Extracorporeal carbon dioxide removal for acute hypercapnic respiratory failure

In the past, the only treatment of acute exacerbations of obstructive diseases with hypercapnic respiratory failure refractory to medical treatment was invasive mechanical ventilation (IMV). Considerable technical improvements transformed extracorporeal techniques for carbon dioxide removal in an attractive option to avoid worsening respiratory failure and respiratory acidosis, and to potentially prevent or shorten the duration of IMV in patients with exacerbation of COPD and asthma. In this review, we will present a summary of the pathophysiological rationale and evidence of ECCO2R in patients with severe exacerbations of these pathologies.


Background
Patients with obstructive lung diseases, such as asthma and chronic obstructive pulmonary disease (COPD), may experience acute exacerbations with severe hypercapnic respiratory failure. Hypercapnia results from acute worsening of expiratory flow limitation caused by the increased small airway resistance with consequent development of dynamic alveolar hyperinflation and intrinsic positive end-expiratory pressure (PEEP). In the most severe cases, these may be refractory to conventional therapies and mechanical ventilation, becoming life-threatening.
Extracorporeal carbon dioxide removal (ECCO 2 R) represents an attractive approach in this setting.
The last decade has seen an increasing interest in the provision of extracorporeal support for respiratory failure, as demonstrated by the progressively increasing number of scientific publications on this topic. In particular, remarkable interest has been focused on extracorporeal carbon dioxide removal (ECCO 2 R), due to the relative ease and efficiency in blood CO 2 clearance granted by extracorporeal gas exchangers as compared to oxygen delivery [1].
In recent years, a new generation of ECCO 2 R devices has been developed. More efficient veno-venous (VV)-ECCO 2 R devices have become available and have replaced the arterio-venous approach, having the advantage of not requiring arterial puncture.
The new VV-ECCO 2 R devices offer lower resistance to blood flow, have smaller priming volumes, and provide a much more efficient gas exchange with relatively low extracorporeal blood flows (0.4-1 L/min) [2]. The technology of these devices is now comparable to that of renal dialysis and has been experimented in several animal and human studies, demonstrating significant reduction in arterial CO 2 and improvement in the work of breathing [3][4][5][6].

Pathophysiological rationale for ECCO 2 R in obstructive lung diseases
In both asthma and COPD exacerbations, diffuse narrowing of the airways results in detrimental physiological consequences. Airway narrowing prevents the lungs from completely emptying ("air trapping") due to resistance to expiratory flow and bronchial closure at higher than normal lung volumes. Air trapping results in dynamic hyperinflation (DHI) [7] which is the excessive increase in end-expiratory lung volume above the relaxation volume of the respiratory system, generating intrinsic positive end-expiratory pressure (auto-PEEP) [8]. As a result, the patient breathes at higher total lung volumes, Open Access *Correspondence: luchomq2077@gmail.com 1 Intensive Care Unit, Hospital Universitario Sagrado Corazón, Barcelona, Spain Full list of author information is available at the end of the article due to increased residual volume [9], which may reduce tidal ventilation. The net effect is that the work of breathing increases significantly. The diaphragm, intercostal muscles, and even the abdominal muscles are overloaded causing respiratory muscle fatigue and dyspnea [10].
Pharmacotherapy with bronchodilators and systemic corticosteroids are part of the medical therapies, administered specifically to reduce the pathophysiological airflow obstruction and improve symptoms.
The recognition for the need for noninvasive ventilation (NIV) is indicated if the patient fails to improve clinically and if the level of pH remains less than 7.32 despite medical therapy [11]. However, NIV fails in up to 20-30% of patients and IMV is indicated with specific ventilation strategies, targeting relative short inspiratory time and longer expiratory time [12,13].
Overall, the goal of mechanical ventilation is to provide adequate gas exchange and reduce the work of breathing while waiting for airflow obstruction to resolve. However, mechanical ventilation itself may aggravate alveolar hyperinflation by worsening DHI, which may lead to worsened hypercapnia, barotrauma, alveolar rupture leading to pneumothorax and further hemodynamic deterioration [14].
Furthermore, if treated with IMV, these patients receive sedatives and likely neuromuscular blockade to facilitate ventilatory support [15]. Sedation and paralysis preclude mobilization, promoting neuromuscular deconditioning, and potentially contributing to the long-term cognitive sequelae of critical illness [16].
When conventional therapeutic options are not successful, novel therapies such as extracorporeal life support are entertained as a possible salvage therapeutic modality.
During exacerbation, relieving the native lung from at least part of the CO 2 elimination with ECCO 2 R could potentially improve the acid-base balance, reduce patient's work of breathing with a consequent reduction in respiratory rate and ventilatory drive, and lower alveolar ventilation. The lower tidal volumes and respiratory rate result in the extension of the expiratory time, suiting better the high expiratory time constant of the respiratory system with expiratory flow limitation. By these physiological mechanisms, ECCO 2 R can counteract the vicious circle of dynamic hyperinflation, and its detrimental respiratory and cardiovascular consequences. The derived beneficial effects on respiratory mechanics, ventilatory muscle efficiency, work of breathing, and cardiovascular function may improve gas exchanges and relieve dyspnea. By these mechanisms, ECCO 2 R thus can potentially prevent NIV failure, facilitate weaning from IMV, and therefore contribute to avoid the unwanted complications of sedation and immobilization.

