Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Extracorporeal membrane oxygenation (venovenous and venoarterial ECMO)
Introduction
Use of extracorporeal life support (ECLS) for respiratory and/or cardiac failure continues to increase, with almost 500 Extracorporeal Life Support Organization (ELSO) centers worldwide and almost 100,000 patients now added to the ELSO registry. In actuality, ECLS use is likely much higher, as many centers do not report data to ELSO or any other international registry. The survival rate for ECLS varies significantly by indication for support and age. Table 40.1 lists the survival rates reported to ELSO for neonates, pediatric, and adult patients. ECLS complications continue to be common and are most frequently the result of bleeding or clotting. As ECLS technology improves and global experience increases, the indications for ECLS continue to expand.
TABLE 40.1
Survival to Decannulation and Discharge or Transfer for Neonatal, Pediatric, and Adult Extracorporeal Membrane Oxygenation (ECMO)
ECPR , Extracorporeal pulmonary resuscitation. Data from the ELSO International Registry Report, July 2020.
| Total Runs | Survival to Decannulation, N (%) | Survival to Hospital Discharge or Transfer, N (%) | |
|---|---|---|---|
| NEONATAL | |||
| Respiratory | 32,634 | 28,627 (87) | 23,860 (73) |
| Cardiac | 8993 | 6216 (69) | 3899 (43) |
| ECPR | 2080 | 1463 (70) | 883 (42) |
| PEDIATRIC | |||
| Respiratory | 10,549 | 7636 (72) | 6347 (60) |
| Cardiac | 12,836 | 9271 (72) | 6854 (53) |
| ECPR | 5086 | 3032 (59) | 2159 (42) |
| ADULT | |||
| Respiratory | 25,631 | 17,832 (69) | 15,741 (60) |
| Cardiac | 27,004 | 16,117 (59) | 11,891 (44) |
| ECPR | 8558 | 3582 (41) | 2549 (29) |
Modes of ECMO
Extracorporeal membrane oxygenation (ECMO) modes are conventionally named for the sites from which the blood is drained and returned. Fig. 40.1 depicts typical cannula placement for neonates, pediatric, and adult patients. There are two major technical types of ECLS: venoarterial (VA) and venovenous (VV). Hybrid modes that incorporate VA and VV together are also used in specific circumstances.
Venoarterial cannulation
In the VA mode, deoxygenated blood is removed from the venous system and pumped to a membrane lung, which removes carbon dioxide and adds oxygen. Oxygenated blood is then returned to the patient via the arterial system. Use of the venous and arterial vessels depends on the patient size and condition and is divided into peripheral and central cannulation.
Venous drainage in older children and adults is either from the internal jugular or femoral vein. As femoral vessel size limits the adequacy of venous and arterial cannulation in neonates and small children, usually up to the age of 2 years, the right or left internal jugular vein is most commonly accessed. Oxygenated blood is usually returned to the patient via the carotid or femoral artery. The femoral artery is the preferred method in adolescents and adults to prevent need to repair or sacrifice the carotid artery at the time of decannulation. An increase in adverse neurologic events resulting from restricted cerebral blood flow secondary to loss of the right carotid artery and from emboli returning from the ECMO circuit to the left carotid are also reasons why carotid access is often avoided. Cerebral venous congestion from internal jugular cannulas is also theorized to add to the risk of neurologic events, although there are no definitive studies that have adequately researched this. Some neonatal centers advocate placement of a distal jugular bulb drainage cannula (also known as a cephalad cannula ), which is then Y’ed into the ECMO venous drainage line to increase venous flow and decrease intracranial venous pressure. This technique, however, is rarely described in adult care. If the femoral artery is used, a reperfusion cannula may need to be employed to provide adequate perfusion distal to the cannula site and thereby mitigate limb ischemia. Continuous noninvasive monitoring by pulse oximetry or near-infrared spectroscopy on lower extremities during femoral VA access may help with earlier identification of limb ischemia if changes in blood flow and oxygenation develop post-cannulation. A side-graft on the femoral artery can also be placed that does not totally occlude distal blood flow, but this process cannot be accomplished via a percutaneous approach and requires surgical intervention. Other arterial access possibilities include cannulation of the subclavian and innominate vessels, facilitated by advances in cannula technology and growing experience. If these arteries are cannulated, care must be taken to avoid either hyperemia of the distal limb (which may require a ligature to restrict distal blood flow) or ischemia that develops after occlusion of distal flow. Side grafts have also been used in subclavian arterial access.
