NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE

Transcription

1 NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE INTERVENTIONAL PROCEDURES PROGRAMME Interventional procedure overview of extracorporeal membrane carbon dioxide removal Using a blood-filtering machine to remove excess carbon dioxide from the blood. Extracorporeal membrane carbon dioxide removal is used to treat respiratory failure (when the lungs do not work effectively) in critically ill patients. The aim is to remove excess carbon dioxide from the blood. The patient will still need to receive oxygen by mechanical ventilation. Blood is drawn from the circulation and passed through a synthetic membrane, where carbon dioxide is removed, before it is returned to the circulation. Introduction The National Institute for Health and Clinical Excellence (NICE) has prepared this overview to help members of the Interventional Procedures Advisory Committee (IPAC) make recommendations about the safety and efficacy of an interventional procedure. It is based on a rapid review of the medical literature and specialist opinion. It should not be regarded as a definitive assessment of the procedure. Date prepared This overview was prepared in September Procedure name Extracorporeal membrane carbon dioxide removal (ECCO 2 R). Specialty societies Association of Anaesthetists of Great Britain and Ireland Intensive Care Society Faculty of Intensive Care Medicine Royal College of Anaesthetists. IP overview: extracorporeal membrane carbon dioxide removal Page 1 of 32

2 Description Indications and current treatment Extracorporeal membrane carbon dioxide removal (ECCO 2 R) is a treatment for patients with severe acute respiratory failure usually used together with mechanical ventilation. Severe acute respiratory failure can result from a number of different disease processes. A common type is associated with acute respiratory distress syndrome (ARDS), which may be caused by sepsis, pneumonia or chest trauma. Patients with severe acute respiratory failure are conventionally treated by mechanical ventilation. However in some patients gas exchange is not good enough at maximal tolerable ventilation pressures. If this is the case, extracorporeal membrane gas exchange systems can be used. They include extracorporeal membrane oxygenation (ECMO) and ECCO 2 R. ECMO systems use a high rate of extracorporeal flow (using a large fraction of cardiac output) to achieve almost complete oxygenation as well as carbon dioxide removal. ECMO systems therefore always require a pump. ECCO 2 R systems remove CO 2 from the blood, but oxygenation is limited. Because only a small percentage of cardiac output is treated, complete pulmonary support is not possible, unlike ECMO. In arteriovenous ECCO 2 R blood flow is maintained by the patient s own blood pressure rather than a pump. ECCO 2 R is associated with a risk of poor limb perfusion. There may also be inadequate flow in patients with low cardiac output. For these reasons, venovenous ECCO 2 R systems have been developed that use a low-flow pump. In some patients, ventilation parameters such as inspired oxygen concentration, tidal volume, and peak and end-expiratory pressures can be reduced after starting ECCO 2 R. This contributes to a reduced risk of ventilation-induced lung injury. Patients may be treated with ECCO 2 R support for up to several weeks, depending on clinical need. What the procedure involves There are two main types of ECCO 2 R: arteriovenous and venovenous. In arteriovenous ECCO 2 R, an artery and vein are cannulated (typically the femoral artery and femoral vein). Imaging may be used to assess the diameter of the femoral artery. The cannulae are connected to a low-resistance synthetic membrane device. Arterial blood flows continuously through the device, where gas exchange occurs, and is returned via the vein. In venovenous ECCO 2 R, cannulation consists either of a single access with a double lumen catheter or a dual access system using two venous catheters, connected to a venovenous circuit driven by a low-flow pump. In both systems, a continuous infusion of heparin is given to reduce the risk of thrombus formation. Extracorporeal flow is monitored by an ultrasound device IP overview: extracorporeal membrane carbon dioxide removal Page 2 of 32

3 and can be modified by clamping the cannulae or by adjusting the pump speed. Arterial blood gases, cannulation sites and lower-limb perfusion are monitored regularly. Because ECCO 2 R can achieve only limited oxygenation, it is used with mechanical ventilation for patients with both hypercapnia and hypoxia. In some cases ventilator settings can be reduced after starting ECCO 2 R. Patients may be treated with ECCO 2 R support for periods up to several weeks, depending on clinical need. Literature review Rapid review of literature The medical literature was searched to identify studies and reviews relevant to extracorporeal membrane carbon dioxide removal. Searches were conducted of the following databases, covering the period from their commencement to 2 August 2011: MEDLINE, PREMEDLINE, EMBASE, Cochrane Library and other databases. Trial registries and the Internet were also searched. No language restriction was applied to the searches (see appendix C for details of search strategy). Relevant published studies identified during consultation or resolution that are published after this date may also be considered for inclusion. The following selection criteria (table 1) were applied to the abstracts identified by the literature search. Where selection criteria could not be determined from the abstracts the full paper was retrieved. Table 1 Inclusion criteria for identification of relevant studies Characteristic Publication type Patient Intervention/test Outcome Language Criteria Clinical studies were included. Emphasis was placed on identifying good quality studies. Abstracts were excluded where no clinical outcomes were reported, or where the paper was a review, editorial, or a laboratory or animal study. Conference abstracts were also excluded because of the difficulty of appraising study methodology, unless they reported specific adverse events that were not available in the published literature. Patients with respiratory failure. Extracorporeal membrane carbon dioxide removal. Articles were retrieved if the abstract contained information relevant to the safety and/or efficacy. Non-English-language articles were excluded unless they were thought to add substantively to the English-language evidence base. IP overview: extracorporeal membrane carbon dioxide removal Page 3 of 32

