Extracorporeal carbon dioxide removal (ECCO2R) is a low-flow extracorporeal lung support (ECLS) technique targeting the removal of CO2 in patients suffering from respiratory failure. Use of ECCO2R prevents injurious mechanical ventilation and ventilator-induced lung injury (VILI), reduces the work of breathing of spontaneously breathing patients, and mitigate respiratory acidosis.1,2 Extracorporeal carbon dioxide removal has been successfully applied in association with low-frequency positive pressure ventilation for the management of patients suffering from acute respiratory distress syndrome (ARDS).3 Until recently, ECCO2R has been used only by major clinical centers in the most complicated patients because of technical challenges, complications, and costs. In the last few years, thanks to technological advances (e.g., high-efficiency polymethylpentene hollow fiber membranes, extracorporeal surface coating, and percutaneous cannulation), new devices with less invasive form factor have been developed to carry out ECCO2R in particular for partial lung support. As a result, applications of ECCO2R are spreading to other forms of respiratory compromise, such as exacerbation of chronic obstructive pulmonary disease,4 refractory status asthmaticus,5 and lung transplant.6
Although respiratory failure is often isolated, it can frequently develop in the context of multiple-organ dysfunction syndrome (MODS).7 Indeed, most of the patients treated with mechanical ventilation or ECLS succumb secondary to multiple-organ failure rather than to compromised gas exchange.8 The development of modular devices capable of providing various forms of extracorporeal blood treatment (e.g., cytokines removal coupled with plasma filtration and absorption or blood purification with renal-replacement therapy) in conjunction with ECCO2R lung support may be beneficial.9,10
Recently, devices combining ECCO2R and renal-replacement technologies have been successfully used for the management of patients suffering from acute respiratory failure.11 The uniqueness of these new devices is the ultrafiltrate recirculation to the inflow of the membrane lung (ML) obtained by a peristaltic pump and their potential to permit modular therapy for patients requiring both lung and renal support.
Until now, effects of ultrafiltrate recirculation on ML CO2 removal per se have not been evaluated. Answering this question can provide useful information for the development of new multiple-organ support devices. We expect ML CO2 removal capabilities not to be altered by ultrafiltrate recirculation. We decided to study the effects of ultrafiltrate recirculation on CO2 removal capabilities of a commercially available ECCO2R device (Hemolung, ALung; Alung Technologies, Pittsburgh, PA) in spontaneously breathing sheep.
This study was approved by the U.S. Army Institute of Surgical Research Institutional Animal Care and Use Committee and was conducted in compliance with the Animal Welfare Act and the Implementing Animal Welfare Regulations and in accordance with the principles of the Guide for the Care and Use of Laboratory Animals.
Three healthy ewes (47 ± 10 kg) were endotracheally intubated after an intramuscular injection of glycopyrrolate (100 μg/kg) and tiletamine-zolazepam (6 mg/kg). During preparation, anesthesia was maintained with isofluorane (expired % 1.0–1.5) and buprenorphine intramuscular injection (10 μg/kg). Antibiotic prophylaxis (ceftriaxone 2 g) was administered before the procedure. Tracheostomy was performed. The left carotid artery was percutaneously cannulated with a 16 G central venous catheter (Arrow International, Reading, PA) for pressure monitoring and blood gas analysis. Two 8.5 Fr sheath introducers (Arrow International) were placed bilaterally into the jugular veins. The left line was kept in situ and used both for infusion of medications and the placement of a Swan-Ganz catheter (Edwards Lifescience, Irvine, CA), which was used to measure pulmonary and wedge pressures, cardiac output, and core temperature and to allow sampling of mixed venous blood. After administering a bolus of unfractionated heparin (100 IU/kg), we used the right introducer as a guide for insertion of a 15.5 Fr dual-lumen femoral catheter (Hemolung, Alung; Alung Technologies). The catheter was then connected to the extracorporeal circuit. Afterwards, heparin was infused continuously to maintain activated clotting time (ACT) level at approximately 200% of the baseline. A Foley catheter was placed into the bladder to measure urinary output.
