Secondary Logo

Journal Logo

Clinical Critical Care

A Retrospective Observational Case Series of Low-Flow Venovenous Extracorporeal Carbon Dioxide Removal Use in Patients with Respiratory Failure

Moss, Caroline E.; Galtrey, Eleanor J.; Camporota, Luigi; Meadows, Chris; Gillon, Stuart; Ioannou, Nicholas; Barrett, Nicholas A.

Author Information
doi: 10.1097/MAT.0000000000000386
  • Free


Extracorporeal carbon dioxide removal (ECCO2R) uses membrane oxygenators at low blood flows (400–600 ml/minute) to provide partial respiratory support by removing carbon dioxide (CO2) while providing minimal systemic oxygenation. This is enabled by the high solubility of CO2 in blood. The earliest generation of dedicated devices were arteriovenous (AV), however, recent technological improvements have led to the development of pumped venovenous (VV) systems. Percutaneously inserted dual lumen venous cannulae are used to access the circulation and CO2 removal is achieved through high efficiency gas exchange membranes.1–3

The goal of selective CO2 removal is to provide partial respiratory support to allow a reduction in alveolar minute ventilation. In patients with acute respiratory distress syndrome (ARDS), tidal volumes of 6 ml/kg and plateau pressures <28–30 cm H2O4 are known to improve mortality. Furthermore higher ventilator driving pressure is associated with increased mortality in patients with ARDS.5 However, even with lung protective strategies, mortality exceeds 50%.6 One of the barriers to using lung protective ventilation is that reducing tidal ventilation can result in hypercapnoea. There is some evidence suggesting that ECCO2R may facilitate lower volumes, lower plateau pressures, and lower driving pressures by controlling hypercapnoea.7–9

The other patient cohort where ECCO2R is promising is in patients with a severe exacerbation of chronic obstructive pulmonary disease (COPD) leading to hypercapnic respiratory failure. In these patients, the current gold standard therapy is noninvasive ventilation (NIV). However, patients who fail NIV and require invasive mechanical ventilation have a significantly higher mortality.10 Extracorporeal carbon dioxide removal has been demonstrated to provide physiologic benefit for patients with COPD by increasing pH, reducing PaCO2, reducing respiratory rate, and minute ventilation. Extracorporeal carbon dioxide removal has been used to prevent intubation and to facilitate early extubation in patients with acute exacerbations of COPD.1,11 Although the impact of CO2 removal on work of breathing has not been directly studied, ECCO2R is associated with reduced subjective dyspnoea.11 Furthermore, although there is yet to be a randomized controlled trial of ECCO2R in COPD, a recent propensity matched retrospective cohort study demonstrated a reduction in mortality associated with ECCO2R for NIV failure in COPD.12

The aim of the current study was to explore the physiologic changes and outcomes in a cohort of patients commenced on ECCO2R for a variety of indications in a tertiary intensive care unit (ICU).

Materials and Methods

We performed a retrospective, observational, electronic case-note review of all patients commenced on ECCO2R by the attending clinician from August 2013 to February 2015. Cases were identified through a prospectively maintained electronic database. Consent was waived by Guy’s & St Thomas’ NHS Foundation Trust Research Department (approval number: 5013).

Demographic data collected included age, gender, ethnicity, primary admission diagnosis, cause of respiratory failure, and any known comorbidities. Physiologic data collected included ventilator settings, associated organ system support (vasopressor therapy or renal replacement therapy), and specific data related to ECCO2R use. Cannulation method and site, pump settings, and anticoagulation method were recorded. Full blood counts, arterial blood gases, and ventilation settings were gathered just before starting ECCO2R, at 4 hourly intervals for the first 24 hours and at 24 hourly intervals until cessation of ECCO2R.

Complications recorded included significant bleeding (defined as a transfusion requirement of more than 2 units of packed red blood cells or intracranial haemorrhage), haemolysis (plasma-free hemoglobin [PfHb] >0.5 g/L), circuit failure, or cannula thrombosis. The method and duration of ventilatory support and overall patient outcome were also recorded.

