The Paracorporeal Artificial Lung Improves 5-Day Outcomes From Lethal Smoke/Burn-Induced Acute Respiratory Distress Syndrome in Sheep

Transcription

1 The Paracorporeal Artificial Lung Improves 5-Day Outcomes From Lethal Smoke/Burn-Induced Acute Respiratory Distress Syndrome in Sheep Joseph B. Zwischenberger, MD, Dongfang Wang, MD, PhD, Scott D. Lick, MD, Donald J. Deyo, DVM, Scott K. Alpard, BS, and Sean D. Chambers, PhD The University of Texas Medical Branch, Galveston, Texas, and MC3 Corporation, Ann Arbor, Michigan Background. Our low-impedence, paracorporeal artificial lung (PAL) prototype is well-tolerated in-series with the normal sheep pulmonary circulation. Using our lethal dose 80% to 100% smoke/burn acute respiratory distress syndrome (ARDS) sheep model, we compared PAL to volume-controlled mechanical ventilation (VCMV) in a prospective, randomized, controlled, unblinded, 5-day outcome study. Methods. Fourteen sheep were randomized to PAL (n 8) versus VCMV (n 6) to assess outcome. For PAL, arterial cannulas were anastomosed to the proximal and distal main pulmonary artery with an interposing snare diverting full flow through a paracorporeal loop. Acute respiratory distress syndrome was induced in both groups (48 breaths smoke insufflation, third degree burn on 40% of total body surface area). When acute respiratory distress syndrome criteria were met (24 to 30 hours after injury), the PAL was interposed in the paracorporeal loop. Both groups were managed with a VCMV algorithm minimizing tidal volume, ventilator rate, and fractional inspired concentration of oxygen (FiO 2 ). Results. Six of eight PAL versus 1 of 6 VCMV sheep survived the 5-day study. In PAL, cardiac output, mean arterial pressure, pulmonary artery pressure, left atrial pressure, and central venous pressure remained stable. Average PAL gas transfer was ml/min O 2 and ml/min CO 2. Ventilator settings 48 hours after lung injury in PAL were significantly lower (p < 0.05) than VCMV (TV 210 versus 425 ml; respiratory rate 6 versus 29 breaths/min; minute ventilation 1.2 versus 10.8 L/min; FiO 2 21 versus 100%). Likewise, PaO 2 / FiO 2 ratio was normalized in PAL and still met acute respiratory distress syndrome criteria in VCMV. The PAL wet/dry ratio was significantly lower than VCMV ( versus ; p 0.008). Conclusions. In a prospective, randomized, controlled, unblinded, outcomes study, PAL decreased ventilatorinduced lung injury in a lethal dose 80% to 100% ARDS model to improve 5-day survival. (Ann Thorac Surg 2002;74:1011 8) 2002 by The Society of Thoracic Surgeons The need for an artificial lung as a bridge to transplantation or recovery persists as demand for donor lungs greatly exceeds supply with a 2-year average wait to transplant and a 30% wait-list mortality. Although mechanical ventilation and extracorporeal membrane oxygenation have been used successfully as bridges to transplant [1 3], each has limitations. Mechanical ventilation allows only partial support, limited by the ventilatory and gas exchange capabilities of the damaged lung parenchyma. Extracorporeal membrane oxygenation is labor-intensive, expensive, time-limited (weeks), and nonambulatory. In collaboration with MC3 Corporation (Ann Arbor, MI), our group has been developing a low-resistance, low-impedance paracorporeal artificial lung (PAL) prototype designed for weeks to months of support in a survival model of implantation in healthy sheep [4 6]. Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 30, Address reprint requests to Dr Zwischenberger, Division of Cardiothoracic Surgery, Department of Surgery, 301 University Blvd, Galveston, TX ; jzwische@utmb.edu. The PAL has undergone several modifications, with improvements in housing and attachments, an inlet compliance chamber, an inlet blood flow separator, and modified outlet geometry [5]. The result is a PAL prototype that functions reliably for up to 7 days. Having achieved the goal of survival in healthy sheep, the next step in development is to test the PAL in a large animal model of severe lung injury. During the past several years, our group has developed a sheep model of acute respiratory distress syndrome (ARDS) using combined smoke inhalation with a third degree cutaneous burn [7]. This model involves a 40% total body surface area third degree burn and varying numbers of smoke breaths from a modified bee smoker through a tracheostomy. The severity of lung injury is proportional to the number of breaths: 36 breaths cause moderately severe ARDS (lethal dose 50% or LD 50 ), whereas 48 breaths causes severe (lethal dose 80% to 100% or LD 100 ) lung injury. Outcome studies with percu- The authors disclose that they