The custom-made extracorporeal circuit used was composed of a Hemolung device (ALung; Alung Technologies) set to maintain a constant blood flow (BF) of 250 ml/min and a fixed gas flow (GF) of 10 L/min of ambient air. In addition, a standard polyethersulfune continuous renal-replacement therapy (CRRT) hemofilter (Purema; NxStage, NxStage Medical, Lawrence, MA) was connected in series after the Hemolung ML. An ultrafiltrate flow (UF) was generated from the hemofilter and recirculated before the ML with a peristaltic pump. Five withdrawal sites were arranged in the circuit: four on the blood side (inlet, pre-ML, post-ML, and outlet) and one on the ultrafiltrate side. Extracorporeal BF was measured by the Hemolung flow meter at the circuitry outlet (Figure 1). Total priming volume of the circuitry was circa 450 ml, as the sum of the priming volume of the Hemolung circuitry (260 ml), the hemofilter (170 ml), and the ultrafiltrate supplementary circuitry (30 ml).
After the surgical procedure, inhalation anesthetic was discontinued. Subsequently, throughout the experiment, analgesia was maintained by intramuscular injections of buprenorphine (10 μg/kg) every 6 hr. Subjects were moved from the surgical table and placed in a metabolic cage, where they were connected to a mechanical ventilator. The sheep were allowed to recover and to breathe spontaneously, with the continuous positive airway pressure set at a fraction of inspired oxygen (FiO2) of 21% and a positive end-expiratory pressure of 5 cm of H2O. During the experiment, animals were awake and fed hay at libitum. Maintenance fluids (Plasma-Lyte A; Baxter International, Deerfield, IL) 2 ml/kg/hr were provided until the end of the experiment.
We evaluated five extracorporeal circuit settings as follows: a baseline step with a GF at 0 L/min and a UF at 0 L/min and four steps with a GF at 10 L/min and four different UF settings (0, 50, 100, and 150 ml/min) (Table 1). As described, extracorporeal BF (measured at the outlet of the extracorporeal circuitry) was fixed at 250 ml/min during these changes in UF. In other words, the actual BF exiting from and returning to the animal was kept constant at 250 ml/min by changing accordingly the rotation of the Hemolung device. Consequently, the flow passing through the Hemolung ML was the sum of the extracorporeal BF and the recirculating UF: 250, 300, 350, and 400 ml/min during UF at 0, 50, 100, and 150 ml/min, respectively.
The baseline step was always performed first, and the remaining steps were randomized. Each step lasted 25 min. A complete block of the five different steps was repeated eight times per animal, totaling 24 repetitions.
Arterial, extracorporeal circuit blood, and ultrafiltrate samples were collected at the end of each step. At the end of baseline steps, mixed venous blood samples were collected to perform mixed venous blood gas analyses and measure hematocrit (Hct) through microhematocrit centrifuge technique. Membrane lung CO2 removal (VCO2ML) was measured by the Hemolung built-in capnometer. The ML CO2 removal efficiency ratio was computed as the ratio between VCO2ML and total CO2 content in the pre-ML blood sample, as previously described.12 Total blood CO2 content was calculated according to a simplified and standardized Henderson–Hesselbach equation, which is commonly used and approved for blood gas analyzers calculations,13 as the sum of the bicarbonate ions concentration and dissolved CO2 obtained by the pCO2 and CO2 solubility coefficient (αCO2 = 0.03 mMol × L−1 × mm Hg−1). Therefore, total CO2 content is calculated as follows:
Hence, ML CO2 removal efficiency can be calculated as follows:
Extracorporeal circuit pressures were monitored in each of the withdrawal site; ultrafiltrate pressure was measured before and after the peristaltic pump. Trans-ML pressure was computed as the difference between pressure at the post-ML site and the pre-ML site. Trans-ML resistance was computed as the ratio between Trans-ML pressure and flow passing through the ML.
