During VV-ECMO, the arterial oxygen content (CaO₂) will depend on the O₂ delivering contributions of both the ML and the NL (Fig. 7.1).

Venous blood returns from the peripheral tissues with low oxygen content (CvO₂). The blood pump generates the extracorporeal blood flow (ECBF) diverting part of the venous return, i.e. cardiac output (CO), towards the ML. The ML loads the ECBF with oxygen, raising the oxygen content from the inlet (CinO₂) to the outlet (CoutO₂) of an amount corresponding to the oxygen delivery of the ML (VO_2ML = ECBF * (CoutO₂ − CinO₂)).

The VO_2ML and then the contribution of ML to total oxygen delivery depend on three main factors:

The oxygenated ECBF is then returned to the venous blood and directed towards the right heart. The oxygen content of the blood returning to the lung (the mixed venous blood, CmvO₂) is the flow-weighted average of CoutO₂ and CvO₂: CmvO₂ = [CoutO₂*ECBF+ CvO₂*(CO − ECBF)]/CO. In practice, the resulting effect of the membrane lung is to increase the O₂ content of the blood returning to the lung from CvO₂ to CmvO₂. With the given CvO₂, CmvO₂ and then the mixed O₂ saturation are directly proportional to ratio between ECBF and CO.

The oxygen contribution of the native lung (VO_2NL = (CaO₂ − CmvO₂)*CO) depends on its residual gas exchange capability which depends on both the severity of the lung disease (i.e. the intrapulmonary shunt fraction) and the ventilator management.

The sum of VO_2NL and VO_2ML, i.e. the total oxygen delivered, corresponds to the total oxygen consumption of the patient (VO₂Tot = VO_2NL + VO_2ML). Consequently, the lower the NL contribution, the higher must be the contribution of the ML and then the oxygen delivery efficiency of the ECMO system. A highly efficient VV-ECMO system, able to deliver a high VO_2ML, requires high ECBF, minimal recirculation and low CinO₂.

While the efficiency in delivering O₂ has several limitations, removal of CO₂ during VV-ECMO is much easier to achieve. While the amount of O₂ delivery is limited by full saturation of arterialized blood and by haemoglobin concentration, a substantial amount of CO₂ can be removed from venous blood. In fact, normal venous blood carries around 50 mL of CO₂/100 mL of blood, most of which in the form of bicarbonate ion. When the ECBF flows through the ML, dissolved CO₂ is transferred from the blood to the gas side, while new dissolved CO₂ is released from carbonic ion assuring a continuous availability of dissolved transferable CO₂. In this process, the main limiting factor to the amount of removable CO₂ is the partial pressure of CO₂ on the gas side, which depends on the ECGF. High ECGF will maintain the partial pressure of CO₂ on the gas side close to zero, thus maintaining a high CO₂ partial pressure gradient. Indeed, contrary to oxygen delivery, efficient CO₂ removal can be obtained with relatively lower ECBF.

Achieving proper vascular access is a fundamental step in implementing an ECMO treatment [9–15]. An important historical step in cannulation for ECMO has been the availability of thin-walled cannulas, in the early 1990s, that allowed to move from surgical to percutaneous cannulation techniques. The percutaneous approach has several advantages: reduced risk of bleeding, shorter operative time and easier mobilization and nursing of the patient. This reason has become the standard technique for VV-ECMO [9–15].

Choice of vessels, cannulas type and size and configuration are mainly dictated by the maximum ECBF needed for the support of the patient, the maximum recirculation tolerable, the patient comfort, the patient anatomical features, the presence of some obstructed vein and local preferences.

The size and position of the drainage cannula are the main determinants of ECBF. For ECBF higher than 3–4 l/min, drainage cannula size to up 23–28 Fr is necessary. The best position is the intrahepatic portion of the inferior vena cava or in the right atrium. Cannulas with multiple holes distributed along the cannulas are available to enhance blood drainage (multiple-stage drainage cannulas). The reimmission cannulas normally have holes only in a short portion near to their extremity.

For VV-ECMO four different cannula configurations are available (Fig. 7.2):

The contact between patient blood and the ECMO foreign surface is associated with activation of coagulation factors and complement factors, platelets and fibrinogen consumption [21–24], which may lead to bleeding [25] and thromboembolic [26] complications. For these reasons, continuous anticoagulation is necessary throughout the ECMO treatment. Anticoagulation during VV-ECMO is commonly achieved by continuous intravenous heparin infusion, targeted to a partial thromboplastin time (aPTT) of 45–60 s and/or to an activated clotting time (ACT) of 1.5–2 times normal. Transfusion policy regarding platelet count varies among centres. Transfusion thresholds between 50,000 and 100,000/μL have been reported. Activation of the coagulation cascade and clot formation in the circuit are accompanied by a progressive decrease in platelet count and fibrinogen level and consumption of coagulation factors that may lead to a syndrome similar to disseminated intravascular coagulation. For this reason, activation of the coagulation and circuit thrombosis, along with oxygenator performance and/or increase of transmembrane pressure, must be monitored to understand the need for a prompt replacement of ECMO circuit. D-dimer levels have been shown to be a reliable variable to detect circuit thrombosis.

