The upper age limit for organ donation for liver transplantation has increased over the past few years. A retrospective case control study was carried out to evaluate the outcome of 36 liver transplants (group A) performed with grafts procured from donors over 70 years old in the period 1996 to April 2000, matched with 36 transplants (group B) chronologically performed thereafter with organs procured from donors below the age of 40 yr. The groups were comparable as regards main clinical characteristics. Mean follow-up was 14.5 months. Clinical and laboratory parameters of the donors, cold ischemia period, intraoperative blood transfusions, 30-d mortality, incidence of primary graft nonfunction, acute rejection episodes, arterial complications and long-term survival of recipients were considered. The main postoperative biochemical parameters were also collected and compared. A liver biopsy was obtained in 20/36 old donors, revealing less than 25% of steatosis in all but one, which showed steatosis involving 70% of the hepatocytes. There were two postoperative deaths (5.6%) in group A and one (2.8%) in group B (p = NS). Seven postoperative arterial complications (19.4%) occurred in group A, leading to the patient's death because of rupture of the hepatic artery in one case, to successful surgical revascularization in three cases and to retransplantation in three cases. Only one patient in group B (2.8%) experienced hepatic artery thrombosis (p = 0.055). One-year patient survival rates were 77.4% for group A and 88.8% for group B (p = NS); 1-yr graft survival rates were 73.3% for group A and 85.7% for group B (p = NS). In conclusion, donors over 70 should not be excluded a priori for liver transplantation in elective settings. Great attention should be paid to the pathological conditions of arterial vessels caused by atherosclerosis, i.e. the presence of calcified plaques on the hepatic artery, which might represent the source of severe complications.
The persistent scarcity of organ donors and the growing number of patients listed every year for liver transplantation has recently prompted transplant teams to evaluate new methods for increasing graft availability, such as xenotransplantation, usage of living donors and organ splitting. However, the more immediate solution seems to be the expansion of the cadaveric donor pool by including the so-called ‘marginal’ subjects, represented by elderly donors or persons with marked liver steatosis.
There is still controversy regarding the upper age limit for liver donation: in conventional opinions, advanced age represents a risk factor, mainly when associated with prolonged ischemia time, liver steatosis or obesity (1,2). On the other hand, recent experiences reported in the world literature have shown the possibility of achieving satisfactory results by performing orthotopic liver transplantations (OLTs), formerly in emergency and then in elective settings, with organs obtained from donors over 60 yr (3–5). These increasing observations have progressively extended the use of elderly donors to those even over 80 yr, also for elective procedures (6,7).
In the present study, we retrospectively evaluated our experience of OLTs performed with grafts procured from donors over 70 years old; we, therefore, compared the outcome of these operations with that obtained with donors below the age of 40 yr, chronologically correlated, in the period 1996–2000.
Materials and Methods
Between December 1996 and April 30, 2000, 36 patients underwent OLT at our institution, with organs from donors over 70 years old (group A); the outcome of these operations was matched with 36 OLT (group B) chronologically performed thereafter with grafts obtained from donors below the age of 40 yr. Emergency procedures for fulminant hepatic failure and retransplantations were excluded from the study. Mean follow-up was 14.5 months (range: 0.2–41.5 months).
The following donor-related clinical parameters were evaluated: age, gender, cause of death, length of the intensive care unit (ICU) stay before liver procurement, aspartate-aminotransferases (AST), alanine-aminotransferases (ALT), total bilirubin (TBIL), γ-glutamyl transferases (γ-GT), prothrombin activity (PT), activated partial thromboplastin time (aPTT) levels and arterial blood pressure (ABP) at the time of procurement, any hypotensive episodes (ABP < 80 mmHg for more than 1 h), cardiac arrests and vasopressive drug infusion prior to and during organ harvesting.
Liver biopsies from donors over 70 years old were initially carried out in cases of macroscopic alterations (i.e. steatosis or fibrosis) of the liver and on the basis of the evaluation of harvesting surgeon, and therefore on a routine basis.
We also analyzed the recipients' age, sex, indication for OLT, status (according to UNOS – United Network of Organ Sharing – classification in use until 1997), cold ischemia period, total transplant operation time, technique of the operation and intraoperative blood replacement.
Selection of the donors
Suitability criteria for the use of a liver graft for transplantation at our center have already been discussed elsewhere (8); nevertheless, the initially referred elderly donors were carefully assessed also on the basis of the biochemical liver function tests (LFTs).