Heart rate and arterial and pulmonary artery pressure were continuously monitored. At the end of the baseline step of each repetition, cardiac output was measured using thermodilution technique. A capnograph (CO2SMO; Novametrix, Wallingford, CT) was used to measure respiratory rate, minute ventilation, and natural lung CO2 removal (VCO2NL). Plasma-free hemoglobin (Hb) concentration was measured during the instrumentation (before connection to ECCO2R circuit) and at the end of the experiment by spectrophotometric analysis. At the conclusion of the experiments, the sheep were euthanized by an intravenous injection of 20 ml Fatal Plus (Vortech Pharmaceuticals, Dearborn, MI).
Data are expressed as mean ± standard deviation. The SigmaPlot 12 statistical program (Systat Software Inc., Chicago, IL) was used for statistical analysis. Data were compared with one-way analysis of variance for repeated measurements with a Tukey’s post hoc correction. The Shapiro–Wilk test was used to test normality. A p value less than 0.05 was considered statistically significant.
Physiological parameters and arterial blood gas data collected during baseline steps (BF = 250 ml/min, GF = 0 L/min, and UF = 0 ml/min) were not statistically different throughout the study (Table 2). Application of ultrafiltrate recirculation did not alter VCO2ML compared with no recirculation. Indeed, VCO2ML was 40.5 ± 4.0, 39.7 ± 4.2, 39.8 ± 4.2, and 39.2 ± 4.1 ml/min, respectively, at UFs 0, 50, 100, and 150 ml/min (p = 0.37) (Figure 2). Rising levels of ultrafiltrate recirculation determined progressive reductions of ML CO2 removal efficiency ratio, from 30% during no UR to 27%, 25%, and 21%, respectively, during UFs 50, 100, and 150 ml/min (Figure 3). Minute ventilation, respiratory rate, and VCO2NL were not influenced by application of ultrafiltrate recirculation (p > 0.45) (Table 3). Furthermore, arterial blood gas analyses were unchanged by the application of ultrafiltrate recirculation (p > 0.35) (Table 4).
The pH, pCO2, and HCO3− levels of the blood and ultrafiltrate samples withdrawn from the circuit are shown in Table 5. Inlet blood values remained constant during the different steps of the experiment. Application of ultrafiltrate recirculation was associated with statistically significant changes only in pre-ML blood gas analyses. Indeed, a reduction in pCO2 and HCO3− and a concomitant rise in the pH levels were observed during ultrafiltrate recirculation. Pre-ML pH during UF 50, 100, and 150 ml/min was higher (p < 0.04) than that during without recirculation. Pre-ML pCO2 was lower at 50, 100, and 150 ml/min compared with that at no UF (p < 0.001); moreover, pCO2 at UF 100 and 150 ml/min was lower than that at UF 50 ml/min (p < 0.03). Pre-ML bicarbonate ion concentration was lower at 50, 100, and 150 ml/min compared with that at no recirculation (p < 0.001) and at UF 100 ml/min compared with 50 ml/min (p < 0.02). In contrast, pH, pCO2, and HCO3− levels in post-ML, outlet, and ultrafiltrate samples were not affected by ultrafiltrate recirculation application.
As expected, regardless of the UF, post-ML and outlet pH were higher (p < 0.001) and pCO2 lower (p < 0.001) relative to pre-ML blood. Ultrafiltrate had a pH similar to post-ML (p > 0.05), lower than outlet (p < 0.009), and higher than inlet and pre-ML at each UF (p < 0.001). Independent of ultrafiltrate recirculation, ultrafiltrate pCO2 was higher than post-ML and at the outlet (p < 0.001) but less than at the inlet (p < 0.001) (Table 5).
Increasing levels of ultrafiltrate recirculation caused progressive rise in the trans-ML pressure, from 110 ± 12 mm Hg at UF 0 ml/min to 115 ± 11, 119 ± 11, and 120 ± 10 mm Hg at 50, 100, and 150 ml/min, respectively. Trans-ML resistance was significantly reduced by recirculation of ultrafiltrate. Indeed, trans-ML resistance was 0.441 ± 0.053, 0.383 ± 0.041, 0.354 ± 0.078, and 0.308 ± 0.057 mm Hg/ml × min with UF = 0, 50, 100, and 150 ml/min, respectively. In effect, trans-ML resistance at UF of 0 ml/min was significantly lower than at UF of 100 and 150 ml/min (p < 0.05) as well as at UF of 50 ml/min was significantly lower than UF of 150 ml/min (p < 0.05). Interposing of the hemofilter in the circuitry resulted in a downstream pressure after the ML equal to 45 ± 16, 49 ± 16, 50 ± 13, and 54 ± 13 mm Hg at UF 0, 50, 100, and 150 ml/min, respectively.
