Secondary Logo

Journal Logo

Kidney/Dialysis/Vascular Access

Combined Pulmonary and Renal Support in a Single Extracorporeal Device

Wiegmann, Bettina; Maurer, Andreas; Zhang, Ruoyu; Zardo, Patrick; Haverich, Axel; Fischer, Stefan

Author Information
doi: 10.1097/MAT.0b013e318292e887
  • Free


In the intensive care unit (ICU), kidney and lung represent the most common targets of multiorgan failure. Because of shared pathophysiological pathways, acute lung injury (ALI) contributes to the development of acute kidney injury (AKI) and vice versa.1–3 Irrespective of the underlying disease, the mortality rate for combined ALI/AKI reaches 80%1 on both surgical and nonsurgical ICUs.

Fluid balance is generally acknowledged as the single most influential outcome predictor for critically ill patients.4,5 Acute kidney injury–associated fluid accumulation can lead to deterioration of respiratory function with prolonged artificial ventilation, nutritional deficits, and metabolic derangement toward catabolic states,6 finally leading to extended ICU stays with increased morbidity and mortality.6,7 Therefore, early implementation (<48 h from ICU admission) of a renal replacement therapy (RRT), in particular by continuous venovenous hemofiltration (CVVH), could potentially improve the outcome.5 It reduces the need for fluid restriction and the need for diuretics and allows delivery of more appropriate nutritional support. Furthermore, CVVH improves cardiovascular and respiratory function by enhancing respiratory compliance and oxygenation.8,9 Through constant fluid homeostasis with improved hemodynamic stability and avoidance of fluid shifts, continuous RRT is largely considered to be the preferred treatment modality.10

Mechanical ventilation is often insufficient in ALI, leading to hypoxemia and hypercapnia, which reduces renal blood flow by increasing renal vascular resistance frequently resulting in inadequate renal function.1 In addition, conventional mechanical ventilation can induce ventilator-associated lung injury and kidney damage via inflammation.11 The key strategy in preventing this cascade is the early implementation of extracorporeal membrane oxygenation (ECMO). Extracorporeal membrane oxygenation therapy in ALI enhances oxygen delivery and allows a switch to protective ventilator settings by decreasing tidal volumes and positive end-expiratory pressures (PEEPs). This strategy has shown to improve hemodynamics by lowering pulmonary vascular resistance and decreasing cardiac preload and postload, thereby also improving renal perfusion.12,13 Increased glomerular filtration rates and free water clearance will finally enhance pulmonary function.14 In recent analyses, simultaneous RRT and ECMO were shown to be associated with an improved fluid balance, amelioration of electrolyte imbalances, increased caloric intake, and reduced administration of diuretics.15 This concept resulted in reduced morbidity and mortality rates. In addition, RRT significantly reduced edema and allowed both, more timely discontinuation of extracorporeal assistance and reduction of the interval to extubation.16,17

Therefore, earlier use of a combined approach might be a preventive strategy for kidneys and lungs, as up to 50% of patients on ECMO support have AKI and require RRT.12,18 Using a conventional approach, however, each device presents its own artificial surfaces to the circulating blood.18 Possible implications include platelet activation, adhesion, and aggregation promoting thrombus formation as well as activation of the complement system and the release of inflammatory mediators potentially resulting in sepsis.19 Combining both extracorporeal systems in one device with only two vascular cannulae exhibits significant advantages. This approach may reduce the required area of artificial surfaces, therefore decreasing the risk of the aforementioned adverse reactions. Furthermore, low-flow resistance of the system could enable long-term function even without a pump. However, the system should be capable to individually modulate continuous gas transfer and fluid balance to maintain sufficient levels of oxygenation and fluid filtration.

This article describes both designs and first in vivo and in vitro evaluation of such a modified prototype (Modified Interventional Lung Assist [iLA] Membrane Ventilator [miLA]), consisting of 50% polymethylpentene (PMP) gas exchange membranes and 50% polyethersulfone (PES) hemofiltration membranes. The construction is based on the original iLA device (Novalung, Heilbronn, Germany), conventionally available as membrane oxygenator, which is well-known for its durable low-flow resistance in clinical practice.20

Material and Methods

miLA Design

The construction of the miLA was based on the original iLA design and included the cannulae, the casing, and the amount of fibers in the device to have a better comparability between the two devices. The miLA fiber arrangement, consisting of 50% PMP gas exchange membranes and 50% PES hemofiltration membranes, is illustrated in Figure 1 and resulted in a total surface area of 1.3 sq m, the same surface area as the original iLA. Because of the square shape of the casing, the PES fibers were crossed right-angled and arranged alternately to the PMP fibers to create separate inflow and outflow tracts for both types of hollow fiber membranes. Thus, the blood flowed in one direction through the device, as both inflow and outflow were positioned in the lower corner of the device, leading to the countercurrent exchange for both pathways, the gas exchange and the hemofiltration. Furthermore, this arrangement resulted in three different pressure areas, which could interact with each other—pressures in PMP (pressure 1) and PES (pressure 3) fibers, as well as in the blood flow area (pressure 2).