Hypoxemia differential
In patients with severe respiratory failure who are cannulated via the femoral artery, development of differential oxygen saturation of the upper and lower extremities can occur because of competing native cardiac output and returning ECMO circuit flow directed retrograde up the aorta. Where the native cardiac output, which is likely desaturated in patients with severe pulmonary gas exchange problems, meets the oxygenated ECMO return flow is often termed the “mixing zone.” Higher native cardiac output and lower ECMO return will move the mixing zone more distal in the aorta, whereas increased ECMO flow may push oxygenated blood farther up the aorta. If desaturated blood from the left heart is primarily perfusing the upper body, including the brain and coronaries, this results in a cyanotic upper body with a well-perfused lower body, a condition for which the terms “harlequin” or “north-south” syndrome is used. Methods to improve upper body oxygenation include placing another venous cannula into the right atrium (RA) which is Y’ed into the arterial ECMO return to direct some oxygenated blood to the RA and improve cardiac oxygen delivery. Care must be taken to ensure that both ECMO return cannulas—especially the femoral cannula—achieve adequate flow to prevent stasis and thrombosis.
For patients cannulated via the femoral artery, use of pulse oximetry or arterial blood gas data from the right hand or use of cerebral near-infrared spectroscopy (NIRS) may help follow oxygen saturation levels. Although there is controversy as to when adequate oxygen is reaching the upper body and brain, most centers use as guidelines the achieving of arterial saturations >80% (sometimes lower) or, conversely, avoidance of signs of inadequate oxygenation, such as mental confusion, electrocardiogram (ECG) changes of ischemia, or high blood lactate levels. These help to determine whether the cannulation strategy needs to be altered to improve upper body oxygenation. If unable to adequately provide respiratory and cardiac support with this configuration, movement of the femoral artery cannula to an upper body site or central cannulation should be established.
Central cannulation
Central cannulation is accomplished via a median sternotomy with cannulation of the RA as the venous drainage and ascending aorta as the arterial return. Advantages of central VA-ECMO include increased venous drainage and antegrade return of oxygenated blood flow to the proximal aorta. Central cannulation is most commonly used for postcardiotomy patients but may benefit septic shock patients when higher flow is required. Disadvantages to central cannulation include increased risk of bleeding and infection and decreased patient mobility.
Venovenous cannulation
In the VV mode, deoxygenated blood is removed from the venous system and sent via an external pump to the membrane lung where oxygen is added and carbon dioxide removed. Oxygenated blood is then returned to the venous system. As the native heart must deliver oxygenated ECMO flow to the systemic circulation, VV requires adequate cardiac function. Venous blood is externally diverted from the patient, commonly by the inferior vena cava (IVC) and usually via the femoral vein, and then reinfused to the RA after gas exchange via the internal jugular vein. The drainage cannula is optimally placed in the hepatic IVC/RA region, as the largest volume of flowing blood for drainage exists here, and this vessel is less prone to collapse than those farther down in the abdomen. The internal jugular reinfusion cannula should be placed in the RA. The two cannulas should be separated by a few centimeters to reduce recirculation of ECMO return being drained back into the ECMO intake circuit without traversing the systemic circulation. Dual-lumen cannulas now exist that drain both the IVC and superior vena cava (SVC) and reinfuse oxygenated blood into the RA. When placed properly, these catheters direct oxygenated return through the tricuspid valve and into the right ventricle, thus reducing recirculation. As it is somewhat difficult to achieve proper placement of the dual-lumen cannula, placement under fluoroscopic or echocardiographic guidance is recommended. Unfortunately, dual-lumen cannulas appropriate for neonates are not readily available; those that are available have concerns regarding cardiac perforation or placement, which limit their use.