There were no fixed criteria for determining organ suitability other than our clinical evaluation. Nevertheless, we adopted more strict criteria for the use of grafts from very aged donors, particularly during the initial phase of this experience.
During the study period, 66 liver donors over 70 years old were referred to our center; two of them (3%) were used for an emergency procedure (fulminant hepatic failure). Twenty-three (34.8%) were judged as unsuitable for transplantation and discarded: 13 of them (19.7%) were grossly abnormal or cirrhotic; three (4.5%) had macroscopic steatosis (in one case an 80% steatosis was documented by biopsy); three (4.5%) were hepatitis C virus (HCV)-positive; two (3%) were HBcAb-positive, without available hepatitis B virus (HBV)-positive patients, one (1.5%) had marked hypertransaminasemia (AST 674, ALT 1845 U/L at the referring time) and one (1.5%) had metastatic nodules from renal tumor, undetected prior to graft harvesting. Furthermore, five donors (7.6%) were not accepted by our center for organizational reasons and were procured by other teams.
The percentage of donors over 70 years old utilized at our center for OLT during the past years is shown in Figure 1.
Characteristics of the donors
Demographic characteristics of the donors are summarized in Table 1. There was a prevalence of males in both groups, while no significant differences in the number of hypotensive episodes, cardiac arrests and dopamine infusion were found. The causes of death were significantly different between the two groups, with a higher incidence of cerebrovascular accidents among older donors (p < 0.05).
Table 1. Demographic profiles of the donors considered in the study
Group A (n = 36)
Group B (n = 36)
74.5 ± 3.9 (70–87)
26.2 ± 6.7 (13–40)
p < 0.05
Cause of death
p < 0.05
3.6 ± 3.4 (0–13)
2.9 ± 3.5 (0–14)
5.4 ± 2.3 (2–15)
6.4 ± 3.4 (0–17)
Biochemical profiles are reported in Table 2. A significant difference was encountered in AST and ALT levels before liver procurement, with group B donors having higher levels. These findings could be explained either by the above-mentioned stricter selection among older donors, who were discarded on the basis of altered LFTs (namely in the earlier cases), and by the higher amount of younger donors who died as a result of trauma.
Table 2. Biochemical profiles of the donors considered in the study
Group A (n = 36)
Group B (n = 36)
46 ± 33.7 (8–149)
81.1 ± 76.4 (9–407)
p < 0.05
46 ± 33.7 (8–149)
81.1 ± 76.4 (9–407)
p < 0.05
1.2 ± 1 (02–4)
1.1 ± 0.9 (0.1–4.7)
55.1 ± 83.2 (7–347)
27.7 ± 29.8 (2–140)
73.6 ± 19.7 (38–121)
69 ± 21.3 (13–104)
39.4 ± 12.7 (20–85)
35.4 ± 16.7 (0.9–99)
Characteristics of the recipients
Demographic characteristics and indications for OLT of the recipients are listed in Table 3. Group A patients were significantly older than group B patients; however, there was no intentional policy to allocate the older graft to the older recipient and we cannot explain the reason for this apparent association.
Table 3. Demographic profiles of the 72 recipients considered in the study
Group A (n = 36)
Group B (n = 36)
54.6 ± 6.7 (37–68)
47.7 ± 8.10 (17–59)
p < 0.05
Indication for OLT
HCC on cirrhosis
No differences were found among the remaining variables.
A liver biopsy was carried out in 20 out of 36 donors over 70 years old. In the remaining 16 cases, a biopsy was not obtained due to normal gross appearance of the liver. In 19 cases we did not observe any pathological findings (including intrahepatic arteriolar disease) and the degree of macrovescicular steatosis was 25% or less; in one case, the histological examination showed steatosis involving more than 70% of the parenchyma. However, in this latter case the graft showed a normal functional recovery.
Twenty-eight organs (77.8%) were flushed with UW (Viaspan, Dupont Pharma®, the Netherlands) solution and eight (22.2%) with Celsior (Imtix, Sangstat®, Lyon, France) solution in each group. No differences were found between the two groups regarding intraoperative transfusion of packed red blood cells (PRBC) (2066 ± 1399 mL in group A vs. 2726 ± 2432 mL in group B) (p = NS). There were comparable hypothermic ischemia periods (495.4 ± 128.6 min vs. 546.5 ± 137.5 min) and total OLT operation time (455.3 ± 70.2 min vs. 478.6 ± 114.9 min). Thirty-one recipients (86.1%) in group A and 28 (77.8%) in group B were operated on using the ‘piggy-back’ technique, while the conventional technique (with or without use of the veno-venous by pass) was adopted in five (13.9%) in group A and eight (22.2%) in group B (p = NS).