ECCO 2 R technical aspects and principle
ECCO 2 R is designed to remove carbon dioxide (CO 2 ) and, unlike extracorporeal membrane oxygen (ECMO), does not provide significant oxygenation.
The device consists of a drainage cannula placed in a large central vein or artery, a membrane lung (artificial gas exchanger), and a return cannula into the venous system ( Fig. 1). Blood is pumped through the membrane lung, and CO 2 is removed by diffusion. A flowing gas known as "sweep gas" containing little or no CO 2 runs along the other side of the membrane, ensuring a diffusion gradient from blood to the other side, hence promoting CO 2 removal.
In contrast to ECMO, where the need for oxygenation requires high blood flow rates, ECCO 2 R requires much lower blood flow rates, due to the significant differences in CO 2 and oxygen (O 2 ) kinetics. Almost all the O 2 in blood is carried by hemoglobin, which displays sigmoidal saturation kinetics. Assuming normal hemoglobin and venous O 2 content, each liter of venous blood can only carry an extra 40-60 mL of O 2 before the hemoglobin is fully saturated. Blood flows of 5-7 L/min through the extracorporeal artificial membrane lung are therefore required to supply enough O 2 for an average adult. Conversely, most CO 2 in blood is dissolved or in the form of bicarbonate, displaying linear kinetics without saturation. Considering that 1 L of blood is transported around 500 mL of CO 2 , in a perfectly efficient system flow of 0.5 L/min would be enough to remove all of the CO 2 produced by an average adult, which is about 250 mL/min [2,17,18]. Also, CO 2 diffuses more readily than O 2 across extracorporeal membranes because of higher solubility. However, in practice, ECCO 2 R is usually able to remove up to 25% of carbon dioxide production given the limitations of blood flow and membrane efficiency [19]. As the rate of CO 2 clearance greatly depends on the fresh sweep gas flow through the membrane lung, this is usually maximized in the low blood flow ECCO 2 R systems. Therefore, the efficiency of CO 2 clearance of the different available devices is critically determined by other important parameters, including the size of the cannula, the rate of recirculation of blood in the circuit, the efficiency of the pump, the total surface area of the artificial lung, and the cardiac preload [20,21]. Importantly, given the low blood flow through the extracorporeal circuit and the even lower flow achieved when the blood is crossing the large surface area of the artificial gas exchanger, the risk of thrombosis plays also a key role in CO 2 clearance.

VV-ECCO 2 R
In the veno-venous configuration, blood is drawn from a central vein by a draining cannula, using a centrifugal or roller pump to generate flow across the membrane. CO 2 is removed by the effect of the "sweep gas, " and blood is then returned into the venous circulation (Fig. 1a). Single site cannulation is possible using a double-lumen cannula. This approach allows ECCO 2 R through the use of smaller cannulas (15-19F), commonly introduced via the right internal jugular vein. The setup is very similar to renal replacement therapy, and in fact, some systems are trying to combine the two in one [22,23] (NCT02590575). One of the advantages of VV-ECCO 2 R compared to the AV approach is the less invasiveness by avoiding arterial cannulation, and the potential for early mobilization of patients. It is also possible to set up an ECCO 2 R system through cannulation of two central veins, one for drainage and the other for reinfusion (e.g., femoral-femoral configuration).

