DEFINITION OF EXTRACORPOREAL LIFE SUPPORT
Extracorporeal life support (ECLS), also known as extracorporeal membrane oxygenation (ECMO), is an artificial form of cardiopulmonary support that allows the heart, the lungs, or both to recover from severe, but potentially reversible, pathologies. ECLS can also function in some cases as a bridge to therapies such as a ventricular assist device or heart or lung transplantation.
BACKGROUND OF EXTRACORPOREAL LIFE SUPPORT
Early work in extracorporeal support dates to 1930 when Dr. John Gibbons built a roller pump device. In 1953, Dr. Gibbons created and successfully used the first heart–lung machine during an atrial septal defect repair. Four years later, silicone rubber membranes replaced the bubbler oxygenator, allowing the prolonged use of extracorporeal machines. These membranes serve as a gas–oxygen interface and thereby prevent severe hemolysis and plasma leakage.1,2
In 1972, Dr. JD Hill announced the first extended use of the extracorporeal circuit outside the operating theater. Dr. Hill’s patient survived posttraumatic respiratory failure after 75 hours of ECLS support. At the same time, Dr. Robert Bartlett and his colleagues at the University of Michigan took the lead in developing and implementing ECLS care, the results of which influenced ECLS care throughout the world.3
After the initial successful attempts, ECLS continued to make progress, albeit slowly. The medical community developed skepticism about the utility of ECLS in adults in the late 1970s when the first randomized controlled trial (RCT) reported poor survival rates.4 Another RCT published in 1994 likewise revealed no survival benefit with ECLS.5 ECLS regained interest in adult patients in 2009 after the publication of several landmark studies, specifically the efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR) trial in the United Kingdom of patients with acute respiratory distress syndrome (ARDS) and observational studies from several different countries of the use of ECLS in respiratory failure caused by H1N1 influenza.6-8 Although both the CESAR trial and the observation influenza pandemic studies had design limitations, these reports stimulated the growth of adult ECLS based on an apparent mortality benefit compared with contemporary conventional mechanical ventilation support.
In an effort to foster and organize ECLS, in 1989, the Extracorporeal Life Support Organization (ELSO) was established to support healthcare professionals and scientists who are involved in ECMO. Among its many activities, ELSO maintains a registry of both facilities trained to provide ECLS care and of patients placed on ECLS. The ELSO patient registry information is used to support clinical research, quality improvement, and regulatory action. ELSO also provides educational programs for ECLS centers and facilities that may be involved in the transfer of patients to higher levels of care.
MODES OF EXTRACORPOREAL LIFE SUPPORT
There are traditionally two modes of ECLS, venovenous (VV ECLS) and venoarterial (VA ECLS). In both modes, blood is drained from a large vein into the ECLS circuit, where gas exchange occurs, and then blood is returned to the venous side (in VV ECLS) or the arterial side (in VA ECLS).
Because ECLS is not a disease-targeted therapy, the selection criteria of the right patient and mode are fundamental to the best outcome. VV ECLS acts as an artificial lung and is connected in series with the cardiopulmonary circulation, allowing complete or partial replacement of the native lung function. It maintains pulmonary blood flow and uses the patient’s own cardiac output (CO). Thus, VV ECLS does not provide hemodynamic support. Conversely, VA ECLS is connected in parallel with the heart and lungs, bypasses the pulmonary circulation, and provides nearly complete hemodynamic support (Fig. 33-1).
Different ECLS configurations with (A) VV-ECLS. Oxygenated blood is returned to the right internal jugular vein while de-oxygenated blood is right femoral vein, with catheter tip ideally near right perihepatic inferior vena cava (IVC). (B) Femoral VA-ECLS. Deoxygenated blood is drained from right femoral vein with catheter tip ideally near right perihepatic inferior vena cava and oxygenated blood returned to right femoral artery. (C) Carotid VA-ECLS. (D) Thoracic VA-ECLS. (Reprinted from Gaffney AM, Wildhirst SM, Griffin MJ, et al. Extracorporeal life support. BMJ. November 2, 2010;341:c5317 with permission from BMJ Publishing Group Ltd.)
INITIATION OF EXTRACORPOREAL LIFE SUPPORT
The ECLS circuit is composed of three essential parts: a gas exchange device, a blood pump, and a heat exchanger device. All parts are connected by tubing along with the intravascular cannula(s) on both ends.
The traditional ECLS oxygenator consists of silicone rubber membranes. They are composed of long spirals that separate the gas phase from the blood phase. The membranes inherently create high resistance to blood flow. Newer hollow, nonporous polymethylpentene (PMP) membranes were developed to mitigate the complications of the high-pressure gradient between the inlet and output circuits. The PMP-based oxygenator device comes in different sizes and shapes and is currently the most widely used oxygenator design.9
Pumps drive blood into the ECLS circuit by one of two means:
Classic roller pumps, in which blood flows passively using gravity siphon
Centrifugal pumps, which are more commonly used and in which blood flows actively through centrifugal force
The heat exchanger consists of hollow, silicone-coated stainless steel tubes that are surrounded by warm water and that function to maintain body temperature as blood flows within the circuit. Heat exchangers are usually located within or after the oxygenator.10
The standard ECLS circuit tubing is made from polyvinylchloride (PVC), a commonly used synthetic polymer. This design tolerates the high pressure generated by the pump and thus prevents circuit rupture. Although there is no ideal biocompatible tubing material, the PVC circuit coated with heparin helps attenuate plasma protein adsorption and reduces the risk of thrombosis.11
There are several kinds of ECLS cannulas. They differ in size, length, lumen(s), and fenestration. Selection of proper cannula depends on the ECLS support mode and size. As a consensus, the cannula size is reported based on the outer diameter. To achieve adequate blood drainage and infusion, it is important to understand the determinants of volume flow, namely the cannula’s length and diameter. Short and wide cannulas are ideal for venous drainage because of reduced resistance.12,13
Insertion sites vary according to the cannula type and designated mode of ECLS. Cannula(s) are usually placed under vascular ultrasound, transesophageal echocardiogram (TEE), or fluoroscopy guidance.
