nurtured an outstanding program, exhibited remarkable vision in how to advance ECLS care, and opened my eyes to its new possibilities.
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 respiratory 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 transplant. 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 efﬁcient 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 decannulation process; and recommendations for crisis management and rehabilitation related to ECLS. The text concludes with a fascinating historical review, showing just how far we’ve 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
from my colleagues at Iowa who strive daily to save the sickest patients; the trainees whose curiosity makes us all want to know more; my contributors who are at the forefront of a truly challenging ﬁeld; and our publisher at Springer-Link who pushed for this important 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 beneﬁt of all our patients. Iowa City, IA, USA
Gregory A. Schmidt, MD
Physiology of Extracorporeal Life Support (ECLS) ........................... Matthew J. Brain, Warwick W. Butt, and Graeme MacLaren
Hypoxemic Respiratory Failure: Evidence, Indications, and Exclusions ......................................................................................... Darryl Abrams, Matthew Bacchetta, and Daniel Brodie
Cardiogenic Shock: Evidence, Indications, and Exclusions................ Nicolas Bréchot and Alain Combes
ECCO2R in Obstructive Diseases: Evidence, Indications, and Exclusions ......................................................................................... Lorenzo Del Sorbo and V. Marco Ranieri
ECLS as a Bridge to Lung Transplantation ......................................... 105 Christian Kuehn
Modes of ECLS ....................................................................................... 117 L. Keith Scott and Benjamin Schmidt
Vascular Access for ECLS ...................................................................... 133 Steven A. Conrad
Circuits, Membranes, and Pumps ......................................................... 147 Bradley H. Rosen
Ventilator Management During ECLS ................................................. 163 Antonio Pesenti, Giacomo Bellani, Giacomo Grasselli, and Tommaso Mauri
Daily Care on ECLS ............................................................................... 181 Giles J. Peek
Crises During ECLS ............................................................................... 193 Cara L. Agerstrand, Linda B. Mongero, Darryl Abrams, Matthew Bacchetta, and Daniel Brodie ix
Mobilization During ECLS .................................................................... 211 Gregory A. Schmidt
ECMO Weaning and Decannulation ..................................................... 223 Sundar Krishnan and Gregory A. Schmidt
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
Darryl Abrams, MD Division of Pulmonary, Allergy and Critical Care, New YorkPresbyterian 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 YorkPresbyterian Hospital/Columbia University Medical Center, New York, NY, USA 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 MilanoBicocca 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 Bréchot, MD, PhD Service de Réanimation Médicale, Hospital Pitié–Salpêtrière, Paris, France Daniel Brodie, MD Division of Pulmonary, Allergy and Critical Care, New YorkPresbyterian Hospital/Columbia University Medical Center, New York, NY, USA Warwick W. Butt, FRACP, FCICM ICU RCH, Department of Paediatrics UoM, Clinical Sciences Theme MCRI, Royal Children’s Hospital, Melbourne, VIC, Australia Paediatric Intensive Care Unit, Parkville, VIC, Australia
Alain Combes, MD, PhD Service de Réanimation Médicale, Institut de Cardiologie, Groupe Hospitalier Pitié-Salpêtrière, 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 Città della Salute e della Scienza di Torino, Università 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 Children’s 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, Glenﬁeld Hospital, Leicester, UK Antonio Pesenti, MD Department of Health Sciences, University of MilanoBicocca, Monza, Italy Department of Anesthesia and Critical Care, San Gerardo Hospital and MilanoBicocca University, Monza, Italy Robert Pollock, BS Division of Cardiothoracic Surgery, Department of Surgery, University of Kentucky College of Medicine, Lexington, KY, USA
V. Marco Ranieri, MD Dipartimento di Anestesiologia e Medicina degli Stati Critici, Ospedale S. Giovanni Battista-Molinette, Università di Torino, Torino, Italy Dipartimento di Anestesiologia e Rianimazione, Azienda Ospedaliera Città della Salute e della Scienza di Torino, Università 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
Physiology of Extracorporeal Life Support (ECLS) Matthew J. Brain, Warwick W. Butt, and Graeme MacLaren
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 thoracotomy. Rotary pumps (without oxygenators) have been miniaturised, allowing development of left and right ventricular assist devices (LVAD and RVAD) respectively. A basic ECMO circuit consists of a blood pump and oxygenator connected by conduits (Fig. 1.1). Other components may be added to this basic configuration, in particular 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  but are not currently in clinical use
1 Physiology of Extracorporeal Life Support (ECLS)
required to prescribe, manage and wean this support and recognise evolving complications of the therapy. Although designed primarily to replace cardiorespiratory 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 progressive 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 metabolism, primarily CO2, or lactate in the case of anaerobic metabolism. These functions are interlinked as the waste products of energy production are generally weak acids and influence local perfusion and oxygen carriage. The quantities of substrate required per unit time will depend on the supported cell mass and its level of metabolic activity as influenced by demand (or stress), temperature, inflammation and hormonal regulation.