Plasma-free Hb was 14.5 ± 5.9 mg/dl before connection to the ECCO2R circuitry and 10.0 ± 2.2 mg/dl at the end of the experiment, with no statistically significant difference between the two. Unfractionated heparin requirements were 22.1 ± 17.0 IU/kg × hr.
No adverse effects, such as bleeding or thromboembolic episodes, and circuitry malfunction were observed.
To the best of our knowledge, this is the first work to evaluate the effects of ultrafiltrate recirculation on ML functionality. In our study, in a conscious spontaneously breathing sheep model undergoing ECCO2R, ultrafiltrate recirculation did not cause changes in ML CO2 removal or the animal’s ventilatory patterns and arterial blood gas values.
Livigni et al.14 recently tested a low-flow ECCO2R device in which ultrafiltrate recirculation was applied. In the circuitry used, the blood was driven by a peristaltic pump providing extracorporeal BF up to 300 ml/min. The circuitry was composed of a neonatal ML coupled in series with a standard hemofilter, such that ultrafiltrate was recirculated to the inflow of the ML by an additional peristaltic pump. The recirculating ultrafiltrate had a flow up to 155 ml/min. The presence of the hemofilter after the ML is supposed to create a downstream resistance and potentially reduce the risk of air bubble formation in the ML. Moreover, Livigni et al. suggested that ultrafiltrate recirculation optimized long-term functionality of the ML and ameliorated CO2 removal. By the application of this new device, an arterial pCO2 reduction of approximately 20% was observed in spontaneously breathing sheep. The same system has been used in a clinical trial11 in 10 patients with ARDS, allowing control of hypercapnia and permitting protective ventilation with very low tidal volumes. Unfortunately, neither study performed measurements of CO2 transfer of the devices, leaving open the question whether the observed consistent effects on ventilation were related to the application of ultrafiltrate recirculation or whether the same results could have been obtained without it. The works by Kolobow et al.15 and a more recent one16 studied the variables affecting CO2 transfer of a ML. These studies proved that while keeping sweep GF constant, VCO2ML is proportional to pre-ML blood pCO2 and to BF. In our experiment, during the application of ultrafiltrate recirculation, both pCO2 and BF were affected, but in opposite ways. Indeed, we observed that application of ultrafiltrate recirculation causes re-direction of already partially decarboxylated ultrafiltrate in front of the ML and hence dilution of pre-ML CO2 blood content. The CO2 trans-membrane gradient is the driving force for ML CO2 removal. An addition of ultrafiltrate recirculation lowers the CO2 concentration in pre-ML blood, leading to decreased ML CO2 removal efficiency. This negative effect is countered by the increased flow passing through the ML. Consequently, there is no effect of ultrafiltrate recirculation on the absolute values of CO2 removed, which indeed was constant regardless of UF. As a direct result of this, we observed no effects of ultrafiltrate recirculation on ventilatory patterns and arterial blood gas values. Ultrafiltrate recirculation did not enhance ML CO2 removal capabilities as compared with a standard ECCO2R.
Common problems of any extracorporeal therapy are the hemostasis and inflammatory response caused by the blood reaction to artificial surfaces.17 Because of the study design, detailed analyses of blood coagulation and inflammatory status could not be performed. Nevertheless, plasma-free hemoglobin was under pathologic thresholds.18 Moreover, ultrafiltrate recirculation progressively reduced trans-ML resistance. These results suggest that ultrafiltrate recirculation is not harmful to blood components and may as well be capable of reducing the shear stress applied to the blood by passage through the ML. Thus, ultrafiltrate recirculation may mitigate hemostasis and inflammatory response to extracorporeal circulation by reducing the coagulability of blood circulating through the ML. Indeed, hemofiltrate is a fluid free of coagulation factors, platelets, and erythrocytes.19 During ultrafiltrate recirculation, the hemofiltrate dilutes the blood entering the ML. The subsequent hemodilution might reduce viscosity and procoagulability.20 Hence, ultrafiltrate recirculation may potentially lengthen ML endurance, reduce coagulation activation, and permit lower anticoagulant requirements.
