Liver transplantation is a highly successful, life-saving modality for treating end-stage liver disease. This success, however, has resulted in an increasing demand for liver transplantation and with it a significant disparity between the number of patients awaiting liver transplantation and the number of organs available. According to the Scientific Registry of Transplant Recipients database, there were 16,749 patients awaiting liver transplantation as of February 2008. In the previous year (2007), there were 5696 donation after brain death (DBD) transplants performed with an additional 244 living donor liver transplantation (LDLT) procedures.1 This discrepancy between supply and demand has resulted in significant morbidity and mortality for patients awaiting liver transplantation and has in turn resulted in the search for alternatives in order to offer liver transplantation to more patients.
Attempts to increase the number of liver transplants using split liver transplantation and LDLT have met with limited success for adults and currently account for only a small percentage of the total number of liver transplants performed annually. LDLT, for example, peaked at 10% of the total number of liver transplants in 2001 and currently has declined to 5% of liver transplants, except in areas in which the average Model for End-Stage Liver Disease (MELD) score at transplant is 25 or greater.2 Split liver transplantation has reduced the waiting list mortality for pediatric patients to less than 5%3, 4 but has done little to affect the waiting list mortality for adults and thus cannot be viewed at the present time as a way to significantly increase the organ pool for adults awaiting liver transplantation. Similarly, although LDLT has been highly effective in children,4 its application to the adult population has raised questions regarding donor morbidity and mortality.5 Although LDLT appears to have a role in countries without a brain death law or in which cultural or religious restrictions prohibit the use of DBD donors, its use in the United Sates appears to be decreasing and is unlikely to affect the organ shortage for adults.
Perhaps the most significant source of organs that would currently expand the donor pool involves the use of extended criteria donors (ECDs). These donors, although not formally characterized, represent a wide spectrum of donors that may have some unfavorable characteristics which have been historically associated with poorer graft and patient survival. Characteristics of this heterogeneous group of donors include advanced age, significant macrovesicular steatosis, hypernatremia, and donation after cardiac death (DCD). Recent attempts to quantify the risk of graft failure when ECD grafts are used have included the donor risk index (DRI).6 DCD donors represent a specific type of ECD for whom death is declared on the basis of cardiopulmonary criteria rather than cessation of whole brain function. DCD subjects the liver to warm ischemia, which may result in intrahepatic biliary strictures (IHBSs), hepatic abscesses, hepatic artery thrombosis (HAT), or primary nonfunction (PNF).7–9 These organs are perceived as being at higher risk for graft failure and have not been widely accepted by the transplant community for liver transplantation.
Since the inception of the program in 1998, Mayo Clinic in Jacksonville has partnered with its affiliated organ procurement organization (OPO), Life Quest, to adopt protocols for the successful utilization of organs from ECDs and DCD donors. To date, we have performed over 150 liver transplants with DCD organs, with a significant number of these coming from donors greater than 60 years of age. These numbers reflect the efforts of the local OPO's DCD program, which has shown a progressive increase in both the absolute number and percentage of DCD donors procured each year.
This retrospective analysis, which represents one of the largest single-center experiences, was undertaken to compare the outcomes for DCD liver transplants to those for DBD liver transplants performed over the same time period. Additional insight is provided into the use of DCD organs from older donors who theoretically have an increased risk of graft failure as measured by the DRI. Given the limitations of the data pertaining to the outcomes of DCD liver transplantation from the transplant databases and the variable results from small single-center experiences regarding the use of DCD organs for liver transplantation, we present a large single-center experience and examine the potential of these grafts to expand the donor pool in a safe and efficient manner.
CIT, cold ischemia time; DBD, donation after brain death; DCD, donation after cardiac death; DRI, donor risk index; ECD, extended criteria donor; HAT, hepatic artery thrombosis; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; IHBS, intrahepatic biliary stricture; LDLT, living donor liver transplantation; MELD, Model for End-Stage Liver Disease; N/A, not applicable; NS, not significant; OPO, organ procurement organization; PNF, primary nonfunction; WIT, warm ischemia time.
PATIENTS AND METHODS
This study represents a single-center, retrospective review of 1436 liver transplants performed between December 1998 and October 2006. Approval for the study was obtained from the Mayo Clinic institutional review board. The study was performed by chart review of all liver transplants using DCD or DBD organs. Recipient information included the age, sex, cause of liver failure, cause of graft loss, presence of hepatocellular carcinoma (HCC), hepatitis C virus (HCV), cold ischemia time (CIT), raw MELD score at the time of transplant, and median follow-up time. Outcomes included the patient and graft survival rates, retransplantation rate, and incidence of PNF, IHBSs, and HAT.