All data were analyzed using Graphpad Prism software. We performed a Shapiro–Wilk normality test on all data to assess distribution. We made no assumption of sphericity of data, and the Geisser-Greenhouse correction was used. For parametrically distributed data, a one-way repeated measure ANOVA analysis was performed, followed by Dunnett’s multiple comparison’s test. For non-normally distributed data, a nonparametric repeat measure tool (Friedman test) was used.


Demographic Data

During our study period, 14 patients were identified who received ECCO2R. Demographic data are in Tables 1 and 2. 50% of patients were female and the overall cohort had a median age of 60.5 years (range: 29–78; Table 1). The etiology of respiratory failure is recorded in Table 2. Just over a third of cases (5/14) were commenced on ECCO2R for an exacerbation of COPD, either for intubation avoidance (3/5) or to facilitate early extubation (2/5). In the remaining cases (9/14), ECCO2R was initiated to facilitate lung protective ventilation. The overall hospital survival was 71% (10/14)—4/5 patients with COPD survived to hospital discharge and 6/9 patients with ARDS survived. The differences between survivors and nonsurvivors are shown in Table 3.

Table 1.
Table 1.:
Demographic and General Data for ARDS and COPD Cohorts of Patients
Table 2.
Table 2.:
Etiology of Respiratory Failure for Study Patients
Table 3.
Table 3.:
Differences Between Survivors and Nonsurvivors

All patients were managed with ECCO2R using Hemolung (ALung Inc, Pittsburgh, PA) via a 15.5 FG percutaneously inserted cannula, of which five (36%) were sited in the femoral vein, and nine (64%) were sited in the internal jugular vein. There were no recorded complications of cannulation.

The median duration of ECCO2R was 5 days (range: 4–19 days). The median total duration of ventilatory support during the ICU stay was 1 day (0–5 days) NIV, and 41.5 days (5–66 days) mechanical ventilation.

Physiologic Data

We initially analyzed the data individually for the subpopulations of COPD and non-COPD patients, but the small sample size precluded any meaningful statistical analysis. Across the entire cohort of 14 patients, using the Shapiro–Wilk normality test, both pH and PaCO2 were shown to be normally distributed. There was a statistically significant improvement in pH (p = 0.012) at all time points except 12 hours post commencement (Figure 1A). The overall change in PaCO2 was nonsignificant (p = 0.262), with the only statistically significant change in PaCO2 observed at t = 4 hours (Figure 1B).

Figure 1.
Figure 1.:
Improvement in pH (A) and pCO2 (kPa) (B) during the first 48 hours of ECCO2R commencement (*p < 0.05, **p < 0.001, p < 0.0001). ECCO2R, extracorporeal carbon dioxide removal; ns, nonsignificant.

For the COPD cohort, there was an apparent reduction in respiratory rate and rise in pH in patients commenced on ECCO2R (Figure 2), although this did not reach statistical significance. Two of three patients avoided intubation, and in the two patients in whom tracheal extubation was the goal, one was extubated within 48 hours, the other within 4 days of ECCO2R commencement. Following the commencement of ECCO2R in patients with ARDS (Figure 3), there was a sustained reduction in peak inspiratory pressures, and an increased positive end-expiratory pressure (PEEP) was required to maintain mean airway pressure and oxygenation, resulting in a reduction in driving pressure. The changes seen were not statistically significant.

Figure 2.
Figure 2.:
Change in RR/minute and pH in COPD patients commenced on ECCO2R (did not reach statistical significance). COPD, chronic obstructive pulmonary disease; ECCO2R, extracorporeal carbon dioxide removal; RR, respiratory rate.
Figure 3.
Figure 3.:
Apparent reduction in driving pressure (plateau pressure-PEEP) in ARDS patients following commencement of ECCO2R. ARDS, acute respiratory distress syndrome; ECCO2R, extracorporeal carbon dioxide removal; PEEP, positive end-expiratory pressure.

Hemoglobin and platelet count were shown to be normally distributed, whereas bilirubin was not. There was a statistically significant reduction in both Hb (p = 0.006) and platelet count (p = 0.0004) in the first 72 hours of ECCO2R therapy (Figure 4A, B). Bilirubin did not change significantly during the first 72 hours (p = 0.065; Figure 4C).

