Ambulatory Oxygenator Right Ventricular Assist Device for Total Right Heart and Respiratory Support

Ambulatory Oxygenator Right Ventricular Assist Device for Total Right Heart and Respiratory Support Dongfang Wang, MD, Scott D. Lick, MD, Xiaoqin Zhou, MD, Xiaojun Liu, MD, Robert J. Benkowski, MD, and Joseph B. Zwischenberger, MD Department of Surgery, University of Kentucky College of Medicine, Lexington, Kentucky; Department of Surgery, The University of Texas Medical Branch, Galveston, and MicroMed Cardiovascular, Inc, Houston, Texas Background. Our goal is ambulatory total respiratory and right heart assistance allowing a bridge to lung transplant. To that end, we have coupled a compact paracorporeal gas exchange device with a right ventricular assist device (RVAD) to create an OxyRVAD. Methods. Through a limited left thoracotomy, the main pulmonary artery (PA) and right atrium (RA) were exposed in 5 anesthetized sheep. After a systemic heparin bolus, a 12-mm outer diameter crimped graft glued to tubing was anastomosed (end to side) to the main PA and a VAD atrial cannula was placed through the RA appendage. The chest was drained and closed, then the PA graft flowed at 1 to 2 L/min as a shunt to the RA overnight. The next day, the animal was reanticoagulated, and the shunt cannulae clamped and divided. The OxyRVAD unit, consisting of commercially available components including an axial flow pump and lowresistance cardiopulmonary bypass gas exchange device, was interposed. Pumping from RA to PA was maintained at 3 L/min. Results. Five consecutive sheep survived the implant, and stood and ate normally after initiation of the OxyR- VAD. Three survived the full 2-week study, and a fourth was sacrificed on day 13 owing to a storm-related power failure. For these 4 sheep, pump flow was stable at 3 L/min. Carbon dioxide removal was constant and total during the experiment at 200 19 ml/min. Oxygen transfer was 144 44 ml/min. One sheep had progressive thrombocytopenia and was sacrificed on day 5 after implant. Conclusions. Our ambulatory OxyRVAD can provide total assistance for the right heart and lungs in normal awake sheep for 14 days. (Ann Thorac Surg 2007;84:1699 703) 2007 by The Society of Thoracic Surgeons Our group previously developed a pumpless paracorporeal artificial lung in a pulmonary artery to pulmonary artery (PA-PA) in-series configuration. This pumpless paracorporeal artificial lung used a snare between two grafts anastomosed to the proximal and distal PA to divert total proximal PA blood flow to the gas exchange device, then back to the distal PA and the native lungs. This configuration preserves native lung perfusion and metabolic function [1 4]. Our paracorporeal artificial lung was applied in a smoke and burn induced lethal adult respiratory distress syndrome sheep model and demonstrated successful total respiratory support for 5 days, with improved survival [5]. However, there are two major limitations for a PA-PA configuration in end-stage human lung disease. First, the human main pulmonary artery is too short (2.5 cm versus 6 cm in sheep) to place two large anastomoses for paracorporeal artificial lung graft attachment [6]. Second, patients with Accepted for publication March 21, 2007. Presented at the Basic Science Forum of the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8 11, 2006. Address correspondence to Dr Zwischenberger, Department of Surgery, University of Kentucky College of Medicine, 800 Rose St, MN-264, Lexington, KY 40536-0298; e-mail: j.zwische@uky.edu. end-stage respiratory disease usually have some degree of pulmonary hypertension and right heart failure, which are aggravated by an in-series PA-PA paracorporeal circuit. Our goal is to develop an ambulatory oxygenator right ventricular assist device (OxyRVAD): a compact paracorporeal gas exchange device coupled with an RVAD for ambulatory total respiratory and right heart support for weeks to serve as a bridge to lung transplant [7, 8]. During a recent survey of 31 lung transplant program directors, most thought 4 weeks of artificial lung support would prove meaningful as a bridge to lung transplant, especially if organ allocation would be prioritized [9]. Such a device should allow lung and right heart recovery, ambulation, and rehabilitation to improve the patient as a lung transplant recipient, much the way a long-term left ventricular assist device does for a heart transplant candidate. This version of our OxyRVAD utilizes commercially available components, which may accelerate clinical availability. Dr Benkowski discloses that he has a financial relationship with MicroMed Cardiovascular, Inc. 2007 by The Society of Thoracic Surgeons 0003-4975/07/$32.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2007.03.068

