By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Due to essential maintenance the subscribe/renew pages will be unavailable on Wednesday 26 October between 02:00- 08:00 BST/ 09:00 – 15:00 SGT/ 21:00- 03:00 EDT. Apologies for the inconvenience.
Live donor liver transplantation (LDLT) has become increasingly common in the United States and around the world. In this study, we compared the outcome of 764 patients who received LDLT in the United States and compared the results with a matched population that received deceased donor transplantation (DDLT) using the United Network for Organ Sharing (UNOS) database. For each LDLT recipient (n = 764), two DDLT recipients (n = 1,470), matched for age, gender, race, diagnosis, and year of transplantation, were selected from the UNOS data after excluding multiple organ transplantation or retransplantation, children, and those with incomplete data. Despite our matching, recipients of LDLT had more stable liver disease, as shown by fewer patients with UNOS status 1 or 2A, in an intensive care unit, or on life support. Creatinine and cold ischemia time were also lower in the LDLT group. Primary graft nonfunction, hyperacute rejection rates, and patient survival by Kaplan-Meier analysis were similar in both groups (2-year survival was 79.0% in LDLT vs. 80.7% in case-controls; P = .5), but graft survival was significantly lower in LDLT (2-year graft survival was 64.4% vs. 73.3%; P < .001). Cox regression (after adjusting for confounding variables) analysis showed that LDLT recipients were 60% more likely to lose their graft compared to DDLT recipients (hazard ratio [HR] 1.6; confidence interval 1.1-2.5). Among hepatitis C virus (HCV) patients, LDLT recipients showed lower graft survival when compared to those who received DDLT. In conclusion, short-term patient survival in LDLT is similar to that in the DDLT group, but graft survival is significantly lower in LDLT recipients. LDLT is a reasonable option for patients who are unlikely to receive DDLT in a timely fashion. (Liver Transpl 2004;10:1263–1268.)
There has been a significant increase in the number of patients listed for liver transplantation in the United States over the past 10 years. The increase in organ demand has exceeded the supply, resulting in longer waiting periods and higher death rates on waiting lists.1–2 Approximately 10 to 20% of patients on the liver transplant list die each year without receiving an organ in a timely fashion.1–3 Transplant physicians have responded to this increased demand by developing several strategies, including the use of older donors, grafts from hepatitis C virus (HCV)-positive donors or those with previous hepatitis B virus infection (positive hepatitis B virus core IgG antibody), domino transplantation (in patients with familial amyloid polyneuropathy), split-liver grafts, and more recently, live donor liver transplant (LDLT).1–3 The advantages of LDLT are the shortened waiting period, the ability to optimize the recipient's health status, elective surgery, decrease in cold ischemia time, and an increase in the organ pool. The major disadvantage of LDLT is the potential morbidity and, in rare cases, mortality of the donor.4
The concept of performing a major operation with a potential for mortality on a healthy person without any direct benefits, other than psychological, is somewhat unique.5 Unlike live donor renal transplantation, donation of the right or left lobe of the liver to another adult is associated with significant donor morbidity and a donor mortality of around .5%.2, 4 One of the key recommendations of the American Society of Transplant Surgeons' position paper on adult-to-adult LDLT was that the risk to the donor should be weighed against the realistic estimate of a successful outcome in the recipient.6 To determine the role of LDLT, and to advise patients and potential donors, it is therefore important to determine the realistic expectation in terms of graft and patient survival in an unselected patient population. In this study, our aim was to compare the outcome of all patients who underwent LDLT in the United States and compare the outcome to a comparable group of patients who underwent deceased donor liver transplantation (DDLT).
LDLT, live donor liver transplantation; DDLT, deceased donor transplantation; UNOS, United Network for Organ Sharing; HCV, hepatitis C virus; HR, hazard ratio; CI, confidence interval; IgG, immunoglobulin G.
Patients and Methods
For this study, we used data from the United Network for Organ Sharing (UNOS) database, which included 50,616 patients who underwent liver transplantation in the United States between 1988 and 2001. During this period, 668 children (<18 years of age) and 764 adults had LDLT, and the rest (n = 49,184) had deceased donor liver transplantation (DDLT). Most of the adult LDLTs were done between 1998 and 2001, as shown in Figure 1. We compared the outcome of these 764 patients who underwent LDLT (LDLT group) with a comparable group of adult patients (case-controls) who underwent DDLT during the same period. For each patient who underwent LDLT, 2 controls were selected by matching for age ± 5 years, gender, race, diagnosis, and year of transplantation ± 2 years. The controls were selected by starting the search in a random fashion matched for year. Using this technique, we could not match controls for 29 patients, and therefore there were only 1,470 adults chosen as controls (DDLT group).