Respiratory failure frequently develops in the context of MODS.7 Patients managed with mechanical ventilation or ECLS die secondary to multiple-organ failure.8 It has been hypothesized that the harmful proinflammatory effects of VILI are not limited to the lungs but that instead lung inflammation elicits systemic endothelial activation.21,22 Extracorporeal lung support itself is associated with activation of the inflammatory response, massive production of proinflammatory cytokines, and subsequent MODS.23,24 Moreover, patients with respiratory failure frequently develop acute kidney injury, which significantly increases their mortality.25,26 Recent meta-analysis reported that up to 52% of patients with ECLS require CRRT27 during their clinical stay. A clinical strategy combining ECLS and CRRT for the early management of patient with ARDS is evaluated by a current ongoing clinical trial.28 Typically, administration of CRRT during ECLS requires placement of an additional catheter and a separate circuit. Alternatively, same treatment can be delivered by a custom-made connection of a CRRT device in parallel with the ECLS circuit. Therefore, development of modular ECLS systems capable of providing various forms of blood treatment may be useful. Coupling of different organ support technologies takes advantage of possible synergistic effects these treatments have to offer. A single modular machine provides a possibility to sustain different organ systems simultaneously, using a single blood catheter and a unique circuitry, hence reducing the overall footprint compared with the application of separate devices. The extracorporeal circuit such as the one presented here, by study design, did not provide renal support, electrolyte homeostasis, or water balance adjustments. Our study was designed to test the CO2 removal capabilities of an ML interconnected with a hemofilter in one of several possible configurations. Notwithstanding, our circuit can be easily supplemented with an additional pump for ultrafiltrate removal or fluid replacement and therefore can provide volume control and blood purification alongside with lung support. Forster et al.29 recently reported implementation of an ML on a low-flow renal-replacement circuit in clinical settings. He demonstrated the feasibility, safety, and CO2 removal efficacy of such an approach in patients suffering from respiratory and renal failure. With additional fluid removal/replacement pumps, our circuit may be similarly efficacious for patients suffering from lung and kidney failure.
Moreover, application of ultrafiltrate recirculation on an ECCO2R device provides an easy access to hemofiltrate, which may be further processed and purified of cytokine and inflammatory mediators. This can be easily achieved through interposition of a cytokine-adsorptive column in the ultrafiltrate circuit. Considering that multiple-organ failure is a leading cause of mortality and morbidity of mechanically ventilated patients affected by respiratory failure,8 development of modular devices capable of simultaneous ECCO2R and immunological modulation is crucial. These devices may offer the possibility of controlling both the local and the systemic adverse effects of ventilation by limiting injurious mechanical stress to the lungs and by controlling systemic inflammatory response caused by VILI.
An important promise of modular extracorporeal life support is that ultrafiltrate recirculation offers the possibility of favorably adjusting the chemical characteristics of blood entering the ML by infusion of acid or hyperosmolar compounds into the UF, thus avoiding the direct contact of blood cellular components with a potentially harmful milieu and minimizing the volume of injected solutes. Such regional blood acidification technique was used previously by Zanella et al.30 Their study proved that blood acidification at the inlet of an ML, converting bicarbonate ions into dissolved carbon dioxide, increased CO2 removal performance of an ML. Moreover, in principle, the ultrafiltrate part of the extracorporeal construct could also serve as a location for infusion of anticoagulants reducing or eliminating the need for systemic anticoagulation.3
Bias in this study may originate from the components used in the setup, which could be arranged in a variety of ways. The Hemolung device is equipped with a central rotating core pump that distributes the blood radially through a stationary hollow fiber polymethylpentene membrane.31 This device was coupled with a standard polyethersulfone hemofilter (Purema; NxStage, NxStage Medical), having a membrane surface of 1.6 m2. We believe that similar results can be achieved regardless of the individual components used for the circuit buildup, and thus the universality of the modular approach is high. Indeed, we showed that the effects of ultrafiltrate recirculation on the CO2 removal are the product of the inherent dilution process rather than specification of a particular technology.
