %) developed clot within the extracorporeal circuit. Although the
authors concluded there was no role for ECCO2R as a therapy for ARDS, there were
several notable limitations of this trial. The intervention group received two differ-
ent experimental treatment modalities (inverse ratio ventilation and ECCO2R),
worldwide experience with extracorporeal techniques was limited prior to study
initiation, and the intervention arm was exposed to notably high airway pressures
despite extracorporeal support. Additionally, survival among those in the control
group was higher than in the previous studies, highlighting the differences in patient
characteristics, therapeutic interventions, or protocols for respiratory care between
this study and previously published data. Furthermore, while all of these studies
used uniform entry criteria that facilitate comparisons between trials, these criteria
may not reflect best practice for the initiation of extracorporeal support.
Starting in the 1980s, there were multiple non-randomized, observational studies
evaluating ECLS for severe ARDS, several of which are listed in Table 2.2 [9,
1322]. The studies listed are expectedly heterogeneous in their demographics,
64 D. Abrams et al.
Table 2.2 Demographics and outcomes of observational ECMO trials for acute respiratory failure
Study Year No. of pts. PaO2:FIO2 Indication Modality Survival (%)
Gattinoni et al. [9] 1986 43 67 ARDS ECCO2R 49
Wagner et al. [13] 1990 76 ARDS ECCO2R 50
Brunet et al. [14] 1993 23 84 ARDS ECCO2R 52
Manert et al. [15] 1996 21 54 ARDS ECMO 81
Kolla et al. [16] 1997 100 56 Severe ARFa ECMO 54
Peek et al. [17] 1997 50 65 Severe ARF ECMO 66
Lewandowski 1997 49 67 ARDS ECMO 55
et al. [18]
Linden et al. [19] 2000 17 46 ARDS ECMO 76
Bartlett et al. [20] 2000 86 55 ARDS ECMO 61
Davies et al. [21] 2009 68 56 ARDS/H1N1 ECMO 75
Freed et al. [25] 2010 6 61 ARDS/H1N1 ECMO 67
Roch et al. [24] 2010 9 52 ARDS/H1N1 ECMO 56
Schmid et al. [22] 2011 176 77 ARDS ECMO 56
ECMO extracorporeal membrane oxygenation, PaO2 partial pressure of arterial oxygen, FIO2 frac-
tion of inspired oxygen, ARDS acute respiratory distress syndrome, ECCO2R extracorporeal car-
bon dioxide removal, H1N1 denotes the 2009 novel influenza A(H1N1) virus, ARF acute respiratory
failure
a
Defined as shunt fraction >30 %, compliance <0.5 mL/cm H2O/kg; life-threatening anatomic
airway obstruction, refractory status asthmaticus, or uncorrectable hypercapnia with pH < 7.0 and
end-inspiratory pressure >45 mmHg
however they all included subjects with marked impairment in oxygenation that
would meet the most recent definition of severe ARDS [1].
Among these studies is the Australia-New Zealand experience with ECMO, pub-
lished by Davies and colleagues, during the influenza A(H1N1) outbreak in 2009
[21]. Sixty-eight subjects were treated with ECMO for ARDS in 15 specialist ICUs
that provided ECMO support during the study period. All cases were confirmed or
strongly suspected to be a result of influenza A(H1N1), with a high severity of ill-
ness as demonstrated by ventilatory parameters: median PaO2 to FIO2 ratio of 56,
median positive end-expiratory pressure (PEEP) of 18 cm H2O, median nadir pH of
7.2, and median highest PaCO2 of 69. Survival in this cohort of subjects was 75 %
[23]. Similar outcomes were demonstrated in other centers during the influenza
pandemic, though with smaller sample sizes (Table 2.2) [9, 1322, 24, 25].
Contemporaneous with the report by Davies et al., a study with similar demo-
graphics and outcomes of subjects with influenza A(H1N1) who were managed
without the use of ECMO was reported by Miller et al. in Utah [26]. Among the 47
subjects admitted to the ICU with a confirmed diagnosis of influenza, 30 (64 %) met
criteria for ARDS, with a median PaO2 to FIO2 ratio of 61 and median PEEP of 22.
Multisystem organ failure was common (87 %). None of the subjects with ARDS
received so-called rescue therapies such as inhaled nitric oxide, prone position-
ing, inhaled epoprostenol, or high-frequency oscillatory ventilation, although neu-
romuscular blocking agents were used in 47 % of subjects. Despite a high severity
of illness (median APACHE II score = 25), survival was 73 %. Comparable survival

to the Australia-New Zealand cohort raised questions about whether ECMO pro-
vides any survival advantage over optimal medical management in cases of severe
ARDS related to influenza A(H1N1).
An attempt to reconcile this issue was made by Noah and colleagues from the
United Kingdom, who compared ECMO-referred subjects with confirmed or
strongly suspected H1N1-related ARDS (n = 80) to non-ECMO-referred subjects
with H1N1-related ARDS who were enrolled in a separate, concurrent prospective
cohort study within the same geographical area and who were potentially eligible
for ECMO (n = 195) [27]. Only 69 of the 80 subjects referred for ECMO received
ECMO (86 %). However, statistical analysis was performed by intention-to-treat
principles. Individual, propensity score, and GenMatch matching were used to
match subjects from each group on the basis of demographic, physiologic, and
comorbidity data that were anticipated a priori to be associated with ECMO use and
hospital mortality (prior duration of mechanical ventilation, PaO2 to FIO2 ratio, age,
SOFA score, body mass index (BMI), pregnancy status, and use of alternative ven-
tilation strategies). After matching for the above variables, the ECMO-referred sub-
jects 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 0.310.84,
p = 0.008). The investigators could not account for differences in ventilatory strate-
gies between groups, nor could they assess the impact of variables not captured in
the database of the cohort of non-ECMO-referred subjects, which may have con-
founded the results of this non-randomized comparison.
Survival rates of subjects receiving extracorporeal support in many of the obser-
vational studies in the late 1990s and 2000s were higher (5481 %) than those
reported for subjects with and without extracorporeal support in the earlier random-
ized trials by Zapol and Morris [1522, 24, 25]. However, there are inherent flaws
in comparing non-randomized studies. Changes in clinical management confound
the comparison of survival rates from different eras. This is evident in the decline in
mortality rates observed in ARDS over the last decade [28], with mortality of 31 %
in the intervention arm of ARMA in 2000 [29], 25 % in the fluid-conservative strat-
egy arm in FACTT in 2006, and 1618 % in the control arms of ALTA and OMEGA
in 2011 [3032]. Likewise, ECMO technology has evolved significantly since the
early randomized trials, with more efficient membranes for gas exchange, the advent
of centrifugal pumps, heparin-coated circuits that can tolerate lower levels of anti-
coagulation resulting in lower bleeding risk, and cannulae that permit single-vessel
access with minimal recirculation. Comparisons of non-randomized ECMO studies
are also confounded by indication, which is influenced by multiple factors, includ-
ing but not limited to patient age, disease severity, concomitant medical conditions,
concurrent therapies, and physicians estimation of prognosis.
In an attempt to estimate the effect of ECMO in ARDS using more advanced
ECMO technology, and coinciding with increasing usage within the critical care
community, the Conventional Ventilation or ECMO for Severe Adult Respiratory
Failure (CESAR) trial was performed [33]. In this prospective, randomized,
controlled trial, 180 subjects, age 1865, with severe but potentially reversible
66 D. Abrams et al.
respiratory failure and a Lung Injury Score (a.k.a. Murray Score, a composite score
based on PaO2 to FIO2 ratio, PEEP, respiratory system compliance, and radiographic
findings [34]) of 3.0 or uncompensated hypercapnia (pH < 7.2) despite optimal
conventional management were randomly assigned to receive ongoing conven-
tional mechanical ventilation at designated treatment centers or be transferred to a
single ECMO center at Glenfield Hospital in Leicester for consideration of treat-
ment with venovenous ECMO. Subjects were excluded from the trial if they had
contraindications to anticoagulation or if they had been on high pressure (peak
inspiratory airway pressure >30 cm H2O) or high FIO2 (>0.8) for greater than 7
days. Hemodynamically stable subjects randomized to the ECMO referral arm were
initially managed on transfer to Glenfield with a standardized management protocol
that included a pressure-restricted ventilation strategy, diuresis to dry weight, trans-
fusion to a hematocrit of 40 %, prone positioning and full nutrition. Those who were
hemodynamically unstable 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 subjects who received ECMO were managed with a lung-
protective ventilation strategy. 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 of a lung-protective ventilation
strategy in the conventional management group (a low-volume, low-pressure strat-
egy was advised) and perhaps because many of these subjects were difficult to ven-
tilate, only 70 % of those subjects were managed with such a strategy at any time
during the study. Subjects were well-matched between groups and the majority
were randomized early in ARDS, with 62 % of ECMO subjects and 66 % of con-
ventionally managed subjects having been on high pressure or high FIO2 for 48 h or
less, with an average of 28 h in each group. Average PaO2 to FIO2 was approxi-
mately 75 in both groups. The primary outcomedeath or severe disability by
6 months after randomizationoccurred in 37 % of the subjects referred for ECMO,
as compared with 53 % of those in the conventional management group, relative
risk 0.69 (95 % confidence interval 0.050.97, p = 0.03).
The results from CESAR, as well as the study by Noah and colleagues, may
reasonably support a strategy of transferring patients with severe ARDS to a center
capable of performing ECMO as part of a standardized management protocol [35].
However, this trial was not a randomized trial of ECMO as compared with
standard-of-care mechanical ventilation. The higher survival in the ECMO-referral
group may 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.
To date, there is no prospective randomized trial comparing modern-day ECMO
technology and techniques to standard-of-care mechanical ventilation in
ARDS. Given that the current body of literature on ECMO in ARDS has been used
to justify or dispute the efficacy of ECMO in cases of severe ARDS, there appears to
be clinical equipoise to perform another prospective, randomized clinical trial in
which subjects randomized to the ECMO arm are guaranteed to receive ECMO and
those randomized to the mechanical ventilation arm are managed with a standardized
ventilation protocol [36, 37]. Centers that already offer ECMO and believe in its util-
ity in ARDS may find it difficult to withhold ECMO from subjects who are failing
mechanical ventilation alone, and have a need to permit crossover within the study,
which may bias results toward the null hypothesis when analyzed by intention to
treat. Nonetheless, a carefully designed clinical trial that adequately matches for
baseline demographics and potential confounders and adheres to treatment protocols
would provide useful additional evidence to help settle the ongoing debate about the
role of ECMO in severe ARDS.
Indications
There is no single set of accepted criteria for the initiation of ECMO in ARDS, and
the threshold for initiation of ECMO varies considerably across studies and guide-
lines. The decision to initiate ECMO would ideally be based on a riskbenefit analy-
sis 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. Unfortunately,
precise data to inform this sort of decision-making do not exist. As riskbenefit
analysis improves, thresholds for ECMO initiation will likely change to reflect risk
benefit tradeoffs compared with standard-of-care mechanical ventilation.
The early randomized trials used two sets of criteria to assess the need for
ECMO: PaO2 less than 50 mmHg for 2 h at FIO2 1.0 and PEEP greater 5 cm H2O
(fast-entry criteria) or PaO2 less than 50 mmHg for greater than 12 h at FIO2 greater
than 0.6, PEEP greater than 5 cm H2O, and shunt fraction (Qs/Qt) greater than 0.3
despite 48 h of maximal medical therapy (slow-entry criteria) [7, 12]. While these
criteria take into account the risks of prolonged hypoxemia and oxygen toxicity,
there is no consideration of hypercapnia with resulting acidemia or plateau airway
pressures as factors that may influence the decision to initiate ECMO. Furthermore,
the specific time and oxygenation cutoffs used for entry criteria have not been inde-
pendently validated or shown to correlate with mortality. Among the best data avail-
able to estimate the prognosis of patients with ARDS without ECMO support comes
from the work of the ARDS Definition Task Force [1], which redefined ARDS based
on the degree of hypoxemia as a predictor of mortality. Lower PaO2 to FIO2 ratio
cutoffs correlate with increased mortality (45 % when PaO2 to FIO2 is less than 100,
vs. 2732 % in mild to moderate ARDS) and longer duration of mechanical ventila-
tion (9 days in severe ARDS vs. 57 days in mild to moderate ARDS). Based on
these data, patients with lower ratios of PaO2 to FIO2 would seem to benefit the most
from aggressive interventions. However, the appropriate cut off for initiating ECMO
has yet to be determined. Given the consensus decision by the ARDSnet trial inves-
tigators to set a goal PaO2 of 5580 mmHg [29], ultimately representing a range of
values within which the amount of PEEP or FIO2 is not deescalated because the
patient is presumed to be near the steep portion of the oxyhemoglobin dissociation
curve, one may consider a PaO2 to FIO2 ratio of 80 as a reasonable cutoff for con-
sideration of ECMO (Table 2.3) [2]. This threshold is similar to the PaO2 to FIO2
68 D. Abrams et al.
Table 2.3 Indications and contraindications for ECMO in ARDS

Indications
Severe hypoxemia (e.g. ratio of PaO2 to FIO2 < 80 despite the application of high levels of PEEP,
typically 1520 cm H2O), in patients with potentially reversible respiratory failure
Uncompensated hypercapnia with acidemia (pH < 7.15) despite optimal ventilator management
Excessively high end-inspiratory plateau pressure (>3545 cm of water, according to the
patients body size) despite optimal ventilator management
Relative contraindications
High-pressure ventilation (plateau pressure >30 cm H2O) for >7 days
High FIO2 requirements (>0.8) for >7 days
Limited vascular access
Any condition or organ dysfunction that would limit the likelihood of overall benefit from
ECMO, such as severe, irreversible brain injury or untreatable metastatic cancer
Any condition that precludes the use of anticoagulation
Thrombotic thrombocytopenic purpura
Absolute contraindications
ECMO as bridge to lung transplantation if transplantation will not be considered
ARDS acute respiratory distress syndrome, ECMO extracorporeal membrane oxygenation, PaO2
partial pressure of arterial oxygen, FIO2 fraction of inspired oxygen, PEEP positive end-expiratory
pressure, TTP thrombotic thrombocytopenic purpura
ratio that was used as part of the slow entry criteria in the early ECMO trials [7, 12],
and at the upper limit of many observational studies [1522, 24, 25]. The PaO2 to
FIO2 ratio is one component of the Lung Injury Score which uses a combination of
physiologic and radiographic characteristics to quantify the extent of lung injury in
ARDS [34]. However, because of its non-physiologic component and a lack of data
to support its ability to predict survival among patients with the most severe forms
of ARDS, it may be less useful as a criterion for initiation of ECMO.
The amount of PEEP that is appropriate to achieve adequate oxygenation in
severe ARDS prior to initiation of ECMO has not been well established and varies
between patients. There was no difference in survival in several studies of high ver-
sus low PEEP strategy in ARDS [3840]. Nonetheless, many of the studies involv-
ing ECMO for ARDS have documented high levels of PEEP to improve oxygenation
prior to the initiation of ECMO. It is reasonable to attempt to achieve levels of PEEP
up to 1520 cm H2O, in conjunction with increases in FIO2, to reach an adequate
level of oxygenation prior to the initiation of ECMO, assuming those levels of PEEP
do not significantly compromise the patients hemodynamic status.
In addition to oxygenation parameters, the ARDSnet protocol specifies a thresh-
old of pH less than 7.15 (usually a result of uncompensated hypercapnia in the set-
ting of poor lung compliance) as the point at which plateau airway pressure targets
may be exceeded in order to increase minute ventilation and correct the acidemia. A
pH less than 7.15 may therefore be a reasonable threshold to initiate ECMO in
attempting to avoid exceeding plateau airway pressure limits that may worsen
ventilator-associated lung injury. Along similar lines, a third indication for the ini-
tiation of ECMO is excessively high plateau airway pressures themselves, despite

adherence to the best accepted standard of care for ventilator management. Plateau
airway pressures exceeding 35 cm H2O have been shown to correlate with signifi-
cantly higher incidence of barotrauma [41]. A reasonable range of plateau airway
pressures in which to consider ECMO initiation is 3545 cm H2O depending on the
patients body habitus [2]. Patients with higher BMI will have higher elastic loads
on the chest wall than those with lower BMI, and thus may require higher plateau
airway pressures to achieve the same degree of alveolar patency, though the trans-
pulmonary pressure gradient may not be higher.
Exclusions
There are few absolute contraindications for ECMO in severe ARDS. Any condi-
tion that precludes the use of anticoagulation is typically considered an absolute
contraindication because of the need for systemic anticoagulant therapy to main-
tain the integrity of the circuit. However, in patients with severe bleeding, antico-
agulation may be withheld for significant periods of time with apparent safety
[42]. Given the advances in technology, the inability to anticoagulate should per-
haps be considered only a relative contraindication to receiving ECMO. Thrombotic
thrombocytopenic purpura (TTP) is another disorder that may be problematic
because of the high risk of thrombosis (within the patient or the circuit) if platelet
transfusion is needed in the setting of life-threatening hemorrhage and thrombocy-
topenia. While this may not be an absolute contraindication, the likelihood of sur-
vival without ECMO would have to be sufficiently low to consider accepting such
a high risk.
The theoretical benefit of ECMO for severe cases of ARDS largely relates to
lung protection: a decreased need for high-pressure ventilation or toxic fractions of
inspired oxygen. Yet, this potential for benefit may be limited if the patient has
already been exposed to high airway pressures and inspired oxygen levels such that
damage is irreversible. Many clinicians think that patients who have been receiving
high-pressure ventilation with plateau pressure exceeding 30 cm H2O for greater
than 7 days are less likely to benefit from ECMO [4345]. Similarly, prolonged
exposure to high fractions of inspired oxygen, which may induce lung inflamma-
tion, could nullify any beneficial effect of ECMO support, though this remains an
area of controversy [46, 47]. Earlier initiation of ECMO, perhaps for these or other
reasons, has been associated with better outcomes in some, but not all, observational
studies [18, 45, 48, 49]. Other relative contraindications include limitations in vas-
cular access that would preclude cannula placement and any conditions in which
ECMO would be unlikely to alter the patients overall prognosis, including but not
limited to advanced malignancy or severe and irreversible brain injury. Finally,
ECMO, when considered as a bridge to lung transplantation, should only be offered
if the patient is actually a candidate for transplantation.
70 D. Abrams et al.
References
1. Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, et al. Acute
respiratory distress syndrome: the Berlin Definition. JAMA. 2012;307(23):252633.
2. Brodie D, Bacchetta M. Extracorporeal membrane oxygenation for ARDS in adults. N Engl J
Med. 2011;365(20):190514.
3. Hill JD, OBrien 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):62934.
4. Gille JP. [Respiratory support by extracorporeal circulation with a membrane artificial lung
(authors transl)]. Bull Physiopathol Respir (Nancy). 1974;10(3):373410.
5. Bartlett RH, Gazzaniga AB, Fong SW, Burns NE. Prolonged extracorporeal cardiopulmonary
support in man. J Thorac Cardiovasc Surg. 1974;68(6):91832.
6. Hill JD, Ratliff JL, Fallat RH, Tucker HJ, Lamy M, Dietrich HP, et al. Prognostic factors in the
treatment of acute respiratory insufficiency with long-term extracorporeal oxygenation.
J Thorac Cardiovasc Surg. 1974;68(6):90517.
7. 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):21936.
8. Gattinoni L, Pesenti A, Caspani ML, Pelizzola A, Mascheroni D, Marcolin R, et al. The role of
total static lung compliance in the management of severe ARDS unresponsive to conventional
treatment. Intensive Care Med. 1984;10(3):1216.
9. Gattinoni L, Pesenti A, Mascheroni D, Marcolin R, Fumagalli R, Rossi F, et al. Low-frequency
positive-pressure ventilation with extracorporeal CO2 removal in severe acute respiratory fail-
ure. JAMA. 1986;256(7):8816.
10. Gattinoni L, Kolobow T, Agostoni A, Damia G, Pelizzola A, Rossi GP, et al. Clinical applica-
tion of low frequency positive pressure ventilation with extracorporeal CO2 removal (LFPPV-
ECCO2R) in treatment of adult respiratory distress syndrome (ARDS). Int J Artif Organs.
1979;2(6):2823.
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(4):
1835.
12. 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):
295305.
13. Wagner PK, Knoch M, Sangmeister C, Muller E, Lennartz H, Rothmund M. Extracorporeal
gas exchange in adult respiratory distress syndrome: associated morbidity and its surgical
treatment. Br J Surg. 1990;77(12):13958.
14. Brunet F, Belghith M, Mira JP, Lanore JJ, Vaxelaire JF, Dall'ava Santucci J, et al. Extracorporeal
carbon dioxide removal and low-frequency positive-pressure ventilation. Improvement in arte-
rial oxygenation with reduction of risk of pulmonary barotrauma in patients with adult respira-
tory distress syndrome. Chest. 1993;104(3):88998.
15. Manert W, Haller M, Briegel J, Hummel T, Kilger E, Polasek J, et al. Venovenous extracorpo-
real membrane oxygenation (ECMO) with a heparin-lock bypass system. An effective addition
in the treatment of acute respiratory failure (ARDS). Anaesthesist. 1996;45(5):43748.
16. Kolla S, Awad SS, Rich PB, Schreiner RJ, Hirschl RB, Bartlett RH. Extracorporeal life support
for 100 adult patients with severe respiratory failure. Ann Surg. 1997;226(4):54464. discus-
sion 565-6.
17. Peek GJ, Moore HM, Moore N, Sosnowski AW, Firmin RK. Extracorporeal membrane oxy-
genation for adult respiratory failure. Chest. 1997;112(3):75964.
18. Lewandowski K, Rossaint R, Pappert D, Gerlach H, Slama KJ, Weidemann H, et al. High
survival rate in 122 ARDS patients managed according to a clinical algorithm including extra-
corporeal membrane oxygenation. Intensive Care Med. 1997;23(8):81935.
19. Linden V, Palmer K, Reinhard J, Westman R, Ehren H, Granholm T, et al. High survival in
adult patients with acute respiratory distress syndrome treated by extracorporeal membrane
oxygenation, minimal sedation, and pressure supported ventilation. Intensive Care Med.
2000;26(11):16307.
20. Bartlett RH, Roloff DW, Custer JR, Younger JG, Hirschl RB. Extracorporeal life support: the
University of Michigan experience. JAMA. 2000;283(7):9048.
21. 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(17):188895.
22. Schmid C, Philipp A, Hilker M, Rupprecht L, Arlt M, Keyser A, et al. Venovenous extracorpo-
real membrane oxygenation for acute lung failure in adults. J Heart Lung Transplant.
2012;31(1):915.
23. Davies A, Jones D, Gattas D. Extracorporeal membrane oxygenation for ARDS due to 2009
influenza A(H1N1)reply. JAMA. 2010;303(10):9412.
24. Roch A, Lepaul-Ercole R, Grisoli D, Bessereau J, Brissy O, Castanier M, et al. Extracorporeal
membrane oxygenation for severe influenza A (H1N1) acute respiratory distress syndrome: a
prospective observational comparative study. Intensive Care Med. 2010;36(11):1899905.
25. Freed DH, Henzler D, White CW, Fowler R, Zarychanski R, Hutchison J, et al. Extracorporeal
lung support for patients who had severe respiratory failure secondary to influenza A (H1N1)
2009 infection in Canada. Can J Anaesth. 2010;57(3):2407.
26. Miller 3rd RR, 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):7528.
27. 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):165968.
28. Matthay MA, Ware LB, Zimmerman GA. The acute respiratory distress syndrome. J Clin
Invest. 2012;122(8):273140.
29. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung
injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome
Network. N Engl J Med. 2000;342(18):13018.
30. Matthay MA, Brower RG, Carson S, Douglas IS, Eisner M, Hite D, et al. Randomized,
placebo-controlled clinical trial of an aerosolized beta(2)-agonist for treatment of acute lung
injury. Am J Respir Crit Care Med. 2011;184(5):5618.
31. Rice TW, Wheeler AP, Thompson BT, deBoisblanc BP, Steingrub J, Rock P. Enteral omega-3
fatty acid, gamma-linolenic acid, and antioxidant supplementation in acute lung injury. JAMA.
2011;306(14):157481.
32. Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, deBoisblanc B, et al.
Comparison of two fluid-management strategies in acute lung injury. N Engl J Med.
2006;354(24):256475.
33. 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):135163.
34. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded definition of the adult respiratory
distress syndrome. Am Rev Respir Dis. 1988;138(3):7203.
35. Zwischenberger JB, Lynch JE. Will CESAR answer the adult ECMO debate? Lancet.
2009;374(9698):13078.
36. Dalton HJ, MacLaren G. Extracorporeal membrane oxygenation in pandemic flu: insufficient
evidence or worth the effort? Crit Care Med. 2010;38(6):14845.
72 D. Abrams et al.
37. Pellegrino VA, Davies AR. CESAR: deliverance or just the beginning? Crit Care Resusc.
2010;12(2):757.
38. Brower RG, Lanken PN, MacIntyre N, Matthay MA, Morris A, Ancukiewicz M, et al. Higher
versus lower positive end-expiratory pressures in patients with the acute respiratory distress
syndrome. N Engl J Med. 2004;351(4):32736.
39. Meade MO, Cook DJ, Guyatt GH, Slutsky AS, Arabi YM, Cooper DJ, et al. Ventilation strat-
egy using low tidal volumes, recruitment maneuvers, and high positive end-expiratory pres-
sure for acute lung injury and acute respiratory distress syndrome: a randomized controlled
trial. JAMA. 2008;299(6):63745.
40. Mercat A, Richard JC, Vielle B, Jaber S, Osman D, Diehl JL, et al. Positive end-expiratory
pressure setting in adults with acute lung injury and acute respiratory distress syndrome: a
randomized controlled trial. JAMA. 2008;299(6):64655.
41. International consensus conferences in intensive care medicine: ventilator-associated lung
injury in ARDS. This official conference report was cosponsored by the American Thoracic
Society, The European Society of Intensive Care Medicine, and The Societe de Reanimation
de Langue Francaise, and was approved by the ATS Board of Directors, July 1999. Am J
Respir Crit Care Med. 1999;160(6):211824.
42. Muellenbach RM, Kredel M, Kunze E, Kranke P, Kuestermann J, Brack A, et al. Prolonged
heparin-free extracorporeal membrane oxygenation in multiple injured acute respiratory
distress syndrome patients with traumatic brain injury. J Trauma Acute Care Surg. 2012;
72(5):14447.
43. Rouby JJ, Brochard L. Tidal recruitment and overinflation in acute respiratory distress syn-
drome: yin and yang. Am J Respir Crit Care Med. 2007;175(2):1046.
44. Pugin J, Verghese G, Widmer MC, Matthay MA. The alveolar space is the site of intense
inflammatory and profibrotic reactions in the early phase of acute respiratory distress syn-
drome. Crit Care Med. 1999;27(2):30412.
45. Pranikoff T, Hirschl RB, Steimle CN, Anderson HL, 3rd, Bartlett RH. Mortality is directly
related to the duration of mechanical ventilation before the initiation of extracorporeal life
support for severe respiratory failure. Crit Care Med. 1997;25(1):2832.
46. Jackson RM. Pulmonary oxygen toxicity. Chest. 1985;88(6):9005.
47. Davis WB, Rennard SI, Bitterman PB, Crystal RG. Pulmonary oxygen toxicity. Early revers-
ible changes in human alveolar structures induced by hyperoxia. N Engl J Med. 1983;
309(15):87883.
48. Beiderlinden M, Eikermann M, Boes T, Breitfeld C, Peters J. Treatment of severe acute respi-
ratory distress syndrome: role of extracorporeal gas exchange. Intensive Care Med.
2006;32(10):162731.
49. Mols G, Loop T, Geiger K, Farthmann E, Benzing A. Extracorporeal membrane oxygenation:
a ten-year experience. Am J Surg. 2000;180(2):14454.
Chapter 3
Cardiogenic Shock: Evidence, Indications,
and Exclusions
Nicolas Brchot and Alain Combes
Introduction
Extracorporeal membrane oxygenation (ECMO) is one of several mechanical circu-

latory support devices used for patients with refractory cardiac failure [1]. Advances
in both extracorporeal technology and cannulation techniques have led to an
improved riskbenet prole, increased the use, and broadened the potential appli-
cations for ECMO in these circumstances. The advantage of ECMO over other per-
cutaneous devices (intraaortic balloon counterpulsation devices [2], TandemHeart
left ventricular assist devices [3] (CardiacAssist, Inc., Pittsburgh, PA), and Impella
devices [4] (Abiomed, Danvers, MA)) results from the rapidity of insertion, the
ability to support right ventricular, left ventricular, or biventricular failure at high
blood ow rates, the potential to support patients with concomitant lung injury
when needed [1, 5], and from its lower cost.
N. Brchot, MD, PhD

Service de Ranimation Mdicale, Hospital Piti-Salptrire,
47, Boulevard de lHpital, Paris 75013, France
A. Combes, MD, PhD (*)
Service de Ranimation Mdicale, Institut de Cardiologie, Groupe Hospitalier
Piti-Salptrire, iCAN, Institute of Cardiometabolism and Nutrition,
47, Boulevard de lHpital, Paris Cedex 13 75651, France
e-mail: alain.combes@psl.aphp.fr

74 N. Brchot and A. Combes
Patient and Disease-Specific Issues
Indication for VA-ECMO
VA-ECMO should be considered in cases of cardiogenic shock refractory to con-

ventional treatment including high-dose inotropes and vasoconstrictors (epineph-
rine > 0.2 g/kg/min or dobutamine > 20 g/kg/min norepinephrine > 0.2 g/kg/
min) and preload optimization. Doppler echocardiography should conrm low car-
diac output (cardiac index <2.2 L/min/m2, or left ventricular ejection fraction
(LVEF) <20 % and aortic velocity-time integral <8 cm) and persistent tissue hypoxia
[1, 59]. Contraindications to ECMO include irreversible heart failure in patients
not candidates for left ventricular assist device (LVAD) or heart transplantation, and
futility due to the patients overall condition. Other classical contraindications (anti-
coagulation, age, chronic organ dysfunction, compliance with medical treatment)
are relative, considering the fatal course of refractory cardiogenic shock without
circulatory assistance.
Configurations and Cannulation Strategies
VA-ECMO drains the blood from the right atrium via a femoral venous or internal
jugular venous cannula or, in patients with an open chest, directly from the right
atrium [5, 10, 11]. The blood is pumped through a membrane oxygenator allowing
oxygen to be added and carbon dioxide to be removed. After passing through the
oxygenator, blood is then actively pumped into the arterial system either via a can-
nula placed in a peripheral artery, usually femoral or subclavian (closed chest), or
directly into the aorta (open chest). VA-ECMO is typically a high blood ow extra-
corporeal circuit that can pump up to 7 L/min and provide full or partial cardiopul-
monary support. VA-ECMO is a closed system, which differs from standard
cardiopulmonary bypass used in the operating room, which is an open system with
a bloodair interface.
ECMO Retrieval Units
Transfer of medically unstable patients to tertiary centers without rst beginning

extracorporeal support may be associated with markedly increased risk. Therefore,
initiating ECMO in remote institutions followed by stabilization and transfer to
tertiary centers under ECMO might signicantly improve survival. Patient retrieval
under ECMO was rst reported by Cornish in 1986 [12]. Since then successful
transportation of patients on cardiopulmonary support has been described for short
and long distances by ambulance, helicopter, and airplane [1317]. Mobile ECMO
retrieval teams for the provision of ECMO support to critically ill patients should be
3 Cardiogenic Shock: Evidence, Indications, and Exclusions 75
created as part of critical care tertiary referral networks of hospitals covering

regional or national areas. A policy directive should be established within the net-
work to create standardized referral indications and to assure 24 h a day, 7 days a
week stafng, equipment, and transport. This strategy should allow safe transporta-
tion under cardiopulmonary support to experienced tertiary centers and might ulti-
mately improve survival of the sickest respiratory or cardiac failure patients initially
treated in centers where ECMO is not possible [13].
ECMO in Specific Situations
Acute Myocardial Infarction
There are no randomized controlled trials comparing ECMO with other mechanical
support systems in myocardial infarction-associated cardiogenic shock; however,
several nonrandomized studies suggest a survival advantage from the early use of
ECMO in such circumstances [8, 18]. A nonrandomized, prospective, observational
study compared patients with ST-segment elevation myocardial infarction-related
profound cardiogenic shock undergoing PCI with (n = 46) and without (n = 25)
ECMO support. Those receiving ECMO had a signicantly lower 30-day mortality
than those who did not receive ECMO (39.1 % vs. 72 %, p = 0.008) [8]. Interpretation
of this data is limited by the fact that each cohort was enrolled over a different period
(19932002 for the non-ECMO cohort versus 20022009 for the ECMO cohort),
potentially leading to discrepancies in both medical and interventional management
between groups, especially given that coronary stents were unavailable at the study
center prior to 1998. Additionally, higher rates of TIMI (thrombolysis in myocardial
infarction) grade 3 ow were achieved in the ECMO group, which may have been a
consequence of improved hemodynamic stability in the catheterization lab; due to
improved PCI technique over time; or other potential explanations. Ultimately, a
prospective randomized controlled trial is needed to determine the true benet, if
any, of VA-ECMO in myocardial infarction-associated cardiogenic shock.
ECMO and Refractory Cardiac Arrest (ECPR)
Extracorporeal CPR (ECPR) relates to the initiation of ECMO during refractory

cardiac arrest. According to the 2010 American Heart Association Guidelines for
Cardiopulmonary Resuscitation and Emergency Cardiovascular Care, ECPR may
be considered when the time without spontaneous circulation is short, resuscitation
attempts are adequate, and the cause of cardiac arrest is potentially reversible or
amenable to heart transplantation or revascularization [19]. The guidelines empha-
size that ECPR use should be restricted to centers in which it is readily available,
and that its initiation and management require highly trained personnel and special-
ized equipment.
Several observational studies comparing conventional CPR and ECPR have been
reported [2024]. In a prospective, non-randomized, observational study of patients
with witnessed in-hospital cardiac arrests of cardiac origin lasting longer than 10
min, propensity analysis matching 46 subjects who received conventional CPR to
46 subjects who received ECPR demonstrated a signicantly higher survival at dis-
charge in the ECPR group (32.6 % vs. 17.4 %; HR 0.51, 95 % CI 0.350.74,
p < 0.0001) and at 1 year (HR 0.53; 95 % CI 0.330.83, p = 0.006) [20]. There was a
trend toward improved neurological outcomes at both time points, however these
results did not reach statistical signicance. Multivariable analysis revealed that an
initial rhythm of ventricular brillation or ventricular tachycardia and use of ECPR
were positively associated with survival to discharge, whereas longer duration of
CPR was negatively associated with survival. Importantly, survival fell to only 17 %
if more than 60 min elapsed from collapse to initiation of ECPR [20]. A more recent
analysis of 406 subjects suffering in-hospital cardiac arrest (of which only 120 sub-
jects underwent matched propensity analysis) demonstrated a signicantly higher
rate of 2-year survival with minimal neurological impairment among those treated
with ECPR compared with conventional CPR (20.0 % vs. 5.0 %; HR = 0.53, 95 %
CI = 0.360.80, p = 0.002). Independent predictors of good neurological outcome
included age less than 65 years (HR = 0.46; 95 % CI = 0.260.81, p = 0.008), CPR
duration of less than 35 min (HR = 0.37, 95 % CI = 0.180.76, p = 0.007), and subse-
quent cardiovascular intervention (HR = 0.36, 95 % CI = 0.180.68, p = 0.002) [24].
Results of ECPR for out-of-hospital cardiac arrest are poorer [18, 22, 2527].
Japanese series suggest that good outcomes can be obtained in 1520 % of the
patients, with provision of ECMO support in less than 60 min [18, 22, 25, 26]. To
attain this objective, out-of-hospital cardiac arrest patients should be scooped and
run as rapidly as possible to the nearest ECMO center [28].
ECMO for Acute Myocarditis
Myocarditis is a disease that may progress rapidly to refractory cardiogenic shock

and death. In such situations, emergent initiation of mechanical circulatory assis-
tance is the only therapeutic option to rescue these dying patients. Considering the
prompt myocardial recovery in most of the patients, the easy and rapid implantation
and explantation of peripheral venoarterial ECMO has made it the elective rst-line
assistance device for those patients [29, 30].
Asaumi and colleagues demonstrated that patients with fulminant myocarditis
who were successfully bridged to recovery with ECMO had long-term prognoses
comparable to hemodynamically stable patients with acute myocarditis [30]. In a
French cohort of 33 patients (mean age of 38 years) receiving peripheral VA-ECMO
for fulminant myocarditis with refractory cardiogenic shock, survival to discharge
was 70 % [10]. This high survival rate is remarkable in light of the extreme disease
severity at the time of ECMO implantation. Higher severity of illness (Simplied
Acute Physiology Score II 56) and elevated cardiac biomarkers (troponin Ic 12
g/L) were independent predictors of intensive care unit mortality and an inability
to wean from ECMO [10]. Of the 26 patients (65 %) with long-term follow-up,
health-related quality of life physical domain scores were lower than age- and
gender-matched control subjects. However, they compared favorably to subjects
from other cohorts who were either bridged with ventricular assist devices to heart
transplantation or were long-term survivors of ARDS [31, 32].
ECMO and Drug Intoxication
Cardiotoxic drug intoxication may rapidly evolve to refractory cardiogenic shock

and cardiac arrest and the reversibility of cardiovascular dysfunction makes
VA-ECMO a recommended procedure in this situation [19, 33]. The largest cohort
to date reported on a series of 62 patients, of whom 14 received peripheral VA-ECMO
[34]. ECMO-treated patients had better survival than other patients (86 % vs. 48 %,
p = 0.02). Of note, none of the patients with persistent cardiac arrest survived with-
out ECMO support. Factors independently associated with lower mortality accord-
ing to multivariable analysis were beta-blocker intoxication, lower SAPS II score,
and ECMO support (OR, 0.18; 95 % CI, 0.030.96; p = 0.04).
ECMO Postcardiotomy
Postcardiotomy cardiogenic shock is an uncommon but dreadful complication of

cardiac surgery [3539]. ECMO may be considered as temporary support postop-
eratively, particularly in those who are unable to be weaned from cardiopulmonary
bypass in the operating room. Mortality among patients requiring this level of sup-
port remains high [37]. In the largest cohort published to date, although more than
half of the patients could be weaned from ECMO, only 24 % were discharged
home, and 1-year survival was 17 %. Age > 70, diabetes, obesity, preoperative renal
insufciency, preoperative LVEF, and preimplantation acidosis were indepen-
dently associated with a poor outcome, while patients who had isolated coronary
artery bypass grafting had better outcomes [37]. A more recent study of 87 patients
who received ECMO after valvular surgery in China reported better outcomes.
ECMO could be explanted in 59 % of the patients, and 49 % were discharged
home. Age >65 and renal replacement therapy were independent predictors of a
poor outcome [40].
ECMO and Heart Transplantation
ECMO may be used as a bridge to heart transplantation or VAD insertion in

INTERMACS class I patients or as a bridge to decision when prognosis is uncertain
[4146]. However, the duration of support that can be provided is shorter than for
VADs, making transplantation or transition to VAD of greater urgency, and patients

receiving ECMO support must remain within the intensive care unit [5]. Success of
ECMO as a bridging therapy varies greatly, and depends in large part on pre-ECMO
patient characteristics and organ availability in the cases where transplantation is the
goal. In a retrospective single-center review of 70 subjects in whom ECMO was
used with the intention to bridge to heart transplantation, 31 subjects (44 %) were
bridged to either heart transplant (n = 15) or VAD (n = 16), though only 11 (73 %)
and 8 (50 %) of the heart transplant and VAD recipients, respectively, survived to
hospital discharge, highlighting the limitations of bridging therapy in this highly
morbid patient population [44]. Age greater than 50, CPR prior to ECMO initiation,
and high sequential organ failure assessment (SOFA) score were identied as inde-
pendent predictors of unsuccessful bridging. Preimplantation CPR as a predictor of
poor outcome was corroborated in a more recent single-center study of 90 patients
who received mechanical circulatory support (VAD or ECMO) for refractory cardio-
genic shock [47]. Forty-nine percent received short-term VAD support as a bridge to
decision and 51 % received ECMO when neurologic status was uncertain or there
was complete hemodynamic collapse or severe coagulopathy. Overall survival was
49 %, with 26 % of patients transitioned to implantable VADs, 18 % recovering suf-
cient native cardiac function, and 11 % bridged to transplantation. Twenty-one
patients (23 %) were receiving CPR at the time of implantation, the occurrence of
which was an independent predictor of in-hospital mortality (OR 5.79, 95 % CI
1.28526.08, p = 0.022). These studies highlight the need for careful consideration
of relative and absolute contraindications to mechanical circulatory support.
ECMO support is also used in the case of primary graft failure (PGF) following
heart transplantation [48, 49]. Overall survival for patients with PGF requiring
ECMO is worse than for those who do not develop PGF. However, patients with
ECMO-supported PGF who survive beyond the early posttransplant period have
comparable long-term survival to non-PGF transplant recipients [49, 50].
ECMO to Prevent Acute Right Ventricular Failure Following

LVAD Implantation
Femoral VA-ECMO [5153] or a percutaneous venous to pulmonary artery form of

venovenous ECMO (either with a graft attached to the main pulmonary artery and
passed through a subxiphoid exit [54, 55] or with a exible outow cannula placed
from the right internal jugular vein to the pulmonary artery [56]) has been success-
fully used to provide peri- and postoperative right ventricular support in patients
with preoperative biventricular dysfunction undergoing LVAD implantation, who
otherwise would have been candidates for biventricular assist device. In this setting,
ECMO can allow time for the already compromised right ventricle to adapt to the
increasing preload, thereby avoiding distension and right ventricular failure leading
to poor lling of the LVAD.
ECMO and Profound Hypothermia
ECMO has been used to rewarm victims of accidental deep hypothermia (<28 C).
Unlike other rewarming techniques, it restores organ perfusion immediately in
patients with inadequate circulation. In a series of 46 patients with accidental hypo-
thermia, 32 had rewarming with ECMO bypass, resulting in 15 long-term survivors
[57]. In most of these patients, deep hypothermia developed after mountaineering
accidents or suicide attempts. The mean interval from discovery of the patient to
rewarming with cardiopulmonary bypass was 141 50 min. At follow-up there were
no hypothermia-related sequelae that impaired quality of life in survivors.
Neurologic and neuropsychological decits observed in the early period after
rewarming had fully or almost completely disappeared. One patient had cerebellar
atrophy on magnetic resonance imaging with mild clinical signs, a condition that
may have been caused by hypothermia. This landmark study demonstrated that
VA-ECMO can rescue and rewarm young, otherwise healthy people who had acci-
dental deep hypothermia with no or minimal cerebral impairment, even following
prolonged circulatory arrest.
ECMO and Massive Pulmonary Embolism
Successful rescue therapy with ECMO has been described in several cases of life-
threatening pulmonary embolism [58, 59], even in patients with persistent cardiac
arrest [60]. Whether adjunctive treatments such as catheter-guided thrombectomy
or surgical embolectomy should be performed in ECMO-treated patients to hasten
recovery is still debated [59, 60], since complete lysis of pulmonary artery clot has
been reported after a few days on ECMO and heparin treatment [58, 61, 62].
ECMO for Refractory Septic Shock
The use of mechanical circulatory support remains controversial for refractory sep-
tic shock. ECMO was shown to be effective in this situation as a salvage therapy in
children [6365]. Alternatively, poor results were reported for adult patients who
received ECMO in the context of sepsis-associated refractory vasoplegia and pre-
served cardiac output [66]. However, profound myocardial dysfunction can also
occur during bacterial septic shock. In a recent series of 14 patients who received
VA-ECMO for refractory cardiovascular failure following bacterial septic shock
[67], 12 (86 %) could be weaned and 10 patients (71 %) were long-term survivors
who recovered normal left ventricular function, and reported a highly preserved
quality of life. Larger studies are needed to determine whether the benet of ECMO
outweighs the risk, especially in cases where septic shock is complicated by marked
disturbances in coagulation.
ECMO-Associated Complications
The benets of ECMO must be weighed against the risks inherent in its use. Because
the ECMO system injects oxygenated blood into the descending aorta in counter-
current fashion, it increases left ventricular (LV) afterload and, in 1520 % of
patients, is associated with severe hydrostatic pulmonary edema. The latter is even
more frequent in patients with little or no residual LV ejection on ECMO. In this
context, hydrostatic pulmonary edema is further aggravated by mitral regurgitation
induced by LV dilation and increased LV end-diastolic pressure, resulting from poor
LV unloading [68]. Restoring pulsatility and decreasing cardiac afterload with an
intra-aortic balloon pump in patients receiving peripheral VA-ECMO was associ-
ated with lower LVED dimensions and pulmonary artery pressures both in animals
[69] and humans [70].
Hemorrhagic complications are among the most frequent adverse events [5]. The
use of lower levels of anticoagulation limits bleeding risks, with some centers tar-
geting activated partial thromboplastin times as low as 4060 s. There are, however,
no universally accepted anticoagulation protocols, and anticoagulation may need to
be adjusted to the specic needs of the individual patient. Thromboses, either within
the circuit or related to the indwelling portions of the cannulae, may result in stroke.
Infectious complications have been reported to varying degrees, with some
reports indicating longer durations of mechanical ventilation, ECMO support, and
hospital stays associated with infection [71]. Limb ischemia and compartment syn-
drome are of concern in VA-ECMO. Insertion of a distal reperfusion cannula into
the supercial femoral artery should be considered to perfuse blood to the extrem-
ity. Other complications that have been associated with ECMO include hemolysis,
thrombocytopenia, acquired von Willebrand syndrome, disseminated intravascular
coagulopathy, and air embolism [72]. In a recent meta-analysis incorporating 1866
patients from 20 studies of ECMO for cardiogenic shock or cardiac arrest between
2000 and 2012, complication rates (95 % CI) were reported as follows: lower
extremity ischemia, 16.9 % (12.522.6); fasciotomy or compartment syndrome,
10.3 % (7.314.5); lower extremity amputation, 4.7 % (2.39.3); stroke, 5.9 %
(4.28.3); major or signicant bleeding, 40.8 % (26.856.6); re-thoracotomy for
postcardiotomy bleeding or tamponade, 41.9 % (24.361.8); and signicant infec-
tion, 30.4 % (19.544.0) [73]. Cardiac or major vascular perforation is a rare but
potentially lethal complication of cannulation, the frequency of which depends on
the type of cannula used. Ultimately, complication rates will vary according to insti-
tutional experience and patient selection.
Outcomes After VA-ECMO
VA-ECMO can rescue up to 80 % of refractory cardiogenic shocks patients [5, 74].

In a series of 81 patients who received ECMO support for medical (n = 55), postcardi-
otomy (n = 16), or posttransplantation (n = 10) cardiogenic shock, early independent
predictors of ICU death included ECMO implantation under CPR, severe liver or
renal failure, and female sex, while myocarditis is associated with better survival
[5]. To improve outcomes, patients with rapidly progressing cardiac dysfunction
should be urgently transferred to experienced centers, where cardiologists, heart
surgeons, and intensivists can implement all the complex medical and surgical inter-
ventions required by these critically ill patients, and can also rapidly diagnose and
treat ECMO-related complications.
Health-related quality of life, evaluated in the 28 long-term survivors of the latter
study [5], revealed persistent problems with work or other daily activities because of
physical health and frequent interferences with normal social activities, while mental
health and vitality were deemed satisfactory. Higher SF-36 scores were also obtained
for patients with longer follow-up, suggesting time-dependent improvement. Lastly,
patients who had peripheral VA-ECMO reported persistent problems at the cannula-
insertion site (lymphocele, late wound-healing, symptoms related to crural nerve
injury such as skin numbness or paresthesia), which tended to regress over time.
Conclusion
Major advances in extracorporeal technology have favorably altered the riskbenet

prole of ECMO for severe cardiovascular failure. The technique is currently exten-
sively used worldwide to rescue dying cardiogenic shock patients. Additional stud-
ies are now needed to rene criteria for ECMO initiation in each of the proposed
indications to ultimately improve short- and long-term outcomes of these patients.
References
1. Werdan K, Gielen S, Ebelt H, Hochman JS. Mechanical circulatory support in cardiogenic