Immunosuppression was induced with cyclosporine in 31 patients (86.1%) of group A and in 32 (88.9%) of group B, and with tacrolimus in five (13.9%) of group A and in four (11.1%) of group B.
Primary end-points of this study were: (1) to assess the perioperative mortality rate, defined as any deaths that occurred within 30 d from OLT; (2) to establish the incidence of primary graft nonfunction [PGNF, defined as a postoperative (p.o.) graft failure leading to the patient's death or to retransplantation within the first p.o. week] and retransplantation; (3) to evaluate long-term survival.
In consideration of the higher prevalence of atherosclerotic changes involving arterial vessels of the older population, p.o. arterial complications of the two groups were also specifically compared.
The secondary aim was to study the p.o. course of grafted patients by analyzing the following biochemical parameters: AST, ALT, TBIL, alkaline phosphatases (AP) γ- GT, PT, aPTT, antithrombin III (AT III), total plasma proteins, albuminemia, creatininemia, uremia levels through the first 14 p.o. d.
Moreover, data on the daily bile secretion through the first 7 p.o. d, ICU and total hospital stay and number of acute rejection episodes were collected and compared.
Results are expressed as mean ± standard deviation. Differences between means were evaluated with Student's t-test. Differences between groups were evaluated with the χ2-test. Postoperative survival was computed from the day of OLT to the last follow-up visit or the date of death. Survival rates were estimated by means of the life table method and differences between survivals compared with the log rank test. Data were managed with the use of the SPSS software packaging (SPSS Inc., Chicago, USA). p-values < 0.05 were considered as statistically relevant.
Three of the 72 patients died within the first 30 p.o. d, representing 4.2% of the total. Causes of death were multiorgan failure (MOF) after retransplantation performed for a PGNF, uncontrollable acidosis with failure of graft function recovery and the rupture of the hepatic artery, respectively. There were two deaths (5.6%) in group A and one death (2.8%) in group B (p = NS).
Incidence of PGNF
A total of four grafts (5.6%) experienced PGNF. One episode (2.8%) of PGNF occurred in group A. Due to the lack of graft function recovery, this patient required a retransplantation 10 d after the first procedure. In group B, three patients (8.3%) experienced a PGNF. There were no differences between the two groups.
Among patients belonging to group B, one suffered from severe intraoperative bleeding and coagulopathy; moreover, the ischemia period exceeded 15 h. This patient died on the 5th p.o. day. The remaining two received grafts obtained from donors who remained in ICU for more than 13 d; they underwent retransplantation on the 2nd and 3rd p.o. days.
Incidence of retransplantation
There were four retransplantations (11.1%) in each group. In group A, one patient received a second graft because of PGNF. The remaining three patients underwent retransplantation 33, 52 and 63 d after OLT as a consequence of hepatic artery thrombosis.
In group B, two patients were retransplanted for PGNF. The third patient received a second graft 181 d after the first procedure, because of HCV infection recurrence and the remaining patient after 8 d due to hepatic artery thrombosis.
In group A, seven vascular complications (19.4%) occurred, mainly related to hepatic artery reconstruction. Two patients were re-operated in the early p.o. period due to hepatic artery thrombosis; a revascularization of the artery was successfully performed. A third patient had a rupture of the hepatic artery at the level of an atherosclerotic plaque of the graft 25 d after OLT and died due to subsequent coagulopathy after emergency laparotomy.
Three already-mentioned patients underwent retransplantation for hepatic artery thrombosis: one died due to MOF 6 months after re-operation and the remaining two are presently alive and well.
The remaining patient presented an aneurysm of the arterial reconstruction performed with the fold-over technique (because of the presence of a right hepatic artery arising from the superior mesenteric). The patient underwent surgical removal of the dilatated tract, followed by an end-to-end anastomosis 6 months after OLT.
Overall, in group A, a vascular complication occurred in half of the cases (four out of eight) where the graft had a right hepatic artery arising from the superior mesenteric artery and requiring the use of the fold-over technique for reconstruction.
In group B, only one patient presented recurrent hepatic artery thrombosis, leading to three consecutive retransplantations, caused by a previously undetected hypercoagulable state (presence of anticardiolipin antibodies).
Even though not statistically relevant (p = 0.055), the difference between the two groups was remarkable.
Hepatic function recovery
The postoperative course of AST, ALT and TBIL is shown in Figure 2. AST level on p.o. days 2 and 4, and ALT level on p.o. day 7 were significantly higher in group A, while TBIL levels on p.o. days 4, 10 and 12 were significantly lower in group B.
Other parameters found as being significantly more favorable in group B compared with group A were: AP on p.o. day 13 (269.1 ± 98.2 U/L vs. 452.2 ± 241 U/L); PT on p.o. days 4 (76.7 ± 17.5% vs. 65.5 ± 19.8%) and 12 (86.6 ± 11.6% vs. 73.1 ± 15.9%); aPTT on p.o. days 1 (1.19 ± 0.16 INR vs. 1.28 ± 0.23 INR), 4 (0.86 ± 0.16 INR vs. 0.97 ± 0.18 INR), 6 (0.79 ± 0.15 INR vs. 0.89 ± 0.12 INR) and 9 (0.78 ± 0.12 INR vs. 0.84 ± 0.11 INR); AT III on p.o. days 1 (62.5 ± 12.6% vs. 55.3 ± 9.9%) and 2 (93.6 ± 104.1% vs. 63.1 ± 12.8%); TP on p.o. day 12 (6.0 ± 0.6 g/dL vs. 5.5 ± 0.5 g/dL); albumin on p.o. days 4 (3.4 ± 0.4 g/dL vs. 3.2 ± 0.5 g/dL) and 6 (3.5 ± 0.5 g/dL vs. 3.2 ± 0.5 g/dL).
No differences were found between the two groups in bile secretion through the T-tube within the first 7 p.o. days.
Mean postoperative ICU stay was 5.2 ± 5.5 d (range: 1–35 d) for patients of group A and 5.7 ± 12.2 d (range: 1–76 d) for those of group B (p = NS). Total hospital stay (excluding patients who died while in ICU) lasted 22.0 ± 13.3 d (range: 10–72 d) for group A and 21.9 ± 17.1 d (range: 8–78 yr) for group B (p = NS).
Acute rejection episodes
No intractable rejection episodes requiring retransplantation occurred. Twenty-two (30.6%) patients experienced one, or more than one, acute rejection episodes, which recovered either by increasing the standard immunosuppression or by steroid bolus administration. Twelve (33.3%) of these patients were from group A and 10 (27.8%) from group B (p = NS). One patient in group A developed a chronic rejection requiring a switch from cyclosporine-based to tacrolimus-based immunosuppression.
Causes and timing of death (after 30 d from transplant) are reported in Table 4. One-year patient survival rates were 77.4% for group A and 88.8% for group B (p = NS); 1-yr graft survival rates were 73.3% for group A and 85.7% for group B (p = NS) (Figures 3 and 4).
Table 4. Causes and timing of death 30 d after OLT
Donor age (yr)
Indication for OLT
Cause of death
Sepsis (untreated erisipela)
HCC on cirrhosis
Sepsis (after HCV recurrence)
Liver failure (HCV recurrence)
HCC on cirrhosis
Liver failure (after retransplantation)
One of the main aspects of organ transplantation remains the donors shortage. In recent years the number of donors has changed very little, at least in most European countries. On the contrary, the waiting list for organ transplantation has grown by around 27% per yr and approximately 24% of patients listed for OLT die without receiving a graft (9).
The potential means of expanding the donor pool include the use of living related donors, xenotransplantation and graft splitting; however, an immediately available method seems to be the usage of ‘marginal’ donors, such as steatosic or elderly patients.
Initial observation of the literature has shown a better survival rate 1 yr after transplantation in patients whose donors were aged between 15 and 45 years old in comparison with those with a donor aged between 45 and 55 years old (9). The European Liver Transplant Registry (ELTR) considers 50 as the cut-off age to analyze graft survival. In its semiannual report, a lower survival rate for those 2619 recipients that received a graft from a donor over 50 years old has been reported (ELTR, Villejuif, France, personal communication).
Data retrieved from the UNOS Registry on 7988 patients who received a first liver graft between 1987 and 1992 show that the 6-month graft failures progressively increased by decades of donor age, with a rate of 50% higher for donors aged over 50 yr by comparison with those aged less than 20 yr (10).
