Primary graft failure (PGF) is a life-threatening condition

Extracorporeal Membrane Oxygenation in Primary Graft Failure After Heart Transplantation ADULT CARDIAC Silvana F. Marasco, FRACS, Matthew Vale, PhD, Vince Pellegrino, FJFICM, Arthur Preovolos, CCP, Angeline Leet, FRACP, Ashley Kras, Elliot Schulberg, Peter Bergin, FRACP, and Donald S. Esmore, FRACS Cardiothoracic Surgery Unit, Intensive Care Unit, Department of Perfusion, and Department of Cardiology, The Alfred Hospital, Melbourne; and Department of Surgery, Monash University, Victoria, Australia Background. The aim of this review was to analyze our results with extracorporeal membrane oxygenation (ECMO) support for primary graft failure (PGF) in heart transplant recipients. Methods. A retrospective review of 239 consecutive patients who underwent heart transplantation between January 2000 and August 2009 was performed. Orthotopic, heterotopic, and heart lung transplants were included in this analysis. Over that time period, 54 patients developed PGF, of whom 39 patients required ECMO support. These 39 patients form the basis of this review. Results. Thirty-four patients (87%) were successfully weaned from ECMO and 29 (74.3%) survived to hospital discharge. There were no significant differences in wean rates or complications between central and peripheral ECMO. Comparison of survival in the 39 ECMO patients to the non-pgf patients (n 185) showed a significantly worse survival in the ECMO group (p 0.007). When those patients who died in the first 30 days were excluded, there was no difference in overall survival between groups (p 0.73). Conclusions. Extracorporeal membrane oxygenation provides excellent circulatory support for patients with PGF after heart transplantation with good wean and survival to discharge rates. (Ann Thorac Surg 2010;90:1541 47) 2010 by The Society of Thoracic Surgeons Primary graft failure (PGF) is a life-threatening condition that is the leading cause of early death after cardiac transplantation. According to International Society of Heart and Lung Transplantation data, PGF accounts for 27.1% of deaths within 30 days of heart transplantation [1]. It is defined as severe dysfunction of the cardiac allograft without obvious anatomic or immunologic cause [2] and is characterized by hypotension, high filling pressures, and refractory low-cardiac output. Primary graft failure covers a spectrum of clinically apparent allograft dysfunction, including right heart dysfunction without evidence of a high afterload, isolated left ventricular dysfunction, and biventricular dysfunction [3]. We have previously reported on our improved outcomes with early institution of mechanical circulatory support in these patients [4, 5]. Due to the limitations of extracorporeal membrane oxygenation (ECMO) a decade ago, we managed these patients with implantation of a Thoratec paracorporeal ventricular assist device (PVAD) (Thoratec Lab, Berkeley, CA) for left ventricular failure, and used a centrifugal pump (Medtronic Biomedicus, Minneapolis, MN) for right ventricular support. With the improvement in available ECMO circuits, and the significant cost reduction in using ECMO compared with a Accepted for publication May 25, 2010. Address correspondence to Dr Marasco, Cardiothoracic Surgery, The Alfred Hospital, Commercial Rd, Prahran, Melbourne, Australia 3181; e-mail:s.marasco@alfred.org.au. dedicated VAD, we now exclusively manage our cardiac allograft PGF with ECMO. Thus the aim of this review was to analyze our results in this current era with ECMO support and to compare our results with central versus peripheral cannulation for ECMO. Patients and Methods A retrospective review of prospectively collected data on 239 patients who underwent heart transplantation at our institution between January 2000 and August 2009 was performed. Orthotopic and heterotopic heart transplants, as well as heart-lung transplants, were included in this analysis. Over that time period, 54 patients developed PGF, of whom 39 patients required ECMO support. These 39 patients form the basis of this review. Institutional ethics approval was granted for this review and individual patient consent was waived. Definition of Primary Graft Failure Primary graft failure was defined as hypotension with a systolic blood pressure less than 90 mm Hg, low cardiac output (cardiac index 2.0 L/minute/m 2 ), and pulmonary capillary wedge pressure greater than 20 mm Hg after coming off cardiopulmonary bypass despite low-dose inotropic support. This analysis includes patients with right or left or biventricular failure in the absence of other causes such as hyperacute rejection. 2010 by The Society of Thoracic Surgeons 0003-4975/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2010.05.066