Figure 1
Figure 1:
miLA design.

In Vitro Tests

Comparison of flow resistance of miLA against iLA and the filtration volumes of miLA were performed in triple, using purified water. To determine delivery volumes, a centrifugal pump (Jostra Rotaflow, Maquet, Hirrlingen, Germany) was used with pump rates from 1,000 to 2,500 rpm. The corresponding delivery volumes were quantified by measuring the delivered volumes with a calibrated cylinder. The centrifugal pump was also used to quantify loss of pressures through both systems at identical pump rates, whereas the pressures were measured at the inflow and outflow cannulae. Pressure differences were calculated as pressure drops. The filtration rates of the miLA were analyzed by setting the pump to 1,000 rpm, pressure 1 to 0 mm Hg and pressure 2 to 50 mm Hg in relation to different back pressures in the PES fibers (Figure 1). Volumetric determination of filtration through the PES fibers was assessed using a calibrated cylinder.

Animal Experiments

The study was approved by local ethics for Animal Care and adhered to guidelines on animal experimentation. Three healthy male pigs (Genetik JSR Hybrid Deutsche Landrasse, Large White × Duroc), weighing 55–60 kg (5–6 months old), were obtained from W&P Agrarhandels GmbH, Oberheldrungen, Germany, and housed in a laboratory room with adequate room temperature (22 ± 1°C). Food and water were provided ad libitum. Animals were fasted for 12 h before the experiment.

After premedication with azaperone (2 mg/kg; Stresnil; Janssen-Cilag, Neuss, Germany) and atropine (0.05 mg/kg; Atropinsulfat; B. Braun, Melsungen, Germany), general anesthesia was induced with ketamine 10% (0.2 ml/kg; Ketanest S; Pfizer, Berlin, Germany) and xylazine 2% (0.1 ml/kg; Rompun; Bayer, Leverkusen, Germany) and kept up with fentanyl (0.005 mg/kg; Fentanyl; Janssen-Cilag), midazolam (0.05 mg/kg; Dormicum; Roche, Germany), and isoflurane (Isoflurane; Baxter, Unterschleißheim, Germany). All surgical procedures were performed under endotracheal intubation and mechanical ventilation (Siemens Servo 900c Ventilator; Siemens-Elema, Solna, Sweden), with an initial inspirational oxygen fraction (FiO2) of 50% after complete muscular relaxation (pancuronium bromide, 0.1 mg/kg). Subsequently, the settings included from tidal volumes of 6–8 ml/kg, a respiratory frequency of 12 per minute, FiO2 0.21 to 1.0, inspiratory/expiratory time ratio 1:2, and PEEP of 5 cm H2O. Lactated Ringer solution was infused at a rate of 20 ml/kg/h. Intravascular catheters were inserted after careful preparation of relevant vessels on neck and groin. A central venous catheter was placed in the left internal jugular vein. Arterial blood pressure was monitored, and regular blood samples were obtained through a catheter in the left internal carotid artery. By maintaining an adequate arterial blood pressure (mean arterial blood pressure [MAP] >60 mm Hg) adequate pump flow rates (>0.5 L/min) could be obtained. Body temperature was kept at 37–38°C. Animals were fully heparinized intravenously with 300 IU/kg BW Na-Heparin (Heparin Immuno; EBEWE Pharma, Unterach, Austria) before miLA cannulation. Activated coagulation time (ACT) was measured hourly and maintained above 200 with heparin bolus injections. After priming the miLA with heparinized normal saline solution, it was connected to cannulae in the right femoral artery and left femoral vein. Extracorporeal circulation was initiated and mechanical ventilation was minimized to an FiO2 of 0.21 with a PEEP of 5 cm H2O. Sweep gas flow through the miLA was set to 2 l oxygen/minute with 100% oxygenation. The miLA was positioned at heart level, so that back pressure in PES fibers was nearly 0 mm Hg. Half-hourly flow rate and loss of pressure rates through miLA were assessed after achieving steady state conditions. Hemodynamic parameters and efficiency of gas exchange and hemofiltration were investigated. For blood gas analysis (PaO2, PaCO2, pH, lactate, base excess), blood samples were obtained from the arterial and venous cannulae and immediately analyzed (ABL 710; Radiometer, Copenhagen, Denmark). Ethylenediaminetetraacetic acid blood samples and serum, obtained from the left carotid artery, as well as collected hemofiltration samples, were centrifuged at 2,000g for 10 min, flash-frozen and stored at −18°C until final analysis. Measured parameters included protein in g/L, albumin in g/L, sodium in mmol/L, potassium in g/L, hemoglobin in g/L, creatinine in µmol/L, urea in mmol/L, lactate dehydrogenase in U/L, bilirubin in µmol/L, and free plasma hemoglobin in g/L. Cardiovascular monitoring (heart rate, arterial pressure [AP], central venous pressure [CVP], peripheral oxygenation) was achieved by an integrated monitor system (Infinity Delta Monitor, Dräger, Germany). After 6 h, all animals were killed through a lethal injection of embutramide (75 mg/kg; T61, Intervet GmbH, Schwabenheim, Germany).