have a financial relationship with MC3 Corp by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 1012 ZWISCHENBERGER ET AL Ann Thorac Surg ARTIFICIAL LUNG IN ARDS 2002;74: taneous arterial venous carbon dioxide removal (AVCO 2 R) demonstrated that AVCO 2 R allowed a reduction in minute ventilation and peak inspiratory pressures [8], lessened barotrauma-induced ARDS [9], and conferred a survival advantage in a LD 50 smoke/burn ARDS model [10]. We applied the PAL to our LD 100 smoke/burn ARDS sheep model in a prospective, randomized, controlled, unblinded outcomes study comparing PAL to volumecontrolled mechanical ventilation (VCMV) for 5 days. Material and Methods All animals received humane care according to Guide for the Care and Use of Laboratory Animals (1996) prepared by the U.S. Department of Health and Human Services and published by the National Institutes of Health (NIH publication 85-23, revised 1996). The study was approved by the Institutional Animal Care and Use Committee of The University of Texas Medical Branch, Galveston, Texas, with strict adherence to the Institutional Animal Care and Use Committee guidelines regarding humane use of animals. Our management of the sheep parallels our standards of patient care. The investigators make daily rounds on the sheep to outline and supervise management. Our management team also consists of a full-time staff veterinary anesthesiologist (DJD) who oversees all anesthesia, sedation, and animal management issues. Medical students volunteer for 24-hour cage-side care 7 days per week. Institutional Animal Care and Use Committee personnel, with no conflict of interest, make daily rounds to check compliance with the animal management protocol. The developmental evolution of the current PAL prototype has been previously described by our group and other investigators [4, 5, 11, 12]. The experimental design of this study was a prospective, randomized, controlled, unblinded outcomes study comparing PAL to VCMV was 5 days or 105 hours (Monday morning through Friday afternoon) (Fig 1). Experimental (PAL) Group (n 8) MONDAY MORNING. Adult free-range ewes undergo general anesthesia induced by intramuscular ketamine (7 to 15 mg/kg) and isofluorane by mask, then are blindly orotracheally intubated and maintained by 4.0% to 5.0% isoflourane titrated to mean arterial pressure of 70 to 110 mm Hg. A tracheostomy is performed. A groin cutdown allows femoral artery and pulmonary artery catheter (Edwards Critical Care, Irvine, CA) placement for measurement of systemic and mean blood pressures and right heart pressures plus thermodilution cardiac output, respectively. The sheep is given 5,000 U of heparin intravenously. Through a left fourth interspace thoracotomy, 16- or 18-mm Dacron arterial grafts bonded to 5/8 in inside diameter silicone tubing are anastomosed end-ofgraft to side-of proximal and distal pulmonary artery. The cannulas are tunneled through a small second left thoracotomy (sixth rib resection) out the skin and attached to a silicone tubing C-loop. The long ovine main Fig 1. Experimental design of the prospective, randomized, controlled, unblinded outcomes study comparing paracorporeal artificial lung (PAL) to volume-controlled mechanical ventilation (Control) in a lethal dose 80% to 100% (LD 100 ) smoke/burn acute respiratory distress syndrome (ARDS) sheep model. (FiO 2 fractional inspired concentration of oxygen; PaO 2 arterial oxygen tension; PA PA pulmonary artery to pulmonary artery.) pulmonary artery, averaging 5.5 cm [13], allows room for a snare between anastomoses. The pulmonary artery snare is tightened, diverting full blood flow through the artificial lung. A left atrial pressure catheter is placed, and the wound is closed over a pleural and pericardial drain. A custom-made Doppler flow probe (Transonic Inc, Ithaca, NY) is placed on the PAL outflow graft to measure flow. Protamine sulfate is titrated to reverse the heparin effect as determined by a normal activated clotting time. Then, under the same general anesthetic, a 40% total body surface area third-degree cutaneous burn is made on both flanks using a propane torch against a template, followed by 48 positive-pressure breaths of smoke from a burning cotton towel delivered through a modified beesmoker by the tracheostomy [7]. The animals are returned to their cages and allowed to awaken. All animals receive mechanical ventilation as previously described for ARDS model development [6]. Intravenous fluids are administered according to the Parkland burn resuscitation formula: Lactated Ringer s solution 4 ml kg 1 h 1 times percentage total body surface area burn for the first 24 hours; half given in the first 8 hours, the rest during the next 16 hours. After 24 hours, Lactated Ringer s is decreased to 1 to 2 ml kg 1 h 1 and titrated downward depending on adequacy of oral intake and urine output. TUESDAY MORNING. All sheep met ARDS criteria before 14:00 (arterial oxygen tension [PaO 2 ]/fractional concentration of oxygen [FiO 2 ] ratio 200). Once ARDS criteria are met, the sheep is given 5,000 U of intravenous heparin, and undergoes a brief mask inhalation anesthetic (isofluorane). The pulmonary artery snare is loosened, and the external portions of the cannulas are clamped. The C-loop is removed, and the PAL interposed

