Article Text

Download PDFPDF
Extracorporeal CO2 removal (ECCO2R) in patients with stable COPD with chronic hypercapnia: a proof-of-concept study
  1. Lara Pisani1,
  2. Stefano Nava1,
  3. Emilia Desiderio2,
  4. Mario Polverino2,
  5. Tommaso Tonetti3,
  6. V Marco Ranieri3
  1. 1 Department of Clinical, Integrated and Experimental Medicine (DIMES), Respiratory and Critical Care Unit, S. Orsola-Malpighi Hospital, Alma Mater Studiorum University of Bologna, Bologna, Emilia-Romagna, Italy
  2. 2 Unità di Pneumologia,Dipartimento di Medicina, Scafati, Italy, Mauro Scarlato Hospital, Scafati, Campania, Italy
  3. 3 Dipartimento di Scienze Mediche e Chirurgiche, Anesthesia and Critical Care Medicine,Policlinico di Sant’Orsola, Bologna, Italy, Alma Mater Studiorum University of Bologna, Bologna, Emilia-Romagna, Italy
  1. Correspondence to Professor Stefano Nava, Department of Clinical, Integrated and Experimental Medicine (DIMES), Respiratory and Critical Care Unit, S. Orsola-Malpighi Hospital, Alma Mater Studiorum University of Bologna, Bologna 40126, Italy; stefanava{at}


Domiciliary non-invasive ventilation (NIV) effectively reduces arterial carbon dioxide pressure (PaCO2) in patients with stable hypercapnic chronic obstructive pulmonary disease, but a consistent percentage of them may remain hypercapnic. We hypothesised that extracorporeal CO2 removal (ECCO2R) may lower their PaCO2. Ten patients hypercapnic despite ≥6 months of NIV underwent a 24-hour trial of ECCO2R. Six patients completed the ECCO2R-trial with a PaCO2 drop ranging between 23% and 47%. Time to return to baseline after interruption ranged 48–96 hours. In four patients, mechanical events led to ECCO2R premature interruption, despite a decreased in PaCO2. This time window ‘free’ from hypercapnia might allow to propose the concept of ‘CO2 dialysis’.

  • COPD ÀÜ mechanisms
  • critical care
  • non invasive ventilation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


To compensate for the severe impairment of respiratory mechanics, patients with chronic obstructive pulmonary disease (COPD) decrease tidal volume and increase respiratory rate. The ‘price’ patients pay for this strategy is hypercapnia.1 In stable conditions, renal and metabolic compensatory mechanisms keep pH in the physiological ranges (chronic hypercapnia).2Although, the relationship between CO2 reduction and clinical benefit in terms of patients-related outcome measures is unclear, studies that have shown clinical benefit of non-invasive ventilation (NIV) all aimed at substantial CO2 reduction.3 However, several patients may not tolerate chronic NIV, are not compliant with the therapy or do not respond in terms of arterial carbon dioxide pressure (PaCO2) reduction.4 5

Extracorporeal CO2 removal (ECCO2R) may decrease PaCO2 and increase pH in patients with acute decompensation of COPD that do not respond to NIV thus avoiding invasive ventilation.6 This ‘proof-of-concept’ study set up to examine the hypothesis that ECCO2R may effectively and safely lower PaCO2 also in patients with chronic hypercapnia unresponsive to domiciliary NIV.


Written informed consent was obtained. Patients with COPD in charge of the domiciliary NIV programme of the ‘Policlinico di Sant’Orsola’ (University of Bologna, Italy) and of the ‘Mauro Scarlato’ (Salerno Italy) hospitals were eligible for inclusion if they were older than 40 years of age, had severe COPD, were clinically stable, as assessed with monthly visit in our outpatients clinic, and had a baseline PaCO2 >50 mm Hg with a pH >7.35. ECCO2R was proposed in patients enrolled in an home care NIV programme for at least 6 months, not responding in terms of PaCO2 reduction (ie, <5% reduction relative to the daytime value observed on spontaneous ventilation before initiation of domiciliary NIV). IPAP was set at 19.3±1.7 and EPAP 4.2±0.42 cmH20 and NIV average use was 5.8±1.1 hours/night. Patients were excluded if contraindication to ECCO2R were present6 or if they had a body mass index >30 or confirmed sleep apnea syndrome.

Patients underwent a 24-hour trial of ECCO2R during spontaneous breathing in a high intensity area of the respiratory ward using the Decap Smart (Hemodec, Salerno) and the ProLung (Estor, Pero) systems both equipped with a polypropylene membrane lung (Euroset, 1.35 m2, Medolla). Heparin was administered to maintain the activated partial thromboplastin time ratio to approximately 1.5. The femoral vein was accessed via a double lumen catheter (14 F; JOLINE).6

Arterial blood gases and respiratory rate were recorded at baseline (time 0) and after 1, 3, 6, 12, 18 and 24 hours of ECCO2R and every 6–8 hours after disconnection until daytime values of PaCO2 returned at baseline level. Patients remained on spontaneous breathing until values of PaCO2 returned to the levels at baseline. NIV was hence restored.