Glycolysis and Aerobic and Anaerobic Metabolism Glucose and other simple carbohydrates enter cells down a concentration gradient through glucose transporters that allow for tissue-specific behaviour such as preferential basal uptake by the brain, concentration-dependent uptake by the liver, concentration- sensing by the insulin-secreting pancreatic β-cells and insulin- dependent uptake in skeletal muscle and fat . 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 phosphorylated (-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 mitochondria and depend on oxygen to restore nicotinamide adenine dinucleotide to its oxidised form (NAD+) for continued cycling. The number of ATP generated depends on the source of reduction power; a single mitochondrial NADH produces 2.5 ATP; however, electrons from cytosolic NADH must be transferred to mitochondrial FADH2 which yields only 1.5 ATP each . 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 oxaloacetate 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
1 Physiology of Extracorporeal Life Support (ECLS)
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 continually 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 electron acceptor is required for efficient restoration of NAD+ so that the cycle can continue. This electron acceptor is oxygen. Oxidative phosphorylation describes the process of restoring NAD+ to perpetuate the TCA cycle. Although oxygen is utilised as an electron acceptor in many enzyme systems, its highest consumption is in this process. Oxidative phosphorylation occurs in the inner mitochondrial matrix and it is to this intracellular destination that oxygen must diffuse in sufficient quantities to sustain ATP generation for 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 and the Respiratory 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 ). RQ =
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)
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
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 palmitoylstearoyl-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 carbohydrate loads are excessive (such as with glucose supplementation exceeding 4 mg·kg−1·min−1 [5.8 g·kg−1·day−1]), lipogenesis occurs with a respiratory quotient as high as 8 [5–7] resulting in a high CO2 burden.
Metabolism in the Stressed 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 α-cells in response to low blood glucose levels and promotes glycogen breakdown and conversion of amino acids
1 Physiology of Extracorporeal Life Support (ECLS)
(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 glycogenolysis 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 mass—primarily amino acids from muscle catabolism —and the immunosuppressive effects of hyperglycaemia and persistently elevated humoral mediators, including catecholamines 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 .
M.J. Brain et al.
The Rapoport–Luebering 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 . 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 , result in decreased intracellular 2,3-DPG concentrations. This is most relevant under conditions of red cell storage where lower glycolysis rates and accumulation of lactic acid can result in minimal 2,3-DPG concentrations at the time of transfusion. Transfused red cells do not restore normal 2,3-DPG concentrations for some time and, given the relatively high transfusion requirements of patients receiving ECLS, this effect may have significant implications for oxygen carriage. The role of 2,3-DPG will be further discussed below when considering oxygen carriage.
Biophysics of Membrane Gas Exchange Mitochondria can couple ATP production to NADH oxidation only if sufficient oxygen 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 introduction is important to understand their operation. Membranes may be arranged in folded sheets or, more commonly, as tubes known as hollow fibre oxygenators (Fig. 1.3). The pores of earlier polypropylene
1 Physiology of Extracorporeal Life Support (ECLS)
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 ε being the porosity—the area of membrane occupied by gas, τ the tortuosity, an index of effective path length for gas to traverse the membrane (a path length is shown but in reality will be unknown), and δ the constrictivity—the 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 . 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 .
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. Figure 1.3 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 of Gas Molecules into a Liquid Phase Concentration of Gases in Solutions Unlike most solutes dissolved in body fluids that are quantified in moles, gas concentrations 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
The universal gas equation: P = the partial pressure, n = number of molecules of gas measured in moles, T is temperature in degrees Kelvin and V is volume of the container in litres. R is the ideal gas constant which in SI units is 8.314 J·K−1·mol−1 or in conventional units: 62.36 mmHg·K−1·mol−1 At a constant temperature, Eq. (1.2) simplifies to P µ n / V . Concentration is defined as moles/unit volume, i.e. n/V; hence pressure is proportional to gas concentration, 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 : O2 ( gas )
K Forward K Reverse
O2 ( dissolved )
[O2 ](gas) ´ K Forward = [O2 ](dissolved ) ´ K Reverse
[O2 ](dissolved ) = SC ´ [O2 ](gas)
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 concentration 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 Henry’s Law. SC is the Bunsen solubility coefficient where SC = KForward/KReverse and is gas- and solvent-specific. SC is affected by other dissolved
1 Physiology of Extracorporeal Life Support (ECLS)
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: 0 °C, 760 mmHg, dry gas), it is customary to report oxygen content in mL·dL−1. Under these conditions, oxygen approximates an ideal gas such that 6.02 × 1023 gas molecules (i.e. 1 mol) occupy 22.414 L at 0 °C. Quantification of human oxygen consumption is performed using STPD rather than BTPS (body temperature and pressure, saturated: Eq. (1.7))  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 relationship between the number of molecules and volume defined in Eq. (1.2).
The Solubility of Respiratory Gases in Solution The Bunsen solubility coefficient of oxygen is 0.003082 mL·dL−1·mmHg−1 and describes the measured solubility corrected to STPD. Utilising this conversion, the solubility coefficient of oxygen in normal plasma is 1.38 × 10−3 mmol·L−1·mmHg−1 while carbon dioxide is nearly 22 times more soluble at 3.08 × 10−2 mmol·L−1·mmHg−1 . Thus, using Henry’s Law (Eq. 1.5) in normal arterial blood the concentration of dissolved carbon dioxide is nearly ten times that of dissolved oxygen:
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 37 °C air that is fully saturated with water vapour at 1 atm. (i.e. BTPS) is: PIO2 = ( 760 mmHg - 47 mmHg ) ´ FiO2
= 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 equivalent summation, i.e. the summation of the partial pressures in solution will not equal atmospheric pressure.