4 List of studies included in the overview This overview is based on approximately 518 patients from 1 randomised controlled trial, 1 review and meta-analysis of 8 case series (2 of which have been described separately in more detail) and 3 additional case series 1 7. Other studies that were considered to be relevant to the procedure but were not included in the main extraction table (table 2) have been listed in appendix A. IP overview: extracorporeal membrane carbon dioxide removal Page 4 of 32

5 Table 2 Summary of key efficacy and safety findings on extracorporeal membrane carbon dioxide removal Abbreviations used: ARDS, acute respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Morris AH (1994) 1 Randomised controlled trial USA Recruitment period: Number of patients analysed: 40 (21 ECCO 2R vs 19 control) Survival rate (at 30 days): ECCO 2R = 33.3% (7/21) (1 patient died before ECCO 2R could be started and 2 patients died within 1 day of starting it.) Control group = 42.1% (8/19), p = 0.56 Major complications : Cardiac dysrhythmia arrest: ECCO 2R = 9.5% (2/21) Control = 10.5% (2/19) Cardiac tamponade: ECCO 2R =4.8% (1/21) Control = 0% (0/19) There were no statistically significant differences in total hospital Intracranial haemorrhage: Study population: patients with length of stay, intensive care unit length of stay or clinical trial time ECCO 2R = 4.8% (1/21) severe ARDS between the 2 groups. Control = 5.3% (1/19) Cerebral arterial gas embolism: n = 40 (21 low-frequency positivepressure ventilation and ECCO 2R ECCO 2R = 58.6 ± 0.3 Control = 5.3% (1/19) Mean PaO 2 (mmhg): ECCO 2R = 0% (0/21) vs 19 continuous positive-pressure Control group = 59.3 ± 0.3 Cerebral hypoxia depression: ventilation alone) ECCO 2R =4.8% (1/21) Mean arterial ph: Mean age: 35 years Control = 5.3% (1/19) ECCO 2R = 7.39 Sex: 43% (17/40) male Extremity ischaemia: Control group = 7.36 ECCO 2R = 0% (0/21) Patient selection criteria: PaO 2 < 0.2, Control = 10.5% (2/19) There were no clinically important differences in blood gas mean bilateral chest radiographic infiltrates, Arterial embolism: values between the 2 groups. total thoracic compliance less than ECCO 2R = 0% (0/21) 50 ml/cmh 2O and no clinical Control = 5.3% (1/19) evidence of heart failure; patients who met ECMO entry and exclusion criteria. Exclusion criteria included age < 12 and > 65 years, and mechanical ventilation > 21 days. Venous thrombosis: ECCO 2R = 4.8% (1/21) Control = 10.5% (2/19) Intrapulmonic haemorrhage: ECCO 2R = 19.1% (4/21) Control = 0% (0/19) Technique: Arterial oxygenation was achieved primarily through the patient s natural lung. Packed red blood cell transfusion > 0.8 l /day: ECCO 2R = 47.6% (10/21) Control = 0% (0/19) Follow-up: 30 days ECCO 2R circuit clotting: 19.1% (4/21) Conflict of interest/source of funding: ECCO 2R had to be discontinued equipment was donated from a in 7 patients because of number of different manufacturers. haemorrhage. Study design issues: Patients were stratified by age and by the presence or absence of trauma. Blinded randomisation with blocking. A sample size of 60 was calculated to detect a survival difference between 9 and 40% (power = 0.80). The trial was discontinued before the sample size was reached because it was deemed that the difference between the 2 therapies was too small for a significant survival difference to be demonstrated with 60 patients. Analysis was by intention to treat. Study population issues: Of the 21 patients assigned to ECCO 2R, 1 patient died before it could be started and another improved without it. There were no statistically significant differences between the 2 patient groups at randomisation. Other issues: The authors noted that the overall survival rate (38%) was higher than expected; they expected a survival rate of 9% in the control group. IP overview: extracorporeal membrane carbon dioxide removal Page 5 of 32

6 Abbreviations used: ARDS, acute respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Walles T (2007) 2 Number of patients analysed: 225 Complications Total = 29% Review and meta-analysis Mean duration of ECCO 2R ranged from 5 to 16 days Arterial thrombus formation = 2.2% (5/225) Germany Success rate (not further defined) = 56% (range 41 83) Venous thrombus formation = 4.9% (11/225) Recruitment period: not reported Oxygenator thrombus formation = 2.7% (6/225) Study population: critically ill patients Limb ischaemia = 9.3% with ventilation-refractory lung failure (21/225) Bleeding at cannulation site n = 225 (8 case series studies, n = 3.6% (8/225) ranged from 5 to 90) Infection = 0.4% (1/225) Plasma leakage = 4.4% (10/225) Age: not reported Sex: not reported Patient selection criteria: not reported Technique: pumpless arteriovenous ECCO 2R using interventional lungassist (ila) device (Novalung, Germany). Follow-up: not reported Conflict of interest/source of funding: not reported This study was included in the original overview for IP408. Study design issues: All included studies were retrospective. Statistical methods not described. Study population issues: The author states that there may have been some overlap between patients in the identified studies. Other issues: The authors noted that cannula thrombosis was not observed following implementation of modified cannulae after The problem of plasma leakage disappeared with a new oxygenator membrane in The authors noted that the following are contraindications to treatment: heart failure, septic shock with low mean arterial pressure, severe arterial occlusive disease and history/presence of heparin-induced thrombocytopenia type II. Unfavourable prognostic factors include acute renal failure, high vasopressor demand, long preceding duration of artificial ventilation, high age, obesity and cancer. IP overview: extracorporeal membrane carbon dioxide removal Page 6 of 32

