- Pictorial Review
- Open Access
Imaging adults on extracorporeal membrane oxygenation (ECMO)
© The Author(s) 2014
- Received: 24 March 2014
- Accepted: 4 September 2014
- Published: 9 October 2014
Extracorporeal membrane oxygenation (ECMO) is increasingly being used in adults following failure to wean from cardiopulmonary bypass, after cardiac surgery or in cases of severe respiratory failure. Knowledge of the different types of ECMO circuits, expected locations of cannulas and imaging appearance of complications is essential for accurate imaging interpretation and diagnosis. Commonly encountered complications are malposition of cannulas, adjacent or distal haemorrhage, stroke, stasis thrombus in access vessels, and distal emboli. This article will describe the imaging appearance of different ECMO circuits in adults as well as commonly encountered complications. If a CT (computed tomography) angiogram is being performed on these patients to evaluate for pulmonary embolism, the scan may be suboptimal from siphoning off of the contrast by the ECMO. In such cases, an optimal image can be obtained by lowering the flow rate of the ECMO circuit or by disabling the circuit for the duration of image acquisition.
• Femoroatrial VV ECMO: femoral vein drainage cannula and right atrial return cannula.
• Femorofemoral VV ECMO: return and drainage cannulas placed in femoral veins.
• Dual-lumen single cannula VV ECMO: via the right IJ/Femoral vein with the tip in the IVC/SVC.
• Peripheral VA ECMO: peripheral venous drainage cannula and peripheral arterial return cannula.
• Central VA ECMO: direct right atrial drainage cannula and aortic return cannula.
- Extracorporeal membrane oxygenation
Extracorporeal membrane oxygenation (ECMO) refers to the life support system utilised in pulmonary or cardiopulmonary support for gas exchange . The goal of the system is to oxygenate the patient’s blood while removing carbon dioxide. Also referred to as extracorporeal life support, ECMO is well established in neonatal respiratory failure. Use in adults has increased and ECMO is commonly used following failure to wean from cardiopulmonary bypass after cardiac surgery or in cases of severe respiratory failure .
Indications and Complications of VA and VV ECMO
After failure to wean from cardiopulmonary bypass
May predispose to aortic stasis thrombus
When cardiac and pulmonary support is required
Large bore arterial cannulas may predispose to occlusion
Carotid cannulation may contribute to stroke
Direct support of gas exchange-respiratory failure; ARDS
DVT in the cannulated limb
Complete discussion of paediatric ECMO is beyond the scope of this article. Neonatal and paediatric ECMO encompasses different cannula sizes, configurations and complications. For example, in neonates arterial cannulation is limited to placement via sternotomy or via cut down technique for carotid cannulation. In such small patients, peripheral cannula placement is exceedingly difficult. Additionally, initiation of VA ECMO usually encompasses carotid or internal jugular vein ligation, a technique that is not well tolerated in adults given the risk of stroke . Similarly, ECMO in the non-neonatal paediatric population may or may not require carotid or jugular ligation. Imaging considerations in paediatric ECMO have been previously discussed in the literature . For the purposes of this discussion, only the imaging of adults on ECMO will be addressed.
In 2009, the CESAR study found improved outcomes in adults treated with ECMO versus conventional therapy . This study randomised patients with severe but reversible respiratory failure to either conventional therapy with intermittent positive pressure ventilation or therapy with ECMO. With the primary end point being severe disability or death at 6 months, the CESAR study demonstrated improved survival and decreased morbidity in patients randomised to ECMO.
There are two types of ECMO: veno-arterial (VA) and veno-venous (VV) . VA ECMO refers to siphoning of deoxygenated blood from a vein with return of oxygenated blood into an arterial vessel. VV ECMO refers to siphoning of deoxygenated blood from a venous vessel and return of oxygenated blood to a systemic venous vessel or the right atrium. The catheters used in ECMO for blood exchange are referred to as ‘cannulas’, a term used in this context to avoid confusion with other vascular catheters. ECMO cannulas may assume variable configurations on radiographs and CTs. Familiarity with differing positions and access sites is important when interpreting images. Additionally, a multitude of complications may arise intrinsic to either ECMO or cannula placement.
The circuit consists of a pump, blender, oxygenator, control console, heater/cooler and two cannulas. The blender mixes oxygen with CO2. This mixture of gas usually consists of 95 % oxygen and 5 % CO2 and is referred to as the ‘sweep gas’. The pump is usually a centrifugal or roller pump and is used to transport blood throughout the circuit. The heater/cooler aids in temperature regulation of extracorporeal blood. The cannulas provide direct transport of oxygenated and deoxygenated blood to and from the patient.