ADULT CARDIAC 1542 MARASCO ET AL Ann Thorac Surg PRIMARY GRAFT FAILURE AND HEART TRANSPLANT 2010;90:1541 47 Abbreviations and Acronyms ECMO extracorporeal membrane oxygenation IABP intraaortic balloon pump INR international normalized ratio LV left ventricle LVAD left ventricular assist device PGF primary graft failure PVAD paracorporeal ventricular assist device RV right ventricle Mechanical Circulatory Support Protocol Our algorithm for the initiation of mechanical circulatory support for a dysfunctional transplanted cardiac allograft is outlined in Figure 1. We have moved away from using an intraaortic balloon pump (IABP) during ECMO as we feel it obstructs ECMO flow in the aorta particularly in patients on peripheral ECMO. We also feel the potential risk to the femoral artery outweighs the benefits that IABP can offer in a patient already fully supported on ECMO. The choice of central or peripheral ECMO cannulation is up to surgeon preference. We currently use the Maquet Quadrox-D oxygenator and Maquet Rotaflow centrifugal pump (Hirrlingen, Germany) with heparin coated circuit tubing and either Biomedicus Carmeda coated peripheral cannulae (Medtronic, Minneapolis, MN) or the Abiomed central cannulae (Abiomed, Danvers, MA). When using peripheral ECMO we always use a small downflow cannula into the superficial femoral artery to ensure arterial flow to the leg is maintained. The advantages of central versus peripheral cannulation have been discussed elsewhere [6]. In brief, the choice of central cannulae allows better flow, antegrade flow through the arch without admixing with deoxygenated blood, avoids lower limb vascular complications, and the Abiomed cannulae allow closure of the sternum which reduces bleeding and infection risk. The drawback is the need to reopen the sternum to remove the cannulae although this does allow evacuation of any pericardial collection, inspection of the heart, and even biopsy if required. The peripheral cannulae allow the sternum to be closed without having to go back in to that surgical field. However, it is more difficult to achieve high flows which may be necessary in larger patients, (retrograde flow is delivered up the descending aorta with the risk of admixing in the arch) and there is always the risk of lower limb vascular complications. Peripheral cannulae are removed under direct vision in the operating room with formal repair of the vessels. Despite the theoretic problem of poor flow with peripheral ECMO, this has rarely been a problem, although in very large patients or those with very small femoral vessels we deploy central rather than peripheral ECMO. The intraoperative management of anticoagulation once the decision is made to institute ECMO is as follows. If heparinization has been reversed with protamine after coming off cardiopulmonary bypass, then ECMO is instituted with a 5,000 U bolus of heparin. If the patient cannot be weaned off cardiopulmonary bypass, then ECMO is instituted in the fully heparinized patient. Heparin is ceased but no protamine is given unless bleeding cannot be controlled surgically, in which case blood products may also be required as indicated by formal blood counts, clotting profile, and thromboelastograph results. Postoperatively, a heparin infusion is commenced within 6 to 12 hours in the nonbleeding patient. We aim for an activated partial thromboplastin time of 60 to 70 seconds, which is measured six hourly. Plasma free hemoglobin is also measured six hourly and daily measurements of hemoglobin, platelet count, and full clotting profile are performed. Blood products are transfused to ensure the platelet count remains above 80,000 per microliter, hemoglobin remains above 8.0 g/dl, and the INR (international normalized ratio) remains below 1.6. In the bleeding patient in whom a surgical cause has been excluded, heparin is ceased and not recommenced until bleeding has stopped for 12 hours. Cryoprecipitate is given to normalize fibrinogen counts, and fresh frozen plasma and Prothrombinex (CSL Limited, Parkville, VIC, Australia) given to achieve an INR less than 1.3. Tranexamic acid infusion is commenced and if bleeding is still an issue, recombinant factor VIIa may be given. The ECMO weaning is first considered after approximately 72 hours (Fig 2). Statistical Analysis All analysis was conducted using STATA 11(StataCorp, College Station, TX). Univariate analysis was performed using 2 tests for equal proportion, and the Student s t test and Wilcoxon rank sum test for nonparametric data. Actuarial survival of patients was presented using Kaplan-Meier curves with p values calculated using logrank tests. A 2-sided p value less than 0.05 was considered to be statistically significant. Fig 1. Algorithm for ECMO institution. (CPB cardiopulmonary bypass; ECMO extracorporeal membrane oxygenation; IABP intraaortic balloon pump; PCWP pulmonary capillary wedge pressures; PGF primary graft failure; SBP systolic blood pressure.)

Ann Thorac Surg MARASCO ET AL 2010;90:1541 47 PRIMARY GRAFT FAILURE AND HEART TRANSPLANT Fig 2. Algorithm for ECMO weaning. (ECMO extracorporeal membrane oxygenation; ICU intensive care unit; SBP systolic blood pressure.) Results Two hundred thirty-nine patients underwent cardiac transplantation between January 2000 and August 2009, comprising 207 orthotopic heart transplants, 19 heterotopic heart transplants, and 13 heart lung transplants. Fifty-four patients developed PGF, of whom 39 required ECMO support (34 orthotopic heart transplants, two heterotopic heart transplants, and three heart lung transplants). In the three heart lung patients, ECMO was instituted primarily for cardiac support as per our algorithm. The demographics of the patient cohort are outlined in Table 1. Details of the donor organs are also outlined in Table 1. Details of the ECMO support and outcomes are outlined in Table 2. Thirty-four patients (87%) were successfully weaned from ECMO. Three patients died while on ECMO. One patient, a heart-lung recipient, died on day 2 from massive bleeding and coagulopathy. Another patient died on day 6 after bleeding complications including a large cerebral bleed. The third death was a 60-year-old male who received a 60-year-old male donor heart. The donor heart never recovered and after 10 days of support treatment was withdrawn at the request of the family. Two of the above patients had been bridged to transplantation with a PVAD. Three patients were bridged from ECMO to another support device. Two of these patients were placed on ECMO early in our experience, prior to the newer ECMO circuits and oxygenators being available. One patient was unable to be supported adequately by ECMO and a Thoratec PVAD was implanted. That patient was successfully discharged from hospital. The other patient was not placed on ECMO until almost 48 hours posttransplant. After eight days of ECMO support, he had an attempted wean of ECMO in theatre. However, due to poor RV function, mechanical support was reinstituted (Biomedicus centrifugal RVAD). The patient died two days later of multiorgan failure. The third patient was bridged to a Thoratec PVAD after 12 days of ECMO support but died 18 days later of multiorgan failure. Five patients died while in hospital, despite successful wean from ECMO. One patient arrested after a cardiac biopsy perforated the right ventricle, two patients died of sepsis, and two patients died of multiorgan failure. Thus, in total 29 patients (74.3%) were alive at discharge. Comparing central with peripheral ECMO support, 24 of 26 (92.3%) patients supported with central ECMO were able to be weaned and 21 (78%) of those were alive at discharge. Of the 16 patients supported with peripheral ECMO, 12 were successfully weaned (75%) and 9 (56%) were alive at discharge from hospital. None of these differences between central and peripheral outcomes were significant (Table 2). Complications in this patient cohort are outlined in Table 2. Circuit complications or exchanges were unusual, reflecting the improved equipment available since 2002. Renal failure and sepsis were the most common complications. Return to theatre for bleeding was also fairly frequent with a very similar incidence between centrally cannulated patients (34.6%) and peripherally cannulated patients (35.3%). There were no significant differences between groups in blood products transfused. Two patients in the peripheral ECMO group and one patient in the central ECMO group required recombinant Table 1. Demographics of ECMO Patients Characteristics No. Recipient Data: Number of patients 39 Age (years) (mean SD) 49.5 12.5 Gender 30M:9F Transplant performed 34 orthotopic heart transplants 2 heterotopic heart transplants 3 heart lung transplants Diagnosis Ischemic heart disease 10 Dilated/idiopathic 15 cardiomyopathy Congenital heart disease 5 Hypertrophic 3 cardiomyopathy Valvular heart disease 3 Other 3 Redo sternotomy 18 VAD explants 7 Donor data: Age (years) (mean SD) 39.3 13.4 Gender 25M:14F Donor organ ischemic time 272 120 (minutes) Donor cause of death 21 cerebrovascular accident 9 trauma 3 anoxic brain injury 6 other ECMO extracorporeal membrane oxygenation; assist device. 1543 VAD ventricular ADULT CARDIAC

ADULT CARDIAC 1544 MARASCO ET AL Ann Thorac Surg PRIMARY GRAFT FAILURE AND HEART TRANSPLANT 2010;90:1541 47 Table 2. Outcome Data ECMO Details All Patients (n 39) a Central Cannulation Only (n 26) Peripheral Cannulation Only (n 16) p Value Central Versus Peripheral Duration of support (days) (mean SD) 6.78 2.64 7.12 2.59 6.42 3.42 0.43 Circuit complications/interventions 2 2 0 0.26 ECMO flow at 4 hours (mean SD) 4.30 0.97 4.46 0.96 3.88 0.91 0.10 ECMO flow at 24 hours (mean SD) 4.69 0.99 4.86 1.04 4.25 0.75 0.11 Blood products transfused b : Packed cells (ml) 5,890 (4,030 9,285) 5,580 (4,340 8,835) 6,820 (3,022 12,445) 0.89 Fresh frozen plasma (ml) 2,250 (1,200 4,275) 2,100 (1,200 3,825) 3,270 (1,350 4,950) 0.45 Platelets (ml) 2,130 (1,275 3525) 2,100 (1,175 3,050) 2,750 (1,663 4,075) 0.31 Cryoprecipitate (ml) 300 (55 450) 300 (55 480) 300 (113 405) 0.96 Patient complications: Return to theatre for bleeding 15 (38.5%) 9 6 0.85 Sternal wound infection 4 (10.3%) 3 1 0.57 Gastrointestinal hemorrhage 5 (12.8%) 2 3 0.28 Other gastrointestinal complication 12 (30.8%) 7 5 0.76 Renal failure 26 (66.6%) 14 12 0.17 Pneumonia 18 (46.2%) 12 6 0.58 Sepsis 19 (48.7%) 9 10 0.08 Multiorgan failure 5 (12.8%) 2 3 0.28 Limb complication 5 (12.8%) 2 3 0.28 Major cerebral complication 3 (7.7%) 1 2 0.29 Patients outcomes: Successful wean 34 (87.2%) 24/26 12/16 0.12 Successful wean or bridge 37 (94.9%) 25/26 14/16 0.29 Successful hospital discharge 29 (74.3%) 21/26 9/16 0.087 a Three patients had their site of cannulation changed. Those three patients have been analyzed twice in this table; once during their central ECMO run and again during their peripheral ECMO run. b Blood products transfused during the ECMO period; reported as median (interquartile range). ECMO extracorporeal membrane oxygenation. factor VIIa for bleeding. Sternal wound infection occurred in three patients who were centrally cannulated but only one patient who was peripherally cannulated. Lower limb complications occurred in three patients who had peripheral femoral cannulation for ECMO. One patient in whom there were difficulties inserting the downflow cannula developed limb ischemia and the peripheral ECMO was reconfigured to central ECMO. Two patients sustained femoral arterial damage as the result of the peripheral cannulation and one of those patients also developed venous obstruction of the lower limb several hours after cannulation requiring resiting of the venous cannula and fasciotomies. The two central cannulation group limb complications were both related to percutaneous insertion of IABP. One patient developed a groin abscess and the other required a below knee amputation. Only nine patients of the entire ECMO cohort had an IABP in the postoperative setting. Overall median intensive care unit length of stay in the entire cohort was 16 days (range 3-53 days), and the overall median hospital length of stay was 23 days (range, 3 to 110 days). Comparison of survival of the 39 ECMO patients with the historically comparable non-ecmo to non-pgf patients (n 185) showed a significantly worse survival in the ECMO group (p 0.007) (Fig 3). However, this difference was driven by deaths in the first 30 days. When those patients who died in the first 30 days were excluded, there was no difference in overall survival between groups (p 0.73). Fig 3. Survival in ECMO versus non-pgf patients after heart transplantation; Kaplan-Meier curve. ( ECMO; --- non-pgf; ECMO extracorporeal membrane oxygenation; PGF primary graft failure.)

Ann Thorac Surg MARASCO ET AL 2010;90:1541 47 PRIMARY GRAFT FAILURE AND HEART TRANSPLANT Comment This series of heart transplant patients requiring ECMO in the immediate postoperative period for PGF represents one of the largest series published. The only published series larger is that by D Alessandro and colleagues [7]. However, this series included patients other than those who sustained primary graft failure (personal communication); specifically it also included acute rejection patients. Our successful wean rate of 85% and survival to discharge rate of 74% also compares favorably with other published series [8 14]. D Alessandro and colleagues have noted an increased early graft failure rate since 2003, which has also been our experience. They note that 2003 was a transitional year during which they faced increased mortality on the recipient waiting list and tried to expand the acceptance criteria for donors. We have faced a very similar problem with a low donation rate in Australia [15] as well as long distances to procure organs, which leads to longer ischemic times [16]. As our waiting list has grown we have also had to accept more marginal donors and we feel this is part of the reason for our increasing PGF rate, although it is likely to be multifactorial. We have analyzed our PGF incidence previously and found that donor age and donor organ ischemic times are the most significant predictors of PGF [4]. In detail, we found that for every hour of extra ischemic time over 4 hours, the risk for developing PGF is increased by 43% (p 0.01). We also found that the risk of developing PGF is increased 2.7% for every extra year of donor age (p 0.04). Thus our increasing incidence of PGF has led us to embrace ECMO as an early postoperative therapy in these patients. The improved ECMO equipment as well as the decreased cost compared with dedicated VAD equipment has meant it has been readily accepted as the short-term mechanical assist device of choice in our institution for these patients [6]. Although our incidence of complications is similar to others, our overall wean and survival to discharge rates are better than most other groups. We have also compared central with peripheral ECMO outcomes in this study and found no significant differences in outcomes. Thus, although we had previously preferred to use central ECMO, we are now using peripheral ECMO more often, for the reasons elucidated in the methods section. However, although there were no significant differences in flows, outcomes, or complications between the two groups, there are some differences which should be highlighted. Three patients (11.5%) in the central ECMO group developed sternal wound infections (versus 6.3% in the peripheral group) and this difference, although not significant, highlights the risks of leaving a sternum open and going back in through this operative field. It is also a much higher incidence than we would accept in elective cardiac surgical cases. The other complication worth highlighting is that of lower limb complications. Three patients in the peripheral cannulation group (18.8%) developed lower limb complications directly related to the femoral vessel cannulation. This 1545 represents a high incidence and a very significant morbidity in these patients. Although two patients in the central cannulation group also developed lower limb complications, both of these cases were directly related to IABP insertion which could have been avoided. This difference highlights the care which must be taken with femoral vessel cannulation. On the basis of this review, we now consider patients with difficult femoral access or very small femoral vessels as candidates for central cannulation in the first instance. Other short-term VADs are now available which have the advantages of left ventricular decompression, lower anticoagulation requirements, and the opportunity to proceed with extubation over ECMO. However, from a purely cost benefit perspective, we are unable to justify the increased cost required to purchase these devices in our government funded hospital system when our outcomes with ECMO seem to be as good as, if not better than, those reported with these short-term VADs. Further, ECMO has the advantage over VADs of providing both right and left ventricular support. Insertion of ECMO is simple and we have trained our intensive care nurses and consultants to look after the ECMO circuitry, reducing the manpower required in the postoperative setting. Right ventricular failure is often the earliest and most common manifestation of PGF. Taghavi and colleagues [17] have previously reported superior results with ECMO for these patients rather than a dedicated RVAD circuit. This has also been our experience. Other groups have reported their results using devices other than ECMO for posttransplant PGF. The Levitronix Centrimag (Levitronix, LLC, Waltham, MA) centrifugal pump has been used in a biventricular configuration which requires one pump to be connected to the LV and another pump to be connected to the RV [18]. The authors note that the embolism risk with this device is low allowing them to commence heparin after 24 hours and that no other anticoagulant or antiplatelet drug is required. In that case report of two patients, both patients survived to hospital discharge. Another study of six patients with PGF postheart transplant reported a mortality rate of 50% [19]. However, that group noted a high rate of thromboembolic events in their nonsurvivors at autopsy contradicting the previously stated low embolism risk. Other reported benefits are the ability to deliver 9.9 L/minute of flow and low hemolysis rates [19]. Despite these purported benefits, the use of such a device, particularly in a biventricular assist configuration, is more expensive than ECMO with lower survival rates than that seen in our cohort. The Abiomed BVS5000 assist device (Abiomed, Inc), has also been used in early graft failure in a small series with good results [20]. However, again two devices are required to be inserted to achieve biventricular support and the devices have to be inserted centrally. The complication profile reported was similar to that seen in our series without the lower limb vascular issues. However, we have used the Abiomed BVS5000 in our institution and although we found it easy to insert, the ADULT CARDIAC

ADULT CARDIAC 1546 MARASCO ET AL Ann Thorac Surg PRIMARY GRAFT FAILURE AND HEART TRANSPLANT 2010;90:1541 47 bedside management was cumbersome with potential for management error. More recently, descriptions of the Impella suite of devices have emerged. The partially implantable Impella Recover RD (Impella Cardiosystems GMbH, Aachen, Germany) device, specifically designed for RV support has been used in four heart transplant patients. Although RV support was achieved, none of the patients survived [21]. The Impella LP 5.0 (Abiomed, Inc), has been successfully used for LV support in acute rejection in a case report utilizing a femoral insertion after cut down onto the artery [22]. These devices may well have a place in post-heart transplant patients in the future although at the moment experience with them for this indication is sparse. The Impella LP 5.0 is also the only device outlined here which can be placed through peripheral access. However, all of these devices are limited in that they only support one ventricle. This may be part of the reason why the results when using them for PGF are suboptimal. Placing two devices for biventricular support seems an unnecessarily cumbersome and expensive method of managing this clinical problem. In conclusion, we have been very satisfied with the use of ECMO as the main temporary ventricular support in our institution. Using the one type of device with a variety of cannulae, we can place a patient on either central or peripheral biventricular support or configure the ECMO as a venovenous circuit for our respiratory failure and lung transplant patients. This enormous versatility, using one type of support, means we have achieved experience and expertise with ECMO quickly, with approximately 200 patients supported on ECMO since 1990. This review demonstrates that ECMO support for PGF post-heart transplantation provides adequate support with either central or peripheral arteriovenous cannulation with excellent weaning and survival to discharge rates. References 1. Taylor DO, Edwards LB, Boucek MM, Trulock EP, Keck BM, Hertz MI. The registry of the International Society for Heart and Lung Transplantation: twenty-first official adult heart transplant report 2004. J Heart Lung Transplant 2004;23: 796 803. 2. Jahania MS, Mullett TW, Sanchez JA, Narayan P, Lasley RD, Mentzer RM Jr. Acute allograft failure in thoracic organ transplantation. J Card Surg 2000;15:122 8. 3. Segovia J, Pulpón LA, Sanmartín M, et al. Primary graft failure in heart transplantation: a multivariate analysis. Transplant Proc 1998;30:1932. 4. Marasco SF, Esmore DE, Negri J, et al. Early institution of mechanical support improves outcomes in primary cardiac allograft failure. J Heart Lung Transplant 2005;24:2037 42. 5. Newcomb A, Esmore D, Rosenfeldt FL, Richardson M, Marasco S. Heterotopic heart transplantation: an expanding role in the 21st century? Ann Thorac Surg 2004;78:1345 51. 6. Marasco SF, Lukas G, McDonald M, McMillan J, Ihle B. Review of ECMO (extra corporeal membrane oxygenation) support in critically ill adult patients. Heart Lung Circ 2008;17(Suppl 4):S41 7. 7. D Alessandro C, Aubert S, Golmard JL, et al. Extra-corporeal membrane oxygenation support for early graft failure after cardiac transplantation. Eur J Cardiothorac Surg 2010;37: 343 9. 8. Arpesella G, Loforte A, Mikus E, Mikus PM. Extracorporeal oxygenation for primary allograft failure. Transplant Proc 2008;40:3596 7. 9. Reiss N, Leprince P, Bonnet N, et al. Results after orthotopic heart transplantation accepting donor hearts 50 years: experience at La Pitie Salpetriere, Paris. Transplant Proc 2007;39:549 53. 10. Chou NK, Chi NH, Ko WJ, et al. Extracorporeal membrane oxygenation for perioperative cardiac allograft failure. ASAIO J 2006;52:100 3. 11. Leprince P, Aubert S, Bonnet N, et al. Peripheral extracorporeal membrane oxygenation (ECMO) in patients with posttransplant cardiac graft failure. Transplant Proc 2005;37: 2879 80. 12. Taghavi S, Ankersmit HJ, Wieselthaler G, et al. Extracorporeal membrane oxygenation for graft failure after heart transplantation: recent Vienna experience. J Thorac Cardiovasc Surg 2001;122:819 20. 13. Fiser SM, Tribble CG, Kaza AK, et al. When to discontinue extracorporeal membrane oxygenation for postcardiolotomy support. Ann Thorac Surg 2001;71:210 4. 14. Ko WJ, Chen YS, Chou NK, Hsu RB, Wang SS, Chu SH. Extracorporeal membrane oxygenation rescue after heart transplantation. Transplant Proc 2000;32:2388 91. 15. Excell L, Russ G, Wride P. ANZOD Registry report. Australia: Australian and New Zealand Organ Donation Registry Adelaide, 2005. 16. Marasco SF, Esmore DS, Richardson M, et al. Prolonged cardiac allograft ischemic time no impact on long-term survival but at what cost? Clin Transplant 2007;21:321 9. 17. Taghavi S, Zuckermann A, Ankersmit J, et al. Extracorporeal membrane oxygenation is superior to right ventricular assist device for acute right ventricular failure after heart transplantation. Ann Thorac Surg 2004;78:1644 9. 18. Santise G, Petrou M, Pepper JR, Deyfus G, Khaghani A, Birks EJ. Levitronix as a short-term salvage treatment for primary graft failure after heart transplantation. J Heart Lung Transplant 2006;25:495 8. 19. Shuhaiber JH, Jenkins D, Berman M, et al. The Papworth experience with the Levitronix CentriMag ventricular assist device. J Heart Lung Transplant 2008;27:158 64. 20. Petrofski JA, Patel VS, Russell SD, Milano CA. BVS5000 support after cardiac transplantation. J Thorac Cardiovasc Surg 2003;126:442 7. 21. Sugiki H, Nakashima K, Vermes E, Loisance D, Kirsch M. Temporary right ventrciular support with Impella Recover RD axial flow pump. Asian Cardiovasc Thorac Ann 2009;17: 395 400. 22. Samoukovic G, Al-Atassi T, Rosu C, Giannetti N, Cecere R. Successful treatment of heart failure due to acute transplant rejection with the Impella LP 5.0. Ann Thorac Surg 2009;88:271 3. INVITED COMMENTARY Waiting expectantly for a transplanted heart to contract, and observing an echocardiogram that shows cardiac function that is worse than the explanted heart is a humbling and sometimes terrifying experience. Managing primary graft dysfunction (PGD) is best accomplished by an organized, standardized, and comprehensive approach that maintains organ function and limb viability while awaiting the cardiac recover that occurs in 2010 by The Society of Thoracic Surgeons 0003-4975/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2010.06.096