All DCD donors were classified as Maastricht type 310 (controlled awaiting cardiac death). Detailed information regarding the DCD donors was obtained from the procurement database of Mayo Clinic in Jacksonville and the OPO records. Donor information that was gathered included the geographic location, age, sex, cause of death, warm ischemia time (WIT), DRI, and liver function tests. WIT was defined as the time from the withdrawal of both ventilator and cardiac support to the cold perfusion of the organ. CIT was defined as the time from the infusion of the cold preservation solution until portal reperfusion of the liver in the recipient.
The recovery of organs from a DCD donor was viewed as a process separate from the consent to withdrawal of support. Withdrawal of support, institution of comfort measures, and declaration of death were in strict compliance with donor hospital policies, and at no time was the transplant team involved in the withdrawal process or in making recommendations of how it should occur. After consent was obtained, the patient was either taken to an anesthesia holding area or brought to the operating room with full cardiopulmonary support in place. An independent physician from the donor hospital, separate from the OPO and the transplant center, was assigned to withdraw artificial life support and provide end-of-life care to the patient. The blood pressure, oxygen saturation, and respiratory rate were recorded at 1-minute intervals. Following the declaration of death by the independent physician, another 5 minutes of mandatory observation was performed as described in the 1997 Institute of Medicine guidelines.11, 12 During the 5-minute waiting period, the patient was transported to the operating room (if not already there) and prepared for organ recovery. Heparin was administered to the patient according to the donor hospital policy. Following the 5-minute wait period, a rapid retrieval technique was performed in which the abdomen was opened with a cruciate incision. The small bowel was reflected superiorly, and the aorta and portal systems were cannulated. Cold preservation fluid, consisting of University of Wisconsin solution, heparin, and glutathione, was then flushed through the abdominal aorta, and the abdomen was packed with ice. After this, the intrathoracic descending aorta was cross-clamped by the opening of either the chest or left hemidiaphragm. Finally, the suprahepatic inferior vena cava was opened to allow venting. The liver was then removed, and the biliary system was flushed on the back-table. Finally, the liver was packaged in ice and transported back to the hospital for implantation. Retrieval of organs from DBD donors was conducted according to a standard technique that has been described elsewhere.
It is standard policy at Mayo Clinic in Jacksonville that recipients be taken into the operating room immediately upon arrival of the donor liver in order to minimize CIT.
All transplants were performed with the piggyback procedure without a portocaval shunt or caval clamping. The standard immunosuppression protocol included tacrolimus, mycophenolate mofetil, and prednisone. Rejection episodes were treated with steroid boluses and elevation of tacrolimus levels. Steroid-resistant rejections were treated with thymoglobulin.
DCD Donor Selection Criteria
No formal criteria were used for accepting or refusing a potential DCD donor when an offer was made. In general, all offered DCD donors were considered to be potential candidates for liver donation and evaluated on an individual basis and as a general rule were considered to be worth evaluating if they would have met criteria for DBD donation. In general, the suitability of a given donor was based on several factors, which included the donor age, body mass index, serum sodium, hepatic function, use of vasopressors, degree of sustained hypoxia and hypotension following withdrawal to cold perfusion, overall WIT, time to a systolic blood pressure less than 50 mm Hg, projected CIT, and appearance of the liver after cold perfusion. Absolute criteria for exclusion were few and included WIT greater than 1 hour, age > 80 years, and a time to a systolic blood pressure less than 50 mm Hg > 30 minutes. These considerations were also applied to DCD donors older than 60 years; again, no formal criteria for acceptance were used.
Recipient Selection Criteria
Although no formal selection criteria were used to determine which recipients would undergo transplantation with DCD organs, some inherent selection bias may have occurred by the avoidance of certain types of recipients. These included those recipients whose clinical history may have resulted in a prolonged CIT or in excessive metabolic demands to be placed on the DCD organ, such as recipients with multiple abdominal surgeries and those with multisytem organ failure.
Statistical analysis was performed with SPSS 14.0 (SPSS, Inc., Chicago, IL). Patient survival and graft survival were determined with the Kaplan-Meier method and life tables, and significance of survival differences was determined with the log-rank test.