Figure 4.
Figure 4.:
Reduction in Hb (g/L) (A), platelet count (×109/L) (B) and bilirubin (μmol/L) (C) during the first 72 hours of ECCO2R therapy (*p < 0.05, **p < 0.001, p < 0.0001). ECCO2R, extracorporeal carbon dioxide removal; ns, nonsignificant.

There were four circuit-related complications recorded. Two circuit changes were undertaken due to falling membrane CO2 clearance, and there were also two episodes of hemolysis that resolved on discontinuation of ECCO2R with no further sequelae. One episode of hemolysis presented within 24 hours as frank hematuria, the other presented at day 6 following ECCO2R commencement. There were no cases of significant bleeding, air entrainment, or cannula thrombosis identified.


There were a total of 14 cases of venovenous ECCO2R for respiratory failure (five COPD, nine ARDS). The approach demonstrated allowed for a statistically significant improvement in pH for both patient cohorts. In addition, there was an apparent reduction in peak and driving pressure for the ARDS group and in respiratory rate for the COPD group, although neither change reached statistical significance. Other than transient hemolysis in two patients, no harm was recorded and there was no lasting adverse effect of therapy. Survival to hospital discharge was excellent at 71%, compared with expected mortality for this population of patients from the literature being 67.5%.13

Although there is good evidence supporting a strategy of reduced plateau pressure and tidal volume to improve mortality,9 the mortality for moderate to severe ARDS remains high.6,14 Furthermore, there is some evidence suggesting that tidal volumes of under 6 ml/kg ideal body weight with a plateau pressure less than 28 cm H2O are beneficial.15 Another important element appears to be the driving pressure (peak pressure-PEEP), which is associated with mortality.5

Recent studies have successfully used ECCO2R to facilitate lung protective ventilation.8,9 Our results are in keeping with this study; although statistical significance was not reached, we demonstrated a trend toward reduced driving pressures in our cohort of patients with ARDS using low blood flow venovenous ECCO2R (median blood flow of 0.5 L/minute). When titrating ECCO2R, sweep gas flow was titrated against CO2 clearance to achieve a pH > 7.35. In our series, there was a median of 10 days of mechanical ventilation before the commencement of ECCO2R. There is growing evidence that delays in commencing extracorporeal membrane oxygenation (ECMO) beyond 6 days is a prognostic indicator for mortality.16 The same data are not available for ECCO2R, however it would seem reasonable that the same relationship applies. It is possible that our cohort may have benefited from earlier application of the therapy. However, due to the lack of evidence supporting the technique, clinicians ensured that all other options were explored first. Given the size of the cohort, it is difficult to make meaningful interpretations of mortality, however it is in keeping with results from other studies.6,14 Finally, it is not currently known how long patients should remain on ECCO2R to receive benefit from the therapy; in this case, clinical benefit was deemed to have taken a median of 5 days, although the range was up to 19 days. This is important for future trials considering the correct “duration” of ECCO2R in the ARDS population.

In the COPD cohort, treatment goals were achieved for 80% patients. The only patient who did not reach their treatment goal was a patient with an influenza A-induced exacerbation of COPD for whom intubation was avoided for 6 days of therapy; however, hemolysis occurred leading to cessation of the device and the patient was subsequently intubated. Ventilatory strategy was not uniform for all COPD patients before commencement of ECCO2R: two patients were already intubated at commencement of therapy, two patients failed NIV, and one patient did not tolerate NIV. Patients were managed with a median blood flow of 0.44 L/minute and for a median of 5 days with the sweep gas flow rate titrated to arterial CO2 for a pH > 7.35. As demonstrated in Figure 2, the respiratory rate and hence work of breathing reduced rapidly following institution of ECCO2R, associated with a reduction in PaCO2 and improvement in pH. This is in keeping with other studies using both AV and VV ECCO2R.2,11 The precise physiologic impact of ECCO2R on the respiratory system is not known, but it is likely that a reduced respiratory rate allows more effective expiration, which in turn allows unloading of the muscles of respiration. Reduced respiratory rate also presumably contributes to reduced patient exhaustion. Therapy was well tolerated by patients and facilitated an oral diet, communication, and mobilization out of bed. These were achieved through the removal of NIV and tracheal extubation. For patients with this severity of COPD, who have failed NIV, the ICU length of stay was relatively short at a median of 7 days.17 It is difficult to make meaningful comments on mortality in this cohort given the size of the study, however it is in keeping with mortality reported in other studies of ECCO2R in severe COPD exacerbation.