shock. Eur Heart J. 2014;35:15667.
2. Thiele H, Zeymer U, Neumann FJ, Ferenc M, Olbrich HG, et al. Intraaortic balloon support for
myocardial infarction with cardiogenic shock. N Engl J Med. 2012;367:128796.
3. Thomas JL, Al-Ameri H, Economides C, Shareghi S, Abad DG, et al. Use of a percutaneous
left ventricular assist device for high-risk cardiac interventions and cardiogenic shock. J
Invasive Cardiol. 2010;22:3604.
4. Engstrom AE, Cocchieri R, Driessen AH, Sjauw KD, Vis MM, et al. The Impella 2.5 and 5.0
devices for ST-elevation myocardial infarction patients presenting with severe and profound
cardiogenic shock: the Academic Medical Center intensive care unit experience. Crit Care
Med. 2011;39:20729.
5. Combes A, Leprince P, Luyt CE, Bonnet N, Trouillet JL, et al. Outcomes and long-term
quality-of-life of patients supported by extracorporeal membrane oxygenation for refractory
cardiogenic shock. Crit Care Med. 2008;36:140411.
6. Bakhtiary F, Keller H, Dogan S, Dzemali O, Oezaslan F, et al. Venoarterial extracorporeal
membrane oxygenation for treatment of cardiogenic shock: clinical experiences in 45 adult
patients. J Thorac Cardiovasc Surg. 2008;135:3828.
7. Idelchik GM, Simpson L, Civitello AB, Loyalka P, Gregoric ID, et al. Use of the percutaneous
left ventricular assist device in patients with severe refractory cardiogenic shock as a bridge to
long-term left ventricular assist device implantation. J Heart Lung Transplant.
2008;27:10611.
8. Sheu JJ, Tsai TH, Lee FY, Fang HY, Sun CK, et al. Early extracorporeal membrane oxygenator-
assisted primary percutaneous coronary intervention improved 30-day clinical outcomes in
patients with ST-segment elevation myocardial infarction complicated with profound cardio-
genic shock. Crit Care Med. 2010;38:18107.
9. Kim H, Lim SH, Hong J, Hong YS, Lee CJ, et al. Efcacy of veno-arterial extracorporeal
membrane oxygenation in acute myocardial infarction with cardiogenic shock. Resuscitation.
2012;83:9715.
10. Mirabel M, Luyt CE, Leprince P, Trouillet JL, Leger P, et al. Outcomes, long-term quality of
life, and psychologic assessment of fulminant myocarditis patients rescued by mechanical
circulatory support. Crit Care Med. 2011;39:102935.
11. Abrams D, Combes A, Brodie D. Whats new in extracorporeal membrane oxygenation for
cardiac failure and cardiac arrest in adults? Intensive Care Med. 2014;40:60912.
12. Cornish J, Gerstmann D, Begnaud M, Null D, Ackerman N. Inight use of extracorporeal
membrane oxygenation for severe neonatal respiratory failure. Perfusion. 1986;1:281287.
13. Beurtheret S, Mordant P, Paoletti X, Marijon E, Celermajer DS, et al. Emergency circulatory
support in refractory cardiogenic shock patients in remote institutions: a pilot study (the
cardiac-RESCUE program). Eur Heart J. 2013;34:11220.
14. Foley DS, Pranikoff T, Younger JG, Swaniker F, Hemmila MR, et al. A review of 100 patients
transported on extracorporeal life support. ASAIO J. 2002;48:6129.
15. Forrest P, Ratchford J, Burns B, Herkes R, Jackson A, et al. Retrieval of critically ill adults
using extracorporeal membrane oxygenation: an Australian experience. Intensive Care Med.
2011;37:82430.
16. Javidfar J, Brodie D, Takayama H, Mongero L, Zwischenberger J, et al. Safe transport of criti-
cally ill adult patients on extracorporeal membrane oxygenation support to a regional extracor-
poreal membrane oxygenation center. ASAIO J. 2011;57:4215.
17. Linden V, Palmer K, Reinhard J, Westman R, Ehren H, et al. Inter-hospital transportation of
patients with severe acute respiratory failure on extracorporeal membrane oxygenation
national and international experience. Intensive Care Med. 2001;27:16438.
18. Sakamoto S, Taniguchi N, Nakajima S, Takahashi A. Extracorporeal life support for cardio-
genic shock or cardiac arrest due to acute coronary syndrome. Ann Thorac Surg.
2012;94:17.
19. Cave DM, Gazmuri RJ, Otto CW, Nadkarni VM, Cheng A, et al. Part 7: CPR techniques and
devices: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and
Emergency Cardiovascular Care. Circulation. 2010;122:S7208.
20. Chen YS, Lin JW, Yu HY, Ko WJ, Jerng JS, et al. Cardiopulmonary resuscitation with assisted
extracorporeal life-support versus conventional cardiopulmonary resuscitation in adults with
in-hospital cardiac arrest: an observational study and propensity analysis. Lancet.
2008;372:55461.
21. Chen YS, Yu HY, Huang SC, Lin JW, Chi NH, et al. Extracorporeal membrane oxygenation
support can extend the duration of cardiopulmonary resuscitation. Crit Care Med.
2008;36:252935.
22. Maekawa K, Tanno K, Hase M, Mori K, Asai Y. Extracorporeal cardiopulmonary resuscitation
for patients with out-of-hospital cardiac arrest of cardiac origin: a propensity-matched study
and predictor analysis. Crit Care Med. 2013;41:118696.
23. Lin JW, Wang MJ, Yu HY, Wang CH, Chang WT, et al. Comparing the survival between extra-
corporeal rescue and conventional resuscitation in adult in-hospital cardiac arrests: propensity
analysis of three-year data. Resuscitation. 2010;81:796803.
24. Shin TG, Choi JH, Jo IJ, Sim MS, Song HG, et al. Extracorporeal cardiopulmonary resuscita-
tion in patients with inhospital cardiac arrest: a comparison with conventional cardiopulmo-
nary resuscitation. Crit Care Med. 2011;39:17.
25. Kagawa E, Dote K, Kato M, Sasaki S, Nakano Y, et al. Should we emergently revascularize
occluded coronaries for cardiac arrest?: rapid-response extracorporeal membrane oxygenation
and intra-arrest percutaneous coronary intervention. Circulation. 2012;126:160513.
26. Kagawa E, Inoue I, Kawagoe T, Ishihara M, Shimatani Y, et al. Assessment of outcomes and
differences between in- and out-of-hospital cardiac arrest patients treated with cardiopulmo-
nary resuscitation using extracorporeal life support. Resuscitation. 2010;81:96873.
27. Le Guen M, Nicolas-Robin A, Carreira S, Raux M, Leprince P, et al. Extracorporeal life sup-
port following out-of-hospital refractory cardiac arrest. Crit Care. 2011;15:R29.
28. Reynolds JC, Frisch A, Rittenberger JC, Callaway CW. Duration of resuscitation efforts and
functional outcome after out-of-hospital cardiac arrest: when should we change to novel thera-
pies? Circulation. 2013;128:248894.
29. Pages ON, Aubert S, Combes A, Luyt CE, Pavie A, et al. Paracorporeal pulsatile biventricular
assist device versus extracorporal membrane oxygenation-extracorporal life support in adult
fulminant myocarditis. J Thorac Cardiovasc Surg. 2009;137:1947.
30. Asaumi Y, Yasuda S, Morii I, Kakuchi H, Otsuka Y, et al. Favourable clinical outcome in
patients with cardiogenic shock due to fulminant myocarditis supported by percutaneous
extracorporeal membrane oxygenation. Eur Heart J. 2005;26:218592.
31. George RS, Yacoub MH, Bowles CT, Hipkin M, Rogers P, et al. Quality of life after removal
of left ventricular assist device for myocardial recovery. J Heart Lung Transplant. 2008;27:
16572.
32. Heyland DK, Groll D, Caeser M. Survivors of acute respiratory distress syndrome: relation-
ship between pulmonary dysfunction and long-term health-related quality of life. Crit Care
Med. 2005;33:154956.
33. Baud FJ, Megarbane B, Deye N, Leprince P. Clinical review: aggressive management and
extracorporeal support for drug-induced cardiotoxicity. Crit Care. 2007;11:207.
34. Masson R, Colas V, Parienti JJ, Lehoux P, Massetti M, et al. A comparison of survival with and
without extracorporeal life support treatment for severe poisoning due to drug intoxication.
Resuscitation. 2012;83:14137.
35. Elsharkawy HA, Li L, Esa WA, Sessler DI, Bashour CA. Outcome in patients who require
venoarterial extracorporeal membrane oxygenation support after cardiac surgery. J Cardiothorac
Vasc Anesth. 2010;24:94651.
36. Kumar TK, Zurakowski D, Dalton H, Talwar S, Allard-Picou A, et al. Extracorporeal mem-
brane oxygenation in postcardiotomy patients: factors inuencing outcome. J Thorac
Cardiovasc Surg. 2010;140:3306. e332.
37. Rastan AJ, Dege A, Mohr M, Doll N, Falk V, et al. Early and late outcomes of 517 consecutive
adult patients treated with extracorporeal membrane oxygenation for refractory postcardiot-
omy cardiogenic shock. J Thorac Cardiovasc Surg. 2010;139:30211. 311 e301.
38. Slottosch I, Liakopoulos O, Kuhn E, Deppe AC, Scherner M, et al. Outcomes after peripheral
extracorporeal membrane oxygenation therapy for postcardiotomy cardiogenic shock: a
single-center experience. J Surg Res. 2013;181:e4755.
39. Wang J, Han J, Jia Y, Zeng W, Shi J, et al. Early and intermediate results of rescue extracorpo-
real membrane oxygenation in adult cardiogenic shock. Ann Thorac Surg. 2009;88:
1897903.
40. Wang JG, Han J, Jia YX, Zeng W, Hou XT, et al. Outcome of veno-arterial extracorporeal
membrane oxygenation for patients undergoing valvular surgery. PLoS One. 2013;8:e63924.
41. Arpesella G, Loforte A, Mikus E, Mikus PM. Extracorporeal membrane oxygenation for pri-
mary allograft failure. Transplant Proc. 2008;40:35967.
42. Chen YS, Ko WJ, Chi NH, Wu IH, Huang SC, et al. Risk factor screening scale to optimize
treatment for potential heart transplant candidates under extracorporeal membrane oxygen-
ation. Am J Transplant. 2004;4:181825.
43. Chou NK, Luo JM, Chi NH, Wu IH, Huang SC, et al. Extracorporeal membrane oxygenation
and Thoratec pneumatic ventricular assist devices as double bridge to heart transplantation.
Transplant Proc. 2012;44:87880.
44. Chung JC, Tsai PR, Chou NK, Chi NH, Wang SS, et al. Extracorporeal membrane oxygenation
bridge to adult heart transplantation. Clin Transplant. 2010;24:37580.
45. Pagani FD, Lynch W, Swaniker F, Dyke DB, Bartlett R, et al. Extracorporeal life support to left
ventricular assist device bridge to heart transplant: a strategy to optimize survival and resource
utilization. Circulation. 1999;100(19 Suppl):II20610.
46. McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, et al. ESC guidelines for the
diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the
Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of
Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the
ESC. Eur Heart J. 2012;33:1787847.
47. Takayama H, Truby L, Koekort M, Uriel N, Colombo P, et al. Clinical outcome of mechanical
circulatory support for refractory cardiogenic shock in the current era. J Heart Lung Transplant.
2013;32:10611.
48. DAlessandro C, Aubert S, Golmard JL, Praschker BL, Luyt CE, et al. Extra-corporeal mem-
brane oxygenation temporary support for early graft failure after cardiac transplantation. Eur J
Cardiothorac Surg. 2010;37:3439.
49. DAlessandro C, Golmard JL, Barreda E, Laali M, Makris R, et al. Predictive risk factors for
primary graft failure requiring temporary extra-corporeal membrane oxygenation support after
cardiac transplantation in adults. Eur J Cardiothorac Surg. 2011;40:9629.
50. Marasco SF, Vale M, Pellegrino V, Preovolos A, Leet A, et al. Extracorporeal membrane oxy-
genation in primary graft failure after heart transplantation. Ann Thorac Surg. 2010;90:
15416.
51. Lebreton G, Nicolescu M, Leger P, Leprince P. Implantation of left ventricular support under
extracorporeal membrane oxygenation. Eur J Cardiothorac Surg. 2011;40:e1657.
52. Scherer M, Moritz A, Martens S. The use of extracorporeal membrane oxygenation in patients
with therapy refractory cardiogenic shock as a bridge to implantable left ventricular assist
device and perioperative right heart support. J Artif Organs. 2009;12:1605.
53. Scherer M, Sirat AS, Moritz A, Martens S. Extracorporeal membrane oxygenation as periop-
erative right ventricular support in patients with biventricular failure undergoing left ventricu-
lar assist device implantation. Eur J Cardiothorac Surg. 2011;39:93944. discussion 944.
54. Haneya A, Philipp A, Diez C, Metterlein T, Puehler T, et al. Successful use of temporary right
ventricular support to avoid implantation of biventricular long-term assist device: a transcuta-
neous approach. ASAIO J. 2011;57:2747.
55. Haneya A, Philipp A, Puehler T, Rupprecht L, Kobuch R, et al. Temporary percutaneous right
ventricular support using a centrifugal pump in patients with postoperative acute refractory
right ventricular failure after left ventricular assist device implantation. Eur J Cardiothorac
Surg. 2012;41:21923.
56. Takayama H, Naka Y, Kodali SK, Vincent JA, Addonizio LJ, et al. A novel approach to percu-
taneous right-ventricular mechanical support. Eur J Cardiothorac Surg. 2012;41:4236.
57. Walpoth BH, Walpoth-Aslan BN, Mattle HP, Radanov BP, Schroth G, et al. Outcome of survi-
vors of accidental deep hypothermia and circulatory arrest treated with extracorporeal blood
warming. N Engl J Med. 1997;337:15005.
58. Omar HR, Miller J, Mangar D, Camporesi EM. Experience with extracorporeal membrane
oxygenation in massive and submassive pulmonary embolism in a tertiary care center. Am J
Emerg Med. 2013;31:16167.
59. Deehring R, Kiss AB, Garrett A, Hillier AG. Extracorporeal membrane oxygenation as a
bridge to surgical embolectomy in acute fulminant pulmonary embolism. Am J Emerg Med.
2006;24:87980.
60. Hsieh PC, Wang SS, Ko WJ, Han YY, Chu SH. Successful resuscitation of acute massive pul-
monary embolism with extracorporeal membrane oxygenation and open embolectomy. Ann
Thorac Surg. 2001;72:2667.
61. Maggio P, Hemmila M, Haft J, Bartlett R. Extracorporeal life support for massive pulmonary
embolism. J Trauma. 2007;62:5706.
62. Davies MJ, Arsiwala SS, Moore HM, Kerr S, Sosnowski AW, et al. Extracorporeal membrane
oxygenation for the treatment of massive pulmonary embolism. Ann Thorac Surg. 1995;
60:18013.
63. Bartlett RH. Extracorporeal support for septic shock. Pediatr Crit Care Med. 2007;8:4989.
64. Maclaren G, Butt W, Best D, Donath S, Taylor A. Extracorporeal membrane oxygenation for
refractory septic shock in children: one institutions experience. Pediatr Crit Care Med.
2007;8:44751.
65. MacLaren G, Butt W, Best D, Donath S. Central extracorporeal membrane oxygenation for
refractory pediatric septic shock. Pediatr Crit Care Med. 2011;12:1336.
66. Huang CT, Tsai YJ, Tsai PR, Ko WJ. Extracorporeal membrane oxygenation resuscitation in
adult patients with refractory septic shock. J Thorac Cardiovasc Surg. 2013;146(5):10416.
67. Brechot N, Luyt CE, Schmidt M, Leprince P, Trouillet JL, et al. Venoarterial extracorporeal
membrane oxygenation support for refractory cardiovascular dysfunction during severe bacte-
rial septic shock. Crit Care Med. 2013;41:161626.
68. Hanna BD. Left atrial decompression: is there a standard during extracorporeal support of the
failing heart? Crit Care Med. 2006;34:26889.
69. Sauren LD, Reesink KD, Selder JL, Beghi C, van der Veen FH, et al. The acute effect of intra-
aortic balloon counterpulsation during extracorporeal life support: an experimental study. Artif
Organs. 2007;31:318.
70. Petroni T, Harrois A, Amour J, Lebreton G, Brechot N, et al. Intra-aortic balloon pump effects
on macrocirculation and microcirculation in cardiogenic shock patients supported by venoar-
terial extracorporeal membrane oxygenation. Crit Care Med. 2014;42(9):207582.
71. Schmidt M, Brechot N, Hariri S, Guiguet M, Luyt CE, et al. Nosocomial infections in adult
cardiogenic shock patients supported by venoarterial extracorporeal membrane oxygenation.
Clin Infect Dis. 2012;55:163341.
72. Heilmann C, Geisen U, Beyersdorf F, Nakamura L, Benk C, et al. Acquired von Willebrand
syndrome in patients with extracorporeal life support (ECLS). Intensive Care Med.
2012;38:628.
73. Cheng R, Hachamovitch R, Kittleson M, Patel J, Arabia F, et al. Complications of extracorpo-
real membrane oxygenation for treatment of cardiogenic shock and cardiac arrest: a meta-
analysis of 1,866 adult patients. Ann Thorac Surg. 2014;97:6106.
74. Paden ML, Conrad SA, Rycus PT, Thiagarajan RR. Extracorporeal Life Support Organization
Registry Report 2012. ASAIO J. 2013;59:20210.
Chapter 4
ECCO2R inObstructive Diseases:
Evidence, Indications, andExclusions
LorenzoDel Sorbo andV.MarcoRanieri
Introduction
In the last decade, there has been an impressive technological and clinical evolution
of extracorporeal life support strategies [14], as demonstrated by the progressively
increasing number of scientific publications on this topic (Fig.4.1). To date, a vari-
ety of modalities of extracorporeal life support have been developed, characterized
by different configurations, different 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) (Fig.4.1) [6, 7], due
to the relative ease and efficiency in blood CO2 clearance granted by extracorporeal
gas exchangers as compared to oxygen delivery. CO2 is transported in blood as
bicarbonate ion or bound to hemoglobin, characterized by a steep linear dissocia-
tion kinetic, and high solubility, features which facilitate rapid diffusion across the
extracorporeal membrane. Therefore, ventilation (sweep gas flow), rather than per-
fusion (extracorporeal blood flow rate), of the artificial lung and its total surface
area are the main determinants of CO2 clearance from the blood. Since whole body
CO2 production averages 200250mL/min, blood flow as low as 0.51L/min
L. Del Sorbo, MD
Inter-departmental Division of Critical Care Medicine, University Health Network,
University of Toronto, Toronto, ON, Canada
V.M. Ranieri, MD (*)
Dipartimento di Anestesiologia e Medicina degli Stati Critici, Ospedale S.Giovanni
Battista-Molinette, Universit di Torino, Corso Dogliotti 14, 10126 Torino, Italy
e-mail: marco.ranieri@unito.it

Respiratory Medicine 16, DOI10.1007/978-1-4939-3005-0_4
88 L. Del Sorbo and V.M. Ranieri
800 20
18
n. publications on ECCO2R
n. publications on ECMO 700
16
600
14
500
12
400 10
300 8
6
200
4
100
2
0 0
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
19
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Year
Fig. 4.1 Number of articles per year published in Pubmed (accessed on January 2014) on extracor-
poreal membrane oxygenation (ECMO) and on extracorporeal carbon dioxide removal (ECCO2R)
through the artificial lung theoretically removes the entire metabolic production,
without however providing significant oxygenation [811].
These features have been translated into clinical strategies that are remarkably
less invasive as compared to full extracorporeal membrane oxygenation, easier to
manage, and as efficient in CO2 clearance [6]. Several ECCO2R devices have been
approved in Europe and are available for clinical use (PALP from Maquet: http://
www.maquet.com; iLA activve from Novalung, http://www.novalung.com;
HemolungRAS from Alung, http://www.alung.com; Decap from Hemodec,
http://www.hemodec.com; and Abylcap from Bellco: www.bellco.net).
In most of these systems blood is drawn from a central vein by a draining can-
nula, driven with a centrifugal or a roller pump through the extracorporeal lung
where CO2 diffuses into the sweep gas, and returned into the venous circulation
(Fig.4.2a). Central venous access with a double-lumen cannula is required by most
of the ECCO2R systems. It is nonetheless possible to set up an ECCO2R circuit so
that two central veins are accessed, one for drainage and the other for reinfusion
(Fig.4.2b). Reducing the dimensions and number of cannulas as well as the dose of
anticoagulation, while maintaining or increasing the efficiency of CO2 removal,
have been key factors to reduce the invasiveness of these systems [6].
An alternative configuration for ECCO2R involves inserting an arteriovenous
(AV) bypass, usually by connecting the femoral artery to the contralateral femoral
vein, equipped with an artificial gas exchanger (Fig.4.2c). In this system, blood
flows through the artificial lung 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
4ECCO2R inObstructive Diseases: Evidence, Indications, andExclusions 89
Fig. 4.2 Possible configurations of extracorporeal carbon dioxide removal (ECCO2R) circuits.
(a)Minimally invasive venovenous ECCO2R circuit with a single venous vascular access through
a double-lumen cannula that can be typically inserted in the internal jugular vein or in the femoral
vein. (b) Venovenous 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 membrane
providing extracorporeal gas exchange in the circuit connecting the femoral artery with the
contralateral femoral vein
resulting from very low tidal volumes during super-protective mechanical ventila-
tion. Zimmermann and colleagues demonstrated that pumpless AV ECCO2R cleared
sufficient CO2 from 51 patients with ARDS as to enable protective mechanical ven-
tilation with tidal volumes lower than 6cc/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 control trial
in patients with ARDS, low tidal volume ventilation (6mL/kg) was compared with
super-protective mechanical ventilation (3mL/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 inflammatory 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
6.30.2 to 4.20.2mL/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 evolu-
tion toward progressively less invasive devices has prompted investigations into a
broader spectrum of disease [1, 7]. In particular, the idea of applying ECCO2R in a
variety of 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 patho-
physiological feature represents the optimal setting for ECCO2R, potentially pro-
viding an alternative, not just an adjunct, to mechanical ventilation as a way to
support respiration [17] until the lung recovers [2022]. This chapter focuses on the
main evidence for the innovative use of ECCO2R in patients with obstructive lung
diseases, along with clinical indications and exclusion criteria.
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
efficacy. However, numerous clinical reports demonstrate the efficiency of this
technologically advanced strategy, which is based on a strong pathophysiological
rationale [17, 23]. The main feature of exacerbations of obstructive diseases is
acute worsening of expiratory flow-limitation caused by increased resistance in
small airways [19, 24]. Increased resistive load is associated with a longer expira-
tory time constant (the product of airway resistance and lung compliance), which
defines the time for exponential fall in lung volume during passive exhalation [25].
The direct consequence of a longer time constant is the development of dynamic
alveolar hyperinflation, 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 (intrinsic-PEEPFig.4.3) [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.4.3) [24, 2628]. Since alveolar ventilation is
inversely proportional to PaCO2 (according to the equation PaCO 2 = V CO / V ,
2 A

where VCO 2 is CO2 production and VA is alveolar ventilation), acute exacerba-
tions of obstructive lung disease 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 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 hemodynamic decom-
pensation [19, 3234].
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
a COPD exacerbation
Worsening of
1. Airflow resistance
and
2. Expiratory flow-limitation
Dynamic alveolar hyperinflation

PEEPi
CDYN VD/VT threshold loading of

inspiratory muscles
pulmonary artery
Neuro-mechanical pressure
uncoupling RV preload
with dyspnea
LV afterload
Rapid shallow
Muscle fatigue breathing
Ventilatory pump failure 1. low VT
2. high RR
Hypercapnia and
respiratory acidosis
ventilatory drive and

VA demand
Fig. 4.3(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
b COPD exacerbation
Worsening of
1. Airflow resistance
and
2. Expiratory flow-limitation
Dynamic alveolar hyperinflation

PEEPi
Rapid shallow
Muscle fatigue breathing
Ventilatory pump failure 1. low VT
2. high RR
Hypercapnia and
respiratory acidosis
ECCO2R
VA demand: lower RR and expiratory time
Resolution of dynamic alveolar hyperinflation

and PEEPi
VD/VT threshold loading of pulmonary artery

inspiratory muscles pressure
RV preload
CDYN Neuro-mechanical
uncoupling LV afterload
Fig. 4.3(continued)
hyperinflated lungs, ECCO2R reduces the volume of 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 interrupt the vicious cycle of dynamic hyperinflation, thereby re-
establishing a more favorable balance between the work of breathing and the respi-
ratory load in patients with acute obstructive lung disease exacerbations [17, 23].
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 cases
describe ECCO2R in patients with near-fatal asthma.
xacerbation ofChronic 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 and projected to become the third-
leading cause by 2020 [35, 36]. Its natural history is characterized by progressive
deterioration of lung function, punctuated by a variable number of acute exacerba-
tions [37, 38]. COPD exacerbations are associated with considerable morbidity and
mortality, which ranges between 8% and 26% for patients admitted to the intensive
care unit [3941].
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 [4244]. Firstly, NIV has become the standard of
care for exacerbations refractory to medical treatment [31, 4345]. 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 breath-
ing and helping to reverse muscle fatigue. Together these effects are instrumental in
decreasing the need for endotracheal intubation and improving survival [4648].
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
[49]. Moreover, MV is associated with a number of undesired side effects [32, 34,
50] which may produce morbidity or mortality [51]. 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,
thereby re-establishing a compensated balance between metabolic production and

the impaired respiratory system. Several reports verify this hypothesis, showing the
successful use of ECCO2R in chronic obstructive respiratory diseases [16, 17, 52].
One of the first reports on the application of ECCO2R to support respiratory func-
tion of a COPD patient was published in 1990 by Pesenti etal. [53]. A 19-year-old
woman with bullous emphysema, recurrent pneumothoraxes, bilateral 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 prolonged MV and the high minute vol-
umes required for CO2 clearance. This vicious cycle was broken by low blood flow
(0.40.6L/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.
Abrams and colleagues [54] reported five older patients (age 738.7years) with
acute COPD exacerbations who failed NIV, requiring MV.After an average of
16.55.9h of MV, ECCO2R was initiated.
In all five patients ECCO2R facilitated extubation within 24h of treatment
(median duration of MV post ECCO2R=4h, range 1.521.5h). The ECCO2R flow
required to accomplish the goal of extubation ranged between 1 and 1.7L/min,
while the sweep gas flow varied from 1 to 7L/min, resulting in a pH of 7.347.48.
Once extubated, patients were intensively rehabilitated while on ECCO2R, with a
mean time to ambulation of 29.412.6h after ECCO2R.
It is worth highlighting that, in these case reports, ECCO2R was applied as an
adjunct to mechanical ventilation, subjecting patients to the potential complica-
tions of two invasive strategies of respiratory support. Excitingly, a new approach
proposes that ECCO2R may be effective in preventing endotracheal intubation
altogether 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, 55]. Avoiding these adverse effects could improve prognosis, especially con-
sidering the severity of underlying lung disease and frequent presence of comor-
bidities [51, 56].
Along this line, Brederlau and coworkers [57] reported their experience with
three patients affected by severe comorbidities and failing NIV for severe COPD
exacerbation. Consent was given to apply a pumpless AV ECCO2R device with the
goal of avoiding endotracheal intubation. Shortly after beginning ECCO2R, PaCO2
fell significantly (from 91, 109, and 142mmHg to 52, 59, and 83mmHg, respec-
tively) in all three patients, while pH rose (from 7.2, 7.19, and 7.06 to 7.41, 7.43,
and 7.34mmHg, respectively). Simultaneously, respiratory rate dropped from 38,
45, and 37 breaths/min to 15, 25, and 18 breaths/min, respectively. The ECCO2R
flow ranged between 1.1 and 1.6L/min, with the sweep gas flow varying from 3 to
8L/min. Invasive MV was prevented in two of these three severely ill patients.
PaCO2 pH Respiratory rate

140 7,6 40
* 35
120 7,5
Respirations / min
30
100 7,4 *
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 9-12 17-20 NIV 0-4 9-12 17-20 NIV 0-4 9-12 17-20
Time intervals
Fig. 4.4 Changes over time of partial pressure of arterial carbon dioxide (PaCO2), pH, and respira-
tory 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 ref. [58]
The same pumpless AV ECCO2R device was studied in a larger case series of 21
patients with acute hypercapnic ventilatory failure not responding to NIV [58]. By
reducing PaCO2 and increasing pH, ECCO2R lowered respiratory rate and prevented
endotracheal intubation in 90% of the patients (Fig.4.4). The ECCO2R-treated
patients were compared to a matched group of controls undergoing MV after a
failed attempt on NIV.Significantly fewer patients in the ECCO2R group required
tracheostomy, although no difference was found in ICU length of stay or mortality.
ECCO2R patients suffered nine bleeding complications, and one case each of femo-
ral artery pseudoaneurysm and heparin-induced thrombocytopenia.
Similar promising results have been reported by other groups using less invasive
ECCO2R devices characterized by venovenous configurations, using a double-lumen
cannula similar in size to a dialysis catheter. In one particularly illustrative case a
72-year-old patient with COPD failing NIV was treated with VV-ECCO2R, while
additional physiological data were collected [23]. Extracorporeal CO2 clearance was
calculated by measuring the CO2 concentration in the sweep gas, native lung CO2
clearance was estimated assuming a respiratory quotient equal to one, and intra-tidal
esophageal pressure swings were recorded as a surrogate for the work of breathing.
The investigators showed the progressive transition of gas exchange from the natu-
ral to the artificial lung. Moreover, they documented the ECCO2R-induced rapid
reduction of esophageal pressure variation and respiratory rate, proving a beneficial
effect on the work of breathing. Avoiding endotracheal intubation and MV, the
patient was able to speak, drink, and eat, despite the severity of the exacerbation.
In another trial, minimally invasive VV-ECCO2R was used in COPD patients

with hypercapnic respiratory failure [59]. Of 20 patients recruited into the trial,
seven were at risk of MV despite NIV; two were difficult to wean from NIV; and 11
had failed liberation from MV.ECCO2R provided a CO2 clearance of 82.515.6mL/
min using an average extracorporeal blood flow of 430.573.7mL/min, allowing
PaCO2 to fall from 78.916.8mmHg to 65.911.5mmHg. None of the patients
failing NIV required endotracheal intubation, and both patients with difficult wean-
ing from NIV were weaned. However, only three of the 11 MV patients were liber-
ated successfully. Moreover, significant complications arose in a number of patients:
bleeding requiring blood transfusion was reported in three patients, deep vein
thrombosis was diagnosed in one patient after removal of the ECCO2R catheter, one
patient experienced pneumothorax due to catheter insertion, and one died from hem-
orrhage when the iliac vein was perforated during ECCO2R catheter placement.
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 (Table4.1) 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. 12 [54]. Smaller devices become potentially more portable and facilitate
patients walking, instead of being bed-ridden, despite the severity of disease [60].
One can even imagine further evolution along these lines, as occurred for ventricu-
lar assist devices [61], so that small and implantable ECCO2R circuits could provide
destination treatment for patients discharged to home.
Acute Severe Asthma
Acute severe asthma is a major health burden, as its prevalence is growing worldwide
with increasing admissions to hospital and intensive care unit [19, 62]. The in-hospital
mortality rate due to acute exacerbations 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 ICU for intubation and MV [19, 63].
The treatment of acute severe asthma consists of measures to reverse airflow
obstruction. Ventilatory assistance is provided to patients refractory to medical treat-
ment and developing hypoxemia or hypoventilation [19, 62]. However, while hyper-
capnia is observed in about 10% of asthma patients in the acute setting, only a small
percentage of these require invasive MV [64]. 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 [65, 66].
Table 4.1 Indications, exclusions, and goals for ECCO2R in COPD

Indications Invasive mechanical
ventilation for COPD
exacerbation with two or
more failed weaning attempts
Failing treatment with NIV NIV for at least 12h with signs of respiratory
for COPD exacerbation distress (respiratory rate 30breaths/min and
use of accessory muscles or paradoxical
abdominal movements) and
PaCO2>55 mmHg and pH<7.25
or pH<7.30 and PaCO2>55 mmHg, with
PaCO2 decrease<20% from baseline
Exclusions No absolute exclusion Anatomical abnormalities or vascular diseases
criteria, but thorough preventing the correct insertion of the
evaluation of risks in each ECCO2R cannula
patient Body mass index>31.1kg/m2 for men
or>32.2 kg/m2 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 (e.g. hemorrhagic disease,
thrombocytopenia, heparin-induced
thrombocytopenia)
History of intracranial bleeding
Inability to receive blood products
History of complications from extracorporeal
support
Relative exclusion criteria Terminal disease
Severe malnutrition or cachexia
Severely debilitated state
Advanced malignancy
Immunocompromised condition
Goals for Physiological goals Reduction of respiratory rate with consequent
ECCO2R increased duration of expiratory time, facilitating
the 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
The accepted indication for MV is progressive hypercapnia with impending

h emodynamic 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 hyperinfla-
tion, thereby causing worsening hypercapnia, barotrauma, pneumothorax, and further
hemodynamic deterioration, with attendant risk of morbidity or mortality [19, 33, 62].
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 [6769]. 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 fail-
ure is profound, ECCO2R may not be sufficient and full extracorporeal membrane
oxygenation (ECMO) may provide a more adequate means of support [7072].
Extracorporeal membrane oxygenation was first applied in near-fatal asthma in
1981, followed by a progressively increasing number of reports [73]. In the interna-
tional Extracorporeal Life Support Organization (ELSO) registry [72], asthma was
the basis for ECMO in 24 out of 1257 adult patients between 1986 and 2006. Before
ECMO, the average pH was 7.170.16, PaCO2 119.758.1mmHg, and PaO2/FiO2
ratio 244180, despite invasive MV.Extracorporeal support was provided for
111.971.2h and 83.3% of patients survived. Complications were described in 19
of the 24 asthmatic subjects (79.2%). Between 1986 and 2007, 64 pediatric asth-
matics were treated with ECMO for life-threatening asthma, and 60 of those sur-
vived (94%) [71]. Prior to initiation of ECMO, median pH was 6.96 (range
6.787.28), median PaCO2 was 123mmHg (range 70237mmHg), and PaO2 was
126mmHg (range 59636mmHg), despite MV.The median time of ECMO support
was 94h, 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 hypoxemia, was the dominant gas exchange problem, suggesting that less inva-
sive 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 [6769]. More recently, a minimally
invasive venovenous ECCO2R device was used as an adjunct to MV in two adults
with asthma refractory to conventional therapies (Fig.4.5) [74]. The first had a pH
of 6.94, PaCO2 of 147mmHg, and PaO2 of 416mmHg despite optimal MV.ECCO2R
was started with a blood flow at 1.3L/min and sweep gas flow of 2.3L/min, mark-
edly improving the respiratory acidosis to a pH of 7.24, PaCO2 of 56mmHg, and
PaO2 of 310mmHg. After 24h of ECCO2R support, the patient was liberated from
MV and progressively weaned from the extracorporeal device, before being success-
fully discharged. The second patient not only had severe respiratory acidosis despite
MV, but required vasopressors for hemodynamic support. ECCO2R contributed to
normalization of pH and PaCO2, while respiratory rate was lowered from 8 to 4
breaths/min and tidal volume from 6 to 4mL/kg. In 1day, intrinsic PEEP resolved,
shock was reversed, and neuromuscular blocking agents were discontinued.
Theimpact of ECCO2R on respiratory system mechanics in these two patients is
shown in Fig.4.5.
a ECCO2R Start Extubation

70
60
50
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Hours
b ECCO2R Start ECCO2R End

60
50
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Days
Peak Airway Pressure (cmH2O) Dynamic Compliance (mL/cmH2O) Auto-PEEP(cmH2O)
Fig. 4.5 Trends in ventilator parameters demonstrating the timeframe of improvement in respira-
tory mechanics of two patients with acute severe asthma refractory to conventional therapy and
invasive mechanical ventilation and therefore treated with extracorporeal carbon dioxide removal
(ECCO2R), as described by Brenner etal. [74]. Reproduced with permission from ref. [74]
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
treatment particularly complex. Since ECCO2R has become a minimally invasive
treatment, well-tolerated, and easy to manage, requiring similar resources as renal
dialysis, every patient with adequate indications should be considered 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 relevant, 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 potential vascular trauma during the
insertion of the circuit cannula, and bleeding due to the need for therapeutic antico-
agulation to avoid thrombosis of the circuit. Therefore, anatomical abnormalities or
vascular diseases preventing the correct insertion of the ECCO2R cannula, pre-
existing risk of bleeding, contraindications to the administration of anticoagulants,
or actual hemorrhagic diseases, represent major conditions that may preclude the
application of ECCO2R (Table4.1).
Moreover, additional exclusion criteria are applied in clinical trials investigating
the efficacy of ECCO2R in improving the recovery from respiratory failure [54, 58,
59]. 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 (Table4.1). An exception would
be a patient with end-stage obstructive lung disease in whom ECCO2R might serve
as a bridge to lung transplant [75].
Funding Ministero della Salute, Programma Ricerca Finalizzata # VMR23a/2004-2006, Regione

Piemonte, Programma Ricerca Finalizzata # CIPE LM002/2005-2007, Ministero dellUniversit,
Programmi di Ricerca di Interesse Nazionale # VMRLM98, 2007-2010.
References
1. Abrams D, Brodie D.Emerging indications for extracorporeal membrane oxygenation in

adults with respiratory failure. Ann Am Thorac Soc. 2013;10:3717.
2. Del Sorbo L, Cypel M, Fan E.Extracorporeal life support for adults with severe acute respira-
tory failure. Lancet Respir Med. 2014;2:15464.
3. Brodie D, Bacchetta M.Extracorporeal membrane oxygenation for ARDS in adults. N Engl J
Med. 2011;365:190514.
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;38:21020.
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;24:16472.
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;18:938.
8. Bartlett RH.Physiology of extracorporeal life support. In: Annich GM, Lynch WR, MacLaren
G, Wilson JM, Barlett RH, Zwischenberger JB, editors. ECMO, extracorporeal cardiopulmo-
nary support in critical care. 4th ed. Ann Arbor: Extracorporeal Life Support Organization;
2012. p.1131.
9. Gattinoni L, Pesenti A, Mascheroni D, etal. Low-frequency positive-pressure ventilation with
extracorporeal CO2 removal in severe acute respiratory failure. JAMA. 1986;256:8816.
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;2:1835.
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;32:50810.
13. Zimmermann M, Bein T, Arlt M, etal. 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, etal. Lower tidal volume strategy (approximately
3ml/kg) combined with extracorporeal CO(2) removal versus conventional protective venti-
lation (6ml/kg) in severe ARDS: the prospective randomized Xtravent-study. Intensive Care
Med. 2013;39(5):84756.
15. Terragni PP, Del Sorbo L, Mascia L, etal. Tidal volume lower than 6ml/kg enhances lung
protection: role of extracorporeal carbon dioxide removal. Anesthesiology. 2009;111:82635.
16. Braune SA, Kluge S.Extracorporeal lung support in patients with chronic obstructive pulmo-
nary disease. Minerva Anestesiol. 2013;79:93443.
17. Lund LW, Federspiel WJ.Removing extra CO2in COPD patients. Curr Respir Care Rep.
2013;2:1318.
18. Barbera JA, Roca J, Ferrer A, etal. Mechanisms of worsening gas exchange during acute
exacerbations of chronic obstructive pulmonary disease. Eur Respir J.1997;10:128591.
19. McFadden Jr ER.Acute severe asthma. Am J Respir Crit Care Med. 2003;168:74059.
20. Del Sorbo L, Ranieri VM.We do not need mechanical ventilation any more. Crit Care Med.
2010;38:S5558.
21. Malagon I, Greenhalgh D.Extracorporeal membrane oxygenation as an alternative to ventila-
tion. Curr Opin Anaesthesiol. 2013;26:4752.
22. Hoeper MM, Wiesner O, Hadem J, etal. Extracorporeal membrane oxygenation instead of
invasive mechanical ventilation in patients with acute respiratory distress syndrome. Intensive
Care Med. 2014;2014:804917.
23. Crotti S, Lissoni A, Tubiolo D, etal. Artificial lung as an alternative to mechanical ventilation
in COPD exacerbation. Eur Respir J.2012;39:2125.
24. Hogg JC.Pathophysiology of airflow limitation in chronic obstructive pulmonary disease.
Lancet. 2004;364:70921.
25. Laghi F, Goyal A.Auto-PEEP in respiratory failure. Minerva Anestesiol. 2012;78:20121.
26. Loring SH, Garcia-Jacques M, Malhotra A.Pulmonary characteristics in COPD and mecha-
nisms of increased work of breathing. J Appl Physiol. 2009;107:30914.
27. Stevenson NJ, Walker PP, Costello RW, Calverley PM.Lung mechanics and dyspnea during
exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med.
2005;172:15106.
28. ODonnell DE, Parker CM.COPD exacerbations. 3: pathophysiology. Thorax. 2006;61:35461.
29. Roberts CM, Stone RA, Buckingham RJ, Pursey NA, Lowe D.Acidosis, non-invasive ventila-
tion and mortality in hospitalised COPD exacerbations. Thorax. 2011;66:438.
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;71:14429.
31. Nava S, Hill N.Non-invasive ventilation in acute respiratory failure. Lancet. 2009;374:2509.
32. Reddy RM, Guntupalli KK.Review of ventilatory techniques to optimize mechanical ventila-
tion in acute exacerbation of chronic obstructive pulmonary disease. Int J Chron Obstruct
Pulmon Dis. 2007;2:44152.
33. Kao CC, Jain S, Guntupalli KK, Bandi V.Mechanical ventilation for asthma: a 10-year experi-
ence. J Asthma. 2008;45:5526.
34. Ward NS, Dushay KM.Clinical concise review: mechanical ventilation of patients with
chronic obstructive pulmonary disease. Crit Care Med. 2008;36:16149.
35. Calverley PM, Walker P.Chronic obstructive pulmonary disease. Lancet. 2003;362:105361.
36. Chapman KR, Mannino DM, Soriano JB, etal. Epidemiology and costs of chronic obstructive
pulmonary disease. Eur Respir J.2006;27:188207.
37. Burge S, Wedzicha JA.COPD exacerbations: definitions and classifications. Eur Respir J
Suppl. 2003;41:46s53.
38. Hurst JR, Vestbo J, Anzueto A, etal. Susceptibility to exacerbation in chronic obstructive
pulmonary disease. N Engl J Med. 2010;363:112838.
39. Patil SP, Krishnan JA, Lechtzin N, Diette GB.In-hospital mortality following acute exacerba-
tions of chronic obstructive pulmonary disease. Arch Intern Med. 2003;163:11806.
40. Donaldson GC, Wedzicha JA.COPD exacerbations. 1: epidemiology. Thorax. 2006;61:

1648.
41. 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;60:92531.
42. Quon BS, Gan WQ, Sin DD.Contemporary management of acute exacerbations of COPD:
asystematic review and metaanalysis. Chest. 2008;133:75666.
43. 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
1:1232.
44. 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;163:125676.
45. International Consensus Conferences in Intensive Care Medicine: noninvasive positive pres-
sure ventilation in acute respiratory failure. Am J Respir Crit Care Med. 2001;163:28391.
46. Brochard L, Mancebo J, Wysocki M, etal. Noninvasive ventilation for acute exacerbations of
chronic obstructive pulmonary disease. N Engl J Med. 1995;333:81722.
47. 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;CD004104.
48. Scala R, Naldi M, Archinucci I, Coniglio G, Nava S.Noninvasive positive pressure ventilation
in patients with acute exacerbations of COPD and varying levels of consciousness. Chest.
2005;128:165766.
49. Chandra D, Stamm JA, Taylor B, etal. Outcomes of noninvasive ventilation for acute exacer-
bations of chronic obstructive pulmonary disease in the United States, 1998-2008. Am J Respir
Crit Care Med. 2012;185:1529.
50. Rello J, Ollendorf DA, Oster G, etal. Epidemiology and outcomes of ventilator-associated
pneumonia in a large US database. Chest. 2002;122:211521.
51. 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;128:51824.
52. Cardenas Jr VJ, Lynch JE, Ates R, Miller L, Zwischenberger JB.Venovenous carbon dioxide
removal in chronic obstructive pulmonary disease: experience in one patient. ASAIO
J.2009;55:4202.
53. 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;72:5713.
54. Abrams DC, Brenner K, Burkart KM, etal. Pilot study of extracorporeal carbon dioxide
removal to facilitate extubation and ambulation in exacerbations of chronic obstructive pulmo-
nary disease. Ann Am Thorac Soc. 2013;10:30714.
55. Bonin F, Sommerwerck U, Lund LW, Teschler H.Avoidance of intubation during acute exacer-
bation of chronic obstructive pulmonary disease for a lung transplant candidate using extracor-
poreal carbon dioxide removal with the Hemolung. J Thorac Cardiovasc Surg. 2013;145:e434.
56. 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:
398405.
57. Brederlau J, Wurmb T, Wilczek S, etal. Extracorporeal lung assist might avoid invasive ven-
tilation in exacerbation of COPD.Eur Respir J.2012;40:7835.
58. Kluge S, Braune SA, Engel M, etal. Avoiding invasive mechanical ventilation by extracorporeal
carbon dioxide removal in patients failing noninvasive ventilation. Intensive Care Med. 2012;38:
16329.
59. Burki NK, Mani RK, Herth FJ, etal. A novel extracorporeal CO(2) removal system: results of
a pilot study of hypercapnic respiratory failure in patients with COPD.Chest. 2013;143:
67886.
60. Garcia JP, Kon ZN, Evans C, etal. Ambulatory veno-venous extracorporeal membrane oxy-
genation: innovation and pitfalls. J Thorac Cardiovasc Surg. 2011;142:75561.
61. 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;30:9714.
62. Phipps P, Garrard CS.The pulmonary physician in critical care. 12: acute severe asthma in the
intensive care unit. Thorax. 2003;58:818.
63. Krishnan V, Diette GB, Rand CS, etal. Mortality in patients hospitalized for asthma exacerba-
tions in the United States. Am J Respir Crit Care Med. 2006;174:6338.
64. McFadden Jr ER, Lyons HA.Arterial-blood gas tension in asthma. N Engl J Med. 1968;278:
102732.
65. Braman SS, Kaemmerlen JT.Intensive care of status asthmaticus. A 10-year experience.
JAMA. 1990;264:3668.
66. 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;119:1159.
67. 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;48:1113.
68. Conrad SA, Green R, Scott LK.Near-fatal pediatric asthma managed with pumpless arteriove-
nous carbon dioxide removal. Crit Care Med. 2007;35:26249.
69. 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;35:9458.
70. Shapiro MB, Kleaveland AC, Bartlett RH.Extracorporeal life support for status asthmaticus.
Chest. 1993;103:16514.
71. 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.
72. 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;55:4752.
73. MacDonnell KF, Moon HS, Sekar TS, Ahluwalia MP.Extracorporeal membrane oxygenator
support in a case of severe status asthmaticus. Ann Thorac Surg. 1981;31:1715.
74. Brenner K, Abrams DC, Agerstrand CL, Brodie D.Extracorporeal carbon dioxide removal for
refractory status asthmaticus: experience in distinct exacerbation phenotypes. Perfusion.
2014;29:268.
75. Del Sorbo L, Boffini M, Rinaldi M, Ranieri VM.Bridging to lung transplantation by extracor-
poreal support. Minerva Anestesiol. 2012;78:24350.
Chapter 5
ECLS as a Bridge to Lung Transplantation
Christian Kuehn
Introduction
Extracorporeal life support (ECLS) has for many years been an established treatment
procedure for patients suffering from acute cardiac, respiratory, or cardiorespira-
tory failure. The procedure can essentially be deployed as a bridge to recovery or
bridge to transplantation where, depending on the underlying disease, a heart,
lung, or combined heartlung transplant is being considered. There has recently
been rapid technical development, and this is nowadays revolutionizing ECLS
treatment before, during, and after lung transplantations. In this chapter, we dis-
cuss technical developments in ECLS treatment and their application to the eld of
lung transplantation.
Historical Aspects
Since the 1980s, lung transplantation has evolved from an experimental technique
to an established option for treatment of severe lung diseases and diseases of the
pulmonary circulation. Technical advances in isolated lung transplantation were
driven especially by Cooper and colleagues in Toronto. In this early era of lung
transplantation, ECLS was viewed as a contraindication to transplantation because
mechanical support was marked by myriad complications and poor outcomes.
Especially troublesome were bleeding, hemolysis, infection, and technical compli-
cations with the hardware then available. Nevertheless, as the number of transplan-
tations rose and transplantation techniques developed further, in 1991 our center
C. Kuehn, MD (*)
Department of Cardiac, Thoracic, Transplantation and Vascular Surgery, Privatdozent Dr.
med., Hannover Medical School, Carl-Neuberg-Str. 1, Hannover 30625, Germany
e-mail: kuehn.christian@mh-hannover.de

106 C. Kuehn
reported the rst use of ECLS as a bridging strategy for re-transplantation in patients
suffering severe early graft failure [1].
This report provoked discussion and controversy. After all, deterioration to the
point of mechanical ventilation or necessity for ECLS signaled the end of transplant
candidacy in most lung transplantation centers. At the time, this was a sensible
approach in light of the existing complications, but also the inability to sustain
patients for more than a few days using ECLS. The rst report in 1992 of successful
ECLS bridging to lung transplantation for post-traumatic ARDS remained, in the
end, only a case report that led to further debate [2]. Thus, from this early era of lung
transplantation, one nds no systematic data and only a few case reports about the
use of ECLS.
Over the ensuing decade, lung transplantation became ever more established and
the number of transplants continued to rise. As transplantation became a realistic
option for patients with end-stage lung disease, the number of patients on the wait-
ing list climbed. Indeed, listed patients rose out of proportion to the available and
suitable organs, so that increasing numbers of patients decompensated while await-
ing a donor. Mechanical ventilation was still viewed as a signicant risk factor for
mortality after lung transplantation [3].
Even after successful intubation many patients had life-threatening hypercapnia and
acidosis despite all attempts to ventilate adequately. Further technical developments in
ECLS and the introduction of pumpless support systems offered a novel option: extra-
corporeal carbon dioxide removal (ECCO2R; Chap. 4) as a bridge to lung transplanta-
tion [4]. These patients could be successfully bridged with a pumpless (arteriovenous)
device, with a notable 1-year survival rate after transplantation of 80 %. The median
duration of ECLS support was over 2 weeks but, because pumpless AV ECLS allows
only low blood ow rates, it is suitable only for CO2 elimination.
For patients with greatly impaired oxygen exchange, adequate saturation of the
blood cannot be attained with low-ow, pumpless circuits (see Chap. 1). By adding
a pump and taking advantage of newer, low-resistance membranes, sufcient blood
ow could be captured to make venovenous ECLS practical for sustaining both CO2
removal and oxygenation [5]. Numerous technical innovations led to a clear reduc-
tion of complications and lowered the threshold for using ECLS early, rather than
only as a last resort. This cast an old tool in a new light: ECLS could be utilized to
bridge even the sickest patients to transplant.
The single, most important technological change was probably the introduction
of coated, polymethylpentene, hollow-ber membranes in the oxygenator, affording
excellent gas transfer [6]. Another crucial advance was to rene the system surface
coating to enhance blood compatibility. Together with powerful centrifugal pumps
engineered to minimize trauma to blood elements, systems could run in a stable
manner for a long duration, some up to 30 days. Based on these technical improve-
ments and the increasing routine around the use of ECLS, the Hanover group pub-
lished the rst series of patients bridged to transplant under awake-ECLS: using no
sedation or mechanical ventilation [7]. This success led to more and more lung
transplant centers adopting ECLS and awake-ECLS as an option for carrying
patients successfully to transplant.
5 ECLS as a Bridge to Lung Transplantation 107
Support Variations
For lung-transplant bridging, three principal ECLS methods are considered. The rst
is to create an arteriovenous shunt from one femoral artery to the (contralateral)
femoral vein, using the systemic blood pressure to drive ow across the gas-
exchanging membrane. This pumpless conguration is referred to as interventional
lung assist (iLA). The second option is a venovenous (VV) ECLS system, which
typically drains from the inferior vena cava and supplies the right atrium. A third
approach is venoarterial (VA) ECLS, whereby blood drains from the right atrium
and feeds into a large artery, such as the femoral artery, subclavian artery, or aorta.
Each of these modes is further described in Chap. 6. In individual cases the implantation
of additional cannulas is considered, depending on the experience of the center.
When choosing a cannulating strategy, the therapy goal (successful bridging to
transplant) should always be kept in mind.
Interventional Lung Assist
As mentioned above, iLA has a special place in bridging to lung transplantation.

In a heterogenous patient population with refractory hypercapnia, it has already
proved successful in removing much of the metabolic production of carbon dioxide,
as shown by Fischer and colleagues [4]. Especially in regard to the duration of
support and the remarkable 1-year survival, these results altered fundamentally the
role for ECLS. Nevertheless, the low blood ow rates that limit application to
hypercapnic lung failure, along with the concern for leg ischemia and arterial com-
plications, led to this mode largely being supplanted by pump-driven, VV-ECLS.
Venovenous ECLS
This classical form of support for respiratory insufciency can take place with two
different variants of cannulation. Using double-lumen cannulas (mostly inserted in the
jugular vein), patients with hypercapnic lung failure can be bridged very well. Smaller
diameter, dual-lumen cannulas afford only modest ow rates, often not sufcient for
the simultaneous saturation of the blood with oxygen. Nevertheless CO2 can be elimi-
nated very well with low ow rates, as with iLA. Another advantage of dual-lumen,
internal jugular cannulas is the fact that only one cannulation is required. With a
single cannulation of the jugular vein (mostly on the right side), patients retain free-
dom of movement at the legs and hips so as to participate in physical therapy and
mobilization. Since the dual-lumen cannula allows sedation, intubation, and
mechanical ventilation to be avoided, many patients can be kept awake and ambula-
tory while awaiting transplant, providing a real advantage compared to mechanical
ventilation. This approach is particularly applicable to patients with cystic brosis.
108 C. Kuehn
For those with severely impaired oxygenation, either in isolation or in addition to

hypercapnia, more traditional VV-ECLS may be needed. In this case, a large venous
cannula is advanced through the femoral vein (usually on the right side) for venous
drainage. Oxygenated blood is then returned to the patient via a single-lumen cannula
in the internal jugular vein (also on the right side). In order to attain excellent
oxygenation, we strive to maintain adequate blood ow and an optimal cannula
separation so as to prevent recirculation of already oxygenated blood.
In a bridge-to-transplant setting, this form of cannulation is most suitable for
patients with pulmonary brosis, sarcoidosis, or chronic bronchiolitis obliterans
complicating prior lung transplantation. The primary approach in these cases is to
establish awake-ECLS earlier, rather than later, to avoid the negative consequences
of mechanical ventilation. At this point we should emphasize that we institute ECLS
only in patients who have been accepted already to the transplant waitlist following
complete evaluation as to suitability. This avoids the challenges of ECLS becoming
a bridge to nowhere as described in Chap. 13.
Venoarterial ECLS
For patients who have isolated lung failure (without concurrent circulatory failure),
the left ventricle continues to eject de-oxygenated blood despite VA-ECLS. Inguinal
cannulation would thus produce an oxygen watershed in the aorta [8], threatening
the oxygen supply to coronary arteries and the upper part of the body, and making
VA-ECLS a less than ideal choice. For patients deteriorating from pulmonary
hypertension, however, gas exchange usually remains acceptable. This advantage
(compared to patients with gas exchange failure) makes peripheral VA-ECLS a
good option for bridging these patients to transplantation. As these patients slide
into circulatory insufciency, VA-ECLS through the femoral artery unloads the
right heart and stabilizes them hemodynamically. As a rule, the efciency of gas
exchange is maintained so that there is no danger of hypoxemia in the area of the
coronary vessels or supra-aortic branches.
This form of support is also established under local anesthesia in the awake
patient. To avoid the danger of distal, arterial leg ischemia, we routinely insert a
small catheter oriented to assure perfusion of the distal leg with oxygenated blood.
We try to implant the cannulas in the groin on only one side: this allows maximal
freedom of movement on the contralateral side, enabling mobilization of the patient.
Other Modes of Support
The most difcult question to answer is how to support a patient with combined gas
exchange failure and right heart dysfunction. Because of the right heart problem,
VV-ECLS is not sufcient. Yet VA-ECLS by itself is not suitable due to the
watershed and danger of hypoxemia in the upper half of the body. One could consider
central ECLS cannulation, but this is clearly more invasive, so that this means rarely
succeeds with local anesthesia in waking patients. In these fragile patients, even a
brief period of anesthesia and the risks of intubation are best avoided.
In these cases we usually begin with VV support. Should this not be sufcient
because of right heart problems, an arterial limb (with distal leg perfusion) is estab-
lished also under local anesthesia, converting to a veno-arterial-venous (VAV)
mode. In the end, blood drains from the caudal atrium and returns through a
Y-connector to both the arterial and venous systems, supporting both gas exchange
and the circulation (see Chap. 6). In these cases, a throttle screw apportions ow
through the arterial and venous limbs; a second ow probe is recommended to regu-
late ow according to demand. This form of support is enormously fragile and
prone to malfunction, demanding a sufciently high expertise of the user, especially
in the context of bridging to lung transplantation.
Bridge to Lung Transplantation: General Considerations
In our interdisciplinary clinic, we examine, evaluate, and extensively educate can-

didates about the theme of transplantation, the uncertain waiting period, and the
inexact long-term prognosis after transplantation. Because deterioration is common
in patients awaiting transplant, we attempt to prepare them in advance for various
possible scenarios. Depending on the primary diagnosis and individual status of
the patient, we may broach the question of bridging strategies in case of decompen-
sation. It is not possible to create a simple algorithm to address these possibilities:
the approach must be tailored to the individual patient and rests, in large part, on the
experience of the transplant center in bridging methods.
Whether a patient is suitable for ECLS as a bridge to transplant (BTT) is based
largely on age and co-morbidities. Usually, we consider only patients with single
organ failure. In daily practice this means that young patients have a very high
potential for recovery, so ECLS often represents a good option. In general, there are
two scenarios in which the use of ECLS for patients awaiting lung transplant is
recommended: gas exchange failure (VV-ECLS) and right-sided heart failure
(VA-ECLS). BTT is a classic example of the use of ECLS to supplant mechanical
ventilation. In the best of circumstances and following full transplant evaluation, the
candidate and clinicians recognize impending failure before intubation, sedation,
mechanical ventilation, and the crisis of ICU admission lead to complications that
threaten the potential for successful transplant. In this best case scenario, we
attempt awake institution of ECLS, giving the optimal chance for continued
rehabilitation and avoidance of new barriers to transplant. Sometimes non-invasive
ventilation can be used to buy time. In this case, we recommend early consideration
for ECLS cannulation before additional deterioration leads to intubation.
If ECLS can be instituted electively, keeping the patient awake, the path to
successful transplant may be somewhat easier (see Chap. 12). Patients can continue
110 C. Kuehn
to eat and drink independently, interact and communicate, and participate in physical
therapy. Mobilization is somewhat restricted, often limiting participation in therapy
to what can be achieved in the bed or beside chair. On the other hand, some patients
can be ambulated successfully, even beyond the connes of the ICU. It is easy to
imagine that such an approach leads to better physical and mental preparation for
the rigors of recovery from transplantation. Meanwhile, all other measures usu-
ally taken to prepare patients for transplant can be conducted in the awake patient
on ECLS.
Often, this ideal scenario is not realized. The patient may deteriorate unexpectedly,
requiring urgent intubation. Alternatively, the awake-ECLS plan may simply fail,
leading to unplanned mechanical ventilation. If a patient has to be intubated acutely,
steps should be taken to facilitate weaning immediately once ECLS is instituted and
gas exchange and right heart function stabilize. Should complete weaning and extu-
bation be impossible, treatment should aim to withdraw sedatives and get the patient
breathing spontaneously. A thorough neurological assessment, a prerequisite for
transplantation, is only possible once the patient is awake and breathing.
Once a transplant candidate in placed on ECLS as a BTT, the outcome is deter-
mined largely by whether new complications set in before an acceptable donor organ
is procured. However, it is important to resist the pressure to accept marginal organs
just because the patient is on ECLS. To generate a good outcome it is important to
accept good organs for these sick patients [9]. Should a good donor organ be accepted,
the operation is usually done over a minimally invasive access through anterolateral
thoracotomy. At the end of the transplantation the decision for ECLS decannulation
lies with the surgeon. Only in patients with pulmonary hypertension who had
VA-ECLS implanted before transplantation is ECLS usually kept for left heart protec-
tion. Should decannulation be attempted, heparin anticoagulation is antagonized,
sometimes supplemented with platelets or coagulation factors. Through these
measures the risk of bleeding in the early postoperative phase can be minimized.
A very challenging scenario arises when complications intervene before the
patient can be transplanted. If absolute contraindications to transplantation develop,
ECLS becomes a bridge to nowhere. In this case, if there is no prospect for recov-
ery short of transplant, ECLS should be terminally withdrawn (see Chap. 13).
Additional end-organ failures (kidney, liver), fulminant infections, and inexorably
rising need for vasoactive infusions are seen by the authors as precluding successful
transplantation. In short: beware of catastrophic transplantations. Nevertheless, the
door to transplantation is not always closed denitively. At times, it may be appropriate
to remove the patient from the active transplant list, pending clinical improvement,
after which the topic of transplantation can be raised anew.
At the same time, one must never lose sight of the goal of ECLS as a BTT: transplant
must have a realistic chance for success. For occasional patients, ECLS may serve
as a bridge to recovery: that is, treatment of an identied precipitant for cardiopul-
monary failure may allow recovery and extubation, after which the patient can be
transplanted from a clinically stable situation. However, once the treating transplant
surgeon, intensivist, and pulmonologist no longer believe the goal of transplant is
realistic, an appropriate conversation should be conducted with the patient or his
surrogates to change the goals of treatment to palliation.
Sometimes ECLS is raised as a possible treatment for a patient with acute

respiratory failure who has not been previously considered for transplantation.
This has been termed bridge to decision. This is rarely advised, and only with
single-organ failure when transplantation is a realistic option. Even then, the patient
must undergo the same extensive evaluation as applied to outpatients with end-stage
lung disease. Further, all involved must acknowledge upfront that ECLS can be
withdrawn if transplantation is not offered and recovery is judged impossible.
Pulmonary Failure
If reduced gas exchange becomes life-threatening, the lung transplant candidate is