In VV ECLS, if two cannulas are used, a typical arrangement is a single-lumen drainage catheter in the common femoral vein with blood return via an infusion cannula in the right internal jugular (RIJ) or femoral vein. Alternatively, if a double-lumen single cannula is used, the RIJ vein can be cannulated (Figs. 33-2 and 33-3). In this configuration, blood is drained from the superior vena cava (SVC) and inferior vena cava (IVC) and returned through the same cannula into the right atrium.
In VA ECLS, venous blood is typically drained from the IVC and infused back into a large artery, commonly the femoral artery. In postcardiotomy operations, the drainage cannula is occasionally placed into the right atrium or SVC, and the infusing cannula is placed in the aorta.
Extracorporeal life support is considered for cardiogenic shock or respiratory failure cases that are severe; refractory to conventional management; and most important, potentially reversible (Table 33-1).14
There are no clearly established absolute contraindications, but the following preexisting conditions are contraindicated by most protocols:
Other relative contraindications:
The cardiopulmonary system is designed to deliver oxygenated blood to the tissues and eliminate carbon dioxide. Therefore, effective management of ECLS requires a robust understanding of this physiology. Oxygen delivery (DO2) can be measured by multiplying the arterial oxygen content (Cao2) by the CO:
Cao2 represents the sum of O2 bound to hemoglobin (Hgb) and is dissolved in plasma. The normal Cao2 is 20 mL/dL. As seen in the formula, CO plays a major role in determining Do2.
Oxygen consumption (Vo2) is the difference between arterial and venous oxygen content, which can be calculated by Fick’s principle:
where Cvo2 (mL/dL) = 1.39 (mL/dL) × Hgb (g) × % Svo2 + 0.003 (mL/dL/mmHg) × Pvo2
Svo2 is the mixed venous oxygen saturation, and Pvo2 is the venous partial pressure of O2, and Hgb is hemoglobin.
Oxygen delivery in ECLS depends on the multiple factors, including flow through the ECLS circuit, the oxygenator membrane surface area, the contact time of the gas–blood interface and O2 gradient, oxygenation by the native lung, and native CO. Of these, flow is the most crucial determinant of blood oxygenation during ECLS. In particular, oxygenation is directly proportional to the circuit blood flow, measured in milliliters per kilogram per minute. In the adult population, a flow of 50 to 80 mL/kg/min is usually sufficient to achieve adequate tissue oxygenation. A typical ECLS circuit can achieve flows of up to 7 to 8 L/min, depending on cannula size. In general, larger flows are not necessary unless the patient has a very high CO (as seen in septic shock) or increased oxygen consumption (eg, fever). Oxygen delivery is also dependent on hemoglobin concentration, and most ECLS centers aim to keep Hgb at 7 to 8 g/dL or greater.
Extracorporeal life support flow is titrated to maintain an arterial oxyhemoglobin saturation of greater than 90% for VA ECLS or greater than 80% to 85% for VV ECLS.
In VV ECLS, the infused blood travels via the right atrium and right ventricle (RV) into the pulmonary circulation. The infused highly oxygenated blood from the ECLS circuit is diluted by the deoxygenated blood returning to the right side of the heart; the ratio of circuit flow to native CO is therefore an important determinant of systemic oxygenation. Because of this ever-present mixing of oxygenated and deoxygenated blood in the RV (shunt physiology), oxygen saturations are typically low in VV ECLS (goal > 80%). Assuming a lack of gas exchange across the native lung, systemic Pao2 is equal to mixed venous Pao2 in the pulmonary artery (PA). A return of native lung gas exchange is indicated by widened gradient between the arterial and central venous saturations. Because blood is both drained and infused on the right side of the heart, there is no hemodynamic effect of VV ECLS.
The circuit augments both cardiac and pulmonary function in VA ECLS. The highly oxygenated infused blood returns to the systemic circulation, bypassing the pulmonary circulation. Depending on the nature of the disease process requiring VA ECLS, there may remain some native left ventricle (LV) CO. If lung function is normal, the oxygen content of blood leaving the LV is sufficient. However, if there is lung disease, blood leaving the LV will be relatively deoxygenated compared with blood infusing via the ECLS circuit. The location in the aorta where these two circulations meet (ie, the mixing point or watershed) varies among patients, depending on LV function and ECLS circuit flow. For this reason, measuring oxygenation in the right hand is typically performed (when the return cannula is in the femoral artery) as a surrogate to the oxygenation of the heart, coronary arteries, and brain given that the brachiocephalic and right subclavian arteries are closest to the heart. If the right hand displays insufficient oxygenation in the setting of lung disease, then the mixing point is distal to the brachiocephalic artery, and these vital organs (heart and brain) may be receiving insufficient oxygenation. The location of the mixing point will differ if the return cannula is not in the femoral artery; the concept of the mixing point should be considered in every patient receiving VA ECLS, depending on cannula location, lung function, and native CO. The adequacy of VA ECLS flow can be measured via markers of systemic oxygenation (lactic acid) or mixed venous saturation.17