In two of the patients who died, although ECCO2R was commenced to facilitate extubation for one case and prevent intubation in the other, once respiratory distress was eased ECCO2R facilitated autonomous decision-making by the patient which resulted in palliative care being instituted at the patient’s request. The use of extracorporeal support allowed the patients to spend quality time with their families before their deaths. It is known that the provision of good end of life care on the patient has a significant impact on relatives, reducing anxiety, stress, and depression.18,19

There is little data available on the impact of pumped VV ECCO2R systems on hematology. However, the hematological effects of ECMO are better understood and include an acquired thrombocytopenia, reduction in a variety of coagulation factors, and potentially an acquired von Willebrand’s factor deficiency.20 These changes are believed to occur as a result of both adsorption of proteins and blood cellular components onto the artificial surface, proinflammatory effects of the circuit, and physical trauma to the blood caused by passage through the pump and membrane lung.21–23 In the current series, there was a progressive thrombocytopenia (defined as a >50% reduction from baseline levels in keeping with diagnostic criteria for other conditions24) in keeping with previous ECMO studies and with previous animal studies using the same device.25 There was no bleeding associated with this degree of thrombocytopenia despite the simultaneous presence of moderate anticoagulation with heparin. A mild anaemia was also noted; this may be due to dilution from the circuit priming volume (approximately 400 ml) and did not appear to be related to haemolysis. Hemolysis was only found in two cases and no change in bilirubin was identified (unfortunately pfHb was not measured consistently in all patients as a measure of hemolysis). Again this appears to be in keeping with other publications using this device25 as well as the general ICU literature where anemia is routinely reported.26 There were no complications of cannulation. Acute cannula thrombosis has occurred in another hospital (personal communication), but did not occur in this series. It is possible that the reported case of thrombosis was a rare event, or that the administration of heparinised saline flush at the time of cannulation prevented this from occurring. No cannula site infections were recorded despite cannulae being in place for a median of 5 days, a maximum of 19 days, and a third of cannulae being placed in femoral vessels. Cannula infection is relatively common in the ICU literature,27 and it is unclear whether the low rate of infection relates to the approach taken for cannulation or to ongoing cannula care.

The current study has significant limitations. It is an uncontrolled, small study with retrospective data collection and therefore has inherent bias, including selection bias and missing information. The study is small preventing meaningful statistical comparisons for much of the data and the clinical diagnoses show significant variability. Nevertheless the cohort is valuable as it is one of the first reports of ECCO2R outcomes in clinical practice.


Our retrospective, observational series of ECCO2R shows that this is a promising technique for the management of patients with severe exacerbations of COPD and for achieving lung protective ventilation in patients with ARDS without refractory hypoxaemia. Venovenous ECCO2R can provide improved physiologic control of gas exchange with minimal complications.