The initial oxygenators consisted of venous blood pools that were oxygenated with oxygen bubbles. These first clinical attempts at extracorporeal life support began in the 1950s and were largely unsuccessful beyond several hours as direct exposure of blood to oxygen led to life-threatening coagulopathy, haemolysis, and multi-organ failure. In 1959, Burns developed the definitive solution involving a silicone membrane that separated blood from oxygen and allowed for indirect oxygenation of blood . This was a major step in extracorporeal life support. In the years following (1972–1976) the first published reports of the successful use of extracorporeal life support emerged . Modern membranes are largely made from polymethyl-pentene, which is lower in resistance and causes less consumption of blood products .
Indications for initiation of adult ECMO
VA ECMO is used for both cardiac and pulmonary support such as acute cardiac failure or failure to wean from cardiopulmonary bypass after cardiac surgery . VV ECMO is used for reversible respiratory failure with normal cardiac function. The most common use is in acute respiratory distress syndrome, which may be secondary to severe pneumonia or influenza . More specifically, eligibility was defined in the CESAR trial by a Murray score >3, a pH of less than 7.20 and having a reversible disease process.
Contraindications to ECMO
Relative contraindications include high pressure ventilation, high FiO2, limited vascular access and organ dysfunction that would lead to poor quality of life. In patients with pre-existing disease such as metastatic cancer or severe irreversible brain injury, ECMO may be contraindicated. Absolute contraindications include any conditions that preclude anticoagulation .
Imaging modalities used in evaluation of ECMO
Transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) play crucial roles during initial cannula placement. A full discussion on the role of echocardiography in ECMO is discussed by Platts et al. in “The Role of Echocardiography in the Management of Patients Supported by Extracorporeal Membrane Oxygenation” .
Radiographs of the chest and abdomen are useful initial examinations for cannula position. They may reveal malposition or unintended migration. They often offer the first clue towards complications such as haemothorax, pneumothorax or mediastinal fluid collections.
Ultrasound is a portable, focussed and readily available modality that can be used to evaluate for insertion site haematomas, peri-cannula thrombus or deep venous thrombus. Spectral colour Doppler is a powerful tool to evaluate distal limb perfusion as the arterial return cannula can lead to obstruction/occlusion with decreased forward flow in the access artery. It can also be used for evaluating thrombus in the access vein. Interpretation of spectral Doppler waveforms with comparison of flow in the contralateral vessel can help in differentiating slow flow, which may be due to a systemic cause such as poor cardiac output, from slow flow in the access artery from either the access cannula itself or thrombus.
Imaging modalities in ECMO
Pneumothorax and haemothorax
Evaluation of quadrants for Murray scoring.
Comparison to priors for change
During initial placement of ECMO cannulas
Evaluate recovery of cardiac function
Ultrasound ± Doppler
Evaluation of vascular patency
Haematoma at cannulation site
Comparison of cannulated and non-cannulated vessels aids in evaluation of abnormal vascular waveforms
Aortic stasis thrombus
Maintain resting respiratory rate to minimise motion artefact
Pulmonary stasis thrombus Aortic stasis thrombus
Switch ECMO circuit to minimal flow status or stop the ECMO pump for the duration of the acquisition
VV ECMO is usually placed percutaneously and peripherally. Several configurations are seen. In the ‘femoroatrial’ configuration the drainage cannula is introduced via the femoral vein and advanced to the level of the diaphragm, below the hepatic veins . The return cannula is introduced via the internal jugular vein and advanced to the level of the superior vena cava (SVC)-atrial junction (Fig. 1). The tip is directed towards the tricuspid valve. This particular configuration is optimal to minimise recirculation , a phenomenon in which a portion of the returning oxygenated blood is prematurely siphoned back to the ECMO circuit. Recirculation will result in less oxygenated blood returning to the pulmonary circulation and thus the systemic circulation.
A configuration in which the drainage cannula and return cannula are introduced via the femoral veins is termed 'femorofemoral ECMO'. In this configuration the drainage cannula may be placed on one side and the return cannula on the other. Alternatively, both cannulas may be introduced on the same side. In either case the drainage cannula is advanced to the distal inferior vena cava (IVC) and the return cannula is advanced to the right atrium .
Summary of common ECMO configurations
Optimal cannula tip position
Direct insertion into the mediastinal vessels
Distal IVC or SVC, before the cavoatrial junction
Proximal femoral artery, axillary artery, subclavian artery
Insertion of cannulas in peripheral vessels
Distal IVC, at the level of the diaphragm
Right atrium via the same or opposite iliofemoral vein
Less recirculation and improved flow
Distal IVC, at the level of the diaphragm
Distal SVC/right atrium via the SVC
Optimal to minimise recirculation
Dual lumen, single cannula
IVC, below the diaphragm
Right atrium via the SVC
For urgent pulmonary support
Gaseous microemboli can be introduced into the ECMO circuit during cannula placement, from inadequate priming of the filter system, from the perfusionist accessing the circuit (for blood draws or medication injection), from the pump or from turbulent flow within the tubing [17, 18]. Only large air emboli may be seen on conventional imaging and only a few cases exist in the literature. Gaseous microemboli are not well seen on conventional imaging and in fact require specialised ultrasound machines for detection in the circuit .