traditionally intubated and ventilated, at least when transplantation remains a realistic
option. It is often impossible to provide adequate mechanical ventilation in patients
with end-stage lung disease, resulting in profound hypercapnia or hypoxemia. Severe
hypercapnia may result in hemodynamic failure often indistinguishable from sepsis.
In such situations, ECCO2R can correct the hypercapnia and acidosis and reduce the
catecholamine dosage [4].
Pulmonary Hypertension and Right Ventricular Failure
In the event of right-sided heart failure due to pulmonary hypertension, VA-ECLS is

needed to ensure that adequate hemodynamics are restored. In clinical terms, right-
sided heart failure is most often detected if it is accompanied by incipient end-organ
failure, such as acute kidney failure or acute liver dysfunction. Invasive measure-
ments indicate high central venous pressures (CVP > 20 mmHg), low central venous
oxygen saturations (ScvO2 < 50 %), and a reduced cardiac output. This life-threatening
situation can be stabilized immediately with VA-ECLS, usually resulting in rapid
recovery of secondary organ dysfunction [10].
A special form of ECLS as BTT for patients with pulmonary hypertension has
been published jointly by the groups from Hanover and Toronto [11]. In this instance
a Novalung oxygenator was cannulated centrally into the pulmonary artery and the
left atrium without a blood pump, a technique that comes quite close to a lung
replacement in terms of anatomical hemodynamics. It was even possible to suc-
cessfully extubate some of these patients after the procedure [12].
Preventing Mechanical Ventilation with ECLS
The vast majority of ECLS procedures are deployed in patients who are intubated
and mechanically ventilated. However, it is well-documented that 1-year survival
rates after lung transplantations are less than 50 % in patients requiring mechanical
112 C. Kuehn
ventilation as BTT [13]. It is questionable whether the additional use of ECLS in a

patient cohort with such a poor prognosis can ever reap the potential benet of this
treatment. Hence high-risk patients may undergo lung transplantation on ventilation
and ECLS, resulting in poor postoperative outcomes. As the total number of lung
transplantations is limited by organ availability, the outcome of aggressive use of
ECLS in a cohort of pre-terminal mechanically ventilated patients could have a
negative impact on the overall survival after lung transplantation.
Therefore, efforts should be channeled in new direction. We have realized that
sedation, intubation, and mechanical ventilation of patients with terminal lung dis-
eases is associated with a very high morbidity and mortality. A new concept is to
avoid mechanical ventilationor at least to facilitate weaning from mechanical
ventilationby using ECLS [14]. The goal is to keep the patient awake, enabling
him to eat, drink, exercise, and communicate. This is the current strategy at Hanover
Medical School. In the meantime, more than 30 non-sedated patients were success-
fully bridged to lung transplantation with VV- or VA-ECLS. The data indicate a
90-day survival rate of 80 % after lung transplantation in patients who were success-
fully bridged. Even in that small number of cases this gure is signicantly better
than for historical patients who (with or without additional ECLS) underwent lung
transplantation on mechanical ventilation [15].
Other Indications for ECLS in Lung Transplantation
Intraoperative VA-ECLS
Advances in ECLS technology have led to a broader application of this technique in

other areas involving lung transplantation. One example is the intraoperative use of
VA-ECLS instead of classical cardiopulmonary bypass (CBP). The practice of using
a heartlung machine in lung transplantation varies greatly from center to center.
Disadvantages in lung transplantation include the need for full heparinization (which
increases the risk of bleeding complications) and the activation of non-specic
inammation (due to the large articial surface area and uncontrolled bloodair
interaction in the reservoir of the heartlung machine). Our center has shifted entirely
from intraoperative CBP to VA-ECLS using inguinal cannulas, resulting in a signi-
cant reduction in the need for blood products [16]. However, this conicts with data
from another group [17] which reported better results with CBP than with ECLS in
a small cohort.
Postoperative ECLS
Selected patients may benet from maintaining VA-ECLS support for a few days
after lung transplantation. For example, these could be patients who show acute
signs of ischemiareperfusion damage at the end of the operation. In the case of
severe ischemiareperfusion, gas exchange in the lung is sharply reduced, so that

adequate oxygenation and CO2 elimination cannot be achieved, even with poten-
tially injurious levels of mechanical ventilation. Moreover, pulmonary vascular
resistance increases, cardiac output falls, and signs of circulatory failure develop,
producing catecholamine dependency and acute kidney failure. VA-ECLS provides
immediate relief. In some cases, the use of ECLS may facilitate a lung-protecting
ventilation strategy, promoting recovery from ischemiareperfusion damage.
Another strategy is planned postoperative VA-ECLS in patients with severe,
long-standing pulmonary hypertension. Early postoperative lung failure has been a
major complication after lung transplantation in these patients in the past, resulting
in unacceptable outcomes. As a consequence, many centers prefer combined heart
lung transplantation for these patients. In a period of limited availability of donor
hearts, there are obvious limitations to this approach. It is now clear that patients
with severe pulmonary hypertension can be treated with double-lung transplanta-
tion when routinely supported by VA-ECLS during the postoperative period. The
reason is rather simple: patients with severe pulmonary hypertension have a small,
stiff left ventricle that is no longer capable of handling normal or higher-than-
normal blood ow rates. Thus, after transplantation, left-sided lling pressures
increase, resulting in acute pulmonary edema. This mechanism frequently results
in a vicious cycle leading to irreversible graft failure. Instead, VA-ECLS is started
at the beginning of lung transplantation, as in any case these patients need CPB
during the procedure. Following implantation, VA-ECLS reduces preload of both
the right and left ventricles, preventing deterioration. The patient is then trans-
ferred from the operating theater to intensive care with the ECLS running, and
extubation is attempted within 48 h. Only after the patient has been successfully
extubated the ECLS is weaned over 5 to 7 days with continuous monitoring of left
ventricular function and lling pressures [18].
ECLS for Organ Preservation
Over the last few years, intensive research has been conducted on the develop-
ment of normothermic perfusion systems for donor lungs. With Ex-Vivo Lung
Perfusion (EVLP), developed in Sweden in 2007, the donor lung is connected to
ventilator and a system of tubes, similar to extracorporeal circulation [19]. With a
normothermic system, it is possible to evaluate marginal donor organs and assess
organ function using parameters such as oxygen uptake, lung elasticity, and perfu-
sion and ventilation pressures, to determine their suitability for the transplantation
process. Moreover, the cold ischemic time can be reduced through the use of this
type of normothermic organ perfusion, thus preventing harm to the donor organ
from a lack of oxygen and nutrients. The further development of the EVLP can be
seen in the standardization of integrated components in a portable casing, the
Organ Care System (OCS). The use of the portable OCS system has been described
in an initial pilot trial conducted by our Hanover group in cooperation with
colleagues from Madrid [20]. This system was used on 12 patients where the
114 C. Kuehn
donor organs, following initial cold perfusion, were connected to the OCS and
transported with warm perfusion and ventilation to the transplant clinic for successful
transplantation.
Conclusions
ECLS treatment of patients before or after lung transplantation is as old as the

procedure itself. Thanks to major technical advances in recent years, complication
rates have fallen signicantly and it has become possible to support patients for
weeks or even months [21]. In the last 5 years ECLS treatment has moved from
being a procedure applied almost exclusively to ventilated patients in cardio-surgical
centers to a highly versatile treatment strategy that can protect patients from intuba-
tion and mechanical ventilation and is essentially able to ensure that critically ill
patients are in a better condition for transplantation. Furthermore, modern ECLS
systems are suitable for use peri-procedurally in thoracic transplantations and
improve results. Therefore, ECLS should be an established part of the treatment
spectrum in modern thoracic transplantation centers [2224].
References
1. Jurmann 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):947. discussion 98.
2. Demertzis S, Haverich A, Ziemer G, Nerlich M, Mller KM, Wagner TO, et al. Successful
lung transplantation for posttraumatic adult respiratory distress syndrome after extracorporeal
membrane oxygenation support. J Heart Lung Transplant. 1992;11(5):10057.
3. Smits JM, Mertens BJ, Van Houwelingen HC, Haverich A, Persijn GG, Laufer G. Predictors
of lung transplant survival in EuroTransplant. Am J Transplant. 2003;3:14006.
4. Fischer S, Simon AR, Welte T, Hoeper MM, Meyer A, Tessmann R, et al. Bridge to lung trans-
plantation with the novel pumpless interventional lung assist device NovaLung. J Thorac
Cardiovasc Surg. 2006 Mar;131(3):71923.
5. Fischer S, Hoeper MM, Tomaszek S, Simon A, Gottlieb J, Welte T, et al. Bridge to lung trans-
plantation with the extracorporeal membrane ventilator Novalung in the veno-venous mode:
the initial Hannover experience. ASAIO J. 2007 Mar-Apr; 53(2):16870.
6. 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.
2006 Sep-Oct; 52(5):50912.
7. Olsson KM, Simon A, Strueber M, Hadem J, Wiesner O, Gottlieb J, et al. Extracorporeal
membrane oxygenation in nonintubated patients as bridge to lung transplantation. Am J
Transplant. 2010 Sep;10(9):21738. Epub 2010 Jul 15.
8. Hoeper MM, Tudorache I, Khn C, Marsch G, Hartung D, Wiesner O, et al. Extracorporeal
membrane oxygenation watershed. Circulation. 2014;130(10):8645.
9. Sommer W, Khn C, Tudorache I, Avsar M, Gottlieb J, Boethig D, et al. Extended criteria
donor lungs and clinical outcome: results of an alternative allocation algorithm. J Heart Lung
Transplant. 2013;32(11):106572.
10. Gregoric ID, Chandra D, Myers TJ, Scheinin SA, Loyalka P, Kar B. Extracorporeal membrane
oxygenation as a bridge to emergency heart-lung transplantation in a patient with idiopathic
pulmonary arterial hypertension. J Heart Lung Transplant. 2008;27:4668 [EBM IV].
11. Strueber M, Hoeper MM, Fischer S, Cypel M, Warnecke G, Gottlieb J, et al. Bridge to thoracic
organ transplantation in patients with pulmonary arterial hypertension using a pumpless lung
assist device. Am J Transplant. 2009 Apr;9(4):8537.
12. de Perrot M, Granton JT, McRae K, Cypel M, Pierre A, Waddell TK, et al. Impact of extracor-
poreal life support on outcome in patients with idiopathic pulmonary arterial hypertension
awaiting lung transplantation. J Heart Lung Transplant. 2011;30(9):9971002.
13. 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
Jan-Feb;54(1):310.
14. Wang D, Zhou X, Lick SD, Liu X, Qian K, Zwischenberger JB. An ambulatory pulmonary and
right heart assist device (OxyRVAD) in an ovine survival model. J Heart Lung Transplant.
2007;26:9749.
15. 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):7638.
16. 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):15106.
17. Bittner HB, Binner C, Lehmann S, Kuntze T, Rastan A, Mohr FW. Replacing cardiopulmonary
bypass with extracorporeal membrane oxygenation in lung transplantation operations. Eur J
Cardiothorac Surg. 2007;31:4627.
18. Tudorache I, Sommer W, Khn C, Wiesner O, Hadem J, Fhner T, et al. Lung transplantation
for severe pulmonary hypertension-awake extracorporeal membrane oxygenation for postop-
erative left ventricular remodelling. Transplantation. 2015;99(2):4518.
19. Steen S, Ingemansson R, Eriksson L, Pierre L, Algotsson L, Wierup P, et al. First human trans-
plantation of a nonacceptable donor lung after reconditioning ex vivo. Ann Thorac Surg.
2007;83(6):21914.
20. Warnecke G, Moradiellos J, Tudorache I, Khn 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):18518.
21. Thiara AP, Hyland V, Norum H, Aasmundstad TA, Karlsen HM, Fiane AE, et al. Extracorporeal
membrane oxygenation support for 59 days without changing the ECMO circuit: a case of
Legionella pneumonia. Perfusion. 2009 Jan; 24(1):457.
22. Strueber M. Extracorporeal support as a bridge to lung transplantation. Curr Opin Crit Care.
2010;16:6973.
23. Rosenzweig EB, Brodie D, Abrams DC, Agerstrand CL, Bacchetta M. Extracorporeal mem-
brane oxygenation as a novel bridging strategy for acute right heart failure in group 1 pulmo-
nary arterial hypertension. ASAIO J. 2014;60(1):12933.
24. Turner DA, Cheifetz IM, Rehder KJ, Williford WL, Bonadonna D, Banuelos SJ, et al. Active
rehabilitation and physical therapy during extracorporeal membrane oxygenation while awaiting
lung transplantation: a practical approach. Crit Care Med. 2011;39(12):25938.
Chapter 6
Modes of ECLS
L. Keith Scott and Benjamin Schmidt
Introduction
Extracorporeal life support (ECLS) requires that blood ow from the body, through
a gas-exchanging membrane, and back again to the body. 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 venovenous (VV) ECLS, meaning
that blood is withdrawn from large central veins or the right atrium, and returned
also to the venous side of the circulation. In venoarterial (VA) ECLS, 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 VAV
ECLS. Finally, blood can move from artery to vein, using the patients native blood
pressure to drive ow through a low-resistance membrane (AV ECLS or AVCO2R).
The rest of this chapter describes the circuit design for each mode, their physiologi-
cal impact, factors leading to mode selection for individual patients, and the
advantages of each mode (Tables 6.1 and 6.2).
L.K. Scott, MD (*)

Department of Anesthesiology, Wake Forest University,
Medical Center Boulevard, Winston-Salem, NC 27105, USA
e-mail: lkscott@wakehealth.edu
B. Schmidt, MD
Department of Surgery, Wake Forest University,
Medical Center Boulevard, Winston-Salem, NC 27105, USA

118 L.K. Scott and B. Schmidt
Table 6.1 Selecting the mode: venovenous (VV), venoarterial (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 pul-
monary disease, eCPR extracorporeal cardiopulmonary resuscitation
Table 6.2 Venovenous (VV) and venoarterial (VA) ECLSadvantages and disadvantages
Factor VV VA
Cannulas One may sufce Two are needed
Gas exchange efciency Lower Higher
PaO2 Lower Higher
Recirculation Moderate None
Circulatory support None Full is possible
RV loading No 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
Venovenous ECLS
The VV Circuit
Cannulation for VV ECLS consists of inserting large-bore drainage cannulas in the infe-
rior (IVC) and superior (SVC) vena cavae, usually via the femoral and internal jugular
veins (Fig. 6.1). Dual-lumen cannulas have channels for both removing and returning
blood, and require only a single insertion site (Fig. 6.2). These are placed so that outow
6 Modes of ECLS 119
Fig. 6.1 Illustration of the

dual-cannula setup for VV
ECLS. Blood is withdrawn
from the venous cannula and
returned into the superior
vena cava
ports are located in both the SVC and IVC (traversing the right atrium; RA), while the
inow port resides within the RA and is directed toward the tricuspid valve [1]. With both
of these circuit congurations, blood ows from the central veins through external tubing
to pump, membrane, warmer, and other circuit components (see Chap. 8), 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 efciency of oxygenation [2]. The
degree of recirculation depends strongly on cannula location, but can be inuenced by
the intravascular volume state, circuit blood ow rates, and patient positioning.
Single-site, dual-lumen cannulation for VV ECLS has some theoretical benets, such
as less recirculation when compared with dual-site cannulation [3]. Downsides include
increased cost, larger cannula diameter, higher risk, and need for continuous visualiza-
tion during insertion (Chap. 7). Dual-lumen cannulas are more expensive to purchase
than two single-lumen cannulas, but these costs are expected to decrease as technology
advances and more vendors enter the marketplace. Dual-lumen cannulas for adult VV
ECLS range in size from 23 to 31 French. Achievable circuit ow depends strongly on
Fig. 6.2 Illustration of the

standard ECLS conguration
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
cannula diameter, so the largest size possible is preferred. A general rule of thumb is that
the vessel diameter (in millimeters; judged by bedside ultrasound) should be at least one-
third the cannula size (in French). Cannulation is more technically challenging and risk-
ier than for the two-site approach. Placement of the guidewire down into the IVC should
be veried and the cannula visualized continuously during insertion to avoid catastrophic
vascular or cardiac injuries, requiring additional personnel, space, and equipment.
Physiological Implications of VV ECLS
The physiology of VV ECLS is fully described in Chap. 1, but several points deserve
emphasis here. First, since oxygenated blood returns to the venous circulation, it
mixes with venous blood that was not captured by the outow cannula. The oxygen
saturation (SO2) of the resulting mixture is often in the mid-80s, even though the oxy-
gen 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
6 Modes of ECLS 121
that is lower than many clinicians are comfortable with and, as discussed below, lower
than is typically achieved during VA ECLS. 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 outow cannulas and directed to the membrane. Thus full support of oxygen-
ation typically requires circuit blood ow rates of 35 L/min, values that sometimes
produce hemolysis, recirculation, or circuit chattering (when the great veins collapse
around the cannula). If the goal of ECLS is largely carbon dioxide removal (ECCO2R;
see Chap. 4), much lower ow rates are possible [4] 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 [5] or status asthmaticus [6], or to support an ultra-lung-protective
mechanical ventilation strategy for patients with severe ARDS [7, 8].
A second physiological consequence of VV ECLS is that it has virtually no
circulatory impact, a result of the fact that equal ows of blood are simultaneously
withdrawn and returned [9]. Because of this, right and left heart function, as well as
the pulmonary circulation, must be reasonably intact for VV ECLS to be an appro-
priate 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 [10] related to the pulmonary vascular effects of lung
injury, hypoxia, thrombosis, and mechanical ventilation. When cor pulmonale is
severe, VA ECLS may be more appropriate. On the other hand, the VV mode has
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 is likely to fall. In the
CESAR Trial of ECLS for severe ARDS, all patients were managed with VV ECLS
and no patient had to transition to a VA circuit during the trial [11]. In contrast to the
VV mode, VA ECLS 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 6.1). 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 inuence the choice of
VV or VA ECLS.
Advantages of VV ECLS
Several advantages are conferred because the VV mode returns blood to the central
veins, rather than the arteries (Table 6.2). These include a reduced risk of systemic
embolism, including catastrophic cerebral or coronary embolism; no injury to
systemic 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 sufce, possibly facilitating mobilization (see Chap.
12). VV ECLS 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.
Venoarterial ECLS
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 positioned
in the RA, SVC, or IVC. Blood is returned through the carotid artery in neonates
and infants (Fig. 6.3) and the descending aorta (via the femoral artery) in older
children and adults (Fig. 6.4). Where to place the return cannula depends partly on
the blood ow needed. If a need for full support is anticipated (for example,
100125 cc/kg/min in a neonate), the excessive resistance of peripheral arteries may
not allow sufcient ow [12]. Central cannulation of the right atrium and the
ascending aorta allows larger cannulas to be used, providing higher ows and
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 [13]. If
the arterial cannula compromises or obstructs nutrient blood ow, 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 ow (Fig. 6.5) [14].
Physiological Implications of VA ECLS
Venoarterial (VA) ECLS can support both respiration and circulation [15]. For
adults with respiratory failure but preserved cardiac function, VV ECLS is pref-
erable, whereas VA ECLS is appropriate for those with circulatory failure
(Table 6.1). With regards to oxygenation and carbon dioxide elimination, the
gas exchange effects of VA ECLS are similar to those of VV ECLS, 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 pro-
vided by the (failing) native circulation. In contrast, with the VV conguration
much venous (desaturated) blood is not captured by the drainage cannula(s) and
passes through the diseased lungs, producing venous admixture and causing
lower PaO2 values. Second, cannula position in VA ECLS precludes the possi-
bility of recirculation, so gas exchange is more efcient. 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.
6 Modes of ECLS 123
Fig. 6.3 Illustration of a

venoarterial (VA)
conguration for neonates
and infants. Blood is
withdrawn from the superior
vena cava and returned into
the carotid artery in neonates
and infants
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 saturation) 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 peculiar-
ity of varying arterial oxygen values may further threaten the circulation or
neurological status.
While the gas exchange differences between VV and VA ECLS are modest,
circulatory physiology is completely different. VV ECLS has essentially no
circulatory effect, yet the basic function of VA ECLS is to provide circulatory
support, returning oxygenated blood to the arterial circulation at physiological
perfusion pressures [16]. Since blood is drawn from the vena cava or RA, VA
ECLS unloads the right ventricle, and this mode has been used for patients with
massive pulmonary embolism, for example. However, the left ventricle does not
benet similarly. Even though the LV receives less blood from the lungs [17],
a failing LV may only eject little stroke volume, especially once extracorporeal
Fig. 6.4 Illustration of VA

cannulation for older children
and adults; CFA and CFV
Fig. 6.5 Antegrade perfusion

catheter (from [14]).
Reproduced with permission
6 Modes of ECLS 125
support raises blood pressure to normal levels. Related to the low native cardiac
output, aortic root ow may be sluggish 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 decom-
press the LV. Such measures should be included in the VA ECLS care pathway
[18]. 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 transcutaneous septoplasty, insertion of a vent cannula through a mini
thoracotomy [19], or placement of an intra-aortic balloon pump (IABP). One
concern with the use of an IABP is that it could obstruct ECLS ow returning
through the descending aorta [20]. However, several studies have shown a ben-
et of the IABP in off-loading the LV [21] and possibly enhancing cerebral [22]
and coronary blood ow [23].
Another difference between VV and VA ECLS relates to systemic arterial
pulsatility. Since pump ow 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 ECLS tends to be signicantly 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 [24]. Mean blood pressure relates to total cir-
cuit ow, 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 met-
rics related to pulse contour analysis (including many modern minimally inva-
sive cardiac output technologies) will not be valid. Still, the return of a pulsatile
waveform on the peripheral arterial transducer can be useful as a gauge of
cardiac recovery.
Advantages of VA ECLS
In comparison with VV ECLS, the VA mode leads to more efcient gas exchange
and no possibility of recirculation. Systemic PO2 is typically much higher than in
VV conguration, although this is sometimes not true of the cerebral, coronary, and
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 [25].
Hybrid ECLS Configurations
The traditional ECLS congurations each have limitations. For example, VA ECLS
may not provide adequate cardiac and cerebral oxygenation, particularly with
peripheral cannulation. Likewise, if a patient on VV ECLS suffers declining heart
function, some form of hemodynamic support may be needed. Hybrid systems aim
to address these challenges. Over the past few years many different congurations
have been reported. We focus on the systems most studied.
Veno-arterio-venous (VAV) ECLS
In this conguration, a venous outow cannula leads to the pump and articial
membrane, after which blood is returned both to the right atrium (or vena cava) and
to the systemic arteries (Fig. 6.6). Often this arises when a patient is started on
Fig. 6.6 Illustration of a

hybrid system that started as
a venoarterial conguration
(femoral vein withdrawal,
femoral artery return) and
was converted to veno-
arterial-venous (VAV)
because of upper body
hypoxemia. This
conguration also arises
following initial VV support
when circulatory support
becomes necessary, requiring
the addition of a venoarterial
component (here, femoral
artery)
6 Modes of ECLS 127
VA ECLS 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
(inow, connected by a Y to the arterial inow) in the IVC, SVC, or RA allows
a stream of fully oxygenated blood to traverse the lungs, raising the saturation of
blood passing through the left heart and into the proximal aorta [26]. A downside
lies in the fact that ow is diverted from the arterial system, possibly reducing
oxygen delivery. Increasing the circuit ow rate may compensate for this but may
be limited by cannula size or the adequacy of venous drainage. Furthermore, after
a venous inow cannula is spliced into the arterial limb of the circuit, much of the
ow may preferentially 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 inow limb of the
circuit (Fig. 6.7).
Fig. 6.7 VAV cannulation

with partially occlusive line
clamp on venous inow
Alternatively, this hybrid conguration may be used when a patient on VV ECLS

develops cardiac dysfunction, for example due to progressive cor pulmonale. Now
the addition of an arterial inow cannula unloads the RV and directly supports the
circulation.
Other hybrid systems have been described, such as combining VV ECLS with
IABP support [27] or performing an atrial septoplasty to unload the RV in a patient
with severe veno-occlusive disease [28]. There are few data on these hybrid modes
and their impact on ECLS duration, complications, or survival. In one small series,
ARDS patients started on or transitioned to VAV ECLS appeared to have a survival
benet [29]. However, the rationale for transitioning from VV to VAV was not
well-dened and appeared to be related to mechanical issues (poor venous drainage),
not to an augmented cardiac output.
AV Pumpless ECLS for CO2 Removal
For some patients, especially those with severe airow obstruction, or to facilitate
lung protection in those with ARDS, extracorporeal therapy focuses on carbon
dioxide removal rather than oxygenation (see Chap. 4). This is termed Arterio-
Venous CO2 Removal (AVCO2R). Mathematical simulation shows that it is possible
to achieve total CO2 removal by directing only 1015 % of cardiac output to the
articial membrane as long as the sweep gas to blood ow ratio is 5 or greater [30].
These modest blood ow requirements can be met with correspondingly smaller
cannulas. Further, modern hollow-ber membranes present such a small resistance
to blood ow that the patients own arterial to venous blood pressure gradient is
sufcient to produce the needed circuit ow. The complete circuit can be very sim-
ple. A 1214Fr arterial catheter is inserted percutaneously using the Seldinger tech-
nique and the vein is cannulated using a larger cannula to assure low resistance and
to maximize blood ow. Oxygen is used as the sweep gas. In such a pumpless sys-
tem, circuit ow is determined by the size of the arterial cannula (the highest resis-
tance component of the circuit), and the pressure gradient between the arterial and
venous systems. As with more complex extracorporeal circuits, carbon dioxide
removal is related to sweep gas and circuit blood ow. In one clinical trial, blood
ow rates above 500 mL/min were seen, producing a mean CO2 removal of 112 mL/
min [31]. This system has shown efcacy in life-threatening asthma, COPD exacer-
bations, and ARDS [32, 33]. Pediatric use has been described, although cannula size
is more likely to be limiting than in adults.
We cared for an intubated 8-year-old male with asthma and severe dynamic
hyperination and marked hypercapnia (pH 6.8; PaCO2 184 mmHg). After sizing
the femoral vessels with ultrasound, a 12F arterial and a 14F venous catheter were
inserted. Using the conguration shown in Fig. 6.8, circuit ow was 750 mL/min.
The evolution of pH and PaCO2 over the rst hours on AV pumpless ECLS is shown
in Fig. 6.9. AVCO2R support allowed us to rapidly reduce the ventilator settings,
improve dynamic hyperination, and avoid barotrauma.
6 Modes of ECLS 129
Fig. 6.8 The conguration of the AVCO2R circuit. A simple, pumpless system using arterial and
venous cannulas attached to a low-resistance, hollow-ber gas exchange membrane
Fig. 6.9 Graph showing the

initiation of AVCO2R. At the
same time, minute ventilation
was lowered to reduce
dynamic hyperination:
nevertheless, within 1 h
PaCO2 fell while pH rose
References
1. Thiagarajan RR, Brogan TV, Scheurer MA, Laussen PC, Rycus PT, Bratton SL. Extracorporeal
membrane oxygenation to support cardiopulmonary resuscitation in adults. Ann Thorac Surg.
2009;87(3):77885.
2. Combes A, Bacchetta M, Brodie D, Mller T, Pellegrino V. Extracorporeal membrane oxygen-
ation for respiratory failure in adults. Curr Opin Crit Care. 2012;18(1):99104.
3. Javidfar J, Wang D, Zwischenberger JB, Costa J, Mongero L, Sonett J, et al. Insertion of bica-
val dual lumen extracorporeal membrane oxygenation catheter with image guidance. ASAIO
J. 2011;57(3):2035.
4. Gattinoni L, Kolobow T, Tomlinson T, Iapichino G, Samaja M, White D, Pierce J. Low-
frequency positive pressure ventilation with extracorporeal carbon dioxide removal (LVPPV-
ECCO2R): an experimental study. Anesth Analg. 1978. http://journals.lww.com/
anesthesia-analgesia/Fulltext/1978/07000/Low_Frequency_Positive_Pressure_Ventilation_
with.18.aspx. Accessed 20 May 2014.
5. Abrams DC, Brenner K, Burkart KM, Agerstrand CL, Thomashow BM, Bacchetta M, et al.
Pilot study of extracorporeal carbon dioxide removal to facilitate extubation and ambulation in
exacerbations of chronic obstructive pulmonary disease. Ann Am Thorac Soc. 2013;10(4):

30714.
6. Brenner K, Abrams DC, Agerstrand CL, Brodie D. Extracorporeal carbon dioxide removal for
refractory status asthmaticus: experience in distinct exacerbation phenotypes. Perfusion. 2014;
29(1):268.
7. Fanelli V, Costamagna A, Terragni PP, Marco Ranieri V. Low-ow ECMO and CO2 removal.
In: Sangalli F, Patroniti N, Pesenti A, editors. ECMO-extracorporeal life support in adults.
Milan: Springer; 2014. p. 30315.
8. Gramaticopolo S, Chronopoulos A, Piccinni P, Nalesso F, Brendolan A, Zanella M, et al.
Extracorporeal CO2 removala way to achieve ultraprotective mechanical ventilation and lung
support: the missing piece of multiple organ support therapy. In: Ronco C, Bellomo R, McCullough
PA, editors. Contributions to nephrology, vol. 165. Basel: Karger; 2010. p. 17484.
Med. 2011;365(20):190514.
10. Vieillard-Baron A, Schmitt JM, Augarde R, Fellahi JL, Prin S, Page B, et al. Acute cor pulmo-
nale in acute respiratory distress syndrome submitted to protective ventilation: incidence,
clinical implications, and prognosis. Crit Care Med. 2001;29(8):15515.
11. Peek GJ, Mugford M, Tiruvoipati R, Wilson A, Allen E, Thalanany MM, et al. Efcacy and
trolled trial. Lancet. 2009;374:135163.
12. Formica F, Mariani S, Paolini G. Surgical cannulation: indication, technique, and complica-
tions. In: Sangalli F, Patroniti N, Pesenti A, editors. ECMO-extracorporeal life support in
adults. Milan: Springer; 2014. p. 4963.
13. Ariyaratnam P, McLean LA, Cale ARJ, Loubani M. Extra-corporeal membrane oxygenation
for the post-cardiotomy patient. Heart Fail Rev. 2014;19(6):71725.
14. Mohite PN, Fatullayev J, Maunz O, Kaul S, Sabashnikov A, Weymann A, et al. Distal limb
perfusion: Achilles heel in peripheral venoarterial extracorporeal membrane oxygenation.
Artif Organs. 2014;38(11):9404.
15. Nguyen DQ, Kulick DM, Bolman III RM, Dunitz JM, Hertz MI, Park SJ. Temporary ECMO sup-
port following lung and heart-lung transplantation. J Heart Lung Transplant. 2000;19(3):3136.
16. Marasco SF, Lukas G, McDonald M, McMillan J, Ihle B. Review of ECMO (Extra Corporeal
Membrane Oxygenation) support in critically ill adult patients. Heart Lung Circ. 2008;17
Suppl 4:S4147.
17. Rubino A, Haddon R, Corti F, Sangalli F. Complications of extracorporeal support and their
management. In: Sangalli F, Patroniti N, Pesenti A, editors. ECMO-extracorporeal life support
in adults. Milan: Springer; 2014. p. 41523.
18. Doll N, Kiaii B, Borger M, Bucerius J, Krmer K, Schmitt DV, et al. Five-year results of 219
consecutive patients treated with extracorporeal membrane oxygenation for refractory postop-
erative cardiogenic shock. Ann Thorac Surg. 2004;77(1):1517.
19. Attisani M, Centofanti P, La Torre M, Bofni M, Ricci D, Ribezzo M, et al. Transapical left
ventricular vent (TLVV) during veno-arterial ECMO support: a bridge to solution in acute
cardiogenic shock. Eur Heart J. 2013;34 Suppl 1:3724.
20. Schroeter T, Vollroth M, Hbartner M, Dhein S, Sahlisch S, Borger MA, Mohr
FW. Extracorporeal membrane oxygenation and intra-aortic balloon pumpan appropriate
combination or useless battle of materials? Thorac Cardiovasc Surg. 2013;61, no.S01. doi:10.
1055/s-0032-1332258.
21. Greco G, Cortinovis B, Avalli L. Left ventricular rest and unloading during VA ECMO. In:
Sangalli F, Patroniti N, Pesenti A, editors. ECMO-extracorporeal life support in adults. Milan:
Springer; 2014. p. 193206.
22. Yang F, Jia ZS, Xing JL, Wang Z, Liu Y, Hao X, et al. Effects of intra-aortic balloon pump on
cerebral blood ow during peripheral venoarterial extracorporeal membrane oxygenation
support. J Transl Med. 2014;12(1):106.
6 Modes of ECLS 131
23. Pierrakos CN, Bonios MJ, Drakos SG, Charitos EI, Tsolakis EJ, Ntalianis A, et al. Mechanical
assistance by intra-aortic balloon pump counterpulsation during reperfusion increases coro-
nary blood ow and mitigates the no-reow phenomenon: an experimental study. Artif Organs.
2011;35:86774. doi:10.1111/j.1525-1594.2011.01241.x.
24. Mohammadzadeh A, Jafari N, Hasanpour M, Sahandifar S, Ghafari M, Alaei V. Effects of
pulsatile perfusion during cardiopulmonary bypass on biochemical markers and kidney func-
tion in patients undergoing cardiac surgeries. Am J Cardiovasc Dis. 2013;3(3):15862.
25. Stub D, Bernard S, Pellegrino V, Smith K, Walker T, Sheldrake J, Hockings L, Shaw J, Duffy
SJ, Burrell A, Cameron P, Smit DV, Kaye DM. Refractory cardiac arrest treated with mechani-
cal CPR, hypothermia, ECMO and early reperfusion (the CHEER trial). Resuscitation. 2014.
pii: S0300-9572(14)00751-5. doi:10.1016/j.resuscitation.2014.09.010. [Epub ahead of print].
26. Moisan M, Lafargue M, Calderon J, Oses P, Ouattara A. Pulmonary alveolar proteinosis
requiring hybrid extracorporeal life support, and complicated by acute necrotizing pneumo-
nia. Ann Fr Anesth Reanim. 2013;32(4):e715.
27. Nakamura H, Yamaguchi H, Amano A, Nakao T. Venovenous extracorporeal membrane
oxygenation is effective against post-cardiotomy acute respiratory failure in adults. Gen
Thorac Cardiovasc Surg. 2013;61(7):4028.
28. Hoopes CW, Gurley JC, Zwischenberger JB, Diaz-Guzman E. Mechanical support for pulmo-
nary veno-occlusive disease: combined atrial septostomy and venovenous extracorporeal
membrane oxygenation. Semin Thorac Cardiovasc Surg. 2012;24(3):2324.
29. Sthr F, Emmert MY, Lachat ML, Stocker R, Maggiorini M, Falk V, et al. Extracorporeal
membrane oxygenation for acute respiratory distress syndrome: is the conguration mode an
important predictor for the outcome? Interact Cardiovasc Thorac Surg. 2011;12(5):67680.
30. Conrad SA, Brown EG, Grier LR, Baier J, Blount J, Heminc T, Zwischenberger JB, Bidani
A. Arteriovenous extracorporeal carbon dioxide removal: a mathematical model and experi-
mental evaluation. ASAIO J. 1998. http://journals.lww.com/asaiojournal/Fulltext/1998/07000/
Arteriovenous_Extracorporeal_Carbon_Dioxide.7.aspx. Accessed 16 May 2014.
31. Conrad SA, Zwischenberger JB, Grier LR, Alpard SK, Bidani A. Total extracorporeal arterio-
venous carbon dioxide removal in acute respiratory failure: a phase I clinical study. Intensive
Care Med. 2001;27:134051.
32. Jr B, Robert L, Tao W, Bidani A, Alpard SK, Traber DL, et al. Prolonged hemodynamic
stability during arteriovenous carbon dioxide removal for severe respiratory failure. J Thorac
Cardiovasc Surg. 1997;114(6):110714.
33. Mauri T, Zanella A, Pesenti A. Extracorporeal gas exchange: present and future. In: Vincent
J-L, editor. Annual update in intensive care and emergency medicine 2013. Berlin Heidelberg:
Springer; 2013. p. 60919.
Chapter 7
Vascular Access for ECLS
Steven A. Conrad
Introduction
Access to the central circulation to provide and maintain blood ow necessary for
adequate gas exchange is one of the most essential aspects for successful extracor-
poreal support. Inadequate extracorporeal ow can lead to failure to deliver suf-
cient support and limit any potential benet of ECLS. Cannulae and cannula
insertion techniques are quite variable, and the choice will depend on the goals and
mode of support, the size of the patient, size of the vessels, as well as institutional
and logistical concerns.
Cannulas for Extracorporeal Support
A variety of cannulae for peripheral vascular access are commercially available.

These cannulae differ with respect to the mode of insertion (percutaneous or surgical),
blood ow direction (drainage or reinfusion), and wall reinforcement, as well as being
available in various lengths and diameters to accommodate the choice of vessel.
Cannulae designed for percutaneous peripheral insertion have some minor fea-
ture differences from those intended for surgical placement. The loading dilator that
accompanies a percutaneous cannula has a long taper and central lumen to accom-
modate a guidewire, whereas the surgical cannula has a blunt dilator with a short tip
and no central lumen. The tip of a percutaneous cannula is designed to t snugly
S.A. Conrad, MD, PhD, MCCM, FCCP (*)

Department of Medicine, Emergency Medicine and Pediatrics, Louisiana State University
Health Sciences Center, 1501 Kings Highway, Shreveport, LA 71103-4228, USA
e-mail: sconrad@lsuhsc.edu

134 S.A. Conrad
against the loading dilator and is tapered to facilitate insertion through tissue,
whereas this feature is optional in surgical cannulas.
Wire-reinforced cannulas contain a layer of metal spiral-wound wire embedded
in the wall of the cannula. This reinforcement allows the cannula to ex without
kinking, resist attening from external compression, and prevent collapse when neg-
ative pressures are applied to the lumen. Since these complications can result in loss
of extracorporeal support, reinforced cannulas are generally 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 drainage or for reinfusion (referred to as
venous and arterial cannulas, respectively). The venous design is characterized by a
greater length (up to approximately 50 cm), greater available diameter (up to 28 Fr),
and a longer distal segment with multiple side holes to facilitate drainage. The
length allows insertion into more central veins such as the superior (SVC) or infe-
rior vena cava (IVC). The arterial 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 ow is not dependent on side holes as is the venous design. An
excess number of side holes can increase the risk of hemolysis in arterial cannulas.
Recently introduced expandable, wire-reinforced cannulas that incorporate a distal
segment of wall-free wire mesh (Smartcanula LLC, Switzerland) are available in some
markets. These cannulas expand to a larger diameter within the vessel to minimize
ow resistance, and the distal mesh maintains vessel patency for improved drainage.
Dual-Lumen Design
Cannulae that incorporate two lumens with two drainage and a single infusion port
are a more recent design that has greatly facilitated the application of venovenous
support for respiratory failure. Although designed for percutaneous insertion, they
can be placed surgically as well. Three fundamental designs are available that have
features to support different needs.
The cavo-atrial design [1, 2] (OriGen, OriGen Biomedical) (Fig. 7.1) is inserted
via the internal jugular vein with the tip positioned in the low right atrium near the
IVC ostium. It has two drainage ports, one distal at the inferior atrium and one
proximately in the superior vena cava, with the reinfusion port in the mid right
atrium directed at the tricuspid valve. The proximity of the distal lumen to the rein-
fusion port allows some recirculation, but effective blood ow remains adequate.
Placement is somewhat easier than the bicaval design since the IVC does not have
to be accessed.
7 Vascular Access for ECLS 135
Fig. 7.1 A dual-lumen venous cannula, designed for drainage from the SVC and inferior right
atrium with reinfusion into the mid right atrium. (Reprinted with permission from OriGen
Biomedical)
The bicaval design (Avalon Elite, Maquet) requires insertion via the internal
jugular vein with the cannula traversing the right atrium and the tip positioned in the
IVC [3, 4] (Fig. 7.2). The drainage lumen extends the length of the cannula with
drainage ports in both the IVC and SVC. The reinfusion lumen is shorter, terminat-
ing in the right atrium with the reinfusion port directed toward the tricuspid valve.
This bicaval drainage design effectively separates the upper and lower venous sys-
tems and results in low recirculation with more effective blood ow.
The third design is similar to a hemodialysis catheter with a single proximal
drainage lumen and a distal reinfusion lumen [5]. This catheter is intended for low-
ow extracorporeal circuits used for carbon dioxide removal (ECCO2R; Chap. 4). If
the cannula ow exceeds the insertion vessel ow, recirculation will limit effective
ow, but placement in the internal jugular or femoral and iliac veins usually assures
adequate blood ow.
Determinants of Cannula Blood Flow
Blood ow through vascular cannulas is driven by the difference between the

pressure at the hub of the cannula and the intravascular pressure at the tip of the can-
nula. Although a cannula is cylindrical in shape, in which the relationship between
ow and pressure is expected to be linear in the presence of laminar ow, the
relationship is only approximately linear over a portion of the ow range (Fig. 7.3).
136 S.A. Conrad
Fig. 7.2 A dual-lumen

venous cannula, designed
for combined drainage
from the IVC and SVC and
reinfusion of blood into the
right atrium for
venovenous extracorporeal
support
Fig. 7.3 Representative pressure-ow relationships for various size single-lumen cannulae. The
graph depicts the nonlinear relationship between ow and pressure due to a combination of lami-
nar and disturbed/turbulent ow resulting from the complex geometry of the catheter
The HagenPoisseulle equation for laminar ow, although not directly applicable
to cannula ow, does illustrate the major determinants of blood ow ( Q ):
DP r 4
Q =
m L
where P is the pressure difference, r and L are the cannula radius and length,
respectively, and is blood viscosity. Maximizing blood ow involves use of the
largest diameter cannulas that can be safely inserted, and keeping the length as short
as possible. The nonlinearity of actual pressure-ow curves most likely comes from
the tip of the cannula, which includes side holes and a tapered tip, causing a depar-
ture from purely laminar ow.
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. Without vessel sizing the
use of too large a cannula can result in venous obstruction, failure to cannulate, or
other complications such as vessel laceration or transection. Too small a cannula
can result in suboptimal blood ow 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 and converting to the French gauge system described by Joseph-
Frdric-Benot Charrire [6] used for sizing cannulas. In the case of vessels with a
circular shape, conversion of vessel diameter in mm to French size is accomplished
with the following simple formula:
Fr = D(mm ) 3
The chosen cannula size should be slightly smaller than the measured vessel to help
assure successful placement and prevent complete obstruction of blood ow.
Infection Control
Infection is not an uncommon risk during prolonged extracorporeal support [7].

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 man-
ufacturers 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.
Peri-procedural prophylaxis with intravenous antibiotics can be considered for
patients who are not receiving antibiotics, and with choice of antibiotic and schedule
138 S.A. Conrad
provided according to the institutions guidelines. Continuation of prophylactic

antibiotics for the duration of ECLS support, other than required for treatment of
underlying infection, is not recommended [8]. 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. Our approach is to perform a 2 min surgical scrub of the site with aqueous
4 % chlorhexidine every 24 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, intervention-
alists, intensivists, and emergency physicians.
Percutaneous
Percutaneous cannulation has been used successfully for venovenous support [9],
and is preferred since it is associated with a lower incidence of cannulation-site
bleeding and infection. It also is non-obstructive, allowing blood ow around the
cannula. It can be used for arterial (other than carotid) as well as venous access.
The same Seldinger technique used for smaller vascular access catheters is used
for ECLS cannulae, but with multiple dilators of no more than 4 Fr difference in size
(typically 12, 16, 20, 24, 28, and 3032 Fr) with the largest size approximately
equal to the size of the cannula to be inserted. Prior to insertion, vessel size is deter-
mined with ultrasound and an appropriate size cannula is chosen. Under adequate
sedation a neuromuscular blocker is administered to prevent respiratory effort.
Under aseptic conditions and following inltration of a local anesthetic, the vessel
is identied with ultrasound and an approach is chosen to avoid injury to neighbor-
ing vessels, since vessels may overlie each other. The access needle is inserted using
ultrasound to guide it into the center of the vessel, and a .035 to .038 guidewire 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 across the atrium.
Following placement of the guidewire, a skin incision is made just large enough
to admit the cannula. The tract is then dilated sequentially to the target size. The
cannula is placed over its tapered loading dilator, and advanced into position. Using
a tubing clamp to control back-bleeding, the guidewire and dilator are removed, and
the cannula ushed with heparinized saline (2 units/mL) to maintain patency until
attached to the ECLS circuit and extracorporeal circulation has begun. The cannula
is sutured to prevent decannulation, taking care to avoid crimping the cannula or
providing a pivot point for cannula kinking.
Percutaneous cannulation of the femoral artery may result in inadequate distal
perfusion and the development lower limb ischemia. This can be managed by per-
cutaneous placement of a retrograde arterial cannula (68 Fr), or surgical cannula-
tion of the posterior tibial artery to assure adequate perfusion of the limb.
Semi-open
A variation of percutaneous cannulation preferred by some surgeons is a technique

which combines percutaneous skin and vessel insertion under direct visualization
through an incision over the vessel entry point. Following sedation and neuromus-
cular blockade, skin preparation and anesthetic inltration, an incision is made over
the expected vessel entry point, and dissection is carried out to visually expose the
vessel. A needle puncture is made distally to the incision, and a tract is created as
for percutaneous insertion. Access into the vessel is performed visually. The subcu-
taneous tissue and skin are closed. Vessel ligation and incision are avoided, and the
skin can be closed without the cannula exiting through the incision, reducing bleed-
ing complications.
Open Surgical
The preferred technique for cervical cannulation when carotid arterial access is
required is the open surgical technique (Fig. 7.4). Following sedation, neuromus-
cular blockade, and skin preparation, an incision is made perpendicular to the axis
of the vessel, and carried down to expose the cervical vessels. The cannula can be
sized by visual comparison with vessel size. The vessels are freed from surround-
ing tissue, and ligatures are placed proximal and distal to control bleeding. An
arteriotomy (or 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 cannula. Following insertion to the proper depth, the ligatures are secured,
typically with pledgets to prevent vessel injury, as vessel repair may be performed
at decannulation. The subcutaneous tissue and skin are closed around the cannu-
lae, taking care to securely close the skin around the cannulae. The above descrip-
tion is generic, and variations are numerous, subject to the surgeons preferences
and experience.
If open cannulation is performed on the femoral artery which, unlike the carotid,
has no collateral circulation, then a smaller retrograde perfusion catheter is placed
140 S.A. Conrad
Fig. 7.4 Technique of surgical cannulation of the cervical vessels for extracorporeal life support.
The technique for the femoral vessels is similar. (Used with permission from [11])
to prevent limb ischemia. An alternative approach to arterial cannulation for the

femoral or subclavian artery is to attach an end-to-side vascular graft to the artery,
and cannulate the graft. This allows use of a large cannula for optimal blood ow
and avoids obstruction and distal ischemia. The graft approach may also be more
suitable for long-term venoarterial support.
Cannulation Configurations
The foremost decision regarding vascular access is the mode of support, which dic-
tates the cannulation conguration. Extracorporeal life support for both respiratory
and cardiac failure was historically performed using only a venoarterial (VA)
conguration. While still the preferred conguration for cardiac failure, other con-
gurations have been developed that are more suitable for other types of support.
Venoarterial
The venoarterial conguration drains blood from the central venous circulation and
returns it to the arterial circulation. The cervical approach is used in neonates and
infants, in whom femoral vessels are small, and since they have adequate collateral
circulation of the cerebral vessels following ligation of the carotid artery. The arte-
rial cannula is placed into the carotid artery and advanced to the proximal innomi-
nate artery. The venous cannula is placed into the internal jugular vein and advanced
into the right atrium. This conguration supplies oxygenated blood to the proximal
aorta, but coronary and right upper extremity blood may be poorly saturated if pul-
monary failure is present and the left ventricle is ejecting (Chap. 6).
Venoarterial cannulation may be performed using the femoral vessels. This con-
guration is 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
Venovenous cannulation was introduced later than venoarterial, and is suitable for
respiratory failure with adequate native cardiovascular function (Chap. 6). It provides
oxygenated blood into the venous system and uses the native heart for oxygen deliv-
ery, making oxygenated blood available to all tissues, including the myocardium.
Cannulae for venovenous support may be placed percutaneously or surgically.
The venovenous conguration was introduced to extracorporeal support using
two single-lumen cannulae, one placed into the femoral vein and advanced to the
intrahepatic inferior vena cava for drainage, and the second placed into the internal
jugular vein and advanced to the superior cavo-atrial junction. An alternative con-
guration to gain better ow and reduce recirculation was to introduce three can-
nulae, two for drainage placed at the superior cavo-atrial junction and distal IVC
respectively, and one for return placed near the inferior cavo-atrial junction.
A major advance in venovenous support was the introduction of the dual-lumen
venovenous cannula. Developed initially for neonates and infants, cannula are now
available for adult and pediatric patients. These cannulae have a single shaft, incor-
porating a drainage lumen with ports in the SVC and IVC (or low right atrium) and
a reinfusion lumen with a port in the mid-right atrium. Recirculation rates with
these cannulae are lower than with the single-lumen congurations, and are negli-
gible with the bicaval design.
142 S.A. Conrad
Veno-arterio-venous
A variation of the venovenous technique is a veno-arterio-venous (VAV) hybrid

mode, which drains from the venous system and returns to both the venous and arte-
rial systems (Chap. 6). This conguration can provide partial cardiac support as
well as oxygenation, and is suitable for patients with respiratory failure who have a
sustained reduction in cardiac function, or cardiac failure who develop respiratory
failure, such as pulmonary edema.
Low-Flow Venovenous
Venovenous support can target carbon dioxide removal (extracorporeal carbon

dioxide removal, ECCO2R) to support lung-protective ventilation in patients for
whom oxygenation can be adequately provided through mechanical ventilation
(Chaps. 4 and 6). A venovenous conguration using smaller single-lumen cannulae
or a dual-lumen cannula with low blood ow (11.5 L/min) can effectively provide
signicant CO2 removal. Commercial systems are emerging which use an integrated
pump and oxygenator and 1516 Fr dual-lumen catheter, similar in design to a
hemodialysis catheter, placed in the jugular or femoral vein.
Arteriovenous
Another approach to extracorporeal carbon dioxide removal is arteriovenous carbon

dioxide removal (AVCO2R), sometimes termed interventional lung assist (iLA).
This conguration involves cannulation of the femoral vessels with a small arterial
cannula (1214 Fr), and a 1618 Fr venous cannula, attached to an oxygenator
using short tubing. The patients arterial blood pressure provides the gradient for
blood ow, avoiding the need for a pump. The major disadvantage is the need for
arterial access, but with smaller cannulae the risk of arterial complications is low. It
is likely that the new generation of dedicated ECCO2R devices will replace the arte-
riovenous conguration, just as continuous venovenous hemoltration (CVVH) has
largely replaced continuous arteriovenous hemoltration (CAVH).
Transthoracic
Although much more invasive, direct cannulation of the right atrium and aortic root
through a sternotomy remains an important approach to vascular access. The most
common use is support of post-cardiotomy failure to wean from cardiopulmonary
bypass (CPB), in which the cardiopulmonary circuit is replaced with the ECLS
circuit. Typically the sternum is left open and draped. The CPB cannulae are large
and support more ow than can be achieved using peripheral access.
The transthoracic approach is associated with more bleeding and infection risk,
so is generally used for patients expected to recover quickly. If prolonged support is
required, the patient may be transitioned to peripheral cannulation, or to a ventricu-
lar assist device. This approach has also been used to provide high-ow support in
patients with severe sepsis [10].
Decannulation
When extracorporeal support is no longer required, the patient is removed from sup-
port by clamping the circuit near the cannulas and removing the circuit.
Anticoagulation is discontinued prior to surgical decannulation or arterial percuta-
neous decannulation, and is held shortly before percutaneous venous decannulation.
The cannulae are ushed to prevent thrombus formation.
Percutaneous venous cannulae are removed by rst placing a purse string suture
in the incision, withdrawing the cannula, and securing the suture. Percutaneous arte-
rial cannulae are removed by withdrawing the cannula and applying pressure until
hemostasis, with care not to fully compress the artery. Arterial puncture closure
devices may be used if appropriately sized. Venous cannulae placed by the semi-
open technique are removed as if placed percutaneously. Short-term anticoagulation
or anti-platelet therapy is used to help prevent venous thrombus formation.
Surgically placed cannulae are removed with an open technique. The skin inci-
sion is re-opened, temporary ligatures placed, and the existing ligatures removed.
The vessel is either repaired or ligated, and the incision closed.
Complications of Cannulation
Recirculation
Recirculation occurs when reinfused blood is aspirated into the drainage cannula,
reducing the effective extracorporeal ow. It is unavoidable with the use of single-
lumen cannulae. Recirculation manifests as a decrease in delivered oxygen and drop
in systemic arterial saturation. Increases in recirculation can occur with displace-
ment of the cannula, and may require radiography to detect. It also increases with
increasing ow, such that high ows may actually reduce delivered oxygen.
Recirculation is less extensive with dual-lumen cannulae. The bicaval design is
associated with the lowest degree of recirculation, often under 3 %. The cavo-atrial
cannula has higher recirculation than the bicaval, but less than the use of two-site
single-lumen cannulation.
144 S.A. Conrad
Insertion Site Bleeding
Bleeding from cannulation sites is the most common bleeding complication. In

most instances it is minimal but can require intervention. Initial approach to man-
agement is to verify appropriate levels of anticoagulation, adequate platelet counts,
normal prothrombin time, and adequate brinogen levels. Reduction of anticoagu-
lation target and the application of topical hemostatic agents may be helpful.
Bleeding can be minimized by limiting the skin incision to snugly t the cannula
when it is inserted. In the case of surgical cannulation, failure of more conservative
measures may require re-exploration of the cannulation site.
Limb Ischemia
Ischemia of the lower limb is one of the major risks associated with cannulation of
the femoral artery. It can usually be managed by placement of a retrograde cannula
either just distal to the cannulation site or in the posterior tibial artery, to provide at
least 150200 mL of blood ow per minute. An alternative percutaneous strategy is
to place two smaller arterial cannulae, one in each femoral, together providing the
total ow of a single larger cannula.
If limb ischemia is not detected in time, sufcient muscle necrosis can occur that
may require fasciotomy or even amputation. Careful clinical examination, Doppler
monitoring of distal pulses, and plethysmographic assessment with pulse oximetry
can help identify this condition early. Many will routinely place a retrograde can-
nula at the time of cannulation to minimize this risk.
Vascular Injury
Injury to the target or adjacent vessel during cannulation can result in inability to
achieve vascular access, transection of a vessel, hemorrhage into areas such as the
retroperitoneal space, exsanguination, and death. Immediate attempts at surgical
repair and completion of cannulation are required, but may not be successful. The
risk is higher with percutaneous cannulation since the vessels are not visible. The
use of ultrasound can mitigate these risks, by allowing for appropriate cannula size,
identication of adjacent vessels, selection of an approach, and guidance of the
puncture to ensure proper entry into the vessel.
Inadequate Flow
The inability to achieve the expected ow can result in the inability to achieve
adequate cardiac support (venoarterial) or persistent hypoxemia and inability to
achieve lung protective settings (venovenous), decreasing the chance of survival.
Three conditions that commonly lead to inadequate ow are placement of cannula

smaller than required, improper placement resulting in impaired venous drainage,
and hypovolemia.
Choice of cannula size should be driven by the ow needed for adequate support,
typically 5075 mL/kg/min for an adult, and higher for pediatric patients, and can
usually be achieved with a single drainage cannula. Uncommonly, two drainage
cannulae may need to be placed. The reinfusion cannula is typically smaller than the
drainage cannula since ow is driven with a much higher pressure gradient, and is
rarely the cause of inadequate ow. Improper placement can be identied by radi-
ography or echocardiography, and corrected.
Hypovolemia is the most common transient cause of inadequate ow. It can
result in chattering of the venous line, in which vascular structures cyclically col-
lapse around the cannula resulting in intermittent ow. Volume expansion with col-
loid (or blood if anemia is also present) resolves the problem.
Infectious Complications
Infection of the cannulation insertion site is a challenging problem, since the patient
may be totally dependent on extracorporeal support and recannulation may be risky
or impossible. Prevention by good skin asepsis at the time of insertion and during
extracorporeal support is important to minimizing this risk. If infection does develop
and appears to be localized, then use of appropriate antibiotics may be successful.
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 may be
required.
References
1. Andrews AF, Zwischenberger JB, Cilley RE, Drake KL. Venovenous extracorporeal mem-
brane oxygenation (ECMO) using a double-lumen cannula. Artif Organs. 1987;11(3):2658.
2. Anderson 3rd HL, Otsu T, Chapman RA, Barlett RH. Venovenous extracorporeal life support
in neonates using a double lumen catheter. ASAIO Trans. 1989;35(3):6503.
3. Wang D, Zhou X, Liu X, Sidor B, Lynch J, Zwischenberger JB. Wang-Zwische double lumen
cannula-toward a percutaneous and ambulatory paracorporeal articial lung. ASAIO
J. 2008;54(6):60611.
4. 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):9915.
5. 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):13827.
146 S.A. Conrad
6. Iserson KV. J.-F.-B. Charriere: the man behind the French gauge. J Emerg Med.
1987;5(6):5458.
7. 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):27781.
8. Extracorporeal Life Support Organization Task Force on Infections. Infection control and
extracorporeal life support 2010. http://elso.org/downloads/resources/committees/infectious-
disease-and-antibiotic/Infection-Control-and-Extracorporeal-Life-Support.pdf.
9. Pranikoff T, Hirschl RB, Remenapp R, Swaniker F, Bartlett RH. Venovenous extracorporeal
life support via percutaneous cannulation in 94 patients. Chest. 1999;115(3):81822.
2007;8(5):44751.
11. Field ML, Al-Alao B, Mediratta N, Sosnowski A. Open and closed chest extrathoracic can-
nulation for cardiopulmonary bypass and extracorporeal life support: methods, indications,
and outcomes. Postgrad Med J. 2006;82(967):32331.
Chapter 8
Circuits, Membranes, and Pumps
Bradley H. Rosen
Introduction
Modern ECLS is based on highly efcient, low-resistance, gas-exchanging mem-

branes. In order to couple the patient and articial lung, vascular access is required
(see Chap. 7), 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, rec-
ognize when something goes awry, and know how to intervene. This chapter
describes the components of the circuit, providing the practitioner with an under-
standing of how they function and interact. It is divided into two large sections: the
rst describes the anatomy of the overall ECLS circuit; the second the physiology
and normal operation of each of the components.
Circuit Anatomy
Overall Circuit Considerations
Circuit designs all attempt to balance efcacy, safety, convenience, and simplicity.
There is no one-size-ts-all solution, however, since varied patients, circumstances,
and clinician preferences may necessitate that safety override simplicity or that por-
tability trump efcacy. For example, inserting multiple stopcocks into the circuit
B.H. Rosen, DO (*)

