Secondary Logo

Journal Logo

Gas Exchange

Extracorporeal Membrane Oxygenation for Perioperative Cardiac Allograft Failure

Chou, Nai-Kuan; Chi, Nai-Hsin; Ko, Wen-Je; Yu, Hsi-Yu; Huang, Shu-Chien; Wang, Shoei-Shen; Lin, Fang-Yue; Chu, Shu-Hsun; Chen, Yih-Sharng

Author Information
doi: 10.1097/01.mat.0000196514.69525.d9
  • Free


Heart transplantation is the standard modality with documented long-term results for patients with end-stage heart failure.1 Because a donor shortage results in patients dying while on the waiting list, more marginal donors, with relative contraindications that might have been rejected previously,2 are being harvested. This may result in an increased incidence of perioperative cardiac allograft failure (PCAGF).3

Perioperative cardiac allograft failure in transplant recipients is a life-threatening condition if not managed properly and promptly. Retransplantation is not recommended based on the data of a 10-year multiinstitutional study,4 and mechanical support sometimes is indicated for this critical situation.3,5,6 Some studies have reported unsatisfactory results with the use of ventricular assist devices (VADs) for PCAGF.3,5,7 Few centers have tried to apply extracorporeal membrane oxygenation (ECMO) in PCAGF.6 ECMO is the first choice for mechanical support beyond intraaortic balloon pumps (IABPs) in our institution, and it has been demonstrated that ECMO might be considered first to provide mechanical support in acute myocarditis.8 The present study describes and analyzes the result of applying ECMO to rescue PCAGF refractory to medical therapeutics and IABP in the early posttransplant period in our group.

Materials and Methods

The Study Was Approved By The Institutional Investigational Review Board.


Retrospective analysis of heart transplantation from 1987 (the first heart transplantation in our institute) to December 2004 was conducted to identify the patients receiving postoperative ECMO support. According to a definition of PCAGF given elsewhere,3 the diagnosis was made by exclusion of possible anastomotic kinking, treatable pulmonary hypertension, and hypoxemia. Pulmonary hypertension was first treated with vasodilators9–11; inotropic support, including type III phosphodiesterase inhibitors,12 and inhaled nitric oxide have been used since 1996. Intraaortic balloon pumps were considered to be the first choice of support, but are sometimes not suitable because of their limited availability for pediatric patients and inadequate support. Furthermore, the left ventricle of a donor heart is assumed to have a normal coronary artery in transplant status. The present study describes and analyzes the result of ECMO application to rescue PCAGF refractory to medical therapy and IABP in the early posttransplant period.

Immunosuppression Agent

A “triple immunosuppression” regimen consisting of cyclosporine, methylprednisolone, and azathioprine was used in the immediate posttransplant period. Rabbit antithymocyte globulin immunoinduction therapy was administered during the first 5 days for induction. For clinical evaluation of cardiac graft rejection, transvenous endomyocardial biopsy was performed three to four times in the first month and every 3 months in the first year, and then the dosage of immunosuppressant was adjusted according to the results of endomyocardial biopsy.

Management of ECMO

The whole ECMO apparatus, including the centrifugal pump and oxygenator (Medtronic Inc., Anaheim, CA), was primed with normal saline alone. Cannulation was performed with a modified open Seldinger method. The femoral vessels were dissected out and the cannulas were inserted. The pressure in the superficial femoral artery was measured after cannulation. If the mean pressure was below 50 mm Hg, a perfusion catheter (8.5 Fr, Super Arrow-Flex central venous catheter) was inserted distally.