7 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Bein T (2006) 3 Germany Recruitment period: Study population: patients with ARDS n = 90 Median age: 44 years Sex: 77% (69/90) male Patient selection criteria: patients with severe respiratory failure considered to be at risk for acute life-threatening hypoxaemia and/or excessive hypercapnia, when other optimised therapies, including mechanical ventilation, have failed. Failure of alternative treatment was defined as insufficient oxygenation (PaO 2 < 55 mmhg at FiO 2 = 1) and/or when hypercapnia had induced arterial acidosis and haemodynamic instability occurred or distinct deterioration was observed. Contraindication: haemodynamic depression of cardiac origin. Technique: pumpless arteriovenous ECCO 2R using interventional lungassist (ila) device (Novalung, Germany). Follow-up: Respiratory variables reported for first 24 hours, plus survival in hospital Number of patients analysed: 90 Survival 59% (53/90) of patients died in hospital; duration of survival and whether patients died while on ECCO 2R was not stated. Causes of death: septic shock or persistent systemic inflammatory response syndrome (49%), multiple organ failure (25%), cardiac failure (19%), cerebral injury (7%) Mean duration of ECCO 2R support = 5 days (SD 5) Changes in arterial blood gases Median (IQR) Time on ECCO 2R Pre 2 hours 24 hours CO 2 removal (ml/min) (85 211) ( ) PaCO 2 (mmhg) 60 (48 80) 36 (30 44)* 34 (30 39)* PaO 2/FiO 2 ratio (mmhg) 58 (47 78) 82 (64 103)* 101 (74 142)* *p < 0.05 vs baseline Change in mechanical ventilation Median Time on ECCO 2R (IQR) Pre 2 hours 24 hours FiO 2 Minute ventilation (l/min) Tidal volume (ml) Respiratory frequency (breaths per min) Peak inspiratory pressure (cm H 2O) 1.0 ( ) 13.0 ( ) 430 ( ) 27 (21 43) 38 (35 40) 0.9 ( )* 11.0 ( )* 410 ( )* 25 (20 40) 36 (32 39) 0.8 ( )* 9.9 ( )* 380 ( )* 23 (17 39)* 35 (31 39)* IP overview: extracorporeal membrane carbon dioxide removal Page 7 of 32 Complications: 24% (22/90) of patients experienced a serious complication. Intracerebral haemorrhage = 1% (1/90) Ischaemia of lower limb = 10% (9/90) (This led to amputation in one patient. In the remaining patients, normal perfusion returned after removal of the cannulae). Compartment syndrome of lower limb = 4% (4/90) Haematoma/aneurysm at cannulation site = 1% (1/90) Haemolysis = 1% (1/90) Diffuse bleeding/shock syndrome during cannulatio% (1/90) This study was included in the original overview for IP408 and is also included in the review above (ref. 2). Study design issues: Retrospective study Study population issues: The authors stated that the patients included were the most severely critically ill, and in retrospect concluded that some were too ill and haemodynamically unstable to benefit from ECCO 2R. They suggest that ECCO 2R should not be used for patients with cancer or refractory shock, or in geriatric patients, very obese patients or those developing multiple organ dysfunction syndrome, who were found to benefit least from the procedure.

8 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Conflict of Interest: The authors declared no financial arrangements with Novalung GmBH PEEP (cmh 2O) *p < (12 17) 15 (13 18) 14 (12 18) IP overview: extracorporeal membrane carbon dioxide removal Page 8 of 32

9 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Flörchinger B (2008) 4 Number of patients analysed: 159 Complications Percutaneous femoral Mean ECCO 2R support interval = 7 days (range 0 33) cannula placement was unsuccessful in 6 patients Germany Successful weaning from ECCO 2R: Overall = 52.2% (83/159) who required an open surgical exposure and Recruitment period: insertion. Study population: patients with severe respiratory insufficiency 59 Mean age: 44 years (range 7 78) Sex: 76% (121/159) male Patient selection criteria: inclusion criteria were: severe hypoxia with a PaO 2/FiO 2 ratio less than 80 mmhg or hypercapnia with PaCO 2 greater than 70 mmhg (despite aggressive mechanical ventilation therapy). An attempt at conservative mechanical ventilator treatment had to be performed before ECCO 2R was initiated. Technique: pumpless arteriovenous ECCO2R using interventional lungassist (ila) device (Novalung, Germany). Follow-up: not reported Conflict of interest/source of funding: not reported. By underlying cause of respiratory insufficiency: ARDS and no trauma = 47.4% (18/38) ARDS after trauma = 64.9% (24/37) ARDS after surgery = 41.4% (12/29) ARDS after chemotherapy = 25.0% (2/8) Bacterial pneumonia = 63.6% (21/33) Viral pneumonia = 25.0% (1/4) Other pneumonia (aspiration) = 62.5% (5/8) Overall survival rate = 34.6% (55/159) 47.2% (75/159) of patients died during ECCO 2R (main causes of death were multi-organ failure, septic shock and low cardiac output syndrome). Variable Death during ECCO 2R Death after ECCO 2R Survivor p value Patients (n) Age (years) 45 ± ± ± Body mass 27.7 ± 27.1 ± 25.4 ± 0.02 index (kg/m 2 ) Support (days) 4.8 ± ± ± Continuous venovenous haemofiltration before ECCO 2R 44 (59%) 10 (34%) 13 (24%) Mechanical ventilation before ECCO 2R 6.6 ± ± ± Lower limb ischaemia = 8.2% (13/159) (In these patients, the arterial cannula was either exchanged with a smaller one or moved to the contralateral femoral vessel. Compartment syndrome necessitating fasciotomy = 2.5% (4/159) (1 patient required lower leg amputation). Thrombus formation = 17.0% (27/159) (required oxygenator exchange; thrombosis of the entire system developed in 8 patients 4 were inadequately anticoagulated, 2 had heparin induced thrombocytopenia type II and there were 2 device failures). Plasma leakage = 4.4% (7/159). Study design issues: Consecutive patients. Study population issues: 70% (112/159) of patients had acute respiratory distress syndrome. In 2 patients, ECCO 2R was used as a bridge to lung transplantation. Other issues: The technology used for ECCO 2R evolved over the 10-year study period. The authors also noted that there was an institutional learning curve with regard to efficacy of ECCO 2R in terms of oxygenation and carbon dioxide removal. Indications for the procedure have changed over the study period. There are some discrepancies between results summarised in the abstract and those presented in the main body of the paper. The percentage of patients dying during ECCO 2R was quoted as 48.7% rather than 47.2% (the numerator is stated to be 75). IP overview: extracorporeal membrane carbon dioxide removal Page 9 of 32