3 Ann Thorac Surg ZWISCHENBERGER ET AL 2002;74: ARTIFICIAL LUNG IN ARDS 1013 distress (systolic blood pressure 60 mm Hg or heart rate 40 beats/min for 1 hour, or PaCO mm Hg or PaO 2 45 mm Hg) was euthanized. FRIDAY 16:00. All remaining sheep are euthanized as described. Autopsies are performed on all animals, with attention toward detecting pulmonary or systemic emboli. Lung wet/dry weights are measured. Fig 2. Illustration of sheep with paracorporeal artificial lung. in its place. The cannulas are unclamped, the snare is again tightened (diverting full blood flow through the PAL), and the animal is allowed to awaken (Fig 2). Intravenous heparin anticoagulation is continued for the duration of the study, titrating the heparin dose to an activated clotting time of 200 to 300 seconds. The sheep remain on a ventilator, treated with VCMV as a lung protective strategy to minimize volutrauma, according to our previously published ventilator management protocol [10]. Briefly, the protocol decreases the initial tidal volume (10 ml/kg) by 20% increments to achieve a peak inspiratory pressure less than 30 mm H 2 O. Respiratory rate is decreased to achieve arterial carbon dioxide tension (PaCO 2 ) less than 40 mm Hg, and FiO 2 is decreased to maintain PaO 2 at more than 60 mm Hg. FRIDAY 16:00. All remaining sheep are euthanized after induction of general anesthesia with intravenous ketamine (8.8 to 14 mg/kg) and xylazine (2 to 4 mg/kg) followed by intravenous potassium chloride. Each PAL prototype is then flushed and shipped to MC3 Corporation to be evaluated for leaks. Results Six of 8 PAL sheep and 1 of 6 control (VCMV) sheep survived the full 5-day, 105-hour (Monday morning through Friday afternoon) study period (Fig 3). In the PAL group, one died at 48 hours due to a sudden pulmonary hypertensive crisis. The other died at 84 hours due to sepsis from pneumonia (hypotension, hypothermia, bradycardia, copious foul-smelling purulent airway secretions, and airway purulence at autopsy). Four control animals were euthanized upon meeting distress criteria of hypoxia (PaO 2 45 mm Hg) at 42, 72, 74, and 96 hours, and one control animal was euthanized upon meeting the bradycardia and hypotension criteria at 96 hours. There were no PAL device failures. There were no reoperations for bleeding, although one sheep in the PAL group required blood transfusion. There were no clinically significant burn wound infections. The pressure gradient across the PAL, measured by twice daily 27- gauge needle sticks through the inflow and outflow cannulas, was mm Hg (0.9 Wood units) with L/min average blood flow. Average PAL gas transfer was ml/min O 2 and ml/min CO 2 (Fig 4). All ventilator settings at 48 hours after smoke/burn injury in the PAL group were significantly (p 0.05) lower than the VCMV group (tidal volume 210 versus 425 ml; respiratory rate 6 versus 29 breaths/min; minute ventilation 1.2 versus 10.8 L/min; FiO 2 21 versus 100%) (Table 1). Likewise, at 48 hours of treatment, the PaO 2 /FiO 2 ratio was normalized in PAL and still met ARDS criteria in VCMV (Fig 5). PAL wet/dry ratio was Control (VCMV) Animals (n 6) MONDAY MORNING. Sheep are anesthetized, intubated, a tracheostomy performed, and the femoral vessels cannulated as in the PAL group. The sheep then undergo a 40% total body surface area burn and 48 breaths of positivepressure smoke insufflation identical to the PAL group. The sheep remain on a ventilator for the duration of the study, according to the same VCMV algorithm as the PAL group. TUESDAY MORNING. All sheep met ARDS criteria before 14:00 of the second day. Once ARDS criteria are met, the sheep are given 5,000 U of intravenous heparin and maintained with a continuous heparin infusion (titrated to activated clotting time 230 to 300 seconds) until completion of the study. Throughout the study, any animal judged to be in Fig 3. Survival curves during the 5-day study comparing paracorporeal artificial lung (PAL) to volume-controlled mechanical ventilation (Control). (ARDS acute respiratory distress syndrome.)