Statistical Analysis

The statistical analysis was performed in Microsoft Excel: Mac 2008 (Microsoft, Redmond, WA). The data were presented as median or mean with the standard deviation (mean ± SD).


miLA Construction and In Vitro Tests

In comparison with iLA, the prototype allowed comparable pump rates versus delivery volume (Figure 2) and loss of pressure (Figure 3), as well as the same low-flow resistance. Figure 4 shows the filtration volumes of miLA in correlation with variable back pressures in PES fibers, whereas potential filtration volume of 7,440 ml/h at a defined back pressure of −20 kPa could be measured.

Figure 2
Figure 2:
In vitro delivery volume versus various pump rates of miLA and iLA (mean ± SD).
Figure 3
Figure 3:
In vitro loss of pressure versus various pump rates of miLA and iLA (mean ± SD).
Figure 4
Figure 4:
In vitro miLA filtration volume versus various back pressures (mean ± SD).

Animal Experiments

During extracorporeal circulation, the CVP was 23 mm Hg (±1.8 mm Hg) at a mean AP of 78 mm Hg (±12 mm Hg) and a heart rate of 90 bpm (±11 bpm). Tidal volume at FiO2 of 0.21 was 426 ml (±50 ml) with a peripheral oxygenation saturation of 96% (±3.0%). Under systemic heparinization, an activating clotting time of 452 (±111) was measured and enabled continuous blood flow through miLA of 1.03 L/min (±0.18 L/min) with 8 mm Hg (±2 mm Hg) loss of pressure. Hemofiltration rates of 70 ml/h (±15 ml/h) were observed. No signs of hemolysis could be detected: lactate dehydrogenase 461 U/L (±55 U/L), conjugated bilirubin 1.04 µmol/L (±0.38 µmol/L), and free hemoglobin 0.11 g/L (±0.047 g/L). Table 1 shows the concentrations of filterable molecules in blood samples and their corresponding concentrations in filtrate samples. No major differences were observed. The measured gas transfer efficiency parameters, including oxygenation (Figure 5, pO2 arterial cannula 84 ± 23 mm Hg, pO2 venous cannula 439 ± 23 mm Hg) and decarboxylation (Figure 6, pCO2 arterial cannula 33 ± 7 mm Hg, pCO2 venous cannula 30 ± 5 mm Hg) indicated a stable gas exchange function with sufficient oxygenation and decarboxylation over the entire course of the experiment.

Figure 5
Figure 5:
miLA oxygenation efficiency at different time points after inducing extracorporeal circulation (mean ± SD).
Figure 6
Figure 6:
miLA decarboxylation efficiency at different time points after inducing extracorporeal circulation (mean ± SD).
Table 1
Table 1:
Concentrations of the Filterable Molecules in the Filtrate Samples and Corresponding Blood Samples at Different Time Points Onset of Extracorporeal Circulation


This study demonstrated the feasibility of designing an efficiently working single extracorporeal device with combined membranes for hemofiltration and gas transfer, which could maintain in vivo fluid balance and sufficient levels of oxygen and carbon dioxide (Figures 46). This innovative concept provides options to continuously modulate individual settings and may potentially improve individually adapted intensive care therapy of critically ill patients with severe ALI/AKI.