1700 WANG ET AL Ann Thorac Surg AMBULATORY OXYGENATOR RVAD 2007;84:1699 703 Material and Methods Our OxyRVAD system includes a commercially available small, portable axial blood pump and a low-resistance membrane gas exchanger. The blood pump is a MicroMed axial flow device (MicroMed, Houston, Texas) capable of 10 L/min blood flow against systemic resistance when used in a left ventricular assist device configuration [10]. One of the authors (R.B.) helped develop the circuit as a collaborator. Although he is employed by MicroMed, we had complete freedom of investigation and of analyzing and reporting the results. The gas exchange device is a low-resistance Affinity Trillium (Medtronic, Minneapolis, Minnesota) coated cardiopulmonary bypass oxygenator [11, 12]. Animal Study All animals received care according to the Guide for the Care and Use of Laboratory Animals prepared by the US Department of Health and Human Services and published by the National Institutes of Health (NIH publication 85-23, revised 1985). The study was approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Texas Medical Branch, Galveston, with strict adherence to the IACUC guidelines regarding humane use of animals. Our management of the sheep parallels our standards for patient care. Our team, including a veterinary anesthesiologist, provides 24-hour bedside care, 7 days a week. Animal Resource Center personnel, with no conflict of interest, make daily rounds to check animal management protocol adherence. Our OxyRVAD system was applied in 5 adult female Suffolk sheep (3 to 4 years old, weighing 35 to 45 kg). After initial sedation with 12.5 mg/kg intramuscular ketamine and mask inhalation 4% halothane (Ohmeda 7000; BOC Health Care, Liberty Corner, New Jersey), sheep were intubated with a 10-mm-diameter endotracheal tube. Anesthesia was maintained with 1% to 2.5% halothane titrated to a heart rate of 75 to 120 beats per minute during surgery. Our operative sterile technique was identical to that used for human procedures. The sheep s neck and groin were prepared and draped in the supine position. A Swan-Ganz catheter (Baxter Health Care, Edwards Critical Care Division, Irvine, California) was introduced percutaneously through the right external jugular vein for PA and central venous pressure monitoring, blood temperature measurement and intravenous infusion. A 16-gauge catheter (Intracath; Becton- Dickinson, Sandy, Utah) was inserted into the right femoral artery through a cut-down for arterial blood pressure monitoring and blood gas sampling. The sheep were repositioned with the left side up, and a musclesparing fourth-interspace left thoracotomy was performed. Pericardial retraction sutures were used to deliver the right atrial appendage into the operative field. After a bolus of intravenous heparin (100 IU/kg), an infusion cannula (3/8-inch outlet cannula with 12-mmdiameter crimped vascular graft) was anastomosed to the main PA (end of graft to side of pulmonary artery). The right atrial appendage was cannulated through a pursestring stitch using an atrial VAD cannula (Thoratec, Pleasanton, California). Both cannulae were brought out through separate stab wounds in the chest wall. The cannulae were connected, allowing PA blood to flow backward to the RA, using a c-clamp to limit flow to 1 to 2 L/min as monitored by a flow probe. A pleural drain was placed and the chest wound closed. The animal was transferred to the intensive care unit and allowed to recover with no further anticoagulation. This staged approach to implantation was used to reverse the systemic anticoagulation and minimize perioperative bleeding. The next morning, the animal was reanticoagulated (heparin 100 IU/kg intravenously), the shunt clamped and divided, and the OxyRVAD system (MicroMed pump and Affinity gas exchange device) interposed. The axial pump inlet was connected to the right atrial VAD cannula, the pump outlet was connected to the gas exchange device, the outlet of which was connected to the crimped vascular graft, perfusing the main PA. The pump was turned on, and venous blood from the RA was pumped into the Fig 1. Sheep with ambulatory paracorporeal oxygenator right ventricular assist device (OxyRVAD), recovered from anesthesia and surgery with free access to water and food.