10 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments (days) Gas exchange Time on ECCO 2R Pre 2 hours 24 hours Before ECCO 2R termination FiO * PaO 2 (mmhg) * PaO 2/FiO (mmhg) PaCO * 39 (mmhg) Arterial saturation of oxygen (%) Venous saturation of oxygen *p = Change in mechanical ventilation Time on ECCO 2R Pre 2 hours 24 hours Before ECCO 2R termination Respiration rate (breaths/min) Peak inspiratory pressure (cm H 2O) Minute ventilation (l/min) IP overview: extracorporeal membrane carbon dioxide removal Page 10 of 32

11 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Liebold A (2002) 5 Germany Recruitment period: not stated Study population: patients with severe pulmonary failure n = 70 Mean age: 41 years (range 8 72) Sex: 79% (55/70) male Patient selection criteria: severest forms of ARDS ; conventional ventilator therapy failing and patient would die without ECCO 2R (physicians consensus); mean arterial pressure > 70 mmhg and cardiac output > 6 l/min. Patients with septic shock were only excluded if haemodynamic requirements could not be met. Patients with cardiac failure were excluded. Technique: Pumpless arteriovenous ECCO 2R using the Nova Breath Lung Assist Device System (Jostra, Germany). Follow-up: Respiratory variables reported for first 24 hours, plus survival in hospital Conflict of interest: not stated Number of patients analysed: 70 Survival 51% (36/70) of patients were weaned; 49% (34/70) died while on ECCO 2R. Of those who were weaned, 11 (16% of all patients) died later from unresolved organ failure; 25 (36% of all patients) were eventually discharged from hospital. Overall 64% (45/70) of patients died. Mean duration of ECCO 2R (days) = 6 (range 1 35) Change in arterial blood gases Mean (SD) Time on ECCO 2R Pre 24 hours PaCO 2 (mmhg) 59 (17) Not stated PaO 2/FiO 2 ratio 110 (SD not 50 (SD not stated) (mmhg) stated) PaO 2 (mmhg) 46 (7) 85 (21) p values not stated Change in mechanical ventilation Median (IQR) Time on ECCO 2R Pre 24 hours FiO (0.1) p values not stated Technical problems occurred in 21.4% (15/70) patients. Thrombus formation Arterial cannula = 7.1% (5/70) Venous cannula = 2.9% (2/70) Membrane gas-exchange device = 1.4% (1/70) Cannulae had to be changed for two patients. The patient with thrombosis in the gas-exchange device was later diagnosed with heparin-induced thrombocytopenia. Ischaemia of lower limb = 4.3% (3/70) (requiring cannula removal) Haemolysis = 0% (0/70) Bleeding = 0% (0/70) Plasma leakage from gasexchange device = 7.1% (5/70), device was replaced; 1 patient received 3 modules The authors stated that none of the deaths was directly attributable to complications with arteriovenous ECCO 2R. This study was included in the original overview for IP408 and is also included in the review above (ref. 2). Study design issues: Consecutive patients Other issues: The authors concluded that ECCO 2R was suitable for patients with ARDS arising from sepsis, as long as cardiac output was adequate. IP overview: extracorporeal membrane carbon dioxide removal Page 11 of 32