Division of Pulmonary, Critical Care, and Occupational Medicine, Department of Internal
Medicine, Carver College of Medicine, University of Iowa Hospitals and Clinics,
200 Hawkins Drive, Iowa City, IA 52242, USA
e-mail: bradley-rosen@uiowa.edu

148 B.H. Rosen
Fig. 8.1 Schematic of a complex circuit design that implements all optional features, including a
compliance chamber, separate non-integrated blood analyzers, a manifold with access sites (closed
unless being accessed), and a bridge
Fig. 8.2 Similar to the previous gure, but simplied with only necessary components: pump,
oxygenator, and ow probe/bubble sensor. The blood analyzers are internalized within the pump
and oxygenator
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 physicians experience and preference with regard to
circuit complexity. Thus, circuits range from rather complex designs incorporating
many safety and monitoring functions (see Fig. 8.1) to minimalistic, simpler layouts
that lack the various bells and whistles (Fig. 8.2).
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 modication
produces a weak point susceptible to rupture or brin 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,
between the pump and the oxygenator. This is intentional: the lack of connectors on
the arterial side reduces the potential for accidental exsanguination, while the simi-
lar lack of connectors proximal to the pump inlet limits the risk of air entrainment
and gas embolism.
8 Circuits, Membranes, and Pumps 149
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 ow (proportional to length according to Poiseuilles
law), necessitating higher circuit pressures and leading to more damage to blood
elements.
Circuit length and complexity also relate to the degree to which blood is exposed
to plastic surfaces and this interaction elicits an inammatory response. It is believed
that induced inammation may further compromise lung function, leading to fur-
ther gas exchange deterioration. It has been hypothesized that heparin coating of
polymethylpentene (PMP) oxygenators serves to reduce this response [1]. An added
consequence of circuit-induced inammation is excessive brinogen production
leading to increased brin deposition on circuit surfaces. Further, this inammation
promotes platelet adherence, elevating the risk of thrombosis which impairs oxy-
genator function [2]. Cellular deposition along the membrane surface (on the blood
side) correlates with a rising resistive pressure across the device [3]. Some brin
deposition within the oxygenator and circuit is unavoidable (apparent rst on the
venous side), but excessive deposition is deleterious to circuit function.
There are several additional implications of circuit length. The greater the surface
area, the more that medications commonly used in the care of critically ill patients
are subject to adsorption. Antibiotics (meropenem, cefazolin, and vancomycin),
sedatives (midazolam), and analgesics (morphine, fentanyl, and acetaminophen) are
all meaningfully adsorbed, to a degree related to the lipophilic nature of the drug.
Antibiotics are only moderately affected (6585 % 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 signicantly to the volume required to
prime the circuit, producing hemodilution. Finally, tubing surface area also relates to
the degree of heat loss. This can be substantial, such that ECLS circuits must incor-
porate a means for temperature control (see Heat Exchanger below).
Circuit Priming
Priming refers to the process by which the gas (ambient air present at manufacture)
is replaced with a physiologically compatible uid. For ECLS in adults, the circuit
is primed with crystalloid uids, such as normal saline, Ringers lactate, or propri-
etary mixed electrolyte solutions (e.g., Plasma-Lyte or Normosol). Purchased cir-
cuits come attached to a large, empty priming bag. The bag is lled with sterile
crystalloid, clamps are opened to the venous and arterial limbs, and the priming bag
is raised to allow gravity to move the uid 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, bridge, manifold, and tubing) and directly relates to the degree of
150 B.H. Rosen
hemodilution that follows. For pediatric and neonatal ECLS, hemodilution is pre-
vented by priming the circuit with blood, but this is not necessary for adults where
the typical priming volume averages 7501000 mL for a complex circuit design and
for a simple 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 regard-
ing shelf life. Priming with a colloid may shorten the shelf life of a primed circuit,
another reason many programs choose a crystalloid prime. A simplied 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 rst with carbon dioxide
(to displace oxygen and nitrogen), hastening the subsequent uid 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 rushes 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 8.1 represents a comprehensive schematic of
a circuit, whereas Fig. 8.2 shows a greatly simplied design with few extraneous
components. In each instance, we describe the circuit beginning with the outow
(venous) cannula, proceeding through the gas-exchanging membrane, and ending
back at the patient through the inow cannula, which may enter an artery (venoarte-
rial or VA ECLS) or vein (venovenous or VV ECLS). Sometimes used for venove-
nous ECLS, dual-lumen cannulas allow blood to exit and enter at the same site but,
for illustration purposes, we have separated these in the gures.
Starting with the outow 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
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. 10). The conduct-
ing tubing should be kept relatively short in order to reduce resistance to blood ow,
surface area of contact with blood, and the priming volume. The conducting tubing
leads to a centrifugal pump (in some designs a compliance chamber or bladder pre-
cedes the pump as in Fig. 8.1) 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 oxygen), being joined to wall oxygen or an E-cylinder through a ow
meter. The newly arterialized blood completes its extracorporeal course through the
inow cannula, delivering oxygenated and warmed blood to the vascular system.
Outow and inow cannulas may be bridged directly by a length of large-diameter
tubing (the bridge; see Fig. 8.1) connected through two high-ow stopcocks.
When opened, the bridge offers a shunt to keep blood owing within the circuit
while clamps are used to isolate the patient. In so doing, the clinician can judge
whether the patient has recovered sufciently to sustain respiration and circulation
without ECLS (see Chap. 13). This is important during VA ECLS, but a bridge is not
needed for VV support since weaning can be conducted by reducing or eliminating
gas ow to the membrane while leaving circuit ow to the patient undisturbed.
During most ECLS operation the bridge remains closed and, because any blood
within is stagnant, blood is generally displaced by saline when the bridge is closed.
In order to monitor circuit function and to prevent complications, devices to mea-
sure pressure and ow and to detect bubbles are included, and information is relayed
to a console. Typically, pressure is measured on the venous side of the pump (P1),
providing information about how much suction is required to draw the needed circuit
blood ow. Two additional pressure transducers (P2 and P3) ank the oxygenator
so that its resistance can be estimated based on the drop in pressure across the
membrane(delta-P). In addition, P3 displays the pressure that drives ow back to
the patient. Circuit blood ow is monitored using an ultrasonic ow probe, since
centrifugal pumps do not guarantee a xed relationship between revolutions per min-
ute and volume displaced, as was true for roller pumps. The ow probe may be inte-
grated 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 ow
measurement and bubble detection. Bubbles distal to the oxygenator can produce
systemic embolism and are especially dangerous in VA modes.
Spectrophotometric sensors allow real-time measurement of such values as PO2,
PCO2, pH, SaO2, SvO2, and hemoglobin concentration, among others. These sensors
must be calibrated periodically by comparing the displayed value against a blood sam-
ple analyzed simultaneously using conventional laboratory methods. The console
receives data from various devices along the circuit, displaying pump speed, ow, pres-
sures, temperature, and other physiological information. The console also may display
alarm notications 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 lock
connections with a three-way stopcock controlling ow to each. The manifold
derives from the region 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 oxygenator. Various infusions of medica-
tions and anticoagulant agents may be connected to the circuit through these ports,
but more often, other vascular access is utilized (see Chaps. 7 and 10). Sufcient
ow 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.
152 B.H. Rosen
Function of the Circuit Components
Cannulas and Tubing
Single- and double-lumen cannulas are described more fully in Chap. 7. 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 rec-
ognize blood color (as a clue to circuit function and recirculation) and to identify
brin, 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 ow.
Compliance Chamber
This device was previously employed when roller head pumps were more common,
as a safety device to dampen any excessive negative pressure generated by the
pump, rather than causing cavitation or hemolysis in the patient. Essentially an
external venous reservoir, this device may also provide information regarding rela-
tive hypovolemia. Collapsing of the compliance chamber would suggest to the clini-
cian that ow through the circuit be slowed or additional uid volume be
administered. With broad use of centrifugal pumps, compliance chambers are gen-
erally felt to be an unnecessary complexity.
The simplest design is a silicone bladder with inlet and outlet ports, placed
between the outow cannula and the pump (Fig. 8.1). A pressure transducer can be
used to signal an alarm or to slow or shut off the pump if negative pressure reaches
a degree that could result in cavitation within the venous system. Various designs
have different port sizes, priming volumes, and orientations. One device is a verti-
cally oriented inline reservoir consisting of a compliance balloon housed within a
PVC chamber. This obviates placing the device on the ground and the lengths of
tubing to the bladder and from it. The vertical orientation may also allow for a more
constant ow assisted by gravity, reducing likelihood of settling blood and throm-
bosis, as well as entrainment of gaseous bubbles at the top of the chamber.
Additionally, this device can be heparin-coated and is FDA-approved for use in
ECLS. It is offered in a standard size with 20 mL priming volume and -in. ports,
as well as a larger version with 115 mL priming volume and 3/8-in. ports [5].
Pumps
There are two types of pumps employed in ECLS circuits: roller/occlusive and cen-
trifugal pumps. Practitioners of modern ECLS have settled on the centrifugal pump
design as the safer of the two.
Roller Head (Occlusive) Pumps: The pump itself has two roller heads within an
enclosure and one head contacts (and variably occludes) the tubing at all times.
These roller heads must be properly adjusted to accurately account for blood ow
[6]. They are placed 180 to each other, to pull blood from the venous limb and
simultaneously push it forward along the circuit path. The length of tubing within
the enclosure is termed the raceway, and that section is advanced or withdrawn
periodically to avoid foci of excessive wear on the tubing. This necessitates a signi-
cantly longer circuit when a roller pump is employed, and accordingly a larger prim-
ing volume. Another limitation that roller pumps impose on circuit design is that
they must form the base of the circuit, requiring more extensive lengths of tubing.
Outside of restricting circuit design, roller head pumps have intrinsic limitations
arising from the fact that the device operates on the principle of positive uid dis-
placement. This allows it to generate excessive positive pressures if, for example,
there is a kink in the outow tubing, or very negative suction (due to intravascular
hypovolemia). It was this attribute that mandated compliance chambers when roller
head pumps were more commonly used. A high positive pressure was required to
drive blood through older silicone rubber membrane lungs, but newer, low-resistance
membranes make this unnecessary.
Flow is calculated based upon the length and diameter of the tubing within the
raceway, combined with the number of pump revolutions. Higher ow (more revo-
lutions) leads to increased wear on tubing and can cause the liberation of micro-
scopic particles of tubing due to material fatigue, termed spallation, and resulting
microembolism to the patient [7]. Specic types of tubing are recommended for use
within the raceway due to their resistance to wear compared with other tubing [8].
In case of emergency events, the pump may be operated manually by a hand-
crank and newer models are equipped with an internal battery backup of limited
duration. Advantages of roller pumps include lower cost; lower direct pump prime
volume (although more tubing is necessary); and afterload-independent ow.
Disadvantages include the need for longer tubing; location at the base of the circuit
(in current designs manufactured); spallation and microembolism; and challenges
in properly setting the occlusion.
Centrifugal pumps: These devices increasingly employ a magnetically driven
impeller within a spiral housing. The impeller 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 ow toward the circumfer-
ence 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 ow is independent of afterload (to the point of causing tubing rupture!),
a centrifugal pump is unable to overcome excessive afterload. Instead, ow 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 outow cannula, as it
is unable to generate sufciently negative pressure. Of course, this means that hypo-
volemia tends to threaten the adequacy of circuit ow when a centrifugal pump is
used. This lack of absolute relationship between pump revolutions and blood ow
necessitates a ow meter. An increasing discrepancy between pump speed and mea-
sured ow is a strong signal of trouble with regard to function of the circuit.
154 B.H. Rosen
Should there be a complete loss of power, these pumps can also be hand-cranked.
Internal batteries provide up to 90 min of support without an external power source.
Advantages of centrifugal pumps include reduced tubing length; safety benets
due to absent issues of spallation, microemboli, and raceway rupture; and greater ex-
ibility in circuit design because the pump does not need to be at the base of the unit.
Disadvantages include direct priming volume required for the pump (although
tubing length is reduced) and blood ow that depends on preload and afterload
(requiring a blood ow probe). Centrifugal pumps were initially reported to produce
unacceptable rates of hemolysis due to heat generation, but engineering improve-
ments have solved this problem [9]. Designs have evolved to reduce hemolysis and
the risk of gaseous microembolism [10, 11], changing the landscape of ECLS cir-
cuits and leading to worldwide adoption of the technology [12, 13].
Membrane Oxygenators
The ideal membrane lung would be highly permeable to relevant gases (O2 and
CO2) while resisting uid transudation from the blood to the gas phase (termed
plasma leak). Blood should ow through the device with minimal resistance,
allowing high ows 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 (see Chap. 14).
Both PMP and polypropylene hollow-ber membranes employ large numbers of
ne capillary tubes to carry the sweep gas while being bathed by owing blood.
This extra-capillary blood ows countercurrent to the direction of gas movement,
increasing the efciency 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 2 m2 of gas exchange surface while the
lung exposes upwards of 70 m2 to blood ow. Additionally, in the best membrane
lungs, oxygen must diffuse 150 m from sweep gas to blood, whereas the compa-
rable distance within the lung is a mere 0.5 m. These disadvantages are offset by
increasing the effective dwell time of the blood within the articial lung. In addi-
tion, so called secondary ows, which describe the mixing of blood around the
gasuid interface due to purposefully created turbulence, further enhances gas
transfer. Blood cells are regularly being brought into close approximation with the
gas-lled capillaries, effectively reducing diffusing distance [14, 15]. This layout
allows for up to a two and a half-fold reduction in surface area necessary for gas
exchange [16]. Typical gas-exchanging capacities for membrane lungs are shown in
Figs. 8.3 (oxygen) and 8.4 (carbon dioxide). Providing sufcient oxygen transfer to
Fig. 8.3 Oxygen transfer

in mL/min as a function of
blood ow in L/min
Fig. 8.4 Carbon dioxide transfer in mL/min as functions of both blood ow in L/min and sweep
gas ow
meet the entire metabolic demand (roughly 250 mL O2/min) requires blood ow of
roughly 4 L/min. In contrast, carbon dioxide transfer is relatively advantaged; so
that much less blood ow is required, especially at very high sweep gas ow rates.
For example, nearly all of the metabolically produced carbon dioxide can be elimi-
nated with only 1 L/min of blood ow, especially at high gas ows [17].
The most common external appearance of PMP oxygenators is an extruded
square evenly balanced on one corner (see Fig. 8.5). At the lower corner, blood
enters the device from the pump under pressure denoted as pre-membrane or
inlet pressure (typical pressures are in the range of 225275 mmHg, certainly less
than 400 mmHg; see Table 8.1). The blood ascends to the opposite corner at the
highest elevation of the device and then ows 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 or outlet pressure) will be
156 B.H. Rosen
Fig. 8.5 Photograph of a

membrane oxygenator
(small adult model).
Identied on this model are
the inlet and outlet for both
blood and sweep gas, as
well as the two connections
for the water heater
Table 8.1 Circuit pressures Location in the circuit Normal operating pressure
Proximal to the pump: P1 100 to 200 mmHg
Oxygenator inlet: P2 225275 mmHg
Oxygenator outlet: P3 190260 mmHg
Delta-P: (P2 minusP3) 1035 mmHg
less than the inlet pressure. The difference between these is called the delta pres-
sure (or delta-P), representing the resistive pressure drop across the membrane at
the current ow. With the current generation of PMP devices, delta-P should range
from the teens to low 30s at typical circuit blood ow rates.
With regard to the pressures and the information that may be gleaned from their
trends, an elevated oxygenator inlet pressure (P2) has variable implications depend-
ing on whether the outlet pressure (P3) is also elevated. In the case where both are
elevated (delta-P is preserved), the circuit should be examined, not the oxygenator:
the distal tubing and circuit may be obstructed by kinks or thrombosis. Similar nd-
ings are seen transiently when patients cough, Valsalva, or are suctioned. When
inlet pressure rises along with an increase in delta-P, resistance within the mem-
brane is excessive, raising concerns for thrombosis, heparin-induced thrombocyto-
penia, or accumulation of brin or cellular elements on the membrane. Ultimately,
this portends failure of the membrane.
Another major advance in modern PMP oxygenators is their low resistance to
blood ow, 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
failure. Additionally, such low resistance to blood ow permits novel applications
of ECLS such as pumpless arteriovenous extracorporeal CO2 removal (ECCO2R)
that rely solely on the difference between arterial and venous blood pressures to
drive ow [18] (see Chap. 4).
In the modern era of ECLS, the centerpiece of the circuit is a PMP oxygenator
and these devices have eclipsed prior generations of articial lung. Nevertheless,
other oxygenator designs are seen occasionally and include both silicone mem-
branes rst developed by Kolobow [19] (most recently marketed as the Medtronic
1-4500-A2) as well as the polypropylene microporous hollow-ber 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 (dened as use for more
than 6 h). However, they require priming with considerable volume (665 mL each
[20]), exhibit a large pressure drop across the membrane that limits the use of
centrifugal pumps, and are relatively inefcient in gas exchange so that at least two
large surface area units are needed per patient. Polypropylene hollow-ber mem-
branes are highly efcient with respect to gas exchange, but tend to develop plasma
leak (described further below). They also present a low resistance to blood ow and
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 ow is increased, greater demands are
placed on the capacity for gas diffusion across the hollow ber barrier. In part, this
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 ows. At sufciently high ows, the
membrane fails to fully saturate the blood. The ow threshold for full oxygenation
is termed the rated ow.
Plasma Leak: Hollow ber membranes should be sufciently permeable to allow
rapid gas diffusion, while remaining impermeable to uid movement. Plasma leak
is the phenomenon whereby plasma phospholipids leak from the circulating whole
blood to the gas compartment of the oxygenator, then serve to propagate further
plasma leakage in a positive feedback cycle [21]. Small amounts of uid will nor-
mally traverse the membrane and a drainage port is provided for egress. Greater
volumes signal a failing membrane, severely impairing the efciency of gas
exchange, accompanied by a rise in delta-P, and eventually requiring exchange of
the device. Important plasma leak can be conrmed by analysis of liquid from the
gas outlet for proteins [22]. Additional implications are a signicant loss of proteins
158 B.H. Rosen
and immunoglobulins, potentially signicant from immunologic and nutritional

standpoints. PMP membranes are much less susceptible to plasma leak than prior
articial lungs, yet they are not fully immune to this complication [23].
Circuit Surface Coatings
The bloodcircuit interface activates neutrophils, produces inammation, and trig-

gers coagulation [24]. Changing from the silicon 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
inammatory response as measured by terminal complement complex formation
and neutrophil elastase concentrations [25]. Complement and platelet activation is
limited as well [25, 26]. The effect in patients seems limited to a reduction in anti-
coagulation requirements [27], reduced plasma-free hemoglobin, and lower tumor
necrosis factor- concentrations [28].
Although it is presumed that there is benet to the practice of heparin-bonding,
the data do not consistently support benet when patient-centered outcomes are the
primary outcome. Coating could be benecial in those who have an elevated pretest
probability or risk of hemorrhage, but overall there is controversy regarding the
value of heparin coatings [29, 30].
Although heparin-induced thrombocytopenia (HIT) is not a common occurrence
in critical care with an estimated incidence of 0.30.6 % [31], there are instances
reported in patients being supported by ECLS [32]. Moreover, the diagnosis of HIT
is fraught because thrombocytopenia is so common during ECLS for myriad rea-
sons. When HIT is conrmed (such as through serotonin-release assay), infused
heparin should be discontinued while anticoagulation is maintained using direct
thrombin inhibitors such as argatroban [33] or bivalirudin [29]. However, the
clinician should be mindful that circuit components are heparin-bonded and that
there is anecdotal evidence that this can be clinically signicant [34]. Manufacturers
generally do have devices available without these biologic coatings for patients with
conrmed HIT.
Heat Exchanger and Heater-Cooler
While on ECLS, the patients blood is exposed to the environment of the outside
world, risking substantial heat loss. Left unchecked, this could produce coagulopa-
thy, 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 inammation or clotting. The blood and water phases are directed
in a countercurrent fashion to maximize efciency of heat transfer. Larger heat
exchangers are more effective in regulating temperature, but require larger priming
volumes, contributing 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 ow is sufciently high, the heat exchanger is highly effec-
tive in maintaining body temperature. Thus, it can be used to prevent hyperthermia
in patients following cardiac arrest who are undergoing targeted temperature man-
agement. At the same time, the clinician must recognize that fever will be masked,
so other indicators of infection must be sought.
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.
Acknowledgment The author thanks the following content consultants: 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 radio-
graphic signs of pulmonary inammation during ECMO between silicon and poly-methyl pen-
tene oxygenators. Perfusion. 2007;22:1522.
2. Peek GJ, Firmin RK. The inammatory and coagulative response to prolonged extracorporeal
membrane oxygenation. ASAIO J. 1999;45:25063.
3. Lehle K, Philipp A, Gleich O, Holzamer A, Muller T, Bein T, et al. Efciency in extracorporeal
membrane oxygenation-cellular deposits on polymethylpentene membranes increase resis-
tance to blood ow and reduce gas exchange capacity. ASAIO J. 2008;54:6127.
4. Wildschut ED, Ahsman MJ, Allegaert K, Mathot RAA, Tibboel D. Determinants of drug
absorption in different ECMO circuits. Intensive Care Med. 2010;36:210916.
5. Tamari Y, Lee-Sensiba K, King S, Hall MH. An improved bladder for pump control during
ECMO procedures. J Extra Corpor Technol. 1999;31:8490.
6. Snyder EJ, Harb HM, Cullen JA, McElwee DL. Setting roller pump occlusion with the
Transonic HT109 owmeter. ASAIO J. 1997;43:604.
7. Peek GJ, Thompson A, Killer HM, Firmin RK. Spallation performance of extracorporeal
membrane oxygenation tubing. Perfusion. 2000;15:45766.
160 B.H. Rosen
8. Peek GJ, Wong K, Morrison C, Killer HM, Firmin RK. Tubing failure during prolonged roller
pump use: a laboratory study. Perfusion. 1999;14:44352.
9. Lawson DS, Ing R, Cheifetz IM, Walczak R, Craig D, Schulman S, et al. Hemolytic character-
istics of three commercially available centrifugal blood pumps. Pediatr Crit Care Med.
2005;6:5737.
10. Burnside J, Gomez D, Preston TJ, Olshove VF, Phillips A. In-vitro quantication of gaseous
microemboli in two extracorporeal life support circuits. J Extra Corpor Technol.
2011;43:1239.
11. Yee S, Qiu F, Su X, Rider A, Kunselman AR, Guan Y, et al. Evaluation of HL-20 roller pump
and Rotaow centrifugal pump on perfusion quality and gaseous microemboli delivery. Artif
Organs. 2010;34:93743.
12. Lawson DS, Lawson AF, Walczak R, McRobb C, McDermott P, Shearer IR, et al. North
American neonatal extracorporeal membrane oxygenation (ECMO) devices and team roles:
2008 survey results of Extracorporeal Life Support Organization (ELSO) centers. J Extra
Corpor Technol. 2008;40:16674.
13. Lawson DS, Walczak R, Lawson AF, Shearer IR, Ing R, Schulman S, et al. North American
neonatal extracorporeal membrane oxygenation (ECMO) devices: 2002 survey results. J Extra
Corpor Technol. 2004;36:1621.
14. Drinker PA, Bartlett RH, Rishon MB, Noyes BS. Augmentation of membrane gas transfer by
induced secondary ows. Surgery. 1969;66:77581.
15. Diller TE, Mikic BB, Drinker PA. Shear-induced augmentation of oxygen transfer in blood. J
Biomech Eng. 1980;102:6772.
16. Gaylor JDS. Membrane oxygenators: current developments in design and application. J
Biomed Eng. 1988;10:5417.
17. Mueller T, Lubnow M, Philipp A, Bein T, Jeron A, Luchner A, et al. Extracorporeal pumpless
interventional lung assist in clinical practice: determinants of efcacy. Eur Respir
J. 2009;33:5519.
18. Johnson P, Frohlich S, Westbrook A. Use of extracorporeal membrane lung assist device
(Novalung) in H1N1 patients. J Card Surg. 2011;26:44952.
19. Kolobow T, Bowman RL. Construction and evaluation of an alveolar membrane articial
heart-lung. Trans Am Soc Artif Intern Organs. 1963;9:23843.
20. Khoshbin E, Roberts N, Harvey C, Machin D, Killer H, Peek GJ, et al. Poly-methyl pentene
oxygenators have improved gas exchange capability and reduced transfusion requirements in
adult extracorporeal membrane oxygenation. ASAIO J. 2005;51:2817.
21. Montoya JP, Shanley CJ, Merz SI, Bartlett RH. Plasma leakage through microporous mem-
branes: role of phospholipids. ASAIO J. 1992;38:M399405.
22. Eash HJ, Jones HM, Hattler BG, Federspiel WJ. Evaluation of plasma resistant hollow ber
membranes for articial lungs. ASAIO J. 2004;50:4917.
23. Puis L, Ampe L, Hertleer R. Case report: plasma leakage in a polymethylpentene oxygenator
during extracorporeal life support. Perfusion. 2009;24:512.
24. Fosse E, Moen O, Johnson E, Semb G, Brockmeier V, Mollnes TE, et al. Reduced complement
and granulocyte activation with heparin-coated cardiopulmonary bypass. Ann Thorac Surg.
1994;58:4727.
25. Jansen PGM, Velthius H, Huybregts RAJM, Paulus R, Bulder ER, van der Spoel HI, et al.
Reduced complement activation and improved postoperative performance after CPB with
heparin-coated circuits. J Thorac Cardiovasc Surg. 1995;110:82934.
26. Pekna M, Hagman L, Halden E, Nilsson UR, Nilsson B, Thelin S. Complement activation dur-
ing cardiopulmonary bypass: effects of immobilized heparin. Ann Thorac Surg.
1994;58:4214.
27. Gravlee GP. Heparin-coated cardiopulmonary bypass circuits. J Cardiothorac Vasc Anesth.
1994;8:21322.
28. Gu YJ, van Oeveren W, Akkerman C, Boonstra PW, Huyzen RJ, Wildevuur CRH. Heparin-
coated circuits reduce the inammatory response to cardiopulmonary bypass. Ann Thorac
Surg. 1993;55:91722.
29. Ranucci M, Ballotta A, Kandil H, Isgro G, Carlucci C, Baryshnikova E, et al. Bivalrudin-based

versus conventional heparin anticoagulation for postcardiotomy extracorporeal membrane
oxygenation. Crit Care. 2011;15:R275.
30. Silvetti S. Do we need heparin coating for extracorporeal membrane oxygenation? New con-
cepts and controversial positions about coating surfaces of extracorporeal circuits. Artif
Organs. 2015;39:1769.
31. Cuker A. Clinical and laboratory diagnosis of heparin-induced thrombocytopenia: an inte-
grated approach. Semin Thromb Hemost. 2014;40:10614.
32. 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;43:512.
33. Beiderlinden M, Treschan T, Gorlinger K, Peters J. Argatroban in extracorporeal membrane
oxygenation. Artif Organs. 2007;31:4615.
34. 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;24:1357.
Chapter 9
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
Extracorporeal membrane oxygenation (ECMO) rst garnered attention when the

National Institutes of Health (NIH) funded the rst prospective controlled random-
ized clinical study on ARDS in 1974 [1]. In this study, the outcomes of severe acute
respiratory failure patients treated by venoarterial (VA) ECMO were compared with
those 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 identied two possible
reasons for the study failure: (a) the VA mode, that could have increased the inci-
dence of pulmonary vascular thrombosis; and (b) the low tidal volume and respira-
tory rate applied in the ECMO group, which might have caused a worsening of
respiratory compliance. However, tidal volume was reduced only modestly in the
ECMO group (from approximately 800 to 600 mL). Moreover, despite this decrease,
plateau pressures remained quite high, in a range between 40 and 50 cmH2O [2]. At
that time, recommended tidal volume was 1015 mL/kg or more, as it was one of the
major determinants of oxygenation [3, 4]. Taking an opposite position, Ted Kolobow
wrote a provocative editorial: Why did ECMO fail? in which he suggested that
A. Pesenti, MD (*) G. Bellani, MD, PhD

Department of Health Sciences, University of Milano-Bicocca, Monza, Italy
Department of Anesthesia and Critical Care, San Gerardo Hospital and Milano-Bicocca
University, Monza, Italy
e-mail: antonio.pesenti@unimib.it
G. Grasselli, MD
Department of Anesthesia and Critical Care, San Gerardo Hospital and Milano-Bicocca
University, Monza, Italy
T. Mauri, MD
Department of Anesthesia and Critical Care, Fondazione IRCCS Ca Granda Ospedale
Maggiore Policlinico, Milan, Italy

164 A. Pesenti et al.
severely diseased lungs have a chance to heal only if the environment remains con-
ducive 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 hypo-
perfusion with severe and lethal lung tissue alkalosis [5]. He had developed in his
laboratory at the NIH a prototype that became the rst commercially available mem-
brane lung, and he was one of the most experienced scientists in this eld [6]. He
pointed out that providing viable blood gases by extracorporeal means was not
enough to foster lung healing. The approach to ventilation should be reevaluated to
take into account that gas exchange via the native lung was no longer essential. High
airway pressures and high tidal volume, a must to optimize gas exchange through
the native lung, were postulated to cause what later became known as VILI [7].
The goal of ECMO was redened from buying time for the lung to heal [8] to
rest the lung [9]. In the same years, Gattinoni and colleagues introduced a novel
ECLS technique called low-frequency positive-pressure ventilation with extracor-
poreal carbon dioxide removal (LFPPV ECCO2R) [10]. The stage was set for a new
approach to the ventilator management of ECLS patients. Over the next 30 years,
ECLS evolved from pure hypoxemia rescue 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]. Still, how best to ventilate ARDS patients undergoing ECLS is based on
scanty data, derived mainly from observational or physiological studies.
Pathophysiology of Gas Exchange During ECLS
We will discuss almost exclusively the pathophysiology of extracorporeal gas

exchange during venovenous bypass, though we will also mention some important
aspects of venoarterial or arteriovenous bypasses.
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 circuit blood ow, hemoglobin concentration (Hb), and oxyhemo-
globin saturation of the blood entering the membrane lung. For any given intrapul-
monary shunt (proportion of the cardiac output that crosses the lung and leaves it
with an unchanged oxygen content equal to the mixed venous one), VV ECMO will
increase the arterial oxygen content in proportion to how much it can raise the
mixed venous oxygen content. For example, if a patient has a mixed venous oxygen
saturation (SvO2) of 50 % and intrapulmonary shunt fraction of 50 %, then the arte-
rial oxygen saturation (SaO2) will roughly be 75 % (given an arteriovenous differ-
ence of 25 %). If ECMO can oxygenate a substantial (say 50 %) proportion of
venous blood, SvO2 rises to 75 % and SaO2 will be the average of 75 % and 100 %
9 Ventilator Management During ECLS 165
(approx 87.5 %; not taking into account dissolved O2). It becomes obvious that the
higher the ratio (extracorporeal blood ow/cardiac output), the higher the effect on
the mixed venous and therefore the arterial oxygen saturation (this is one of the
reasons to advocate -blockers for some patients undergoing VV ECMO) [12].
Real life is much more complicated than that, mainly because of the negative
effect of bypass recirculation (which increases by increasing blood ow), and
because intrapulmonary shunt is not xed, but rather tends to increase along with
SvO2 as hypoxic vasoconstriction is relieved.
The ability to provide oxygen through ECMO is limited by the fact that venous
blood already contains much oxygen (patients normally have, and probably require,
a mixed venous oxygen saturation around 70 %) and there is only so much hemo-
globin available to bind oxygen. This basic principle [13] shows why oxygenation
depends mainly on circuit blood ow and why this ow must be very similar to the
normal cardiac output.
Carbon Dioxide
In contrast, CO2 removal requires much lower blood ow [13]. This is due to the
very high carbon dioxide content of blood (mostly carried as bicarbonate ion). The
carbon dioxide content of normal venous blood is approximately 600 mL/L (50 %
more in compensated chronic hypercapnia), meaning that 1 L of blood contains two
to three times the entire bodys metabolic CO2 production. Total CO2 removal (i.e.,
the removal of the minute CO2 production) can be achieved at a blood ow in the
range of 0.5 L/min, as opposed to the 5 L/min required to provide the 250 mL/min
of oxygen consumed physiologically.
In the second half of the 1970s, the pioneering work of Kolobow and Gattinoni
provided the pathophysiological foundations to suggest a clinical role for
ECCO2R. This extraordinary pair of scientist-inventor and scientist-clinician real-
ized that the membrane oxygenator was indeed a membrane lung, exchanging both
oxygen and CO2. In a little more than a year they (a) described a membrane lung
optimized for CO2 removal (the so-called CDML: the carbon dioxide membrane
lung) [14]; (b) showed that it was possible to control the ventilation of an awake
sheep by removing CO2 at incremental rates [15], even to complete apnea [16]; and
(c) developed a mode of ventilation that would maintain lung volume while using
very low respiratory rates [17]. They claried that, during apneic oxygenation, alve-
olar PO2 is a function of alveolar PN2, in turn at equilibrium with the PN2 of the gas
ventilating the membrane lung [16]. Another observation dened the role of the
ratio of CO2 eliminated and oxygen consumed by the natural lung (respiratory quo-
tient, RQ) in determining alveolar PO2, of particular importance in the spontane-
ously breathing subject. This stresses the need to raise FiO2 to maintain the alveolar
fraction of oxygen (FaO2) when the alveolar gas equation is faced with RQs much
lower than 1 (as is the case when CO2 is also removed by the membrane lung).
When no CO2 is exchanged through the natural lung (apneic oxygenation), FaO2 can
be maintained constant only when the FiO2 is 1 [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 articial lung allows
dramatically reducing the load imposed on the natural lung by mechanical ventila-
tion and implementing a truly protective ventilatory strategy (low pressure, tidal
volume, and respiratory rate) [19]. These reductions will lower mean airway pres-
sure, potentially causing alveolar derecruitment. If necessary, this can be prevented
by an appropriate increase in positive end-expiratory pressure (PEEP). While these
principles seem evident, no consensus or formal evidence exists on how to set the
ventilator during the early oxygenation rescue phase of ECMO treatment (Table
9.1) [2022].
Walking the Tightrope Between Lung Recruitment

and Lung Rest
Three major goals are important in supporting the severely hypoxemic ARDS
patient: (a) rescue from hypoxia; (b) lung recruitment (open lung strategy); and (c)
VILI prevention (lung rest). During conventional mechanical ventilation a high
price is paid in terms of airway pressures (PEEP, plateau, and mean airway pres-
sure) to maintain arterial oxygenation, while alveolar ventilation must be sufcient
to eliminate CO2. VV-ECMO is capable of providing the entire oxygen consump-
tion, even when native lung oxygen transfer is nil (100 % intrapulmonary shunt). To
achieve this, circuit blood ow must be maximized and relatively low arterial oxy-
gen saturation levels are to be accepted, sometimes around 80 %. While this level of
oxygenation might not be better than before bypass, at least it affords viable oxy-
genation in concert with lung protection. However, a compromise is often recom-
mended (Table 9.1), mostly because some uncertainties exist as to whether a
recruited lung will heal faster than a collapsed one. Airway pressures are set at
Table 9.1 Guidelines proposed by different ECMO centers to guide ventilation settings in the
early ECMO phase
Ventilation Tidal volume RR (per PEEP
Guidelines mode (mL/kg) min) (cmH2O) FiO2
ELSO (ref. 20) PCV To reach Peak 45 10 Not reported
of 20 cmH2O
CESAR trial (ref. 21) PCV To reach Peak 10 1015 0.3
of 2025 cmH2O
Karolinska (ref. 22) PCV or PSV To reach Peak of Not 510 0.4
2025 cmH2O reported
RR respiratory rate, PEEP positive end-expiratory pressure, FiO2 inspired O2 fraction, PCV
pressure-controlled ventilation, PSV pressure support ventilation
compromise values, with the aim of maintaining a reasonable recruitment status of

the native lung. A rather different approach is taken by some, as exemplied by the
guidelines applied by the Karolinska ECMO Center [22].
We suspect a major aspect of this alternative strategy resides on the arterial oxy-
gen saturation one is prepared to accept. The reader will appreciate that the authors
have been keen to accept 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 ventila-
tion was used very early in this series).
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
(46 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 [23], while Bein
and co-workers applied ultra-protective ventilation at 3 mL/kg tidal volume in a
series of ARDS patients [24]. Others have reported the occasional use of high-
frequency ventilation combined with ECMO. Unable to distinguish between cause
and effect, Pham and co-authors [25] reported that in the rst day of ECMO the
highest plateau pressures were observed in the more severe patients, as judged by
their highest mortality rate.
Turning to the best level for PEEP, no uniform guidelines can be found either. As
a personal rule, at the start of ECMO we decrease plateau pressures and respiratory
rate, but raise 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 (or even higher). An abrupt decrease to 15 cmH2O might produce
sudden hypervolemia with the risk of lung ooding. Certainly, when deciding the
PEEP level we select between two opposing priorities: (a) to privilege lung volume
avoiding atelectasis (higher PEEP) or (b) to minimize alveolar strain (lower PEEP).
Most often, a compromise is reached, this time rather independently from oxygen-
ation needs, since oxygenation is accomplished extracorporeally.
The lack of consensus on how to set ventilation during ECMO is reected again
in Table 9.2, which reports ventilator management data during the early ECMO
phase from three different groups [2628].
Specific Considerations About Venoarterial (VA) Mode
Right heart dysfunction or failure is a very common occurrence in severe ARDS

[29], and it is an indication for venoarterial rather than venovenous ECMO. It is
important to recognize that, as a general rule, high intra-thoracic pressure consti-
tutes a hemodynamic obstacle in right ventricular failure, while the opposite may be
true for left ventricular failure. Lung collapse at very low airway pressures might
however contribute to increase pulmonary vascular resistance. Venoarterial ECMO
contributes one more degree of freedom in selecting ventilator settings in
Table 9.2 Consensus on how to set ventilation during the early ECMO phase is lacking: this table
reports main ventilator settings from three different published case series
Tidal volume RR (per PEEP
(mL/kg) min) (cmH2O) FiO2
Brogan et al. Before ECLS Not reported 20 [15;25] 12 [10;17] 1 [1;1]
(ref. 26) 24 h of ECLS Not reported 10 [10;15] 10 [8;14] 0.5 [0.4;0.5]
Zimmermann Before ECLS 6.6 [5.3;7.2] 25 [22;27] 17 [14;20] 1 [0.8;1]
et al. (ref. 27) 24 h of ECLS 4.4 [3.4;5.4] 21 [18;26] 17 [14;20] 0.7 [0.6;0.9]
Patroniti Before ECLS 6.2 [4.7;7.7] 28 [20;33] 16 [14;19] 1 [1;1]
et al. (ref. 28) 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
extracorporeal lung support
hypoxemic patients with severe right ventricular dysfunction or failure. Severe pul-
monary hypoperfusion may however favor thrombosis of the pulmonary circulation,
as originally suggested by Ratliff [30]. Hyperventilation of hypoperfused regions is
a hallmark of ARDS [31], and VA bypass might contribute to worsen it, leading to
severe tissue alkalosis and hemorrhagic edema [32]. In summary, protective ventila-
tion is all the more important during VA ECMO.
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 [33, 34]. 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 [35]. In the next paragraphs, we briey review the mechanism of action
of the different rescue therapies and, when available, the clinical data on their appli-
cation in ECMO patients.
Prone Positioning
Prone positioning was rst proposed in 1974 as a form of respiratory physiokinetic

therapy for pediatric patients [36]. Since then, its use has progressively increased
and it is now recommended as a complement to conventional ventilation in severely
hypoxemic patients [37]. An increase in oxygenation is observed in 6070 % of
patients [38], probably obtained through the combination of: more homogeneous
distribution of ventilation, recruitment of dorsal lung segments, and improved
ventilationperfusion matching [39]. This may ultimately lead to optimal recruit-

ment at a given level of PEEP with a reduced risk of VILI [34, 40].
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 results were published [37], the sustained increase in
oxygenation obtained with proning could not be translated to reduced mortality.
However, most prior trials were underpowered, included patients of different sever-
ity, and differed in the duration of proning and the details of ventilatory settings. A
number of meta-analyses supported prone positioning as effective in decreasing
mortality in the most severe ARDS patients, but producing more complications in
less severely ill patients [41]. 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 duration of proning. Prone positioning signi-
cantly decreased mortality only in ARDS patients, while a longer duration of venti-
lation in the prone position may have produced a stronger effect [42]. The PROSEVA
trial nally demonstrated that, when applied in severe ARDS patients for a suf-
ciently high number of hours per day, prone position lowers mortality [37].
Prone positioning is associated with well-known complications, such as pressure
sores, endotracheal tube removal or obstruction, dislodgement of tubes or vascular
catheters, and hemodynamic instability. It is easy to understand that these complica-
tions may be particularly important in ECMO patients, in whom compression or
inadvertent removal of the vascular cannulas may lead to reduction or interruption
of the extracorporeal support.
Very few studies provide information on the application of prone positioning dur-
ing ECMO. Hemmila and colleagues at the University of Michigan reported a case-
series of 255 patients placed on extracorporeal support for severe ARDS in whom
prone positioning for 1218 h/day was applied routinely [43]. No information was
provided on the type or number of complications attributable to position changes.
Haefner and coauthors described 63 pediatric patients (median age 12 months,
median weight 9.8 kg) receiving intermittent prone positioning while on ECMO for
respiratory failure [44]. The only complications attributable to proning were bleed-
ing from cannulation sites in 18 % of patients and chest tube dislodgement 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. retrospectively analyzed 10 ECMO patients and 42 patients on
continuous renal replacement therapy (CRRT) to assess the risk of cannula-related
complications related to proning [45]: no patient experienced inadvertent cannula
removal and only in two cases did extracorporeal ow fall during positioning; the
site of cannula insertion (jugular or femoral) did not affect the risk of malfunction.
Finally, Litmathe and coworkers reported two morbidly obese ARDS patients
(body mass index of 61 and 51 kg/m2) in whom prone positioning was applied dur-
ing ECMO: no major complications were observed, both patients improved after
proning, and both were discharged successfully from the ICU [46].
In summary, the rationale of performing a trial of prone positioning during

ECMO is to optimize alveolar recruitment and ventilationperfusion matching
without further increasing airway pressure, and possibly to reduce the risk of
VILI. It must be remembered that the effect of proning on oxygenation is unpredict-
able in the individual patient; moreover, any improvement in oxygenation may take
several hours, so a long duration of prone position seems to be advisable [37]. Based
on the available data, prone positioning in ECMO patients seems feasible and safe
but, due to the risk of potentially disastrous complications, should be performed
only in centers with extensive experience.
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-
constriction; on the contrary, inhaled delivery of short-acting pulmonary vasodilators
such as nitric oxide (NO) or prostacyclin selectively increases pulmonary blood
ow only to ventilated lung units, thus improving ventilationperfusion matching
[47]. In patients with a high shunt fraction, this can signicantly raise arterial oxy-
genation and reduce pulmonary artery pressure [48].
Inhaled NO (iNO) is the best-studied. Several trials have demonstrated its efcacy
in improving oxygenation and pulmonary vascular resistances in ARDS patients.
These effects are short-lived, however, and not associated with improved outcome: a
meta-analysis of 12 RCTs including a total of 1237 patients conrmed 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 benet in survival [49]. To
date, no studies have been published focusing on the use of iNO during ECMO: Ullrich
and colleagues described the use of iNO and ECMO as part of an integrated approach
to ARDS [35], but no information was provided on the combination. However, since
iNO can raise oxygenation rapidly, with negligible acute toxic effects, a trial (dose
range 540 ppm) can be attempted in patients who remain severely hypoxemic despite
ECMO, especially when there is concomitant pulmonary artery hypertension.
Inhaled aerosolized prostacyclin has effects similar to iNO on the pulmonary
vasculature [50] but few studies have investigated its use in ARDS patients [5153].
To our knowledge, no data have been reported on the use of inhaled prostacyclin in
ECMO patients.
High-Frequency Oscillatory Ventilation (HFOV)
HFOV combines a bias ow and an oscillating piston to deliver tidal volumes lower
than the anatomical dead space (around 12 mL/kg) at very high respiratory fre-
quency (usually 36 Hz in adults) [54]. The rapid oscillations of gas are delivered
above and below the mean airway pressure (Paw), usually set 5 cmH2O higher than
the mean observed during conventional ventilation. Oxygenation depends on the set
level of Paw and FiO2, while CO2 removal is inuenced by the pressure amplitude
and frequency of oscillation [54]. Compared to conventional ventilation, the small
tidal volumes should limit alveolar overdistension while the higher Paw promotes
alveolar recruitment and limits cyclical alveolar collapse, thus improving gas
exchange and lung protection.
HFOV is used extensively in neonates with respiratory distress syndrome [55],
while the experience in adult ARDS patients is more limited and controversial.
Recently, the results of two large, multicenter, randomized trials comparing HFOV
to conventional ventilation in adult ARDS patients have been published. The
Oscillation for Acute Respiratory Distress Syndrome Treated Early (OSCILLATE)
trial was stopped after inclusion of 548 of a planned 1200 patients because the mor-
tality of patients treated with HFOV was signicantly 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 [56]. 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 %) [57]. Despite some important methodological differences, the results of
these trials argue against widespread application of HFOV in ARDS patients.
Subjects enrolled, however, were less hypoxemic than would typically qualify for
rescue measures, so these results leave open the possibility that HFOV could be
benecial.
Few case reports describe the application of HFOV during ECMO. Banach and
coauthors reported the case of a 35-year-old man with lobar pneumonia and
refractory hypoxemia who required discontinuation of ECMO because of hemor-
rhagic complications: ECMO was then replaced by HFOV coupled with pumpless
arteriovenous extracorporeal lung assist (PECLA) for CO2 removal [58].
Muellenbach and colleagues showed in both an animal model [59] and in a patient
with post-traumatic ARDS [60] that the combination of an arteriovenous extracor-
poreal lung assist with HFOV allowed the use of oscillatory frequencies higher
than those usually applied (up to 1015 Hz), thus minimizing the risk of baro- and
volutrauma.
Management of Pneumothoraces
Finally, we discuss ventilation during ECMO for patients with established baro-
trauma (pneumothorax, pneumomediastinum, subcutaneous emphysema). In such
cases we have learned [23], like others, that the best way to manage pneumothora-
ces is to decrease ventilation further, lower airway pressure as much as possible, and
wait for the gas to be reabsorbed. Chest drainage becomes mandatory only when
pneumothorax causes hemodynamic impairment.
Assisted Ventilation During ECLS
Following the phase of initial ECMO application and when the patients condition is
more stable, mechanical ventilation is switched from control to assisted mode. The
aim is then a progressive decrease in support, ECLS discontinuation and, nally,
extubation. The pace toward extubation, though dictated mainly by the individual
patient evolution (some patients stay days on bypass, but a few stay months), is
approached differently in various centers. An early switch to protective, assisted MV
may improve respiratory muscle function and gas exchange, decrease the need for
sedation, and aid weaning from ventilator [61]. In recent years, important techno-
logical advances contributed to signicantly improve the safety of ECLS. This led
many investigators to a change in strategy, moving to wean the patient rst from the
ventilator, and even extubate, then later from bypass [62]. Our own approach is to
start the weaning process (i.e. assisted breathing rather than controlled) as soon as
the patients safety and comfort allow. While most groups delay assisted ventilation
until the native lung improves substantially, others almost immediately switch the
ventilator to pressure support mode [63]. For example, the Karolinska group reported
their experience in 17 adult ARDS patients treated by ECLS between 1995 and
1999. They discontinued muscle relaxation and set the ventilator to pressure support
right after starting ECLS. Sedation was lightened and, soon, patients were awake
during the day, able to interact with staff and family, and participating in quiet activi-
ties like watching TV. Spontaneous assisted breathing was enhanced, when needed,
by adding 5 % CO2 to the sweep gas. Pressure support ventilation was used for all
patients, but no data are reported to indicate how pressures were titrated. Tidal vol-
ume fell from a range of 450947 mL before ECLS to 122662 mL during extracor-
poreal treatment. Mortality was surprisingly low (24 %) given the baseline severity.
From this study we conclude that in severe ARDS patients ECLS coupled with mini-
mal sedation, PSV, and low tidal volumes may be associated with high survival.
Our group previously showed that PSV may be difcult to implement in ARDS
patients with very low respiratory system compliance (Crs), likely because inspira-
tory ow falls rapidly and the ow-triggered expiratory phase of PSV starts while
the patient is still inspiring (premature expiratory cycling) [64]. Thus, these patients
are at high risk of patientventilator asynchrony. Asynchrony is a serious threat, as
higher asynchrony is associated with iatrogenic injury and delayed weaning from
the ventilator. To this end, we compared PSV with neurally adjusted ventilator assist
(NAVA) [65] in a group of severe ARDS patients with Crs ranging between 7 and 31
mL/cmH2O undergoing venovenous ECLS. In this selected group of patients, low
tidal volume PSV (34 mL/kg) led to premature expiratory cycling and high patient
ventilator asynchrony. In contrast, NAVA improved patientventilator interaction,
the more so in patients with the lowest Crs values [62].
Other modes of assisted ventilation have been applied during ECMO. For example,
airway pressure release ventilation (APRV) has been used [66], a mode of ventilation
that alternates two pressure levels at xed inspiratory and expiratory times. APRV
allows spontaneous breathing throughout the respiratory cycle, potentially avoiding
respiratory muscle atrophy and increasing ventilation to dependent lung zones.
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 patients
awaiting lung transplant (a few of whom were affected by severe ARDS) are so
numerous that it may be considered standard of care. Most notably, Fuehner and
coauthors published a retrospective analysis of 26 patients treated with ECLS at the
onset of end-stage respiratory failure [67]. Of these, 16 patients did not require sub-
sequent intubation and were transplanted after a median of 11 days on ECMO. Two
died from septic multiorgan failure after transplantation. Another patient died 60
days after transplantation from lung cancer. The remaining 13 patients were dis-
charged from the hospital 2087 days after transplantation and all of these patients
remained alive during the follow-up period (739 months). In comparison, an his-
torical control group underwent MV without ECMO while waiting transplant: those
who survived until hospital discharge required longer times on the ventilator, in the
ICU, and in hospital [67]. Overall survival was 62 % in the awake ECMO group and
35 % in the MV group (intention-to-treat populations at 6 months; 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
(p = 0.02). Thus, ECMO should be considered before intubation in patients with
end-stage respiratory failure waiting for lung transplant, but only in highly special-
ized centers with prepared and dedicated staff.
Is an awake ECMO approach feasible in severe ARDS patients? Hoeper and
colleagues reported the results of a single-center, uncontrolled, pilot trial designed
to assess the feasibility of venovenous ECMO in awake, non-intubated, spontane-
ously breathing patients with ARDS [68]. Six patients were enrolled, four of whom
were immunocompromised. Three were ultimately weaned from ECMO without
being intubated and discharged from hospital alive. On the other hand, the other
three required invasive MV, and two died in hospital. The authors concluded that
awake ECMO appears feasible in selected patients with ARDS and deserves fur-
ther evaluation.
Avoiding intubation is particularly attractive for specic situations, such as in
immunocompromised patients with ARDS, so that ECMO could be considered a
primary approach rather than only a rescue tactic. This concept is very promising,
but systematic clinical experience is still lacking.
Disconnection from ECLS
No universally accepted rules exist as to when disconnect a patient from ECMO. The
risk inherent with extracorporeal support has to be evaluated against the risk of the
patient deteriorating without such support. For example, new bleeding may dictate
earlier than optimal disconnection in order to stop anticoagulation. On the other
hand, a new infection may challenge the marginal patients physiological reserve,
making discontinuation of ECMO unduly risky. We consider very useful the guide-
line proposed by Palmer: a test of at least 4 h is conducted with the sweep gas ow
shut off; the ventilator in PSV mode; peak inspiratory pressure not more than 25
cmH2O; mean airway pressure between 10 and 15 cmH2O; FiO2 0.5 or less; and
PEEP between 5 and 10 cmH2O. If the patient passes the test, we proceed with
decannulation under local anesthesia in combination with mild sedation [63].
Partial Extracorporeal CO2 Removal (PECOR)

and Ventilator Management
Modern extracorporeal CO2 removal techniques are characterized by very low

