Extracorporeal Membrane Oxygenation as a Bridge to Lung Transplant: Midterm Outcomes

Transcription

1 Extracorporeal Membrane Oxygenation as a Bridge to Lung Transplant: Midterm Outcomes Christian A. Bermudez, MD, Rodolfo V. Rocha, MD, Diana Zaldonis, MPH, BSN, Jay K. Bhama, MD, Maria M. Crespo, MD, Norihisa Shigemura, MD, PhD, Joseph M. Pilewski, MD, Penny L. Sappington, MD, Arthur J. Boujoukos, MD, and Yoshiya Toyoda, MD, PhD Departments of Cardiothoracic Surgery, Critical Care Medicine, and Pulmonary, Allergy and Critical Care Medicine, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania Background. Extracorporeal membrane oxygenation (ECMO) is used occasionally as a bridge to lung transplantation. The impact on mid-term survival is unknown. We analyzed outcomes after lung transplant over a 19-year period in patients who received ECMO support. Methods. From March 1991 to October 2010, 1,305 lung transplants were performed at our institution. Seventeen patients (1.3%) were supported with ECMO before lung transplant. Diagnoses included retransplantation (n 6), pulmonary fibrosis (n 6), cystic fibrosis (n 4), and chronic obstructive pulmonary disease (n 1). Fifteen patients underwent double lung transplant, one patient had single left lung transplant and one patient had a heart-lung transplant. Venovenous and venoarterial ECMO were implanted in eight and nine cases, respectively. Median duration of support was 3.2 days (range, 1 to 49 days). Mean patient follow-up was 2.3 years. Results. Thirty-day, 1-year, and 3-year survivals were 81%, 74%, and 65%, respectively, for the supported patients and 93%, 78%, and 62% in the control group (p 0.56). Two-year survival was not affected by ECMO type, with survival of five out of nine patients supported by venoarterial ECMO vs seven out of eight patients supported by venovenous ECMO (p 0.17). At 1- year follow-up, allograft function for the ECMO-supported patients did not differ from the control group (forced expiratory volume in one second, 2.35 L vs 2.09 L, p 0.39) (forced vital capacity, 3.06 L vs 2.71 L, p 0.34). Conclusions. Extracorporeal membrane oxygenation as a bridge to lung transplantation is associated with higher perioperative mortality but acceptable mid-term survival in carefully selected patients. Late allograft function did not differ in patients who received ECMO support before lung transplant from those who did not receive ECMO. (Ann Thorac Surg 2011;92: ) 2011 by The Society of Thoracic Surgeons Accepted for publication April 29, Presented at the Forty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31-Feb 2, Address correspondence to Dr Bermudez, UPMC-Presbyterian University Hospital, 200 Lothrop St, Ste C900, Department of Cardiothoracic Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA 15213; bermudezc@upmc.edu. Lung transplant is considered an established treatment option for patients with advanced chronic respiratory failure. Even though an increasing number of lung transplants are being performed worldwide, the scarcity of available organs and regional factors are associated with a considerable mortality on the waiting list, ranging from 5% to 10% [1]. Efforts to prioritize urgency and improve organ allocation have improved organ utilization, but simultaneously, the increasing demand for lung transplants has forced the clinicians to consider patients in more advanced stages of respiratory failure, who were historically considered poor transplant candidates. Recent reports of transplant candidates requiring mechanical ventilation (MV) and extracorporeal membrane oxygenation (ECMO) showed that despite higher early mortality, surviving patients may have comparable long-term outcomes as those who did not require mechanical support [2]. Extracorporeal membrane oxygenation has been historically utilized mainly as a rescue therapy to support patients after lung transplant in cases of severe primary graft dysfunction (PGD) [3, 4]. Based on the initial poor outcomes reported with the use of ECMO in the treatment of severe adult respiratory failure [5], ECMO was rarely considered to support patients before lung transplantation. Since the first successful clinical series of lung transplants, reported in 1987 [6], sporadic cases have been published using ECMO as a bridge to primary or redo lung transplant, with mixed results [7 11]. In the last decade, important advances in the technology for ECMO support have reignited its use in patients with advanced cardiac and respiratory failure [12]. The significant improvement in outcomes observed in recent randomized trials of ECMO use for severe adult respiratory failure [13] in association with the improvements in ECMO technology has impacted the strategy utilized by some transplant centers, making them willing to consider ECMO support as a mechanism to bridge to lung transplant in some patients. In this single-center study, we 2011 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg BERMUDEZ ET AL 2011;92: ECMO AS A BRIDGE TO LUNG TRANSPLANT analyzed our experience with ECMO support as a bridge to lung transplant, assessing surgical considerations, mid-term outcomes, and causes of early and late mortality. Patients and Methods Study Protocol We performed a retrospective analysis of patients, from March 1991 to October 2010, who underwent lung transplant (primary and re-transplantation), while on ECMO support. Data was obtained from the University of Pittsburgh Medical Center transplant database and patient charts. This study was approved by the Total Quality Council of our institution. The informed consent requirement was waived. Patient Selection Extracorporeal membrane oxygenation was selectively implanted to support patients with advanced respiratory failure unresponsive to maximal MV support, including elevated inspired oxygen fraction, positive end-expiratory pressure and use of nitric oxide. Lung transplant was considered in patients supported on ECMO for respiratory failure in selected cases in which the patient presented a rapid deterioration of a chronic lung disease while on the waiting list or during the lung transplant evaluation process and with no standard contraindication for transplantation. Occasionally, patients on ECMO support for PGD after a primary lung transplant, with the absence of lung recovery despite adequate treatment, were considered for redo lung transplant. Programmatically, the decision was made not to include patients with other established organ dysfunctions, including chronic or established renal failure on hemodialysis, severe liver failure, or the presence of a severe systemic infection. Patients who had lung transplant without the use of preoperative ECMO during the period analyzed served as a control group. Donor Criteria Standard procurement techniques were used in all cases. Donors considered included those with adequate oxygenation, lung compliance, and absence of signs of active infection or aspiration on bronchoscopy. Mean age of the donors used for patients supported on ECMO was 28 years ( 15 years) versus 34 years ( 14 years) for the control group ( p 0.11). The mean PaO 2 before procurement, for the patients supported on ECMO, was 399 mm Hg ( 118 mm Hg) versus 392 mm Hg ( 145 mm Hg) for the control group (p 0.85). The preservation solution was University of Wisconsin solution in the early phase of the program and Perfadex (Vitrolife, Englewood, CO) after Donors with extended criteria, including age 55 years and smoking history, were not excluded from this study. ECMO Considerations Venoarterial (VA) and venovenous (VV) ECMO techniques were used depending on patient hemodynamic stability and presence of severe pulmonary hypertension at the time of implantation. Peripheral cannulation was the preferred method of support with the VA and VV configurations and was used in 16 of 17 patients (94%). Dual VV ECMO cannulation, through the right internal jugular and right femoral vein, was used in 5 patients. A single VV ECMO cannulation strategy, through the right internal jugular vein using the Avalon Elite Bicaval Dual Lumen Catheter (Rancho Dominguez, CA) was used in 3 patients according to surgeon s preference [14]. More recently, we have favored the VV strategy, preferentially using a peripheral approach because of its potential advantages over VA support. Peripheral VA ECMO was used in femoral-femoral configuration in 8 patients and central aorta-right atrium configuration in 1 patient. Recently, we have only used percutaneous VA ECMO in the presence of hemodynamic deterioration and in patients with severe pulmonary hypertension and significant right ventricular dysfunction. In such cases, VV ECMO may be associated with poor response. The case with central cannulation, in a patient with recent lung transplant and PGD, was considered for redo lung transplant during the early phase of our program. Technical aspects of ECMO implantation at our institution have been previously published [3, 15]. A Sci-Med Oxygenator (Sci-Med Life Systems Inc, Minneapolis, MN) was used early in the study period in 5 patients (30%). Since 2001, a Medtronic Carmeda (Minneapolis, MN) heparin-bonded Affinity Oxygenator was used alone 2 cases (12%). A QuadroxD (Maquet Cardiopulmonary, Rastatt, Germany ) has been our preferred oxygenator since 2008, used alone in 4 patients (23%). A combination between Medtronic Carmeda heparin-bonded Affinity Oxygenator and the QuadroxD was used in 6 patients (35%). A centrifugal pump (BP-80; Medtronic, Biomedicus, Eden Praire, MN) was used in 14 patients. Levitronix Centrimag (Levitronix LLC, Waltham, MA) was used in 2 patients and a Jostra RotaFlow pump (Maquet Cardiopulmonary, Rastatt, Germany) was used in 1 patient. Heparin (5000 U) was administered at the time of ECMO implantation and then titrated to maintain an activated clotting time of 140 seconds in the VV group and seconds in the VA group. Data Analysis Continuous variables are shown as mean SD, median (range). Actuarial survival estimates were calculated using Kaplan-Meier life table analysis. The log rank statistic was used to determine whether the survival curves differed, with p 0.05 considered statistically significant. Statistical analysis was performed using STATA, version 8.2 (STATA Corp, College Station, TX). Results 1227 Of the 1305 lung and heart-lung transplants performed during the study period, 17 (1.3%) received ECMO support undergoing subsequently lung transplantation (ECMO group) and 1288 had no ECMO support before transplant (control group). Three patients died while on GENERAL THORACIC

3 1228 BERMUDEZ ET AL Ann Thorac Surg ECMO AS A BRIDGE TO LUNG TRANSPLANT 2011;92: support and were not transplanted, with an average time of support of 15 hours (range, 1 to 41 hours). The reasons for ECMO withdrawal were presence of neurologic dysfunction due to severe hypoxia previous to ECMO implantation in one case; inability to obtain adequate flow immediately after ECMO implantation due to system thrombosis; and high degree of sensitization with multiple strong antibodies that, associated with a small chest cavity size, made the chances of obtaining a suitable donor unlikely, in the third case. Within the ECMO group, 5/17 (29%) patients had ECMO as a bridge to lung transplant in the 1990s (between 1991 and 1993). The other 12/17 (71%) patients had pre-transplant ECMO after Patients in the ECMO group were younger and more had redo lung transplant than in the control group (Table 1). In the ECMO group, more patients underwent double lung transplant (15/17; 88%) than in the control group (696/1288; 54%). The mean ischemic time was higher in the ECMO group. The average perioperative (intraoperative and first 24 hours postoperative) red blood cell, fresh frozen plasma and platelet transfusion was 11.5 ( Table 1. Patient Characteristics Characteristics ECMO Support No ECMO Support p Value Total patients 17 1,288 Mean age (years) Sex (%) Male 7 (41%) 659 (51%) 0.47 Female 10 (59%) 629 (49%) Underlying disease Cystic fibrosis 4 (23%) 159 (12%) 0.25 Pulmonary fibrosis 6 (35%) 273 (21%) 0.22 COPD 1 (6%) 412 (32%) 0.01 PH 0 92 (7%) 0.62 Retransplant 6 (35%) 60 (4%) Other (23%) 0.01 Type of transplant Single Right Lung (17%) 0.05 Single Left Lung 1 (6%) 309 (24%) 0.09 Double Lung 15 (88%) 696 (54%) Heart-Lung 1 (6%) 58 (5%) 0.17 Ischemic time Mean (minutes) Median, Range 369, , Type of ECMO VA 9 VV 8 Time of ECMO support before LTx (hours) Mean 177 Median, Range 78, COPD chronic obstructive pulmonary disease; ECMO extracorporeal membrane oxygenation; LTx lung transplant; PH pulmonary hypertension; VA venoarterial; VV venovenous. 4.7) units/patient, 5.8 ( 4.4) units/patient and 2.1 ( 1.7) units/patient, respectively. Significant acute cellular rejection, A2, was present in 2/12 (16%) patients, and was treated with intravenous steroids (Solumedrol) boluses, without subsequent clinical manifestations of obliterative bronchiolitis syndrome. This data was not available for the 5 patients transplanted between 1991 to Before ECMO implantation, respiratory stabilization with standard MV support was attempted in all 17 patients in the ECMO group, with a median time of MV support of 36 hours (range, 1 to 624 hours). Arterial blood gas measurements in recipients at the time of ECMO implantation were: mean hydrogen ion concentration (ph) 7.2 ( 0.06), mean partial pressure of carbon dioxide (PaCO 2) 77 mm Hg ( 37 mm Hg), and mean alveolar oxygen pressure (PaO 2 ) 57 mm Hg ( 11 mm Hg) on maximal MV support. The median ECMO support time prior to lung transplant in the 17 patients was 78 hours (range, 1 to 1176 hours). Median support time was 120 hours (range, 3 to 1176 hours) for the 8 patients undergoing VV ECMO versus 48 hours (range, 1 to 196 hours) for the 9 patients undergoing VA ECMO (p 0.13). During ECMO support, a mean arterial clotting time (ACT) of 143 seconds ( 3 seconds) was obtained. Eight patients (48%) who received ECMO before transplant also required ECMO support for PGD after lung transplant, with a median time of support posttransplantation of 109 hours (range, 19 to 436 hours), versus 95 patients (7.3%) in the control group that required ECMO support posttransplantation (p ). Median intensive care unit (ICU) length of stay, after lung transplantation was 16 days (range, 3 to 40 days) in the ECMO group versus 6 days (range, 1 to 272 days) in the control group (p 0.99). The median time of postoperative MV for the ECMO group was 12 days (range, 2 to 20 days). Prolonged postoperative MV ( 5 days) was required in 8 of 17 (48%) in the ECMO group versus 347 of 1,288 (27%) patients in the control group (p 0.09). The mean follow-up time for patients receiving ECMO support before lung transplant was 2.3 years. Adverse events in the ECMO group included renal failure in 6 of 17 patients (35%), with 4 (23%) requiring temporary dialysis. Nine patients (52%) presented associated pulmonary infections requiring antibiotic treatment. Sepsis of differing etiologies occurred in the postoperative course of 7 of 17 (41%) patients, and was the direct cause of in-hospital mortality in 3 (17%). Bacterial infection was the cause of sepsis in 3 patients and fungal infection was the cause in 4 patients. Aspergillus was identified in 3 of 4 fungal septicemias. Tracheostomy was required in 7 of 17 patients (41%). Three episodes of distal digital ischemia and subsequent gangrene occurred, 2 (11%) in the upper extremities (one of them related to arterial line complication) and 1 (5%) in the lower extremities. Postoperative gastrointestinal bleeding secondary to colonic ulcer occurred in 1 patient, leading to an emergency partial colectomy. One patient developed a large hemorrhagic cerebrovascular accident late (47 days) after transplant while recovering from the surgery. Adverse events re-

4 Ann Thorac Surg BERMUDEZ ET AL 2011;92: ECMO AS A BRIDGE TO LUNG TRANSPLANT Table 2. One-Year Posttransplant Pulmonary Function Tests Respiratory Test ECMO No ECMO p Value FVC FEV FEV 1 /FVC ECMO extracorporeal membrane oxygenation; FEV 1 forced expiratory volume in 1 second; FVC forced vital capacity; LTx lung transplant GENERAL THORACIC lated to the ECMO support included significant bleeding from the arterial femoral cannulation site requiring intervention in 1 patient and transient encephalopathy of unclear etiology while on ECMO support with spontaneous resolution in 1 patient. Survival One year after transplant, lung function, assessed by pulmonary functions tests, did not differ between the patients requiring ECMO before lung transplant and patients who did not require ECMO support (Table 2). Unadjusted survival at 30 days, 1-year, and 3-years was 81%, 74%, and 65%, respectively, in the ECMO group and 93%, 78%, and 62%, respectively, in the control group (p 0.562) (Fig 1). Survival was improved in patients treated since 2005 as compared with the patients in the 1990 s (p ) (Fig 2). Thirty-day, 1-year, and 2-year survival was 63%, 63%, and 51% for the 9 of 17 patients supported with VA ECMO and 100%, 85%, and 85% for the 8 of 17 patients supported by VV ECMO (p ) (Fig 3). Thirty-day, 1-year and 3-year survival for the patients that had ECMO support before primary LTx (11 patients) was 100%, 89%, and 74% versus 50%, 50%, and 50% for the patients that required ECMO before retransplantation (6 patients) (p ). The causes of early ( 30 days) and late mortality for the ECMO group are depicted in Table 3. One case of Fig 2. Kaplan-Meier 3-year survival in patients supported by extracorporeal membrane oxygneation (ECMO) before lung transplantation in the 1990s versus 2000s. severe Aspergillus infection occurred 411 days after ECMO, leading to sepsis and subsequent death. The other late mortality happened 47 days after ECMO, after a lung transplant postoperative course complicated by cerebral vascular accident, respiratory insufficiency and multiple sepsis events, culminating with multiple organ failure. Comment We report, in this study, our single-center experience over a 19-year period with the use of ECMO support as a bridge to lung transplant. Despite higher perioperative morbidity and mortality, in a well-selected population, mid-term patient outcomes and allograft function were similar to those in patients who did not receive ECMO support before lung transplant. The median ECMO support time in this series was limited compared with other published series [8 11], and is probably an important factor in the improved outcomes observed in this series, Fig 1. Kaplan-Meier 3-year survival in patients supported by extracorporeal membrane oxygenation before lung transplantation (ECMO pre-ltx) versus patients not supported by ECMO before lung transplantation (control). Fig 3. Kaplan-Meier 2-year survival in patients supported by venovenous (VV) extracorporeal membrane oxygenation (ECMO) before lung transplantation versus patients supported by venoarterial (VA) ECMO before lung transplantation.

5 1230 BERMUDEZ ET AL Ann Thorac Surg ECMO AS A BRIDGE TO LUNG TRANSPLANT 2011;92: Table 3. Causes of Early and Late Mortality Among Patients Supported With Extracorporeal Membrane Oxygenation Before Lung Transplantation Mortality Number of Patients Causes of Mortality Early (0 to 30 days) 2 Infection 1 Refractory cardiogenic shock Late ( 30 days) 1 Infection 1 Multiple organ failure indirectly reflecting the importance of institutional experience and transplant volume when considering this type of strategy. Due to the notable donor scarcity, historically, the use of mechanical support, including ECMO, has been considered as a contraindication to lung transplant by some physicians. Since the first report of a patient bridged on ECMO to lung transplant in 1975 [16], a few case reports and limited institutional series have demonstrated the feasibility of this strategy [7 11]. However, this technique has not been consistently considered a valid alternative in most lung transplant centers. The lack of uniform ECMO support criteria and the prolonged time on the waiting list needed in most low-volume and mid-volume centers, make this alternative impractical in many instances. The recent improvements in ECMO technologies, including the use of more biocompatible polymethylpentene oxygenators, the new generation of centrifugal pumps and heparin-coated circuits and cannulas, have facilitated ECMO support in all patients, including those requiring lung transplants. This newer ECMO technologies have led to less frequent oxygenator failure and less hemolysis and blood factors activation, allowing more prolonged support time without the physiologic and inflammatory derangements seen in the early days of ECMO support. Other strategies, including the use of pumpless oxygenator support have also emerged as an alternative to bridge patients to lung transplant with acceptable outcomes, especially in less advanced cases with partial respiratory failure and CO 2 retention [17, 18]. In this series, we considered only patients active on our institutional waiting list or who decompensated during the lung transplant evaluation process. Programmatically, we have not considered ECMO use as a bridge to lung transplant in acute respiratory failure syndromes or in patients transferred on support from other institutions, in light of the ethical and organ allocation considerations, although recent reports have shown acceptable outcome in this situation [9]. A limited group of patients were considered for lung re-transplant following severe allograft dysfunction. During the early phase of clinical lung transplantation, ECMO was mainly utilized by some centers to support patients with severe PGD and in some cases as a bridge to re-transplantation, as previously presented by the Hannover group [19]. This was also the case in our institution, in which early re-transplantation was considered just days following primary graft dysfunction. We performed four re-transplantation cases during the early 1990 s. In the current era, lung retransplantation has been considered only in cases with severe parenchymal involvement, in the absence of infection, rejection and other organ dysfunction (kidney or liver). Two cases were considered after 2005, with ECMO support of 21 and 44 days. Similar to the report presented by Mason and coworkers [2], our patients supported on ECMO before lung transplant were younger than patients who did not receive ECMO support, representing an evident selection bias for ECMO use in patients with higher chances of survival. Patients with COPD were less prevalent in the group supported on ECMO differing from other previous publications [2]. In the early programmatic phase ( ), pulmonary re-transplant was the primary diagnosis considered for pre-transplant support with a clear change in the recent period, after 2005, when patients with idiopathic pulmonary fibrosis or cystic fibrosis were the most frequently supported patients. Similiar to other [9, 11], not surprisingly, the patients who received ECMO support had a significantly higher rate of double lung transplant. This may be a reflection of the common belief that patients with more advanced lung disease may do better and have a less complicated postoperative course when receiving double lung transplant. Another important observation was that despite the use standard donors in all cases, the ECMO group had higher rate of PGD and a higher incidence of postoperative ECMO than the control group. Factors that may have influenced this higher PGD rate are the more prolonged ischemic time and the need for cardiopulmonary bypass or ECMO support to complete the allograft implantation in all cases. Relevant morbidities, including pulmonary and other infections, renal failure requiring dialysis and vascular complications, were not infrequent and were comparable to morbidity on ECMO support reported previously [11]. Although patients on the ECMO group had significant perioperative morbidity, short- and mid-term patient outcomes were not significantly different, in this small series, between the patients who received ECMO and the patients who did not. Additionally, mid-term allograft function in patients undergoing ECMO before lung transplant did not differ from the control group and is consistent with data previously presented by our group and others [3, 11], emphasizing the absence of a direct long-term effect of ECMO on pulmonary structure and function. The type of ECMO support utilized to bridge patients (VA or VV) did not affect mid-term survival, similar to the previously published data with the use of ECMO for PGD in our institution [3]. The need for more prolonged ECMO support before lung transplant in this patient population observed in other institutions should be considered and could favor the use of VV ECMO, with the objectives of minimizing vascular and neurological complications, which have been previously reported with VA ECMO [20]. The use of pumpless oxygenators, requiring smaller size cannulas, for less sick patients could become

6 Ann Thorac Surg BERMUDEZ ET AL 2011;92: ECMO AS A BRIDGE TO LUNG TRANSPLANT a viable alternative to VA ECMO support, especially in selected patients with severe pulmonary hypertension, as recently reported [17]. There were notable outcome differences between our early and late experiences, probably reflecting improvement in patient management and selection, as well as in the ECMO technology. Despite this promising result, we continue to consider this strategy only in highly selected cases, generally in young patients with recent rapid deterioration of a chronic pulmonary disease, which had permitted adequate functional status prior to deterioration, and in the absence of other chronic organ dysfunction. The consideration for the use of ECMO as a bridge to lung transplantation in low-volume centers should be judiciously considered, especially in cases with prolonged expected waiting times, to prevent futile support. This strategy should also not be considered a standard procedure in high-volume centers experienced in management of ECMO support, as the implications on longterm patient and allograft outcomes and organ allocation are still not well elucidated. Limitations of this study include the small number of patients included, retrospective nature of the analysis, inclusion of patients during different periods of the program and differences in support including VA and VV ECMO. A carefully designed, prospective multicenter study would allow more definitive conclusions. In conclusion, we demonstrate that ECMO support can be used as a bridge to lung transplantation and is associated with acceptable mid-term outcomes, despite higher perioperative morbidity and mortality, in carefully selected patients. Late allograft function did not differ in patients who received ECMO support before lung transplant as compared with those who did not. References 1. Accessed November Mason D, Thuita L, Nowicki E, et al. Should lung transplantation be performed for patients on mechanical respiratory support? The US experience. J Thorac Cardiovasc Surg 2010;139: Bermudez C, Adusumilli P, McCurry K, et al. Extracorporeal membrane oxygenation for primary graft dysfunction after lung transplantation: long-term survival. Ann Thorac Surg 2009;87: Fischer S, Bohn D, Rycus P, et al. Extracorporeal membrane oxygenation for primary graft dysfunction after lung transplantation: analysis of the Extracorporeal Life Support Organization (ELSO) registry. J Heart Lung Transplant 2007;26: Zapol W, Snider M, Hill J, et al. Extracorporeal membrane oxygenation in severe acute respiratory failure. A randomized prospective study. JAMA 1979;242: Cooper J, Pearson F, Patterson G, et al. Technique of successful lung transplantation in humans. J Thorac Cardiovasc Surg 1987;93: Jackson A, Cropper J, Pye R, et al. Use of extracorporeal membrane oxygenation as a bridge to primary lung transplant: 3 consecutive, successful cases and a review of the literature. J Heart Lung Transplant 2008;27: Olsson KM, Simon A, Strueber M, et al. Extracorporeal membrane oxygenation in nonintubated patients as bridge to lung transplantation. Am J Transplant 2010 ;10: Haneya A, Philipp A, Mueller T, et al. Extracorporeal circulatory systems as a bridge to lung transplantation at remote transplant centers. Ann Thorac Surg 2011 ;91: Garcia J, Iacono A, Kon Z, et al. Ambulatory extracorporeal membrane oxygenation: a new approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg 2010;139:e Hämmäinen P, Schersten H, Lemström K, et al. Usefulness of extracorporeal membrane oxygenation as a bridge to lung transplantation: a descriptive study. J Heart Lung Transplant 2011;30: elso.med.umich.edu/registry.html.accessed November Peek G, Mugford M, Tiruvoipati R, et al. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet 2009;374: Bermudez C, Rocha R, Sappington P, et al. Initial experience with single cannulation for venovenous extracorporeal oxygenation in adults. Ann Thorac Surg 2010;90: Glassman LR, Keenan RJ, Fabrizio MC, et al. Extracorporeal membrane oxygenation as an adjunct treatment for primary graft failure in adult lung transplant recipients. J Thorac Cadiovsc Surg 1995;110: Veith FJ. Lung transplantation. Transplant Proc 1977;9: Fischer S, Simon AR, Welte T, et al. Bridge to lung transplantation with the novel pumpless interventional lung assist device NovaLung. J Thorac Cardiovasc Surg 2006;131: Fischer S, Hoeper M, Tomaszek S, et al. Bridge to lung transplantation with the extracorporeal membrane ventilator Novalung in the veno-venous mode: the initial Hannover experience. ASAIO J 2007;53: Jurmann MJ, Haverich A, Demertzis S. Extracorporeal membrane oxygenation as a bridge to lung transplantation. Eur J Cardiothorac Surg 1991;5: Hartwig M, Appel J 3rd, Cantu E 3rd, et al. Improved results treating lung allograft failure with venovenous extracorporeal membrane oxygenation. Ann Thorac Surg 2005;80: GENERAL THORACIC DISCUSSION DR SETH FORCE (Atlanta, GA): Doctor Mathisen, Dr Reed. I would first like to congratulate Dr Bermudez and the University of Pittsburgh lung transplant program for their excellent presentation and their outstanding results. This study, like other similar recent studies, highlights the need for artificial lung support and represents another step forward towards the pursuit of a bridge towards lung transplantation. Doctor Bermudez, I have two questions. My first question, in your results you showed one- year and three-year survivals of 88% and 74%, respectively, in the extracorporeal membrane oxygenation (ECMO) group, however, only 12 out of 17 patients, or 70%, survived to discharge. Could you please explain how your one-year and three- year survival rates were higher than your survival rates to discharge in this patient population?

7 1232 BERMUDEZ ET AL Ann Thorac Surg ECMO AS A BRIDGE TO LUNG TRANSPLANT 2011;92: DR BERMUDEZ: We had three patients that died within 30 days. Our two late mortalities were in a patient that was not able to be discharged and died 47 days after her transplant and the other case that was discharged died 15 months later. DR FORCE: Just to clarify, there were 5 patients who died in the ECMO group out of 17 patients before they were discharged from the hospital, survival to discharge. DR BERMUDEZ: The reality is that only one of them died after discharge from the hospital, late mortality, from a fungal infection. DR FORCE: In the abstract and the manuscript there was between a 66% and 70% survival to discharge, which is significantly lower than the one-year and three-year survival rates that are presented. I think it is something you just need to go back to. DR BERMUDEZ: We definitely will review the survival presented on the manuscript version sent to you as we may have a transcription error. We mentioned on the early abstract survival to discharge of 65%, but really meant overall survival. To clarify, our 30-day survival was 81%, 1-year survival was 74%, and 3-year survival was 65%. Again, we apologize for the transcription error. DR FORCE: Let me get to my second question. Part of the reason for your good results is your short ECMO duration, a median of three days and a mean of seven days. Other recent reports such as one out of Sweden that was published in last month s Journal of Heart and Lung Transplantation had a mean ECMO duration of 17 days and three patients died on ECMO while waiting for a lung transplantation. This is two-part question. First, what is your cutoff as for how long you would leave a patient on ECMO while waiting for organs? Second, how do you explain your ability to find organs so quickly? For example, did you have to broaden your donor criteria to get organs more quickly and do you see this as a limiting factor in the feasibility of ECMO as usefulness for a bridge to lung transplantation? Thank you. DR BERMUDEZ: I concur with all your comments, and this is one of the things that we were very curious to analyze when we went back to review these patients. Yes, our waiting time on ECMO was shorter than other series recently published, and I think it is basically related to the activity of the program. We performed this year probably 130 cases with evidently large number of local and external offers. We did not consider any different approach or expanded our donor criteria for patients supported on ECMO and may be reflective only of our local practice. Even though we had one patient that stayed up to 46 days on support, frequently we were able to get lungs earlier compared to other centers, and that is really brings up a different question. Should this type of support pre-lung transplant be considered in small volume centers? I am not sure about that answer. The majority of the recent reports have been by highvolume centers, so I think this is something to keep in mind for future considerations. In terms of the donors, again, I want to emphasize that the donors that we took were regular donors that we could have considered in any other high LAS Score patient. We did not make a difference in patient on ECMO support. Having said that, we have recently favored the use of VV ECMO, due to the lower risk of vascular complications because we understand, especially in IPF patients, that the support may be longer due to size restrictions. In terms of duration of support, we don t have a preconceived time of support; we withdraw ECMO in the presence of progressive multi-organ dysfunction or major vascular or hemorrhagic complications. The moment the patient goes into renal or hepatic failure or we see important sepsis, we generally discontinue support, and we have three of 20 patients that we supported on ECMO as a bridge to lung transplant during this period that were discontinued for these reasons and were not included in the study as they did not undergo transplant. DR MICHAEL MADANI (San Diego, CA): Christian, I really enjoyed your presentation. One quick question. In terms of inclusion criteria, can you give us an idea of the patient population? Did you have a specific age cutoff or any particular diagnosis? DR BERMUDEZ: Thank you for your question Dr Madani. When we started doing this in the second phase, after 2005, we were not clear about the age cutoff. We started with 55 years old but we actually have moved up in age limit and are willing to consider some older patients, with no significant co morbidities, in well selected cases. We are selecting patients that are on the list with an acute decompensation, previously active, ideally, with only recent hospitalization. The average ventilator support of these patients was 36 hours. So patients were recently intubated with very acute decompensation. The reality is that, this point still a matter of debate within the group, considering the limited experience, and I think would be an important discussion among the transplant centers that are considering this strategy. We should be able to collect data and potentially use registries like Extracorporeal Life Support Organization or others to be able to answer this question.