Acute respiratory failure is one of the most common reasons for intensive care admission1 so that patients can be supported with mechanical ventilation. Mechanical ventilation with excessive airway pressure may result in ventilator-induced lung injury (VILI). VILI is known to worsen patient mortality. The concept of lung protective ventilation (LPV) was introduced to avoid VILI by limiting the peak inspiratory pressure (PIP) and tidal volume (TV). Mechanical ventilation with low TV (6–8ml/kg calculated to estimated body weight) reduces the incidence of VILI, thus improving patient survival (note that as all mechanical ventilation may be damaging, we prefer to use the term Least Damaging Lung Ventilation (LDLV)). The disadvantage of low minute volume ventilation is that it is known to lessen CO2 clearance leading to respiratory acidosis. Acidosis has deleterious effect on the cardiovascular system2.
Extracorporeal membrane oxygenators (ECMO) have been used to support patients in respiratory failure with hypoxia. ECMO is a way to provide gas exchange outside the body. Large intravascular cannulae allow a high volume of blood to flow across an oxygenator to effectively resolve hypoxia and hypercarbia. Smaller cannula decrease the incidence of complications but only allow a lower blood flow to circulate, this limits smaller systems to only removing CO2, due to the multiple factors determining the amount of O2 and CO2 that can be added or cleared by an extracorporeal system.
Extracorporeal CO2 removal (ECCO2R), a form of ECMO using lower blood flow, may provide a good solution for patients presenting with hypercarbia without profound hypoxia. CO2 can be cleared safely while oxygenation is not required. ECCO2R allows a decrease in ventilation that may in turn reduce the risk of VILI.
Physiology of extracorporeal CO2 removal
Extracorporeal gas exchange
In the healthy person O2 and CO2 diffuse through the alveolar membrane along their partial pressure gradient.
The artificial lung gas transfer follows the same law of diffusion but the surface area of the membrane is much less than the human lung, and only a proportion of the patient’s cardiac output can circulate through the system.
Moreover, the amount of O2 that may be added to the blood is limited by how oxygen is captured by the haemoglobin (losing efficiency when the saturation is greater than 75%). CO2 has greater lipid solubility and diffuses rapidly. CO2 removal will then primarily be affected by the venous partial pressure and the fresh gas flow through the membrane lung (sweep gas).
As only around 200ml of CO2 is produced per minute by the human body, a system removing this amount would maintain equilibrium. This can in effect be achieved with blood flows as low as 0.5L/min through an extracorporeal circuit. If the CO2 production increases (such as in sepsis or pregnancy) the blood volume may need to be increased and this may not be possible with an ECCO2R system.
Allowing least damaging lung ventilation (LDLV)
Avoidance of severe respiratory acidosis with extracorporeal CO2 removal allows the clinician to alter mechanical ventilation settings to achieve LDLV. Removing CO2 will in turn impact on oxygenation. The reduction in CO2 reaching the alveoli results in an increased alveolar concentration in O2. This can aid strategies to improve oxygenation such as increasing PEEP or application of inverse inspiratory-expiratory ratio.
The Membrane Lung
The membrane oxygenator consists of hollow fibres woven into a complex configuration contained in a plastic module. To reduce thrombus formation and improve biocompatibility the membrane surfaces are coated with an anticoagulant (usually heparin). Blood flows over the external surface of the fibres with sweep gas flowing over the internal surface. Sweep gas flow is adjusted to achieve an adequate partial pressure gradient.
Arterio-venous ECCO2R systems
The interventional lung assist (iLA®) is an arterio-venous ECCO2R system that does not incorporate mechanical pumps to maintain blood flow. The drainage cannula is inserted into the femoral artery with the return cannula into the contra-lateral femoral vein (see Figure 1). Blood is driven through the low resistance circuit by the patient’s arterio-venous pressure gradient. To achieve the maximal flow (1–2L/min) adequate systemic mean arterial pressure is required.
Fig. 1: Artery-venous ECCO2R using iLA® Novalung.
Veno-venous ECCO2R systems
Veno-venous ECCO2R systems incorporate a pump rather than relying on patients own systemic blood pressure to generate blood flow through the circuit. The pump uses the principle of centrifugal force created by the rotation of vanes. Flow depends on the rotations per minute (RMP-revolution per minute). The membrane oxygenator is placed downstream to the centrifugal pump. Sweep gas is attached to the membrane lung.
Novalung is a veno-venous ECCO2R system which incorporates a centrifugal pump able to achieve blood flows of 0.1–7L/min. Mono or dual lumen cannulae are available. Mono cannula are inserted into separate central veins (femoral or internal jugular vein). Dual lumen cannula allow both inflow and outflow, and are commonly placed into the internal jugular vein. Internal diameter 18, 22 and 24 French (Fr) are available. The combination of pump and larger cannula allows higher blood flow (0.1–7L/min). An integrated heat exchanger allows active temperature control.
The ALung has an integral centrifugal pump pushing blood through the circuit (see Figure 2). Access is via a dual lumen cannula (size 15.5Fr), inserted into the internal jugular or femoral vein. The small internal diameter of the cannula limits flow to 330–500ml/min. The control panel allows real-time observation of blood flow and rate of CO2 removal.
Fig. 2: ALung oxygenator with integral centrifugal pump.
Initiation and termination of ECCO2R
ECCO2R offers temporary respiratory support in reversible hypercapnic respiratory failure while providing complete or partial rest for the native lung. Reversibility of the underlying pathology must be assessed prior to initiation of ECCO2R. Destination device therapy is not available in respiratory failure thus extracorporeal support may be used to bridge the patient to recovery or transplantation.