(<500 mL/min), or intermediate blood ow (<1.5 L/min). Clinical experience and
data are very scanty and, at time of writing, insufcient to provide any guideline or
recommendation. The eld of extracorporeal CO2 removal still remains an experi-
mental one. It is important, however, to review the principles upon which the physi-
ology of extracorporeal CO2 removal is based in order to understand its side effects,
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
conducted in Kolobows laboratory at the National Institutes of Health in the previ-
ous years [69]. The technique was a hybrid between ECMO and pure ECCO2R. Based
on venovenous bypass at 1.52.5 L/min, oxygenation was achieved mostly through
the native lung, kept inated by PEEP and a very low respiratory rate (46 breaths/
min) with limited plateau pressure. At that time [9] we claimed that the technique
seems to overcome the ventilation/perfusion mismatching due to ventilation mald-
istribution 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-dened.
The rst real low-ow, partial CO2 removal technique was applied in a patient
with large, bilateral air leaks following bilateral pleurectomy [70]. In this case can-
nulation was percutaneous, blood ow ranged from 0.4 to 0.6 L/min, and between 22
and 40 % of the total carbon dioxide production was removed. Thanks to the decreased
ventilator needs afforded by ECCO2R, the patient could be switched to CPAP in a few
days, and was eventually extubated with complete resolution of the air leak.
Ventilatory Control During ECLS/PECOR
Control of breathing becomes increasingly important as increasing numbers of

ECMO patients are allowed to breathe spontaneously or on assisted modes. Kolobow
and Gattinoni showed that ECCO2R could control spontaneous breathing and the
need for mechanical ventilation in sheep [15, 18]. When we faced the rst patients
Fig. 9.1 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 ow (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)
recovering from severe ARDS, we could show that they decreased their ventilation
(on CPAP) in proportion to the amount of CO2 removed by the membrane lung [71].
The same effect has been observed in ARDS patients ventilated with NAVA while
on ECMO [72].
While supporting the possibility of targeting ventilatory needs and respiratory
drive in ARDS patients, these data should be interpreted with caution. Experience
shows that in the early, acute interstitial edema phase, ARDS patients have a very
high respiratory drive characterized by severe dyspnea and tachypnea. Blood gas
values have only a minor impact on respiratory drive. In ARDS patients on ECMO,
respiratory rate remains high and inspiratory effort may still generate very negative
intrathoracic pressures (and high transpulmonary pressures risking barotrauma and
VILI) despite low values for PaCO2 (2528 mmHg; Fig. 9.1). These observations
contrast with what has been reported in acutely decompensated COPD patients, in
whom respiratory drive (and the need for intubation) can be effectively controlled
by ECCO2R [73, 74].
One possible role for ECCO2R in ARDS patients, then, revolves around the
ultra-protective ventilator approach proposed by Terragni and colleagues [75].
These investigators showed that 2530 % of ARDS patients showed evidence of
severe hyperination on CT scan, even while ventilated according to the NIH lung-
protective recommendations (Fig. 9.2) [76]. They reasoned that VILI risk could be
reduced by lowering tidal volume from 6 to 4 mL/kg, while maintaining nearly
normal arterial pH and PCO2 through low-ow (250400 mL/min)
ECCO2R. Although their study was not designed to provide outcome data, it showed
the feasibility of such an approach. Two case reports suggest the possibility of using
ultra-protective ventilation (down to 2 mL/kg) facilitated by extracorporeal gas
exchange ([23, 77], case report 2). Finally, the Xtravent study reported on using
tidal volumes of 3 mL/kg, supported by ECCO2R, in a group of severe ARDS
patients [24]. Although the study was stopped due to slow enrolment, post-hoc anal-
ysis revealed signicantly earlier weaning from MV in the ultra-protective arm.
We can foresee the use of minimally invasive, low-ow ECCO2R devices in the
early phase of ARDS, either to avoid or delay intubation, possibly by extending the
Fig. 9.2 Lung CT scan from one representative patient showing signs of diffuse alveolar over-
distension (red areas) despite being ventilated according to the NIH recommendations to achieve
lung protective ventilation (modied from ref. 76)
range of application of NIV or CPAP. Even though muscle paralysis and controlled
mechanical ventilation are certain to maintain a role in ARDS management, judi-
cious application of ECCO2R may lead to earlier assisted breathing, weaning, and
extubation. The possibility of less sedation, greater capacity to interact and com-
municate, and briefer intubation (or none at all) is certainly appealing, and could
reduce the incidence of ventilator-associated pneumonia and ICU-acquired weak-
ness. The most promising applications of low-ow ECCO2R appear to be related
however to the concept of ultra-protective ventilation in ARDS [76] and to the treat-
ment of COPD patients [78].
Conclusions
The concept of VILI and the need for protective ventilation has gained, following
the early experimental work of Kolobow and Dreyfuss, widespread acceptance hav-
ing been clearly validated in the clinical eld.
The practical application of these concepts in the patient undergoing ECLS,
though based on solid pathophysiological grounds, lacks however enough scientic
evidence to allow the drawing of guidelines or recommendations. Clinical experi-
ence suggests two alternative approaches, possibly applicable at different times in
the same patient: both recommend protective ventilation with tidal volumes lower
than 6 mL/kg, one looks for optimized lung recruitment by a higher PEEP level, 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 neces-
sary for the lung to recover.
References
1. Zapol WM, Snider MT, Hill JD, Fallat RJ, Bartlett RH, Edmunds LH, et al. Extracorporeal
JAMA. 1979;242(20):21936.
2. National Heart, Lung, and Blood Institute: Division of Lung Disease. Section four: hemody-
namics, gas exchange and mechanics. In: US Department of Health, Education and Welfare,
editor. Extracorporeal support for respiratory insufciency. A collaborative study in response
to RFP-NHLI-73-20. Washington, DC: US Department of Health, Education and Welfare;
1979.
3. Pontoppidan H, Gefn B, Lowenstein E. Acute respiratory failure in the adult. 3. N Engl J
Med. 1972;287(16):799806.
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):15862.
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):589.
7. Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies.
Am J Respir Crit Care Med. 1998;157(1):294323.
8. Zapol WM, Kitz RJ. Buying time with articial lungs. N Engl J Med. 1972;286(12):6578.
9. Gattinoni L, Agostoni A, Pesenti A, Pelizzola A, Rossi GP, Langer M, et al. Treatment of acute
respiratory failure with low-frequency positive-pressure ventilation and extracorporeal
removal of CO2. Lancet. 1980;2(8189):2924.
10. Gattinoni L, Kolobow T, Agostoni A, Damia G, Pelizzola A, Rossi GP, et al. Clinical applica-
tion of low frequency positive pressure ventilation with extracorporeal CO2 removal (LFPPV-
ECCO2R) in treatment of adult respiratory distress syndrome (ARDS). Int J Artif Organs.
1979;2(6):2823.
of the Bramson membrane lung. N Engl J Med. 1972;286(12):62934.
12. Guarracino F, Zangrillo A, Ruggeri L, Pieri M, Calabr MG, Landoni G, et al. -Blockers to
optimize peripheral oxygenation during extracorporeal membrane oxygenation: a case series.
J Cardiothorac Vasc Anesth. 2012;26(1):5863.
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):97117.
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;23:1721.
15. Kolobow T, Gattinoni L, Tomlinson TA, Pierce JE. Control of breathing using an extracorpo-
real membrane lung. Anesthesiology. 1977;46(2):13841.
16. Kolobow T, Gattinoni L, Tomlinson T, Pierce JE. An alternative to breathing. J Thorac
Cardiovasc Surg. 1978;75(2):2616.
17. Gattinoni L, Kolobow T, Tomlinson T, Iapichino G, Samaja M, White D, et al. Low-frequency
positive pressure ventilation with extracorporeal carbon dioxide removal (LFPPV-ECCO2R):
an experimental study. Anesth Analg. 1978;57(4):4707.
18. Gattinoni L, Kolobow T, Tomlinson T, White D, Pierce J. Control of intermittent positive pres-
sure breathing (IPPB) by extracorporeal removal of carbon dioxide. Br J Anaesth.
1978;50(8):7538.
Med. 2011;365(20):190514.
20. ELSO Patient Specic guidelines. http://www.elsonet.org/index.php?option=com_phocadow
nload&view=category&id=4&Itemid=627.
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):135163.
22. Holzgraefe B, Broom M, Kalzn H, Konrad D, Palmr K, Frenckner B. Extracorporeal mem-
brane oxygenation for pandemic H1N1 2009 respiratory failure. Minerva Anestesiol.
2010;76(12):104351.
23. Mauri T, Bellani G, Foti G, Grasselli G, Pesenti A. Successful use of neutrally adjusted ventila-
tory assist in a patient with extremely low respiratory system compliance undergoing
ECMO. Intensive Care Med. 2011;37(1):1667.
24. Bein T, Weber-Carstens S, Goldmann A, Mller T, Staudinger T, Brederlau J, et al. Lower tidal
volume strategy (3 ml/kg) combined with extracorporeal CO2 removal versus conventional
protective ventilation (6 ml/kg) in severe ARDS: the prospective randomized Xtravent-study.
Intensive Care Med. 2013;39(5):84756.
25. Pham T, Combes A, Roze H, Chevret S, Mercat A, Roch A, et al. Extracorporeal membrane
oxygenation for pandemic inuenza A(H1N1)-induced acute respiratory distress syndrome: a
cohort study and propensity matched analysis. Am J Respir Crit Care Med.
2013;187(3):27685.
26. 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):210514.
27. Zimmermann M, Bein T, Arlt M, Philipp A, Rupprecht L, Mueller T, et al. Pumpless extracor-
poreal interventional lung assist in patients with acute respiratory distress syndrome: a pro-
spective pilot study. Crit Care. 2009;13(1):R10.
28. Patroniti N, Zangrillo A, Pappalardo F, Peris A, Cianchi G, Braschi A, et al. The Italian ECMO
network experience during the 2009 inuenza A(H1N1) pandemic: preparation for severe
respiratory emergency outbreaks. Intensive Care Med. 2011;37(9):144757.
29. Repess X, Charron C, Vieillard-Baron A. Right ventricular failure in acute lung injury and
acute respiratory distress syndrome. Minerva Anestesiol. 2012;78(8):9418.
30. Ratliff JL, Hill JD, Fallat RJ, Parrot J, Tucker HJ. Complications associated with membrane
lung support by venoarterial perfusion. Ann Thorac Surg. 1975;19(5):5379.
31. Nuckton TJ, Alonso JA, Kallet RH, Daniel BM, Pittet JF, Eisner MD, et al. Pulmonary dead-
space fraction as a risk factor for death in the acute respiratory distress syndrome. N Engl J
Med. 2002;346(17):12816.
32. Kolobow T, Spragg RG, Pierce JE. Massive pulmonary infarction during total cardiopulmo-
nary bypass in unanesthetized spontaneously breathing lambs. Int J Artif Organs.
1981;4(2):7681.
33. Pipeling MR, Fan E. Therapies for refractory hypoxemia in acute respiratory distress syn-
drome. JAMA. 2010;304(22):25217.
34. Patroniti N, Isgr S, Zanella A. Clinical management of severely hypoxemic patients. Curr
Opin Crit Care. 2011;17(1):506.
35. Ullrich R, Lorber C, Rder G, Urak G, Faryniak B, Sladen RN, et al. Controlled airway pres-
sure therapy, nitric oxide inhalation, prone position, and extracorporeal membrane oxygen-
ation (ECMO) as components of an integrated approach to ARDS. Anesthesiology.
1999;91(6):157786.
36. Bryan AC. Conference on the scientic basis of respiratory therapy. Pulmonary physiotherapy
in the pediatric age group. Comments of a devils advocate. Am Rev Respir Dis. 1974;110(6
Pt 2):1434.
37. Gurin C, Reignier J, Richard JC, Beuret P, Gacouin A, Boulain T, et al. Prone positioning in
severe acute respiratory distress syndrome. N Engl J Med. 2013;368(23):215968.
38. Hemmila MR, Napolitano LM. Severe respiratory failure: advanced treatment options. Crit
Care Med. 2006;34(9 Suppl):S27890.
39. Pelosi P, Brazzi L, Gattinoni L. Prone position in acute respiratory distress syndrome. Eur
Respir J. 2002;20(4):101728.
40. Galiatsou E, Kostanti E, Svarna E, Kitsakos A, Koulouras V, Efremidis SC, et al. Prone posi-
tion augments recruitment and prevents alveolar overination in acute lung injury. Am J Respir
Crit Care Med. 2006;174(2):18797.
41. Sud S, Friedrich JO, Taccone P, Polli F, Adhikari NK, Latini R, et al. Prone ventilation reduces
mortality in patients with acute respiratory failure and severe hypoxemia: systematic review
and meta-analysis. Intensive Care Med. 2010;36(4):58599.
42. 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.
43. Hemmila MR, Rowe SA, Boules TN, Miskulin J, McGillicuddy JW, Schuerer DJ, et al.
Extracorporeal life support for severe acute respiratory distress syndrome in adults. Ann Surg.
2004;240(4):595607.
44. 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):158994.
45. 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):4525.
46. Litmathe J, Sucker C, Easo J, Wigger L, Dapunt O. Prone and ECMOa contradiction per se?
Perfusion. 2012;27(1):7882.
47. Grifths MJ, Evans TW. Inhaled nitric oxide therapy in adults. N Engl J Med.
2005;353(25):268395.
48. Rossaint R, Falke KJ, Lpez F, Slama K, Pison U, Zapol WM. Inhaled nitric oxide for the adult
respiratory distress syndrome. N Engl J Med. 1993;328(6):399405.
49. 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.
50. 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):7806.
51. 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):9916.
52. Walmrath D, Schneider T, Pilch J, Grimminger F, Seeger W. Aerosolised prostacyclin in adult
respiratory distress syndrome. Lancet. 1993;342(8877):9612.
53. 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):81927.
54. Hager DN. High-frequency oscillatory ventilation in adults with acute respiratory distress syn-
drome. Curr Opin Anaesthesiol. 2012;25(1):1723.
55. Courtney SE, Durand DJ, Asselin JM, Hudak ML, Aschner JL, Shoemaker CT, et al. High-
frequency oscillatory ventilation versus conventional mechanical ventilation for very-low-
birth-weight infants. N Engl J Med. 2002;347(9):64352.
56. Ferguson ND, Cook DJ, Guyatt GH, Mehta S, Hand L, Austin P, et al. High-frequency oscilla-
tion in early acute respiratory distress syndrome. N Engl J Med. 2013;368(9):795805.
57. Young D, Lamb SE, Shah S, MacKenzie I, Tunnicliffe W, Lall R, et al. High-frequency oscil-
lation for acute respiratory distress syndrome. N Engl J Med. 2013;368(9):80613.
58. 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):2015.
59. Muellenbach RM, Kuestermann J, Kredel M, Johannes A, Wolfsteiner U, Schuster F, et al.
Arteriovenous extracorporeal lung assist allows for maximization of oscillatory frequencies: a
large-animal model of respiratory distress. BMC Anesthesiol. 2008;8:7.
60. Muellenbach RM, Wunder C, Nuechter DC, Smul T, Trautner H, Kredel M, et al. Early treat-
ment 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):7669.
61. Putensen C, Muders T, Varelmann D, Wrigge H. The impact of spontaneous breathing during
mechanical ventilation. Curr Opin Crit Care. 2006;12(1):138.
62. Mauri T, Bellani G, Grasselli G, Confalonieri A, Rona R, Patroniti N, et al. 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):28291.
63. Lindn V, Palmr K, Reinhard J, Westman R, Ehrn H, Granholm T, et al. High survival in
adult patients with acute respiratory distress syndrome treated by extracorporeal membrane
oxygenation, minimal sedation, and pressure supported ventilation. Intensive Care Med.
2000;26(11):16307.
64. Cereda M, Foti G, Marcora B, Gili M, Giacomini M, Sparacino ME, et al. Pressure support
ventilation in patients with acute lung injury. Crit Care Med. 2000;28(5):126975.
65. Sinderby C. Neurally adjusted ventilatory assist (NAVA). Minerva Anestesiol.
2002;68(5):37880.
66. Combes A. Extracorporeal membrane oxygenation (ECMO) for severe acute respiratory dis-
tress syndrome (ARDS). The EOLIA (ECMO to rescue Lung Injury in severe ARDS) trial: a
multicenter, international, randomized, controlled open trial. Ranimation. 2011;20:4961.
67. 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):7638.
68. Hoeper MM, Wiesner O, Hadem J, Wahl O, Suhling H, Duesberg C, et al. Extracorporeal
membrane oxygenation instead of invasive mechanical ventilation in patients with acute respi-
ratory distress syndrome. Intensive Care Med. 2013;39(11):20567.
69. 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):1835.
70. 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):5713.
71. 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):50810.
72. Karagiannidis C, Lubnow M, Philipp A, Riegger GA, Schmid C, Pfeifer M, et al. Autoregulation
of ventilation with neurally adjusted ventilatory assist on extracorporeal lung support. Intensive
Care Med. 2010;36(12):203844.
73. Burki NK, Mani RK, Herth FJ, Schmidt W, Teschler H, Bonin F, et al. A novel extracorporeal
CO(2) removal system: results of a pilot study of hypercapnic respiratory failure in patients
with COPD. Chest. 2013;143(3):67886.
74. Cardenas Jr VJ, Lynch JE, Ates R, Miller L, Zwischenberger JB. Venovenous carbon dioxide
removal in chronic obstructive pulmonary disease: experience in one patient. ASAIO
J. 2009;55(4):4202.
75. Terragni PP, Del Sorbo L, Mascia L, Urbino R, Martin EL, Birocco A, et al. Tidal volume
lower than 6 ml/kg enhances lung protection: role of extracorporeal carbon dioxide removal.
Anesthesiology. 2009;111(4):82635.
76. Terragni PP, Rosboch G, Tealdi A, Corno E, Menaldo E, Davini O, et al. Tidal hyperination
during low tidal volume ventilation in acute respiratory distress syndrome. Am J Respir Crit
Care Med. 2007;175(2):1606.
77. Bein T, Osborn E, Hofmann HS, Zimmermann M, Philipp A, Schlitt HJ, et al. Successful treat-
ment of a severely injured soldier from Afghanistan with pumpless extracorporeal lung assist
and neurally adjusted ventilatory support. Int J Emerg Med. 2010;3(3):1779.
78. Kluge S, Braune SA, Engel M, Nierhaus A, Frings D, Ebelt H, et al. Avoiding invasive
mechanical ventilation by extracorporeal carbon dioxide removal in patients failing noninva-
sive ventilation. Intensive Care Med. 2012;38(10):16329.
Chapter 10
Daily Care on ECLS
Giles J. Peek
Introduction
Daily management of the patient on ECLS requires a careful assessment of the cir-
cuit, the patient, and the interaction between them. Robust communication between
ECLS specialists, intensivists, surgeons, nurses, respiratory therapists, and others is
essential to providing safe and effective care. Once the patient has been established
on ECLS their care passes through three phases of stabilization, stability, and nal
improvement before they can be weaned off ECLS and decannulated. Accordingly,
this chapter will address daily care in four sections: (a) housekeeping of the cir-
cuit; (b) initial stabilization of gas exchange and cardiovascular homeostasis; (c)
meticulous progress toward regaining organ function and limiting complications
during a phase of relative stability; and, nally, (d) judgments regarding when to
withdraw ECLS as the patient improves.
Circuit Housekeeping
The ECLS specialist examines the circuit from end-to-end at the beginning of each
shift and rechecks it hourly to ensure its safe operation and detect problems before
they become circuit-threatening. A checklist should be used to ensure that nothing
is omitted. We begin by examining the exit site of the drainage cannula (or double-
lumen cannula) for signs of infection, bleeding, or visible recirculation, while also
assuring that the sutures are secure. A transparent lm dressing is helpful in
G.J. Peek, MD, FRCS CTh, FFICM (*)

Heartlink ECMO Centre, Gleneld Hospital, Groby Road, Leicester LE3 9QP, UK
e-mail: gilespeek@hotmail.co.uk

182 G.J. Peek
allowing the cannulation site to be easily viewed. After 2 or 3 weeks the skin under
these dressings can become sweaty and macerated so we often switch to a breath-
able dressing. We swab the cannulation site two or three times per week for micro-
biological surveillance. If the cannulation site is breaking down (which can happen
after several months on ECLS), we may apply an antiseptic ointment (chloram-
phenicol or betadine) sparingly.
Next the connector joining the cannula to the venous line is checked: this is espe-
cially important during the rst circuit check following cannulation in case steps
were overlooked in error. The connection on the cannula side is bonded in the fac-
tory and does not require tie strapping. The connector should also be examined for
cracks and clots with a bright light. This light is then used to follow the circuit past
the bridge (if present), along to the compliance chamber (if used), and at the inlet
pressure measurement port (if used). Each connector should be checked for secu-
rity, tie straps, cracks, and clots. Luer connectors should have their bungs or pigtails
checked for integrity and the pigtails themselves should be examined for clots. Any
pigtails that are not being used should either be ushed with heparin-containing
saline or removed altogether (a simpler circuit is a safer circuit). If a bridge is used
it should be examined carefully for clots and the tubing for cracks; constant clamp-
ing and unclamping can lead to tubing failure. It is becoming increasingly common
to omit the bridge to remove a source of clot formation. Having either a stopcock or
removable bridge is a good compromise between a streamlined circuit with few
connectors and the security of being able to use a bridge if necessary.
If a venous compliance chamber is being used, its function should be tested by
clamping the inlet tubing and watching the bladder collapse. It should be examined
for clots and the housing examined for cracks. Air can be added or withdrawn to
adjust the level of pressure in the chamber so that a dimple is seen in the compliance
tubing. If the inlet pressure is being measured from the chamber, the transducer
should be zeroed (the transducer tubing and system should be air-lled for this pur-
pose). If a compliance chamber is not being used and the inlet pressure is being
measured directly from the circuit, the (saline-lled) transducer system should be
zeroed with caution to prevent air being entrained into the circuit. Because of this
risk, it is increasingly common to omit measurement of inlet pressure.
The tubing is then followed to the pump which is checked with the light for clots.
The pump should be listened to for abnormal sounds and felt for excessive heat
generation or vibration. The console should be surveyed according to the manufac-
turers instructions. The alarms, alarm limits, pressure measurements (pre- and
post-membrane), pump speed, ow, battery, and power supply should be veried. If
inlet pressure is not being measured, the relationship of pump speed and circuit ow
should be checked to ensure that excessive rpms are not being used. This relation-
ship varies from pump to pump so that each ECLS center should create a nomogram
of what ow to expect at a given rpm setting. For the 3/8 Thoratec Centrimag the
blood ow in mL/min should exceed the rpm; thus a blood ow of 4.5 L/min with a
pump speed of 3250 rpm indicates unobstructed venous outow. A lower ratio of
ow to rpm signals higher afterload, as when the membrane is failing or during
venoarterial (VA) ECLS.
10 Daily Care on ECLS 183
The tubing is then followed to the oxygenator. If pigtails are used to access the
circulation, they are best placed between pump and oxygenator where inadvertant
opening forces blood out rather than sucks air in (hypovolemia being much easier to
correct than a circuit full of air). If a continuous venovenous hemoltration (CVVH)
machine is connected to the ECLS circuit at this point, the inlet and outlet pressure
limits for the CVVH access must be adjusted to account for the positive pressure in
the ECLS circuit.
The membrane lung (oxygenator) should be examined for clots and cracks, and
all taps, bungs, and connections examined for cracks and checked for tightness. The
pressure gradient across the membrane should be measured and compared to prior
values; with polymethylpentene membranes this should be on the order of 1050
mmHg, depending on ow, with higher values suggesting clot build up and incipient
oxygenator failure. Membrane function should be checked by noting the color of
blood exiting the membrane, which should be bright red. The PO2 of a blood sample
taken from the outlet pigtail of the oxygenator will be on the order of greater than
30 kPa (225 mmHg), assuming 100 % oxygen as the sweep gas. Falling post-
membrane PO2 values may signal a failing oxygenator, but its ability to support gas
exchange is not impaired until the PO2 drops below about 15 kPa (112 mmHg) and
the saturation begins to fall. The decision to change out the oxygenator is based on
the rate of falling post-membrane PO2 and the trajectory of the patients improve-
ment. If the patient is likely to come off ECLS in the next 48 h and membrane func-
tion is deteriorating slowly, it would be better to push on with the current oxygenator
(since replacement has risks, costs, and may induce additional systemic inamma-
tion, delaying lung recovery). Membrane life can be prolonged by increasing the
degree of anticoagulation and maintaining high circuit blood ow (at least 2 L/min).
However, if the patient will not recover soon, it is better to replace a failing oxygen-
ator electively rather than put the patient at risk of a sudden deterioration and emer-
gency replacement (see Chap. 11).
The sweep gas ow and the FIO2 selector on the blender should be veried.
Condensation can build up in the gas phase of the oxygenator and lead to impairment
of gas exchange. This can be seen as clear water dripping from the gas exhaust port
and will be reected in a poor post-oxygenator PO2, but a normal pressure gradient.
To prevent uid accumulation, the membrane should be ushed with high ow
sweep gas for a few seconds every shift, or more often if uid is clearly building up.
In addition, the membrane can be burped by occluding the gas exhaust port with a
nger for a few seconds and allowing the gas pressure to build up, then releasing.
Care should be exercised to prevent excessive pressure: some centers use a safety
valve on the sweep gas inlet to prevent this. Burping should not be attempted if using
micro-porous polypropylene oxygenators, because these membranes develop plasma
leakage characterized by bubbling of amber-colored plasma from the gas exhaust.
Water heater function is judged by noting whether the tubing is warm, the pump
turning, the reservoir still contains water, and the patients temperature is stable.
Reservoir water can become colonized with pseudomonas: we send samples for
microbiological culture periodically.
184 G.J. Peek
Circuit housekeeping concludes with careful appraisal of the blood tubing from
oxygenator past the bridge (if used) to the patient. If single-lumen cannulae are
being used, the return cannula is checked in the same way as the drainage cannula
was; if a double-lumen cannula is being used the assessment is nished at the
connector.
Phase I: Stabilization on ECLS
ECLS ow is initiated and built up slowly over a few minutes, until further increases
in pump speed do not produce more ow. The speed is then reduced by 1020 % to
prevent excessive negative inlet pressure. If adequate ow (around 60 mL/kg/min in
adults) cannot be achieved readily, the patient may be hypovolemic or cannula posi-
tion may be suboptimal. During initiation of ECLS the patient may become hypo-
tensive due to sudden changes in concentrations of potassium, calcium, or
magnesium. The risk of this seems less if the circuit is primed using a balanced
electrolyte solution. If hypotension occurs administration of calcium will often
prove effective.
During the early minutes we keep the sweep gas ow low (around 2 L/min) in
order to prevent a sudden reduction in PaCO2 resulting in cerebral vasoconstriction.
Once circuit blood ow has been optimized, we set the sweep gas ow at a similar
value (in L/min) and start reducing the ventilation as described in Chap. 9 in order
to reduce the risk of ventilator-induced lung injury (VILI) [2]. Our own practice is
to use pressure-controlled ventilation with the inspiratory pressure initially below
30 cm H2O, but reduced over the next several hours to 10 cm H2O above PEEP [3].
We generally apply 10 or 15 cm H2O PEEP and a respiratory rate of 10/min while
the patient is paralyzed, then convert to pressure support of 10 cm H2O over PEEP
after paralysis is discontinued. As ventilator settings are reduced, sweep gas ow is
adjusted to achieve the target PaCO2 (usually 46 kPa; 3045 mmHg) based on arte-
rial blood gas results. For patients with severe hypercapnia, rapid reduction in
PaCO2 may not be desired and a relatively normal arterial pH is a more appropriate
target for the rst 1224 h as the metabolic compensation resolves.
The patients systemic oxygenation depends largely on how much of the cardiac
output can be directed to the ECLS circuit, as discussed in Chap. 1. Once circuit
ow is adequate, it should be possible to discontinue other oxygenation rescue ther-
apies, such as inhaled prostacyclin or nitric oxide. Ventilator FIO2 can be reduced as
long as circuit ow and membrane function are adequate, with a goal SpO2 gener-
ally above 80 %. We aim to reduce the FiO2 to 30 % over the rst 1224 h. If this is
not possible, one of two problems is likely:
1. The ECLS circuit is inefcient because of recirculation, poor cannula position,
or inadequate size of the venous cannula.
2. The patient has a very high cardiac output and the ECLS circuit ow is capturing
an insufcient proportion of this resulting in an effective right to left shunt (since
the lungs are not working well).
Solutions may involve cannula repositioning, inserting an additional venous

drainage cannula, weaning inotropes, or mild systemic cooling (35.5 C). Since a
primary aim of ECLS is to achieve lung rest, one should strive to get the FiO2 below
60 % and inspiratory airway pressures below 30 cm H2O within 24 h. The stabiliza-
tion phase has not ended until lung rest is achieved.
For ARDS patients treated with venovenous (VV) ECLS, hemodynamic stability
is usually achieved readilyindeed, the VV mode is chosen only when patients have
acceptable cardiovascular function. For patients with circulatory failure placed on
VA ECLS, initiation of ow usually produces rapid improvements in blood pressure
and measures of perfusion. In either case, vasoactive drug infusions can usually be
reduced rapidly as long as intravascular volume is adequate. Moreover, as ECLS
augments systemic oxygenation, facilitates lower ventilator pressures (and the atten-
dant cardiopulmonary interactions that contribute to shock [4]), and lowers right
ventricular afterload (see Chap. 1), there is an opportunity to rapidly reduce hemody-
namic support. Patients with refractory shock are at very high risk of death but may
benet from conversion to veno-arterio-venous (VAV) ECLS to provide some direct
circulatory support (see Chap. 6). If urine ow is inadequate despite hemodynamic
stabilization, we initiate CVVH aiming for zero or slightly negative uid balance.
During cannulation a bolus of heparin will have been administered, typically
around 50100 U/kg. An activated clotting time (ACT) sample is drawn to conrm
prolongation as a result of the heparin bolus, and unfractionated heparin is infused
at around 1015 U/kg/h targeting an ACT of 160220 s. We recheck the ACT hourly
and adjust the rate of heparin infusion as described more fully below. Blood is
drawn for estimation of platelet count, blood count, brinogen, international nor-
malized ratio (INR), and, in specialized circumstances, thromboelastography
(TEG). Transfusion of blood products is guided by these values.
At the conclusion of this management phase, the patient has been on ECLS for 612
h; ventilation has been reduced to allow lung rest; vasoactive drugs have been reduced
or stopped; paralytic drugs have been discontinued; renal function or renal replacement
therapy has been established; and hematological targets have been achieved.
Phase II: Stability on ECLS
This phase of the illness is far from the passive period one might expect, the aim
being to actively move the patient in the direction of recovery from underlying ill-
ness, iatrogenic VILI, microcirculatory dysfunction, and ICU-acquired weakness.
During this phase we hope that capillary leak resolves, edema improves, and that the
catabolism of injury is replaced by repair and healing. Lung protection may allow
radiographic clearing and improvements in lung mechanics. A coherent plan of
management is crucial, using a systematic approach and institutionally agreed pro-
tocols. We aim to use as many of the standard ICU bundles as possible to ensure that
normal intensive care management proceeds, but with some specic provisos dis-
cussed in more detail below.
186 G.J. Peek
Lung Management
After stabilization, the patient should be on low pressure (or volume) settings with
a machine rate of 10 breaths/min but triggering the ventilator (usually on pressure
support of 10 cm H2O). The tidal volume may be as low as 0.5 mL/kg and the
patient may be quite tachypneic (2030 breaths/min or even higher) as a conse-
quence of the disease process. Raising the sweep gas ow to enhance CO2 removal
has little apparent impact on dyspnea in some patients. At this stage, the intensivist
is confronted with several options for continued ventilator support: extubation [5],
tracheostomy, or continued conventional ventilation through the endotracheal tube.
Patients with little or no sputum, particularly those with ARDS rather than
pneumonia, and with good neurological function, may be best extubated. This
approach prioritizes preventing ventilator-associated pneumonia by removing the
endotracheal tube; avoiding sedatives; and facilitating communication and mobili-
zation, as described in Chap. 12. Potential downsides include the uncertain conse-
quences of derecruitment on lung healing; the potential impact of spontaneous
breathing on progression of VILI [6]; and the risk of aspiration. Most patients,
however, end up undergoing a tracheostomy. This can usually be done percutane-
ously at the bedside with bronchoscopic control, but surgical tracheostomy is also
used. In either case it is essential that the operator is both experienced and an inte-
gral part of the ECLS team, since bleeding complications are best prevented. If
tracheostomy site bleeding occurs, surgical exploration may be needed. An alterna-
tive approach is to re-intubate while packing the stoma for 48 h. For some patients,
we choose to continue ventilation through a conventional endotracheal tube.
Younger patients who have only been ventilated for 1 or 2 days prior to ECLS may
recover sufciently quickly to avoid a tracheostomy. We also leave the endotra-
cheal tube in patients with severe air leak syndrome and surgical emphysema as a
means to apply high-frequency oscillatory ventilation at a low pressure (mean air-
way pressure of 13 cm H2O and no attempt to recruit), hoping to allow the lung to
heal while minimizing ventilator alarms.
Whichever approach to airway management is chosen, patients are positioned
semi-upright to reduce the risk of ventilator-associated pneumonia and facilitate
ventilation and recruitment of the dependent lung. We employ regular physiother-
apy with bagging, suction, lavage, nebulized medications, and bronchoscopy as
needed to clear secretions. We occasionally manage ECLS patients in the prone
position to augment lung protection; unload a compromised right ventricle [7]; or
aid recruitment when CT demonstrates dependent consolidation. It is our practice to
prone patients for around 18 h in each 24, usually turning prone in the afternoon and
un-proning after morning rounds [8]. Of course, sedating a patient who is awake so
that they can be proned is not very sensible, as an awake, upright patient, breathing
spontaneously and with a good cough is preferred.
Pleural effusions and pleural air should be drained to expand the lung. As with
tracheostomy, experienced operators who are an integral part of the ECLS team
should be chosen to insert chest tubes. It is essential that the pleural space is kept
empty because the lung will not recover if it is surrounded by infected uid or incar-
cerated by brosis. It is misguided to refrain from placing a chest tube because of
fear of bleeding; the result will often be a thoracotomy if the lung becomes trapped
in a collapsed state. In the event of persistent failure of the lung to expand it is likely
that there is either a necrotic segment with persistent bronchopleural stula, or a
collection which will not drain. In either event a CT chest with contrast is helpful in
determining the best treatment. This may be a targeted drain or an extensive formal
exploration of the chest with decortication and resection of the necrotic portion of
the lung [9]. Finally if a lung biopsy is needed on ECLS, it must be done surgically,
not transbronchially, for obvious reasons.
Sedation: Brain, Nerves, Muscles, and Joints
This may seem an eclectic mix but they are all intimately related. When ECLS is
initiated, most patients are paralyzed and receiving infusions of an opiate and a
sedative. Many are on multiple drugs as they are difcult to sedate. We rst stop
the paralytic and, once it has worn off, simplify and reduce the sedative regime. As
with other mechanically ventilated patients, we use a protocol to aid daily interrup-
tion of sedation and allow us to assess neurologic function prior to restarting the
infusions at a reduced dose [10]. It is often necessary to maintain the opiate infusion
to keep the patient comfortable or even increase it as the sedative is withdrawn.
Having the patient comfortable but awake allows a closer eye to be kept on neuro-
logical function as the patient can move and obey commands. Moreover, this is also
important in maintaining movement of joints and recovery of muscles. Physical
therapy, splinting, and passive and active range-of-motion exercises may prevent
foot drop, exion deformities, and other forms of neuromuscular debility. Ultimately,
it is possible to have patients undertaking exercise while on ECLS and even mobi-
lizing out of bed (Chap. 12).
Delirium is common, and usual ICU approaches to prevention are recommended,
including the ABCDE steps (Assess and manage pain; Both daily sedative interrup-
tion and spontaneous breathing trials; Choice of analgesia and sedatives; Delirium
monitoring; and Early mobility) [11]. Some patients may become agitated at night
despite having been normal during the daytime; treatable causes of delirium should
be excluded before resorting to pharmacologic approaches. We strive to modulate
room lighting and staff activity to help re-establish circadian rhythm, and a window
with a view of the sky is invaluable, especially since these patients may be critically
ill for many weeks.
Many patients complain of being cold on ECLS and develop uncontrollable shiv-
ering despite having a core temperature of 37 C. A warm air blanket can provide
considerable symptomatic relief. Of course patients on ECLS are at risk of cerebral
complications both of their disease and of anticoagulation, so we have a low thresh-
old for obtaining a head CT in patients with new or persisting neurological ndings.
188 G.J. Peek
Heart
One of the most important heart-related tasks is to ensure that the non-cardiogenic
pulmonary edema of the patients ARDS is not, in fact, cardiogenic. Endocarditis or
acute valve failure can result in severe respiratory failure and a radiographic appear-
ance identical to ARDS. Obtaining a high-quality echocardiogram (transthoracic or
transoesophageal) is a high priority early in the management of patients on
ECLS. H1N1 inuenza causes a myocarditis in a small proportion of patients, mani-
festing as severe bradycardia but amenable to temporary trans-venous pacing and
lasting for 12 weeks.
Feeding and the Gastrointestinal Tract
We aim to provide full nutrition via an enteral route but, if this is not possible, total
parenteral nutrition (TPN) is used. Normal nasogastric feeding regimes can be used
but great care should be exercised when inserting the nasogastric tube as even a
small abrasion in the throat during insertion has proved fatal. When inserting lines
for TPN, we recommend avoiding the subclavian route unless one is extraordinarily
competent in line insertion, ultrasound guidance, and control of bleeding from the
subclavian vein and artery. Peripherally inserted central lines are ideal for TPN dur-
ing ECLS as it is rarely needed for more than a week or 2. If access for TPN really
is impossible to obtain, the ECLS circuit can be used. Feeding is the best prophy-
laxis against gastric ulceration, but local care bundles to prevent stress ulceration
should be used. Fully awake extubated patents can eat as normal but may benet
from continued enteral feeding to ensure sufcient intake.
Kidneys
Patients will often already be in acute renal failure when referred for ECLS, or may
go on to develop it while on ECLS. These patients should receive CVVH: this can
be done via a standard double-lumen CVVH catheter, allowing the same protocol to
be used in both ECLS and non-ECLS patients. Additional anticoagulation is not
required if heparin is being given to the patient to prevent clotting in the ECLS cir-
cuit. The subclavian route should be avoided as bleeding is very challenging in this
area. If access is impossible the circuit can be used with wide-bore pigtails between
the pump and the oxygenator (to prevent air embolus). In this case, the access pres-
sures on the CVVH machine will need to be recongured to account for the positive
line pressure which is much greater than the default settings on the machine are
used to. In addition, the tap on the red inlet line to the CVVH machine may need to
be turned slightly to reduce the inlet pressure (as this pressure is usually negative
during normal CVVH). As long as the patient is hemodynamically stable, diuresis

or ultraltration is usually indicated to mobilize edema since, in patients with
ARDS, this improves gas exchange, respiratory mechanics, and speeds liberation
from the ventilator [12].
Hematology, Anticoagulation, and Transfusion
Heparin is used to maintain a degree of slight anticoagulation in the ECLS circuit.

It can be given either to the patient (ideal as it reduces the number of access points
on the circuit) or, if access is a problem, to the circuit. The heparin infusion means
that low molecular weight heparin thromboprophylaxis is not required. It is tradi-
tional to monitor heparin dose using the ACT. The exact number of seconds will
depend on the manufacturer of the device and the tubes being used. It is being
increasingly recognized that ACT monitoring is not particularly accurate, but is
familiar to staff and is in widespread use. The usual target ACT is between 160 and
220 s, but there are no data on which to base rm recommendations. The aim is to
give enough heparin to prevent clotting in the circuit without causing bleeding in the
patient. A dose of 13 U/kg/h will usually achieve an activated partial thromboplastin
(aPTT) ratio of about 1.5 to 2; an ACT of 180 s; and a satisfactory prolongation of
the TEG r-time to about twice the heparinase r-time. Many patients, however, will
need much more heparin than this, some as much as 40 U/kg/h (such doses should
prompt measurement of AT-III levels). The concentration of anti-thrombin III is
important for effective heparin anticoagulation, so we monitor this level every other
day and replace as needed. If the patient is bleeding the circuit will run for days on
a xed dose of 5 or 10 U/kg/h without signicant clotting; indeed, heparin can be
stopped completely in the event of uncontrolled bleeding with an expectation that
the circuit will continue to function for a day or 2 as long as high ow is maintained
(at least 2 L/min). Other methods of monitoring heparin effect include TEG, acti-
vated partial thromboplastin time, anti-Xa levels, and heparin levels.
Some centers use argatroban instead of heparin because of the possibility of
heparin-induced thrombocytopenia (HIT) and adjust the infusion according to the
aPTT ratio. In reality HIT is very rare and very obvious (the platelet count will be
close to zero, will not improve with transfusion of platelets, and will be restored
with platelet transfusion after all heparin is discontinued).
There is increasing recognition that maintaining a completely normal hematocrit
during ECLS, as was initially recommended, is not necessary. In many areas of
intensive care medicine, conservative transfusion protocols reduce the total number
of transfusions, adverse events, and (in some groups) mortality. The optimal trans-
fusion target during ECLS has yet to be fully answered, but a level of 10 g/dL seems
clearly safe and some groups transfuse only for values below 7 g/dL as in most other
critically ill patients [13]. To maintain tissue oxygen delivery at low hemoglobin
concentrations requires a commensurate increase in cardiac output, a fact one
should bear in mind when exploring the outer reaches of the tissue oxygen delivery
190 G.J. Peek
envelope. If lactic acidosis, tachycardia, and reduced tissue oxygen saturation (by
near-infrared spectroscopy) remain unresponsive to treatment, it is reasonable to
transfuse, even when the hemoglobin concentration is above 8 g/dL.
To control bleeding on ECLS, especially when surgery is required, antibrinolyt-
ics such as aprotinin, tranexamic acid, and epsilon-aminocaproic acid can be useful
adjuncts. We transfuse fresh frozen plasma, cryoprecipitate, and platelets aiming to
maintain values of INR <1.5, brinogen >200 mg/dL, and platelet count >150/mm3
when patients are bleeding but these values can be relaxed somewhat when patients
are not bleeding.
Phase III: Improvement
One hopes that clear signs of improvement are seen after 9 days or so of
ECLS. Recovery may be signaled through better gas exchange (for oxygen or CO2),
radiographic clearing, improving respiratory mechanics (increased tidal volumes
for any given pressure; or falling pressures for any given tidal volume), or general
hemodynamic stability (see Chap. 13). Even if there is no clinical improvement
after 9 days, excellent outcomes are still possible. Many patients simply take longer
to recover. The chest radiograph often lags behind other signs of improvement and
so should not be relied on in isolation. If a month passes with no apparent healing,
a stringent search for non-reversible causes should be undertaken such as a full body
CT scan, bone marrow aspiration, repeat echocardiography (and estimation of PA
pressure), and possibly a lung biopsy. If all of these investigations are unrevealing
and there are no other extenuating factors, further patience is recommended.
If the patient is ready to trial off ECLS, the lung mechanics will have improved to
the point where the tidal volume is around the 4 mL/kg on rest settings and gas
exchange is clearly improving. In order to test whether the patient is ready to come
off ECLS a trial is conducted as described in Chap. 13, generally by returning to
normal lung-protective ventilator settings, then turning the sweep gas ow off. When
this is well-tolerated, the patient can be decannulated. For patients extubated while
on ECLS, the criteria for passing a trial off support are even simpler: they are identi-
cal to those for judging a spontaneous breathing trial [14]. If there is uncertainty it is
perfectly reasonable to prolong the trial off for several more hours or even overnight
as patients often slowly improve as they re-establish spontaneous breathing.
Conclusion
For successful ECLS care, it is important to view the patients management as a

coherent package rather than ECLS being a bolt-on accessory to the ventilator. This
will only be achieved when the care of the patient is supervised by an experienced
ECLS team communicating fully with the intensivist, surgeon, nurses, therapists
and others who have invested their efforts in this demanding task.
References
1. Butt W, Heard M, Peek GJ. Clinical management of the extracorporeal membrane oxygenation
circuit. Pediatr Crit Care Med. 2013;14(5 Suppl 1):S139.
2. Brower RG, Matthay MA, Morris A, et al. 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;342:13018.
3. Peek GJ, Mugford M, Tiruvoipati R, et al. Efcacy and economic assessment of conventional
ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory
failure (CESAR): a multicentre randomised controlled trial. Lancet. 2009;374:135163.
4. Schmidt GA. Cardiopulmonary interactions in acute lung injury. Curr Opin Crit Care.
2013;19:516.
5. Hoeper MM, Wiesner O, Hadem J, et al. Extracorporeal membrane oxygenation instead of
invasive mechanical ventilation in patients with acute respiratory distress syndrome. Intensive
Care Med. 2013;39:20567.
6. Yoshida T, Torsani V, Gomes S, et al. Spontaneous effort causes occult pendelluft during
mechanical ventilation. Am J Respir Crit Care Med. 2013;188:14207.
7. Jozwiak M, Teboul J-L, Anguel N, et al. Benecial hemodynamic effects of prone positioning
in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med.
2013;188:142833.
8. Gurin C, Reignier J, Richard J-C, et al. Prone positioning in severe acute respiratory distress
syndrome. N Engl J Med. 2013;368:215968.
9. Joshi V, Harvey C, Nakas A, et al. The need for thoracic surgery in adult patients receiving
extracorporeal membrane oxygenation: a 16-year experience. Perfusion. 2013;28:32832.
10. Kress JP, Pohlman AS, OConnor MF, Hall JB. Daily interruption of sedative infusions in criti-
cally ill patients undergoing mechanical ventilation. N Engl J Med. 2000;342:14717.
11. Pandharipande P, Banerjee A, McGrane S, et al. Liberation and animation for ventilated ICU
patients: the ABCDE bundle for the back-end of critical care. Crit Care. 2010;14:1579.
12. Wiedemann HP, Wheeler AP, Bernard GB, et al. Comparison of two uid-management strate-
gies in acute lung injury. N Engl J Med. 2006;354:256475.
13. Hbert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial
of transfusion requirements in critical care. N Engl J Med. 1999;340:40917.
14. Ely EW, Baker AM, Dunagan DP, et al. Effect on the duration of mechanical ventilation of
identifying patients capable of breathing spontaneously. N Engl J Med. 1996;335:18649.
Chapter 11
Crises During ECLS
Cara L. Agerstrand, Linda B. Mongero, Darryl Abrams,

Matthew Bacchetta, and Daniel Brodie
Introduction
The use and success of extracorporeal life support (ECLS) has historically been
limited by frequent and potentially severe complications. Increased experience
among providers and improvements in circuit technology have reduced, but not
eliminated, these complications in the modern era of ECLS. Accurate estimates of
current complication rates are difcult to quantify. Published rates and those found
in registries typically reect the use of both older and newer technology and prac-
tice, tending to overestimate the current risks. The Extracorporeal Life Support
Organization (ELSO), which currently compiles the most extensive database on
complications related to ECLS, includes ECLS cases with both modern and out-
dated technology, as well as centers with greater and lesser degrees of experience
C.L. Agerstrand, MD (*)

Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine,
Columbia University College of Physicians and Surgeons/New York-Presbyterian Hospital,
622 West 168th Street, PH 8-101, New York, NY 10032, USA
e-mail: Ca2264@columbia.edu
L.B. Mongero, CCP, BS
Department of Clinical Perfusion, New York Presbyterian-Columbia University Medical
Center, 4 Brook Row, Locust Valley, NY 11560, USA
e-mail: mongero@nyp.org
D. Abrams, MD D. Brodie, MD
Department of Medicine, New York-Presbyterian Hospital/Columbia University Medical
Center, 622 West 168th Street, PH 8 East, Room 101, New York, NY 10032, USA
e-mail: da2256@columbia.edu; hdb5@columbia.edu
M. Bacchetta, MD, MBA, MA
Department of Surgery, New York-Presbyterian Hospital/Columbia University Medical
Center, 161 Fort Washington Ave, Room 336, New York, NY 10032, USA
e-mail: mb781@columbia.edu

194 C.L. Agerstrand et al.
with ECLS using a variety of ECLS congurations. Determining the expected rate
of complications in any given circumstance should be attempted with caution.
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.
Circuit Crises
Circuit Air
Any loss of circuit integrity may result in the entrainment of air into the circuit. Air
in the circuit is an emergency situation that must be addressed immediately; how-
ever, the risk to the patient is highest when air is present on the post-oxygenator side
of the circuit since the oxygenator itself may serve as an air trap. In venovenous
ECLS, circuit air risks venous air embolism. In venoarterial ECLS or venovenous
ECLS in the presence of an atrial septal defect or patent foramen ovale, there is a
risk of arterial air embolism.
Pre-oxygenator air can originate from peripheral or central venous lines that are
inadvertently left open or accessed for intravenous infusions. Because circuit pressure
is negative between the outow cannula and the pump, cracked stopcocks or tubing, or
loose connections can entrain air, as can partial dislodgement of the drainage cannula
resulting in exposure of a side port. Air can arise within the oxygenator itself from
improper priming, oxygenator membrane rupture, or occlusion of the gas exhaust out-
let. Post-oxygenator air is seen if air passes through the oxygenator from the pre-ox
ygenator side or is directly entrained through defects in the circuit after the
oxygenator.
Another mechanism through which air can enter the circuit is known as cavita-
tion, which occurs when a markedly negative pressure within the venous drainage
cannula and tubing draws a large amount of gas out of solution. Extreme changes in
pressure induce hydrodynamic phenomena, including bubble nucleation, growth,
and collapse. This process can be minimized by avoiding extremes in negative pres-
sure on the venous side of the circuit. Cavitation can also occur in the setting of a
high sweep gas ow rate and a low blood ow rate.
In order to minimize introduction of air into the circuit, all medication and blood
products should be given directly to the patient through separate venous access. If
the circuit must be used, the pre-oxygenator limb should be accessed so that the
oxygenator can trap any air inadvertently introduced. A bubble detector may also be
placed on the post-oxygenator tubing, which can be set to stop the pump immedi-
ately if air is detected.
Small amounts of air on the venous (outow) side can often be removed without
ceasing ECLS by carefully withdrawing the air into a syringe; however, 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 the circuit and
isolating it from the patient. When air appears on the post-oxygenator side, the out-
11 Crises During ECLS 195
Fig. 11.1 Appropriate clamp position when isolating patient from circuit. During an ECLS crises in
which the integrity of the circuit is at risk or in question the patient may be separated from the circuit
by clamping the venous and arterial limbs. Reprinted with permission from CollectedMed.com
ow and inow lines should be clamped immediately, thereby halting ECLS

(Fig. 11.1), and the ventilator adjusted to provide 100 % oxygen and sufcient ven-
tilation until ECLS can be reinstituted. The patient should be placed in the
Trendelenburg position so that any embolization of air is directed toward the lower
extremities.
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 ows 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 uid is not administered through a second access
point. When large volumes are withdrawn, signicantly negative pressure risks
hemolysis 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 signi-
cant amount of air gains access to the oxygenator, it may have to be re-primed.
A cardiotomy reservoir spliced into the circuit can allow the circuit to be re-primed
more efciently. In the event the circuit cannot be de-aired quickly and easily, the
oxygenator, or possibly the entire circuit, may need to be replaced.
Table 11.1 Circuit-related emergencies and management

Circuit crises Diagnostics Management
Circuit air
Pre-oxygenator Inspect for defects in Isolate air within tubing at closest
pre-oxygenator tubing and access point
access points Aspirate air through syringe
Check drainage cannula for Replace blood volume removed
exposed drainage ports during aspiration
Repair or replace circuit defects
Oxygenator Inspect for defects in Remove oxygenator from its
pre-oxygenator tubing and holder
access points Tap oxygenator to isolate air
Check drainage cannula for superiorly
exposed drainage ports Aspirate air through syringe
Replace blood volume removed
during aspiration
If large amount of air, the
oxygenator may need to be
re-primed or replaced
Post-oxygenator Inspect for defects in Clamp arterial and venous limbs
pre-oxygenator, oxygenator, of the circuit
and post-oxygenator tubing Reduce RPMs
and access points Place patient on emergency
Check drainage cannula for ventilator settings
exposed drainage ports Isolate air within tubing at closest
access point
Aspirate air through syringe
Replace blood volume removed
during aspiration
Thrombosis Assess oxygenator function; Ensure adequate systemic
follow pre- and post- anticoagulation
oxygenator PO2 and pressures Treat patient causes of
Assess for patient causes of hypercoagulability
hypercoagulability If HITT suspected
Stop heparin
Start direct thrombin inhibitor
Measure platelet factor-4
antibody and conrm with
SRA
Evaluate for patient
thrombosis
Oxygenator failure Assess oxygenator function; Clamp arterial and venous limbs
follow pre- and post- of the circuit
oxygenator PO2 and pressures Place patient on emergency
ventilator settings
Replace oxygenator
Pump failure Check power supply Initiate hand-cranking
Check pump head integrity Replace power supply if
Assess for inadequate preload necessary
and/or excessive afterload Replace pump head if necessary
If no cause identied, consider
replacing motor
(continued)
Table 11.1 (continued)