4 1014 ZWISCHENBERGER ET AL Ann Thorac Surg ARTIFICIAL LUNG IN ARDS 2002;74: Fig 4. Paracorporeal artificial lung (PAL) gas transfer throughout the 5-day study. Calculated from device inflow and outflow blood gases and paracorporeal artificial lung blood flow. (ARDS acute respiratory distress syndrome.) Fig 5. Arterial oxygen tension (PaO 2 )/fractional concentration of oxygen (FiO 2 ) ratio in paracorporeal artificial lung (PAL) versus volume-controlled mechanical ventilation (Control) throughout 5-day study. (ARDS acute respiratory distress syndrome.) also significantly lower than the VCMV controls ( versus ; p 0.008). Autopsies showed no pulmonary or systemic emboli in any animal from either group. Hemodynamic parameters remained stable in the PAL group throughout the study (Fig 6). In the VCMV control group, the average transpulmonary gradient (mean pulmonary artery pressure pulmonary capillary wedge pressure) went from 7 to 11 cm H 2 O, cardiac output remained stable at an average of 4.7 L/min, and pulmonary vascular resistance (transpulmonary gradient/ cardiac output) increased from 1.5 to 2.4 Wood units after smoke/burn injury. In the PAL group, transpulmonary gradient (mean pulmonary artery pressure downstream of the PAL PCWP) before and after interposition remained between 7 and 8 cm H 2 O, and average cardiac output with the PAL remained stable at 4.0 L/min. However, direct comparison of cardiac output or pulmonary vascular resistance between the two groups is impossible because cardiac output in the PAL group was measured by a flow probe on the PAL outflow cannula, whereas cardiac output in the closed-chest VCMV controls was measured by pulmonary artery thermodilution. Comment An artificial lung will likely make its first clinical impact as a short-term (weeks to months) bridge to transplant. The ideal patient will have already spent several months on the waiting list (currently still a single-tiered system) but be deteriorating and not expected to live to transplant. The candidate should need a single lung trans- Table 1. Ventilator Variables at 48th Hour After Onset of ARDS Variables PAL Control TV 210 ml 425 ml RR 6/min 29/min MV 1.2 L/min 10.8 L/min PIP 24 cm H 2 O 40 cm H 2 O FiO 2 21% 100% PaO mm Hg 63 mm Hg PaCO 2 34 mm Hg 61 mm Hg All ventilator settings in the PAL group are significantly lower (p 0.05) than the VCMV (Control) group. ARDS acute respiratory distress syndrome; FiO 2 fractional concentration of oxygen; MV minute ventilation; PaCO 2 arterial carbon dioxide tension; PAL paracorporeal artificial lung; PaO 2 arterial oxygen tension; PIP peak inspiratory pressure; RR respiratory rate; TV tidal volume. Fig 6. Hemodynamic variables in the paracorporeal artificial lung (PAL) group throughout the 5-day study. (CO cardiac output; CVP central venous pressure; MAP mean arterial pressure; PAP mean pulmonary artery pressure.)

5 Ann Thorac Surg ZWISCHENBERGER ET AL 2002;74: ARTIFICIAL LUNG IN ARDS 1015 plant, which would allow both flexibility in organ acceptance and a shorter wait time. As the artificial lung evolves and becomes safe, it will enter a second phase, that of longer term bridge-to-transplant, and the focus will change to patient rehabilitation and ambulation. Once long-term bridge-to-transplant becomes a reality, the artificial lung will enter a third phase of evolution: alternative to transplant. The results of this study suggest a fourth application of the PAL: bridge to recovery from a lethal lung injury. A large animal survival model is important to show that an artificial lung is life sustaining, and not just compatible with animal survival. Simply removing an animal s native lungs after PAL implant is not a clinically relevant model. The native lungs need to remain in the animal, and receive at least some pulmonary blood flow for two major reasons. First, the native lungs serve as a compliance chamber, allowing gradual filling of the left atrium. Second, the pulmonary circulation performs many vital metabolic functions, including deactivation of vasoactive compounds (bradykinin, serotonin, leukotrienes, and norepinephrine), and production of angiotensin II and prostacyclin [14]. The metabolism of these compounds decreases linearly with blood flow, and below 10% pulmonary flow an animal develops cardiovascular collapse [15]. In addition, our prior studies have shown that the metabolic rate and CO 2 production are elevated early in ARDS [8]. In the current study, PAL gas exchange reached an equilibrium a few hours after onset of ARDS after which CO 2 transfer and oxygenation varied less than 10%. An artificial lung can be configured in-parallel (pulmonary artery [PA] left atrium [LA]) or in-series (PA PA) with the native circulation. The PA LA configuration can either be complete (PA snared distally) or competitive (no snare). The complete PA LA configuration not only denies the pulmonary bed of blood flow necessary for metabolic functions, but the excluded pulmonary bed is at risk for diffuse thrombosis. The competitive PA LA configuration has the advantage of unloading the right ventricle by offering a parallel resistance bed. This configuration is recommended for pulmonary hypertension. The University of Michigan group has pursued the PA LA configuration with their own MC3 prototype, showing