12 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Zimmermann M (2009) 6 Number of patients analysed: 51 Complications Ischaemia of lower limb = Survival rate = 50.9% (26/51) Non-survivors were statistically significantly older than survivors 5.9% (3/51) (removal of the cannula resulted in Germany Recruitment period: but there were no differences between the groups for severity of disease and severity of lung injury. normalisation of distal perfusion) Cannula thrombosis = 1.9% Study population: patients with ARDS (1/51) (mostly due to pneumonia, trauma or Bleeding during cannulation sepsis) = 1.9% (1/51) n = 51 Compartment syndrome (limb) = 1.9% (1/51) (required surgical intervention) Median age: 52 years Sex: 84% (43/51) male Patient selection criteria: persisting impairment in pulmonary gas exchange after a stabilisation period of hours (PaO 2/FiO mmhg with PEEP of 10 cmh 2O or more and/or arterial ph < 7.25 because of respiratory acidosis). Contraindications included cardiac insufficiency, severe peripheral vascular disease, and the need for continuous high doses of vasoactive or inotropic agents. Technique: pumpless arteriovenous ECCO 2R using interventional lungassist (ila) device (Novalung, Germany). Reduced diameter cannulae were used, inserted with the aid of ultrasound. Follow-up: to hospital discharge Conflict of interest/source of funding: 1 author received lecture honorary from manufacturer. Changes in gas exchange, cardiovascular and respiratory variables before and during ECCO 2R Median (IQR) Time on ECCO 2R Pre 2 hours 24 hours PaO 2/FiO * 110* (62 130) (70 127) (86 160) PaCO 2 (mmhg) 73 44** 41** (61 86) (36 54) (34 48) Arterial ph ** 7.44** ( ) ( ) ( ) Mean arterial pressure (mmhg) 73 (65 80) 83** (75 91) FiO 2 1 ( ) Minute 11.5 ventilation (l/min) ( ) Tidal volume 6.6 (ml/kg) ( ) Respiratory rate 25 (breaths per min) (22 27) Plateau pressure 35 (cmh 2O) (31 38) PEEP (cmh 2O) 17 (14 20) * p < 0.05 in comparison with pre-ecco 2R ** p < 0.01 in comparison with pre-ecco 2R 0.8** ( ) 8.6** ( ) 5.0** ( ) 23 (20 30) 34 (30 37) 15* (11 19) 81 (76 90) 0.7** ( ) 6.6** ( ) 4.4** ( ) 21 (18 26) 30** (26 34) 17 (14 20) Study design issues: Prospective study. Other issues: In more severe cases of hypoxaemia, a pump-driven venovenous extracorporeal membrane oxygenation (ECMO) was preferably initiated. The authors note that the data in this study resulted from a strict algorithm with defined indications and contraindications, in contrast to their previous study (Bein et al. 2006). The procedure was withdrawn from life-threatening rescue-situations towards the support of lung-protective ventilation in acute lung injury or early acute respiratory distress syndrome. The authors note there is ongoing technical evolution of smaller cannulae, more efficient gas exchange membranes and easy system handling. IP overview: extracorporeal membrane carbon dioxide removal Page 12 of 32

13 Abbreviations used: ARDS, adult respiratory distress syndrome; CO 2, carbon dioxide; ECCO 2R, extracorporeal carbon dioxide removal; ila, interventional lung-assist; IQR, interquartile range; FiO 2, inspired fraction of oxygen; PaCO 2, partial pressure of CO 2 in arterial blood; PaO 2, partial pressure of oxygen in arterial blood; PEEP, positive end expiratory pressure; SD, standard deviation Study details Key efficacy findings Key safety findings Comments Gattinoni L (1986) 7 Number of patients analysed: 43 The only ECCO 2R-related adverse event was bleeding. An Survival rate = 48.8% (21/43) (the lowest survival rate was in average of 1800 ± 850 ml/d of patients with post-traumatic acute respiratory failure [22%]) whole blood was transfused. Italy Recruitment period: Study population: patients with acute respiratory failure n = 43 Mean age: 26 years (range 2 56) Sex: 42% (18/43) male Patient selection criteria: PaO 2 less than 50 mmhg for more than 2 hours when measured at FiO 2 of 1.0 at PEEP of 5 cm H 2O or greater; after 48 hours of maximal therapy, a PaO 2 less than 50 mmhg for more than 12 hours when measured at FiO 2 of 0.6 or greater, a PEEP of 5 cm H 2O or greater, with a right to left shunt greater than 30% of cardiac output; total static lung compliance lower than 30 ml. Exclusion criteria included pulmonary capillary wedge pressure greater than 25 mmhg, chronic systemic disease, severe chronic pulmonary disease, terminal cancer, and major burns. Technique: venovenous bypass was used. Follow-up: not reported Conflict of interest/source of funding: extracorporeal apparatus was supplied by Kontron, Italy. The severity of acute respiratory failure immediately before ECCO 2R was comparable in survivors and non-survivors. Improved lung function, if any, was always seen within the first 48 hours after starting ECCO 2R. During the first 6 hours of the procedure, cardiac index and mean pulmonary artery pressure decreased significantly (p < 0.05) towards normal values. Irrespective of ultimate survival, 72.1% (31/43) of patients had improved lung function during ECCO 2R (improved PaO 2, right to left shunt, total static lung compliance, and chest X-ray films). There was no statistically significant relationship between survival rate and age. In 3 patients, intrapulmonary bleeding was the terminal event. Minor bleeding, often requiring a surgical revision every 2 to 3 days, occurred at the cannulation sites. Major bleeding occurred from chest tubes, or during pulmonary surgery while on bypass (usually to treat a bronchopleural fistula; 3 patients, 1 survivor). On 6 occasions, the membrane lungs were replaced because of either reduced efficiency in gas exchange and/or the presence of disseminated intravascular coagulation. Study design issues: Prospective study. Other issues: The authors noted that this procedure was only ever offered as a last resort, after conventional continuous positive-pressure ventilation and other nonconventional forms of treatment, such as inverted-ratio ventilation or high-frequency jet ventilation, had failed to improve gas exchange. IP overview: extracorporeal membrane carbon dioxide removal Page 13 of 32