In patients cannulated for respiratory failure via VV-ECMO, sudden right ventricular failure can occur, often after days to a week on ECMO support. Although causation of such events is unclear, etiologies may include development of progressive pulmonary fibrosis and associated increased pulmonary vascular resistance, pulmonary emboli, fluid overload, or failure of the right ventricle from constant exposure to the relatively high-pressure jet returning from the ECMO circuit. Deterioration can occur rapidly, and if not medically managed successfully or the patient is not transitioned to VA support, right ventricular failure may lead to cardiac arrest. Periodic echocardiographic evaluation of right ventricular function coupled with clinical examination aimed at detecting venous distention or other evidence of right heart failure are advisable.
Physical systems for ECMO
The ECLS circuity is composed of cannulas, a pump, an oxygenator (which is also referred to as the membrane lung ), and tubing. Several technologic changes have been made to the ECLS circuit, including creation of centrifugal pumps with streamlined circuity and decreased priming volume; advancements in artificial membrane technology; and integrated systems that allow for ongoing assessment of venous saturations, blood gas values, temperature, hemoglobin, ECMO flow, ECMO pressures, and alarms. Newer systems also have enhanced portability and improved safety. ,
Two pump types exist for ECLS: centrifugal and roller-head. With the advent of newer centrifugal systems that are not associated with the severe hemolysis and plasma leakage issues of the past, most centers now use centrifugal pump and hollow-fiber/polymethlypentene membrane lung systems. Roller-pump devices generate high return pressure that poses justified concern for circuit rupture. Given this hazard and their dependence on gravity for venous drainage, many centers now have abandoned them. Centrifugal systems are also easier to set up, require less priming volume, may be safer, are easier to move, and facilitate patient mobility. In addition, centrifugal systems tend to require less immediate bedside technician monitoring, which reduces staffing needs and costs. Moreover, newer systems are thought to minimize the hemolysis encountered with earlier versions of centrifugal pumps. At this writing, newer centrifugal and membrane lung systems are appearing on the market, each with potential advantages and more miniaturization; however, whether these innovations truly affect outcome is yet unknown. Of interest, use of roller-pump devices continues in the United States for neonatal patients, whereas the rest of the world has changed to centrifugal setups. Several reports of an increase in hemolysis, hyperbilirubinemia, renal injury, and a decrease in survival in children with centrifugal pumps compared with roller devices have appeared, although whether these events are related to lack of experience with centrifugal systems or to other factors is unclear.
Much like the pumps, oxygenator technology has improved over time. Use of the silicone membrane lung, the workhorse for ECMO for many years, provided good gas exchange but was cumbersome to prime, stimulated a large inflammatory response because of its large surface area, and promoted thrombosis which required replacement. Its use has almost disappeared, and most centers have transitioned to microporous hollow-fiber oxygenators that employ polymethylpentene fiber technology. Hollow-fiber oxygenators are also advantageous because of their lower priming volume, smaller surface area, and improved gas exchange. These devices now exist in a variety of sizes, which makes them applicable to patients of varying size—neonates through adults.
Most ECMO systems also incorporate measurement of flow, revolutions per minute, pressure and alarm limits, and a variety of other indicators. Devices to measure flow, blood gases, and other parameters can also be placed on the circuit noninvasively. These even afford remote monitoring and adjustments in flow, which may allow multiple patients to be managed at distant sites. One new recently Food and Drug Administration (FDA)–approved system by Abiomed provides an oxygen concentrator, which obviates need for an external oxygen source, and a battery life of 3 hours, which may allow ambulation and facilitate care of patients outside the intensive care environment.