that in a competitive PA LA configuration, on average 47% of the right heart output will go through the University of Michigan Artificial Lung in healthy sheep [11] and 71% in a model of pulmonary hypertension [12]. However, the competitive PA LA configuration inherently provides only partial gas exchange, and device change-out or malfunction risks left-sided embolism. More important, during times of intolerance to anticoagulation, the PAL cannot be removed from the circuit and replaced with a closed loop because of the resultant massive right-to-left shunting. For these reasons, our group has continued to pursue the PA PA configuration. Although the additive impedance of the PAL and the injured lung could evoke right heart strain, the right heart (and cardiac output) tolerated the PAL with a LD 100 lung injury model. In part, this could be due to the pulmonary vasodilating effect of the fully oxygensaturated blood from the PAL into the pulmonary artery bed during lung injury [16]. The MC3/UTMB PAL is an ultra low-resistance, lowimpedance membrane gas exchanger. Unlike traditional cardiopulmonary bypass oxygenators, blood enters a central port and distributes radially to an outer collection chamber to prevent shunt and improve efficiency [4]. The PAL prototype was originally made with sampling ports and slip-on connectors for acute use. Initial experience with our large animal survival model quickly showed prototype design weaknesses such as bleeding through broken ports or slipped connectors. The ports were deleted and the connectors redesigned to a more stable collet-nut design. Survival of some animals in our first series did achieve the target of 7 days [4]. However, the immediate right heart failure rate was 50%. The in-series configuration inherently places the highest strain on the right ventricle of any artificial lung configuration [17]. Three modifications were then incorporated into the PAL design: a silicone compliance chamber was added to the inflow port (to simulate the capacitance of the pulmonary arterial bed), an inflow dispersion cone was placed to more evenly distribute blood flow, and the outlet port geometry was modified. As a result, the PAL steady-flow gradient at 5 L/min blood flow dropped from 8 to 6 mm Hg. The modified PAL in normal sheep showed an increased cardiac output from 2.8 L/min to 4.2 L/min, with mean central venous pressure 6.8, and a decrease from 2.5 to 0.79 Wood units [5]. The PAL devices are made by a rapid prototyping technique using polyester resins inside a computergenerated silicone rubber mold allowing multiple design iterations and prototype fabrication. Because these are hand-built prototypes, the necessary biomaterials are less durable, and the tolerances less precise relative to commercial membrane oxygenators. We have not studied the blood/surface interactions in these early generation PAL prototypes because the final device will be constructed from more blood/surface compatible biomaterials (heparin-bonded or coated surfaces, silicone impermeable fibers, etc.). Similarly, we have not investigated the pathophysiology of lung protective ventilation compared to the PAL at this early phase of development. However, we have evaluated the pathophysiology of lung injury in the smoke/burn sheep model treated with AVCO 2 R, which uses a commercially available membrane gas exchanger (Affinity AVECOR, Minneapolis, MN). We found reduced interleukin-8 mrna and protein (possibly due to less lung stretch), fewer pulmonary neutrophils, and reduced pulmonary myeloperoxidase activity in the smoke/burn AVCO 2 R group versus smoke/burn controls [18]. The AVCO 2 R has demonstrated safety in a phase 1 clinical study [19] and is currently undergoing a phase II multicenter clinical trial as treatment for ARDS. Like our AVCO 2 R studies, the PAL reduced ventilator-induced lung stretch and presumably subsequent lung injury. In the near future when injection molded preproduction PAL devices with blood/surface compatible materials are

6 1016 ZWISCHENBERGER ET AL Ann Thorac Surg ARTIFICIAL LUNG IN ARDS 2002;74: available, we plan to study pathology at 48 hours with concurrent controls. Remarkably, in this study, the PAL demonstrates a survival benefit for severe (LD 100 ) lung injury. We believe that these data imply a new potential application for the PAL as a bridge to recovery in early ARDS too severe for ventilator management or AVCO 2 R and too long-standing ( 7 to 10 days) for extracorporeal membrane oxygenation. Heparin has been shown to ameliorate lung injury if present before a smoke/burn injury [20, 21], possibly through antithrombin III-induced endothelial release of prostacyclin [22]. To allow ARDS development in our model, heparin was reversed immediately after cannula implant and chest closure, and before the smoke/burn injury. In addition, heparin reversal provided a window without any heparinization to allow early postoperative hemostasis. The VCMV control and PAL sheep were heparinized on the second study day, after ARDS criteria was met (PaO 2 /FiO 2 200). Both groups received identical heparin exposure after induction of lung injury for the remainder of the study to eliminate heparin as an outcomes variable. Based upon our success with the development of the PAL and the