Between December 1998 and October 2006, 1436 liver transplants were performed by the transplant program at Mayo Clinic in Jacksonville. Of these, 108 (7.5%) involved DCD donors, and 1328 (92.5%) involved DBD donors. The percentage of DCD liver transplants performed each year increased steadily over the study period, culminating in 15.3% of all liver transplants in 2006 (Fig. 1). No LDLT was performed during the study period.
The recipients of the 108 DCD donors included 9 patients who underwent combined liver/kidney transplantation (both organs came from the same DCD donor) and 7 patients who underwent retransplantation and for whom the organ for the primary transplant was procured from a DBD donor. Fifty-five percent of the DCD donors were procured locally, 37% were procured regionally, and 7.4% were procured nationally.
The DBD donors were significantly older than the DCD donors (48 ± 20 for DBD versus 41 ± 17 for DCD, P = 0.004). A greater percentage of DBD donors were older than 60 years of age (28% versus 17.6%, P = 0.018). The mean DRI was significantly higher in the DCD group (2.02 ± 0.57 for DCD versus 1.67 ± 0.45 for DBD, P < 0.001). CIT was significantly lower in the DCD group (6.3 ± 1.7 versus 7.15 ± 2.1 hours, P = <0.005). The mean WIT for DCD donors was 22.3 minutes (range, 5-59 minutes).
Table 1. Donor and Recipient Characteristics by Donor Type (DCD Versus DBD)
DCD (n = 108)
DBD (n = 1328)
Abbreviations: CIT, cold ischemia time; DBD, donation after brain death; DCD, donation after cardiac death; DRI, donor risk index; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; MELD, Model for End-Stage Liver Disease; N/A, not applicable; WIT, warm ischemia time.
There was no significant difference with respect to recipient age or MELD score. There was, however, a statistically higher incidence of recipients with HCC and HCV in the DCD group (HCC, 26% versus 16.3%, P = 0.016; HCV, 48% versus 37.5%, P = 0.031).
The median follow-up for all recipients was 48 months (34.5 months for DCD and 50 months for DBD). The 1-, 3-, and 5-year patient survival and graft survival for DCD donors were 91.5%, 88.1%, and 88.1% and 79.3%, 74.5%, and 71.0, respectively. The 1-, 3-, and 5-year patient survival and graft survival for DBD donors were 87.3%, 81.1%, and 77.2% and 81.6%, 74.7%, and 69.1%, respectively
Patient Death, Graft Failure, and Retransplantation (Table 2)
Thirty-three (30.5%) DCD grafts and 362 (27.3%) DBD grafts were lost over the study period (P = not significant). For DCD grafts, causes of graft loss included IHBSs (9, 8.3%), PNF (4, 3.7%), recurrent HCV (4, 3.7%), HAT (1, 0.9%), rejection (2, 1.9%), and patient death/other (13, 12%). Only IHBS rates were significantly different between the DCD and DBD groups as a cause of graft loss [9 (8.3%) for DCD versus 25 (1.9%) for DBD, P = 0.008]. Sixteen (14.8%) recipients of a DCD graft underwent retransplantation versus 123 (9.3%) recipients of a DBD graft (P = 0.107). Graft and patient survival was 100% for patients undergoing DCD retransplantation, all of whom had received DBD grafts.
Table 2. Causes of Graft Failure and Patient Death
Graft Failure/Patient Death
Abbreviations: DBD, donation after brain death; DCD, donation after cardiac death; HAT, hepatic artery thrombosis; HCV, hepatitis C virus; NS, not significant; PNF, primary nonfunction.
In order to evaluate the impact of donor age on outcome, both the DCD and DBD donors were divided into those less than 60 years of age and those greater than 60 years of age. For donors older than 60 years of age, DBD donors had a higher mean age than DCD donors (70.9 ± 7.1 for DBD versus 65.5 ± 5.3 for DCD, P < 0.001). DCD donors had a significantly higher DRI than DBD donors for all age groups. Similarly, DCD donors > 60 years of age had a significantly higher DRI than DCD donors < 60 years of age (2.67 ± 0.44 for DCD > 60 versus 1.87 ± 0.49 for DCD < 60, P < 0.001).
Table 3. Donor and Recipient Characteristics by Donor Type (DCD Versus DBD) for Donors < 60 Years Old and Donors > 60 Years Old
DCD: Donor Age < 60
DBD: Donor Age < 60
DCD: Donor Age > 60
DBD: Donor Age > 60
Abbreviations: CIT, cold ischemia time; DBD, donation after brain death; DCD, donation after cardiac death; DRI, donor risk index; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; MELD, Model for End-Stage Liver Disease; N/A, not applicable; WIT, warm ischemia time
WIT for DCD < 60 versus DCD > 60 was nonsignificant (P = 0.273).