After the ECMO was hooked up to the patients, hemodilution was corrected by packed red blood cell transfusion. The hematocrit and blood oxygen saturation in the preoxygenator and postoxygenator circuits were continuously monitored. Hematocrit was kept between 30% and 35%. Lower hematocrit compromised oxygen delivery, but higher hematocrit increased complications of clot formation and hemolysis by the centrifugal pump. Continuous monitoring of postoxygenator blood oxygenation saturation provided an indicator for the gas-exchange function of the oxygenator. Systemic heparinization was not needed on the first day of ECMO support, when the risk of bleeding was the highest immediately after the heart transplantation operation. Afterwards, heparin infusion was not used until the bleeding was controlled, which usually took 1 to 2 days. When bleeding decreased, heparin drip was started to keep the activated clotting time between 160 and 180 seconds. The ECMO apparatus was changed when oxygenator dysfunction, clot formation, or hemolysis was found. Symptoms and signs of low cardiac output were usually resolved after the ECMO support was initiated, and catecholamine infusion could be tapered accordingly. Arterial pulse pressure wave contour, serial echocardiography, and blood oxygenation saturation in the preoxygenator circuit were used to monitor the recovery of cardiac allograft. If hemodynamics could be well maintained by reduced ECMO blood flow at 0.5 l/min for 1 hour, ECMO was removed at bedside. The wound was primarily repaired.

Data Collection and Analysis

Donor-, surgery-, and ECMO-related variables were evaluated for association with operative mortality, success of weaning, and survival rate. Continuous variables were expressed as means and standard deviations, and means were compared by independent-sample Student’s t test. Nominal variables were expressed as percentages and analyzed by the χ2 test. Statistical significance was assumed at a p value of less than 0.05.


During this period, 219 transplants were performed, of which 19 patients received ECMO for PCAGF (Table 1). The mean age of transplantation in transplants excluding PCAGF was 37.6 ± 22.5 years, and the mean age of the PCAGF group (n = 19), which included 4 pediatric patients (<18 years), was 44.2 ± 17.3 years. The etiology of indication for transplantation included dilated cardiomyopathy (DCM, n = 6), ischemic cardiomyopathy (ICM, n = 9, 1 was a case of pediatric Kawasaki disease), valvular cardiomyopathy (VCM, n = 2, 1 was a case of corrected transposition of great arteries), restrictive cardiomyopathy (n = 1), and posttransplant coronary artery disease (n = 1). Mean donor age was 29.3 ± 14.3 years for the PCAGF group, which was not very different from that of patients without PCAGF (30.5 ± 12.4 years). The ischemic time of 206.8 ± 96.1 minutes was significantly longer in PCAGF compared with the 150.3 ± 57.5 minutes in the group without PCAGF (p = 0.021) (Table 2).

Table 1
Table 1:
Perioperative Graft Failure Data
Table 2
Table 2:
Comparison of Pretransplant and Posttransplant (1 week) Data

The possible etiology of PCAGF included primary graft failure,7 right heart failure secondary to pulmonary hypertension,10 and sepsis.2 All patients were rescued with ECMO. Ten patients were set up with ECMO in the operating room because of failure of weaning from cardiopulmonary bypass. Six patients were started with ECMO within 12 hours of leaving the operating room, and 3 patients were set up 3 to 7 days after transplantation (Table 1).

The mean duration of support was 157 ± 129 hours. Sixteen patients were able to be weaned from ECMO (84.2%), in which one received re-transplantation, but he died from subsequent rejection. The 30-day survival was 57.8% (n = 11). Ten of 16 successfully decannulated patients survived to discharge (62.5%), which was 52.6% of the whole PCAGF group.

The preoperative bilirubin and creatinine levels were not significantly different between the PCAGF and non-PCAGF groups. The bilirubin, but not creatinine, level 1 week after transplantation was significantly elevated when compared with preoperative data (Table 2, p < 0.05). The elevation of bilirubin was more significantly different in the PCAGF group (p < 0.05), indicating that PCAGF patients had more severe liver damage, possibly due to the high incidence of right heart failure.

Follow-up revealed 80% survival at discharge (8 of 10 patients), and 42% survival in the whole PCAGF group. We lost another patient during the fourth year after rescue because of vasculopathy.