AV-ECCO 2 R
In the AV-ECCO 2 R configuration, the blood flows from the femoral artery, usually instrumented with percutaneous cannulation, to the contralateral femoral vein, creating a pumpless arterio-venous (AV) bypass, equipped with an artificial gas exchanger across the AV shunt, which allows the "sweep gas" to remove CO 2 (Fig. 1b). This pumpless systems require an arterio-venous pressure gradient ≥ 60 mmHg and a cardiac index > 3 L/min/ m 2 , which is unsuitable for hemodynamically unstable patients [18,24]. Furthermore, cannulation of a major artery can result in distal ischemia [25], although measuring the artery diameter with ultrasound and selecting a cannula that occupies less than 70% of the lumen reduces this risk [26].

Chronic obstructive pulmonary disease
Chronic obstructive pulmonary disease (COPD) is a significant worldwide health burden. Currently, it is the fourth leading cause of death worldwide, and is the only leading cause of death that is rising, and will likely become the third cause of death by 2020 [27,28].
Acute exacerbations of COPD (aeCOPD) constitute a significant cause of morbidity and mortality among these patients. Patients with moderate to severe acute exacerbations develop alveolar hyperinflation that may lead to increased work of breathing, muscle fatigue, and hypercapnia, creating a vicious loop refractory to medical treatment [29][30][31]. The standard respiratory support in this setting is NIV, which however, fails in up to 30% of patients with aeCOPD, prompting intubation and IMV [32][33][34]. In recent meta-analysis and observational studies, it has been reported that the in-hospital mortality of patients with aeCOPD requiring IMV is as high as 25-39% [35][36][37][38]. Patients with COPD requiring IMV develop a considerable reduction in respiratory muscle strength, with higher risk of prolonged weaning and/or failure to wean, compared to other causes of acute hypercapnic respiratory failure. Up to 60% of the ventilatory time in these patients is spent for weaning [39] and is very likely to require a tracheotomy. The prolonged time on IMV results in an increased potential risk of ventilator-induced lung injury, ventilator-associated pneumonia, and ventilator-induced diaphragmatic dysfunction, in addition to the abovementioned complications associated with prolonged sedation and immobilization.

Evidence and clinical trials of ECCO 2 R in aeCOPD to date
One of the first reports on the application of ECCO 2 R to support respiratory function of a COPD patient was published in Pesenti et al. [40]. However, the technique was abandoned due to technical complications.
As the medical community regained interest in ECCO 2 R, investigators began applying the technique to prevent intubation or to assist weaning from IMV in this patients' population. Several studies in both VV and AV configurations were published, including a meta-analysis (Table 1).

