Kopp, Ruedger*; Bensberg, Ralf†; Arens, Jutta‡; Steinseifer, Ulrich‡; Schmitz-Rode, Thomas‡; Rossaint, Rolf†; Henzler, Dietrich§
Venovenous extracorporeal membrane oxygenation (ECMO) may be used in most severe acute respiratory distress syndrome (ARDS) with persistent hypoxemia despite optimized lung protective mechanical ventilation and adjunctive therapy.1 A recent controlled randomized trial demonstrated a survival benefit for patients with ARDS treated with venovenous ECMO as part of a advanced treatment concept, when compared with conventional therapy without ECMO.2 So far, the complexity, size, and risk of device specific complications have limited the application of ECMO to specialized centers and a rescue therapy for refractory hypoxemia.3 Few specifically “pulmonary” ECMO devices are available, which means that most centers use modified cardiopulmonary bypass circuits with high filling volume and large blood contacting surfaces. As a result, hemodilution and activation of plasmatic coagulation, platelets, and immune system contribute to device associated morbidity.4
Different approaches have been taken to integrate an oxygenator and a blood pump in one device to simplify management of these devices5 and/or to miniaturize the circuit. Coupling of the pump to the oxygenator and placing the device beside the patient with shorter tubes is one way to reduce filling volume. Commercially available devices combine conventional components of cardiopulmonary support by individual coupling of miniaturized blood pump and oxygenator6 or by permanent connection of a centrifugal pump and an oxygenator.7 Although these devices have demonstrated easier handling and reduced filling volumes in the range of 300 ml, they use conventional technology without miniaturization of oxygenator or blood pump.
New concepts seem desirable to incorporate specifically optimized, miniaturized blood pumps and oxygenators in one housing, to decrease filling volume, and to reduce blood contacting surfaces and minimize heat loss. The rational to develop such devices includes the potential for increased mobility. A highly integrated ECMO devices could simplify management and increase safety for inter- and intrahospital transport of severely ill patients connected to ECMO.8–10 Further applications could potentially include patients with very low blood volume (i.e., children and neonates), where miniaturization could significantly reduce hemodilution and blood transfusion rate.
On the basis of these considerations, the highly integrated extracorporeal membrane oxygenator (HEXMO) was developed and tested in vitro as described previously.11 The HEXMO combines approximately 0.9 m2 of hollow fiber membrane surface with an integrated rotary blood pump in a single housing (Figure 1). The drive unit of the blood pump is placed concentrically inside the device but has no blood contact. The rotor is driven by a magnetic coupling. To enable such compact housing, a new blood flow concept had to be developed for the oxygenator, whereby the blood flow changes direction inside the oxygenator by 180 degrees to perfuse an inner and an outer cylindrical bundle of fibers sequentially (Figure 2). We tested a configuration without a heat exchanger, with the assumption that heat produced by the blood pump would be transferred into the extracorporeal flow and the placement of HEXMO close the patient with short connective tubing would further optimize heat preservation. The pump-oxygenator unit is working temperature neutral. As a result of the compact design, the priming volume was reduced to 125 ml in total.
The aim of the study was to test the efficiency and hemocompatibility of this prototype in an in vivo animal study.
Materials and Methods
Six HEXMO modules were tested in female pigs with a bodyweight of 37 ± 1 kg. Respiratory failure was simulated by mechanical ventilation with a hypoxic gas mixture, and the operating time of the HEXMO was planned to be 4 hours. The experiments were approved by the appropriate governmental animal care committee, and the principles of laboratory animal care were followed.
The oxygenator modules (filling volume 150 ml) were wound and potted by hand, using polypropylene membranes (Membrana Oxyphan PP50/200) and a cohesive two-part addition-curing silicone rubber (Wacker Silicones, Elastosil RT625 A/B). Hand-made production of oxygenation modules was chosen to avoid the delay for outsourced production. Two housing units were used, which were dissembled, thoroughly cleaned, and inspected after each experiment and equipped with a new fiber bundle.
Before connecting the device to the animal, the circuit was primed with hydroxyethyl-starch (HES) solution (200/0.5, 10%) and Ringer's lactate solution 1:1 and carefully deaired.