CIT did not differ significantly between DCD donors > 60 years of age and those < 60 years of age; however, CIT was significantly higher for DBD donors < 60 years of age in comparison with DCD donors < 60 years of age (7.24 ± 2.11 for DBD versus 6.26 ± 1.64 for DCD, P < 0.001). WIT was not significantly different between DCD donors less than 60 years of age and those greater than 60 years of age (22.7 ± 13.1 for DCD > 60 versus 24.5 ± 11.5 for DCD < 60, P = 0.540).
Recipient MELD scores were equivalent among all donor groups, with the exception of the DCD > 60 years of age group, for which the MELD score of the recipient was significantly less than those of the other groups.
Recipients from donors > 60 years of age, regardless of type, were significantly less likely to have HCV as a cause for liver failure. This reflects a change in our center's allocation policy in 2003, when our own data and those of others showed poorer outcomes in HCV recipients receiving older grafts.13
DCD donors > 60 years of age had 1- and 3-year patient and graft survival of 89.5% and 89.5% and 74.9% and 72.4%, respectively. For DCD donors < 60 years old, the 1- and 3-year patient survival and graft survival were 91.9% and 87.6% and 79.4% and 72.4%, respectively (P = not significant).
Similarly, when DBD donors were divided into donors > years old and donors < than 60 years old, there was no statistical difference in either patient or graft survival in comparison with DCD donors < 60 years of age or DCD donors > 60 years of age.
The use of DCD organs has the potential to increase the availability of organs for liver transplantation. In doing so, it may reduce the waiting list mortality and allow the transplant community to rely less on alternatives such as LDLT, which may subject donors to significant morbidity and mortality.5 Currently, the use of DCD donors for organ transplantation is heavily weighted toward the use of kidneys, for which there can be no doubt that the outcomes for DCD donors are equivalent to those for DBD donors.14, 15 However, the last decade has witnessed a significant increase in both the absolute number of liver transplants performed with DCD liver allografts and the number of institutions performing DCD liver transplants. In 1995, there were 18 DCD liver transplants performed nationally, which accounted for 0.4% of all liver transplants. This number rose to 388 DCD liver transplants for 2005, accounting for 5.8% of all liver transplants.1 Similarly, the number of transplant programs performing at least 1 DCD liver transplant increased from 7 in 1996 to 33 in 2005.2 These data suggest an interest in the use of DCD organs for liver transplantation.
The true potential for DCD liver transplantation is difficult to evaluate. A recent analysis of our local OPO examined the death records of patients from 10 local hospitals and determined that the number of potentially viable liver allografts retrieved could be increased by 53% if all potential DCD donors who were identified as possible donors were used (unpublished data).
The single-center experience with DCD liver transplants reported herein is the largest reported to date. The median follow-up in this study was 48 months, which was sufficient to assess the long-term outcomes of these grafts and make meaningful conclusions. Currently, DCD liver transplantation accounts for over 15% of all liver transplants performed yearly at our institution. Fifty-five percent of these were procured locally with a mean CIT of only 6.3 ± 1.7 hours, which was significantly less than the CIT for DBD transplants (mean CIT, 7.15 ± 2.1 hours). Both the low CIT and the mean WIT of only 24 minutes may have contributed to the acceptable graft outcomes.
Recipients of DCD liver transplants were not selected on the basis of specific criteria, and there were no exclusions from the analysis. With the exception of recipients from older DCD donors, who had lower MELD scores than the other groups, the clinical and demographic features for both DCD and DBD recipients demonstrated no significant difference in the MELD scores or age of the recipient; this is contrary to the belief that DCD organs are placed into healthier patients. On the contrary, the DCD recipients included 9 recipients of combined liver/kidney transplants, 7 recipients of retransplants, and a significantly greater number of HCC and HCV patients. When donor demographics were examined, DCD donors were found to be significantly younger than DBD donors, even though 17.6% of DCD donors were older than 60 years of age; this emphasized the aggressive approach to using ECDs even within the DBD group. In addition, DRI was significantly higher in the DCD donors, reflecting the negative weighting of DCD in the DRI equation.