improved survival in this series, we plan additional large animal studies. First we will assess PAL long-term survival (weeks) in normal sheep. Next, we will conduct prospective, randomized, controlled, outcome studies in a LD 100 model of ARDS comparing PAL to VCMV for 10 days. Finally, we will investigate the pathophysiology of PAL in a LD 100 model of ARDS. Our goal is to develop clinically relevant PAL management protocols for long-term application in humans as a bridge-to-recovery or lung transplant. In conclusion, PAL provides total gas exchange with hemodynamic stability in a PA PA configuration to significantly decrease ventilator support with improved PaO 2 /FiO 2 ratio and wet/dry ratio during lung injury. In a prospective, randomized, controlled, unblinded outcome study comparing PAL to VCMV, PAL decreased ventilator-induced lung injury in a smoke/burn LD 100 sheep model of ARDS to improve survival. On the basis of this study, we propose a potential application for the PAL as a bridge to recovery in ARDS too severe for ventilator management or AVCO 2 R and too longstanding for extracorporeal membrane oxygenation. Supported in part by NIH STTR grant R41 H , Shriners Hospitals for Children grant 8700, and MC3 Corporation. References 1. Meyers BF, Lynch JP, Battafarano RJ, et al. Lung transplantation is warranted for stable, ventilator-dependent patients. Ann Thorac Surg 2000;70: Jurmann MJ, Schaefers HJ, Demertzis S, Haverich A, Wahlers T, Borst HG. Emergency lung transplantation after extracorporeal membrane oxygenation. ASAIO J 1993;39: M 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: 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: 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: Lynch WR, Montoya JP, Brant DO, Schreiner RJ, Iannettoni MD, Bartlett RH. Hemodynamic effect of a low-resistance artificial lung in series with the native lungs of sheep. Ann Thorac Surg 2000;69: Alpard SK, Zwischenberger JB, Tao W, Deyo DJ, Traber DL, Bidani A. New clinically relevant sheep model of severe respiratory failure secondary to combined smoke inhalation/ cutaneous flame burn injury. Crit Care Med 2000;28: Tao W, Brunston RL Jr, Bidani A, et al. Significant reduction in minute ventilation and peak inspiratory pressures with arteriovenous CO 2 removal during severe respiratory failure. Crit Care Med 1997;25: 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: Zwischenberger JB, Alpard SK, Tao W, Deyo DJ, Bidani A. Percutaneous extracorporeal arteriovenous carbon dioxide removal improves survival in respiratory distress syndrome: a prospective randomized outcomes study in adult sheep. J Thorac Cardiovasc Surg 2001;121: Lynch WR, Haft JW, Montoya JP, et al. Partial respiratory support with an artificial lung perfused by the right ventricle: chronic studies in an active animal model. ASAIO Abstracts 2000;46: Haft JW, Montoya P, Alnajjar O, et al. An artificial lung reduces pulmonary impedance and improves right ventricular efficiency in pulmonary hypertension. J Thorac Cardiovasc Surg 2001;122: 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: Said SI. Pulmonary metabolism of prostaglandins and vasoactive peptides. Annu Rev Physiol 1982;44: Takewa Y, Tatsumi E, Taenaka Y, et al. Hemodynamic and humoral conditions in stepwise reduction of pulmonary blood flow during venoarterial bypass in awake goats. ASAIO J 1997;43:M Lazar EI, Weinstein S, Stark CJ. Lung injury does not increase vascular resistance if pulmonary blood is fully saturated. ACS Surgical Forum 1993: 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: Schmalstieg FC, Chow J, Savage C, Rudloff HE, Palkowetz KH, Zwischenberger JB. Interleukin-8, aquaporin-1, and inducible nitric oxide synthase in smoke and burn injured sheep treated with percutaneous carbon dioxide removal. ASAIO J 2001;47: 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. Intensive Care Med 2001;27: Cox CS Jr, Zwischenberger JB, Traber DL, Traber LD, Haque AK, Herndon DN. Heparin improves oxygenation and minimizes barotrauma after severe smoke inhalation in an ovine model. Surg Gynecol Obstet 1993;176: Temmesfeld-Wollbruck B, Walmrath D, Grimminger F, Seeger W. Prevention and therapy of the adult respiratory distress syndrome. Lung 1995;173: Uchiba M, Okajima K. Antithrombin III (AT III) prevents LPS-induced pulmonary vascular injury: novel biological activity of AT III. Semin Thromb Hemost 1997;23:

7 Ann Thorac Surg ZWISCHENBERGER ET AL 2002;74: ARTIFICIAL LUNG IN ARDS 1017 DISCUSSION DR RIYAD C. KARMY-JONES (Seattle, WA): Trauma results in anywhere from 70,000 to 110,000 deaths a year, or two Vietnams a year at least as far as the United States is concerned. For each patient who dies, there are 5 to 7 patients who are critically injured, and it has become apparent, considering the trinodal pattern of death, that the only two areas of intervention that have significant promise are preventive strategies, or dealing with 15% to 20% of patients who die subsequently from multiple organ failure manifested predominantly by acute respiratory distress syndrome, with or without sepsis. It has been a pleasure to review this manuscript, which is part of an ongoing process that Dr Zwischenberger and his colleagues have been pursuing for more than 15 years, looking both for nontrauma