Circuit crises Diagnostics Management
Tubing rupture/cannula fracture
Pre-pump Identify site of rupture Clamp arterial and venous limbs
of the circuit
Reduce RPMs
Place patient on emergency
ventilator settings
Repair tubing at site of defect
Post-pump Identify site of rupture Clamp arterial and venous limbs
of the circuit
Reduce RPMs
Place patient on emergency
ventilator settings
Replace tubing at site of defect
Heat exchanger Assess for water in the Replace heat exchanger
malfunction oxygenator or blood in the Replace oxygenator if evidence
water bath of water or blood leakage
Inadvertent decannulation
Partial Assess cannula for exposure Reinforce cannula with sutures
dislodgement of drainage port If drainage port exposed, replace
with preserved cannula
blood ow Do not reintroduce exposed
portion of cannula
Complete Clamp arterial and venous limbs
dislodgement or of the circuit
partial Reduce RPMs
dislodgement Place patient on emergency
with ventilator settings
compromised Control cannulation site bleeding
blood ow Discontinue anticoagulation
Replace cannulae
RPMs revolutions per minute, PO2 partial pressure of oxygen, HITT heparin-induced thrombocy-
topenia with thrombosis, SRA serotonin release assay
If air reaches the patient, management is mostly supportive, although venous air
may be aspirated under some circumstances. (Table 11.1 highlights diagnosis and
management of circuit-related crises.)
Thrombosis
The interaction between blood and the ECLS circuit may alter both procoagulant
and anticoagulant pathways, potentially resulting in thrombosis in the circuit or the
patient. Thrombosis in the ECLS circuit appears as dark red clot or white strands of
Fig. 11.2 Picture of blood and brin clot in oxygenator. Arrows depict blood and brin clot forma-
tion on the oxygenator. Reprinted with permission from CollectedMed.com
brin (Fig. 11.2). Inadequate anticoagulation, disseminated intravascular coagu-

lopathy, heparin-induced thrombocytopenia, anti-thrombin III deciency, and other
hypercoagulable states may further increase risk of circuit thrombosis, especially at
sites of low or turbulent ow. A simplied circuit with only the essential compo-
nents 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 [1, 2]. Although there are no univer-
sally accepted anticoagulation guidelines, the administration of heparin with a goal
activated partial thromboplastin time of 4060 s has been suggested as sufcient to
maintain circuit patency [3, 4].
If thrombosis has developed in the setting of thrombocytopenia, a diagnosis of
heparin-induced thrombocytopenia with thrombosis should be considered [5]. If
there is a reasonable suspicion for heparin-induced thrombocytopenia with throm-
bosis, heparin should be stopped immediately and anticoagulation continued with a
direct thrombin inhibitor [6, 7]. Measurement of platelet factor-4 antibody levels is
the initial serologic screening test, which, if positive, should be followed by a sero-
tonin release assay [8].
Oxygenator Failure
Polymethylpentene hollow ber oxygenators are more efcient and durable than
older polypropylene or silicone oxygenators. Despite improvements in durability,
oxygenator effectiveness is critical to the successful management of ECLS and its
function should be assessed frequently. A decreasing post-oxygenator PO2, an
increasing trans-membrane pressure gradient, or a need to steadily increase sweep
gas ow rate to manage PaCO2 may be indications of a failing oxygenator.
Oxygenator thrombosis is a common cause of an increase in the trans-membrane
pressure gradient and oxygenator malfunction. This phenomenon can be assessed
by visual inspection of the oxygenator during a circuit check. If the pressure gradi-
ent increases over time, a signicant amount of thrombus may be developing.
Raising the level of anticoagulation may stabilize the membrane but, otherwise, it
will need replacement.
Membrane leakage, with blood passing across the membrane and into the gas
exhaust line, is another potential complication resulting in oxygenator failure. If
blood is found in the gas exhaust line, the oxygenator should be replaced. However,
because oxygenator replacement is not without risk, close monitoring alone may be
considered if the leakage is felt to be minimal. The gas exhaust outlet must remain
vented to the atmosphere; otherwise, signicant increase in oxygenator gas pressure
could place the patient at risk for air embolism across the compromised membrane.
Oxygenator replacement requires that the patient temporarily be removed from
ECLS. Experienced teams can change out an oxygenator in approximately 30 s.
However, this takes careful coordination and planning. Patients should be placed on
emergency ventilator settings and the pump turned off or to a very low ow rate.
Vasopressors and inotropic agents should be immediately available to manage any
hemodynamic compromise. The circuit is clamped near the inow and outow
ports of the oxygenator, which is then removed and replaced. Gas and heater lines
are connected to the new oxygenator and ECLS support is resumed.
Pump Failure
Pump failure is uncommon but has been reported [9]. If pump failure occurs for any
reason, a patient who is dependent on ECLS should be placed on emergency ventila-
tor settings. During VA ECLS it is imperative to clamp the circuit to prevent reversal
of ow from the patients arterial system through the ECLS circuit into the venous
system (a large arteriovenous shunt), leading to cardiopulmonary collapse. A hand
crank, which should be kept alongside ECLS circuits, is used to generate blood ow
while the source of the problem is identied and corrected. If disruption of the elec-
trical power supply is the cause of pump failure, then battery power or a backup
power supply should be obtained. If the ECLS circuit was inadvertently powered
off, it should be restarted immediately with an assessment of the alarm functions.
If the circuit has adequate power supply and the pump is rotating, but there is no
ow, 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 conrmed by thor-
ough visual inspection, assessing for cracks or thromboses. If the pump head is
intact, it may be placed back onto the motor and the pump restarted. In the event
there is still no ow, the motor should be replaced.
Tubing Rupture/Cannula Fracture
Tubing rupture or cannula fracture constitutes an emergency and can result in exsan-
guination or entrainment of air into the circuit and patient. Risk of exsanguination
is greatest if the rupture or fracture occurs post-pump, where the circuit is under
positive pressure. The risk of air entrainment is greatest when tubing rupture or can-
nula fracture occurs pre-pump, under negative pressure conditions. For post-pump
rupture or fracture, the circuit should be immediately clamped off, which temporar-
ily discontinues ECLS and requires placing the patient on emergency ventilator
settings, while the defective tubing or cannula is isolated and replaced. For small
cracks in the tubing or cannula that occur pre-pump, it may be possible to continue
ECLS support while repairing the problem if the patient is on VV-ECMO and there
is no atrial septal defect or other potential right-to-left circulatory shunt.
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 ber 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. Increased tension on the cannula and sutures may occur during patient trans-
port or during any procedures when the patient or 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 reintroduced
into the patient as this could result in vascular injury or infection.
Complete decannulation or partial decannulation to the point of compromised
blood ow is a surgical emergency. If complete decannulation occurs, the remaining
cannulae should be clamped immediately to prevent exsanguination and the patient
should come off ECLS support and be placed on emergency ventilator settings.
Direct, rm pressure should be applied to the vascular site and anticoagulation
should be discontinued.
Education and in-servicing of staff not familiar with ECLS may help reduce the
risk of inadvertent decannulation. Furthermore, the securing sutures on the cannulae
should be checked on a routine basis to ensure their integrity. An experienced pro-
vider should monitor the patient and circuit during radiologic or other bedside pro-
cedures 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, at times, in up to 3040 % of cases [9]. The etiology and severity of
ECLS-related bleeding varies from minor oozing at surgical sites to fatal intracere-
bral hemorrhage; however the risk of severe hemorrhage appears to have decreased
over time (Table 11.2). 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
Table 11.2 Hemorrhagic Bleeding site Frequency (%)

complications during ECLS
Cannulation-site 15.7
Surgical-site 14.6
Pulmonary 7.6
Gastrointestinal 5.5
Intracranial 4.0
Hemorrhagic complications during use of
ECLS for adult respiratory failure, from the
Extracorporeal Life Support Organization
(ELSO) ECLS Registry Report, International
Summary, July 2013, based on self-reporting
from ELSO member centers
50 % of patients and transfusion requirements of greater than 6 units per day in

select series [10, 11].
Modern ECLS circuitry and evolving approaches to patient management, par-
ticularly regarding the amount of anticoagulation used during ECLS, have reduced
both the risk of hemorrhage and the need for transfusion of packed red blood cells
(pRBCs) in patients receiving extracorporeal support. Two cohorts that used
modern ECLS technology during the inuenza A(H1N1) pandemic reported vary-
ing degrees of hemorrhage, with one study reporting that 71 % of deaths were
bleeding-related, whereas a second study reported only one bleeding-related
death in a similarly sized cohort [1214]. The only modern randomized controlled
trial of ECLS in severe forms of the acute respiratory distress syndrome (ARDS),
known as the Efcacy and economic assessment of conventional ventilator sup-
port versus extracorporeal membrane oxygenation for severe adult respiratory
failure (CESAR) trial, which included adult patients with ARDS of any etiology,
reported only one death related to hemorrhage out of 68 subjects who received
ECLS in the trial [15].
Bleeding during ECLS may be difcult to manage due to both the potential coag-
ulopathy induced by the ECLS circuit and the need to provide systemic
anticoagulation to prevent thrombus formation within the circuit. Some degree of
thrombocytopenia is common in the setting of ECLS. However, clinically signi-
cant thrombocytopenia due solely to platelet interactions with the ECLS circuit may
be less common. Modern circuits are more biocompatible, requiring a lesser degree
of anticoagulation than older ECLS technology, thereby reducing the risk and sever-
ity of bleeding. Concurrently, evolving trends in critical care have reduced the
hemoglobin threshold at which transfusions are thought to be necessary [1618].
Therefore, while transfusions remain common during ECLS support, and severe or
hemodynamically unstable bleeding requires transfusion of pRBCs, it may not be
necessary to transfuse for mild bleeding or oozing for the sake of maintaining a
hemoglobin level in the normal range as has been traditionally practiced at many
ECLS centers [19].
The optimal hemoglobin for patients receiving ECLS is not known and practice
varies widely by center. Abundant evidence shows that liberal transfusion strategies
targeting higher hemoglobin values result in increased morbidity and mortality in
critically ill medical and surgical patients when compared with a more conservative
approach [16, 17, 20, 21]. This includes higher rates of transfusion-related acute
lung injury, worsened ARDS, volume overload, infections, sepsis, poor wound
healing, and intensive care unit, hospital and 1-year mortality [20, 22, 23].
Transfusion-associated morbidity extends to patients receiving even small amounts
of blood [24]. Adverse events associated with transfusions increase in a dose-
dependent manner and provide evidence suggesting that even as little as 12 units
of blood may increase morbidity and mortality considerably [25].
A comparison between liberal and conservative transfusion protocols has not
been conducted in patients receiving ECLS. However, 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 bene-
cial to lower the transfusion threshold in patients receiving ECLS and tolerate a
moderate degree of anemia [2628]. Similar to our practice with other critically ill
patients, our protocol 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 donorrecipient cross-matching.
Thrombocytopenia also is a common complication of ECLS, with platelet activa-
tion and consumption potentially resulting in a notable amount of thrombocytopenia
over the rst 45 days of ECLS support [2932]. 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 undergoing
an invasive procedure. An exception to this approach may be in patients with severe
pulmonary hypertension, in whom the risk of platelet sequestration in the pulmo-
nary vasculature, which can precipitate pulmonary hypertensive crises, may out-
weigh the benet 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
biocompatible and less prone to thrombosis, low levels of systemic anticoagulation
may be administered without a signicant 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 undetermined
periods of time without clinically signicant thrombosis, although anticoagulation
should be resumed as soon as possible [33]. Fresh frozen plasma can be adminis-
tered as needed. Fibrinogen levels may be maintained above 100 mg/dL with the
transfusion of cryoprecipitate.
If bleeding from surgical or cannulation sites 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 denitive interventions.
Uncontrollable, life-threatening bleeding has been treated successfully with anti-
brinolytics such as aminocaproic acid (AMICAR) and tranexamic acid [3436].
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 [3740]. However, with any of these agents, the risk of oxygenator and
circuit thromboses, including fatal clots, may be increased [41, 42]. These agents
should only be considered in life-threatening situations when other measures have fail-
ure. Prior to using of any of these therapies, consultation with a hematologist is recom-
mended. (Table 11.3 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
Table 11.3 Patient-related emergencies and management

Patient crises Diagnostics Management
Bleeding
Mildmoderate Monitor hemoglobin Local bleeding control
Check platelets, PT, aPTT, Maintain target hemoglobin
brinogen (typically 7.0 g/dL)
Maintain platelet count 50,000/mL
Transfuse cryoprecipitate for
brinogen <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 platelets, PT, aPTT, Maintain target hemoglobin
brinogen (typically 7.0 g/dL)
Maintain platelet count 50,000/mL
Discontinue anticoagulation
Transfuse FFP for elevated PT
Transfuse cryoprecipitate for
brinogen <100 mg/dL
Administer DDAVP if concern for
platelet dysfunction
Discontinue anticoagulation
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
bilirubin, free hemoglobin negative than 50 mmHg
Check haptoglobin and Minimize circuit access points
urinalysis Correct tubing kinks
Assess for tubing kinks Replace heat exchanger if
Assess for circuit malfunctioning
thromboses Correct patient-centered causes of
Check heat exchanger hemolysis
temperature
Assess for cause of
hemolysis in the patient
Refractory Assess for recirculation; If recirculation present, consider
hypoxemia monitor pre-oxygenator PO2 reducing pump RPMs or
Assess oxygenator function; repositioning cannulae
check pre- and post- If oxygenator failure, replace
oxygenator PO2 and oxygenator
pressures Increase blood ow rate, augment
Assess circuit blood ow maximal ow with additional venous
rate drainage cannula
Compare upper and lower If femoral venoarterial conguration
body PaO2 and discrepancy in upper and lower
body oxygenation, consider addition
of venous reinfusion cannula to
upper body
(continued)
Table 11.3 (continued)

Patient crises Diagnostics Management
Shock
Low-output state Check TTE If venovenous ECLS, consider
CVP conversion to venoarterial
Central or mixed venous conguration
oxygen saturation If venoarterial ECLS, augment
Lactate circuit blood ow, and/or consider
Other diagnostics as per other mechanical cardiac support
routine care Consider the use of inotropes
High-output Check TTE Augment ECLS blood ow rate
state CVP, central or mixed
venous oxygen saturation
Lactate
Other diagnostics as per
routine care
PT prothrombin time, aPTT activated partial thromboplastin time, DDAVP desmopressin, FFP
fresh frozen plasma, LDH lactate dehydrogenase, PO2 partial pressure of oxygen, PaO2 partial
pressure of oxygen in arterial blood, RPMs revolutions per minute, TTE transthoracic echocardio-
gram, CVP central venous pressure, ECLS extracorporeal life support
[4348]. Despite these advances, hemolysis can still occur from extremes in nega-
tive pressure in the venous drainage cannula and tubing, typically when pressures
exceed 100 mmHg (and perhaps even from pressures simply exceeding
50 mmHg), resulting in cavitation and red blood cell destruction. Hemolysis can
occur at sites of turbulent, non-laminar blood ow in the circuit or patient, includ-
ing at points where tubing may be kinked or intentionally split, or where vascular
access points or hemoltration devices are connected to the circuit. In situ throm-
boses can also cause non-laminar ow and result in hemolysis, as can overheated
water baths.
The components of any ECLS circuit depend on institutional preference, how-
ever, a simplied circuit without a bridge may reduce the rate of hemolysis.
Likewise, externally applied, non-invasive monitors can reduce the necessity for
access points, further minimizing the amount of turbulent ow. The use of a sim-
plied 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,
signicant hemolysis does not typically occur with the use of a simplied circuit
design.
If severe hemolysis is identied that cannot be attributed to kinks in cannula
tubing or a systemic process in the patient, and cannot be corrected by reducing
the temperature of the water bath, then the circuit or pump head may need to be
replaced. With any occurrence, a source of non-circuit-related hemolysis should
be excluded clinically.
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. Factors that determine the relative contribution of the circuit to oxygen-
ation include the severity of lung disease, the patients native cardiac output, the
amount of recirculation, the rate of extracorporeal blood ow, the membranes gas
transfer efciency, oxygen extraction in the patients tissues, and the metabolic
demand of the patient.
Accurate estimation of blood ow and oxygen requirements is essential prior to
ECLS initiation in order to choose properly sized cannulae to meet the patients
metabolic demand. While this calculation can be estimated by body surface area
(BSA), patient-specic physiologic characteristics, such as pregnancy, vasodilatory
shock, or other high cardiac output states, may necessitate blood ows that exceed
those calculated by BSA [49]. In venovenous ECLS, the addition of a second venous
drainage cannula may be used to augment blood ow if the initial conguration can-
not achieve adequate blood ow rates.
In venoarterial support, most commonly via the femoral vein and artery, rein-
fused, oxygenated blood ows retrograde in the aorta, where it mixes with blood
ejected from the left ventricle. Depending on the relative contribution of circuit
blood ow and native cardiac output, regional perfusion and ow differences can
occur, especially when cardiac dysfunction is not severe. In these circumstances,
the lower body receives a relatively higher proportion of the reinfused, oxygenated
blood, while the upper body, including the carotid and coronary arterial beds, may
receive a relatively lower proportion of the reinfused blood. For this reason, sys-
temic 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 ow rate may improve delivery of oxygenated blood to the aortic arch.
Alternatively, the addition of a second reinfusion cannula into an internal jugular
vein, thereby creating a venoarterial-venous (VAV) conguration, can improve
oxygen delivery to the upper body by utilizing the patients native cardiac function
to deliver reinfused, oxygenated blood into the ascending aorta (see Chap. 6). A
venoarterial ECLS conguration via the axillary or subclavian artery may preclude
the need for VAV conversion, though this approach is more technically difcult and
not suitable for emergent situations [50].
Hypoxemia can also result from circuit malfunction, disconnection of the gas line
tubing, recirculation, or a failing oxygenator. Recirculation, in which reinfused, oxy-
genated blood is returned to the circuit without having passed through the patient, is
more common in the traditional two-site venovenous ECLS conguration, especially
when the drainage and reinfusion ports are in close proximity to one another or the
pump ows are too high and the venous cannula drains a disproportionate amount of
reinfused blood. Recirculation should be suspected when a high pre-oxygenator PO2
is observed. However, clinical judgment must be used to distinguish whether the
high pre-oxygenator PO2 is due to recirculation or whether it is a consequence of

either pulmonary recovery or poor-tissue oxygen extraction in the setting of sepsis.
Repositioning of the cannulae or use of a bicaval dual-lumen cannula may reduce
recirculation [51]. Oxygenator function can be assessed by monitoring pre- and post-
oxygenator blood gases and the pressure gradient across the oxygenator. Provided
that the blood ow is less than the rated ow of the oxygenator, the post-oxygenator
oxygen saturation should be greater than 95 % in a functioning oxygenator.
Shock
Management of shock states during venovenous ECLS is treated the same as shock
without ECLS because venovenous congurations do not provide circulatory sup-
port. If cardiogenic shock develops in the setting of venovenous ECLS, conversion
to venoarterial ECLS, which provides circulatory support, should be considered,
although the upper body oxygenation issues with femoral arterial reinfusion men-
tioned above, must be kept in mind. Inotropic medications may also be considered.
If a high-output shock state develops on ECLS of any conguration, the extracor-
poreal blood ow rate will account for a lower percentage of the patients cardiac
output. Consequently, the proportion of blood being oxygenated by the circuit will
be lower, which may 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 for adult patients with septic shock in order to augment
cardiopulmonary function in the setting of vasodilation, reduced cardiac output, and
high metabolic demand. Judicious patient selection is paramount to maximize the
benet of ECLS in the setting of shock [5256].
Conclusion
Successful application of ECLS requires expeditious identication and intervention

during both patient and circuit-related crises. Through clinical instruction and water
drills, ECLS centers must train providers to recognize and respond to both common and
uncommon complications in order to maximize patient safety and improve outcomes.
References
1. Kopp R, Mottaghy K, Kirschnk M. Mechanism of complement activation during extracorpo-

real blood-biomaterial interaction: effects of heparin coated and uncoated surfaces. ASAIO J.
2002;48(6):598605.
2. Deptula J, Glogowski K, Merrigan K, Hanson K, Felix D, Hammel J, et al. Evaluation of bio-
compatible cardiopulmonary bypass circuit use during pediatric open heart surgery. J Extra
Corpor Technol. 2006;38(1):226.
3. Combes A, Bacchetta M, Brodie D, Muller T, Pellegrino V. Extracorporeal membrane oxygen-

ation for respiratory failure in adults. Curr Opin Crit Care. 2012;18(1):99104.
Med. 2011;365(20):190514.
5. 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):512.
6. Koster A, Weng Y, Bottcher W, Gromann T, Kuppe H, Hetzer R. Successful use of bivalirudin
as anticoagulant for ECMO in a patient with acute HIT. Ann Thorac Surg. 2007;83(5):
18657.
7. Lee GM, Arepally GM. Diagnosis and management of heparin-induced thrombocytopenia.
Hematol Oncol Clin North Am. 2013;27(3):54163.
8. Warkentin TE, Greinacher A, Gruel Y, Aster RH, Chong BH. Laboratory testing for heparin-
induced thrombocytopenia: a conceptual framework and implications for diagnosis. J Thromb
Haemost. 2011;9(12):2498500.
9. Extracorporeal Life Support Organization. ECLS registry report, international summary.
January 2013.
10. Sell LL, Cullen ML, Whittlesey GC, Yedlin ST, Philippart AI, Bedard MP, et al. Hemorrhagic
complications during extracorporeal membrane oxygenation: prevention and treatment. J
Pediatr Surg. 1986;21(12):108791.
11. Gattinoni L, Pesenti A, Mascheroni D, Marcolin R, Fumagalli R, Rossi F, et al. Low-frequency
positive-pressure ventilation with extracorporeal CO removal in severe acute respiratory fail-
ure. JAMA. 1986;256(7):8816.
12. Davies A, Jones D, Bailey M, Beca J, Bellomo R, Blackwell N, et al. Extracorporeal mem-
brane oxygenation for 2009 inuenza A(H1N1) acute respiratory distress syndrome. JAMA.
2009;302(17):188895.
13. Davies A, Jones D, Gattas D. Extracorporeal membrane oxygenation for ARDS due to 2009
inuenza A(H1N1)reply. JAMA. 2010;303(10):9412.
14. Patroniti N, Zangrillo A, Pappalardo F, Peris A, Cianchi G, Braschi A, et al. The Italian ECMO
network experience during the 2009 inuenza A(H1N1) pandemic: preparation for severe
respiratory emergency outbreaks. Intensive Care Med. 2011;37(9):144757.
trolled trial. Lancet. 2009;374(9698):135163.
16. Vincent JL, Baron JF, Reinhart K, Gattinoni L, Thijs L, Webb A, et al. Anemia and blood
transfusion in critically ill patients. JAMA. 2002;288(12):1499507.
17. Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. A multicenter,
randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion
Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J
Med. 1999;340(6):40917.
18. Hebert PC, Fergusson DA. Red blood cell transfusions in critically ill patients. JAMA.
2002;288(12):15256.
19. Butch SH, Kna P, Oberman HA, Bartlett RH. Blood utilization in adult patients undergoing
extracorporeal membrane oxygenated therapy. Transfusion. 1996;36(1):613.
20. Koch CG, Li L, Sessler DI, Figueroa P, Hoeltge GA, Mihaljevic T, et al. Duration of red-cell
storage and complications after cardiac surgery. N Engl J Med. 2008;358(12):122939.
21.Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Abraham E, et al. The CRIT study:
anemia and blood transfusion in the critically illcurrent clinical practice in the United States.
Crit Care Med. 2004;32(1):3952.
22. Marik PE, Corwin HL. Acute lung injury following blood transfusion: expanding the deni-
tion. Crit Care Med. 2008;36(11):30804.
23. Shorr AF, Duh MS, Kelly KM, Kollef MH. Red blood cell transfusion and ventilator-associ-
ated pneumonia: a potential link? Crit Care Med. 2004;32(3):66674.
24. 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(6):167480. discussion 1680.
25. 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. 2012;147(1):
4955.
26. Simon T, Snyder EL, Solheim BG. Rossis principles of transfusion medicine. 4th ed. West
Sussex, UK: Wiley-Blackwell; 2009.
27. Roback JD, Neuman RB, Quyyumi A, Suliff R. Insufcient nitric oxide bioavailability: a
hypothesis to explain adverse effects of red blood cell transfusion. Transfusion. 2011;51(4):
85966.
28. Carson JL, Grossman BJ, Kleinman S, Tinmouth AT, Marques MB, Fung MK, et al. Red blood
cell transfusion: a clinical practice guideline from the AABB. Ann Intern Med. 2012;157(1):
4958.
29. Plotz FB, van Oeveren W, Bartlett RH, Wildevuur CR. Blood activation during neonatal extra-
corporeal life support. J Thorac Cardiovasc Surg. 1993;105(5):82332.
30. Anderson III HL, Cilley RE, Zwischenberger JB, Bartlett RH. Thrombocytopenia in neonates
after extracorporeal membrane oxygenation. ASAIO Trans. 1986;32(1):5347.
31.Robinson TM, Kickler TS, Walker LK, Ness P, Bell W. Effect of extracorporeal membrane
oxygenation on platelets in newborns. Crit Care Med. 1993;21(7):102934.
32. Stallion A, Cofer BR, Rafferty JA, Ziegler MM, Ryckman FC. The signicant relationship
between platelet count and haemorrhagic complications on ECMO. Perfusion. 1994;9(4):
2659.
33. Lamb KM, Cowan SW, Evans N, Pitcher H, Moritz T, Lazar M, et al. Successful management
of bleeding complications in patients supported with extracorporeal membrane oxygenation
with primary respiratory failure. Perfusion. 2013;28(2):12531.
34. Downard CD, Betit P, Chang RW, Garza JJ, Arnold JH, Wilson JM. Impact of AMICAR on
hemorrhagic complications of ECMO: a ten-year review. J Pediatr Surg. 2003;38(8):12126.
35. 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(4):53640. discussion 5401.
36. van der Staak FH, de Haan AF, 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(4):5949.
37. Velik-Salchner C, Sergi C, Fries D, Moser P, Streif W, Kolbitsch C. Use of recombinant factor
VIIa (novoseven) in combination with other coagulation products led to a thrombotic occlu-
sion of the truncus brachiocephalicus in a neonate supported by extracorporal membrane oxy-
genation. Anesth Analg. 2005;101(3):924.
38. Wittenstein B, Ng C, Ravin H, Goldman A. Recombinant factor VII for severe bleeding during
extracorporeal membrane oxygenation following open heart surgery. Pediatr Crit Care Med.
2005;6(4):4736.
39. Dominguez TE, Mitchell M, Friess SH, Huh JW, Manno CS, Ravishankar C. Use of recombi-
nant factor VIIa for refractory hemorrhage during extracorporeal membrane oxygenation.
Pediatr Crit Care Med. 2005;6(3):34851.
40. Repesse X, Au SM, Brechot N, Trouillet JL, Leprince P, Chastre J. Recombinant factor VIIa
for uncontrollable bleeding in patients with extracorporeal membrane oxygenation: report on
15 cases and literature review. Crit Care. 2013;17(2):R55.
41. 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(4):8524.
42. 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(5):16435.
43. Morgan IS, Codispoti M, Sanger K, Mankad PS. Superiority of centrifugal pump over roller
pump in paediatric cardiac surgery: prospective randomised trial. Eur J Cardiothorac Surg.
1998;13(5):52632.
44. Klein M, Dauben HP, Schulte HD, Gams E. Centrifugal pumping during routine open heart
surgery improves clinical outcome. Artif Organs. 1998;22(4):32636.
45. Parolari A, Alamanni F, Naliato M, Spirito R, Franze V, Pompilio G, et al. Adult cardiac sur-
gery outcomes: role of the pump type. Eur J Cardiothorac Surg. 2000;18(5):57582.
46. Khoshbin E, Roberts N, Harvey C, Machin D, Killer H, Peek GJ, et al. Poly-methyl pentene
oxygenators have improved gas exchange capability and reduced transfusion requirements in
adult extracorporeal membrane oxygenation. ASAIO J. 2005;51(3):2817.
47. Dogal NM, Mathis RK, Lin J, Qiu F, Kunselman A, Undar A. Evaluation of three hollow-ber
membrane oxygenators without integrated arterial lters for neonatal cardiopulmonary bypass.
Perfusion. 2012;27(2):13240.
48. 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 ber and at sheet membrane oxygenators. Artif Organs. 2000;24(1):438.
49. Mongero L, Beck J. On bypass: advanced perfusion techniques. New York: Springer; 2008.
50. Javidfar J, Brodie D, Costa J, LaVelle M, Miller J, Newmark A, et al. Subclavian artery can-
nulation for venoarterial extracorporeal membrane oxygenation. ASAIO J.
2012;58(5):4948.
cannula-toward a percutaneous and ambulatory paracorporeal articial lung. ASAIO J.
2008;54(6):60611.
52. Vohra HA, Adamson L, Weeden DF, Haw MP. Use of extracorporeal membrane oxygenation
in the management of septic shock with severe cardiac dysfunction after Ravitch procedure.
Ann Thorac Surg. 2009;87(1):e45.
53. 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(5):70710.
2007;8(5):44751.
55. Huang CT, Tsai YJ, Tsai PR, Ko WJ. Extracorporeal membrane oxygenation resuscitation in
adult patients with refractory septic shock. J Thorac Cardiovasc Surg. 2013;146(5):10416.
56. Brechot N, Luyt CE, Schmidt M, Leprince P, Trouillet JL, Leger P, et al. Venoarterial extracor-
poreal membrane oxygenation support for refractory cardiovascular dysfunction during severe
bacterial septic shock. Crit Care Med. 2013;41(7):161626.
Chapter 12
Mobilization During ECLS
Gregory A. Schmidt
Introduction
The history of ECLS is a gripping story of creativity, persistence, and ceaseless

innovation (see Chap. 14). Yet, until recently, patients were deeply sedated, often
paralyzed, always mechanically ventilated, and moved as gingerly as possible. Now,
technological innovations in gas-exchanging membranes, pumps, and cannulas,
combined with the recognition that life-sustaining gas exchange can be achieved
through the venovenous (VV) mode, have spurred enthusiasm to mobilize ECLS
patients, as we do other critically ill patients [1, 2]. These changes have also forced
us to reconsider the role of ECLS in severe ARDS. Is it merely to rescue the desper-
ately ill once mechanical ventilation, paralysis, prone positioning, and inhaled vaso-
dilators 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 intensivists
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 veno-
arterial (VA), mode. For example, in the Conventional Ventilation or ECMO for
Severe Acute Respiratory Failure (CESAR) trial, all ECLS patients were treated
with the VV mode and none required conversion to VA ECLS [3] (and R. Firmin,
G.A. Schmidt, MD (*)

Division of Pulmonary Diseases, Critical Care, and Occupational Medicine, Department of
Internal Medicine, University of Iowa, 200 Hawkins Drive, Iowa City, IA 52242, USA
e-mail: gregory-a-schmidt@uiowa.edu

212 G.A. Schmidt
personal communication). VV ECLS offers three major advantages. First, returning

blood from the gas exchange membrane to the veins, rather than systemic arteries,
requires much lower pressures and correspondingly simpler, safer circuits (see
Chap. 6). Second, since only venous access is required, mobilization-related risks of
large vessel arterial cannulation (hemorrhage, limb ischemia, dislodgement) are
avoided. Finally, by its very nature, VA ECLS requires at least two cannulas, adding
to the complexity of movement, whereas VV support typically needs only a single
cannula (see Chap. 7).
ECLS devices, too, have evolved in a direction to make mobilization more fea-
sible. Modern polymethylpentene (PMP) gas-exchanging membranes present a
much lower resistance to blood ow than earlier technologies (see Chap. 8). This
has two key advantages: (a) higher circuit blood ow is possible, capturing a greater
fraction of the native cardiac output, making it easier to attain acceptable oxygen-
ation, as described in Chap. 1; and (b) the lower resistance requires less pressure to
drive blood ow, 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 ECLS patients. It is also hard to over-
state 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 mobi-
lize ECLS patients, we dedicate one member of the team to focus entirely on the
cannula, since the patients 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 15 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], post-traumatic 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
12 Mobilization During ECLS 213
as early as the rst day of critical illness; and (2) weakness may persist for years.
These ndings suggest that early approaches to treatment and prevention might
yield the most gains.
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 up-
regulated. Also like in peripheral muscle, active contraction (that is, active breath-
ing) can effectively modify the degree of catabolism, helping to maintain contractile
function [12]. This has potentially important implications for the role of ECLS. By
freeing the diseased lungs from the burden of gas exchange, extracorporeal support
could allow less ventilation (even extubation, see Chap. 9) and, therefore, less seda-
tion, 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 inuences requires
changing our view of the critically ill from fragile beings in need of (over-) protec-
tion 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 ECLS.
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 signicantly lower ICU and hospital
length of stay [14].
The rst randomized, controlled, multi-centered study of early mobilization
enrolled subjects within 72 h of intubation, allowing physical therapy to be pro-
vided, on average, within 2 days of intubation [1]. Intervention subjects achieved
important milestones previously thought unattainable during mechanical ventila-
tion: 43 % transferred to chair and 24 % were able to walksome as far as 30 m.
Subjects who received the intervention had less delirium, more ventilator-free days,
214 G.A. Schmidt
and were more likely to be functionally independent at hospital discharge. When

combining Awakening, Breathing Coordination, Delirium monitoring and manage-
ment, and Early exercise and mobility in a bundled approach (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]. In one of the rst 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 pre-transplant rehabilitation, ECLS was instituted
through a dual-lumen cannula in the right internal jugular vein [17]. Mechanical venti-
lation was weaned by 24 h and, using a wheeled intravenous pole to carry the pump,
oxygenator, and oxygen tank, the patient was able to ambulate throughout the hospital,
use a treadmill, and ride an exercise bicycle. He was successfully transplanted and
discharged home. These same authors subsequently described ten subjects in whom
ambulatory ECLS was attempted for severe respiratory failure due to ARDS, idiopathic
pulmonary brosis, COPD, or pulmonary arterial hypertension [18]. Six were ulti-
mately liberated from ventilatory support and four of these ambulated while on ECLS.
Much of the ambulatory ECLS experience has been gained in patients awaiting
transplantation. Because pre-transplant mechanical ventilation increases the risk of
complications and predicts worse outcomes following surgery, using ECLS 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 [19]. Active rehabilitation, physical therapy, and ambulation
were achieved pre-transplant 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 brosis patients awaiting transplant [20]. In a small, retrospective
series, patients bridged with ECLS who participated in rehabilitation had shorter
post-transplant mechanical ventilation, incidence of ICU-AW, and length of stay in
the ICU and in the hospital [21]. ECLS has also been used to avoid intubation alto-
gether. In small cohort, ve patients with cardiopulmonary failure due to pulmonary
hypertension were placed on ECLS using local anesthesia [22]. All stabilized imme-
diately and three survived to transplantation (1835 days later), breathing spontane-
ously, eating and drinking, and participating actively in physical therapy in the
interim. This series is notable in that VA ECLS was the chosen mode, in light of the
underlying illness. In a larger series, 31 patients were transplanted from ECLS
(range 253 days) and 19 were ambulatory at the time of surgery [23]. Roughly half
of these patients were managed with VA (or VAV) ECLS.
This early experience has led many centers, not only to consider and trial ambu-
latory ECLS, but to actively develop and nurture specialized teams to aid mobiliza-
tion. Following the creation of a mobilization program aimed at ECLS patients, 35
of 100 patients with refractory respiratory or cardiac failure could be actively
rehabilitated [24]. Since bridge-to-recovery patients tend to be more ill and suffer-
ing 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 ECLS through a dual-lumen cannula in
the internal jugular vein, four patients were on VA ECLS (internal jugular vein to
subclavian artery) and eight had femoral access. Of the 16 bridge-to-recovery
patients, 14 survived, most going directly home following discharge. It should be
emphasized that these results followed an explicit plan to create and train a special-
ized team including physical and occupational therapists, perfusionists, critical care
RNs, critical care nurse practitioners, respiratory therapists, and ECLS intensivists
and surgeons.
Barriers to Mobilization During ECLS
It is no small task to safely mobilize patients during ECLS. Barriers include severity
of illness, resource and stafng limitations, equipment needs, risks of catastrophic
complications, and cultural barriers. There is broad agreement that mobilization is
safe for critically ill patients, including those with endotracheal tubes requiring
mechanical ventilation [25], but still very limited condence that we can safely do
this with the very sickest. Before mobilization, patients should be screened for con-
traindications, such as those described in the trial by Schweickert and colleagues as
modied for ECLS in Table 12.1 [26]. Many ECLS patients will have neurological
failure as a consequence of the underlying illness or its treatment, precluding extu-
bation, ambulation, or any meaningful therapy. Severe circulatory failure may also
be limiting, although many patients in shock and on vasoactive infusion can sit,
stand, and even walk [1]. In others, mobilization threatens 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-specic 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 reassuring. For example, in the small series cited above [17, 18] serious com-
plications were not seen. In the larger series of bridge-to-transplant and bridge-to-
recovery, there were no complications related to physical therapy [24]. Even in
the broader experience gained in generally critically ill patients, physical therapy
and mobilization have been remarkably complication-free [14, 27]. 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 peri-
ods of patientventilator dyssynchrony. Falls to knees without injury did not
216 G.A. Schmidt
Table 12.1 Safety screen: Mean arterial blood pressure <65 mmHg
contraindications to
Heart rate <40 or >130/min
mobilization
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 ECLS cannula
Malfunction of cannula or circuit with movement
Modied from [26]
occur in two studies [1, 14], but occurred rarely in another [13]. Accidental extu-
bations 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 arrhythmia) and only 4 of 1110 required minimal additional treatment
or cost [27].
A nal barrier relates to each ICUs 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 stafng to include full-time
physical and occupational therapists succeeded in improving delirium, functional
mobility, and ICU length of stay [28]. 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 stafng is an effective means to change practice [29]. 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 benets are subtle and delayed, leading
some clinicians to urge rst 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 ECLS 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 nger-pointing.
Mobilization: Team and Methods
Once a patient passes the safety screen described above, the mobility team should
be assembled. Safely ushering an ECLS 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: four caregivers at a mini-
mum and often several more (Fig. 12.1). 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,
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. 8. 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 ECLS 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
Fig. 12.1 Patient with

severe ARDS making a trip
outdoors, accompanied by
caregivers and family
218 G.A. Schmidt
Table 12.2 Terminating a Mean arterial blood pressure <65 mmHg

therapy session
Heart rate <40 or >130/min
Respiratory rate >40 that is sustained
Critical hypoxemia (SaO2 falling more than 5 % from
baseline)
New arrhythmia
Signicant ventilator dyssynchrony
Signicant new symptoms, such as chest pain
Inability to maintain adequate circuit ow
Circuit malfunction or alarming
Device dislodgement or instability
Subjective distress: verbal or facial cues, overt agitation
Unsteadiness sufcient to risk safely continuing
Fall
during cardiopulmonary resuscitation [30]. 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 rm 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 ow, 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 ow should be
disconnected from wall oxygen, using a transport cylinder. Sweep gas ow 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 wheel-
chair 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 [31]. 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 distress, as
well as the ability to measure blood pressure. Any deterioration, including those
listed in Table 12.2, should prompt an immediate halt to mobilization, reassessment
of safety, and a decision whether to terminate the session. Returning the patient to
the ICU bed requires careful planning and may benet 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 [32]. Electrical stimulation of muscles has been hypothesized to
improve strength and limit wasting, but efcacy remains unproved [33]. As the
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
ECLS, 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, Nigos C, Pawlik AJ, Esbrook CL, et al. Early
physical and occupational therapy in mechanically ventilated, critically ill patients: a random-
ized controlled trial. Lancet. 2009;373:187482.
2. Engel HJ, Needham DM, Morris PE, Gropper MA. ICU early mobilization: from recommen-
dation to implementation at three medical centers. Crit Care Med. 2013;41:S6980.
trolled trial. Lancet. 2009;374:135163.
4. Kress JP, Pohlman AS, OConnor MF, Hall JB. Daily interruption of sedative infusions in criti-
cally ill patients undergoing mechanical ventilation. N Engl J Med. 2000;342:14717.
5. Riker RR, Shehabi Y, Bokesch PM, Ceraso D, Wisemandle W, Koura F, et al.
Dexmedetomidine vs. midazolam for sedation of critically ill patients: a randomized trial.
JAMA. 2009;301:48999.
6. Kress JP, Gehlbach B, Lacy M, Pliskin N, Pohlman AS, Hall JB. The long-term psychological
effects of daily sedative interruption on critically ill patients. Am J Respir Crit Care Med.
2003;168:145761.
7. Shehabi Y, Bellomo R, Reade MC, Bailey M, Bass F, Howe B, et al. Early intensive care seda-
tion predicts long-term mortality in ventilated critically ill patients. Am J Respir Crit Care
Med. 2012;186:72431.
8. Girard TD, Kress JP, Fuchs BD, Thomason JW, Schweickert WD, Pun BT, et al. Efcacy and
safety of a paired sedation and ventilator weaning protocol for mechanically ventilated patients
in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial.
Lancet. 2008;371:12634.
9. Fan E, Cheek F, Chlan L, Gosselink R, Hart N, Herridge MS, et al. An ofcial American
Thoracic Society clinical practice guideline: the diagnosis of intensive care unit-acquired
weakness in adults. Am J Respir Crit Care Med. 2014;190:143746.
220 G.A. Schmidt
10. Demoule A, Jung B, Prodanovic H, Molinari N, Chanques G, Coirault C, et al. Diaphragm

dysfunction on admission to the intensive care unit. Prevalence, risk factors, and prognostic
impacta prospective study. Am J Respir Crit Care Med. 2013;188:2139.
11. Wollersheim T, Woehlecke J, Krebs M, Hamati J, Lodka D, Luther-Schroeder A, et al.
Dynamics of myosin degradation in intensive care unit-acquired weakness during severe
critical illness. Intensive Care Med. 2014;40:52838.
12. Sassoon CSH, Zhu E, Caiozzo VJ. Assist-control mechanical ventilation attenuates ventilator-
induced diaphragm dysfunction. Am J Respir Crit Care Med. 2004;170:62632.
13. Bailey P, Thomsen GE, Spuhler VJ, Blair R, Jewkes J, Bezdjian L, et al. Early activity is
feasible and safe in respiratory failure patients. Crit Care Med. 2007;35:13945.
14. Morris PE, Goad A, Thompson C, Taylor K, Harry B, Passmore L, et al. Early intensive care
unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med.
2008;36:223843.
15. Balas MC, Vasilevskis EE, Olsen KM, Schmid KK, Shostrom V, Cohen MZ, et.al. Effectiveness
and safety of the awakening and breathing coordination, delirium monitoring/management,
and early exercise/mobility bundle. Crit Care Med. 2014;42:102436.
16. Thiagarajan RR, Teele SA, Teele KP, Beke DM. Physical therapy and rehabilitation issues for
patients supported with extracorporeal membrane oxygenation. J Pediatr Rehabil Med.
2012;5:4752.
17. Garcia JP, Iacono A, Kon ZN, Grifth BP. Ambulatory extracorporeal membrane oxygenation:
a new approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg. 2010;
139:e1379.
18. Garcia JP, Kon ZN, Evans C, Wu Z, Iacono AT, McCormick B, et al. Ambulatory veno-venous
extracorporeal membrane oxygenation: innovation and pitfalls. J Thorac Cardiovasc Surg.
2011;142:75561.
19. Turner DA, Cheifetz IM, Rehder KJ, Williford WL, Bonadonna D, Banuelos SJ, et al. Active
rehabilitation and physical therapy during extracorporeal membrane oxygenation while
awaiting lung transplantation: a practical approach. Crit Care Med. 2011;39:25938.
20. Hayes D Jr, Kukreja J, Tobias JD, Ballard HO, Hoopes CW. Ambulatory venovenous extracor-
poreal respiratory support as a bridge for cystic brosis patients to emergent lung transplanta-
tion. J Cyst Fibros. 2012;11:405.
21. Rehder KJ, Turner DA, Hartwig MG, Williford WL, Bonadonna D, Walczak RJ Jr, et al.
Active rehabilitation during extracorporeal membrane oxygenation as a bridge to lung trans-
plantation. Respir Care. 2013;58:12918.
22. Olsson KM, Simon A, Strueber M, Hadem J, Wiesner O, Gottlieb J, et al. Extracorporeal
membrane oxygenation in nonintubated patients as bridge to lung transplantation. Am J
Transplant. 2010;10:21738.
23. Hoopes CW, Kukreja J, Golden J, Davenport DL, Diaz-Guzman E, Zwischenberger JB.
Extracorporeal membrane oxygenation as a bridge to pulmonary transplantation. J Thorac
Cardiovasc Surg. 2013;145:8628.
24. Hoopes CW, Kukreja J, Golden J, Davenport DL, Diaz-Guzman E, Zwischenberger JB. Early
mobilization of patients receiving extracorporeal membrane oxygenation: a retrospective
cohort study. Crit Care. 2014;18:R38.
25. Hodgson CL, Stiller K, Needham DM, Tipping CJ, Harrold M, Baldwin CE, et al Expert con-
sensus and recommendations on safety criteria for active mobilization of mechanically venti-
lated critically ill adults. Crit Care. 2014;18:658. doi:10.1186/s13054-014-0658-y.
26. Pohlman MC, Schweickert WD, Pohlman AS, Nigos C, Pawlik AJ, Esbrook CL, et al.
Feasibility of physical and occupational therapy beginning from initiation of mechanical
ventilation. Crit Care Med. 2010;38:208994.
27. Sricharoenchai T, Parker AM, Zanni JM, Nelliot A, Dinglas VD, Needham DM. Safety of
physical therapy interventions in critically ill patients: a single-center prospective evaluation of
1110 intensive care unit admissions. J Crit Care. 2014;29:395400.
28. Needham DM, Korupolu R, Zanni JM, Pradhan P, Colantuoni E, Palmer JB, et al. Early physi-
cal medicine and rehabilitation for patients with acute respiratory failure: a quality improve-
ment project. Arch Phys Med Rehabil. 2010;91:53642.
29. Garzon-Serrano J, Ryan C, Waak K, Hirshberg R, Tully S, Bittner EA, et al. Early mobilization
in critically ill patients: patients mobilization level depends on health care providers profes-
sion. PM R. 2011;3:30713.
30. Yeung JH, Ong GJ, Davies RP, Gao F, Perkins GD. Factors affecting team leadership skill
and their relationship with quality of cardiopulmonary resuscitation. Crit Care Med.
2012;40:261721.
31. Hickmann CE, Roeseler J, Castanares-Zapatero D, Herrera EI, Mongodin A, Laterre PF.
Energy expenditure in the critically ill performing early physical therapy. Intensive Care Med.
2014;40:54855.
32. Burtin C, Clerckx B, Robbeets C, Ferdinande P, Langer D, Troosters T, et al. Early exercise in
critically ill patients enhances short-term functional recovery. Crit Care Med. 2009;37:
2499505.
33. Kho ME, Truong AD, Zanni JM, Ciesla ND, Brower RG, Palmer JB, et al. Neuromuscular
electrical stimulation in mechanically ventilated patients: a randomized, sham-controlled pilot
trial with blinded outcome assessment. J Crit Care. 2015;30:329.
Chapter 13
ECMO Weaning and Decannulation
Sundar Krishnan and Gregory A. Schmidt
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,
oxygenator sweep gas, and ECMO ow during the weaning process; and cardiopul-
monary assessment during weaning. The process of weaning and trialing-off extra-
corporeal support is somewhat different for venovenous (VV) and venoarterial
(VA) ECMO, because of the difference in physiological support provided by the
two modalities. Further, we will discuss the procedure of decannulation and the
follow-up of patients after decannulation. Finally, we review the termination of
ECMO in patients where further therapy is deemed unrecoverable.
S. Krishnan, MBBS (*)

Department of Anesthesia, University of Iowa Hospitals and Clinics,
200 Hawkins Drive, Iowa City, IA 52242, USA
e-mail: sundar-krishnan@uiowa.edu
G.A. Schmidt, MD
Division of Pulmonary Diseases, Critical Care, and Occupational Medicine, Department of
Internal Medicine, University of Iowa, 200 Hawkins Drive, Iowa City, IA 52242, USA
e-mail: gregory-a-schmidt@uiowa.edu

224 S. Krishnan and G.A. Schmidt
Signs of Readiness
Therapy directed toward improving pulmonary function starts soon after ARDS
patients are placed on ECMO support. Modalities that might help hasten recovery
in most patients include diuresis, antibiotic therapy, bronchodilators, and bronchos-
copy and airway evacuation. Ventilator management involves limiting FiO2 to 30 %,
maintaining PEEP to preserve recruitment, and limiting tidal volumes with low lev-
els of support.
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 typi-
cally used during extracorporeal support, an increase in PaO2 or a decrease in
PaCO2 might be noticed. Increased levels of CO2 in capnometry might also provide
a clue to enhanced alveolar gas exchange. Improved pulmonary compliance will
lead to increased tidal volumes on unchanged pressure-control ventilator settings.
Usually, respiratory system compliance >20 cc/cm H2O suggests adequate recovery
to proceed with weaning. An improvement in pulmonary aeration on chest radio-
graph often accompanies improving alveolar recruitment. Signs of improving lung
function are listed in Table 13.1.
Once signicant aeration is achieved, as assessed by improving tidal volumes
(>100150 cc) on rest settings and improvements in chest radiograph, the Cilley test
[1] should be performed daily. The test involves increasing the FiO2 to 100 %, with
no other changes in ventilator settings. A positive test is marked by a rapid increase
in oxygen saturation within a couple of minutes, and is a marker for readiness to
advance ventilator settings.
In ARDS patients on VA-ECMO support for cardiac failure in addition to respira-
tory failure, cardiac recovery would usually follow with improved pulmonary func-
tion, especially if the cardiac failure is related to acute right ventricular dysfunction.
Improved gas exchange, decreased ventilatory airway pressures (on volume-preset
Table 13.1 Signs of improving lung function

Improved pulmonary mechanics Improved lung compliance, as observed in ventilator data
Increased functional residual capacity seen on chest
radiography
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
Rapid improvement in SpO2 with increase in FiO2 to 100 %
(Cilley test)
Ancillary signs Improved lung aeration on chest X-ray
Resolving B-lines on lung ultrasound
13 ECMO Weaning and Decannulation 225
modes), and decreased pulmonary vascular resistance will lead to enhanced right
ventricular function. As ventricular function improves and native cardiac output
carries blood forward, an increase in pulse pressure becomes evident.
ECMO Weaning and Liberation
ARDS patients with improving pulmonary function can be weaned and liberated
from extracorporeal support. On occasion, the decision to wean off extracorporeal
support might be hastened by bleeding complications related to the anticoagulation
necessary to allow blood ow through the extracorporeal circuit. Weaning and lib-
eration from ECMO is managed with continuous cardiopulmonary assessments,
while changes are made in ventilator settings, ECMO ow, and oxygenator sweep
gas ow (Fig. 13.1). Weaning involves decreasing extracorporeal support and evalu-
ating cardiopulmonary recovery without ECMO. Trialing-off involves temporary
discontinuation of extracorporeal support to prove that cardiopulmonary function
has recovered to the point that it can be supported without ECMO. Finally, libera-
tion 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 function, it is possible to trial off ECMO [2].
Ventilator Management
When the decision is made to advance from rest settings, recruitment maneuvers
might be required before lung-protective ventilator settings are employed. Further
weaning is only possible when the patient is able to achieve adequate tidal volumes
(approximately 6 mL/kg) with acceptable airway pressure (plateau airway pressure
<30 cm H2O). Adjusting sedation to allow spontaneous breathing assists in improv-
ing lung recruitment and weaning. Adequate minute ventilation will have to be
established before patients can be weaned or trialed off extracorporeal support.
As ECMO support for gas exchange is decreased, the patients respiratory rate
will likely increase. The magnitude of increase in work of breathing during weaning
is an indication of the amount of extracorporeal support the patient was receiving in
addition to their native lung function. A high respiratory rate (>30/min), low tidal
volumes, or visible distress are signs that the patient is not ready for trialing off
ECMO. Ideally, the patient should be supportable with minimal ventilator settings,
with the ability to increase ventilatory support if needed, once extracorporeal sup-
port is removed.
In some patients, it might be possible to decrease sedation to the point of com-
plete wakefulness while maintaining adequate gas exchange on extracorporeal
support. At this point, the endotracheal tube remains the only major indication for
sedation, and extubation can be considered before weaning ECMO ow.
Improved pulmonary function
Recruitment maneuvers
Are
No respiratory
mechanics
acceptable?
Yes
Maintain sweep gas FiO2 at 100%. Decrease sweep gas flow in steps to 1 lpm
Are
No hemodynamic
and respiratory
parameters
stable?
Yes
Turn off oxygenator sweep gas

flow, maintain ECMO blood flow
Patient is not ready to be Over the next two

weaned off extracorporeal hours, are gas Patient is ready for discontinuation
support. Continue VV ECMO No exchange, respiratory Yes of ECMO support.
support. Maintain adequate mechanics, and
hemodynamic Plan for decannulation
oxygenator sweep gas flow
and ECMO blood flow parameters
acceptable?
Fig. 13.1 Algorithm for weaning ECLS in the ventilated ARDS patient on VV-ECMO
Alternatively, tracheostomy might be considered early in patients whose recovery is

likely to be prolonged. Liberation from sedation allows patients with ARDS to con-
tinue physical therapy and avoid the complications associated with immobility
(Chap. 12). In extubated patients, weaning from ECMO might involve instituting
non-invasive ventilatory support.
The decision to remove ventilator support rst versus weaning off ECMO rst
depends on multiple factors. As mentioned previously, hemorrhagic complications
might lead to discontinuing ECMO sooner than planned. On the other hand, for
patients who require signicant sedation to tolerate the endotracheal tube, extuba-
tion with continued ECMO support might assist in mobilization and enhanced
recovery. The institutions ability to provide prolonged ECMO support must be
taken into account in patients expected to require respiratory support for a long
duration. In addition, the condition of the ECMO circuit is a paramount consider-
ation before withdrawing ventilator support. Finally, the intensive care unit should
be able to provide emergency ventilatory support in case of ECMO failure in an
extubated patient.
Oxygenator Sweep Gas Flow Management
Once adequate tidal volumes are possible on lung-protective ventilator settings, the
sweep gas ow is reduced in steps to 1 L/min and sweep FiO2 is maintained at 100
%. Alternatively, it is also acceptable to decrease sweep gas ow to 2 L/min then
decrease sweep FiO2 to maintain SaO2 >95 %. When gas exchange is stable at these
settings, the patient is trialed off ECMO by turning off the sweep gas altogether and
capping off the oxygenator. Gas exchange, work of breathing, and hemodynamic
status should be assessed during this trial. There needs to be no interruption in blood
ow in the extracorporeal circuit during a trial off VV-ECMO. The return tubing
from the oxygenator would soon change in color from oxygenated blood to deoxy-
genated blood, similar to the tubing draining blood from the patient. If the patient
remains stable over the next hour or 2, they have successfully passed their trial off
ECMO and are ready for decannulation. If the patient shows signs of inadequate
native pulmonary function, the sweep gas ow is simply turned back on and extra-
corporeal support is resumed. Ventilatory settings and sedation can be managed to
the patients comfort, until the next trial-off period.
In patients on VA-ECMO support, oxygenator sweep gas ow is only titrated to
PaCO2, and not decreased for purposes of weaning and assessing the degree of car-
diopulmonary recovery. Decreasing oxygenator sweep gas ow would lead to inad-
equate oxygenation of blood owing through the VA-ECMO circuit, causing
right-to-left shunting.
ECMO Flow Management
Circuit blood ow does not need to be changed while attempting weaning from
VV-ECMO. However, as the patients intravascular volume is decreased with diure-
sis, it might be difcult to maintain a high rate of blood ow through the extracor-
poreal circuit. As the patients pulmonary gas exchange improves, it becomes
possible to decrease circuit blood ow. A minimum ow rate of 5001000 mL/min
needs to be maintained to prevent stasis of blood in the cannulae, circuit, pump, and
oxygenator.
In patients on VA-ECMO support, inotropic and ventilatory therapy is estab-

lished to support cardiopulmonary function before weaning. ECMO pump blood
ow is decreased slowly. The circuit needs to be assessed vigilantly for clots. If the
patient maintains adequate hemodynamics at idle ow (approximately 1000 mL/
min) for 12 h, then the patient might be ready for a trial-off. Trialing-off on
VA-ECMO involves clamping the inow and outow cannulae from the patient, and
owing blood through the bridge in the circuit. Clamping has to be performed in a
VBA (vein-before-artery) sequence to prevent a high afterload for the ECMO pump.
This trial-off can be done for 23 h. During a VA-ECMO trial-off, a procedure
called ashing the cannulae is performed every 1020 min for 15 s, whereby the
clamps on the inow and outow cannulae are released and blood is allowed to ow
through the cannulae to prevent prolonged stasis.
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 ow 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.
For VA-ECMO, full anticoagulation must be maintained throughout the weaning
and trial-off period, because the decreased circuit ows increase the chances of
stasis and clot formation.
Cardiopulmonary Assessment
As patients are weaned off extracorporeal support, cardiac and pulmonary workload
increases. Cardiopulmonary function should be closely monitored during this pro-
cess. As alluded to earlier in this chapter, respiratory rate, tidal volume, gas
exchange, and respiratory distress are vital clues to pulmonary function during
weaning and trialing off from ECMO, much as one would judge the success of a
spontaneous breathing trial. It is important to note that in patients on VA-ECMO
support, ashing introduces oxygenated blood into the patient. Hence gas exchange
assessments should be made prior to ashing the cannulae.
In patients on VA-ECMO, cardiac function assessments are made as ECMO ow
is decreased. On echocardiography, the heart is empty while the patient is on full
ECMO ow. As extracorporeal ow is reduced, more volume is left in the right

atrium, and this should lead to increased native cardiac output and pulse pressure.
Inotropic therapy will likely be required to maintain cardiac function. Cardiac
chamber size and function on echocardiography, hypotension, and low cardiac
index, as well as downstream parameters of the adequacy of tissue perfusion (cen-
tral venous oxygen saturation, lactate, and urine output) should be monitored as
ECMO ow is decreased.
Even in patients on VV-ECMO, increased work of breathing, altered gas exchange,
and increased airway pressures might lead to cardiac decompensation during wean-
ing and the trial-off period. In addition to standard hemodynamic monitoring and
assessment of the need for inotropic therapy, echocardiography can be a valuable
tool in assessing cardiac function. Right ventricular dilation when oxygenator sweep
gas ow is decreased or turned off can be a clue toward hemodynamic instability.
Decannulation
Decannulation is performed after a trial-off proves that recovery is adequate to

maintain life support through more conventional means. Decannulation is usually
performed as a planned bedside procedure. Attention needs to be paid to anticoagu-
lation management, hemostasis, and patient positioning.
Before decannulation, cardiopulmonary stability since the trial-off needs to be
conrmed. Inotropic and ventilatory therapy is reestablished if they were reduced
since the trial-off. Heparin is discontinued at least 3060 min before the procedure.
Extracorporeal circuit ow 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 signicant negative pressure gradient into the venous
system. After removal of percutaneously placed cannulae, pressure needs to be held
at the venous puncture sites for at least 30 min. For cannulae placed via surgical
cut-down, a surgical approach and repair might be necessary.
Following Decannulation
Since patients have demonstrated adequate gas exchange (and hemodynamic stabil-
ity) 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. Procedures that were withheld because of the anti-
coagulation 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 (uid overload, thromboembolism), performing appropriate diagnos-
tic studies, and using the ventilator to support respiratory function (or, for the extu-
bated patient, non-invasive ventilation or re-intubation). In the 2448 h after
decannulation, patients also need to be assessed for venous thrombosis at the can-
nula site.
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
illness, such as nosocomial pneumonia, septic shock, or acute renal failure turn a
desperate situation into an unrecoverable downward spiral (Table 13.2). 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 [3, 4], perhaps more so in ECLS patients [5, 6]. Difcult 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 [6]. Laying out goals for
Table 13.2 Common reasons for failure of ECLS in ARDS

Category of failure Reason for failure
Primary illness progresses or fails to remit Proliferative ARDS, lung brosis
Cor pulmonale
Multi-organ 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
patients who are being bridged to recovery is more challenging, because recovery
can be protracted, hard to predict accurately, and the course may be punctuated with
periodic downturns that make survival seem unlikely. At the time that consent for
ECLS is obtained, it is appropriate to raise the possibility that it will fail to achieve
its goals and may have to be withdrawn. During the treatment course, new compli-
cations or additional organ failures should serve as opportunities to revisit the goals,
judging anew whether they are still attainable. Even without such events, 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) [7]
can be helpful for patients and families.
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. For a patient with decisional capacity who does not
accept the recommendation to withdraw ECLS, this choice should be respected,
even when it seems a poor use of resources [6].
Occasionally, the balance of benets and burdens shifts unfavorably as the
patient or family succumbs to the stress of chronic critical illness. Some patients
will request withdrawal of a life-sustaining device even while the clinicians judge
that the overall picture is improving, a situation that carries a high emotional burden
for both patient and caregivers [8]. Ethicists generally accept the moral equivalence
of withdrawing and withholding life-sustaining treatments for the critically ill, such
as dialysis and mechanical ventilation. Whether similar reasoning applies to circula-
tory support devices is less clear, especially when these treatments are highly effec-
tive in sustaining life for many weeks or months. For example, a substantial minority
of health care providers and many patients consider deactivating implantable car-
diac devices such as pacemakers or total articial hearts to be euthanasia or
physician-assisted suicide [9]. Since ECLS is never instituted as destination therapy,
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.
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 [10]. Patient
comfort should be assured, using analgesics and sedatives as needed, and anticipat-
ing that withdrawal of ECLS may provoke or heighten dyspnea. We then turn off the
sweep gas, preventing further gas exchange in the membrane and, assuming ECLS
was necessary, leading to death. For those on VA ECLS, circuit blood ow can be
reduced as vasoactive drug infusions are stopped.
Acknowledgements Dr Krishnan has no conicts to disclose.