Intramuscular premedication with 4 mg · kg−1 azaperone and 10 mg · kg−1 ketamine was followed by tracheal intubation and intravenous anesthesia with 5–10 mg · kg−1 · h−1 thiopental and 8–12 μg · kg−1 · h−1 fentanyl infusions. Animals were mechanically ventilated in volume-controlled mode in supine position (Servo 300A Ventilator, Siemens Elema, Lund, Sweden). Inspiratory oxygen fraction (FiO2) was 1.0, and respiratory rate (RR) was adjusted to achieve normocapnia. An active heating system (warming blanket) was used during animal preparation. After connection to the HEXMO, the animals were covered with conventional sheets without active warming.
Ringer's solution and HES solution (200/0.5, 10%) were infused to maintain adequate hydration, and urine output was measured by transurethral bladder catheterization.
A 16-G arterial catheter (Vygon, Ecouen, France) was percutaneously introduced into a femoral artery and an 8.5-Fr venous sheath with a Swan-Ganz catheter positioned in a pulmonary artery (Arrow, Erding, Germany) by the femoral vein. Blood pressures were directly transduced, and cardiac output was measured by transpulmonary thermodilution and calculated by a computer monitor (AS/3 Compact Datex- Ohmeda, Helsinki, Finland).
Vascular access for the HEXMO was achieved by cannulation of opposite femoral vein and the right external jugular vein with 17–19 Fr cannulas. Blood flow through the extracorporeal circuit was measured by ultrasound (HT 110 Transonic Systems, Maastricht, Netherlands).
The FiO2 was reduced to achieve a hypoxic inspiratory mixture targeting for an arterial oxygen saturation (SaO2) below 85%, and then the animals were connected to HEXMO. The pump flow was adjusted to 30%–40% of cardiac output.6,12In vitro testing had resulted in stable flow characteristics up to 8,000 rpm and 3.0 L/min.13 During pilot experiments, we had achieved in vivo blood flows up to 2.8 L/min with optimized fluid management and cannula position at a maximum of 8,000 rpm. Intravenous heparin was repeatedly given to maintain an activated clotting time (ACT) of ≥150 seconds. After the experiments, the animals were killed by barbiturate injection.
Hemodynamics, ventilation, and performance of the extracorporeal gas exchanger were measured after induction of hypoxemia (T 0) and after 2 (T 2) and 4 hours (T 4) of ECMO. Blood gas analysis was performed on arterial and mixed-venous blood and blood taken from the circuit before and after the HEXMO (ABL 510 and OSM 3, Radiometer, Copenhagen, Denmark). Blood cell count, plasma-hemoglobin, fibrinogen, and thrombin-antithrombin III complex (TAT) were measured in arterial blood.
Blood cell count was analyzed from ethylenediaminetetraacetic acid blood using an automated hematology analyzer (MEK-6108G, Nihon Kohden Europe, Bad Homburg, Germany), and plasma-hemoglobin concentration was measured from citrate-anticoagulated plasma using a previously published protocol.14 Freshly taken whole blood was used for analysis of ACT (Hemochron 401 with FTCA510 tubes, ITC, Edison, NJ). Fibrinogen concentration was measured in citrate plasma (Multifibren, Dade Behring, Marburg, Germany) with a coagulometer (KC 4A, Heinrich Amelung GmbH, Lemgo, Germany). TAT was quantified with a commercially available human enzyme-linked immune sorbent assay (ELISA) demonstrating crossreactivity to pig species (Enzygnost TAT ELISA, Dade Behring).15
Data are presented as mean ± standard deviation (SD) and median. Data were compared with nonparametric repeated measures analysis of variance using Friedman's test (InStat Statistical Software version 3.06, GraphPad, San Diego, CA). A p value <0.05 was considered significant.
The animals were mechanically ventilated with a mean tidal volume of 9 ± 1 ml · kg−1 bodyweight (median: 8 ml · kg−1) during the entire experiment. Reducing the FiO2 to 0.20 caused severe hypoxemia with a SaO2 of 79% ± 5%. The cannulas were placed without problems, and all animals were connected to the primed HEXMO devices.