Access vessel obstruction/occlusion
Vascular complications are seen in 10–16.9 % of patients on ECMO . Clinical presentation may include pallor, diminished or absent pulses, compartment syndrome or gangrene. In our experience, spectral Doppler ultrasound is the imaging modality of choice for evaluation of peripheral vessels in this setting.
In venous obstruction, colour Doppler may show a lack of flow or absence of normal phasicity on spectral Doppler imaging. Although much consideration is given to arterial perfusion, venous drainage should be an equally important consideration. If the venous cannula is large enough to obstruct venous return, severe limb oedema and ischaemia result . As such, one may see an additional catheter in the distal access vein that assists with venous drainage. Comparing flow in the non-cannulated extremity vessel will again help elucidate whether the suspected flow abnormality is due to poor cardiac function or true occlusion.
Venous and arterial thrombus
The incidence of deep venous, pulmonary and arterial thrombosis in patients on ECMO is not well documented in the literature. In fact experience with these entities is restricted mainly to case reports. In cases of acute arterial thrombus in the setting of peripheral VA ECMO, spectral Doppler imaging will show lack of flow, absent waveforms and an abrupt cutoff of distal vessels. In venous occlusion by thrombus, spectral Doppler imaging may show thrombus within or surrounding the cannula. There may also be enlargement of the affected vein and extension of the thrombus distally.
Cerebral ischaemia and stroke
Distal haematoma may be seen as haemothorax or haemorrhagic ascites. It is usually multifactorial. Factors such as anticoagulation, consumption of complement and prothrombotic substances in addition to recent cannulation may cause subacute haemothorax to develop over time. Radiographs may demonstrate new effusion and/or displacement of mediastinal structures. Ultrasound will demonstrate echogenic fluid that can be confirmed by CT as high-attenuation fluid.
The first is to reduce pump flow to the ECMO circuit typically to 500 cc/min for 15–25 s, inject IV (intravenous) contrast and place the region of interest locator at the main pulmonary artery. Minimal flow will preclude thrombus formation in the cannulas. With the ECMO circuit in low-flow status, IV contrast will opacify the pulmonary arteries . As cardiac function may be poor in these patients, CT acquisition may be manually triggered when the region of interest demonstrates a peak or a plateau in the opacification of the main pulmonary artery. This plateau will vary according to right ventricle function and may be different for each patient. Triggering the scan will vary accordingly. Scan parameters are as follows: kVp and mAs are as determined from the initial scouts. IV contrast is injected at 4–5 cc/s (contrast volume = 4–5 cc × [delay time of the scanner from initiation of bolus trigger + scan duration]) .
If pulmonary arterial opacification is suboptimal with the above technique, a repeat scan should be considered with the circuit completely disabled for the duration of the CT acquisition. This will typically be 10–15 s for a 64-slice MDCT. The pulmonary arteries will be opacified in a near anatomic manner depending on cardiac function. Bolus tracking as described above may be used.
ECMO cannula configurations are extremely variable. The radiologist should be familiar with the expected locations of cannulas in VV ECMO, peripheral VA ECMO and central VA ECMO. As such, the radiologist should be able to comment on malposition and prospectively search for complications such as vascular obstruction. Additionally, there should be a high index of suspicion for haemorrhage, venous thrombus, stasis thrombus and cerebral ischaemia. Familiarity with the appropriate imaging modalities and common complications will be useful for the interpreting radiologist.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.
- Allen S et al (2011) A review of the fundamental principles and evidence base in the use of extracorporeal membrane oxygenation (ECMO) in critically ill adult patients. J Intensive Care Med 26(1):13–26View ArticlePubMedGoogle Scholar
- Organization ELS (2014) ELSO website, under registry information. [website] 2014 [cited 2014; Available from: http://www.elso.org/index.php?option=com_content&view=article&id=95&Itemid=478.