The design itself was associated with essential benefits, whereas the reduction of the artificial surfaces was achieved in two aspects. On one hand, and because the miLA layout was based on the iLA design, 50% of the original PMP gas exchange membranes was replaced by PES hemofiltration fibers (Figure 1). Therefore, by combining these two systems in one device, surface-associated adverse reactions,21 such as thrombus formation and inflammation, might decrease. On the other hand, instead of four, only two vascular access sites were needed for large-caliber catheters, also reducing the artificial surfaces. In addition, potential risks of cannulation such as dispuncture, dislocation, dissection, bleeding, and infection were reduced. This combined system needed a much more precise anticoagulation regimen. However, keeping ACT at therapeutic levels might be easier than performing repeated bolus applications of heparin when starting separate continuous/intermittent RRT. Nevertheless, one of the additional important challenges in miLA optimization will be the miniaturization of the system and the improvement of the fiber arrangement, obtaining the ideal balance between a minimized artificial surface, which at the same time allows sufficient gas and hemofiltration exchange capacities, and the present low-flow resistance of the system.

The miLA was designed to maintain the same essential and well-established advantages of iLA—low-flow resistance and durability. The results of this study demonstrated that the PMP/PES fiber arrangement have had an iLA-like low-flow resistance without any signs of hemolysis (Figures 2 and 3). Stable in vitro features concerning flow properties indicated the absence of thrombus formation in miLA. This implies that miLA may work without a pump, driven by the patient’s blood pressure. Because venovenous application appears to be the first choice in respiratory failure and significantly decreases mortality and morbidity (e.g., circuit rupture, brain death, and renal insufficiency), which occur more commonly among patients started on venoarterial ECMO,12 application of miLA with an additional pump system also needs to be investigated. Moreover, and with regard to the durability and potential long-term benefits of the system, testing in a chronic animal model would be beneficial.

To assess the hemofiltration capacity, further investigation of the complex interactions of the three separate pressures in the miLA (Figure 1), which modify the filtration volumes and lead to different in vitro and in vivo results, needs to be conducted. More precisely, in vitro pressures were fixed, but varied in vivo with the additional PMP fiber pressure, which might have influenced the filtration rates. Moreover, the purified water used in the study is not directly comparable with blood in the clinical setting. However, comparable in vivo and in vitro filtration volumes and concentrations of filterable molecules in the filtrate to blood indicate the general feasibility of hemofiltration with the miLA. Nevertheless, more near-physiologic in vitro experiments need to be done in future studies between miLA and conventional hemofiltration systems, whereas inter alia the feasibility of modulating hemofiltration rates via variable pore sizes and back pressures, especially in miLA PES fibers, should be analyzed. Besides this, an in vivo comparison between both gas exchange capacities of the miLA and iLA will be done in further experiments, which would emphasize our in vitro based statement that combining 50% PMP membranes with 50% PES membranes in one single device enables a sufficient gas exchange.

One of the general problems of extracorporeal devices remains the possibility of technical problems, including clotting and fiber rupture, which need to be detected by physicians and nursing staff. Therefore, thorough monitoring of the system for blood pressures, circuit pressures, and ultrafiltrate volumes should be added to improve control and safety. However, as significant advancements in indication, timing, and utilization of ECMO and RRT are ongoing22 and technical and material properties of extracorporeal systems constantly improve, these mentioned potential drawbacks of miLA would not be insurmountable.

Based on our initial results, there is a significant potential for further miLA optimization. Dedicated systems might improve intensive care therapy in novel ways. It is conceivable that certain clinical situations may require a different approach to miLA functionality. In the near future, specific designs of devices as “disease-specific prototypes” may be feasible. For sepsis or septic shock, an extracorporeal prototype may consist of gas exchange membranes and individually adapted filtration membranes with corresponding pore sizes. Such a design may allow for modulation of the inflammatory response in septic shock using high flux membranes to eliminate chemokines or cytokines by convection. In addition, a reduction of IL-8 and IL-10 helping to influence systemic consequences of sepsis and septic shock23 may be achievable. A differential disease-specific arrangement may be the combination of gas exchange and plasmapheresis membranes, allowing simultaneous gas exchange and decreasing antibody levels. This may be an option for patients with severe lung disease on ECMO waiting for lung transplantation (e.g., with donor-specific antibodies), thereby enabling successful transplantation by preventing acute rejection.24 Therefore, the combinations of these different membranes in one single device will be tested in future experiments because of their high potential of individually adapted intensive care therapy.


In conclusion, our results provide a proof-of-concept for the feasibility to combine a device for pulmonary and renal support in a single extracorporeal device. Novel device designs expanding on this concept may potentially improve individually tailored therapy of critically ill patients under various disease-specific conditions.