Long-term applications of ECMO support are an evolution of heart-lung machine use in cardiac surgery. The first successful use of prolonged ECMO in an adult ARDS patient was reported by Hill et al. in 1972 [27]. After only 2 years from first successful cases, the National Institute of Health commissioned a multicentre randomized clinical trial on prolonged ECMO for adults with ARDS [28]. At the time, the main indication for ECMO in ARDS patients was to correct the hypoxaemia while buying time for the lungs to heal [29]. As such, high ECMO blood flow rates were necessary to improve oxygenation, and the veno-arterial configuration was the standard. Following the negative results of the study, which failed to show any survival benefit from ECMO, only few investigators continued studying and improving the technique. Contrary to adult applications, applications in newborn showed encouraging results [30–35]. This contributed to maintain some interest on ECMO.

A landmark contribution to the evolution of adult ECMO came from the pioneering work done in the laboratory of T. Kolobow at the NIH. In a series of study, Kolobow and Gattinoni showed that [36–40] (1) nearly all the metabolic CO₂ production could be removed through an artificial lung using much lower extracorporeal blood flows than those required to oxygenate the blood and (2) removal of CO₂ through the membrane lung allows to reduce, virtually to zero, ventilation of the native lung. Based on these observations, they developed the concept of extracorporeal CO₂ removal (ECCO₂R) and proposed the use of a VV bypass configuration instead of the classical VA mode and the use of low-frequency ventilation to allow lung rest [40]. Unfortunately, a single-centre, randomized controlled trial conducted by Morris et al. in 40 ARDS patients failed to show a benefit from this technique [41]. Nevertheless, few centres around the world continued to apply ECMO in adult patients, contributing to improve the ECMO technique for long-term applications [42–45].

Following the concept of lung rest developed by Kolobow et al., several animal and human studies have demonstrated that, though necessary to preserve life, mechanical ventilation (MV) can exacerbate lung damage eventually contributing to mortality [46–54]. This has led to the concept of ventilator-induced lung injury (VILI) [54] and of “protective ventilatory strategy”, mainly consisting of low Vt (6–8 mL/Kg PBW) and limitation of plateau airway pressures (Pplat <28–30 cmH₂O). ECMO is a unique technique to master protective ventilation and lung rest since, as Kolobow foresaw, it allows to minimize the need for mechanical ventilation allowing for using extremely low tidal volumes and to apply safe plateau pressures even in the most severe patients. Finally, in 2009, Peek et al. published the results of the Conventional Ventilation or ECMO for Severe Adult Respiratory Failure (CESAR) trial [55]. The study enrolled 180 adult patients (18–65 years) with severe but potentially reversible acute respiratory failure, defined as a Murray score ≥3 or uncompensated hypercapnia with a pH < 7.20. Patients randomized to receive ECMO were transferred to the Leicester ECMO centre, while controls remained in designated treatment centres. The trial showed a survival advantage (47% vs 63% at 6 months) associated with the use of ECMO in adults. Unfortunately, the enthusiasm for the positive results has been mitigated by important limitations of the study. First, not all patients allocated to the ECMO group received ECMO, because they died before or during transportation (five patients) or they improved with conventional treatment after transportation to the ECMO centre (17 patients). Second, as there was no standardized protocol for mechanical ventilation in the control group, significantly fewer patients in the control group received a protective ventilatory strategy. However, despite these limitations, the results of the study have led to an increase in interest in ECMO worldwide. Incidentally, publication of the study almost coincided with the outbreak of H1N1 influenza in 2009. Following the successful use of ECMO in patients with H1N1-induced severe ARDS reported from the Australian and New Zealand experience [50], several countries recurred to the use of ECMO, even organizing national network able to centralize the most severe patients. The published case series from these experiences report survival rates ranging from 68 to 83% [56–63].

Since 2009, the use of respiratory ECMO in adults has continued to grow. The relative simplification of the technique together with the acquisition of experience has push more and more centres to institute specialized ECMO teams able to employ different ECMO techniques for different indications.

There are two main indications to VV-ECMO in ARDS (Fig. 7.3):

Total respiratory extracorporeal support refers to ECMO application in the most severe acute respiratory failure patients with hypoxaemia refractory to lung-protective ventilation and adjunctive therapies like inhaled nitric oxide, prone positioning and recruitment manoeuvres. The aim is to provide viable oxygenation when the natural lung becomes unable to provide it and prevent the mechanical ventilation-associated lung damage. Therefore, total extracorporeal respiratory support is viewed as a rescue therapy for severely hypoxaemic ARDS patients.

The reason to institute ECMO in severe ARDS is to break the “vicious cycle” of tissue hypoxia (Fig. 7.3) [64]. Refractory hypoxaemia leads in the most severe cases to tissue hypoxia; in this setting the weapons that conventional mechanical ventilation uses to relieve hypoxaemia are the increase of FiO₂ up to 1 and the increase of mean airway pressure (PEEP, I:E ratio manipulation, plateau, recruitment manoeuvres). The increase in intrathoracic pressures leads in turn to haemodynamic compromise defacing, in the most severe cases, the effects of the relief of hypoxaemia on tissue hypoxia. In this setting, institution of high-flow total ECMO support can provide viable delivery of oxygen to the tissues also when gas exchange through the natural lung is completely abolished.

Indeed, high-flow ECMO provides also total CO₂ removal, abolishing the need of tidal ventilation of the natural lung to remove CO₂. During ECMO the natural lungs can be maintained “at rest” avoiding high plateau pressures and high minute ventilation requirements that stretch the low-compliance “baby lungs” of severe ARDS leading to barotrauma, volutrauma and biotrauma and breaking therefore the vicious cycle of ventilator-associated lung injury (VALI or VILI) and the final dismal result of multiple organ failures (MOFs) leading to death [65].