When there is no hope of lung recovery, discontinuation of extracorporeal respiratory support should be considered. Withdrawing support of an alert and conversant patient who is dependent on ECCO2R and has irreversible lung condition may raise ethical concerns. End of life decisions should involve the patient, their family and the multidisciplinary team.
Bedside care includes monitoring and hourly recording of gas flow and blood flow across the membrane. Gas flows are titrated to ensure adequate CO2 removal. Low blood flow through the circuit will increase the risk of clots developing. Low flows may be caused by kinking of the tubing which is reduced by repositioning of the patient or the tubing.
The entire circuit, including the oxygenator, should be regularly checked for visible clots. If clots are noticed in the return arm the risk of them entering the patient necessitates a change of circuit.
Cannula sites are checked regularly for bleeding, to ensure fixation of cannula and detect early signs of infection. If femoral vessels are cannulated distal limb observations including pedal pulses are documented hourly.
Use of dual lumen cannula decreases the risks associated with immobility by allowing rehabilitation of the patient, including mobilising during ECCO2R support.
Initiation of ECCO2 removal is not without risk. Available literature is limited, highlighting the need to assess safety.
Blood clot formation is induced by foreign surface–blood interaction. Thrombi can be formed in any part of the ECCO2R circuit. Appropriate anticoagulation and heparin-coated surfaces reduce the risk.
There are no cases of air embolism reported in literature with ECCO2R. However, awareness is essential as accidental opening of the negative pressure side may lead to air entering the circuit.
Pump malfunction is a risk and backup circuits should be available and staff trained in emergency change of circuit.
Despite lower anticoagulation requirements, haemorrhage is the most common complication reported in patients supported on ECCO2R and can be explained by several mechanisms:
– Vascular damage: use of ultrasound guidance for percutaneous cannulation can reduce vascular complications
– Anticoagulation: systemic heparin is commonly used to avoid clot formation. Gastrointestinal and respiratory tract bleeding are reported in literature. One patient had severe retroperitoneal haemorrhage following removal of ECCO2R cannula.
In the case of AV-ECCO2R limb ischaemia is well-recognised. To reduce the risk, the arterial lumen should be 1.5 times the size of the arterial cannula used. Ischaemia necessitating limb amputation was reported in three cases in the literature. When arterial cannulation is performed, vessel wall damage can cause severe haemorrhage. Difficulty controlling bleeding may result in compartment syndrome.4,5
Acute respiratory distress syndrome (ARDS)
ARDS continues to carry the risk of high mortality and long-term morbidity.
LDLV is a well-established strategy to reduce VILI. Lowering minute volume may effect ventilation inducing severe respiratory acidosis.
A recently published review article identified two randomised controlled trials (RCT) aiming to explore the survival benefit of ECCO2R in ARDS. Several positive effects have been identified: ECCO2R device was found to facilitate reduction in mechanical ventilation. Intervention groups also had less sedation requirements.
Definitive evidence of improved mortality has not been provided. Both RCTs were terminated before adequate patient numbers were achieved as statistical analysis suggested that a significant difference between the control and interventional group was unlikely.6 Ventilator-free days were set as secondary outcome of the Xtravent study and these were significantly higher in the ECCO2R group in moderate to severe ARDS (PaO2/FiO2<150mmHg).
Sufficient clinical data to support the routine use of ECCO2R apparatus in ARDS are lacking.7
Two ongoing RCTs are aiming to establish efficacy and effect on patient mortality of ECCO2R. REST trial (protective ventilation with veno-venous lung assist in respiratory failure) is conducted in the United Kingdom and SUPERNOVA (a strategy of ultra-protective lung ventilation with ECCO2R for new onset moderate to severe ARDS) pilot study has been started in Europe. Results are awaited.
Chronic Obstructive Pulmonary Disease (COPD)
COPD is increasingly common in the developing world. Acute respiratory failure in COPD patients may be characterised by acute respiratory acidosis. Initial treatment with non-invasive ventilation (NIV) is a gold standard. However, NIV failure occurs in 40% of patients necessitating endotracheal intubation. Mechanical ventilation is known to increase intensive care stay and inhospital mortality.8
A recent systematic review on efficacy and safety of ECCO2R in acute respiratory failure in patients with COPD included 10 studies to explore two primary end-points: firstly, if ECCO2R was able to prevent endotracheal intubation following NIV failure and secondly the role of ECCO2R systems in early extubation. Results are encouraging: ECCO2R was found to reduce PaCO2 and improve pH, however due to heterogeneity of the studies included, no recommendation can be made about routine use of ECCO2R in this patient group. There has been no RCT conducted to provide robust evidence to date, the need for higher quality studies is clear. Ongoing research aims to establish the significance of ECCO2R in preventing endotracheal intubation as well as reducing mechanically ventilated days following NIV failure.9,10
Extracorporeal support may be instituted to provide bridge to lung transplantation in patients with progressive pulmonary disease.
A recently published case series of four patients evaluating low flow ECCO2R in the perioperative period of lung transplantation shows encouraging results. An ECCO2R device was initiated intraoperatively and in the postoperative period for treatment of hypercapnic respiratory acidosis. Significant improvement in PaCO2 and pH were observed.11
ECCO2R circuits have undergone rapid development since first described in 1970s. Improving technology targets simple, mobile, biocompatible circuits. Simplified technology aims to promote the application of ECCO2R in units other than those in specialised centres. Widespread use may result in improved complication rates.
Studies in preclinical and clinical settings have demonstrated that partial CO2 removal improves respiratory acidosis in patients with acute respiratory failure, however robust evidence on improvement of patient outcome is missing. Further clinical trials to explore efficacy in the clinical setting, cost effectiveness and safety are needed to gain a better established position in the management of acute respiratory failure.