and trauma applications of lung support. The animal model that he described took 3 intensive years to develop. It is a clinically relevant model, particularly in the fact that it combines a cutaneous and thermal injury, and also the control group was managed by contemporary low volume protective ventilator strategies. The initial Achilles heel had been acute right heart failure, which they appear to have controlled, predominantly with the use of a compliance chamber and also taking advantage of the protective effect of hyperoxygenation with subsequent pulmonary vasodilation. And, Jay, you still prefer the pulmonary artery to pulmonary artery configuration over pulmonary artery to left atrium configuration for the reasons that you mentioned, including more reliable gas exchange, but the system is still susceptible to pulmonary hypertension and other vagaries of pulmonary circulation that might occur with sepsis and more prolonged acute respiratory distress syndrome. And you note that you will be doing further long-term studies and that this is limited to a 5-day survival study, but within those constraints I have the following questions. First, given that burn wound sepsis starts to become a problem between 3 to 5 days, were any of these animals showing signs of wound infection? Second, can you anticipate how sepsis will affect the function of the paracorporeal artificial lung? Third, given the need for heparinization, how do you anticipate handling burn wound excision in such patients? Fourth, you favor the sheep model because the long pulmonary artery allows you to do two anastomoses of the paracorporeal artificial lung to the main pulmonary artery. How do you anticipate modifying this for the human condition with a much shorter pulmonary artery? And finally, you have quoted in the manuscript and in another areas that the ideal gas exchange would be as high as 800 ml. You quote 200 ml here. I take it that is just a function of the size of the animal. Thank you very much for the privilege of discussing this paper. DR JURO J. WADA (Tokyo, Japan): Lethal smoke insufflation oftentimes seen in big fire or jet pilot shutdown just about time of landing on the ground. (My work in Hyperbaric Medicine coincided with the Vietnam War.) In such a situation, resuscitation is on the site. The best is portable, collapsible hyperbaric chamber available in emergency ambulance. Oxygen insufflation is simple to apply on site. Paracorporeal artificial lung or another resulting method to follow would be in order. (Hyperbaric Medicine, Proceedings of IV International Congress of Hyperbaric Medicine took place in Sapporo, Japan, edited by J. Wada.) Such a way of treating seawater (salty water) drowning during swimming would be another clinical usefulness. (Judson Mc- Namara, University of Hawaii Cardiac Surgeon). Listening to this paper, it is certainly a needed way to preserve donor lung; however, I get the impression from this paper that it is too drastic, a lot of danger of contamination. In early 1960s I was heavily involved in use of hyperbaric oxygenation, which has been forgotten by cardiothoracic surgeons. Nobody mentioned about its use. That is, those burn together with smoke and so forth. It happens to jet plane pilots when they come down to the ground surrounded by fire. There he dies. Those are the best candidates just at the time of a fresh burned trachea can have it. The hyperbaric chamber is no longer a giant iron chamber weighing three tons, this is outdated. I had developed a collapsible carry-in hyperbaric chamber. A fire-exposed, dying person is put in the hyperbaric chamber without touching anything and oxygen is given into the bloodstream through the skin. Quickly, the simple collapsible hyperbaric chamber is put in the ambulance with the patient inside, brought to the hospital, then opened, and whatever the surgeons have to do to preserve the heart or make the end of life more peaceful. It would be ideal. I am too young to practice this, but in hearing this paper, I would like to stress to everyone something good we had under the name of hyperbaric oxygenation with the use of more than the plastic collapsible chamber available in hand in the surgeon. Thank you. DR ZWISCHENBERGER: I greatly appreciate the comments and the discussions. First of all, those of you who have worked with large animal models of acute respiratory distress syndrome fully appreciate the fact that we have no perfect large animal model of acute respiratory distress syndrome, and the fact that 90% of human acute respiratory distress syndrome is a secondary condition, not a primary condition, underlines the complexity of this problem. However, we tried to develop a model that was clinically pertinent, which is why we developed the burn and smoke inhalation combination model in our sheep, and it is with that model that we perform our outcomes studies. I have to admit that we are pushing the envelope with 5- and 10- and 15-day studies to try to investigate the respiratory injury when in fact it is multifaceted and also involves a synergistic effect from the burn. So we may run into problems with the burn. We had one of the sheep in this series develop sepsis, and that was the cause of death, which was outside the pulmonary injury. Therefore I share the discussant s concern about the burn wound being a problem, and we are currently addressing the fact that we are going to do early excision of the burn wound. Now, how are we going to do that in the face of heparin? Well, these prototypes that I describe are all handmade using a rapid prototyping fabrication technique, and the reason is because we have yet to come up with the perfect design to where we can invest in the molds necessary for final type of design and final type of material use. So I freely admit that these rapid prototype oxygenators are still embryonic in their development, and we are not yet able to fully appreciate the blood surface interaction and how we can best control that. Be that as it may, we are going