The equivalent patient survival and graft survival across donation type are promising and suggest that better outcomes are possible in comparison with recent transplant registry analysis of DCD outcomes. One such analysis of United Network for Organ Sharing data from 1998 to 2004 suggested that the use of DCD organs from donors less than 60 years of age was associated with poorer outcomes in comparison with DBD organs from donors less than 60 years old. However, the DCD donors less than 60 years of age had patient survival and graft survival equivalent to those of DBD donors greater than 60 years of age.15 Similarly, data from the Scientific Registry of Transplant Recipients database demonstrated inferior 1- and 3-year graft survival of 70.2% and 63.3% for DCD donors when compared to DBD donors.16 At first glance, the registry data appear to indicate poorer outcomes from DCD liver transplants. However, when the data are controlled for factors such as recipient selection, center experience, Maastricht type of DCD donor, CIT, and WIT, the results are more encouraging. Mateo et al.17 examined the United Network for Organ Sharing database from 1996 to 2003 and reported DCD graft survival of 71% at 1 year and 60% at 3 years; this was significantly inferior to DBD donors (80% at 1 year and 72% at 3 years, P < 0.001). However, when low-risk recipients were combined with low-risk DCD organs (WIT < 30 minutes and CIT < 10 hours, n = 226), graft survival rates increased to 81% and 67% at 1 and 3 years, respectively, and these rates were not significantly different from those of recipients with DBD grafts (80% and 72% at 1 and 3 years, respectively).
Single-center data also have limitations because many of these reports are based on small numbers of patient spread over long time periods. Most series reported to date include fewer than 30 patients with 1-year graft survival rates ranging from 67% to 84%.8, 18–22 These wide variations in results may reflect different selection criteria for donors, recipients, CIT, and WIT that are acceptable for each institution. However, they clearly demonstrate that in certain circumstances, the results from DCD organs can match those from DBD transplants. This success is even more significant when the results of older DCD donors are considered. Our results suggest that if CIT, WIT, and recipient selection are optimal, excellent graft survival can be obtained from this group of DCD donors. Our results agree with previously published data reported by Fukumori et al.20
Over the 10-year history of our program, 33 (30.5%) of 108 DCD grafts were ultimately lost versus 362 (27.3%) of DBD grafts (P = not significant), and this indicated no significant difference in the overall graft loss rate. Specifically, there was not a significant increase in the incidence of PNF or HAT, both of which have been previously reported in the literature as significant causes of graft failure.23 Our results confirm the findings of others and suggest that the rates of PNF and HAT are not significantly higher in DCD grafts compared to DBD grafts.8, 19–21 This may reflect a general trend toward using DCD donors more selectively and ensuring that both the CIT and WIT are kept to a minimum.
Currently, although there still exist some ethical, financial, and logistical barriers to the widespread use of DCD liver transplantation, the development of biliary complications is perhaps of the most concern. Although there are many types of biliary complications, the development of IHBSs is most damaging and is the true Achilles heel of DCD liver transplantation.7, 9 Our series reported a significantly higher rate of biliary necrosis (8.3% versus 1.9%) resulting in graft failure and a higher, though statistically not significant, rate of retransplantation in the DCD group (14.8% versus 9.3%).
IHBSs are difficult to treat and often result in retransplantation. Some institutions have opted to treat these complications with a variety of interventional radiology techniques and thus delay the inevitable retransplant. We have adopted an aggressive approach to IHBSs, opting for early retransplantation for all patients who have become symptomatic or develop deteriorating allograft function. This low threshold for retransplantation is reflected in our higher retransplant rate among recipients of DCD organs. Efforts to prevent IHBSs should be approached at the clinical and basic science levels. These interventions may include the use of alternative low-viscosity preservation fluids24, 25 and the use of thrombolytic agents, which may aid in flushing the microscopic biliary vasculature and thus possibly prevent microvascular thrombosis in the biliary tree.26
In order for DCD organs to be used more widely by the transplant community, various logistical, financial, and ethical issues need to be addressed by the governing bodies of transplantation. Clearly, under certain circumstances and with increasing experience, acceptable patient and graft survival can be obtained with these organs even from older donors who were once felt to be poor candidates for DCD liver transplantation.
Collaboration between centers with a specific interest in DCD liver transplantation needs to be encouraged. This would promote protocol development, standardize the procurement process, and develop acceptable definitions and guidelines for WIT. In addition, allocation policies regarding recipient selection need to be addressed to enable optimal use of these organs.