1. Burki NK, Mani RK, Herth FJ, et al.: A novel extracorporeal CO(2) removal system: Results of a pilot study of hypercapnic respiratory failure in patients with COPD. Chest 2013.143: 678686.
2. Cole S, Barrett N, Glover G, et al.: Extracorporeal carbon dioxide removal as an alternative to endotracheal intubation for non-invasive ventilation failure in acute exacerbation of COPD. JICS 2014.15: 14.
3. MacLaren G, Combes A, Bartlett RH: Contemporary extracorporeal membrane oxygenation for adult respiratory failure: Life support in the new era. Intensive Care Med 2012.38: 210220.
4. Brower RG, Matthay MA, Morris A, et al.: The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome.N Engl J Med, 2000.342: 13011308.
5. Amato MB, Meade MO, Slutsky AS, et al.: Driving pressure and survival in the acute respiratory distress syndrome. N Engl J Med 2015.372: 747755.
6. Villar J, Blanco J, Añón JM, et al.; ALIEN Network: The ALIEN study: Incidence and outcome of acute respiratory distress syndrome in the era of lung protective ventilation. Intensive Care Med 2011.37: 19321941.
7. Pesenti A, Patroniti N, Fumagalli R: Carbon dioxide dialysis will save the lung. Crit Care Med 2010.38(10 suppl): S549S554.
8. Terragni PP, Del Sorbo L, Mascia L, et al. Tidal volume lower than 6 ml/kg enhances lung protection: Role of extracorporeal carbon dioxide removal. Anesthesiology 2009.111: 826835.
9. Bein T, Weber-Carstens S, Goldmann A, et al. Lower tidal volume strategy (approximately 3 ml/kg) combined with extracorporeal CO2 removal versus ‘conventional’ protective ventilation (6 ml/kg) in severe ARDS: The prospective randomized Xtravent-study. Intensive Care Med 2013.39: 847856.
10. Stefan MS, Nathanson BH, Higgins TL, et al. Comparative effectiveness of noninvasive and invasive ventilation in critically ill patients with acute exacerbation of chronic obstructive pulmonary disease. Crit Care Med 2015.43: 1386394.
11. Kluge S, Braune SA, Engel M, et al.: Avoiding invasive mechanical ventilation by extracorporeal carbon dioxide removal in patients failing noninvasive ventilation. Intensive Care Med 2012.38: 16321639.
12. Del Sorbo L, Pisani L, Filippini C, et al. Extracorporeal CO2 removal in hypercapnic patients at risk of noninvasive ventilation failure: A matched cohort study with historical control. Crit Care Med 2015.43: 120127.
13. Goldhill DR, Sumner A: Outcome of intensive care patients in a group of British intensive care units. Crit Care Med 1998.26: 13371345.
14. Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory distress syndrome: The Berlin Definition. JAMA 2012.307: 25262533.
15. Hager DN, Krishnan JA, Hayden DL, Brower RG. Tidal volume reduction in patients with acute lung injury when plateau pressures are not high. Am J Respir Crit Care Med 2005.172: 12411245.
16. Schmidt M, Zogheib E, Rozé H, et al.: The PRESERVE mortality risk score and analysis of long-term outcomes after extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. Intensive Care Med 2013.39: 17041713.
17. Braune S, Burchardi H, Engel M, et al.: The use of extracorporeal carbon dioxide removal to avoid intubation in patients failing non-invasive ventilation–a cost analysis. BMC Anesthesiol 2015.15: 160.
18. Hetherington G, Freeman R: End of life care strategy: Second Annual Report. UK Department of Health, 2010. Available at: t/uploads/system/uploads/attachment_data/file/213816/dh_118955.pdf.
19. Detering KM, Hancock AD, Reade MC, Silvester W: The impact of advance care planning on end of life care in elderly patients: Randomised controlled trial. BMJ 2010.340: c1345.
20. Murphy DA, Hockings LE, Andrews RK, et al.: Extracorporeal membrane oxygenation-hemostatic complications. Transfus Med Rev 2015.29: 90101.
21. Oliver WC: Anticoagulation and coagulation management for ECMO. Semin Cardiothorac Vasc Anesth 2009.13: 154175.
22. Annich GM. Extracorporeal life support: the precarious balance of hemostasis. J Thromb Haemost. 2015;13(suppl 1):S336S342.
23. Kalbhenn J, Wittau N, Schmutz A, Zieger B, Schmidt R: Identification of acquired coagulation disorders and effects of target-controlled coagulation factor substitution on the incidence and severity of spontaneous intracranial bleeding during veno-venous ECMO therapy. Perfusion 2015.30: 675682.
24. Greinacher A: Heparin-Induced Thrombocytopenia. N Engl J Med 2015.373: 18831884.
25. Wearden PD, Federspiel WJ, Morley SW, et al.: Respiratory dialysis with an active-mixing extracorporeal carbon dioxide removal system in a chronic sheep study. Intensive Care Med 2012.38: 17051711.
26. Astin R, Puthucheary Z: Anaemia secondary to critical illness: an unexplained phenomenon. Extrem Physiol Med 2014.3: 4.
27. Miller SE, Maragakis LL: Central line-associated bloodstream infection prevention. Curr Opin Infect Dis 2012.25: 412422.

venovenous extracorporeal carbon dioxide removal; respiratory failure

Copyright © 2016 by the American Society for Artificial Internal Organs