Management of anticoagulation and blood transfusion
During ECMO initiation, a massive inflammatory response occurs because of contact of patient blood with the foreign ECMO circuit, upregulating the coagulation system. Both activation and consumption of platelets occur, which induce changes in the coagulation cascade and promote a prothrombotic state. Anticoagulation is administered during cannulation and is continued throughout the ECMO run to mitigate the risk of thrombosis in both the circuit and the patient. Unfractionated heparin (UFH) remains the standard anticoagulant for ECLS patients, although newer anticoagulants, including direct thrombin inhibitors such as bivalirudin and argatroban, have been used successfully as well. No randomized trial exists currently to compare UFH with these newer anticoagulants to demonstrate if the anticoagulants are equivalent or if one approach is better than another. In a pilot single-center adult study, use of an algorithm-based heparin titration based on activated partial thromboplastin time (aPTT) was compared with a weight-dosed arm of 10 units/kg/hour with no titration for 10 patients on VV-ECMO. No differences were noted in need for circuit exchange, bleeding events, or amount of transfused blood. Other case series in patients with high risk for bleeding, including trauma or postcardiotomy, have shown that low-dose or no heparin during ECMO may be safe, but continuous monitoring of the circuit for thrombus formation is warranted. Other means of reducing platelet activation, aggregation, and circuit thrombosis such as surface coatings on the ECMO tubing, administering nitric oxide across the membrane lung, and use of antiplatelet agents, are areas of clinical and bench research.
Laboratory monitoring for UFH on ECLS varies among centers, and no single test or combination of tests has been found to be superior to another regarding bleeding or thrombosis. Although much effort and research have occurred and continue to try to define the optimal anticoagulation testing regimen for patients receiving ECLS, the answer is not currently known. Activated clotting time (ACT) is historically the most common laboratory test for monitoring UFH, but other tests, such as anti–factor Xa and viscoelastic assays, are being employed at many ECLS institutions. ACT is a whole-blood test that measures the time of whole blood to form a fibrin clot. It can be performed at the bedside and is simple to do, but its accuracy is affected by several factors, including thrombocytopenia, platelet dysfunction, hemodilution, hypofibrinogenemia, temperature, and technical factors. Anti–factor Xa assay is a chromogenic assay that measures the UFH inhibition of the factor Xa activity, and thus should be a better test for providing information specific to heparin effect. Viscoelastic testing methods such as thromboelastography (TEG) and rotational thromboelastometry (ROTEM) are whole-blood tests that measure the viscoelastic properties of a clot, including time to initiate clot formation, extent of fibrin formation, clot strength, and clot lysis. Although viscoelastic tests seem to provide a more comprehensive view of the quantity and function of coagulation factors, platelets, and fibrinogen, there currently are minimal data to support their use in ECLS, although more data are emerging. Bedside viscoelastic testing is now available at many centers, but it is currently unknown if it will improve anticoagulation management or if it will decrease the incidence of thrombotic or bleeding events. An understanding of what each anticoagulant test measures and its limitations is crucial to manage ECMO anticoagulation. Many studies suggest that a combination of probes, including a clotting time test such as ACT or TEG/ROTEM, plus a plasma-based test such as anti–factor Xa or aPTT, present a more comprehensive picture of the coagulation status and anticoagulant effect in ECLS patients.
Blood product transfusion thresholds vary based on patient age and type of ECLS (VV versus VA). Table 40.2 displays common transfusion thresholds recommended by ELSO. Other more detailed manuscripts on this topic are included in the references. , Adult patients tend to have lower hemoglobin goals than children, although reasons for this are unclear. Red blood cells are transfused with the primary goal to increase oxygen-carrying capacity and augment tissue oxygenation, but study results vary as to whether increasing hemoglobin improves clinical status or if the potential adverse effects of transfusion outweigh any potential benefit. Similar issues with adverse effects of platelet transfusion to achieve goal platelet count have been documented in ECMO patients. , , One common agreement regards the mandate to maintain normal fibrinogen levels during ECMO.
TABLE 40.2
Transfusion Thresholds for Blood Products in Adults and Children
| Adult | Pediatric | |
|---|---|---|
| Hemoglobin | 7 g/dL | 8–10 * g/dL |
| Platelet |
|
|
| FFP | >1.5–2.0 if bleeding | >1.5–2.0 if bleeding |
| Fibrinogen | <100–150 mg/dL | <100–150 mg/dL |
FFP: Fresh frozen plasma.
* Consider higher hemoglobin goal for cyanotic congenital heart disease.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here