Potential adverse events were classified as mechanical and patient related6 and daily assessed for 15 days after the ECCO2R trial. Occurrence of any adverse event during the ECCO2R trial led to treatment interruption and catheter removal.


ECCO2R was implemented in ten patients. Baseline characteristics of patients and ECCO2R settings are shown in table 1.

Table 1

Baseline values

Twenty-four hours of ECCO2R were completed in six patients causing a reduction in PaCO2 ranging between 23% and 47% (figure 1A). Following interruption of ECCO2R, the time required to return to baseline values of PaCO2 ranged between 48 and 96 hours (table 1). In four patients, mechanical related adverse event (circuit clotting, catheter displacement and pump malfunctioning) led to ECCO2R interruption (figure 1B) after 2–23 hours of treatment (table 1). Of note, values of PaCO2 immediately before ECCO2R interruption were 23%–33% lower than PaCO2 at baseline. Values of pH, respiratory rate and HCO3- remained stable during the time course of the trial (figure 2).

Figure 1

Individual levels of PaCO2 during ECCO2R in all patients completing the study (PANEL A) or interrupting the trial for a technical problem (PANEL B). ECCO2R, extracorporeal CO2 removal; PaCO2, arterial carbon dioxide pressure.

Figure 2

Individual values of pH, HCO3-, PaO2/FiO2 and respiratory rate, during ECCO2R in representative patients completing the study (upper part) or interrupting the trial for a technical problem (lower part). ECCO2R, extracorporeal CO2 removal; HCO3-: bicarbonate; PaO2/FiO2, arterial oxygen tension/fractional inspired oxygen.

No patient-related adverse events6 were observed for 15 days following ECCO2R trial.


This study shows that it is possible to safely lower PaCO2 with ECCO2R in stable patients with COPD with chronic hypercapnia refractory to chronic NIV. In the group of patients able to complete the 24 hours treatment, the effect was retained for 48–96 hours after discontinuation of ECCO2R.

The pathophysiological hallmark of COPD is the combination of the impairment of respiratory mechanics with the weakness of the inspiratory muscles. Under these circumstances, the patients reduce alveolar ventilation,1 so that hypercapnia occurs.2 Compensatory mechanisms such as bicarbonate production by body buffers keep pH in the physiological ranges stabilising the clinical manifestations of COPD and leading to chronic hypercapnia.7 In these patients, NIV has been shown to effectively improve outcome preventing acute exacerbation.8 However, chronic NIV may fail to decrease PaCO2.4 5 We, therefore, challenged the hypothesis that ECCO2R, may improve CO2 clearance in patients with chronic hypercapnia unresponsive to NIV. Regardless of the expected reduction of PaCO2 with ECCO2R, we observed that values PaCO2 remained lower than those observed at baseline after a relatively long period of time after ECCO2R was interrupted. Interestingly the patients with higher forced expiratory volume in 1 s (n.3 and 5), returned to PaCO2 baseline values later than those with more limited ventilator capacity, and this suggest, that this subset of patients, may better respond to ECCO2R. Nevertheless our results might allow to propose the concept of ‘CO2 dialysis’ in analogy with renal dialysis, using a ferula as venous access.

The mechanisms underlying this response are not clearly identified. However, it might be speculated that, since in the case of chronic hypercapnia, tissues CO2 store capacities able to stabilise CO2 tension in the circulating blood are saturated,9 the observation that following interruption of extracorporeal support, PaCO2 returned to baseline values in 48–96 hours may suggest that removing CO2 through the extracorporeal circuit may empty parts of the CO2 buffers. This empty space may therefore re-establish the CO2 storing capacity allowing transient normocapnia during ‘unassisted’ breathing.

The proof-of-concept nature of this study is the major limitation. In fact, we focused exclusively to detect a signal whether ECCO2R may provide, a time window ‘free’ from hypercapnia in patients. The study demonstrated however the feasibility of the hypothesis and therefore launch future investigations aimed to assess the ‘dose–response curve’ (how many hours of ECCO2R are needed to provide the longest time window ‘free’ from hypercapnia) and selection of ideal inclusion/exclusion criteria. Further step will be tackle safety, the evidence of clinical benefits that include the reduction of occurrence of episodes of acute decompensation, the improvement of dyspnoea, exercise capacity and health-related quality of life. Finally, further research is needed in this area to get information about the sample size required for a larger trial.

In conclusion, this study shows that in patients with stable hypercapnic COPD not responding to home NIV, ECCO2R lowers PaCO2 that returned to baseline values in >48 hour after suspension. These data might allow to propose the concept of ‘CO2 dialysis’ and support the need of further studies of CO2 dialysis.



  • Contributors LP performing the experimental trial—writing paper. SN study design, performing experimental trial, writing paper. ED performing the experimental trial. MP performing the experimental trial, revising paper. TT statistical analysis. VMR study design, writing paper.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval Review boards approved the protocol.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles