Switch From Venoarterial Extracorporeal Membrane Oxygenation to Arteriovenous Pumpless Extracorporeal Lung Assist

Switch From Venoarterial Extracorporeal Membrane Oxygenation to Arteriovenous Pumpless Extracorporeal Lung Assist Bernhard Floerchinger, MD, Alois Philipp, ECCP, Maik Foltan, ECCP, Leopold Rupprecht, MD, Alexander Klose, MD, Daniele Camboni, MD, Frank Bruenger, MD, Simon Schopka, MD, Mathias Arlt, MD, Michael Hilker, MD, and Christof Schmid, MD Departments of Cardiothoracic Surgery and Anesthesiology, University Hospital of Regensburg, Regensburg, Germany Background. Extracorporeal membrane oxygenation is an effective rescue tool to treat cardiopulmonary failure. Pumpless systems treat lung failure only; they require adequate cardiac output. Methods. We report on 18 patients initially provided with venoarterial extracorporeal membrane oxygenation and then downgraded to a pumpless arteriovenous shunt with a membrane oxygenator by removal of the pump from the circuit after hemodynamic stabilization in the face of persisting pulmonary failure. Main underlying diseases were adult respiratory distress syndrome (44%) and pneumonia (28%). Mean patient age was 44 years, and mean body mass index was 25.7 kg/m 2. Anticoagulation, hemodynamic, and respiratory variables were analyzed. Results. All patients exhibited severe cardiopulmonary failure with a mean oxygenation ratio (partial pressure of oxygen to fraction of inspired oxygen ratio) of 74 43 mm Hg (mean partial pressure of oxygen, 70 33 mm Hg) and a mean partial pressure of carbon dioxide of 68 32 mm Hg despite maximal (ventilatory) conservative therapy (fraction of inspired oxygen, 0.98 0.08). Initial serum lactate was 51 43 mg/dl. The sequential organ failure assessment score averaged 11.8 2.47, and the lung injury score was 3.1 0.58. Total mechanical respiratory support was performed for a mean of 13.6 15.7 days. After 24 hours an improvement in oxygenation and a decrease in carbon dioxide was achieved with a mean partial pressure of carbon dioxide of 40 11 mm Hg (p < 0.001) and a partial pressure of oxygen of 86 26 mm Hg (p 0.031). After 6 3 days of extracorporeal membrane oxygenation, patients were hemodynamically stabilized. Extracorporeal membrane oxygenation was downgraded to pumpless extracorporeal lung assist for another 10 15 days (range, 2 to 71 days). Twelve patients (66.7%) could be weaned, with a 30-day mortality of 55.6%. Norepinephrine dosage could be reduced significantly within 24 hours (3.2 1.8 versus 1.5 1.5 mg/h; p 0.008). Conclusions. Respiratory support by an extracorporeal device used as last resort therapy allows rapid stabilization of patients with acute lung failure. Early replacement of extracorporeal membrane oxygenation by pumpless extracorporeal lung assist minimizes the negative side effects of extracorporeal circulation. (Ann Thorac Surg 2010;89:125 31) 2010 by The Society of Thoracic Surgeons Acute lung failure is associated with mortality rates ranging between 40% and 50% regardless of treatment [1, 2]. The main underlying pathophysiology is adult respiratory distress syndrome (ARDS), which is characterized by respiratory insufficiency with severe hypoxemia or hypercapnia. The treatment strategy is based primarily on mechanical ventilation, including low tidal volume and positive end-expiratory pressure ventilation, spontaneous breathing, and prone positioning [3]. Inhalation of nitric oxide and use of glucocorticoids are Accepted for publication Sept 1, 2009. Address correspondence to Dr Schmid, Department of Cardiothoracic Surgery, University Hospital of Regensburg, Franz-Josef-Strauss-Allee 11, Regensburg, 93053, Germany; e-mail: christof.schmid@klinik.uniregensburg.de. further therapeutic options, but a clear benefit has not been demonstrated [4, 5]. Since its first clinical use by Hill and O Brien [6] and Bartlett and colleagues [7] in the 1970s, extracorporeal membrane oxygenation (ECMO) support has became a lifesaving procedure. Initially, the best outcomes were achieved in patients with postoperative cardiac failure of newborn and adults with capillary leak syndrome [6, 7]. Extracorporeal membrane oxygenation is now widely used in cases of cardiac and respiratory failure [8]. An important indication for ECMO has become severe ARDS regardless of underlying disease, for which it is considered a last resort therapy [9, 10]. Although ECMO ensures adequate oxygenation and carbon dioxide removal through pump-driven extracorporeal respiratory support, ventilation variables can be reduced to mini- 2010 by The Society of Thoracic Surgeons 0003-4975/10/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2009.09.007

126 FLOERCHINGER ET AL Ann Thorac Surg EXTRACORPOREAL SUPPORT IN LUNG FAILURE 2010;89:125 31 Table 1. Patient Characteristics and Outcome, Sequential Organ Failure Assessment Score, and Lung Injury Score Before Extracorporeal Support Patient No. Diagnosis Age (y) BMI SOFA Score LIS Outcome 1 Wegener granulomatosis 21 23 15 3 Weaned 2 Waterhouse-Friderichsen syndrome 17 24 14 3.33 Weaned 3 ARDS after trauma 39 25 14 2.67 Weaned 4 Pneumonia after bilobectomy 69 23 10 3.75 Weaned 5 ARDS after bilobectomy 60 22 10 3.25 Died on system 6 ARDS after general surgery 70 26 14 3.67 Died on system 7 ARDS after aortic surgery 49 28 13 2 Died on system 8 Lung fibrosis, pneumonia 37 35 8 3.5 Died on system 9 Pneumonia 20 20 13 4 Weaned 10 ARDS after chemotherapy 41 25 11 2.5 Weaned 11 Cystic fibrosis 25 19 8 3.33 Weaned 12 Tracheal injury 46 26 13 4 Weaned 13 Pneumonia 46 34 15 3.67 Weaned 14 ARDS after SAH, neurosurgery 50 24 11 3 Weaned 15 ARDS after thoracic trauma, pneumectomy 62 26 11 2.67 Weaned 16 Pulmonary hypertension 38 23 12 2.67 Weaned 17 Pneumonia 60 30 7 2.67 Died on system 18 ARDS after trauma 33 30 14 2.41 Died on system ARDS acute respiratory distress syndrome; BMI body mass index; LIS lung injury score; SAH subarachnoid hemorrhage; SOFA sequential organ failure assessment. mize ventilator-associated lung injury while supporting pulmonary recovery. Pumpless extracorporeal lung assist therapy (PECLA, also called interventional lung assist, or ila; NovaLung GmbH, Hechingen, Germany) uses a low-resistance membrane oxygenator interposed in an arteriovenous shunt. The system uses the blood pressure difference between arterial and venous circulations as the driving force to lead blood through an oxygenator [11, 12]. Accordingly, adequate cardiac output has to be ensured before institution of the pumpless device, and low cardiac output and hypotension are contraindications. In 1998 PECLA was first used in Regensburg for 10 days in a patient with acute pancreatitis and lung failure [11]. As oxygenation through the PECLA device is limited by blood flow and blood pressure, hypoxia still remains an indication for ECMO [12]. The advantages of the PECLA device include less artificial surface contact, with fewer side effects of pump-driven ECMO such as coagulation disorders, which are reported in 16% to 32% of those on the therapy. Also, use of a pumpless device may avoid the technical complications that occur in 2% to 6% of ECMO patients [13, 14]. The aim of the current study was to evaluate whether downgrading the extracorporeal circuit to a pumpless device and restoring physiologic pulsatile perfusion improve outcomes and reduce complications among patients with ARDS who need extracorporeal support. We report our experience with 18 patients experiencing severe circulatory failure and additional ARDS who were initially provided with pump-driven ECMO and were then switched to PECLA after hemodynamic recovery. Material and Methods From 1998 to 2007, 18 patients (14 male, 4 female) exhibiting acute respiratory failure were treated with venoarterial ECMO followed by PECLA at the University Medical Center of Regensburg, Germany. Mean age was 44 years (range, 17 to 70 years), and body weight varied from 52 to 110 kg with a mean body mass index of 25.7 4.3 kg/m 2. Treatment data were collected prospectively (Table 1). The study was approved by the institutional ethics committee. As the patients had a life-threatening condition, written consent for extracorporeal support was given by next of kin entrusted with care. The consent to downgrade the device was not required by the institu- Fig 1. Pump-driven venoarterial extracorporeal membrane oxygenation, femoral access.

Ann Thorac Surg FLOERCHINGER ET AL 2010;89:125 31 EXTRACORPOREAL SUPPORT IN LUNG FAILURE 127 using the Seldinger technique. After cannula placement, the centrifugal blood pump (Rotaflow; Jostra Inc, Hirrlingen, Germany) and a membrane oxygenator (Quadrox BE, or a Quadrox PLS Bioline Coated; Maquet Inc, Hirrlingen, Germany) were connected by means of heparin-coated tubes (Fig 1). The Quadrox membrane oxygenator is made of polymethylpentene and features a flow resistance of approximately 15 mm Hg at a blood flow of 2.5 L/min. It has an inner surface of 1.3 m 2. Oxygen can be applied up to 12 L/min. The priming volume of the entire system amounted to 600 ml of gelatin-based plasma expander. Fig 2. Pumpless arteriovenous extracorporeal circuit, femoral access. tional ethics committee, as there is an extensive experience of pulmonary support in our hospital. Extracorporeal Membrane Oxygenation Therapy All patients presented with deteriorated pulmonary function after ARDS. Pre-ECMO treatment included positive end-expiratory pressure ventilation and prone position ventilation. Nitric oxide ventilation was performed in 8 patients, whereas 1 patient had surfactant augmentation. Indication for venoarterial ECMO was based on hypoxia with an arterial partial pressure of oxygen (Po 2 ) to fraction of inspired oxygen (Fio 2 ) ratio less than 80 mm Hg and hypercapnia with an arterial partial pressure of carbon dioxide (Pco 2 ) greater than 70 mm Hg despite maximal mechanical ventilatory support. Criteria for primary use of a pump-driven system were a cardiac index less than 2.5 L min 1 m 2 despite high-dosage catecholamine support (epinephrine 4 g/kg/min, norepinephrine 5 g/kg/min), with a mean arterial blood pressure less than 70 mm Hg. Extracorporeal membrane oxygenation therapy usually allowed the reduction of mechanical ventilation, with a positive end-expiratory pressure less than 10 cm H 2 O and an Fio 2 less than or equal to 0.6 at most. Before initiation of ECMO, a patient s hemodynamic status was assessed either by Swan-Ganz catheterization or echocardiography. Whenever possible, the diameters of the femoral vessels were measured by ultrasound, and the femoral arteries additionally scanned for arteriosclerotic changes. The cannula sizes chosen were 20% smaller than the diameter of the respective vessel. The arterial cannula (19F to 21F) was placed into the femoral or subclavian artery using the Seldinger technique, or after suturing a Dacron prosthesis (DuPont, Wilmington, DE) end-to-side to the respective vessel. The venous cannula (21F) was inserted into the femoral vein, also Pumpless Extracorporeal Lung Assist Therapy The pump-driven ECMO was downgraded to a PECLA system by removal of the centrifugal blood pump, leaving an arteriovenous shunt with a membrane oxygenator in between. This was performed after stabilization of cardiac function and improvement of oxygenation in the face of persistent high carbon dioxide levels (Fig 2). As a consequence, there was a flow reversal within the circuit from venoarterial to arteriovenous. Prerequisites were a stabilized hemodynamic status with a cardiac index greater than 2.5 L min 1 m 2 estimated by echocardiography, with ECMO flow reduced to 2 L/min. Further requirements for system downgrade were a mean arterial blood pressure of greater than 70 mm Hg with only a low catecholamine dosage, as well as an arterial oxygen pressure greater than 70 mm Hg for at least 24 hours. During PECLA therapy, mechanical ventilation was continued at a noninvasive and lung-protective level with a positive end-expiratory pressure less than 10 cm H 2 O and an Fio 2 less than 0.6. Pumpless extracorporeal lung assist support was terminated when patients had stable respiratory status for greater than 24 hours with an arterial Po 2 greater than 80 mm Hg, a Pco 2 less than 40 mm Hg, and an Fio 2 less than 0.45. Anticoagulation In all patients, cannulas and blood lines were coated with high-molecular-weight heparin (Bioline Coating; Maquet Inc) to avoid thrombus deposition in the system. A continuous infusion of high-molecular-weight heparin was added with the intention of maintaining a partial thromboplastin time of 60 to 70 seconds in pump-driven devices. After the centrifugal pump was removed, the target range of thromboplastin time was reduced to 50 to 60 seconds. Activated clotting time target range was 130 to 150 seconds, respectively. Measured Variables Fraction of inspired oxygen, arterial oxygen and carbon dioxide pressures, oxygen saturation, and serum lactate and ph were measured before establishing an extracorporeal circuit, 2 hours and 24 hours after initiation, immediately before termination, and 24 hours after termination. Flow rates in the device and respirator settings (tidal volume, peak inspiratory pressure, respirator breath rate) were noted as well. Blood samples were

128 FLOERCHINGER ET AL Ann Thorac Surg EXTRACORPOREAL SUPPORT IN LUNG FAILURE 2010;89:125 31 Table 2. Course of Arterial Blood Gases, ph, Inspired Oxygen, Norepinephrine, Mean Arterial Blood Pressure, and Serum Lactate Variable Preimplant 2 h 24 h Preexplant 1 d Postexplant p Value a p Value b Po 2 (mm Hg) 70 33 86 46 86 26 87 23 89 20 0.031 0.857 Pco 2 (mm Hg) 68 32 44 18 40 11 36 11 41 15 0.001 0.686 Fio 2 0.98 0.08 0.77 0.22 0.64 0.21 0.48 0.1 0.52 0.15 0.001 0.75 Po 2 /Fio 2 (mm Hg) 74 43 145 166 156 82 190 62 185 65 0.001 0.839 Lac (mg/dl) 51 43 77 59 65 80 24 19 25 24 0.812 0.696 ph 7.24 0.14 7.33 0.13 7.43 0.05 7.44 0.06 7.39 0.1 0.001 0.371 Nor (mg/h) 3.2 1.8 2.0 1.9 1.5 1.5 0.3 0.3 0.9 1.6 0.008 0.931 MAP (mm Hg) 62 20 73 25 74 23 74 28 75 29 0.012 0.91 a Statistical difference between before and after 24-hour support. b Statistical difference between support stop and 24 hours after stop. Fio 2 fraction of inspired oxygen; Lac serum lactate; MAP mean arterial blood pressure; Nor norepinephrine; Pco 2 partial pressure of carbon dioxide; Po 2 partial pressure of oxygen. analyzed daily to monitor anticoagulation, hemoglobin, creatinine kinase, and hemogram levels. Statistical Analysis Statistical analyses were performed with Sigma-Stat 3.1 (Scientific Solutions Inc, Lausanne, Switzerland) and SPSS (SPSS Inc, Chicago, IL). Differences between survivors and nonsurvivors were calculated by Mann- Whitney-Wilcoxon test. Results Pre Extracorporeal Membrane Oxygenation Status Before the initiation of ECMO, all patients showed respiratory failure with severe hypoxia and hypercapnia despite maximum ventilation therapy with a mean Fio 2 of 0.98 0.08 (Po 2 /Fio 2 ratio, 74 43 mm Hg). Arterial Po 2 before device placement was 70 33 mm Hg and Pco 2 was 68 32 mm Hg. Mean ph value was 7.24 0.14, and serum lactate levels were 51 43 mg/dl. Mean ventilation time was 2.6 4.9 days. Main underlying diseases were ARDS in 44% (n 8) of patients, caused by either previous surgery or trauma, and pneumonia in 28% of patients (n 5). The average sequential organ failure assessment score was 11.8 2.47, and the mean lung injury score was 3.1 0.58. Four patients had to be resuscitated before or during venoarterial ECMO placement. In 3 patients, ECMO devices had been placed in another hospital, after which patients were transferred to our institution. Extracorporeal Membrane Oxygenation Therapy The femoral artery was used for arterial cannulation in 14 patients, among 6 of whom an open surgical access was necessary. The right subclavian artery was deemed favorable in 2 patients, and a Dacron prosthesis (6 mm) was sutured end-to-side to the vessel. Two patients had an intrathoracic cannulation of the ascending aorta and right atrium. In the other 16 patients, the venous cannula was inserted into a femoral vessel. Extracorporeal membrane oxygenation caused a decline of arterial Pco 2 to 44 18 mm Hg (p 0.002) within 2 hours, whereas Po 2 was increased to 86 46 mm Hg (p 0.214). After 24 hours, a significant improvement of gas exchange was achieved in all but 1 patient; Po 2 was kept constant (86 26 mm Hg; p 0.031), and Pco 2 was decreased to 40 11 mm Hg (p 0.001). Blood ph levels normalized to 7.43 0.05 (p 0.001), and serum lactate remained at the same level (65 80 mg/dl; p 0.812). Respirator therapy could be downgraded to a lungprotective mode (Fio 2 approximately 0.64 0.21; p 0.001) after the first day. The Po 2 /Fio 2 ratio accordingly increased significantly (156 82 mm Hg; p 0.001). Norepinephrine dosage could be also reduced significantly (3.2 1.8 versus 1.5 1.5 mg/h; p 0.008), and mean arterial blood pressure rose significantly (62 20 versus 74 23 mm Hg; p 0.012) within 24 hours (Table 2). Pumpless Extracorporeal Lung Assist Therapy After 6 3 days of venoarterial ECMO cardiopulmonary stabilization, the extracorporeal circuit could be downgraded to a PECLA system by removing the centrifugal pump. At this point blood gas analysis revealed a mean Po 2 of 85 22 mm Hg and a mean Pco 2 of 38 7mmHg. The Fio 2 averaged 0.48 0.09. Pumpless extracorporeal lung assist was then maintained for 10 15 days (range, 2 to 71 days), after which conventional ventilator therapy was necessary for another 2 to 30 days (10 12 days). When extracorporeal support was stopped, the Po 2 / Fio 2 ratio averaged 190 62 mm Hg (Po 2 approximately 87 23 mm Hg, Pco 2 approximately 36 11 mm Hg, Fio 2 approximately 0.48 0.1). Twenty-four hours later, Po 2 was 89 20 mm Hg (p 0.857) and Pco 2 was 41 15 mm Hg (p 0.686). At that time respirator therapy was performed with a mean Fio 2 of 0.52 0.15 (p 0.75), and the Po 2 /Fio 2 ratio remained at 185 65 mm Hg (p 0.839). Survival Twelve patients were successfully weaned from extracorporeal support (66.7%), but 5 patients died after weaning: 2 patients underwent multiorgan failure (after 1 and 10 days, respectively), and in 1 patient ECMO had to be

Ann Thorac Surg FLOERCHINGER ET AL 2010;89:125 31 EXTRACORPOREAL SUPPORT IN LUNG FAILURE reestablished owing to recurrent respiratory failure on day 39 after primary successful weaning from the extracorporeal support. One patient died of circulatory collapse after 17 days, and 1 patient died as a result of general cachexia 33 days after weaning. Six patients died during extracorporeal support: 3 patients died while on PECLA support owing to both septic shock and multiorgan-failure (2 patients) or because of acute circulatory failure (1 patient). In 3 patients, the ECMO system had to be reinstituted after initial downgrade owing to circulatory depression. All these patients in whom ECMO had to be reinstituted died during pump-driven extracorporeal support. The entire project showed a 30-day mortality of 55.6% (10 patients), and a survival to hospital discharge of 33.3% (6 patients). One patient who initially had been provided with an ECMO connected to the ascending aorta and the right atrium was downgraded after 9 days to a PECLA connected to the pulmonary artery and the left atrium to work as an artificial lung. After extracorporeal support for 62 days, this patient underwent double-lung transplantation [15]. A mean of 1.4 membrane oxygenators were used per patient, the most common reason for oxygenator exchange being thrombus formation (6 patients). In 3 patients thrombotic deposits were detected while receiving pump-driven support: 2 patients (1 patient with heparininduced thrombocytopenia) showed thrombus formation both in the oxygenator and centrifugal pump. In 1 patient a system stop was caused by dislocation of the arterial cannula. During pumpless lung assistance, a clotted oxygenator had to be exchanged in 3 patients. Kinking of the tubes caused a system stop and required an exchange in 1 patient. The patient provided with a PECLA connected to the pulmonary artery and left atrium required four oxygenator exchanges during 62 days of pumpless support. In 1 patient thrombotic deposition and air leakage occurred simultaneously. Oxygenator air leakage was noted in 2 patients. Because ischemia of the lower limb occurred in 1 patient after insertion of the arterial cannula in the femoral artery, arterial access was switched to the opposite side. As the leg ischemia spontaneously decreased thereafter, no surgical reconstruction was required. No ischemia was noted in patients with a subclavian access. Removal of the arterial cannula had to be performed surgically in 8 patients (2 subclavian, 6 femoral artery). Comment Pump-driven ECMO devices, invented in the early 1970s, are now used routinely in most cardiothoracic surgical departments. Various ECMO systems are now available for patients with respiratory and circulatory failure. Pump-driven ECMO provides effective oxygenation and carbon dioxide removal and allows hemodynamic stabilization when initiated in the venoarterial setting. Survival rates are best in newborns (approximately 75%) but considerably lower in adult patients (30% to 50%) [14, 16]. 129 Pumpless extracorporeal lung assist devices allow effective carbon dioxide removal, but their ability to improve oxygenation is limited by blood flow through the system. A sufficient cardiac output and a mean arterial blood pressure more than 70 mm Hg are required for proper function of the system, but augmentation of cardiac output and blood pressure by high doses of catecholamines is not recommended. In this patient cohort, each of whom was provided with ECMO initially and downgraded to a pumpless device afterward, the overall survival appears to be disappointing. Therefore the obvious question might be when or why to use a PECLA instead of an ECMO system. Venoarterial ECMO may seem preferable, but the systems should be tailored to the individual patient s condition. If patients with severe cardiopulmonary compromise have partially recovered and present with stable cardiac pump function and improved oxygenation, their high carbon dioxide levels must still be treated. Aggressive mechanical ventilation, such as high tidal volumes, does not appear to be sufficient for removing carbon dioxide in the presence of pulmonary inflammation, which may be caused by these factors, potentially resulting in systemic multiple organ failure. Similarly, the use of high end-expiratory pressures improves oxygenation but has no impact on carbon dioxide removal and has not been shown to reduce mortality rates from lung failure [17]. Likewise, better efficacy in oxygenation can be achieved by high-frequency oscillatory ventilation, but a beneficial effect regarding carbon dioxide removal cannot be shown [18]. Additionally, like ECMO support, invasive mechanical ventilation treatment requires sedation of the patient. In the case of hypercapnia treated solely with PECLA, sedatives can be discontinued. Even ventilation therapy might be reduced and mobilization might be performed as reported previously [15]. Patients can be weaned from the respirator, even with PECLA in place. However, when patients gas exchange strictly depends on the extracorporeal device, technical complications may cause acute hypoxia and hypercapnia. Bleeding complications and technical complications are common adverse side effects during extracorporeal support [13, 14]. These effects can be reduced by PECLA support. To avoid potential technical complications, removal of the blood pump appears to be beneficial. After downgrading the extracorporeal circuit, less anticoagulation is necessary. Subsequently there is a lower risk of bleeding complications, especially important among this patient population, which frequently presents an impaired coagulation system owing to multiorgan failure. Even withdrawal of heparin infusion for 2 to 3 days is made possible by an easily performed oxygenator exchange. The most appropriate site for insertion of the arterial cannula remains a matter of discussion. Most patients in this report underwent femoral insertion of the arterial cannula with only one instance of femoral ischemia afterward; no ischemia occurred in patients who were provided with a subclavian arterial cannula. We favor

130 FLOERCHINGER ET AL Ann Thorac Surg EXTRACORPOREAL SUPPORT IN LUNG FAILURE 2010;89:125 31 subclavian access, as sufficient perfusion of the supraaortic vessels with oxygenated blood is ensured. However, before insertion an anastomosis with a Dacron prosthesis is recommended to avoid ischemia of the upper limb. The Seldinger technique may be hazardous in the subclavian site because of potential vascular laceration. The benefit of a femoral insertion site is the simple and quick accessibility, making it useful in resuscitation settings. Insertion of a cannula by the Seldinger technique is relatively easy, and low complication rates comparable to those associated with coronary angiography and percutaneous coronary interventions are reported [19, 20]. An arterial return through a femoral artery offers optimal gas exchange to the lower body, but the upper half of the body (including the coronary arteries and the supraaortic vessels) is still provided with blood that has passed through the native lungs. Thus, a high Pco 2 and a lower ph within the coronary system and the brain vessels may prevent optimal recovery of the respective end-organ function. The alternatives are to relocate the arterial return from the femoral to the subclavian artery or to implement a venovenous system. Both ECMO variations work quite well, but the inherent complications of a pump-driven device remain. Regarding comparative costs of treatment, we leave the Quadrox BE oxygenator in the circuit (Fig 2) with higher flow resistance than the ila Membranventilator (Novalung Inc, Hechingen, Germany), which is initially used in pumpless devices. Because oxygenator resistance depends on blood flow through the system, an estimated mean flow of 2.0 L/min yields comparable resistance both in the ila and the Quadrox BE oxygenator. In summary, extracorporeal support both in oxygenation and carbon dioxide removal is provided to patients in whom conventional ventilation therapy is failing. For these patients with respiratory failure and an impaired hemodynamic status, a pump-driven device has to be considered to resolve cardiovascular failure and to assure oxygenation. After hemodynamic stabilization, downgrading of the extracorporeal circuit is reasonable to restore physiologic pulsatile perfusion and to reduce the adverse effects of extracorporeal devices. Commonly it is multiorgan dysfunction or sepsis, not hypoxia or hypocapnia, that are reported as leading causes of death in these patients [21]. Therefore therapy should be focused on resolving septic shock and multiorgan failure, as cardiovascular and acute renal failure are frequent complications in nonsurviving patients [22]. Nevertheless a rather accurate selection of patients represents a substantial point for application of an extracorporeal device. Our 10-year experience reflects an overall survival rate of 35%, with best outcomes in patients after trauma compared with other disorders leading to respiratory failure [23]. In 33% of our patients, ARDS occurred after extended surgical procedures. Three patients had undergone thoracic surgery, whereas in only 1 patient aortic surgery with cardiopulmonary bypass had been performed. However, because the PECLA device is based on an arteriovenous shunt, and sufficient cardiac output as well as a mean arterial blood pressure greater than 70 mm Hg are required, we do not approve its use immediately after cardiac surgery or before stable hemodynamic status is ensured. Despite constant efforts in improving extracorporeal devices, their use does not appear to imminently improve survival rates in ARDS [13]. At first glance, the overall mortality rate of 67% in this study may appear rather high, but the patients impaired status with a mean sequential organ failure assessment score of 12 and a mean lung injury score of 3.1 has to be considered. Ferreira and associates [24] report mortality rates of 80% and higher with an initial sequential organ failure assessment score of more than 11 in 352 consecutive patients. A high mortality rate (exceeding 50%) was reported by Eachempati and colleagues [25] in a comparable study of ARDS patients with an initial lung injury score of 3.3. Even though no imminent benefit in survival is detectable for patients with respiratory or circulatory failure, ECMO and PECLA devices as well as life-support devices remain highly efficient tools for rescuing patients from imminent death. Therefore extracorporeal support is to be legitimately administered. References 1. The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301 8. 2. Phua J, Badia JR, Adhikari NK, et al. Has mortality from acute respiratory distress syndrome decreased over time? Am J Respir Crit Care Med 2009;179:220 7. 3. Girard TD, Bernard GR. Mechanical ventilation in ARDS. Chest 2007;131:921 9. 4. Deja M, Hommel M, Weber-Carstens S, et al. Evidencebased therapy of severe acute respiratory distress syndrome: an algorithm-guided approach. J Int Med Res 2008;36:211 21. 5. Bream-Rouwenhorst HR, Beltz EA, Ross MB, Moores KG. Recent developments in the management of acute respiratory distress syndrome in adults. Am J Health Syst Pharm 2008;65:29 36. 6. Hill JD, O Brien TG. Prolonged extracorporeal oxygenation for acute posttraumatic failure; use of Branson lung. N Engl J Med 1972;286:629 34. 7. Bartlett RH, Gazzaniga AB, Fong SW, Jefferies MR, Roohk HV, Haiduc N. Extracorporeal membrane oxygenator support for cardiopulmonary failure. Experience in 28 cases. Thorac Cardiovas Surg 1977;73:375 86. 8. Schuerer DJE, Kolovos NS, Boyd KV, Coopersmith CM. Extracorporeal membrane oxygenation, current clinical practice, coding, and reimbursement. Chest 2008;134: 179 84. 9. Hemmila MR, Rowe SA, Boules TN, et al. Extracorporeal life support for severe acute respiratory distress syndrome in adults. Ann Surg 2004;240:595 607. 10. Wigfield CH, Lindsey JD, Steffens TG, Edwards NM, Love RB. Early institution of extracorporeal membrane oxygenation for primary graft dysfunction after lung transplantation improves outcome. J Heart Lung Transplant 2007;26: 331 8. 11. Philipp A, Behr R, Reng M, Kaiser M, Birnbaum D. Pumpless extracorporal lung assist. J Extra Corpor Technol 1998;30: 38 41. 12. Mueller T, Lubnow M, Philipp A, et al. Extracorporeal pumpless interventional lung assist in clinical practice: determinants of efficacy. Eur Respir J 2009;33:551 8.

Ann Thorac Surg FLOERCHINGER ET AL 2010;89:125 31 EXTRACORPOREAL SUPPORT IN LUNG FAILURE 131 13. Mols G, Loop T, Geiger K. Extracorporeal membrane oxygenation: a ten-year experience. Am J Surg 2000;180:144 54. 14. Conrad SA, Rycus PT, Dalton H. Extracorporeal Life Support Registry Report 2004. ASAIO J 2005;51:4 10. 15. Schmid C, Philipp A, Hilker M, et al. Bridge to lung transplantation through a pulmonary artery to left atrial oxygenator circuit. Ann Thorac Surg 2008;85:1202 5. 16. Haines NM, Rycus PT, Zwischenberger JB, Bartlett RH, Undar A. Extracorporeal Life Support Registry Report 2008: neonatal and pediatric cardiac cases. ASAIO J 2009;55:111 6. 17. Villar J, Kacmarek RM, Perez-Mendez L, Aguirre-Jaime A, for The National Heart Lung and Blood Institute ARDS Clinical Trials Network. Higher versus lower positive endexpiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med 2004;351:327 36. 18. Derdak S, Mehta S, Stewart TE, et al. High-frequency oscillatory ventilation for acute respiratory distress syndrome in adults: a randomized, controlled trial. Am J Respir Crit Care Med 2002;166:801 8. 19. Chue CD, Hudsmith LE, Stumper O. Femoral vascular access complications in adult congenital heart disease patients: audit from a single tertiary center. Congenit Heart Dis 2008;3:336 40. 20. Luciani N, Anselmi A, De Geest R. Extracorporeal circulation by peripheral cannulation before redo sternotomy: indications and results. J Thorac Cardiovasc Surg 2008;136: 572 7. 21. Estenssoro E, Dubin A, Laffaire E, et al. Incidence, clinical course, and outcome in 217 patients with acute respiratory distress syndrome. Crit Care Med 2002;30:2450 6. 22. Ferring M, Vincent JL. Is outcome from ARDS related to the severity of respiratory failure? Eur Respir J 1997;10:1297 300. 23. Floerchinger B, Philipp A, Klose A. Pumpless extracorporeal lung assist: a 10-year institutional experience. Ann Thorac Surg 2008;86:410 7. 24. Ferreira FL, Bota DP, Bross A, et al. Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA 2001;286:1754 8. 25. Eachempati SR, Hydo LJ, Shou J. Outcomes of acute respiratory distress syndrome (ARDS) in elderly patients. J Trauma 2007;63:344 50. INVITED COMMENTARY Floerchinger and colleagues [1] present a case series of 18 patients who were converted from conventional venoarterial (VA) extracorporeal membrane oxygenation (ECMO) to a pumpless arteriovenous (AV) configuration called a pumpless extracorporeal lung assist (PECLA). The authors contend that by removing the pump, the system becomes less complex with fewer complications while still primarily performing CO 2 removal, which allows a gentler ventilatory strategy to be used. The goal of PECLA is to allow early weaning off the ventilator by managing CO 2 removal. The authors report a survival to discharge of 34% (6 of 18). The population presented is a heterogeneous group of 18 patients with severe ARDS (Pao 2 /Fio 2 ratio 80) who were placed on standard VA ECMO with a centrifugal pump and membrane oxygenator. As it was believed that the patients would improve, the pump was removed and the blood flow was reversed through an AV loop. The concept of converting from VA ECMO to a pumpless AV system is unique. This conversion from a VA pumped system to an arterial-venous passive system allows for reductions of the level of anticoagulation, eliminates the risk of arterial thrombo-emboli, and removes the pump as a potential complication. Although the patients would still be heparinized, by far the riskiest part of ECMO, the authors lowered their target activated clotting time (ACT) to 130 to 150 seconds, which should help decrease bleeding complications. The authors also contend that this simplified configuration would not require a dedicated team of specialists. Our experience would dictate that as long as you have a large bore cannula in an artery, most institutions will want a dedicated specialist at the bedside. Although outcomes data is not supportive as to how much this weaning technique will decrease complications or improve outcomes, this approach provides for an interesting and easy alternative to either continued VA ECMO support or discontinuation of ECMO to an injurious ventilatory strategy in patients with CO 2 retention. James E. Lynch, BS, RRT University of Texas Medical Branch Galveston, TX 77555 e-mail: jelynch@utmb.edu Joseph B. Zwischenberger, MD Department of Surgery University of Kentucky College of Medicine 800 Rose St Lexington, KY 40536 Reference 1. Floerchinger B, Philipp A, Foltan M, et al. Switch from venoarterial extracorporeal membrane oxygenation to arteriovenous pumpless extracorporeal lung assist. Ann Thorac Surg 2010;89:125 31. Dr Zwischenberger discloses that he has a financial relationship with Novalung. 2010 by The Society of Thoracic Surgeons 0003-4975/10/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2009.09.067