ECCO 2 R to avoid IMV
Brederlau et al. [41] described their experience in three patients that failed NIV for severe aeCOPD. They applied a pumpless AV ECCO 2 R device with the goal of avoiding endotracheal intubation. The ECCO 2 R flow ranged between 1.1 and 1.6 L/min, with the sweep gas flow varying from 3 to 10 L/min. Shortly after beginning ECCO 2 R, the PaCO 2 fell significantly, and also the respiratory rate dropped from 38, 45, and 37 breaths/min to 15, 25, and 18 breaths/min, respectively.
Kluge et al. [5] in the same year evaluated the safety and efficacy of using an AV pumpless extracorporeal lung assist in 21 COPD patients who did not respond to NIV and compared them to 21 matched controls treated with IMV. The use of AV ECCO 2 R resulted in the decrease of PaCO 2 after 24 h and obviated the need for IMV in 90% of the experimental arm. Although the experimental group had a shorter hospital length of stay, there was no significant difference in mortality at 28 days (19% with ECCO 2 R vs. 24% without ECCO 2 R) or 6 months (both groups 33%) compared to the control group.
Burki et al. [42] treated 20 hypercapnic COPD patients with VV ECCO 2 R through a 15.5-Fr dual-lumen cannula achieving a mean blood flow of 430 mL/min. Of the 20 patients, seven were at risk of failing NIV, two were difficult to wean from NIV, and 11 had failed liberation from MV. With ECCO 2 R, none of the patients failing NIV required endotracheal intubation, and both patients with difficult weaning from NIV were weaned. However, only three of the 11 IMV patients were liberated successfully. Moreover, significant complications arose in a number of patients: bleeding requiring blood transfusion was reported in three patients, deep vein thrombosis was diagnosed in one patient after removal of the ECCO 2 R catheter, one patient experienced pneumothorax due to catheter insertion, and one died from hemorrhage when the iliac vein was perforated during ECCO 2 R catheter placement.
Del Sorbo et al. [4] compared 25 patients with aeCOPD treated with NIV + VV ECCO 2 R versus 21 historical controls treated with NIV alone with regard to the cumulative incidence of intubation. They reported that ECCO 2 R with a 14-Fr dual-lumen catheter and blood flow rates of 177-333 mL/min not only improved respiratory acidosis but also reduced the need for intubation by 75% (12% vs. 33%; p = 0.047) and significantly reduced the in-hospital mortality (8% vs. 35%; p = 0.035). However, this came with a cost of 52% prevalence of ECCO 2 R-related side effects and led the authors to suggest the end point of future studies should be long-term mortality.
Braune et al. [43] in the ECLAIR study showed that IMV was avoided in 56% of cases treated with VV ECCO 2 R, which was associated with a high incidence of complications. However, in this study, there was an inclusion of patients with relative contraindications to NIV, and there was an unexpectedly high incidence of hypoxemic patients [44].
Finally, Morelli et al. [45] confirmed the efficacy of VV ECCO 2 R (with a flow rate of 250-450 mL/min through a 13-Fr dual-lumen cannula) in reducing the PaCO 2 in a series of 30 patients with acute hypercapnic respiratory failure due to aeCOPD, who refused endotracheal intubation after failing NIV. The duration of ECCO 2 R was 2-16 days, and it was possible to prevent endotracheal intubation in 27 patients.

ECCO 2 R to facilitate weaning from IMV
Cardenas et al. [46] made the first attempt to use modern ECLS components for VV ECCO 2 R in a patient with aeCOPD. They demonstrated a successful reduction in PaCO 2 , minute ventilation, and ventilator pressures.
Burki et al. [42] in a subgroup of 11 patients receiving IMV, ECCO 2 R allowed the weaning from mechanical ventilator in only three patients.
Abrams et al. [3] reported five older patients (age 73 ± 8.7 years) with aeCOPD who failed NIV, requiring IMV. After an average of 16.5 ± 5.9 h of IMV, ECCO 2 R was initiated. By using a dual-lumen cannula (20-23 Fr) with blood flow rates of 1-1.7 Lt/min, with a sweep Roncon-Albuquerque Jr. et al. [47] using a pediatric VV ECMO system (with blood flow rates of 0.9 L/min through a 19 Fr dual-lumen cannula placed in the right jugular vein) in two patients with aeCOPD reported early extubation after 72 h and patient mobilization out of bed at day 6.

Future studies on ECCO 2 R for COPD
More data will be forthcoming on the application of ECCO 2 R in the management of patients with COPD exacerbations from a number of ongoing or planned clinical trials ( Table 2).
Despite the strong physiological rationale, the existing data are not sufficient to support the routine use of ECCO 2 R in patients with aeCOPD, as randomized controlled trials investigating the efficacy of ECCO 2 R in improving important patient centered outcome are lacking, and the intervention is associated with a high rate of complications.
Furthermore, the relevant incidence of ECCO 2 R-related complications considerably affects the choice of the target patient population of randomized controlled trials, and hence their inclusions and exclusion criteria. The application of ECCO 2 R to prevent IMV in aeCOPD patients at high risk of NIV failure has a remarkable potential clinical impact, but exposes a number of patients, who will not require IMV, to the unnecessary risk of ECCO 2 R-related complications. The application of ECCO 2 R in aeCOPD patients intubated after NIV failure to accelerate liberation from IMV exposed patients simultaneously to the complications of two invasive treatments. In both scenarios, given the high mortality rate associated with IMV in this obviously vulnerable patient population, these studies should be powered to demonstrate a mortality benefit.
The development of new ECCO 2 R technology with less associated complications will allow the study of ECCO 2 R also in patients with milder severity of aeCOPD or even in stable COPD patients to prevent the occurrence of exacerbations.

Severe acute asthma
Asthma is an inflammatory disorder of the airways characterized by airway hyperactivity with bronchospasm, mucosal swelling, and mucus production. The standard treatment of severe acute asthma consists of measures to reverse airflow obstruction. β2 agonists and steroids are the mainstays of treatment [12]. Other available adjunct therapies including anticholinergics, magnesium sulfate, methylxanthines, ketamine, and heliox have been utilized with varying results [48].
Despite advances in asthma therapy, asthma mortality has remained stable in recent years. One reason is the occurrence of status asthmaticus, which can be unresponsive to initial treatment and may lead to hypercapnic respiratory failure despite maximal therapy, and in the most severe cases requires IMV.
Approximately, 4% of all patients hospitalized for acute asthma require IMV, which is associated with increased in-hospital mortality compared with patients who do not require mechanical ventilation (7 vs. 0.2%) [49].
Although necessary, mechanical ventilation may aggravate alveolar hyperinflation as it was described above. To prevent these potential detrimental effects, ECCO 2 R has been applied as rescue therapy. ECCO 2 R as an adjunct to IMV for refractory asthma was first reported in 1981 [50]. Subsequently, several case series have been reported (Table 3) [51][52][53][54][55]. In the international Extracorporeal Life Support Organization (ELSO) registry, the use of ECMO for asthma has been reported in 24 adult patients between 1986 and 2006. Hypercapnia, rather than hypoxemia, was the main gas exchange derangement treated with ECMO, suggesting that a less invasive approach, such as low flow ECCO 2 R, could also be suitable in these cases. Indeed, the use of ECCO 2 R in patients with asthmatic exacerbation has been reported, although in a limited number of cases.

Complications
Although ECCO 2 R seems to be effective in improving or mitigating hypercapnic acidosis and possibly in reducing the rate of endotracheal intubation, its use is associated with a range of vascular, hematological, and other complications (Table 4).
Arterial cannulation is associated with higher risk than venous catheterization, with specific complications including distal limb ischemia, compartment syndrome of the lower limb requiring fasciotomy or limb amputation, as devastating consequences [18].
The occurrence of bleeding events is the most frequent complication of ECCO 2 R. The low flow renders systemic anticoagulation mandatory, increasing the risk of significant bleeding including cerebral, gastrointestinal, and nasopharyngeal bleeds. In the studies on ECCO 2 R for COPD to date, the rate of clinically significant hemorrhagic complications ranges between 2% and 50% [56].
Thrombocytopenia is also commonly observed, as well as hemolysis.

100%)
Thirty-four hours after initiating ECCO 2 R, the patient was weaned entirely from NIV, and the cannula could be removed without any complication. On day 4, the patient was discharged from the ICU without the need for supplemental oxygen and 6 days later, discharged from hospital without any impairment Conversely, thrombus formation is higher at lower blood flow rates because of increased exposure time to the membrane lung and circuit. Clots may detach and enter the patient's bloodstream, plugging the membrane or obstructing the cannula if anticoagulation is not achieved.

Conclusion
In the past, ECCO 2 R was a complex technique requiring intensive monitoring and surgical expertise. Due to a high rate of complications, it was avoided by all but few high expertise centers. With newer simplified systems, ECCO 2 R devices can be easily used and can be initiated by most intensivists. However, given the lack of conclusive clinical evidence and the relatively high rate of associated complications, its use should be restricted to investigational applications in specific cohorts of patients.
In summary, minimally invasive ECCO 2 R appears very promising for patients with acute exacerbation of obstructive diseases refractory to conventional treatment, but systematic evaluation is needed to prove its clinical efficacy.