To delineate the role of different mechanical supports, we included two recent studies concerning mechanical support for posttransplant graft failure necessitating ECMO or VAD rescue,3,6 and analyzed the weaning rate, survival rate, and graft survival. The meta-analysis (Table 3) revealed that ECMO had a higher weaning rate (81% vs. 27%, p < 0.001), and a higher graft survival rate (53% vs. 18%, p = 0.003). The patient survival rate was also higher in the ECMO group, though not significantly so (p = 0.053). The overall weaning rate in mechanically supported PCAGF was 53.8%, and the survival rate was 41.5%.

Table 3
Table 3:
Meta-analysis of Different Types of Mechanical Support for PCAGF

For PCAGF, different centers had different policies in applying mechanical support.3,5,6 The meta-analysis showed that ECMO had a higher weaning rate (p < 0.001), higher graft survival rate (p = 0.003), and a higher patient survival rate. The overall weaning rate in mechanically supported PCAGF was 53.8%, and the survival rate was 41.5%.Our meta-analysis revealed better weaning and graft survival rates for PCAGF in the ECMO-treated group (Table 3).


Early primary allograft failure in the perioperative period after cardiac transplantation carries high mortality and morbidity.4 High mortality in this category has been observed in some centers,3 and immediate retransplantation is not recommended.4 An early series reported uniformly poor results and strongly advised against the use of mechanical support after early graft failure.13 Because of the advancement and improvement of technology, the utility of mechanical support in pretransplant stabilization14,15 and postcardiotomy shock16,17 is well established, and is gradually being accepted as a proper choice of treatment for this category.3,6

Ventricular assist devices and ECMO are two major mechanical supports for critical patients. VAD is considered to be superior to ECMO as mechanical support for long-term support in pretransplant stabilization, because it affords fewer complications and longer support duration. For postcardiotomy shock, the different results reveal the existing controversy. We have advocated that ECMO may have a potential advantage for acute myocarditis and severe shock necessitating mechanical support.8

The etiologies of PCAGF usually involve poor myocardial protection, acute rejection, and right heart failure. The transplanted donor heart is usually highly selected with preserved function. Poor myocardial protection usually leads to stunning, and short-term mechanical support with fewer complications is adequate for this purpose. ECMO has the advantage of easy application and feasibility. Taghavi et al. found ECMO to be superior to right VADs for acute right ventricular failure after heart transplantation.6 Results of their retrospective study comparing right VADs (n = 15) with ECMO (n = 13) showed an overall in-hospital survival of 42% with only 2 (13%) right VAD patients weaned compared with 10 (77%) ECMO patients weaned. ECMO was used for 10 patients in our group for right heart failure, 7 patients for primary graft failure, and 2 patients for sepsis. Survival rate was 7 patients (70%) for right heart failure, 2 patients (28.5%) for primary graft failure, and 1 patient (50%) for sepsis.

A retrospective study by Kavarana et al.3 compared four devices: a right VAD (n = 11), left VAD (n = 4), biventricular assist device (n = 3), and IABP (n = 2), and reported a 45% wean rate with duration of device support from 2 to 965 hours. Only 28% survived from VAD after heart transplantation for PCAGF. Previous data also demonstrated longer support times in the VAD group.6 According to our data, the duration of support in the majority of survivors in our study was less than 6 days. ECMO support could be minimized with myocardial injury by poor protection. We lost another patient during the fourth year after rescue because of posttransplantation coronary vasculopathy. In our group, the rates of 1 to 5 years of actuarial freedom from the presence of coronary vasculopathy were 97%, 93%, 86%, 80%, and 69%.18

Acute rejection sometimes leads to sudden hemodynamic collapse, and can be reversed by megadose steroid if mechanical support is applied in time to assist the perfusion. ECMO seems to be the most suitable device because of its feasibility and simplicity. Right heart failure is usually caused by poor preoperative status, coagulopathy, and massive bleeding/transfusion, all of which are usually correctable. VADs might not provide adequate oxygenation after massive transfusion, and direct right ventriculotomy of VAD will increase the potential problems of hemostasis. Our heparin-free protocol used in the first 2 days decreases the complications of bleeding, but early oxygenator failure with replacement is expected. Early replacement of all circuits in the first 2 days is necessary for safety.