Dr Schmidt has no conicts to disclose.
References
1. ELSO Adult Respiratory Failure Supplement to the ELSO General Guidelines. December
2013. https://www.elso.org/Portals/0/IGD/Archive/FileManager/989d4d4d14cusersshyerdoc
umentselsoguidelinesforadultrespiratoryfailure1.3.pdf. Accessed Jan 2015.
2. Lynch WR. Weaning, trialing and futility. In: ECMO: extracorporeal cardiopulmonary support
in critical care. 4th ed. ELSO; 2012.
3. Luce JM. A history of resolving conicts over end-of-life care in intensive care units in the
United States. Crit Care Med. 2010;38:16239.
4. Siegel MD. End-of-life decision making in the ICU. Clin Chest Med. 2009;30:18194.
5. 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;59:64250.
6. Abrams DC, Prager K, Blinderman CD, Burkart KM, Brodie D. Ethical dilemmas encountered
with the use of extracorporeal membrane oxygenation in adults. Chest. 2014;145:87682.
7. Schenker Y, Tiver GA, Hong SY, White DB. Discussion of treatment trials in intensive care. J
Crit Care. 2013;28:8629.
8. Bruce CR, Allen NG, Fahy BN, Gordon HL, Suarez EE, Bruckner BA. Challenges in deacti-
vating a total articial heart for a patient with capacity. Chest. 2014;145:62531.
9. Rady MY, Verheijde JL. Ethical challenges with deactivation of durable mechanical circula-
tory support at the end of life: left ventricular assist devices and total articial hearts. J
Intensive Care Med. 2014;29:312.
10. Cook D, Rocker G. Dying with dignity in the intensive care unit. N Engl J Med.
2014;370:250614.
Chapter 14
The Story of ECLS: History and Future
J. Ann Morris, Robert Pollock, Brittany A. Zwischenberger,

Cherry Ballard-Croft, and Joseph B. Zwischenberger
Introduction
Cardiopulmonary bypass (CPB) for cardiac surgery was rst introduced in the 1950s
[1]. In the 1970s, CPB use at the bedside for pulmonary support began and was termed
extracorporeal membrane oxygenation (ECMO) [2]. Later, in the 1980s, applications
for extracorporeal technology expanded, and the term extracorporeal life support
(ECLS) was coined by researchers at University of Michigan as an umbrella term for
all forms of extracorporeal technologies. These technologies include oxygenation,
CO2 removal, and hemodynamic support, as well as supportive techniques for the
kidneys and liver. Extracorporeal membrane carbon dioxide removal (ECCO2R) is the
term used for the selective removal of CO2 [3]. Despite the other terms, ECMO is most
often used to refer to all forms of extracorporeal support if gas exchange is involved.
ECMO is a rapidly evolving eld that was developed for critically ill patients
with respiratory failure more than 40 years ago. ECMO currently serves as a form
of life support for critically ill patients when traditional supportive care is no longer
effective. Recently, it has also become a bridge to transplantation and a way to man-
age acute shock. Other patient groups with diverse diseases, such as H1N1, are also
receiving ECMO as part of their treatment program for acute respiratory distress
syndrome (ARDS). While technological advances and better management of criti-
cally ill patients allow the option of ventilator support in some patients, ECMO is
still the main option for temporary cardiopulmonary support in numerous patient
populations [4]. In general, ECMO should be considered if the disease or trauma
meets the following criteria: (1) severe, (2) acute, and (3) potentially reversible [5].
J.A. Morris, BS R. Pollock, BS B.A. Zwischenberger, MD C. Ballard-Croft, PhD

J.B. Zwischenberger, MD (*)
Division of Cardiothoracic Surgery, Department of Surgery,
University of Kentucky College of Medicine, Lexington, KY 40536, USA
e-mail: Jzwis2@uky.edu

234 J.A. Morris et al.
Foundation for ECMO Development: Cardiopulmonary

Bypass
The development of cardiopulmonary bypass (CPB) laid the groundwork for the
establishment of ECMO as a viable treatment option (Table 14.1) [6]. With CPB,
the blood circulation and gas exchange (CO2 removal and oxygenation) are tempo-
rarily supported by an external mechanical device, allowing surgeons to operate on
a nonbeating, bloodless heart. The medical community soon realized the necessity
of a bypass machine to support lung and heart function during more complex sur-
geries, such as congenital heart defects.
CPB was rst devised for mechanical cardiopulmonary support in the operating
room by John Gibbon (Fig. 14.1). During his time in Boston in the early 1930s,
Gibbon monitored a patient that became hypotensive and short of breath following
a cholecystectomy. Gibbon thought that the patient might recover if her venous
blood was oxygenated by an external source. Although the patient did not survive,
Gibbon persisted. In 1934, he constructed a rudimentary heartlung machine capa-
ble of fully supporting circulation in a feline model for 30 min.
On September 2, 1952, the rst successful intracardiac procedure with direct
vision was performed by Dr. F. John Lewis and Dr. C. Walton Lillehei at the
University of Minnesota [7]. Dr. Lewis repaired an atrial septal defect (ASD) in a
5-year-old patient via full-body hypothermia and in-ow stasis. The previous year,
Dr. Clarence Dennis performed two unsuccessful intracardiac procedures at the
University of Minnesota. These were the rst attempts at total CPB using a heart
lung device developed by Dr. Dennis and his colleagues [8]. According to Dr.
Lillehei, these attempts of Dr. Dennis failed because of the high perfusion rates
required by the heartlung device [9]. Other complications included excessive
bleeding which obscured the visual eld, incorrect preoperative diagnoses, and air
emboli from inadvertent evacuation of total arterial reservoir volume.
Gibbons heartlung machine (Fig. 14.2) was used successfully on May 6, 1953,
at the Jefferson Hospital in Philadelphia during the operation of an 18-year-old
woman with an ASD [10]. Interestingly, Gibbon once stated, After we nally got
ready, it was ridiculously easy to repair the ASD [8]. Yet, the operation was not
quite that simple. At one point during the procedure, there was reduced extracorpo-
real blood ow, loss of circulating volume, and thrombosis in the CPB circuit,
resulting in prolonged hypotension, hypoxia, and acidosis [10]. Despite these com-
plications, the patient survived. Dr. Gibbon subsequently applied his CPB pump
with a screen/lm oxygenator four more times, all of which were failures [10]. Once
again, these failures related directly to incorrect preoperative diagnoses and intra-
cardiac bleeding which visually obscured the surgical eld. These pioneering expe-
riences highlighted the need for further research to develop both the CPB procedure,
a safer CPB machine with diminished thrombosis risk, and improved diagnostics.
Nevertheless, the events of May 6, 1953, laid the foundation for the future of extra-
corporeal membrane oxygenators.
14 The Story of ECLS: History and Future 235
Table 14.1 Timeline of ECMO development

Year Landmark event
1953 Gibbon performed the rst successful open heart surgery that utilized a heartlung
machine
1954 Lillehei repaired congenital heart disorders using cross-circulation methods
1955 Kirklin et al. further developed Gibbons CPB machine and repaired an atrial septal
defect with it (Mayo Clinic)
1956 G. Clowes developed the rst successful membrane oxygenator
1957 Dewall-Lillehei helix reservoir disposable bubble oxygenator is developed at the
University of Minnesota and used in 250 patients. This work enhanced the safety of
CPB
1964 Gott et al. published a paper on the techniques of applying a graphite-benzalkonium-
heparin coating to certain plastics and metals
1972 Hill et al. published the rst successful use of ECMO for respiratory failure
1975 NIH sponsored the rst randomized multicenter ECMO trial in adult patients suffering
from ARDS. The dismal study ndings were published in 1979
1975 Five brands of membrane oxygengators for ECMO are built and used (Bramson,
Kolobow Teo, Kolobow Sci-Med, Lande-Edwards, Pierce-GE)
1975 Bartlett et al. reported the rst successful use of ECMO in a neonatal patient
(Esperanza)
1978 The BioMedicus Biopump disposable centrifugal pump was made commerically
available as an improvement to the widely used roller pump for CPB
1982 Kolobow published a paper on extracorporeal carbon dioxide removal from the blood
1982 Bartlett et al. published an ECMO study on 45 neonates in which a survival rate
of >50 % was achieved in patients with a 10 % chance of survival on conventional
therapies
1985 Bartlett et al. randomized-play-the-winner study demonstrated the efcacy of ECMO
in neonates
1985 Fifty percent of all CPB procedures in the United States utilized disposable membrane
oxygenators and the other half used bubble oxygenators
1986 Gattinoni reported a 49 % survival rate in adult EC carbon dioxide removal (ECCO2R)
1987 Bard cardiopulmonary support system (CPS) was introduced for rapid deployment,
portable, percutaneous access for cardiac emergencies
1989 ELSO and ELSO registry are established
1989 Two-phase trial to study ECMO versus conventional medical therapy (CMT) showed
that 97 % of newborns survived on ECMO, while only 60 % survived CMT
1994 Randomized clinical trial found no signicant difference in survival between ARDS
patients treated with traditional therapies versus EC CO2 removal following pressure-
controlled inverse ratio ventilation
1996 UK Collaborative ECMO Trial Group published a large-scale randomized trial of
neonatal ECMO in severe respiratory failure patients. The study indicated 68 % and
41 % survival rates for ECMO and CMT groups, respectively
1997 Kolla et al. reported a 54 % survival (n = 100) in adult ARDS patients treated with
ECMO versus conventional therapies
2009 Multicenter randomized controlled trial (CESAR) showed a 16 % increase in 6-month
survival without severe disability for the ECMO versus conventional treatment groups
(63 % vs. 47 %, respectively)
2009 VV ECMO successfully utilized in ARDS patients suffering from H1N1. One study
reported a 68 % survival in ECMO-treated patients (n = 60)
Reproduced with permission of ref. [3]
Fig. 14.1 John H. Gibbon,

MD. Courtesy of Thomas
Jefferson University,
Archives and Special
Collections
Fig. 14.2 The Gibbon-IBM

heartlung machine model
II. This CPB machine was
used between 1952 and 1953
and weighed more than
2000 lb. This cumbersome
machine was used on the rst
4 patients placed on CPB
before newer models were
developed (Stoney [20]).
Courtesy of Thomas
Jefferson University,
Archives and Special
Collections
In the early 1950s, there was considerable pessimism regarding the feasibility
of CPB. Of the reported cases from 1952 to 1954 that utilized total CPB, only 1 out
of 18 cases (5.5 %) had been successful. Many felt the magnitude of the proce-
dures performed and the multiple comorbidities of patients requiring open heart
surgery made recovery of the surgically repaired heart impossible. Thus, the idea
of long-term cardiopulmonary support emerged. Clinicians realized that if long-
term cardiopulmonary support could be achieved, cardiac repair outcomes could
be greatly improved, and other previously untreatable cardiorespiratory diseases
could be overcome.
Along with the reported failures of early CPB machines was the discovery of the
azygous ow concept (low ow concept) [1, 9]. Lillehei and his colleagues utilized
canine experimentation models to prove that the accepted cardiac output ow rate
for CPB of 100165 mL/kg/min was much higher than actually necessary. Canines
were surgically oriented such that only the azygous vein would return blood to the
heart with ow rates of only 814 mL/kg/min. Internal organs of these animals were
well protected during these trials. From this work, Lillehei concluded that success-
ful perfusion could be achieved with 2025 mL/kg/min [9].
Reduced perfusion rates enabled clinicians to further rene the practice of using
autogenous lungs [11]. The use of nonhuman donor lungs was pioneered by Dr.
William Mustard who made use of monkey lungs in the CPB device that he and his
team applied unsuccessfully ve times from 1951 to 1953 [12]. Swift development
of pulmonary edema often followed any perturbations of the many venous catheters
required in the surgical eld during such procedures. In order to preclude such com-
plications, Lillehei and colleagues initially found during experimental trials that they
could use secondary canine subjects as cross-circulatory vehicles during intracardiac
procedures on their primary subjects. Other studies showed that cross-circulation
could provide adequate support based on the azygous (low) ow concept [1, 9].
The application of cross-circulation to human cases allowed the idea of direct-
vision intracardiac procedures to gain acceptance within the medical community.
Until then, only hypothermia with inow stasis had positive outcomes. The cross-
circulation method, which involved the use of a volunteer as a living oxygenator,
was used exclusively and extensively from its inception in March 1954 for all but the
simplest cases, which still utilized either early bubble oxygenators or autogenous
lung support. From March 26, 1954 until July 9, 1955, Dr. C. Walton Lillehei and
coworkers utilized cross-circulation with human lung donors to achieve blood oxy-
genation in 45 infants and children undergoing operations for cardiac anomalies.
The human donors lungs served as the extracorporeal oxygenator, allowing Lillehei
to operate on a large range of patients. The following abnormalities were repaired:
ventricular septal defect (VSD), patent ductus arteriosus (PDA) (with pulmonary
hypertension), tetralogy of Fallot, atrioventricular communis, isolated infundibular
pulmonary stenosis, pulmonary stenosis, ASD, and anomalous pulmonary venous
return. Remarkably, 28 patients survived Lilleheis cutting-edge surgery. Throughout
these trials, perfusion rates of 2530 mL/kg/min adequately supported the patients
[1317]. The widespread acceptance that gross congenital heart defects could be
effectively repaired when provided with adequate extracorporeal support spurred
many clinicians and researchers to further develop existing technology.
ECMO Technology: Direct Contact Oxygenators
Dr. Gibbon demonstrated that a metal lm oxygenator was effective [1, 18]. This
device was composed of a metal cylinder which rotated while partially submerged
in a blood reservoir. This allowed blood to lm over the rotating cylinder, exposing
the blood to an oxygen-rich atmosphere in which the cylinder and blood reservoir
were enclosed. Gibbons device was then modied for use in humans. The upgrade
that allowed for larger perfusion volumes consisted of a vertically-oriented sequence
of stationary metal discs over which blood was allowed to ow downwards. This
set-up created a large surface area of blood lm which was directly exposed to oxy-
gen in an enclosed atmosphere. This device was successfully applied in the rst
total body perfusion via external oxygenator pump during an intracardiac repair
[19]. Gibbon successfully applied this version of his pump only once.
Dr. J.W. Kirklin and his colleagues at the Mayo Clinic in Rochester, Minnesota
obtained a Gibbon pump apparatus and modied it to meet their updated specica-
tions [17]. In March 1955, Kirklin and his colleagues used their Mayo-Gibbon
pump oxygenator in a series of landmark intracardiac procedures (245 patients by
September 1958) and were so successful that the device was commercialized [15,
16, 20, 21]. In contrast, Lillehei and others reported that the number of teams who
could effectively deploy the Mayo-Gibbon pump oxygenator was extremely limited
[20, 21]. This was due to the pumps complexity, cost-prohibitive construction,
complicated maintenance/sterilization routines, and the requirement for very expe-
rienced technicians for even basic operation [17].
To achieve clinical acceptance, a new device must be manufactured, distributed,
sterilized, and deployed cheaply. In May 1953, the DeWall-Lillehei bubble oxygen-
ator was the rst device that satised these requirements [22]. Lillehei and DeWall
based their device on the earlier designs of Clarke, Golan, and Gupta which had been
introduced on a smaller scale beginning in 1950 [23, 24]. The original DeWall-
Lillehei bubble oxygenator functioned as follows: (1) blood and oxygen are sepa-
rately introduced into a vertical oxygenating chamber; (2) the blood and oxygen mix
as they ow upwards and enter a defoaming chamber where the surface tension in the
bubbles present in the blood is released; (3) the blood is then pumped down into a
reservoir submersed in a liquid medium which acts as the heat exchanger; and (4)
excess bubbles are lost in an upward diffusion as the blood is pumped down before
exiting the reservoir and reentering the body [22, 23]. A Sigma motor pump was used
with the DeWall-Lillehei bubble oxygenator. This pump was composed of several
actuating rods/ngers and was so named for the undulating, sinusoidal S-shape
taken on by the rods while they massaged the circuit tubing to induce blood ow [23].
The safety, reliability, and effectiveness of the DeWall-Lillehei oxygenator
were shown in numerous reports leading to rapid acceptance [22, 23, 25]. Dr.
Lillehei states (and others have conrmed) that the DeWall-Lillehei bubble oxy-
genator achieved instant success because it was efcient, inexpensive, heat
sterilizable, easy to assemble, had no moving parts, and was disposable [26].
Improved versions of the DeWall-Lillehei bubble oxygenator consisted of a single,
unitized plastic sheet which came as pre-sterilized, ready-to-assemble, and com-
pletely disposable kits [27, 28]. At the University of Minnesota Hospital between
1953 and 1966, a total of 2715 cardiotomy procedures were performed with 2581
of these cases utilizing bubble oxygenators [29]. In 1966, DeWall and colleagues
introduced the Temptrol (Bentley) disposable blood oxygenator which was an inte-
grated device that combined oxygenator and heat exchanger into one small device.
This new, hard-shell device was composed of a pre-sterilized, disposable polycar-

bonate plastic shell casing that now characterizes almost all ECMO devices [22].
This newest generation oxygenator was implemented on such a wide scale that the
Mayo Clinic converted to it almost exclusively by 1971 [21].
Most early, direct-contact extracorporeal oxygenators were severely limited
since both bubble and lm oxygenators experience signicant rates of complication
when used for more than 4 h. A third type of oxygenator that could be used for
longer periods was available, but it was not as widely used as the Mayo-Gibbon or
DeWall-Lillehei devices. The rotating disc oxygenator operated on a principle simi-
lar to the Mayo-Gibbon device, but it more closely resembled earlier, rotating cyl-
inder devices. The oxygenating chamber contained a series of vertical discs arrayed
horizontally along an axle or spindle. These discs were partially submersed in the
blood reservoir and rotated along their central axes to raise a lm of blood onto their
surface, thereby oxygenating the blood [30, 31]. This type of device was rst pio-
neered experimentally by Bjork in 1950 and successfully used clinically by Melrose
in 1953 [32]. By 1956, Dr. Frederick Cross and Dr. Earl Kay had completed their
Kay-Cross rotating disc oxygenator which gained some acceptance in the United
States [33]. Later improvements were completed in 1960 by Osborn, Bramson, and
Gerbode [31]. Rotating disc oxygenators enjoyed continued use up through the
1960s and 1970s. The rotating disc oxygenators became even more efcient and
safe with improvements such as Teon-coated disks, silicone-coated Pyrex
encasements, and more affordable disposable, pre-sterilized discs [31]. These adap-
tations allowed the rotating disc oxygenators to become the preferred gas exchanger
for long duration cases before the advent of membrane lungs.
Despite early success during the 1950s, 1960s, and 1970s, there are a number of
disadvantages related to the use of direct-contact oxygenating systems. The Mayo-
Gibbon and Kay-Cross oxygenators were very difcult to maintain and sterilize
between use due to their sheer size and complexity. Also, both of these oxygenators
required large priming volumes of blood. Morbidity and mortality were shown to
increase in direct correlation with intraoperative blood transfusions as few as one
unit [34]. By 1961, N. Zuhdi and colleagues introduced the practice of hemodilution
whereby hypothermic dextrose (or saline) solutions were mixed with blood for cir-
cuit priming volumes [35, 36]. Hemodilution improved the general course of total
body perfusion [37, 38].
ECMO Technology: Silicone Rubber Membrane Lungs
Early membrane oxygenators had proven gas exchange ability. However, these
devices suffered from limitations of existing materials: (1) poor gas permeability,
(2) weak structural integrity, (3) difculty manufacturing, (4) blood-surface incom-
patibility, (5) membrane acting as an additional barrier to diffusive exchange, and
(6) difculty achieving evenly distributed blood ow. Researchers are still strug-
gling with these obstacles today.
In 1944, Kolff and Berk were testing their hemodialysis machine design when
they observed gas exchange in blood via passive diffusion across the large surface
area cellophane tubing [39]. By 1956, this observation led to creation of a mem-
brane gas exchanger composed of polyethylene [34]. Encouraged by the 1944 nd-
ings of Kolff and Berk, Clowes and colleagues were experimenting with materials
such as ethyl cellulose and polyethylene in the hopes of creating a membrane gas
exchanger [40, 41]. In 1956, they used Teon for their rst clinically applicable
gas exchanger [42].
Early membrane oxygenators were less traumatic to blood than direct-contact
oxygenators, but also had limitations. The early materials used to create lung mem-
branes were hydrophilic, and this greatly exacerbated plasma leakage with extended
use. Therefore, designers began to incorporate hydrophobic materials, such as
Teon, to prevent excessive plasma leakage [42, 43]. Unfortunately, the solubility
of CO2 in hydrophobic materials is much less than in hydrophilic materials. The
term membrane lung was then used to emphasize the primary goal of membrane
oxygenator designers which was efcient O2 and CO2 exchange similar to that of
anatomical lungs, not just the exchange of O2 alone.
In 1957, Kammermeyer reported that O2 and CO2 are 40 times more permeable
in silicone rubber (dimethylpolysiloxane polymer) than in Teon and that CO2
permeability is 8 times that of O2 in silicone rubber [43]. Designers quickly began
experimenting with silicone rubber to nd a solution to its low structural capacity
and tendency for forming pinholes. In 1959, Thomas developed a method to support
a thin, silicone rubber membrane on a mesh of nylon ber using a dip-coating
method. Two membranes were used in conjunction (layered) to alleviate the prob-
lem of plasma leakage through the pinhole defects. That same year, Dr. Nora Burns
was able to produce adequately thin, pinhole-free, and structurally sound silicone
rubber under the tutelage of Professor Dennis Melrose at Hammersmith Hospital,
London [44]. This material was so revolutionary and effective that its modern equiv-
alent is still employed today in long-term use devices that are used both experimen-
tally and clinically.
Very shortly after Burns innovation, many prominent researchers began to
employ silicone rubber membranes in their devices, either supported by nylon or
integral silicate material [45]. In 1960, E. Converse Pierce II introduced an updated
version of Clowes device [46]. Kolobow and Bowman published their work in
1963 on what they termed an alveolar membrane articial heartlung [47]. The
term alveolar membrane refers specically to the ability of silicone rubber
membranes to effectively replicate natural alveolar tissue. Bramson et al. produced
a silicone rubber membrane lung (SRML) oxygenator with an integrated heat-
exchanger in 1965 [48]. Land et al. submitted their experience utilizing membrane
materials in their oxygenator-dialyzer in 1967 [49]. That same year, Day et al. pub-
lished their investigations of cellophane and silicone rubber membranes [50].
These investigations, and many others, demonstrated that SRMLs were not
achieving expected rates of gas exchange. Oxygen is capable of being exchanged at
a rate of 12,100 cc O2/m2/min through a 0.001 in. thick silicone rubber membrane
and 300 cc O2/m2/min through a 0.004 in. thick membrane (standard thickness for
the thin membranes introduced by Burns). CO2 is exchanged at a much higher rate.
Regardless of the physical deposition of each devices membranes (sheets or chan-
nels), researchers were only able to achieve around 30 cc O2/m2/min [50]. The arti-
cial materials used were not themselves the source of this discrepancy. Rather, it
was the blood that was creating this gas exchange barrier. In 1960, Marx et al.
showed that blood diffusion into a lm is proportional to the square of the thick-
ness of the blood lm and the diffusion resistance of the boundary layer [51]. More
succinctly, gas exchanges easily into the surface of a blood volume but diffusivity of
gas through blood itself is much slower. In order to facilitate more efcient gas dif-
fusion, designers sought a way to narrow the blood channels within SRML oxygen-
ator devices and to create secondary ow pathways.
Bodell et al. provided an innovative solution to one half of this problem in 1963
with the rst capillary oxygenator which utilized tubes between 100 and 500 m in
diameter as channels to alleviate the problem of thick blood streams as a barrier to
blood diffusivity [52]. Two orientations were investigated. In the extracapillary ori-
entation, gas was directed into capillaries while blood was directed around capillar-
ies. In the intracapillary orientation, blood was directed into capillaries and gas was
directed around capillaries. By 1965, Wilson et al. determined that an intracapillary
orientation was more effective given the restrictions of available materials [53]. The
intracapillary hollow-ber membrane lung delivered a twofold advantage. Gas dif-
fusivity through constricted channels was akin to the one-cell thick capillaries of the
natural alveolus, and there was more exacting control over the volumes in the blood
and gas chambers.
Various methods were devised in an effort to solve the other half of the blood
diffusivity issue and induce secondary ows throughout SRML oxygenators. Two
teams, the Dorson and Peirce groups, sought to create secondary ows passively.
Dorsons group directed blood into a helical tube arrangement while Peirces
group forced blood around intrusive conical elements, regularly interspersed
0.040 in. apart, in an envelope-style membrane [54, 55]. These efforts were efca-
cious, but lacked the dramatic results reported by other research teams operating at
the same time.
Kolobow et al. demonstrated foresight when they found a solution to the problem
of secondary ow generation as early as 1963 [56]. This group investigated the use
of hypobaric oxygenation to increase membrane surface area which had the added
benet of actively inducing secondary blood ows. Application of intermittent suc-
tion allowed for cyclic deformation of the exible membrane material into the gas
chamber of the device, creating larger surface areas for diffusion and inducing sec-
ondary ows that dramatically increased oxygen diffusion rates. Kolobow et al.
reported oxygen diffusion rates in excess of 200 cc O2/m2/min [56]. The Kolobow
device was manufactured rst by SciMed, then AveCor, and nally MedTronic. It
became the only commercially available solid silicone rubber membrane lung
(SRML) device to be consistently utilized over several decades. At their peak,
SRMLs were so effective they were used in 90 % of all reported ECMO treatments
conducted worldwide.
With efcient SRMLs in hand, the issues of extended anesthetization, vascular

access, bleeding, clotting, and device longevity could be addressed. In 1968,
Kolobow et al. reported that extrathoracic cannulation could supply sufcient blood
ow to sustain a subject, removing the need for more invasive central/thoracic can-
nulation [56]. Moreover, Bartlett et al. developed a method for controlling the titra-
tion of heparin to specic clotting times which resulted in more effective systemic
anticoagulation [57, 58]. In 1969, Bartlett et al. developed an updated oxygenator
design to actively induce secondary ows, a feature which increased efciency and
proved worthy of further renement [57].
ECMO Technology: Microporous Devices
In 1960, teams began experimenting with alternative microporous materials [59].

The micropores allowed for more direct contact between oxygen and blood, thereby
increasing efciency above that observed with the SRMLs. Initially, polyethylene
and hydrophobic cellulose acetate polymers were used to provide 110 m pores.
Later, 0.02 m pores were achieved with polypropylene, which resulted in direct
blood contact and much less plasma leakage [60]. These smaller micropores pre-
vented plasma leakage by surface tension interactions around the pore orice.
Microporous devices underwent a period of renement and improvement similar to
the SRMLs. By the 1980s, these simple, efcient, low blood impact, microporous
devices replaced bubble oxygenators for short-term surgical use with silicone rub-
ber devices remaining in use for longer-term ECMO applications.
An exciting spinoff technology was the intravascular oxygenator/carbon dioxide
removal device developed by Mortensen et al. in 1987 [61]. In 1991, building on
Mortensens design, Cox et al. modied the device to be inserted directly into the
patients vena cava [62]. The IVOX device consists of a bundle of hollow micropo-
rous polypropylene bers, coated in silicone rubber to prevent plasma leakage, with
circulatory support provided via a dual-lumen gas conduit. The bers are directly
exposed to the natural circulation of the body and can achieve up to 30 % of the total
gas exchange requirement of subjects [62]. Although gas exchange proved inade-
quate in clinical application, much was learned about the surface area limits of the
vena cava, insertion techniques, and gas exchange strategies in a device that accom-
plished mostly CO2 removal [63, 64].
ECMO Technology: Nonmicroporous Devices
Hollow-ber membranes composed of newer nonmicroporous materials provide

oxygenator performance previously unobtainable in silicone rubber or polypropyl-
ene membranes. In 2002, an investigation of poly-4-methyl-1-pentene (PMP)
showed that there was a signicant reduction in platelet consumption and resistance
to ow with hollow-ber, nonmicroporous PMP oxygenators when compared to
SRML equivalents [65]. PMP devices with infused heparin coatings signicantly
outperformed existing modalities to such a degree that some ECMO centers
switched to their use exclusively. Moreover, in 2005, Toomasian et al. at the
University of Michigan reported that PMP devices had better gas exchange, more
platelet preservation, and appreciably lower pressure gradients than commercially
available descendants of the Kolobow coil-type SRML [66]. These advantages
make PMP membranes ideal for extended ECMO/ECLS. The low resistance of
PMP oxygenators makes them ideally suited for use in tandem with centrifugal
pumps or even for use in pumpless arteriovenous ECMO protocols [65, 66].
The success of new nonmicroporous materials such as PMP has encouraged
innovation along similar lines in older nonmicroporous materials. Motomura et al.
from Fuji Systems Corporation in Yokohama, Japan, in conjunction with their asso-
ciates at Baylor College of Medicine, Texas, have developed a silicone rubber
hollow-ber oxygenator (SRHF) with ultra-thin membrane walls (35 m thick)
[67]. Their device, the Fuji-B.C.M. ECMO oxygenator module, outperforms tradi-
tional coil-type SRML oxygenators, both in terms of gas exchange and prevention
of hemolysis while still maintaining blood compatibility. Also, the plasma leakage
that characterizes microporous polypropylene membranes during long-term use is
absent in their device. This ultra-thin SRHF device provides an alternative to the
successful PMP devices [65, 67].
ECMO Configurations
Application and management of the different modalities of ECMO are described in

detail in Chap. 6. We will briey outline the modalities from a historical perspective.
In the original VA ECMO conguration, venous blood is drained and returned
through a central artery (especially the right carotid in neonates), providing heart
and lung support. As ECMO was further developed and critical care improved, there
was a need for different circuit congurations to provide distinct types of circulatory
support for different diseases and conditions. In the VV ECMO mode, the venous
blood is drained and returned to the venous system to achieve total O2 and CO2 gas
exchange without cardiac support. A third type of ECMO is arteriovenous (AV),
which can achieve partial to total CO2 removal at ows of 1020 % of cardiac output
[68]. While the disease status of the patient dictates which ECMO conguration to
use, all ECMO subtypes provide gas exchange or circulatory support, allowing fail-
ing organs to rest and potentially recover while the primary condition is treated.
VA ECMO
In the early days of ECMO, the VA mode was used almost exclusively since ECMO
developed from cardiopulmonary bypass. In VA ECMO, deoxygenated blood drains
from the central venous system most often accessing the right internal jugular vein,
is pumped through the gas exchanger, and then returned to the patient through the
right common carotid artery into the aortic arch [69]. VA ECMO provides both
cardiac and respiratory support and is frequently used in patients who are hemody-
namically unstable, are post-cardiotomy, or have heart failure.
VV ECMO
In VV ECMO, deoxygenated blood is drained from the superior or inferior vena

cava, pumped through the gas exchanger, and returned through the venous system
toward the heart. VV ECMO only supports the lungs and can be employed only
when the patients heart is strong enough to support circulation. In the past, VV
ECMO utilized an internal jugular vein cannula and a femoral vein cannula which
prohibited patient movement. This two-cannula blood access typically requires
patient sedation while on ECMO. Variations include a double-lumen cannula
(DLC), especially for use in neonates and children [70, 71]. Although clinically
successful, the limitations of the rst generation DLC included recirculation and
limited ows. More recently, the FDA-approved AvalonElite double-lumen
cannula (DLC) for adults (Avalon Laboratories, LLC, Los Angeles, CA, USA)
[72] and the OriGen double-lumen cannula (OriGen Biomedical, Austin, TX)
[73] have shown promise as a replacement for the two-cannula VV ECMO tech-
nique [74, 75]. With this DLC technique, the patient can be extubated and con-
scious while on ECMO, allowing patient mobility, or in select cases, bedside
exercise [76, 77].
Extracorporeal Cardiopulmonary Resuscitation (ECPR)
Extracorporeal cardiopulmonary resuscitation (ECPR) refers to the use of VA car-

diopulmonary bypass (usually using percutaneous access) to provide short-term life
support for patients following cardiopulmonary arrest [78]. This application of
ECMO is frequently utilized in the emergency room or cardiac catheterization lab
to support acute shock for the time they need to improve from an acute event or
reach the operating room for surgery. Patients in need of ECPR often only stay on
the circuit for a short time because the circuit parts (oxygenator) are designed for
short-term, less than 1-day use. Highly trained teams that can effectively deploy
ECPR in minutes are essential to success in these acute settings. As ECPR criteria
and outcomes are evaluated to assess the risks and benets, application for the treat-
ment or triage of cardiogenic shock should increase.
Clinical ECMO: Adult Respiratory Distress Syndrome
Recent technological advances in ECMO technology have allowed physicians to

initiate ECMO protocols earlier in ARDS, which appears to improve long-term lung
function [7982]. By initiating ECMO earlier in the course of ARDS disease pro-
gression, lung function could be preserved by minimizing mechanical ventilation-
induced lung damage, diminishing systemic stress generated by ARDS, and
potentially disrupting disease progression.
ECMO was rst used successfully on an adult acute ARDS patient in 1972 by
Hill et al. [83], and the rst attempt at a multicenter, randomized ECMO trial in
adult ARDS patients was completed in 1979 by Zapol et al. [84]. This study found
no signicant difference in survival between patients receiving ECMO versus those
receiving conventional therapy for ARDS [84]. In fact, the trial was aborted before
one-third of the planned patient population had participated in the trial because of
the dismal 9.5 % (4 out of 42) survivability rate. According to one of the investiga-
tors, Dr. Robert Bartlett, the trial was an example of how not to do a trial because
it was too early in the development of ECMO and participating centers did not have
enough training to properly execute the recommended ECMO protocol (Bartlett,
2013; unpublished interview). Additional issues with the trial included frequent
uncontrolled bleeding and a high percentage of inuenza/pneumonia patients in the
study groups [85]. Although this study failed to reveal the potential of ECMO as a
viable intervention for critically ill patients and inhibited the use of ECMO in adults
for decades, it highlighted the need for proper patient selection, better training pro-
tocols for ECMO, and technology improvement [85].
A qualied team utilizing a standard treatment protocol was determined neces-
sary to accrue the best possible outcomes in ECMO treatments [85, 86]. The ECMO
team not only consists of physicians (thoracic surgeons and/or pulmonary/critical
care physicians), but also perfusionists, registered nurses (RNs), and registered
respiratory therapists (RRTs). Those involved in the deployment and management
of ECMO must also be experienced in critical care medicine and fully comprehend
the respiratory, hemodynamic, renal, and coagulation pathophysiology associated
with ECMO [85, 86].
For over a decade following the 1979 Zapol study, there were no further clinical
studies of ECMO use in adult ARDS patients. In 1994, a randomized clinical trial
evaluated the effect of extracorporeal CO2 removal following pressure-controlled
inverse-ratio ventilation on ARDS patient survival [87]. This 1994 clinical trial
found no signicant difference in survival between the experimental and control
groups (33 % vs. 42 %). The negative results of this trial, along with the great cost
and training necessary to run adult ECMO protocols, have inhibited ECMO from
becoming a mainstream therapy in adult patients to this day [88].
After failure to show clinical signicance of ECMO in adult ARDS, patient man-
agement largely focused on the conventional therapies of lung-protective ventila-
tion, prone positioning, corticosteroids, and nitric oxide [89]. Despite improvements
in the care of critically ill patients, other than low tidal volume ventilation, ventilator
strategies have not improved long-term survival of ARDS patients. Following the
success of ECMO in the neonatal population (see below) and its expanded use
throughout the late 1970s and 1980s, physicians turned to ECMO once again in the
treatment of adult ARDS in the late 1980s and early 1990s [85]. Further support for
ECMO in adult patients came from: (1) the positive results of a European study that
analyzed the use of a pumpless extracorporeal lung assist device over the course of
a decade [90], (2) advancements in ECMO technology and management practices,
and (3) general improvements in critical care [89]. Together, these factors spurred
clinicians to reevaluate ECMO in adult ARDS patients in the late 1990s.
Initially, small studies and case reports predominated, with studies such as one by
Kolla et al. [91] that reported a 54 % survival (n = 100) in adult ARDS patients. More
conclusive evidence supporting ECMO as a management therapy in ARDS came in
2009, when a large multicenter randomized controlled trial, the CESAR
(Conventional Ventilatory Support versus Extracorporeal Membrane Oxygenation
for Severe Adult Respiratory Failure) trial, showed a 16 % increase in 6-month sur-
vival without severe disability for the ECMO group versus the conventional treat-
ment group (63 % vs. 47 %) [92]. However, this study was not without limitations.
All patients were treated by best therapy at the local hospital, or referred to a sin-
gle ECMO center and thus treated by the same staff [92]. Utilizing a single referral
center diminished variability in administering ECMO, but imposed transportation
morbidity and mortality and is not a realistic approach for countries with more than
one ECMO center [88]. Moreover, many ARDS trials have shown that survival with
a low tidal volume critical-care protocol is approximately 60 % [93, 94]. Critics
noted the modest benets seen in the CESAR trial may not support the inclusion of
ECMO in the standard critical care protocols for adult ARDS patients.
During the 2009 H1N1 pandemic, many hospitals successfully employed VV
ECMO as a rescue treatment for ARDS [9598]. In this VV ECMO experience,
68 % of the H1N1 patients survived following ECMO treatment [98]. Among these
patients, mortality risk was correlated with extrapulmonary organ function just prior
to receiving the VV ECMO treatment [99]. Moreover, this study highlighted the
feasibility and efcacy of using a national VV ECMO protocol at 14 different ter-
tiary centers [98]. A recent upsurge in H1N1 patients in winter of 20132014 has
seen a much broader application of VV ECMO treatment at an earlier stage of
ARDS. The initial results appear promising but outcomes data is not yet available.
Clinical ECMO Application: Neonatal Acute Respiratory

Distress Syndrome
In 1975, an illiterate Mexican teenager crossed the U.S. border to give birth to a
baby girl. On her way to Los Angeles from Baja, Mexico, the young mother went
into labor. After a long and difcult labor at Orange County Medical Center, the
newborn was hypoxic, hypotensive, and experiencing seizures. Her arterial pO2 was
Fig. 14.3 Dr. Robert

H. Bartlett was the rst
physician to successfully
utilize ECMO in neonates.
Photo courtesy of Dr.
Bartlett, with permission
12 [99]. At the time, Dr. Bartlett and his colleagues were working on developing
ECMO in adult patients. Until that night, ECMO had been perceived as
unsuccessfulapproximately 150 adult patients had been placed on ECMO, with
only a 1015 % survival rate [100]. Moreover, ECMO had never been employed
successfully in children. Yet, the baby, Esperanza, had run out of options and Dr.
Bartlett was called [99] (Fig. 14.3).
This rst neonatal case would change the face of ECMO forever. After the
infants mother consented, the cyanotic infant was placed on VA ECMO using
right jugular vein-carotid artery cannulation. At rst, Esperanzas lung function
was not recovering while on ECMO. Bartlett ligated the patent ductus arterio-
sus. Once Esperanzas pulmonary hypertension subsided, she was slowly taken
off life support. Now, almost 40 years later, Esperanza has grown into a mature
woman (Fig. 14.4). Had it not been for this opportunistic neonatal case, Dr. Bartlett
believes that their ECMO research would have been abandoned (Bartlett, unpub-
lished interview).
Following the success of ECMO in Esperanzas case, Bartlett published an
ECMO study on 45 neonates in 1982 [101]. In this study, the newborns selected for
ECMO treatment had already failed to recover on traditional therapies and were
labeled as moribund. Bartlett reported a greater than 50 % survival in these newborn
Fig. 14.4 (a) Photograph of Esperanza while she was on ECMO in 1975. (b) Photograph of Dr.
Bartlett and Esperanza at age 34. Photos courtesy of Dr. Bartlett, with permission
patients that had only a 10 % chance of survival using conventional treatments.

Bartlett was subsequently recruited to University of Michigan which became the
home of ECMO in the United States for the next several decades.
As neonatal ECMO use continued to expand, many in the medical eld called for
a randomized clinical trial. Yet, how do you ethically randomize a neonatal patient
to a control group when, in expert hands the ECMO treatment appeared signicantly
more benecial than traditional treatment? Furthermore, how can one justify treat-
ing persistent pulmonary hypertension of the newborn (PPHN) with conventional
therapies that have established 80 % or greater mortality rates when ECMO centers
in the mid-1980s had survival rates of 80 % or greater when treating PPHN [102]?
Bartletts group decided on a play-the-winner strategy to analyze neonatal
ECMO and conventional treatment side-by-side in order to optimize the number of
patients that might receive the most effective treatment [103]. At the beginning of
the treatment protocol, the randomized play-the-winner method provided an
equal chance of the patient being assigned to the experimental or control groups. As
the study progressed, each previous patients outcome inuenced the odds of subse-
quent patients being assigned to a particular group. For instance, the rst patient had
a 50:50 chance of receiving ECMO. If this patient was randomly assigned to the
ECMO group and survives, another chance at ECMO was added to the randomiza-
tion pool. If the patient did not survive the ECMO treatment, then another chance at
the conventional therapy was added. This protocol of adding another chance of
either ECMO or conventional therapy continued until statistical signicance was
reached. In this play-the-winner neonatal study, 11 neonates received ECMO and
survived, while one received conventional therapy and died [103]. This prospective,
controlled, randomized study showed that ECMO provides lungs with the necessary
support to heal while improving outcomes in neonates when compared to standard
ventilator therapy in the treatment of severe respiratory failure.
Following the publication of Bartletts play-the-winner neonatal ECMO trial,
many called for more data on the efcacy of ECMO. A retrospective review of
PPHN treatments between 1982 and 1983 at two Harvard hospitals revealed that
11 of 13 (85 %) infants suffering from PPHN and treated with conventional ventila-
tor therapies had died [104]. These ndings suggested that ECMO could signi-
cantly increase survival rates when compared to conventional ventilator therapies in
specic patient populations. Consequently, Pearl ORourke conducted a two-phase
trial to study ECMO versus conventional medical therapy (CMT). In total, 39 new-
borns with severe PPH and respiratory failure that had an 85 % chance of dying
were randomly assigned to either ECMO or conventional medical treatment (CMT)
groups [104]. After the fourth death in either treatment group, randomization would
be suspended and future patients would receive the superior treatment during phase
II. Phase II would similarly be suspended if four deaths occurred or until it was
shown that the phase II treatment was signicantly better than the suspended treat-
ment protocol. During the rst phase, all nine infants survived the ECMO treatment
while only six out of ten infants survived CMT. After the fourth CMT death, phase
II began in which all patients received ECMO. In the second phase, 19 out of 20
infants treated with ECMO survived. Overall, 97 % (28/29) of the patients survived
ECMO treatment, while just 60 % (6/10) survived CMT. ORourke encountered
constant criticism from either those who thought the study should be a 50:50 trial or
those who thought the control group was unethical.
Between 1993 and 1995, a large-scale randomized trial of neonatal ECMO in
mature (at least 35 weeks gestation and birth weight of at least 2 kg) newborn infants
with severe respiratory failure was undertaken in the United Kingdom (UK) [105].
This study enrolled 185 patients from 55 hospitals in which patients were either
maintained at their original hospital for Conventional Medical Therapy (CMT) or
referred to one of ve specialist centers for ECMO treatment. Importantly, there were
numerous causes of respiratory failure among this patient population, allowing
observation of subgroup trends. Overall, 63 of 93 (68 %) ECMO patients survived,
while only 38 of 92 (41 %) CMT patients survived. Moreover, the difference in sur-
vival rates remained regardless of the primary diagnosis, disease severity, and referral
center. The results of this study indicated that ECMO produces a survival benet for
neonates suffering from a variety of potentially reversible respiratory failure [105].
A majority of newborn respiratory failure conditions are potentially reversible
[103]. Moreover, neonatal lungs might be more capable of recovery than older lungs
because they are typically placed on ECMO at a much earlier time point in their
disease progression. All of these factors play a role in the increased neonatal sur-
vival rates on ECMO when compared to pediatric and adult cases.
Clinical ECMO Application: Pediatric Acute

Respiratory Distress Syndrome
ECMO was rst successfully used in adult and neonatal ARDS patients during
the 1970s [86, 106]. In the 1980s and 1990s, as adult ECMO fell out of favor;
pediatric application rapidly evolved for severe acute cardiopulmonary failure.
ECMO became standard for complex congenital cardiac surgery whenever the
patent failed to wean from CPB or experienced postoperative cardiac or respira-
tory failure. Across the United States, all of the major childrens hospitals have
incorporated ECMO capability into their pediatric heart, pediatric ICU, and neo-
natal programs [107].
Among the pediatric population, ECMO is used in the management of respira-
tory failure associated with ARDS, pneumonia, sepsis, burns, drowning, inhalation
injury, and other potentially reversible pulmonary diseases, trauma, or injury
[69, 108]. By January 2013, 16.3 % of all reported ECMO cases (8674 out of 53,190
total cases) were in the pediatric population (ELSO Registry, July 2012). Despite
the widespread application, no randomized controlled trial has been completed that
specically evaluates ECMO protocols in pediatric patients. Furthermore, pediatric
indications for ECMO tend to be much more varied and judgmental than in the
neonatal population. Retrospective studies and single-center reports indicate that
pediatric ECMO benets high-risk respiratory failure patients and decreases mor-
tality [103105, 109, 110]. In one study at the University of Michigan, data from 25
non-neonatal pediatric patients treated with ECLS for severe respiratory failure
(from 1991 to 1993) were analyzed [110]. Of the 25 selected patients, 22 survived
to discharge, which represents a survival rate of 88 %. In a larger, multicenter, ret-
rospective study, 331 pediatric patients admitted during 1991 from 32 hospitals
were examined [109]. This study revealed that ECMO was associated with improved
survival in acute respiratory failure patients. Moreover, once patients were matched
according to diagnosis and risk into pairs, the ECMO-treated patients had a signi-
cantly lower mortality rate than non-ECMO patients that received conventional
therapies (26.4 % vs. 47.2 %) [109]. Collectively, these studies of the early 1990s
generated more enthusiasm toward pediatric ECMO in respiratory failure patients.
Clinical ECMO Application: Bridge to Lung Transplantation
Every year, numerous patients on the United Network of Organ Sharing (UNOS)
waitlist die before they can receive a lung transplant [111]. Moreover, the longer
patients spend on the waitlist, the higher their chance of death post-transplant due to
their deconditioning prior to transplantation. Since the lung allocation scoring
(LAS) system was implemented, waitlist times have decreased, but only for specic
diseases [80]. Individuals awaiting lung transplantation from diseases such as pul-
monary hypertension secondary to pulmonary brosis still have long waitlist times
[77]. While awaiting transplant, such patients often deteriorate requiring mechani-
cal ventilation [81]. Yet, mechanical ventilation is associated with an increased risk
of death in the rst 6 months following lung transplantation [112].
In the past 5 years, the outcomes of using VV ECMO as a bridge to lung trans-
plantation have vastly improved, especially in patients with a high LAS score
(Hoopes, 2013, unpublished interview). Patients bridged to a lung transplant with
ECMO had a survival advantage compared to patients with a LAS score greater than
60 that were in the national Scientic Registry of Transplant Recipients database in

a recent retrospective, multicenter study [80]. A critical part of the improved out-
comes was proper patient selection and early ECMO initiation. Utilizing ECMO
early as a bridge toward lung transplant is critical for patients because respiratory
failure requiring mechanical ventilation is associated with increased morbidity and
mortality following lung transplant. Moreover, ECMO can facilitate the weaning of
lung transplant patients off mechanical ventilation. Decreased time on mechanical
ventilators, ambulation, and physical therapy can increase the patients condition-
ing, decrease frailty, and decrease ventilator acquired pneumonia [112].
Recent technology has allowed for great strides in bridge-to-transplant strategies
with patient ambulation, especially for adult patients. The Wang-Zwische double-
lumen cannula (DLC) was introduced in 2008 and was commercialized as the
Avalon-Elite DLC (Avalon Laboratories LLC, Rancho Dominguez, CA). Since
FDA approval in 2009, this DLC has increased the opportunity for extubation and
ambulation. ECMO diminishes time on mechanical ventilation and can allow
patients to remain conscious facilitating physical therapy and potentially ambula-
tion (Fig. 14.5) [113]. Substantially longer periods of cannulation are possible with
fewer complications than multiple site cannulation strategies.
Fig. 14.5 Example of ECMO ambulation: (a) extubated, bed-ridden patient on ECMO; (b) patient
on ECMO and able to sit up in bed and ambulate. Reproduced from: Tulman DB, Stawicki SPA,
Whitson BA, Gupta SC, Tripathi RS, Firstenberg MS, et al. Veno-venous ECMO: a synopsis of
nine key potential challenges, considerations, and controversies. BMC Anesthesiol 2014; 14, 65
(Open Access article)
In a 2011 retrospective study by Garcia et al. [76], three patients were ambulatory
while on VV ECMO, attempting bedside exercises, while one patient was able to
walk on a treadmill. In 2012, ambulatory single cannula-venous VV ECMO was
used successfully in four cystic brosis patients awaiting bilateral lung transplanta-
tion [77]. Renements to the other ECMO circuit components including pumps,
circuit coatings, and decreased bulk, allow for long-term patient support with poten-
tial ambulation, until an organ becomes available for transplantation.
Clinical application of ambulatory ECMO as a bridge to lung transplant has rap-
idly gained acceptance. Several authors have reported series with 1020 patients
with approximately an 80 % survival after transplant. Hoopes et al. [79] reporting
on the USC/UK experience reported an 83 % overall survival with 25 of the last 27
alive to discharge following lung transplant. Most of the later patients had a percu-
taneous double-lumen catheter that allowed the patient out-of-bed, off the ventila-
tor, with ability to exercise on a treadmill. Ambulatory ECMO as a bridge is
supported by the recent medical literature regarding the avoidance of frailty in the
respiratory failure population.
Clinical ECMO Application: Supportive Technique