8 1018 ZWISCHENBERGER ET AL Ann Thorac Surg ARTIFICIAL LUNG IN ARDS 2002;74: to try to use heparin-coated devices and we are going to try to use lower doses of heparin, which will allow us to control the burn wound. Next is the fact that the sheep is a wonderful model for doing a pulmonary artery to pulmonary artery configuration, because we published that the sheep has a 5.4-cm pulmonary artery, and those of you who do cardiac surgery know that that is a very generous pulmonary artery, and humans do not have one that long. We have currently received Institutional Review Board approval to use fresh cadavers in humans to do pulmonary artery to pulmonary artery dissections and try to show how we can best orient that configuration in a human, and the congenital heart surgeons can easily tell you that the bifurcation of the pulmonary artery will lend itself to being able to use one of the branches for the distal anastomosis. So I think that that is going to work. Finally, I share Dr Wada s enthusiasm for alternate therapies in respiratory failure. I have spent many years trying to figure out whether it is oxygenation, hypercapnia, lung destruction, or all the inflammatory mediators that seem so popular these days as the cause of the etiology of respiratory failure. Quite frankly, I do not know the answer, Dr Wada, but I certainly share your enthusiasm, and I hope as we continue this research we can apply this technology to human use. Thank you. Requirements for Recertification/Maintenance of Certification in 2003 Diplomates of the American Board of Thoracic Surgery who plan to participate in the Recertification/ Maintenance of Certification process in 2003 must hold an active medical license and must hold clinical privileges in thoracic surgery. In addition, a valid certificate is an absolute requirement for entrance into the recertification/maintenance of certification process. If your certificate has expired, the only pathway for renewal of a certificate is to take and pass the Part I (written) and the Part II (oral) certifying examinations. The American Board of Thoracic Surgery will no longer publish the names of individuals who have not recertified in the American Board of Medical Specialties directories. The Diplomate s name will be published upon successful completion of the recertification/maintenance of certification process. The CME requirements are 70 Category I credits in either cardiothoracic surgery or general surgery earned during the 2 years prior to application. SESATS and SESAPS are the only self-instructional materials allowed for credit. Category II credits are not allowed. The Physicians Recognition Award for recertifying in general surgery is not allowed in fulfillment of the CME requirements. Interested individuals should refer to the 2003 Booklet of Information for a complete description of acceptable CME credits. Diplomates should maintain a documented list of their major cases performed during the year prior to application for recertification. This practice review should consist of 1 year s consecutive major operative experiences. If more than 100 cases occur in 1 year, only 100 should be listed. Candidates for recertification/maintanance of certification will be required to complete all sections of the SESATS self-assessment examination. It is not necessary for candidates to purchase SESATS individually because it will be sent to candidates after their application has been approved. Diplomates may recertify the year their certificate expires, or if they wish to do so, they may recertify up to two years before it expires. However, the new certificate will be dated 10 years from the date of expiration of their original certificate or most recent recertification certificate. In other words, recertifying early does not alter the 10-year validation. Recertification/maintenance of certification is also open to Diplomates with an unlimited certificate and will in no way affect the validity of their original certificate. The deadline for submission of applications for the recertification/maintenance of certification process is May 10 each year. A brochure outlining the rules and requirements for recertification/maintenance of certification in thoracic surgery is available upon request from the American Board of Thoracic Surgery, One Rotary Center, Suite 803, Evanston, IL 60201; telephone number: (847) ; fax: (847) ; ; abts_ evanston@msn.com. This booklet is also published on the website: by The Society of Thoracic Surgeons Ann Thorac Surg 2002;74: /02/$22.00 Published by Elsevier Science Inc