14 Efficacy Survival A randomised controlled trial of 40 patients treated by low-frequency positivepressure ventilation ECCO 2 R or continuous positive-pressure ventilation reported 30-day survival rates of 33% (7/21) and 42% (8/19), respectively (p = 0.56) 1. A case series of 159 patients reported that 52% (83/159) of patients were successfully weaned from ECCO 2 R and 35% (55/159) of patients survived 4. A case series of 51 patients reported a survival rate of 51% (26/51) 6. A case series of 43 patients treated by venovenous ECCO 2 R reported a survival rate of 49% (21/43) 7. Change in concentration of blood gases Carbon dioxide concentration Three case series reported a statistically significant reduction in the partial pressure of carbon dioxide in arterial blood (PaCO 2 ) within 24 hours of initiating ECCO 2 R support, compared with baseline 3,4,6. In the first case series of 90 patients, PaCO 2 decreased from a median of 60 mmhg to 34 mmhg at 24 hours (p < 0.05) 3. In the second case series of 159 patients, PaCO 2 decreased from 67 mmhg to 35 mmhg at 24 hours (p = 0.001) 4. In the case series of 51 patients, PaCO 2 decreased from 73 mmhg at baseline to 41 mmhg at 24 hours (p < 0.01) 6. Oxygen concentration (PaO 2 /FiO 2 ratio) A case series of 90 patients reported a statistically significant increase in PaO 2 /FiO 2 ratio from 58 mmhg at baseline to 101 mmhg at 24 hours (p < 0.05) 3. A case series of 51 patients reported a statistically significant increase from 75 mmhg at baseline to 110 mmhg at 24 hours (p < 0.05) 6. Two further case series of 159 and 70 patients reported increases from 72 to 111 mmhg and 50 to 110 mmhg, respectively (p values not stated) 4,5. Change in mechanical ventilation settings Minute ventilation (MV) Three case series reported a reduction in MV during the first 24 hours of ECCO 2 R. In the case series of 90 patients, median MV decreased from 13.0 litres/min at baseline to 9.9 litres/min at 24 hours 3 (p < 0.05). In the case series of 159 patients, mean MV decreased from 13.8 litres/min to 11.6 litres/min (p value not stated) 7. In the case series of 51 patients, median MV decreased from 11.5 litres/min to 6.6 litres/min (p < 0.01) 6. IP overview: extracorporeal membrane carbon dioxide removal Page 14 of 32

15 Respiratory frequency The case series of 90 patients reported that median respiratory frequency decreased from 27 breaths per minute at baseline to 23 breaths per minute at 24 hours (p < 0.05) 3. The case series of 159 patients reported a decrease from 32 breaths per minute at baseline to 29 breaths per minute at 24 hours (p value not stated) 4. The case series of 51 patients reported a decrease from 25 breaths per minute to 21 breaths per minute at 24 hours (p value not stated) 6. Safety Limb ischaemia and compartment syndrome In a review of 8 case series including a total of 225 patients, limb ischaemia was reported in 9% (21/225) of patients 2. In one of the case series included in this review, 1 out of 90 patients had ischaemia of the lower limb that led to amputation; 4% (4/90) of patients developed compartment syndrome of a lower limb 3. In a case series of 159 patients, lower limb ischaemia was reported in 8% (13/159) of patients and compartment syndrome necessitating fasciotomy was reported in 3% (4/159) of patients, 1 of whom required lower leg amputation 4. In a case series of 51 patients, ischaemia of a lower limb was reported in 6% (3/51) of patients (removal of the cannula restored distal perfusion); 1 patient developed compartment syndrome requiring surgical intervention 6. Thrombus formation In the review of 8 case series including a total of 225 patients, arterial thrombus formation was reported in 2% (5/225) of patients, venous thrombus formation in 5% (11/225) of patients and oxygenator thrombus formation was reported in 3% (6/225) of patients 2. Bleeding complications In the review of 8 case series including a total of 225 patients, bleeding at the cannulation site was reported in 4% (8/225) of patients 2. Plasma leakage from the gas-exchange device In the review of 8 case series including a total of 225 patients, plasma leakage was reported in 4% (10/225) of patients 2. In the case series of 159 patients, plasma leakage was reported in 4% (7/159) of patients 4. IP overview: extracorporeal membrane carbon dioxide removal Page 15 of 32

16 Validity and generalisability of the studies The review of 8 case series only includes studies that used the interventional lung-assist (ila) Membrane Ventilator device (Novalung ) 2. Most of the evidence included in table 2 comes from Germany. The studies include patients with respiratory failure resulting from various causes. The evidence presented relates largely to arteriovenous systems; one study used venovenous ECCO 2 R 7. There are 9 additional studies on venovenous ECCO 2 R in the appendix (mainly case reports). There have been technological developments since the procedure was first used, including the use of smaller cannulae and improved gas exchange membranes. Existing assessments of this procedure An Ontario Health Technology Assessment on extracorporeal lung support technologies was published in April It included 7 case series studies on Interventional Lung Assist, all of which are included in this overview (either in table 2 or appendix A). The report recommended that any approval for bridge to transplantation or bridge to recovery in adults for ILA or ECMO should be conditional on evidence development, since there is insufficient evidence that either technology improves survival rates. Given the fact that there is moderate quality evidence that these technologies improve intermediate outcomes, from a social values perspective and in terms of biological plausibility, controlled funding should be considered as there are no alternative technologies for these patients. Related NICE guidance Below is a list of NICE guidance related to this procedure. Appendix B gives details of the recommendations made in each piece of guidance listed. Interventional procedures extracorporeal membrane carbon dioxide removal. NICE interventional procedures guidance 250 (2008). This guidance is currently under review and is expected to be updated in For more information, see Extracorporeal membrane oxygenation for severe acute respiratory failure in adults. NICE interventional procedures guidance 391 (2011). Available from Extracorporeal membrane oxygenation (ECMO) in postneonatal children. NICE interventional procedures guidance 38 (2004). Available from IP overview: extracorporeal membrane carbon dioxide removal Page 16 of 32