One experiment was terminated after 10 minutes due to breakdown of the magnetic motor coupling of the integrated diagonal blood pump. Five experiments were completed successfully for the planned time without complications.
The ventilation settings remained unchanged after start of HEXMO (Table 1), and hypoxemia was reversed in all animals, as indicated by an increase in SaO2 and PaO2 (Figure 3). All animals were hemodynamically stable with a trend toward an increased cardiac output (Table 1).
The mean extracorporeal blood flow was maintained at 32% ± 6% of cardiac output with a gas to blood flow ratio of 1.2 ± 0.3. The body temperature remained constant (Table 2).
To maintain adequate intravascular volume, 6.2 ± 4.4 ml · kg−1 · h−1 of crystalloid and 1.9 ± 2.0 ml · kg−1 · h−1 of colloid solutions were infused. This caused a mild hemodilution as indicated by reduced hematocrit and hemoglobin values (Table 3). No significant hemolysis could be detected as plasma-hemoglobin levels did not increase over time. However, although ACT was in the target range, increased levels of TAT and a decrease in platelet count suggest significant activation of the coagulation system (Table 3). Some thrombus formation was detected in the outflow tract of the HEXMO after one experiment, although we could not observe a flow limitation during that experiment.
During the extracorporeal flow phase, the oxygen transfer rate by the HEXMO was unchanged, but the pulmonary oxygen transfer decreased (Table 4, Figure 4). A small decrease in PaCO2 was not significant (Figure 3), although the combined carbon dioxide elimination by the lung and the HEXMO increased (Table 4, Figure 5).
For patients with most severe ARDS, ECMO is a rescue therapy with demonstrated benefit in case of refractory hypoxemia, if it is integrated into an advanced treatment concept.2 Despite advances in oxygenator and blood pump design including a reduced filling volume of approximately 300 ml for both components, the application of ECMO is associated with a number of specific complications, such as hemorrhage, thromboembolism, or technical failure of components.3 The more complex a system, the more vulnerable it is to technical and application failure. The logical consequence is to reduce complexity, which is achieved by the HEXMO by integrating oxygenator and blood pump into a single housing with a reduced filling volume of 150 ml. This new design facilitates full ECMO capacity embedded into a device only slightly bigger than a can of pop. We have tested a prototype development mainly to assess performance and biocompatibility in an in vivo setting.
The HEXMO was effective in reversing hypoxemia, and the lungs were unloaded from pulmonary oxygen transfer as the extracorporeal oxygen transfer became effective. We could not demonstrate a significant reduction in the PaCO2, although the extracorporeal carbon dioxide elimination was clearly present. This may be attributed in part to the experimental design, which does not impair the gas exchange capabilities of the lungs. Thus, native lungs and HEXMO were “competing” for carbon dioxide elimination, a situation which would have not occurred in a model with significant ventilation/perfusion mismatch. We speculate that in a situation where carbon dioxide elimination by the lungs is impaired, the effect of the HEXMO on PaCO2 might have been greater.
Blood flow through the oxygenator is crucial for oxygenation. For example, 300 ml/min was sufficient to reverse hypercapnia but was not effective for oxygen transfer.16 In another new development with an integrated paracorporeal rotating oxygenator, the oxygen transfer rate was limited to 52 ml/min despite higher blood flow rates.17 Carbon dioxide transfer was 64 ml/min mean. The HEXMO has similar oxygen transfer efficiency, but with a gas/blood flow ratio of approximately 1.2, carbon dioxide elimination was limited. We speculate that increasing the gas flow higher than 2 L/min might increase efficacy, as it is recommended similarly for a pumpless extracorporeal device with similar blood flow and gas exchange characteristics.18
During experiments, no hemolysis was observed, and hemodilution after connecting the ECMO to the animals was marginal. Activation of plasmatic coagulation occurred as indicated by increasing concentrations of TAT. Of note, this prototype's surfaces were not heparin coated, which has been shown to be a major advantage for hemocompatibility.4 A previous animal study resulted in lower TAT concentration after 4 hours of ECMO, but all surfaces were heparin coated.12 A further problem was significant thrombus formation in a dead water area at the blood outlet part of the housing. For future development of the HEXMO, a redesign of the blood outlet and heparin coating has to be considered.