Ann Thorac Surg WANG ET AL 2007;84:1699 703 AMBULATORY OXYGENATOR RVAD 1701 gas exchange device, and returned to the pulmonary artery system. Sweep gas was 100% oxygen at approximately 6 L/min. A continuous heparin infusion was titrated to activated clotting time of 200 to 300 s. For the duration of the study, the animals were allowed to stand, with free access to food and water (Fig 1). Hemodynamic variables were continuously monitored using a HP 78534B monitor, including central venous pressure, mean arterial pressure, and pulmonary arterial pressure. Circuit blood flow was continuously monitored by a Transonic 9XL tubing flowsensor and HT110 flowmeter (Transonic System, Ithaca, New York). Systemic arterial, pre, and post gas exchange device blood gasses, and exhaust sweep gas were analyzed by Synthesis 15 (Instrumentation Laboratory, Lexington, Massachusetts). The CO 2 concentration in the exhaust sweep gas was directly measured, then expressed as volume percentage. Before and after gas exchange device pressures were measured and documented every 24 hours. Sweep gas flow was regulated by an oxygen flowmeter (Datex- Omeda, Madison, Wisconsin). Since the sweep gas is pure oxygen, exhaust sweep gas CO 2 equals CO 2 removed, derived from the following formula: CO 2 removal (ml min) sweep gas flow CO 2 concentration in exhaust sweep gas. Oxygen transfer was calculated from the change of blood O 2 saturation across the gas exchange device, blood flow, and hemoglobin (Hb) level: O 2 transfer 1.34 Hb 10 delta O 2 sat Qblood 0.003 deltapao 2 Results All 5 sheep survived the OxyRVAD implantation and were transferred to the intensive care unit. Four remained healthy during the 2-week study, eating and drinking normally. One of these 4 was sacrificed at the Fig 2. Gas exchange: carbon dioxice removal (solid circles) and oxygen transfer (open circles) across the gas exchange device. Fig 3. Plasma-free hemoglobin throughout the experiment. (Solid line animal 1; short dash line animal 2; long dash line animal 3; dotted line animal 4.) end of day 13 because of a storm-related power failure, and the others on day 14. The fifth animal began the study thrombocytopenic (platelet count 56,000/cc), with continuously worsening thrombocytopenia (to 24,000) and diffuse bleeding and was sacrificed on day 5 of the study. Although this sheep was consistent with the other animals, it was censored from the following data owing to early withdrawal. Results on the remaining 4 sheep follow. Hemodynamics remained stable for the 2 weeks, with no need for inotropic or vasopressor support. The OxyR- VAD pump flow remained stable throughout the experiment, at approximately 3 L/min. Gas Exchange Device Performance Carbon dioxide removal remained relatively constant during the experiment, at 200 19 ml/min (Fig 2). The O 2 transfer was more variable throughout the study, at 144 43 ml/min, and appeared to vary directly with increasing level of sheep activity (Fig 2). Upon examination of the gas exchange devices, there was a thin layer of thrombus in the distal bottom of each gas exchanger. In one case, the pump stopped owing to a storm-caused power failure on day 13, but was working normally until then. In another, pump flow decreased on day 14, shortly before planned sacrifice. This sheep remained stable while surviving on native right heart and lung support. Analysis of the pumps was performed after the experiments were terminated. The one failed pump (an older uncoated version of the current coated VAD) had a large thrombus causing decreased performance. The thrombus presumably came from the pump adapters that connect the pump housing to the tubing. The other pumps did not exhibit any thrombus that altered pump performance. Blood loss during surgery was less than 50 ml. Hemoglobin remained 8 to 9 mg/dl throughout the experiment (sheep normal is 10 mg/dl). Chest drainage the first few hours after implant became clear and remained so while

1702 WANG ET AL Ann Thorac Surg AMBULATORY OXYGENATOR RVAD 2007;84:1699 703 the tubes were in place, usually 7 days. Plasma free hemoglobin was usually approximately 20 mg/ml (Fig 3). Generated pump pressure was approximately 80 mm Hg. Trans gas exchange device pressure drop (pre gas exchange device mean pressure mean PA pressure) remained approximately 60 mm Hg at 3 L/min blood flow. Comment Unlike dialysis, which functions as a bridge to renal transplantation, or ventricular assist devices, which serve as a bridge to cardiac transplantation, no suitable ambulatory bridge to lung transplantation exists. Mechanical ventilation allows only partial support, and is limited by the ventilatory and gas exchange capabilities of the damaged native lung. Intravenacaval devices for intravascular oxygenation [13] are nonambulatory, and are surface-area limited, and thus do not provide enough gas exchange for near-total support. Pumpless peripheral lung support using arteriovenous CO 2 removal eliminates near total CO 2 production, allowing gentle ventilation to treat adult respiratory distress syndrome [14 17], but does not adequately treat hypoxia. Extracorporeal membrane oxygenation (ECMO) can provide total gas exchange for weeks, resulting in recovery from severe respiratory failure in 90% of newborn infants, 50% of children, and 40% of adults in more than 30,000 patients to date [18, 19]. Although occasionally successful for persistent respiratory failure and as a bridge to lung transplant [20, 21], ECMO is too complex, bulky, costly, labor intensive, and blood traumatic for routine bridge-to-transplant use [22]. The ECMO systems, whether in a venoarterial (VA ECMO) or venovenous (VV ECMO) configuration, are traumatic to blood because of a bulky silicone membrane gas exchanger with high blood resistance; long tubing connecting the patient to the unit, and high-resistance peripheral blood vessel cannulae. Consequently, a powerful pump is needed to generate high pressure to drive flow, resulting in significant blood trauma and sheer. For lung recipients with primary graft failure, VV ECMO allows blood flow through the transplant lungs, and appears to provide better outcomes with fewer complications than VA ECMO [23]. However, VV ECMO is subject to recirculation of already oxygenated blood (mixing) and incomplete capture of blood at the entry to the right heart (shunting). Also, VV ECMO requires normal right ventricular function. Both mixing and shunting are minimized with the OxyRVAD: mixing is prevented by the pulmonic and tricuspid valves, and shunting is minimized by efficient capture of right atrial blood using a large-bore atrial cannula connected to the axial flow pump. Finally, ECMO in any configuration is labor intensive, nonambulatory, and requires frequent blood transfusions. An artificial lung can be designed to be placed in parallel or in series with the native pulmonary circulation [24, 25]. The in-parallel configuration (PA to left atrium [LA]) has the lowest total strain on the right heart, because the gas exchange device acts as a pop-off valve to the pulmonary circulation; therefore, the higher the native lung resistance, the more blood is shunted through the artificial lung. However, this configuration allows only partial support as the percent shunted blood is variable, and during low flow circuit, thrombosis can occur. This configuration also allows a direct shunt into the systemic circulation, increasing the probability of systemic emboli. The in-series configuration (PA-PA) captures all the pulmonary blood, but has the highest right heart strain. As most patients awaiting a lung transplant will have some right heart strain, the in-series configuration has a likelihood of right heart failure in the clinical setting. In either configuration (PA-PA or PA- LA), the addition of an inflow compliance chamber can lower impedance and improve right heart function; unfortunately, this greatly adds to the complexity of the design [26, 27]. A hybrid configuration has been proposed, with blood flow from the proximal PA to both the distal PA and the LA, to achieve lower right heart strain and preserve lung perfusion [28]. The problems associated with the variable shunt, resulting in frequent incomplete gas exchange and systemic emboli in the case of the in-parallel configuration, or the right heart strain, resulting in right heart failure in the case of the in-series configuration, or a combination of both problems with the hybrid configuration, have directed our group to pursue a right atrial to pulmonary artery pump-assisted artificial lung design, which we term OxyRVAD. The OxyRVAD design allows total right heart and respiratory support with total blood flow to the native lung. Our choice to place the OxyRVAD circuit in a paracorporeal orientation parallels the development of cardiac assist devices and allows continuous monitoring of the individual components, with simple device change-out when failure occurs. We have developed our new OxyRVAD system to address the shortcomings of both the in-parallel and in-series configurations and ECMO. Like ECMO, the OxyRVAD provides total respiratory support. Our previously developed pumpless, ambulatory PA-PA artificial lung provided complete gas exchange, but did not address the pulmonary hypertension and right heart failure commonly found in end-stage lung disease, and also requires a main PA longer than that found in humans. Unlike ECMO, the OxyRVAD allows ambulation and should be much less traumatic to the blood. The OxyRVAD is a low-resistance circuit attached directly to the heart (right atrium) and large vessel (main PA) by short, large-bore VAD-type cannulae. Our compact axial pump is powerful enough to drive the total cardiac output through this short, low-resistance circuit. A pulseless, axial flow pump was chosen, instead of a pulsatile device, to eliminate the need for a compliance chamber and simplify overall design. The Affinity gas exchange device was chosen because it has a relatively low resistance for a cardiopulmonary bypass oxygenator, as reflected by our transoxygenator pressure drop of approximately 60 mm Hg at 3 L/min.