17 Specialist Advisers opinions Specialist advice was sought from consultants who have been nominated or ratified by their Specialist Society or Royal College. The advice received is their individual opinion and does not represent the view of the society. Andrew Bodenham (Association of Anaesthetists of Great Britain and Ireland); David Hunter (Royal College of Anaesthetists); Andrew Rhodes, Carl Waldmann, Bob Winter, Duncan Wyncoll (Intensive Care Society). Five Specialist Advisers had performed the procedure at least once; one had never performed it. Five Advisers considered the procedure to be novel and of uncertain safety and efficacy; one Adviser stated that arteriovenous ECCO 2 R was no longer new but pumped venovenous ECCO 2 R had not yet been used in the UK. One Adviser stated that there is no clear comparator because it is usually considered to be a rescue therapy. Three Advisers stated that the comparator would be conventional ventilation with lung protective strategy and one stated that the comparator would be ECMO. Anecdotal adverse events include vascular damage, bleeding (including intracerebral haemorrhage), thrombosis and infection. Additional theoretical adverse events include gas embolism, haemolysis and heparin-induced thrombocytopenia. One Adviser noted that they had seen 1 death and several complications requiring major surgery that were related to poor insertion techniques for cannulation of the femoral vessels. Reported adverse events include leg ischaemia, cannula thrombosis, bleeding and compartment syndrome. Key efficacy outcomes include survival (one Adviser specified good quality survival at 6 months and 1 year), length of mechanical ventilation, length of stay in intensive care unit, length of hospital stay, reduction in ventilated minute volume and airway pressures, carbon dioxide removal (mls per minute), and avoidance of mechanical ventilation. Several Advisers noted that there was uncertainty about the efficacy of the procedure. One Adviser stated that efficacy of a lung protective ventilatory strategy was not fully proven in acute lung injury. Another Adviser stated that there was no proof to date that this technology or ECMO actually improve survival in adults with acute lung injury requiring supportive assisted ventilation. One Adviser stated that full training in ultrasound insertion of large bore vascular access devices is required. Further training in the use of X-ray screening is required if bi-caval venous cannulae need inserting. Vascular backup is required, and cardiac backup is required for the insertion of bi-caval devices. Four Specialist Advisers thought that the procedure was likely to have a minor impact on the NHS, in terms of resources and numbers of patients; two Advisers thought that the impact would be moderate. IP overview: extracorporeal membrane carbon dioxide removal Page 17 of 32

18 Patient Commentators opinions NICE s Patient and Public Involvement Programme was unable to gather patient commentary for this procedure. Issues for consideration by IPAC A German prospective randomised multicentre study was completed in January 2011 (start date September 2007), with an estimated enrolment of 120 patients (NCT ). The principal investigator has stated that the data are currently with the statistician and he expects the paper to be ready for review by December IP overview: extracorporeal membrane carbon dioxide removal Page 18 of 32

19 References 1. Morris AH, Wallace CJ, Menlove RL et al. (1994) Randomized clinical trial of pressure-controlled inverse ratio ventilation and extracorporeal CO 2 removal for adult respiratory distress syndrome. American Journal of Respiratory and Critical Care Medicine 149: Walles T. (2007) Clinical experience with the ila Membrane Ventilator pumpless extracorporeal lung-assist device. Expert Review of Medical Devices 4: Bein T, Weber F, Philipp A et al. (2006) A new pumpless extracorporeal interventional lung assist in critical hypoxemia/hypercapnia. Critical Care Medicine 34: Flörchinger B, Philipp A, Klose A et al. (2008) Pumpless extracorporeal lung assist: a 10-year institutional experience. Annals of Thoracic Surgery 86: Liebold A, Philipp A, Kaiser M et al. (2002) Pumpless extracorporeal lung assist using an arterio-venous shunt. Applications and limitations. Minerva Anestesiologica 68: Zimmermann M, Bein T, Arlt M et al. (2009) Pumpless extracorporeal interventional lung assist in patients with acute respiratory distress syndrome: a prospective pilot study. Critical Care 13: R Gattinoni L, Pesenti A, Mascheroni D et al. (1986) Low-frequency positivepressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 256: Medical Advisory Secretariat. (2010) Extracorporeal lung support technologies - bridge to recovery and bridge to lung transplantation in adult patients: an evidence-based analysis. Ontario Health Technology Assessment Series 10: IP overview: extracorporeal membrane carbon dioxide removal Page 19 of 32

20 Appendix A: Additional papers on extracorporeal membrane carbon dioxide removal The following table outlines the studies that are considered potentially relevant to the overview but were not included in the main data extraction table (table 2). It is by no means an exhaustive list of potentially relevant studies. Venovenous system IP overview: extracorporeal membrane carbon dioxide removal Page 20 of 32

21 Article Number of patients/ follow-up Direction of conclusions Reasons for non-inclusion in table 2 Borg UR, Reynolds HN, Habashi NM. (1998) Veno-venous extracorporeal lung assist with concurrent distal aortic perfusion: repair of ruptured aorta in a patient with dense ARDS. International Journal of Artificial Organs 21: Venovenous ECCO 2R allowed surgical repair of a ruptured descending thoracic aorta to be performed in a patient with profound respiratory failure. Brunet F, Mira JP, Belghith M et al. (1992) Effects of aprotinin on hemorrhagic complications in ARDS patients during prolonged extracorporeal CO2 removal. Intensive Care Medicine 18: n = 2 Venovenous After aprotinin infusion combined with heparin, bleeding vanished until the end of bypass. Cardenas VJ Jr, Lynch JE, Ates R et al. (2009) Venovenous carbon dioxide removal in chronic obstructive pulmonary disease: experience in one patient. ASAIO Journal 55: Venovenous First case using venovenous ECCO 2R resulted in reduction in minute ventilation to 30% of baseline with improved arterial blood gases, reduction in peak airway pressures and improvement in hyperinflation. Everts P, Franssen T, Schipper R et al. (1989) Venovenous long term extracorporeal CO2 removal with biopump and hollow fiber membrane oxygenator for the failing lung. Journal of Extra-Corporeal Technology 21: Venovenous The patient died because of irreversible traumatic cerebral damage and ECCO 2R was terminated after 60 hours. Knoch M, Kollen B, Dietrich G et al. (1992) Progress in veno-venous longterm bypass techniques for the treatment of ARDS. Controlled clinical trial with the heparin-coated bypass circuit. International Journal of Artificial Organs 15: RCT (comparing heparincoated circuit with non-heparin coated circuit) Venovenous There were fewer bleeding complications and more patients survived in the group with a heparin-coated circuit. 8 Moscatelli A, Ottonello G, Nahum L et al. (2010) Non-invasive ventilation and low-flow veno-venous extracorporeal carbon dioxide removal as a bridge to lung transplantation in a child with refractory hypercapnic respiratory failure due to bronchiolitis obliterans. Pediatric Critical Care Medicine 11: e8 12. Venovenous Non-invasive mechanical ventilation and venovenous ECCO 2R were efficacious in managing refractory hypercapnic respiratory failure in a pediatric patient awaiting lung transplantation. IP overview: extracorporeal membrane carbon dioxide removal Page 21 of 32