For PCAGF, our meta-analysis revealed better results in the ECMO-treated group for PCAGF. Therefore, ECMO is highly recommended as first-line choice for this group of patients, though several possible issues remain to be discussed.


Perioperative cardiac allograft failure has a high mortality rate if no suitable mechanical support is applied for rescue. Long ischemic time is the major risk factor for this category. Right heart failure and primary graft failure are the major etiologies of PCAGF. ECMO affords comparable and even higher weaning and survival rates when compared with VADs. Therefore, ECMO instead of VAD is highly recommended for PCAGF rescue.


1. Shumway NE: Thoracic transplantation. World J Surg 24: 811–814, 2000.
2. Sweeney MS, Lammermeier DE, Frazier OH, et al: Extension of donor criteria in cardiac transplantation: Surgical risk versus supply-side economics. Ann Thorac Surg 50: 7–10, 1990.
3. Kavarana MN, Sinha P, Naka Y, et al: Mechanical support for the failing cardiac allograft: A single center experience. J Heart Lung Transplant 22: 542–547, 2003.
4. Radovancevic B, McGiffin DC, KobashigawaJA, et al: Retransplantation in 7,290 primary transplant patients: A 10- year multi-institutional study. J Heart Lung Transplant 22: 862–868, 2003.
5. Minev PA, El-Banayosy A, MinamiK, et al: Differential indication for mechanical circulatory support following heart transplantation. Intensive Care Med 27: 1321–1327, 2001.
6. 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 78: 1644–1649, 2004.
7. Noon GP, Ball JW, Short HD: Bio-Medicus centrifugal ventricular support for postcardiotomy cardiac failure: A review of 129 cases. Ann Thorac Surg 61: 291–295, 1996.
8. Chen YS, Yu HY, Huang SC, et al: Experience and Result of Extracorporeal Membrane Oxygenation in Treating Fulminant Myocarditis with Shock – What mechanical support should be considered first? J Heart & Lung Transpl 24: 81–7, 2005.
9. Bigatello LM, Hurford WE: Inhaled nitric oxide and pulmonary vascular resistance. Crit Care Med 27: 2060–2061, 1999.
10. Bigatello LM: Nitric oxide: modulation of the pulmonary circulation. Minerva Anesthesiol 66: 307–313, 2000.
11. Girard C, Fargnoli JM, Godin-Ribuot D, et al: Inhaled nitric oxide: effects on hemodynamics, myocardial contractility, and regional blood flow in dogs with mechanically induced pulmonary artery hypertension. J Heart Lung Transplant 15: 700–708, 1996.
12. Pamboukian SV, Carere RG, Webb JG, et al: The use of milrinone in pre-transplant assessment of patients with congestive heart failure and pulmonary hypertension. J Heart Lung Transplant 18: 367–371, 1999.
13. Tenderich G, Koerner MM, Stuettgen B, et al: Mechanical circulatory support after orthotopic heart transplantation. Int J Artif Organs 21: 414–416, 1998.
14. Frazier OH, Rose EA, Oz MC, et al: Multicenter clinical evaluation of the HeartMate vented electric left ventricular assist system in patients awaiting heart transplantation. J Heart Lung Transplant 20: 201–202, 2001.
15. Farrar DJ: The thoratec ventricular assist device: a paracorporeal pump for treating acute and chronic heart failure. Sem Thorac Cardiovasc Surg 12: 243–250, 2000.
16. Pennington DG, McBride LR, Miller LW, Swartz MT: Eleven years’ experience with the Pierce–Donachy ventricular assist device. J Heart Lung Transplant 13: 803–810, 1994.
17. Pennock JL, Pierce WS, Wisman CB, et al: Survival and complications following ventricular assist pumping for cardiogenic shock. Ann Surg 198: 469–478, 1983.
18. Wang HJ, Ko WJ, Lee CM, et al: Delayed occurrence of cardiac allograft vasculopathy in Chinese heart transplant recipients compared with their western counterparts. Clin Transplant 18: 21–27, 2004.
Copyright © 2006 by the American Society for Artificial Internal Organs