The first 4 hours of ECMO seem to be sufficient to provide a reliable estimate of hemocompatibility,19 thus it is reasonable to assume that the HEXMO has the potential to provide long-term support. Assembly and handling of the device is greatly simplified, and the technology can be managed by “turning of one knob.” This will increase ease and safety of patients transfer.
This study has several limitations. To allow comparison with previous studies, we used a porcine model for testing of extracorporeal lung assist.6 We chose ventilation with a hypoxic gas mixture to produce stable severe hypoxemic conditions with a SaO2 <85%. Other methods to create hypoxemia, for example, induction of acute lung injury by surfactant depletion or oleic acid application,12,20 may have direct implications on organ function as the inflammatory component of lung injury is missing in the hypoxia model.21
Because the hypoxia model does not reproduce physiological conditions seen in a clinical situation, it allows to study the effects of the ECMO system independently from inflammatory response and multiorgan failure.
In this study, we tested the first hand-made production in vivo, and with the knowledge gained from these preliminary experiments, the HEXMO prototype can be further developed to improve gas exchange capacity and to optimize blood flow characteristics in the device.
Other available miniaturized ECMO systems usually combine individual coupling of components with a small filling volume12 or the fixed combination of conventional centrifugal pump and oxygenator.22 In these closely to the bedside placed systems, the priming volumes are reduced mainly by shortening of the tubing. A downsized extracorporeal circulation system (Emergency Life Support System-ELS, Maquet, Hirrlingen, Germany) demonstrated safe and simple application for the transport of patients on ECMO compared with conventional, not integrated systems.7 In patients with high bleeding risk, a completely heparin-coated, miniaturized system had demonstrated safe application without increased thromboembolism,23 in the absence of systemic heparinization.
Hemocompatibility and sustainability of the used materials will determine the ability to perform long-term support without failure. As there was some clot formation, the flow design has to be improved. We did not subject the components, blood pump and hollow fibers, to additional material testing, as we only used hand-crafted prototypes for proof of concept. However, this study clearly shows the importance of doing early in vivo tests to enable corrections of the design in the early development stage.
These considerations apply to extracorporeal lung support, but efforts to develop intravascular oxygenators (IVOXs) toward the artificial lung have been made. The first clinically applicable IVOX demonstrated significant gas exchange capabilities in an uncontrolled clinical study with 160 patients, but the device did not prevail clinically due to a high complication rate of 29%.24 Newer concepts integrating a miniaturized blood pump11 or a pulsating balloon in the center of the device25 could overcome the limitations and problems of IVOX.
The HEXMO demonstrated constant gas exchange and blood pumping performance over a period of 4 hours, which was sufficient to improve oxygenation in a model of hypoxic lung failure without causing hemodynamic instability or major hemoincompatibility. Prototype-typical problems with rotor-pump coupling and intradevice formation of blood clots were encountered in <20% and will have to lead to improvements in the design. This new device has demonstrated promising results suggesting further development and testing in various scenarios and over longer time periods.
Supported by START of RWTH Aachen University and Novalung Hechingen, Germany. The authors thank late Prof. Helmut Reul, Helmholtz Institute of Biomedical Technology, RWTH Aachen, who had had substantial intellectual input in the development of the device.
1. Kopp R, Dembinski R, Kuhlen R: Role of extracorporeal lung assist in the treatment of acute respiratory failure. Minerva Anestesiol 72: 587–595, 2006.
2. Peek GJ, Mugford M, Tiruvoipati R, et al: Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): A multicentre randomised controlled trial. Lancet 374: 1351–1363, 2009.
3. Brogan TV, Thiagarajan RR, Rycus PT, et al: Extracorporeal membrane oxygenation in adults with severe respiratory failure: A multi-center database. Intensive Care Med 35: 2105–2114, 2009.