We collected the following information on recipients: age, gender, race, height and weight at the time of transplantation, serum creatinine at the time of transplantation, etiology of liver disease, ABO blood type matching status, UNOS listing status, cold ischemia time, primary graft nonfunction, and graft and patient survival. In addition, we collected information on donor age, gender, and race, and relationship of the donor to the recipient. Body mass index was calculated by dividing the weight in kilograms (at the time of transplantation) by the square of the height in meters (kg/m2). The definition of UNOS listing status was stable during the study period: UNOS status 1 (fulminant hepatic failure), UNOS status 2A (patient in the intensive care unit or requiring intensive monitoring, in addition to Childs-Turcotte-Pugh score of 10 or more), and others (2B, 3, etc.). We classified the cause of liver disease as alcoholic liver disease, HCV, hepatitis B virus, cryptogenic cirrhosis, primary biliary cirrhosis, and other causes. Based on ABO blood type matching, patients were divided into 3 groups: matched when donor and recipient were of same blood type; compatible when donor blood type was O and recipient blood type was other than O, or when recipient blood type was AB and donor blood type was either A or B; mismatched when recipient blood type was A, B, or O and the donor's blood type was other than O or different from that of the donor.
The continuous variables were compared by Student's t-test and categorical variables by chi-squared tests. The graft and patient survival was determined by Kaplan-Meier survival analysis, and groups were compared by log-rank test. The survival was adjusted for the confounding variables by Cox regression analysis. For all analyses, a 2-tailed P value of .05 or less was considered significant. The statistical analyses were performed using SPSS software version 11.0.1 (SPSS, Chicago, IL).
The demographics of patients who received LDLT and DDLT are shown in Table 1. As per our study design, age, gender, race, diagnosis, and year of transplantation (not shown) were similar in both groups (Table 1). Serum creatinine and cold ischemia time (predictably) were higher in the DDLT group. There were more patients in the DDLT group who were on life support (9.3 vs. 3.6%), and in an intensive care unit (22 vs. 4.7%). In addition, 27.9% of patients in the DDLT group were UNOS status 1 or 2A compared to only 3.8% in the LDLT group, confirming that the DDLT group was considerably sicker than the LDLT group at the time of liver transplantation (Table 1).
Table 1. Recipient Characteristics
Live Donor Liver Transplant (n = 764)
Deceased Donor Liver Transplant (n = 1,470)
Abbreviations: ALD, alcoholic liver disease; HCV, hepatitis C; HBV, hepatitis B; Crypto, cryptogenic; PBC, primary biliary cirrhosis; UNOS, United Network for Organ Sharing; BMI, body mass index; ICU, intensive care unit.
49.7 ± 5.2
49.8 ± 10.8
Gender (female %)
26.2 ± 5.2
27.6 ± 5.4
1.02 ± .8
1.25 ± 1.1
Cold ischemia time (hours)
3.9 ± 7.3
8.4 ± 4.5
UNOS listing (1/2A)
Days on transplant list
306 ± 377
317 ± 380
The relationship of the donor to the recipient was recorded in only 470 of 764 donors; 8% of donors were children, 23% were siblings, 8% spouses, 3% parents, and 10% other relatives, but 18% were unrelated. A total of 90% donated the right lobe, and the rest donated the left lobe or left lateral segments. The donor characteristics are shown in Table 2. The donor age was lower in the LDLT group when compared with the DDLT group, and there were fewer African American donors in the LDLT group (Table 2).
Table 2. Donor Characteristics
Live Donor Liver Transplant (n = 764)
Deceased Donor Liver Transplant (n = 1,470)
Abbreviations: F, female; M, male; ABO, ABO blood type.
Donor age (years)
35.8 ± 10.4
38.9 ± 18.1
Donor gender (female %)
Gender match (% match)
F → M
F → F
M → F
M → M
The surgical complications were not reported to UNOS in the majority (≈90%) of patients in both groups, and of the reported complications, infection rate (25.4 vs. 14.3%; P = .05) was higher in the LDLT group. Biliary complications (8.5 vs. 4.2%) were twice as common in the LDLT group, but it was not statistically significant because of the small sample size. Primary graft nonfunction was reported in 27 (27 / 764; 3.5%) of the LDLT group and 49 (49 / 1,470; 3.3%) of the DDLT group.