Ann Thorac Surg WANG ET AL 2007;84:1699 703 AMBULATORY OXYGENATOR RVAD 1703 Unlike ECMO, the OxyRVAD uses VAD-type cannulae, which allow for ambulation, and hence patient rehabilitation. The large-bore Thoratec venous cannula used in the OxyRVAD is externally textured, allowing tissue ingrowth. Such VAD cannulae have been shown in the bridge-to-heart-transplant experience to resist infection for many months. We chose to limit this proof-of-concept study to 2 weeks because of cost and manpower limitations. The separate paracorporeal components of the OxyRVAD (blood pump and gas exchange device) could be simply and safely replaced as needed without risk of systemic embolization because the entire circuit is confined to the right circulation. As the axial flow pump, gas exchange device, and cannulae were not designed for each other, all connections are manual and potential foci of thrombosis. Moreover, heparin-bonded surface technology was not used for the cannulae, tubing, or connections. In our next phase of development, we plan to use commercial-quality component connections, or possibly a lowresistance gas exchange device specifically designed for this purpose. In conclusion, we have shown the feasibility of an ambulatory compact OxyRVAD system, using a compact axial pump coupled to a low-resistance gas exchange device, to achieve 14-day total respiratory and right heart support in healthy large animals. Further refinement of the components could allow longer-term support, toward a goal of a stable 4-week (or longer) bridge to lung transplant [29, 30]. Supported in part by a National Institutes of Health STTR Grant and Shriners Hospital of North America, and contracts with MicroMed Cardiovascular, Inc, Houston, Texas. References 1. Lick SD, Zwischenberger JB, Alpard SK, Witt SA, Deyo DM, Merz SI. Development of an ambulatory artificial lung in an ovine survival model. ASAIO J 2001;47:486 91. 2. Lick SD, Zwischenberger JB, Wang D, Deyo DJ, Alpard SK, Chambers SD. Improved right heart function with a compliant inflow artificial lung in series with the pulmonary circulation. Ann Thorac Surg 2001;72:899 904. 3. Lick SD, Deyo DJ, Wang D, et al. Paracorporeal artificial lung: perioperative management for survival study in sheep. J Invest Surg 2003;16:177 84. 4. Zwischenberger JB, Anderson CM, Cook KE, Lick SD, Mockros LF, Bartlett RH. Development of an implantable artificial lung: challenges and progress. ASAIO J 2001;47:316 20. 5. Zwischenberger JB, Wang D, Lick SD, Deyo DJ, Alpard SK, Chambers SD. The paracorporeal artificial lung improves 5-day outcomes from lethal smoke/burn-induced acute respiratory distress syndrome in sheep. Ann Thorac Surg 2002;74:1011 6. 6. Harper DD, Alpard SK, Deyo DJ, Lick SD, Traber DL, Zwischenberger JB. Anatomic study of the pulmonary artery as a conduit for an artificial lung ASAIO J 2001;47:34 6. 7. Lick SD, Zwischenberger JB. Artificial lung: bench toward bedside. ASAIO J 2004;50:2 5. 8. Schmalsteig F, Zwischenberger JB. Artificial lung inflammatory interface: a refocus of the problem. ASAIO J 2004;50:6 8. 9. Haft JW, Griffith BP, Hirschel RB, Bartlett RH. Results of an artificial-lung survey to lung transplant program directors. J Heart Lung Transplant 2002;21:467 73. 