4. Wendel HP, Ziemer G: Coating-techniques to improve the hemocompatibility of artificial devices used for extracorporeal circulation. Eur J Cardiothorac Surg 16: 342–350, 1999.
5. Mehlhorn U, Brieske M, Fischer UM, et al: LIFEBRIDGE: A portable, modular, rapidly available “plug-and-play” mechanical circulatory support system. Ann Thorac Surg 80: 1887–1892, 2005.
6. Dembinski R, Kopp R, Henzler D, et al: Extracorporeal gas exchange with the DeltaStream rotary blood pump in experimental lung injury. Artif Organs 27: 530–536, 2003.
7. Arlt M, Philipp A, Zimmermann M, et al: First experiences with a new miniaturised life support system for mobile percutaneous cardiopulmonary bypass. Resuscitation 77: 345–350, 2008.
8. Rossaint R, Pappert D, Gerlach H, et al: Extracorporeal membrane oxygenation for transport of hypoxaemic patients with severe ARDS. Br J Anaesth 78: 241–246, 1997.
9. Clement KC, Fiser RT, Fiser WP, et al: Single-institution experience with interhospital extracorporeal membrane oxygenation transport: A descriptive study. Pediatr Crit Care Med 11: 509–513, 2010.
10. Prodhan P, Fiser RT, Cenac S, et al: Intrahospital transport of children on extracorporeal membrane oxygenation: Indications, process, interventions, and effectiveness. Pediatr Crit Care Med 11: 227–233, 2010.
11. Cattaneo G, Strauss A, Reul H: Compact intra- and extracorporeal oxygenator developments. Perfusion 19: 251–255, 2004.
12. Kopp R, Bensberg R, Henzler D, et al: Hemocompatibility of a miniaturized extracorporeal membrane oxygenation and a pumpless interventional lung assist in experimental lung injury. Artif Organs 34: 13–21, 2010.
13. Strauss A, Cattaneo G, Henzler D, Reul H. Int J Artif Organs 27: 634, 2004 (abstract).
14. Richterich R (ed): Klinische Chemie: Theorie und Praxis. Basel, Switzerland, Karger, 1968.
15. Ravanat C, Freund M, Dol F, et al: Cross-reactivity of human molecular markers for detection of prethrombotic states in various animal species. Blood Coagul Fibrinolysis 6: 446–455, 1995.
16. 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 10: R151, 2006.
17. Svitek RG, Frankowski BJ, Federspiel WJ: Evaluation of a pumping assist lung that uses a rotating fiber bundle. ASAIO J 51: 773–780, 2005.
18. Bein T, Weber F, Philipp A, et al: A new pumpless extracorporeal interventional lung assist in critical hypoxemia/hypercapnia. Crit Care Med 34: 1372–1377, 2006.
19. Kopp R, Mottaghy K, Kirschfink M: Mechanism of complement activation during extracorporeal blood-biomaterial interaction: Effects of heparin coated and uncoated surfaces. ASAIO J 48: 598–605, 2002.
20. Dembinski R, Max M, Bensberg R, et al: High-frequency oscillatory ventilation in experimental lung injury: Effects on gas exchange. Intensive Care Med 28: 768–774, 2002.
21. Fries M, Bickenbach J, Henzler D, et al: S-100 protein and neurohistopathologic changes in a porcine model of acute lung injury. Anesthesiology 102: 761–767, 2005.
22. Muller T, Philipp A, Luchner A, et al: A new miniaturized system for extracorporeal membrane oxygenation in adult respiratory failure. Crit Care 13: R205, 2009.
23. Arlt M, Philipp A, Voelkel S, et al: Extracorporeal membrane oxygenation in severe trauma patients with bleeding shock. Resuscitation 2010;81: 804–809.
24. Zwischenberger JB, Tao W, Bidani A: Intravascular membrane oxygenator and carbon dioxide removal devices: A review of performance and improvements. ASAIO J 45: 41–46, 1999.
25. Budilarto SG, Frankowski BJ, Hattler BG, Federspiel WJ: Flow visualization study of a novel respiratory assist catheter. Artif Organs 33: 411–418, 2009.