Regarding the effects of ultrafiltrate recirculation on blood coagulability, because of the repeated measurement design of our experiment, it was not possible to evaluate the effects of ultrafiltrate recirculation on coagulation, inflammatory status of the animals, or ML durability. Further studies are necessary to evaluate the possible influence on coagulation activation and blood hemostasis because of application of ultrafiltrate recirculation on an ECCO2R device as well as the potential to exclusively anticoagulate a subject extracorporeally.
A limitation of our study is the low number of animals used, which can potentially lead to type II error. However, we believe that this problem has been minimized by randomizing the same test eight times on each animal and by proving that the baseline measurements remained consistent throughout the experiment. Other limitations may come from the characteristics of the experimental animals. Because the experiment aim was to evaluate specifically the effects on extracorporeal blood and ML function of ultrafiltrate recirculation, we used healthy animals to limit further confounding factors (e.g., different levels of hypercapnia). Studies applying this technique in injured model are necessary to evaluate its clinical efficacy.
In conclusion, ultrafiltrate recirculation did not affect CO2 removal of an ECCO2R device. Combination of extracorporeal lung and kidney support circuitry by ultrafiltrate recirculation may provide a suitable modular platform for coupling CO2 removal therapy with various forms of dialysis and blood purification. Future studies in clinically relevant injured models are necessary to evaluate the therapeutic efficacy and clinical advantages of combination of single-organ support technologies into modular multiple-organ support devices.
1. Gattinoni L, Pesenti A, Mascheroni D, et al. Low-frequency positive-pressure ventilation with extracorporeal CO2
removal in severe acute respiratory failure. JAMA. 1986;256:881–886
2. Bein T, Weber-Carstens S, Goldmann A, et al. Lower tidal volume strategy (≈3 ml/kg) combined with extracorporeal CO2
“conventional” protective ventilation (6 ml/kg) in severe ARDS: The prospective randomized Xtravent-study. Intensive Care Med. 2013;39:847–856
3. Pesenti A, Pelizzola A, Mascheroni D, et al. Low frequency positive pressure ventilation with extracorporeal CO2
removal (LEPPV-ECCO2R) in acute respiratory failure (ARF): Technique. Trans Am Soc Artif Intern Organs. 1981;27:263–266
4. Cardenas VJ Jr, Lynch JE, Ates R, Miller L, Zwischenberger JB. Venovenous carbon dioxide removal in chronic obstructive pulmonary disease: Experience in one patient. ASAIO J. 2009;55:420–422
5. Brenner K, Abrams DC, Agerstrand CL, Brodie D. Extracorporeal carbon dioxide removal
for refractory status asthmaticus: Experience in distinct exacerbation phenotypes. Perfusion. 2014;29:26–28
6. Javidfar J, Bacchetta M. Bridge to lung transplantation with extracorporeal membrane oxygenation support. Curr Opin Organ Transplant. 2012;17:496–502
7. Estenssoro E, Dubin A, Laffaire E, et al. Incidence, clinical course, and outcome in 217 patients with acute respiratory distress syndrome. Crit Care Med. 2002;30:2450–2456
8. Vincent JL, Zambon M. Why do patients who have acute lung injury/acute respiratory distress syndrome die from multiple organ dysfunction syndrome? Implications for management. Clin Chest Med. 2006;27:725–731; abstract x
9. Cruz D, Bellomo R, Kellum J a, de Cal M, Ronco C. The future of extracorporeal support. Crit Care Med. 2008;36(4 suppl):S243–S252
10. Gramaticopolo S, Chronopoulos A, Piccinni P, et al. Extracorporeal CO2
removal—A way to achieve ultraprotective mechanical ventilation and lung support: The missing piece of multiple organ support therapy. Contrib Nephrol. 2010;165:174–184