Following Transplantation
Despite a positive experience with bridge-to-transplant, many lung transplant spe-

cialists treat ECMO as a rescue technique for patients suffering from severe primary
graft dysfunction (PGD) following lung transplantation [113]. Studies have shown
that VV ECMO successfully supports patients with PGD until they recover suf-
cient lung function [113, 114]. Thus, ECMO appears to provide favorable support
for transplant patients even when ECMO support is deployed shortly after lung
transplantation.
The ELSO Registry
Following the successes of ECMO toward neonatal cardiorespiratory failure, in the

early 1980s several ECMO centers formed a club to exchange ideas and experiences
to improve outcomes. Physicians collected data, analyzed cases and technology,
shared experiences, and discussed complications [85]. In 1989, the steering commit-
tee and bylaws established the Extracorporeal Life Support Organization (ELSO)
[85]. Clinicians from across the United States gathered in Ann Arbor, Michigan to
determine criteria for ECMO and the future of ELSO (Fig. 14.6). Central to ELSO
is the ELSO Registry, which pools voluntarily reported ECMO cases into a single
database. Since 1990, the registry has seen the number of ECMO-capable centers
grow from 83 to 200 (Fig. 14.7) (ELSO Registry, Jan 2013). Over the past few
decades, the ELSO Registry has been a vital resource that provides clinicians a
Fig. 14.6 ELSO charter meeting in Ann Arbor, Michigan. Photo courtesy of ELSO, used with
permission
Fig. 14.7 Since 1990, the number of active ECLS centers has more than doubled. Beginning in
2003, the number of centers has steadily increased, increasing access to ECMO. Data collected
from the ELSO Registry January 2013
Table 14.2 Overall patient outcomes from ELSO registry (January 2013)
Total Survived ECLS Survived to Discharge
Neonatal
Respiratory 26,205 22,145 85 % 19,559 75 %
Cardiac 4987 3058 61 % 2010 40 %
ECPR 851 540 63 % 331 39 %
Pediatric
Respiratory 5656 3692 65 % 3183 56 %
Cardiac 6225 4034 65 % 3054 49 %
ECPR 1745 948 54 % 708 41 %
Adult
Respiratory 3761 2400 64 % 2084 55 %
Cardiac 2884 1581 55 % 1132 39 %
ECPR 876 325 37 % 241 28 %
Overall total 53,190 38,723 73 % 32,302 61 %
potent tool for retro-analysis. As of January 2013, the ELSO Registry contained
53,190 ECMO cases. Of these cases, 32,043 were neonates, 13,626 were children,
and 7521 were adults (Table 14.2) (ELSO Registry, January 2013). The ECLS pro-
cedure was survived by 38,723 patients (73 %) and 32,302 patients (61 %) survived
to decannulation or transfer. These statistics are remarkable when compared to the
results of the earliest attempts to establish ECMO as a practical and safe technique.
Conclusions
Since the 1970s, the ECMO eld has evolved and expanded exponentially. Through
technology improvements and the knowledge gained from the collective member-
ship of ELSO, published reports, and shared experiences including the ELSO
Registry, ECMO has ourished and saved thousands of patients that were otherwise
moribund. Use of the ELSO Registry and careful analysis of ECMO studies will
remain central to maintaining positive progress in the eld. The potential survival
benets of early ECMO deployment in the emergency room setting may make
ECMO the standard of care for patients with any form of potentially reversible car-
diogenic shock (Bartlett, unpublished interview, 2013). Moreover, the development
of circuits that do not require continuous anticoagulation would not only revolution-
ize modern-day ECMO, but would vastly impact the medical eld in general. As
technology continues to develop and researchers push the boundaries of ECMO,
more lives will be saved, but at a cost of more capital equipment and critical care
expertise.
References
1. Gibbon Jr JH, Miller BJ, Dobell AR, Engell HC, Voigt GB. The closure of interventricular
septal defects in dogs during open cardiotomy with the maintenance of the cardiorespiratory
functions by a pump-oxygenator. J Thorac Surg. 1954;28(3):23540.
2. Gille JP, Bagniewski AM. Ten years of use of extracorporeal membrane oxygenation (ECMO)
in the treatment of acute respiratory insufciency (ARI). Trans Am Soc Artif Intern Organs.
1976;22:1029.
3. Gaffney AM, Wildhirt SM, Grifn MJ, Annich GM, Radomski MW. Extracorporeal life
support. BMJ. 2010;341:c5317.
4. Dalton HJ, Butt WW. Extracorporeal life support: an update of Rogers Textbook of Pediatric
Intensive Care. Pediatr Crit Care Med. 2012;13(4):46171.
5. Froelich C. Pediatric ECMO: old dog or new trick? Pediatric Grand Rounds
6. Gibbon Jr JH. Application of a mechanical heart and lung apparatus to cardiac surgery. Minn
Med. 1954;37(3):17185.
7. Cohen M, Lillehei CW. A quantitative study of the azygos factor during vena caval occlusion
in the dog. Surg Gynecol Obstet. 1954;98(2):22532.
8. Dennis C. A heart-lung machine for open-heart operations. How it came about. ASAIO Trans.
1989;35(4):76777.
9. Lillehei CW. Controlled cross circulation for direct-vision intracardiac surgery; correction of
ventricular septal defects, atrioventricularis communis, and tetralogy of Fallot. Postgrad
Med. 1955;17(5):38896.
10. Kurusz M. May 6, 1953: the untold story. ASAIO J. 2012;58(1):25.
11. Warden HE, Dewall RA, Read RC, Aust JB, Cohen M, Varco RL, et al. Total cardiac by-pass
utilizing continuous perfusion from reservoir of oxygenated blood. Proc Soc Exp Biol Med.
1955;90(1):24650.
12. Fazel S, Williams WG, Goldman BS. Thinned blood, monkey lungs, and the cold heart.
J Card Surg. 2004;19(3):2758.
13. Gibbon Jr JH. An oxygenator with large surface area to Volume ratio. J Lab Clin Med.
1939;24:11928.
14. Miller BJ, Gibbon JH, Fineburg C. An improved mechanical heart and lung apparatus; its use
during open cardiotomy in experimental animals. Med Clin North Am. 1953;1:160324.
15. Jones RE, Donald DE, Swan JC, Harshbarger HG, Kirklin JW, Wood EH. Apparatus of the
Gibbon type for mechanical bypass of the heart and lungs; preliminary report. Proc Staff
Meet Mayo Clin. 1955;30:10513.
16. Kirklin JW, Dushane JW, Patrick RT, Donald DE, Terry Jr HR, Burchell HB. et al. Intracardiac
surgery with the aid of a mechanical pump-oxygenator system (gibbon type): report of eight
cases. Proc Staff Meet Mayo Clin. 1955;30(10):2016.
17. Kirklin JW, Donald DE, Harshbarger HG, Hetzel PS, Patrick RT, Swan HJ, et al. Studies in
extracorporeal circulation. I. Applicability of Gibbon-type pumpoxygenator to human intra-
cardiac surgery: 40 cases. Ann Surg. 1956;144:28.
18. Dennis C. Perspective in review. One groups struggle with development of a pump-
oxygenator. Trans Am Soc Artif Intern Organs. 1985;31:111.
19. Stoney WS. Evolution of cardiopulmonary bypass. Circulation. 2009;119:284453.
20. Kirklin JW, Donald DE, Harshbarger HG, Hetzel PS, Patrick RT, Swan HJ. Studies in extra-
corporeal circulation. I. Applicability of Gibbon-type pump-oxygenator to human intracar-
diac surgery: 40 cases. Ann Surg. 1956;144(1):28.
21. Kirklin JW. Open-heart surgery at the Mayo Clinic. The 25th anniversary. Mayo Clin Proc.
1980;55(5):33941.
22. DeWall RA, Naja H, Roden T. A hard-shell temperature-controlling disposable blood
oxygenator. Use in clinical open-heart surgery. JAMA. 1966;197(13):10658.
23. Dewall RA, Gott VL, Lillehei CW, Read RC, Varco RL, Warden HE. Total body perfusion for
open cardiotomy utilizing the bubble oxygenator; physiologic responses in man. J Thorac
Surg. 1956;32(5):591603.
24. Clarke Jr LC, Gollan F, Gupta VB. The oxygenation of blood by gas dispersion. Science.
1950;111(2874):857.
25. DeWall RA, Bentley DJ, Hirose M, Battung V, Naja H, Roden T. A temperature controlling
(omnithermic) disposable bubble oxygenator for total body perfusion. Dis Chest. 1966;49(2):
20711.
26. Lillehei CW. A personalized history of extracorporeal circulation. Trans Am Soc Artif Intern
Organs. 1982;28:516.
27. Gott VL, Dewall RA, Paneth M, Zuhdi MN, Weirich W, Varco RI, Lillehei CW. A selfcon-
tained disposable oxygenator of plastic sheet for intracardiac surgery. Thorax. 1957;12:19.
28. Gott VL, Sellers RD, Dewall RA, Varco RL, Lillehei CW. A disposable unitized plastic sheet
oxygenator for open heart surgery. Dis Chest. 1957;32(6):61525.
29. Naef AP. The story of thoracic surgery. Toronto: Hogrefe and Huber; 1990.
30. Cross ES, Berne RM, Hirose Y. Evaluation of a rotating disc type oxygenator. Proc Soc Exp
Biol Med. 1956;93:2104.
31. Osborn JJ, Bramson ML, Gerbode F. A rotating disc blood oxygenator and integral heat
exchanger of improved inherent efciency. J Thorac Cardiovasc Surg. 1960;39:42737.
32. Melrose DG, Bramson ML, Osborn JJ, Gerbode F. The membrane oxygenator; some aspects
of oxygen and carbon dioxide transport across polyethylene lm. Lancet. 1958;1:10501.
33. Cross FS, Berne RM, Hirose Y, Jones RD, Kay EB. Description and evaluation of a rotating
disc type reservoir-oxygenator. Surg Forum. 1957;7:2748.
34. Kolff WJ, Efer DB, Groves LK, Peereboom G, Moraca PP. Disposable membrane oxygen-
ator (heart-lung machine) and its use in experimental surgery. Cleve Clin Q. 1956;23:
6997.
35. Zuhdi N, McCollough B, Carey J, Krieger C, Greer A. Hypothermic perfusion for open-heart
surgical procedures. Report on the use of a heart-lung machine primed with ve per cent
dextrose in water inducing hemodilution. J Int Coll Surg. 1961;35:31926.
36. Zuhdi N, McCollough B, Carey J, Greer A. Double-belical reservoir heart-lung machine
designed for hypothermic perfusion; primed with 5 per cent glucose in water; inducing hemo-
dilution. Arch Surg. 1961;82:3205.
37. Dewall RA, Lillehei CW. Simplied total body perfusion. Reduced ows, moderate
hypothermia, and hemodilution. JAMA. 1962;179:4304.
38. Dewall RA, Lillehei RC, Sellers RD. Hemodilution perfusions for open-heart surgery. Use of
ve per cent dextrose in water for the priming volume. N Engl J Med. 1962;266:107884.
39. Kolff WJ, Baltzer R. The articial coil lung. Trans Am Soc Artif Intern Organs. 1955;
1:3942.
40. Clowes Jr GHA, Hopkins AL, Kolobow T. Oxygen diffusion through plastic lms. Trans Am
Soc Artif Intern Organs. 1955;1:234.
41. Clowes Jr GH, Hopkins AL, Neville WE. An articial lung dependent upon diffusion of
oxygen and carbon dioxide through plastic membranes. J Thorac Surg. 1956;32:6307.
42. Clowes Jr GH, Neville WE. Further development of a blood oxygenator dependent upon the
diffusion of gases through plastic membranes. Trans Am Soc Artif Intern Organs.
1957;3:528.
43. Kammermeyer K. Silicone rubber as a selective barrier. Ind Eng Chem. 1957;49:16856.
44. Burns N. Production of a silicone rubber lm for the membrane lung. Biomed Eng.
1969;4:3569.
45. Iwahashi H, Yuri K, Nose Y. Development of the oxygenator: past, present, and future. J Artif
Organs. 2004;7:11120.
46. Peirce 2nd EC. A modication of the Clowes membrane lung. J Thorac Cardiovasc Surg.
1960;39:43848.
48. Bramson ML, Osborn JJ, Main FB, et al. A new disposable membrane oxygenator with inte-
gral heat exchange. J Thorac Cardiovasc Surg. 1965;50:391400.
49. Land AJ, Dos SJ, Carlson RG, et al. A new membrane oxygenator-dialyzer. Surg Clin North
Am. 1967;47(6):146170.
50. Day SW, Crystal DK, Wagner CL, Kranz JM. Properties of synthetic membranes in extracor-
poreal circuits. Am J Surg. 1967;114(2):3149.
51. Marx TI, Snyder WE, St John AD, Moeller CE. Diffusion of oxygen into a lm of whole
blood. J Appl Physiol. 1960;15:11239.
52. Bodell BR, Head JM, Head LR, Formolo AJ, Head JR. A capillary membrane oxygenator.
J Thoracic Cardiovasc Surg. 1963;46:63950.
53. Wilson R, Shepley DJ, Llewellyn-Thomas E. A membrane oxygenator with a low priming
volume for extra-corporeal circuit. Can J Surg. 1965;8:30911.
54. Dorson Jr W, Baker E, Hull H. A shell and tube oxygenator. Trans Am Soc Artif Intern
Organs. 1968;14:2429.
55. Peirce 2nd EC, Dibelius NR. The membrane lung: studies with a new high permeability co-
polymer membrane. Trans Am Soc Artif Intern Organs. 1968;14:2206.
56. Kolobow T, Zapol W, Pierce JE, Keeley AF, Replogle RL, Haller A. Partial extracorporeal
gas exchange in alert newborn lambs with a membrane articial lung perfused via an A-V
shunt for periods up to 96 hours. Trans Am Soc Artif Intern Organs. 1968;14:32834.
57. Bartlett RH, Isherwood J, Moss RA, Olszewski WL, Polet H, Drinker PA. A toroidal ow
membrane oxygenator: four day partial bypass in dogs. Surg Forum. 1969;20:1523.
58. Bartlett RH, Drinker PA, Burns NE, et al. The toroidal membrane oxygenator: design, perfor-
mance, and prolonged bypass testing of a clinical model. Trans Am Soc Artif Intern Organs.
1972;18:36974.
59. McCaughan JS, Weeder R, Schuder JC, Blakemore WS. Evaluation of new nonwettable mac-
roporous membranes with high permeability coefcients for possible use in a membrane
oxygenator. J Thorac Cardiovasc Surg. 1960;40:57481.
60. Lim MW. The history of extracorporeal oxygenators. Anaesthesia. 2006;61(10):98495.
61. Mortensen JD. An intravenacaval blood gas exchange (IVCBGE) device. A preliminary
report. ASAIO Trans. 1987;33(3):5703.
62. Cox Jr CS, Zwischenberger JB, Traber DL, Traber LD, Haque AK, Herndon DN. Use of an
intravascular oxygenator/carbon dioxide removal device in an ovine smoke inhalation injury
model. ASAIO Trans. 1991;37(3):M4113.
63. Kaushik M, Wojewodzka-Zelezniakowicz M, Cruz DN, Ferrer-Nadal A, Teixeira C, Iglesias
E, et al. Extracorporeal carbon dioxide removal: the future of lung support lies in the history.
Blood Purif. 2012;34(2):94106. doi:10.1159/000341904. Epub 2012 Oct 24.
64. Zwischenberger JB, Tao W, Bidani A. Intravascular membrane oxygenator and carbon
dioxide removal devices: a review of performance and improvements. ASAIO J. 1999;
45(1):416.
65. Peek GJ, Killer HM, Reeves R, Sosnowski AW, Firmin RK. Early experience with a
polymethyl pentene oxygenator for adult extracorporeal life support. ASAIO J. 2002;48(5):
4802.
66. Toomasian JM, Schreiner RJ, Meyer DE, Schmidt ME, Hagan SE, Grifth GW, et al. A poly-
methylpentene ber gas exchanger for long-term extracorporeal life support. ASAIO J.
2005;51(4):3907.
67. Motomura T, Maeda T, Kawahito S, Takano T, Nonaka K, Linneweber J, et al. Development
of silicone rubber hollow ber membrane oxygenator for ECMO. Artif Organs. 2003;27(11):
10503.
68. Bartlett RH. The development of prolonged extracorporeal circulation. In: Zwischenberger
JB, Steinhorn RH, Bartlett RH, editors. ECMO: extracorporeal cardiopulmonary support in
critical care. Michigan, Ann Arbor: Extracorporeal Life Support Organization; Ann Arbor,
Michigan, 2000.
69. Crawford D, Harvey B. A review of extracorporeal membrane oxygenation in the UK. Nurs
Child Young People. 2012;24(8):1822.
70. Anderson 3rd HL, Otsu T, Chapman RA, Barlett RH. Venovenous extracorporeal life support
in neonates using a double lumen catheter. ASAIO Trans. 1989;35(3):6503.
71. Otsu T, Merz SI, Hultquist KA, Attorri RJ, Anderson 3rd HL, Schefer DE, et al. Laboratory
evaluation of a double lumen catheter for venovenous neonatal ECMO. ASAIO Trans.
1989;35(3):64750.
cannula-toward a percutaneous and ambulatory paracorporeal articial lung. ASAIO J.
2008;54(6):60611.
73. Rais-Bahrami K, Walton DM, Sell JE, Rivera O, Mikesell GT, Short BL. Improved
oxygenation with reduced recirculation during venovenous ECMO: comparison of two
catheters. Perfusion. 2002;17(6):4159.
74. Garcia JP, Iacono A, Kon ZN, Grifth BP. Ambulatory extracorporeal membrane oxygen-
ation: a new approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg. 2010;
139:e1379.
75. Reeb J, Falcoz PE, Santelmo N, Massard G. Double lumen bi-cava cannula for veno-venous
extracorporeal membrane oxygenation as bridge to lung transplantation in non-intubated
patient. Interact Cardiovasc Thorac Surg. 2012;14:1257.
76. Garcia JP, Kon ZN, Wu Z, Iacnon AT, McCormick B, Grifth BP. Ambulatory veno-venous
extracorporeal membrane oxygenation: innovation and pitfalls. J Thorac Cardiovasc Surg.
2011;142(4):75561.
77. Hayes Jr D, Kukreja J, Tobias JD, Ballard HO, Hoopes CW. Ambulatory venovenous
extracorporeal respiratory support as a bridge for cystic brosis patients to emergent kung
transplantation. J Cyst Fibros. 2012;11(1):405.
78. Bellezzo JM, Shinar Z, Davis DP, Jaski BE, Chillcott S, Stahovich M, et al. Emergency
physician-initiated extracorporeal cardiopulmonary resuscitation. Resuscitation. 2012;83(8):
96670.
79. Hoopes CW, Kukreja J, Golden J, Davenport DL, Diaz-Guzman E, Zwischenberger JB.
Extracorporeal membrane oxygenation as a bridge to pulmonary transplantation. J Thorac
Cardiovasc Surg. 2013;145(3):8627. discussion 8678.
80. Diaz-Guzman E, Hoopes CW, Zwischenberger JB. The evolution of extracorporeal life
support as a bridge to lung transplantation. ASAIO J. 2013;59(1):310.
81. Iglesias M, Jungebluth P, Petit C, Matute MP, Rovira I, Martnez E, et al. Extracorporeal lung
membrane provides better lung protection than conventional treatment for severe postpneu-
monectomy noncardiogenic acute respiratory distress syndrome. J Thorac Cardiovasc Surg.
2008;135:136271.
82. Meyer A, Struber M, Fischer S. Advances in extracorporeal ventilation. Anesthesiol Clin.
2008;26:38191.
of the Bramson membrane lung. N Engl J Med. 1972;286:62934.
84. Zapol WM, Snider MT, Hill JD, Fallat RJ, Edmunds LH, Morris AH, et al. Extracorporeal
JAMA. 1979;242(20):21936.
85. Zwischenberger JB, Bartlett RH. Extracorporeal life support: an overview. In: Van Meurs K,
Lally KP, Peek G, Zwischenberger JB, editors. ECMO extracorporeal cardiopulmonary support
in critical care. 3rd ed. Ann Arbor: Extracorporeal Life Support Organization; 2005. p. 127.
86. Kanto WP, Shapiro MB. The development of prolonged extracorporeal circulation. In:
Zwishcenberger JB, Steinhron RH, Bartlett RH, editors. ECMO: Extracorporeal cardiopul-
monary support in critical care. 2nd ed. Extracorporeal Life Support Organization; 2000.
87. Morris AH, Wallace CJ, Menlove RL, Clemmer TP, Orme Jr JF, 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):
295305.
88. Zwischenberger JB, Lynch JE. Will CESAR answer the adult ECMO debate? Lancet.
2009;374:13078.
89. Sadahiro T, Oda S, Nakamura M, Watanabe E, Abe R, Nakada TA, et al. Trends in and per-
spectives on extracorporeal membrane oxygenation for severe adult respiratory failure. Gen
Thorac Cardiovasc Surg. 2012;60:192201.
90. Flrchinger B, Philipp A, Klose A, Hilker M, Kobuch R, Rupprecht L, et al. Pumpless extra-
corporeal lung assist: a 10-year institutional experience. Ann Thorac Surg. 2008;86(2):410
7. discussion 417.
91. Kolla S, Awad SS, Rich PB, Schreiner RJ, Hirsch RB, Bartlett RH. Extracorporeal life
support for 100 adult patients with severe respiratory failure. Ann Surg. 1997;226:54464.
discussion 6566.
economic assessment of conventional ventilator support versus extracorporeal membrane
oxygenation for severe adult respiratory failure (CESAR): a multicenter randomized
controlled trial. Lancet. 2009;374:135163.
93. Ketoconazole for early treatment of acute lung injury and acute respiratory distress
syndrome: a randomized controlled trial. The ARDS Network. JAMA. 2000;283(15):
19952002.
94. Eisner MD, Thompson T, Hudson LD, Luce JM, Hayden D, Schoenfeld D, et al. Efcacy of
low tidal volume ventilation in patients with different clinical risk factors for acute lung
injury and the acute respiratory distress syndrome. Am J Respir Crit Care Med. 2001;164(2):
2316.
95. Davies A, Jones D, Balley M, Beca J, Bellomo R, Blackwell N, et al. Extracorporeal mem-
brane oxygenation for 2009 Inuenza A (H1N1) acute respiratory distress syndrome. JAMA.
2009;302:188895.
96. Freed DH, Henzler D, White CW, Fowler R, Zarychanski R, Hutchison J, et al. Extracorporeal
lung support for patients who had severe respiratory failure secondary to inuenza A (H1N1)
2009 infection in Canada. Can J Anaesth. 2010;57:2407.
97. Roch A, Lepaul-Ercole R, Grisoli D, Bessereau J, Brissy O, Castanier M, et al. Extracorporeal
membrane oxygenation for severe inuenza A (H1N1) acute respiratory distress syndrome: a
prospective observational comparative study. Intensive Care Med. 2010;36:1899905.
98. Pappalardo F, Pieri M, Greco T, Patroniti N, Pesenti A, Arcadipane A, et al. Predicting mor-
tality risk in patients undergoing venovenous ECMO for ARDS due to inuenza A (H1N1)
pneumonia: the ECMO net score. Intensive Care Med. 2013;39:27581.
99. Bartlett RH. Esperanza: presidential address. Trans Am Soc Artif Intern Organs. 1985;31(1):
7236.
100. Mortensen JD. Extracorporeal membrane oxygenation for pulmonary assist in patients with
ART. Chest. 1975;67(1):2930.
101. Bartlett RH, Andrews AF, Toomasian JM, Haiduc NJ, Gazzaniga AB. Extracorporeal
membrane oxygenation for newborn respiratory failure: forty-ve cases. Surgery. 1982;92(2):
42533.
102. Ware JH. Investigating therapies of potentially great benet: ECMO. Statist Sci.
1989;4(4):298306.
103. Bartlett RH, Roloff DW, Cornell RG, Andrews AF, Dillon PW, Zwischenberger
JB. Extracorporeal circulation in neonatal respiratory failure: a prospective randomized
study. Pediatrics. 1985;76:47987.
104. ORourke PP, Crone RK, Vacanti JP, Ware JH, Lillehei CW, Parad RB, et al. Extracorporeal
membrane oxygenation and conventional medical therapy in neonates with persistent pulmo-
nary hypertension of the newborn: a prospective randomized study. Pediatrics. 1989;84:
95763.
105. UK Collaborative ECMO Trial Group. UK collaborative randomized trial of neonatal extra-
corporeal membrane oxygenation. Lancet. 1996;348(9020):7582.
106. Bartlett RH, Harken DE. Instrumentation for cardiopulmonary bypasspast, present, and
future. Med Instrum. 1976;10(2):11924.
107. Bartlett RH. Extracorporeal life support: history and new directions. ASAIO J. 2005;
51(5):879.
108. Friedman DF, Montenegro LM. Extracorporeal membrane oxygenation and cardiopulmo-
nary bypass. In: Handbook of pediatric transfusion medicine; 2004. p. 1819.
109. Green TP, Timmons OD, Fackler JC, Moler FW, Thompson AE, Sweeney MF. The impact of
extracorporeal membrane oxygenation on survival in pediatric patients with acute respiratory
failure. Pediatric Critical Care Study Group. Crit Care Med. 1996;24(2):3239.
110. Moler FW, Custer JR, Bartlett RH, Palmisano JM, Akingbola O, Taylor RP, et al.
Extracorporeal life support for severe pediatric respiratory failure: an updated experience
19911993. J Pediatr. 1994;124(6):87580.
111. Kreider M, Hadjiliadis D, Kotloff RM. Candidate selection, timing of listing, and choice of
procedure for lung transplantation. Clin Chest Med. 2011;32(2):199211.
112. Singer JP, Blanc PD, Hoopes C, Golden JA, Koff JL, Leard LE, et al. The impact of pretrans-
plant mechanical ventilation on short- and long-term survival after lung transplantation. Am
J Transplant. 2011;11(10):2197204.
113. Diaz-Guzman E, Davenport DL, Zwischenberger JB, Hoopes CW. Lung function and ECMO
after lung transplantation. Ann Thorac Surg. 2012;94:68694.
114. Cypel M, Keshavjee S. Extracorporeal life support as a bridge to lung transplantation. Clin
Chest Med. 2011;32:24551.
Index
A Arteriovenous carbon dioxide removal

Activated clotting time (ACT), 185, 189 (AVCO2R)
Acute hypercapnic respiratory failure, 36, 95, 96 configuration, pumpless system, 128, 129
Acute myocardial infarction, 75 mathematical simulation, 128
Acute myocarditis, 7677 pH and PaCO2, 128, 129
Acute respiratory distress syndrome (ARDS), ASD. See Atrial septal defect (ASD)
202, 206 Assisted mode, ventilation, 172
CESAR, 246 Asthma
Cilley test, 224 AVCO2R, 128
conventional therapies, 245246 hospital and intensive care unit, 96
ECCO2R, 175, 176 MV, progressive hypercapnia, 98
ECMO support, 224 near-fatal asthma refractory, 99
extracorporeal CO2 removal, 245 respiratory system mechanics, 98, 99
HFOV, 171 severe respiratory acidosis, 98
iNO, 170 treatment, 96
lung function, 224225 Atrial septal defect (ASD), 234, 237
mechanical ventilation AVCO2R. See Arteriovenous carbon dioxide
early ECMO phase, 166 removal (AVCO2R)
lung recruitment and lung rest, 166167 Axial flow pumps, 42
venoarterial (VA) mode, 167168
prone positioning, 169, 170
randomized ECMO trial, 245 B
RNs and RRTs, 245 Barotrauma, 174, 175
treatment protocol, 245 Bernoulli equation, 45
VA-ECMO support, 224225 Blood flow
VV ECMO treatment, 246 for gas transfer
weaning ECLS, 226 Bernoulli equation, 50
Adenosine triphosphate (ATP) molecules, 5 conservation of energy, ECMO circuit,
Aerobic and anaerobic metabolism, 3, 5 5052
Airway pressure release ventilation (APRV), 172 membrane oxygenator, 48
Ambulatory, 214215 oxygenator as resistor, 4950
APRV. See Airway pressure release ventilation Reynolds number, 51
(APRV) laminar flow, resistance to, 4647
ARDS. See Acute respiratory distress Reynolds number, 4748
syndrome (ARDS) turbulent flow, 47

Respiratory Medicine 16, DOI 10.1007/978-1-4939-3005-0
262 Index
Blood pump Cardiopulmonary bypass (CPB)

positive displacement pumps, 41 ASD, 234
roller pumps, 41 autogenous lungs, 237
velocity pumps, 42 canine experimentation models, 237
Bohr effect, 19 congenital heart defects, 234
Bridge to transplant (BTT) cross-circulation method, 237
age and co-morbidities, 109 Gibbons heart-lung machine model,
bridge to recovery, 110 234, 236
chronic bronchiolitis, 108 human donors lungs, 237
mechanical ventilation, 109, 111112 John Gibbon, 234, 236
organ donation, 110 lung transplantation, 113
physical therapy, 215 magnitude, 236
pulmonary fibrosis, 108 pulmonary edema, 237
pulmonary hypertension, 111 timeline, ECMO development, 234, 235
sarcoidosis, 108 Cardiorespiratory function, 12
BTT. See Bridge to transplant (BTT) Cardiotoxic drug intoxication, 77
Cellular metabolism
aerobic and anaerobic metabolism, 3
C glycolysis, 35
Cannulation, vascular access CESAR. See Conventional ventilatory support
bleeding, 144 versus extracorporeal membrane
configurations oxygenation for severe adult
arteriovenous, 142 respiratory failure (CESAR )
low-flow venovenous, 142 Chronic obstructive pulmonary disease (COPD)
transthoracic, 142143 alveolar hyperinflation, 93
venoarterial, 141 dynamic lung hyperinflation, 96
veno-arterio-venous, 142 endotracheal intubation, 95
venovenous, 141 exacerbation, 93
decannulation, 143 femoral artery pseudoaneurysm, 95
extracorporeal support (see Extracorporeal heparin-induced thrombocytopenia, 95
support, cannulation) hypercapnic respiratory failure, 96
inadequate flow, 144145 indications, exclusions and goals, 96, 97
infection, 145 invasive strategies, respiratory support, 94
insertion technique ventilator-induced diaphragmatic
open surgical, 139140 dysfunction, 94
percutaneous, 138139 Circuit air
semi-open, 139 blood and fibrin, 194195
limb ischemia, 144 cardiotomy reservoir, 195
patient preparation cavitation, 194
infection control, 137138 diagnosis and management, 196197
vessel size, 137 pre-oxygenator, 194
recirculation, 143 tubing, 195
vascular injury, 144 Circuit crises
Carbon dioxide partial pressure circuit air, 194197
blood pump, types, 41, 42 heat exchanger, 200
ECCO2R, 3640 inadvertent decannulation, 200201
extracorporeal blood path, 41 oxygenator failure, 199
Carbon dioxide transport pump failure, 199200
arteriovenous, 3435 thrombosis, 197198
carbamino carriage, 3435 tubing rupture/cannula fracture, 200
carbonic anhydrase, 34 Circuits
hydrogen ion concentration, effect, 3234 air entrainment and gas embolism, 148
in solution, 3132 antibiotics, sedatives and analgesics, 149
Carbonic anhydrase, 34 artery and vein, 150
Index 263
blood flow, 151 D

cannulas and tubing, 152 Decannulation
cellular deposition, 149 gas exchange, 229230
centrifugal pump, 150 inadvertent, 200201
complex circuit design, 148 respiratory function, ventilator, 230
compliance chamber, 152 trialing-off, 229
components and connectors, 148 and weaning (see Weaning and
dual-lumen cannulas, 150 liberation, ECMO)
heat exchanger and heater-cooler, Diffusion
158159 of gas molecules into liquid phase
inflow and outflow cannulas, 151 biophysics of membrane oxygenation,
length and complexity, 149 1213
medications and anticoagulant agents, 151 concentration of gases in solutions,
membrane oxygenators, 154157 1011
outflow cannula, 150 Ficks Law of Diffusion, 12
polymethylpentene (PMP) oxygenators, 149 porous media, 13
ports, 151 respiratory gases in solution, solubility
pressure, 151 of, 1112
priming, 149150 transport of gases
pumps, 152154 blood pressure and systemic vascular
renal replacement therapy, 147, 148 resistance, 14
spectrophotometric sensors, 151 effect of turbulence and haematocrit,
surface coatings, 158 1617
ultrasound probes, 151 Ficks Law, 13
Coefficient of ultrafiltration, 1617 Henrys Law, 13
Colour flow Doppler techniques, membrane exposure time, 14
4546 relative flow direction, 15
Complications Double-lumen cannula (DLC), 244, 251
BTT, 109, 110
cannulation, 143145
catastrophic, 215 E
ECCO2R, 89, 9496 ECCO2-R. See Extracorporeal carbon dioxide
ECMO, 80 removal (ECCO2-R)
hemorrhagic, 201 ECLS configuration, VAV. See Veno-arterio-
HFOV, 171 venous (VAV) ECLS
LV distension, 125 ECLS crises
postcardiotomy, 77 bleeding and transfusions
prone positioning, 169170 adverse events, 202
vessel laceration/transection, 137 anticoagulation, 202, 203
Constrained vortex pumps, 42 bronchoscopy, 203
Continuous venovenous hemofiltration diagnosis and management, 204205
(CVVH), 142, 183, 185, 188189 hemoglobin, 202203
Conventional medical therapy (CMT), hemorrhagic complications, 201
235, 249 patient management, 202
Conventional ventilatory support versus thrombocytopenia, 203
extracorporeal membrane thrombosis, 203
oxygenation for severe adult circuit air, 194197
respiratory failure (CESAR ), 65, heat exchanger, 200
66, 202, 212, 246 hemolysis, 203, 205
COPD. See Chronic obstructive pulmonary inadvertent decannulation, 200201
disease (COPD) oxygenator failure, 199
CPB. See Cardiopulmonary bypass (CPB) pump failure, 199200
CVVH. See Continuous venovenous refractory hypoxemia, 206207
hemofiltration (CVVH) shock, 207
264 Index
ECLS crises (cont.) CPB (see Cardiopulmonary bypass (CPB))

thrombosis, 197198 DeWall-Lillehei bubble oxygenator, 238
tubing rupture/cannula fracture, 200 direct-contact extracorporeal
ECLS emergencies oxygenators, 239
circuit-related, 194201 disease, 233
patient-related, 201207 drainage cannula, 181
ECMO. See Extracorporeal membrane ECMO, 233
oxygenation (ECMO) excessive pressure, 183
ECMO configuration. See Types of ECMO feeding and gastrointestinal tract, 188
ECMO hybrids. See Veno-arterio-venous hard-shell device, 239
(VAV) ECLS heart, 188
ECPR. See Extracorporeal cardiopulmonary hematology, anticoagulation and
resuscitation (ECPR) transfusion, 189190
ELSO. See Extracorporeal Life Support hemodilution, 239
Organization (ELSO) hypotension, 184
ELSO registry inhaled prostacyclin/nitric oxide, 184
charter meeting, 252253 inlet pressure, 184
number of active ECLS centers, 252, 253 kidneys, 188189
overall patient outcomes, 254 lung management, 186187
Erythrocyte metabolism, 78 lung transplantation, 105114
EVLP. See Ex-vivo lung perfusion (EVLP) management phase, 185
Extracorporeal carbon dioxide removal Mayo-Gibbon pump oxygenator, 238
(ECCO2-R) metal film oxygenator, 237
acute severe asthma, 9699 morbidity and mortality, 239
and alveolar oxygen concentrations, 3940 oxygenator, 183
ARDS, 88, 245 polymethylpentene membranes, 183
arteriovenous (AV) bypass, 88 pressure-controlled ventilation, 184
asthma, 90 pump, 182
blood flow robust communication, 181
rates, 63 rotating disc oxygenator, 239
requirements, 3637 sedation, 187
clinical indications and exclusion criteria, 90 Sigma motor pump, 238
configurations, 88, 89 single and double-lumen cannulae, 184
COPD, 90 sweep gas flow and FIO2 selector, 183
integration with pulmonary oxygenation, 38 transparent film dressing, 181182
Pubmed, 87, 88 venoarterial (VA), 182
pumpless arteriovenous, 157 veno-arterio-venous (VAV), 185
recirculation, 3738 venous compliance chamber, 182
technological progress and marketing, 88 venovenous (VV), 185
trinsic-PEEP, 90 water heater function, 183
ventilation strategies, 3839 Extracorporeal Life Support Organization
Extracorporeal cardiopulmonary resuscitation (ELSO), 193, 201
(ECPR) Extracorporeal membrane oxygenation (ECMO)
application, 244 acute myocardial infarction, 75
and ECMO, 7576 for acute myocarditis, 7677
VA cardiopulmonary bypass, 244 acute respiratory failure, 163
Extracorporeal gas exchange, 14, 87, 89, 164, advantages, 73
175176 ARDS (see Acute respiratory distress
Extracorporeal life support (ECLS) syndrome (ARDS))
ACT, 185 avoidance of intubation, 173
blood film, 237238 cavitation, 52
cerebral vasoconstriction, 184 complications, 80
clinical improvement, 190 configurations, 2
connector joining, 182 conservation of energy, 5051
Index 265
decannulation, 229230 H
disconnection, 173 Haemoglobin
and drug intoxication, 77 arterial oxygen saturations, 27
and ECPR, 7576 oxygen affinity, 19
flow management, 227 venous oxygen saturation, 28
flow regurgitation, 52 Haldane effect, 19
and heart transplantation, 7778 Heart transplantation, 7778
high airway pressures and high tidal Heat exchanger, 200
volumes, 164 Hemolysis, 202, 205
LVAD implantation, 78 Hemorrhage, 201, 202
and massive pulmonary embolism, 79 Henrys Law, 11
microporous devices, 242 Heparin-induced thrombocytopenia (HIT),
nonmicroporous devices, 242243 158, 189, 196198
occlusion of flow path, 5152 High-frequency oscillatory ventilation
outcomes after VA-ECMO, 8081 (HFOV), 170171
oxygenation, 164165 HIT. See Heparin-induced thrombocytopenia
patient and disease-specific issues (HIT)
configurations and cannulation Hydrogen ion concentration, 3234
strategies, 74 Hypercapnic acidosis, 39
retrieval units, 7475 Hypovolemia, 155
VA-ECMO indication, 74 Hypoxemia, 208209
PECOR, 174176
PGD, 252
postcardiotomy cardiogenic shock, 77 I
and profound hypothermia, 79 IABP. See Intra-aortic balloon pump (IABP)
refractory septic shock, 79 iLA. See Interventional lung assist (iLA)
termination of ECLS, 230231 Inadvertent decannulation, 200201
VA ECMO, 243244 Interventional lung assist (iLA), 107
VV ECMO, 244 Intra-aortic balloon pump (IABP), 125, 128
Extracorporeal support, cannulation
blood flow, 135137
dual-lumen L
bicaval design (Avalon Elite, Maquet), LAS system. See Lung allocation scoring
134, 135 (LAS) system
cavo-atrial design (OriGen, OriGen Left ventricular assist devices (LVAD), 2, 5354
Biomedical), 134 Left ventricular end diastolic pressure (LVEDP),
hemodialysis catheter, 135 5657
percutaneous cannula, 133134 Liberation, 218. See also Weaning and
single lumen design, 134 liberation, ECMO
wire-reinforced cannulas, 134 Limb ischemia, 144
Ex-vivo lung perfusion (EVLP), 113 Lung allocation scoring (LAS) system, 250
Lung transplantation
bleeding, hemolysis and infection, 105
F bridge to decision, 111
Flashing the cannulae, 228 BTT (see Bridge to transplant (BTT))
Frank-Starling mechanism, 56 centrifugal pumps, 106
complications, 110
diseases, 250
G DLC, 251
Glycolysis iLA, 107
ATP molecules, 5 intraoperative VA-ECLS, 112
glycogen storage, 34 intubation, 110
intracellular metabolism, 4 LAS system, 250251
TCA cycle, 5 mechanical ventilation, 106, 111112
266 Index
Lung transplantation (cont.) erythrocyte, 78

mobilization, 110 in stressed state, 67
organ preservation, 113114 Mobilization
oxygen exchange, 106 ambulatory, 214215
physical therapy and potentially barriers, 215216
ambulation, 251252 early, 213214
postoperative ECLS, 112113 health costs of immobility, 212213
post-traumatic ARDS, 106 myopathy and weakness, 213
primary diagnosis, 109 team and methods
pulmonary failure, 111 ambulation, 219
pulmonary hypertension, 111 caregivers, 217
right heart problems, 109 complexity, 217218
right ventricular failure, 111 cycle ergometry, 219
technical complications, 105 physical activity, 218219
technical developments, 106 responsibilities, 218
VA-ECLS, 108 technological advances, 211212
VV-ECLS, 107108 weaning, 214
VV ECMO, 250 Modes of ECLS
LVAD. See Left ventricular assist devices AVCO2R, 128129
(LVAD) VA, 117, 118, 122125
VAV, 126128
VV, 118122
M
Massive pulmonary embolism, 79
Mayo-Gibbon pump oxygenator, 238 N
Mechanical ventilation, ARDS Neonatal acute respiratory distress syndrome
early ECMO phase, 166 arterial pO2, 246, 247
lung recruitment and lung rest, CMT, 249
166167 Esperanzas lung function, 247, 248
venoarterial (VA) mode, 167168 play-the-winner strategy, 248249
Membrane exposure time, 14 PPHN, 248
Membrane oxygenation randomized clinical trial, 248
biophysics, 1213
extra-capillary flow, 9
heat loss, 10 O
hollow fibre oxygenators, 89 Obstructive diseases
Membrane oxygenators acute severe asthma, 9699
blood and gas phases, 154 ARDS, 88
carbon dioxide transfer, 155 COPD, 9396
circuit pressures, 155, 156 dynamic alveolar hyperinflation, 90
delta pressure, 156 exclusions, 99100
extra-capillary blood flows, 154 expiratory flow-limitation, 90
oxygen transfer, 154155 extracorporeal life support strategies, 87
plasma leak, 154, 157158 NIV, 90
PMP and polypropylene hollow-fiber pathophysiological mechanisms, 9092
membranes, 154 respiratory acidosis, 90
PMP oxygenators, 155157 Organ care system (OCS), 113114
polypropylene hollow-fiber, 157 Osmotic diuresis, 7
rated flow, 157 Outcome assessment, after VA-ECMO, 8081
secondary flows, 154 Oxygenator failure, 199
thrombosis, 156 Oxygen carriage
Metabolism cardiac output, arterial saturations, 2325
aerobic and anaerobic, 35, 12 effect of oxygen carrying capacity, 26
cellular, 35 haemoglobin oxygen affinity, 1719
Index 267
mixed-venous saturation, 26, 29 levels, 167

oxygen-haemoglobin dissociation curve, ventilation, 3839, 167, 168
1923 VILI, 169
recirculation, 28 Postcardiotomy cardiogenic shock, 77
venous oxyhaemoglobin saturation, 30 Primary graft dysfunction (PGD), 252
VO2 and cardiac output, 26 Profound hypothermia, 79
VV-ECMO, 2728, 30 Pulmonary hypertension, 39
Oxygen-haemoglobin dissociation curve, 18 Pumps
oxygen partial pressures, 2021 centrifugal, 153154
right-shifted curve, 20 failure, 199200
VV-ECMO and oxygen transport, 2123 roller head, 153
Oxygen transport, 17 velocity (see Velocity pumps)
Oxyhemoglobin saturation, 30
R
P Rapoport-Luebering shunt, 8
Partial extracorporeal CO2 removal (PECOR) Registered nurses (RNs), 245
barotrauma, 174, 175 Registered respiratory therapists
bilateral pleurectomy, 174 (RRTs), 245
clinical application, 174 Rehabilitation, 214, 216
ventilatory control, 174176 Renal replacement therapy (RRT), 151
Passive oxygenation, 39 Respiratory quotient (RQ), 56
Patent ductus arteriosus (PDA), 237 Respiratory system compliance (Crs), 172
Patient crises Reynolds number, 4748
bleeding and transfusions Right ventricular assist devices (RVAD), 2
adverse events, 202 RNs. See Registered nurses (RNs)
anticoagulation, 202, 203 RRT. See Renal replacement therapy
bronchoscopy, 203 (RRT)
diagnosis and management, 204205 RRTs. See Registered respiratory therapists
hemoglobin, 201203 (RRTs)
hemorrhagic complications, 201 RVAD. See Right ventricular assist devices
patient management, 202 (RVAD)
thrombocytopenia, 203
thrombosis, 203
hemolysis, 203, 205 S
refractory hypoxemia, 206207 Septic shock, 79
shock, 207 Shock, 207
Patient-ventilator interaction, 172 Silicone rubber membrane lungs (SRML)
PECOR. See Partial extracorporeal CO2 alveolar membrane, 240
removal (PECOR) extracapillary orientation, 241
Pediatric acute respiratory distress syndrome, gas exchange, 240241
249250 helical tube arrangement, 241
PEEP. See Positive end-expiratory pressure hemodialysis machine design, 240
(PEEP) heparin, 242
Persistent pulmonary hypertension of the hypobaric oxygenation, 241
newborn (PPHN), 248 incorporate hydrophobic materials, 240
Physical therapy, mobilization intracapillary orientation, 241
ambulation, 214, 215 limitations, 239
bridge-to-transplant, 215 O2 and CO2, 240
intubation, 213 SRML. See Silicone rubber membrane lungs
resources, ICU, 217 (SRML)
Positive end-expiratory pressure (PEEP) Systematic review, 99, 150
airway pressures, 166 Systemic inflammatory response syndrome
ECCO2R, 174 (SIRS), 3
268 Index
T ischemia-reperfusion damage, 112113

TCA. See Tricarboxylic acid (TCA) cycle local anesthesia, 108
Termination, ECLS neonates and infants, 122, 123
ARDS, 230 older children and adults, 122, 125
critical care, 230 oxygenation and carbon dioxide
treatment goals, 231 elimination, 122123
withdrawal of life-sustaining device, 231 pulmonary hypertension, 108, 110, 111
Thrombosis, 197198 respiration and circulation, 122
Total parenteral nutrition (TPN), 188 right-sided heart failure, 109
Transplant team. See Lung transplantation stroke, 123, 125
Trialing-off, 225 Veno-arterial ECMO (VA-ECMO)
decannulation, 229 differential cyanosis, 5556
VA-ECMO, 228 flow reversal, 55
VV-ECMO, 228229 valvular incompetence, 56
Tricarboxylic acid (TCA) cycle, 5 Veno-arterio-venous (VAV) ECLS
Turbulence and haematocrit cardiac dysfunction, 128
gas exchange membrane, 16 configuration, 126127
resistances to diffusive transport, 16 IABP, 128
ultrafiltration of plasma water, 1617 partially occlusive clamp, 127
Types of ECMO Venous oxyhaemoglobin saturation, 30
VA ECMO, 243244 Venovenous (VV) ECLS, 107108
VV ECMO, 244 advantages, 117, 118, 121122
dual-lumen cannulation, 119120
inferior (IVC) and superior (SVC) vena
U cavae, 118119
Ultrafiltration rate, 17 physiological implications, 120121
United Network of Organ Sharing (UNOS), 250 Veno-venous ECMO (VV-ECMO)
arterial oxygen saturations, 27
oxygenation, 30
V and oxygen transport
VADs. See Ventricular assist devices (VADs) access configurations, 22
VA ECLS. See Venoarterial (VA) ECLS anaerobic metabolism, 23
Vascular access. See Cannulation, vascular arterial (SaO2) and venous oxygen
access saturations (SvO2), 2325
VAV ECLS. See Veno-arterio-venous cardiac output, 2122
(VAV) ECLS ECMO circuit, 22
Velocity pumps mixing blood streams, 23
systemic circulation oxyhaemoglobin dissociation curve, 23
thrombus formation, 58 venous haemoglobin oxygen
VADs, 5354 saturation, 23
VA-ECMO, 5556 physiology, 30
ventricular preload vs. device preload, recirculation, 28
5658 venous saturations, 27, 28
types Ventilator-induced lung injury (VILI)
axial flow pumps, 42 ARDS, 175
centrifugal constrained vortex pumps, 42 description, 164
monitoring pump output, 42 prevention, 166
Venoarterial (VA) ECLS prone positioning, 169, 170
advantages, 117, 118, 125 Ventilator management
antegrade perfusion catheter, 122, 125 ARDS (see Acute respiratory distress
blood pressures, 125 syndrome (ARDS))
circulatory effect, 123 assisted mode, 172
IABP, 125 avoidance of intubation, ECMO, 173
intraoperative, 112 disconnection, 173174
Index 269
gas exchange Viscoelasticity

carbon dioxide (CO2), 165 fluid and conduit, 4446
oxygenation, 164165 relative viscosity of blood, 4344
PECOR, 174176 viscous friction, 44
rescue therapies, hypoxemia vortex centrifugal pumps, 43
HFOV, 171 VSD. See Ventricular septal defect (VSD)
inhaled NO (iNO), 170 VV ECLS. See Venovenous (VV) ECLS
pneumothoraces, 171
prone positioning, 168170
Ventricular assist devices (VADs), 5354 W
Ventricular preload vs. device preload Weaning and liberation, ECMO, 94, 96, 110,
Frank-Starling mechanism, 56 172, 214
inflow cannula, 57 anticoagulation management, 228
LVEDP, 5657 cardiopulmonary assessment, 228229
mitral inflow restriction, 58 ECMO flow management, 227228
right ventricular function and pulmonary oxygenator sweep gas flow
haemodynamics, 58 management, 227
Ventricular septal defect (VSD), 237 in ventilated ARDS patient, 226
VILI. See Ventilator-induced lung injury (VILI) ventilator management, 225227

@anesthesia Books 2016 Extracorporeal

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

@anesthesia Books 2016 Extracorporeal

Загружено:

Авторское право:

Доступные форматы

Respiratory Medicine

Series Editor: Sharon I.S. Rounds

Gregory A. Schmidt Editor

More information about this series at http://www.springer.com/series/7665

Extracorporeal Life Support

ISSN 2197-7372 ISSN 2197-7380 (electronic)

Library of Congress Control Number: 2015950466

Springer New York Heidelberg Dordrecht London

Printed on acid-free paper

Humana Press is a brand of Springer

Extracorporeal life support (ECLS) consists of using an external gas-exchanging

Iowa City, IA, USA Gregory A. Schmidt, MD

1 Physiology of Extracorporeal Life Support (ECLS) ........................... 1

12 Mobilization During ECLS .................................................................... 211

Index ................................................................................................................. 261

Alain Combes, MD, PhD Service de Ranimation Mdicale, Institut de

V. Marco Ranieri, MD Dipartimento di Anestesiologia e Medicina degli Stati

MatthewJ.Brain, WarwickW.Butt, andGraemeMacLaren

Extracorporeal life support (ECLS) and related implantable circulatory assistance

M.J. Brain, MBBS (Hons), FRACP, FCICM, DDU (*)

Springer Science+Business Media New York 2016 1

systemic blood flow is augmented by the extracorporeal blood pump, while

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

Glycolysis andAerobic andAnaerobic Metabolism

Carbon Dioxide Production andtheRespiratory Quotient

without increasing mitochondrial NADH.As the rate of oxygen consumption

Lipogenesis* : 13.83C6 H12 O6 + 5O2 C55 H104 O6 + 28CO2 + 31H 2 O RQ = 5.6

Metabolism intheStressed State

The RapoportLuebering shunt (Fig.1.2) describes the pathway for 2,3-DPG

Biophysics ofMembrane Gas Exchange

Membrane Oxygenator Construction

Extracorporeal membrane oxygenators consist of a high surface area blood path

Diffusion ofGas Molecules into aLiquid Phase

Concentration ofGases inSolutions

The Solubility ofRespiratory Gases inSolution

The Bunsen solubility coefficient of oxygen is 0.003082mLdL1mmHg1 and

[O2 ] = 1.38 10-3 90 mmHg

Since the partial pressure of a gas is proportional to the concentration in solution

Biophysics ofMembrane Oxygenation

orders of magnitude greater than coefficients in liquids. Low-molecular-weight

The Driving Force forDiffusive Transport ofGases

After a brief period of operation any extracorporeal gas exchange membrane

Membrane Exposure Time

Partial Pressure of Gas Phase

Partial Pressure Liquid Phase

Relative Flow Direction

 ffect ofTurbulence andHaematocrit onLocal

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

Resistances toDiffusive Transport

Area oftheGas Exchange Membrane

Ultrafiltration ofPlasma Water Over theOxygenator Membrane [15]

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

As outlined above, oxygen has a limited solubility in plasma of 1.39103mmolmmHg1

Oxygen carriage in blood is massively increased by the presence of haemoglobin, a

Haemoglobin is a spherical molecule consisting of four globin subunits (two

Oxygen Hemoglobin Dissociation and Oxygen Content

SaO2 at pH 7.44 Temp 37.5 & BE 0

DO2 = Q B CaO2 10 (1.16)

Modulation ofHaemoglobins Affinity forOxygen

The OxygenHaemoglobin Dissociation Curve

The oxygenhaemoglobin dissociation curve is characterised by an upper plateau at

Mixing Blood ofDiffering Oxygen Partial Pressures

The oxygenhaemoglobin dissociation curve is of importance when considering

Glycolysis andAerobic andAnaerobic Metabolism

Carbon Dioxide Production andtheRespiratory Quotient

Metabolism intheStressed State

Biophysics ofMembrane Gas Exchange

Membrane Oxygenator Construction

Diffusion ofGas Molecules into aLiquid Phase

Concentration ofGases inSolutions

The Solubility ofRespiratory Gases inSolution

Biophysics ofMembrane Oxygenation

The Driving Force forDiffusive Transport ofGases

Membrane Exposure Time

Relative Flow Direction

ffect ofTurbulence andHaematocrit onLocal

Resistances toDiffusive Transport

Area oftheGas Exchange Membrane

Ultrafiltration ofPlasma Water Over theOxygenator Membrane [15]

Modulation ofHaemoglobins Affinity forOxygen

The OxygenHaemoglobin Dissociation Curve

Mixing Blood ofDiffering Oxygen Partial Pressures

Veno-venous ECMO andOxygen Transport

rterial Saturations Are Dependent ontheFraction

ixed Venous Saturations Are Determined by VO2

The Effect ofOxygen Carrying Capacity

Further Consideration ofRecirculation

The Importance ofMixed-Venous Saturation

Venous Oxyhaemoglobin Saturation

Limitations Imposed onVV-ECMO Oxygenation

Carbon Dioxide Physiology

Carbon Dioxide Transport inPhysical Solution

Reactions ofCarbon Dioxide inSolution

Effect oftheHydrogen Ion Concentration

The Arteriovenous-CO2 Difference andCarbamino Carriage

Determinants ofCarbon Dioxide Partial Pressure

Extracorporeal Carbon Dioxide Removal: ECCO2R

Blood Flow Requirements

Recirculation withSmaller Access Catheters

Oxygenation inECCO2R: Integration withPulmonary Oxygenation

Ventilation Strategies withECCO2R

ECCO2R andAlveolar Oxygen Concentrations

Biophysics oftheExtracorporeal Blood Path

Types ofBlood Pump

Positive Displacement Pumps

Physical Properties ofBlood Relevant toFlow

Flow ofIdeal Fluids inSolid Conduits

Resistance toLaminar Flow

Optimising Blood Flow forGas Transfer

The Oxygenator asaResistor

Conservation ofEnergy Across theECLS Circuit