1. Ko GJ, Rabb H, Hassoun HT. Kidney-lung crosstalk in the critically ill patient. Blood Purif. 2009;28:75–83
2. Koyner JL, Murray PT. Mechanical ventilation and the kidney. Blood Purif. 2010;29:52–68
3. Murray PT. The kidney in respiratory failure and mechanical ventilation. Contrib Nephrol. 2010;165:159–165
4. Bouchard J, Soroko SB, Chertow GM, et al.Program to Improve Care in Acute Renal Disease (PICARD) Study Group. Fluid accumulation, survival and recovery of kidney function in critically ill patients with acute kidney injury. Kidney Int. 2009;76:422–427
5. Payen D, de Pont AC, Sakr Y, Spies C, Reinhart K, Vincent JLSepsis Occurrence in Acutely Ill Patients (SOAP) Investigators. . A positive fluid balance is associated with a worse outcome in patients with acute renal failure. Crit Care. 2008;12:R74
6. Mehta RL, Pascual MT, Soroko S, Chertow GMPICARD Study Group. . Diuretics, mortality, and nonrecovery of renal function in acute renal failure. JAMA. 2002;288:2547–2553
7. Arieff AI. Fatal postoperative pulmonary edema: Pathogenesis and literature review. Chest. 1999;115:1371–1377
8. Marenzi G, Lauri G, Grazi M, Assanelli E, Campodonico J, Agostoni P. Circulatory response to fluid overload removal by extracorporeal ultrafiltration in refractory congestive heart failure. J Am Coll Cardiol. 2001;38:963–968
9. Wiedemann HP, Wheeler AP, Bernard GR, et al.National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006;354:2564–2575
10. Bellomo R, Farmer M, Wright C, Parkin G, Boyce N. Treatment of sepsis-associated severe acute renal failure with continuous hemodiafiltration: Clinical experience and comparison with conventional dialysis. Blood Purif. 1995;13:246–254
11. Pinhu L, Whitehead T, Evans T, Griffiths M. Ventilator-associated lung injury. Lancet. 2003;361:332–340
12. Brogan TV, Thiagarajan RR, Rycus PT, Bartlett RH, Bratton SL. Extracorporeal membrane oxygenation in adults with severe respiratory failure: A multi-center database. Intensive Care Med. 2009;35:2105–2114
13. . Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000;342:1301–1308
14. Kuiper JW, Groeneveld AB, Slutsky AS, Plötz FB. Mechanical ventilation and acute renal failure. Crit Care Med. 2005;33:1408–1415
15. Hoover NG, Heard M, Reid C, et al. Enhanced fluid management with continuous venovenous hemofiltration in pediatric respiratory failure patients receiving extracorporeal membrane oxygenation support. Intensive Care Med. 2008;34:2241–2247
16. Kelly RE Jr, Phillips JD, Foglia RP, et al. Pulmonary edema and fluid mobilization as determinants of the duration of ECMO support. J Pediatr Surg. 1991;26:1016–1022
17. Roy BJ, Cornish JD, Clark RH. Venovenous extracorporeal membrane oxygenation affects renal function. Pediatrics. 1995;95:573–578
18. Santiago MJ, Sánchez A, López-Herce J, et al. The use of continuous renal replacement therapy in series with extracorporeal membrane oxygenation. Kidney Int. 2009;76:1289–1292
19. Peek GJ, Firmin RK. The inflammatory and coagulative response to prolonged extracorporeal membrane oxygenation. ASAIO J. 1999;45:250–263
20. Bein T, Weber F, Philipp A, et al. A new pumpless extracorporeal interventional lung assist in critical hypoxemia/hypercapnia. Crit Care Med. 2006;34:1372–1377
21. McGuigan AP, Sefton MV. The influence of biomaterials on endothelial cell thrombogenicity. Biomaterials. 2007;28:2547–2571
22. Marasco SF, Lukas G, McDonald M, McMillan J, Ihle B. Review of ECMO (extracorporeal membrane oxygenation) support in critically ill patients. Heart Lung Circ. 2008;17(suppl 4):S41–S47
23. Joannidis M. Continuous renal replacement therapy in sepsis and multisystem organ failure. Semin Dial. 2009;22:160–164
24. Martinu T, Chen DF, Palmer SM. Acute rejection and humoral sensitization in lung transplant recipients. Proc Am Thorac Soc. 2009;6:54–65

extracorporeal membrane oxygenation; device; acute respiratory distress syndrome; intensive care

Copyright © 2013 by the American Society for Artificial Internal Organs