11. 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;111:826–835
12. Gille JP, Lautier A, Tousseul B. EC CO2R: Oxygenator or hemodialyzer? An in vitro
study. Int J Artif Organs. 1992;15:229–233
13. CLSI.Blood Gas and pH Analysis and Related Measurements, second ed., C46-A2. 2009 Wayne, PA Clinical and Laboratory Standards Institute:3
14. Livigni S, Maio M, Ferretti E, et al. Efficacy and safety of a low-flow veno-venous carbon dioxide removal device: Results of an experimental study in adult sheep. Crit Care. 2006;10:R151
15. Kolobow T, Gattinoni L, Tomlinson T, White D, Pierce J, Iapichino G. The carbon dioxide membrane lung (CDML): A new concept. Trans Am Soc Artif Intern Organs. 1977;23:17–21
16. Park M, Costa ELV, Maciel AT, et al. Determinants of oxygen and carbon dioxide transfer during extracorporeal membrane oxygenation in an experimental model of multiple organ dysfunction syndrome. PLoS One. 2013;8:e54954
17. Gorbet MB, Sefton MV. Biomaterial-associated thrombosis: Roles of coagulation factors, complement, platelets and leukocytes. Biomaterials. 2004;25:5681–5703
18. Lou S, Maclaren G, Best D, Delzoppo C, Butt W. Hemolysis in pediatric patients receiving centrifugal-pump extracorporeal membrane oxygenation: Prevalence, risk factors, and outcomes. Crit Care Med. 2013
19. Guth HJ, Klingbeil A, Wiedenhöft I, Rose HJ, Kraatz G. Presence of factor-VII and -XIII activity in ultrafiltrate during hemofiltration
. Int J Artif Organs. 1999;22:482–487
20. Ogawa S, Ohnishi T, Hosokawa K, Szlam F, Chen EP, Tanaka KA. Haemodilution-induced changes in coagulation and effects of haemostatic components under flow conditions. Br J Anaesth. 2013;111:1013–1023
21. Parsons PE, Eisner MD, Thompson BT, et al. Lower tidal volume ventilation and plasma cytokine markers of inflammation in patients with acute lung injury. Crit Care Med. 2005;33:1–6; discussion 230–232
22. Stüber F, Wrigge H, Schroeder S, et al. Kinetic and reversibility of mechanical ventilation-associated pulmonary and systemic inflammatory response in patients with acute lung injury. Intensive Care Med. 2002;28:834–841
23. Adrian K, Mellgren K, Skogby M, Friberg LG, Mellgren G, Wadenvik H. Cytokine release during long-term extracorporeal circulation in an experimental model. Artif Organs. 1998;22:859–863
24. Edmunds LH. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg. 1998;66(5 suppl):S12–S16; discussion S25–8
25. Cooke CR, Kahn JM, Caldwell E, et al. Predictors of hospital mortality in a population-based cohort of patients with acute lung injury. Crit Care Med. 2008;36:1412–1420
26. Liu KD, Glidden DV, Eisner MD, et al. Predictive and pathogenetic value of plasma biomarkers for acute kidney injury in patients with acute lung injury. Crit Care Med. 2007;35:2755–2761–XXX
27. Zangrillo A, Landoni G, Biondi-Zoccai G, et al. A meta-analysis of complications and mortality of extracorporeal membrane oxygenation. Crit Care Resusc. 2013;15:172–178
28. Hôpital Ambroise Paré. . Pulmonary and renal support during acute respiratory distress syndrome (PARSA). Clin [Internet] Bethesda Natl Libr Med. Available at: http://clinicaltrials.gov/show/NCT01239966
. Accessed Feb 10, 2014
29. Forster C, Schriewer J, John S, Eckardt KU, Willam C. Low-flow CO2
removal integrated into a renal-replacement circuit can reduce acidosis and decrease vasopressor requirements. Crit Care. 2013;17:R154
30. Zanella A, Patroniti N, Isgrò S, et al. Blood acidification enhances carbon dioxide removal of membrane lung: An experimental study. Intensive Care Med. 2009;35:1484–1487
31. Batchinsky AI, Jordan BS, Regn D, et al. Respiratory dialysis
: Reduction in dependence on mechanical ventilation by venovenous extracorporeal CO2
removal. Crit Care Med. 2011;39:1382–1387