22 Article Nelson J, Cairns B, Charles A. (2009) Early extracorporeal life support as rescue therapy for severe acute respiratory distress syndrome after inhalation injury. Journal of Burn Care & Research 30: Ruberto F, Pugliese F, D'Alio A et al. (2009) Extracorporeal removal CO2 using a venovenous, low-flow system (Decapsmart) in a lung transplanted patient: a case report. Transplantation Proceedings 41: Ryan DP, Doody DP. (1992) Treatment of acute pulmonary failure with extracorporeal support: 100% survival in a pediatric population. Journal of Pediatric Surgery 27: Number of patients/ follow-up n = 2 (ECCO 2R) Direction of conclusions Venovenous By using rest ventilator settings and venovenous ECCO 2R, the patient remained stable for a total of 6.5 days and was then successfully transitioned to a conventional ventilator and decannulated. Venovenous No adverse events occurred. From baseline to 48 hours, ph values increased and partial pressure of CO2 reduced. At the same time ventilatory support was reduced. Venovenous Both children survived and were successfully weaned from the ventilator. Reasons for non-inclusion in table 2 or pumpless system (including studies where the type of system has not been specified) Bartosik W, Egan JJ, Wood AE. (2011) The Novalung interventional lung assist as bridge to lung transplantation for self-ventilating patients - initial experience. Interactive Cardiovascular & Thoracic Surgery 13: Beed M, Jayamaha J, Sherman R et al. (2006) Successful use of portable extracorporeal carbon-dioxide removal in a patient with uncontrollable hypercapnoea. British Journal of Intensive Care 16: Bein T, Scherer MN, Philipp A et al. (2005) Pumpless extracorporeal lung assist (PECLA) in patients with acute respiratory distress syndrome and severe brain injury. Journal of Trauma 58: n = 2 n = 5 Self-ventilating patients awaiting lung transplantation. 1 patient was successfully transplanted after 140 days. The other was weaned off ECCO 2R after a short period of time. Successful outcome for patient with pneumonia. Hypercapnia was eliminated and minute volume of artificial ventilation could be reduced. 4 out of 5 patients survived showing a good neurologic function. (included in table 2 of previous overview) Bein T, Osborn E, Hofmann HS et al. (2010) Successful treatment of a severely injured soldier from Afghanistan with pumpless extracorporeal lung assist and neurally adjusted ventilatory support. International Journal of Emergency Medicine 3: ECCO 2R enabled safe transportation and lung protective ventilation. IP overview: extracorporeal membrane carbon dioxide removal Page 22 of 32

23 Article Number of patients/ follow-up Direction of conclusions Reasons for non-inclusion in table 2 Bein T, Zimmermann M, Philipp A et al. (2011) Addition of acetylsalicylic acid to heparin for anticoagulation management during pumpless extracorporeal lung assist. ASAIO Journal 57: Nonrandomised comparative study (comparing 2 anticoagulant regimens) Supplementation of low-dose acetylsalicylic acid is safe and might preserve the function of oxygen transfer. n = 30 Bombino M, Patroniti N, Foti G et al. (2011) Bronchopleural fistulae and pulmonary ossification in posttraumatic acute respiratory distress syndrome: successful treatment with extracorporeal support. ASAIO Journal 57: Lung rest, achieved by ECCO 2R, allowed weaning from mechanical ventilation, closure of bronchopleural fistula, and resumption of spontaneous breathing. Brunet F, Belghith M, Mira JP et al. (1993) Extracorporeal carbon dioxide removal and low-frequency positivepressure ventilation. Improvement in arterial oxygenation with reduction of risk of pulmonary barotrauma in patients with adult respiratory distress syndrome. Chest 104: n = 23 Survival rate = 50% Bleeding was the only complication related to the procedure and was the cause of death in 4 patients. Brunet F, Mira JP, Belghith M et al. (1994) Extracorporeal carbon dioxide removal technique improves oxygenation without causing overinflation. American Journal of Respiratory & Critical Care Medicine 149: ECCO 2R with low frequency positive pressure ventilation improves gas exchange without causing lung overinflation in a majority of patients with adult respiratory distress syndrome. Conrad SA, Zwischenberger JB, Grier LR et al. (2001) Total extracorporeal arteriovenous carbon dioxide removal in acute respiratory failure: a phase I clinical study. Intensive Care Medicine 27: n = 8 Applied in conjunction with mechanical ventilation and permissive hypercapnia, ECCO 2R resulted in normalisation of arterial PCO 2 and ph and permitted significant reductions in the level of mechanical ventilation. IP overview: extracorporeal membrane carbon dioxide removal Page 23 of 32