- Zapol WM et al (1979) Extracorporeal membrane oxygenation in severe acute respiratory failure. A randomized prospective study. JAMA 242(20):2193–2196View ArticlePubMedGoogle Scholar
- Maslach-Hubbard A, Bratton SL (2013) Extracorporeal membrane oxygenation for pediatric respiratory failure: history, development and current status. World J Crit Care Med 2(4):29–39PubMed CentralView ArticlePubMedGoogle Scholar
- Fenik JC, Rais-Bahrami K (2009) Neonatal cerebral oximetry monitoring during ECMO cannulation. J Perinatol 29(5):376–381View ArticlePubMedGoogle Scholar
- Goodwin SJ et al (2014) Chest computed tomography in children undergoing extra-corporeal membrane oxygenation: a 9-year single-centre experience. Pediatr Radiol 44(6):750–760View ArticlePubMedGoogle Scholar
- Peek GJ et al (2009) Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet 374(9698):1351–1363View ArticlePubMedGoogle Scholar
- Brodie D, Bacchetta M (2011) Extracorporeal membrane oxygenation for ARDS in adults. N Engl J Med 365(20):1905–1914View ArticlePubMedGoogle Scholar
- Lim MW (2006) The history of extracorporeal oxygenators. Anaesthesia 61(10):984–995View ArticlePubMedGoogle Scholar
- Khoshbin E et al (2005) Performance of polymethyl pentene oxygenators for neonatal extracorporeal membrane oxygenation: a comparison with silicone membrane oxygenators. Perfusion 20(3):129–134View ArticlePubMedGoogle Scholar
- Sidebotham D et al (2009) Extracorporeal membrane oxygenation for treating severe cardiac and respiratory disease in adults: Part 1–overview of extracorporeal membrane oxygenation. J Cardiothorac Vasc Anesth 23(6):886–892View ArticlePubMedGoogle Scholar
- Murray JF et al (1988) An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis 138(3):720–723View ArticlePubMedGoogle Scholar
- Platts DG et al (2012) The role of echocardiography in the management of patients supported by extracorporeal membrane oxygenation. J Am Soc Echocardiogr 25(2):131–141View ArticlePubMedGoogle Scholar
- Sidebotham D et al (2010) Extracorporeal membrane oxygenation for treating severe cardiac and respiratory failure in adults: part 2-technical considerations. J Cardiothorac Vasc Anesth 24(1):164–172View ArticlePubMedGoogle Scholar
- Kanji HD et al (2010) Peripheral versus central cannulation for extracorporeal membrane oxygenation: a comparison of limb ischemia and transfusion requirements. Thorac Cardiovasc Surg 58(8):459–462View ArticlePubMedGoogle Scholar
- Rupprecht L et al (2013) Cardiac decompression on extracorporeal life support: a review and discussion of the literature. ASAIO J 59(6):547–553View ArticlePubMedGoogle Scholar
- Lou S et al (2011) Generation, detection and prevention of gaseous microemboli during cardiopulmonary bypass procedure. Int J Artif Organs 34(11):1039–1051View ArticlePubMedGoogle Scholar
- Win KN, Wang S, Undar A (2008) Microemboli generation, detection and characterization during CPB procedures in neonates, infants, and small children. ASAIO J 54(5):486–490View ArticlePubMedGoogle Scholar
- Cheng R et al (2014) Complications of extracorporeal membrane oxygenation for treatment of cardiogenic shock and cardiac arrest: a meta-analysis of 1,866 adult patients. Ann Thorac Surg 97(2):610–616View ArticlePubMedGoogle Scholar
- Rao AS et al (2010) A novel percutaneous solution to limb ischemia due to arterial occlusion from a femoral artery ECMO cannula. J Endovasc Ther 17(1):51–54View ArticlePubMedGoogle Scholar
- Chauhan S, Subin S (2011) Extracorporeal membrane oxygenation, an anesthesiologist’s perspective: physiology and principles. Part 1. Ann Card Anaesth 14(3):218–229View ArticlePubMedGoogle Scholar
- Kasirajan V et al (2002) Technique to prevent limb ischemia during peripheral cannulation for extracorporeal membrane oxygenation. Perfusion 17(6):427–428View ArticlePubMedGoogle Scholar
- Russo CF et al (2009) Prevention of limb ischemia and edema during peripheral venoarterial extracorporeal membrane oxygenation in adults. J Card Surg 24(2):185–187View ArticlePubMedGoogle Scholar
- Madershahian N et al (2013) Thrombosis of the aortic root and ascending aorta during extracorporeal membrane oxygenation. Intensive Care MedGoogle Scholar
- Mateen FJ et al (2011) Neurological injury in adults treated with extracorporeal membrane oxygenation. Arch Neurol 68(12):1543–1549View ArticlePubMedGoogle Scholar
- Auzinger G et al (2013) Computed Tomographic Imaging in Peripheral VA-ECMO: Where Has All the Contrast Gone? J Cardiothorac Vasc AnesthGoogle Scholar
- Liu KL et al (2014) Multislice CT scans in patients on extracorporeal membrane oxygenation: emphasis on hemodynamic changes and imaging pitfalls. Korean J Radiol 15(3):322–329PubMed CentralView ArticlePubMedGoogle Scholar
- Khadir MM et al (2014) Looking beyond the thrombus: essentials of pulmonary artery imaging on CT. Insights Imaging 5(4):493–506PubMed CentralView ArticlePubMedGoogle Scholar