10. Song X, Throckmorton AL, Untaroiu A, et al. Axial flow blood pumps. ASAIO J 2003;49:355 64. 11. Goodin MS, Thor EJ, Haworth WS. Use of computational fluid dynamics in the design of the Avecor Affinity oxygenator. Perfusion 1994;9:217 22. 12. Palanzo DA, Zarro DL, Montesano RM, et al. Effect of trillium biopassive surface coating of the oxygenator on platelet count drop during cardiopulmonary bypass. Perfusion 1999;14:473 9. 13. Conrad SA, Zwischenberger JB, Eggerstedt JM, Bidani A. In vivo gas transfer performance of the intravascular oxygenator in acute respiratory failure. Artif Organs 1994;18:840 5. 14. Alpard SK, Bidani A, Conrad SA, Zwischenberger JB. Arteriovenous carbon dioxide removal. ASAIO J 1998;44:223 4. 15. Zwischenberger JB, Conrad SA, Alpard SK, Grier LR, Bidani A. Percutaneous extracorporeal arteriovenous CO2 removal for severe respiratory failure. Ann Thorac Surg 1999;68: 181 7. 16. Alpard SK, Zwischenberger JB, Tao W, Deyo DJ, Bidani A. Reduced ventilator pressure and improved P/F ratio during percutaneous arteriovenous carbon dioxide removal for severe respiratory failure. Ann Surg 1999;230:215 24. 17. Conrad SA, Zwischenberger JB, Grier LR, Alpard SK, Bidani A. Total extracorporeal arteriovenous carbon dioxide removal in acute respiratory failure: a phase I clinical study. Intens Care Med 2001;27:1340 51. 18. Bartlett RH. Extracorporeal life support: history and new directions. Semin Perinatol 2005;29:2 7. 19. Extracorporeal life support organization (ELSO). Registry report, July 2006. 20. Jurmann MJ, Schaefers HJ, Demertzis S, Haverich A, Wahlers T, Borst HG. Emergency lung transplantation after extracorporeal membrane oxygenation. ASAIO J 1993;39: M448 52. 21. Demertzis S, Haverich A, Ziemer G, et al. Successful lung transplantation for posttraumatic adult respiratory distress syndrome after extracorporeal membrane oxygenation support. J Heart Lung Transplant 1992;11:1005 7. 22. Zwischenberger JB, Nguyen TT, Upp JR Jr, et al. Complications of neonatal extracorporeal membrane oxygenation. Collective experience from the Extracorporeal Life Support Organization. J Thorac Cardiovasc Surg 1994;107:838 48. 23. Hartwig M, Appel JZ III, Cantu E III, et al. Improved results treating lung allograft failure with venovenous extracorporeal membrane oxygenation. Ann Thorac Surg 2005;80: 1872 80. 24. Fazzalari FL, Bartlett RH, Bonnell MR, Montoya JP. An intrapleural lung prosthesis: rationale, design, and testing. Artif Organs 1994;18:801 5. 25. Zwischenberger JB, Alpard SK. Artificial lungs: a new inspiration. Perfusion 2002;17:253 68. 26. Cook KE, Perlman CE, Seipelt R, Backer CL, Mavroudis C, Mockros LF. Hemodynamic and gas transfer properties of a compliant thoracic artificial lung. ASAIO J 2005;51:404 11. 27. Haft JW, Alnajjar O, Bull JL, Bartlett RH, Hirschl RB. Effect of artificial lung compliance on right ventricular load. ASAIO J 2005;51:769 72. 28. Boschetti F, Perlman CE, Cook KE, Mockros LF. Hemodynamic effects of attachment modes and device design of a thoracic artificial lung. ASAIO J 2000;46:42 8. 29. Zwischenberger JB. Artificial lung. The next 50 years of extracorporeal circulation. Proc Am Acad Cardiovasc Perfusion 2003;24:76 81. 30. Zwischenberger JB. Future of artificial lungs. ASAIO J 2004; 50:xlix-li.