Kaplan-Meier patient and graft survival curves are shown in Figure 2. Although patient survival was similar, graft survival was significantly lower in the LDLT group (P < .001 by log-rank test). Since there were differences in the severity of liver disease, creatinine, and UNOS status between the groups, we adjusted the survival by Cox regression analysis. Patients who had LDLT were 60% more likely to lose the graft within 2 years (mostly in the 1st year) compared to DDLT (hazard ratio [HR] 1.6; confidence interval 1.1-2.5).
Figure 3 shows the Kaplan-Meier survival curves in HCV and non-HCV patients who received LDLT. There was no difference between the groups. Figure 4 shows the survival curves in HCV patients who received LDLT and HCV patients who received DDLT. Graft survival was lower in HCV patients who received LDLT (P < .01 by log-rank test), but patient survival was similar.
In this case-controlled study, we have shown that LDLT patients have similar survival when compared with a matched population that received DDLT. However, 1-year graft survival was significantly lower (≈15%) in the LDLT group, indicating that a significant number of patients had retransplantation. Moreover, these results were achieved in relatively stable patients with cirrhosis, a group that could be expected to have a 1-year graft survival between 90 and 95% with cadaver transplantation. Compared to the LDLT group, our DDLT group had many negative predictors of survival including older donor age, longer cold ischemia time, higher serum creatinine, and a higher proportion of patients in the intensive care unit, on life support, or with UNOS status 1 or 2A.7 Therefore, we may have underestimated the true difference in graft and patient survival between the LDLT and DDLT groups. Despite these reservations, on the basis of comparable patient survival, we believe that LDLT is justified in patients who are unlikely to receive DDLT in a timely fashion in this era of acute cadaver organ shortage.
Our study also showed a trend towards similar survival in HCV patients that received LDLT when compared with non-HCV recipients who received LDLT. It is premature to make any judgements in HCV patients based on these trends, but we believe that the role of LDLT in HCV patients should be defined based on long-term (5-year) follow-up data. Five-year survival rates in HCV recipients who receive DDLT are relatively poor in comparison with non-HCV patients, and a worse outcome with LDLT may prove to be unacceptably low to justify the costs, considering that many of them may have to undergo retransplantation.8 In addition, the outcome of retransplantation, especially in those with HCV, is worse than the outcome of primary transplantation.9
The reasons for low graft survival are not explained by our study. It is possible that the relatively better survival reported in the literature might reflect a reporting or publication bias. It is also possible that the better, published results may have come from centers with more experience. Moreover, our data may reflect a learning curve since LDLT was done predominantly since 1999. A study from Japan had shown that in patients who undergo LDLT, those with poor risk factors, as in DDLT, had a worse outcome compared to those with “good” risk factors.7, 10 We do not believe that the lower graft survival in the LDLT group was secondary to poor recipient risk factors, since those who received LDLT had fewer risk factors compared to our DDLT group. In a series of 157 adult LDLTs performed in Korea from 1997 to 2000, Lee et al.11 reported a sharp decrease in mortality as they gained more experience, 29% in 1997 to 5.7% in 2000. It is possible that such a learning curve, since the data came from many centers in the United States, was one of the factors that determined lower graft survival. It has been shown that a graft-to-recipient body weight ratio of less than .8% is associated with an increase in graft loss.12, 13 Whether lower graft-to-recipient body weight was a cause of increased graft loss could not be determined in our study, since this information was not available to us. However, these possibilities do not diminish the relevance of the findings of our study, and we believe that our findings are a reflection of the current status of LDLT in the United States. It is possible that these results might improve with more experience, but that remains speculative in a surgery of this nature in which there is considerable risk to the donor in addition to the recipient.
Our observations show that short-term patient survival after LDLT is comparable to that of DDLT. The main advantage to the recipient is a decrease in waiting time, which in many instances can be life saving. The positive aspects extend globally by increasing the organ pool, especially in countries where there is no history of cadaver organ transplantation. However, the concern for donor safety and lower graft survival remains the critical issue. The success of LDLT should not result in inertia in exploring ways to increase deceased donor organ donation. It is clear that the majority of donors are satisfied with the process of LDLT, but the attitudes of the lay public reflect an unrealistic tolerance for adverse donor outcomes, more than we as a medical community deem reasonable.14, 15 This is a clear signal that we should continue to strive to improve outcomes and availability of LDLT, and it also reinforces the importance of having a multidisciplinary team approach with objective criteria for the selection of donors and recipients for adult LDLT.
We thank UNOS for providing us with the data for this study.