In the largest published analysis, Marino et al. showed in a univariate fashion that the only two donor-related factors significantly associated with poor results were older age (mean 43.9 ± 1.4 years old) and female donor gender. In the subsequent multivariate analysis, graft loss within 90 d independently correlated with donor age, while both advanced age and female gender were among the variables independently associated with late failures (11).
These latter reports reflected the current opinions and expertise available during the study periods (early 1990s) and included only a limited proportion of truly older donors (10,11).
Until a few years ago donor age up to 60 yr was considered as the maximum threshold for using a graft, and older donors were almost exclusively evaluated for urgent procedures (12,13,14).
In elderly subjects parenchymal abnormalities can be observed more frequently, in particular steatosis. This latter represents a risk factor for the outcome of OLT, being a potential cause of PGNF when involving more than 30% of the hepatocytes and with a macrovescicular histological pattern (1,3,15).
It has been demonstrated that aging does not per se affect either the parenchyma or the intrahepatic vascular structures (as is the case with other organs) (6). Recent series of OLTs performed with the use of donors older than 60 years old (3–5,16,17) and, in one case, an 87-year-old (7), show that graft recovery can be satisfactory.
Adam et al. (18) compared a group of patients whose donors were aged over 50 yr with a group receiving an organ from donors below the age of 50 yr, showing no substantial differences in terms of liver function recovery, incidence of PGNF or retransplantation and graft survival up to 6 months. The authors applied more restrictive criteria in older donor selection on the basis of LFTs and monoethylglycinexylidide (MEGX) test. Hoofnagle et al. (19) showed that grafts from older (> 50 yr) donors judged of ‘good’ quality had a survival similar to that of livers from younger donors, even though the older ones were more likely to be assessed as ‘poor’ and graft survival was significantly worse in the whole series.
In a group of 36 patients transplanted using donors over 70 years old, Emre et al. (16) reported a 1-yr actuarial graft and patient survival of 85% and 91%, respectively. Intraoperative blood losses, early graft function and ICU and ward stay were not different compared with those of patients grafted with organ donors aged below 70 yr. Unsatisfactory results were obtained in the event of donor obesity, AST levels over 120 U/L, serum osmolality > 330 mOsm, γ- GT > 100 U/L, ICU stay > 5 d, any hypotensive episodes (ABP < 80 mmHg for more than 1 h) or cardiac arrests, and high-dose dopamine infusion.
Up to the end of the period of the present study, only 12 OLTs have been performed in Europe using donors aged over 80 yr and just five using donors aged over 87 yr; two of the respective recipients are alive at present (ELTR, Villejuif, France, personal communication). According to UNOS Registry data, 28 liver transplants were carried out between 1993 and September 1998 in the United States with donors over 80 yr, four of them with donors aged over 85 yr (UNOS, Richmond, VA, USA, personal communication).
The greater experience gained during the second half of the 1990s in the treatment and selection of organ donors led to extend both indications for OLT and acceptance criteria for organ donors, with the final results of increasing the number of procedures performed.
In our center, we have been using very old donors for elective OLTs since the end of 1996, and their number has progressively increased (Figure 1).
In the evaluation of old donors, we tried to maintain advanced age as a single risk factor. This policy included a careful assessment of LFTs normality, hemodynamic stability prior to procurement, absence of parenchymal alterations of the graft and absence of previous hepatitis.
Among the older donors, mean ICU stay before harvesting was short (3–4 d), they had received low doses of dopamine (5.4 γ/kg/min) and the incidence of hemodynamic alterations was minimal. We tried to reduce the ischemia period as much as possible, although there was no difference from that of the younger patient group. The lack of significant structural alterations found in the grafts should also be outlined.
Liver biopsies were systematically carried out each time a parenchymal abnormality was suspected; following that, we never encountered a fatty infiltration involving more than 25% of the hepatocytes, except in one case, when a very steatotic (70%) graft nevertheless showed an excellent postoperative recovery.
Rather than a functional defect, the main problem that emerges from our study is the intriguing technical demand of arterial reconstruction. A marked and widespread atherosclerosis represents an almost constant finding in older patients and it can determine an increased risk of surgical complications due to the difficulty in performing a safe vascular anastomosis in the presence of a thick and hard arterial wall. This has led to a significant increase in the amount of our postoperative complications which, in some cases, have progressed to the patient's death; moreover, it definitively affected long-term graft and patient survival, contributing to the difference of more than 10% at the 1-yr follow-up. Even if no statistical difference was revealed, it must be taken into consideration that the number of patients included in the present study is relatively small.
In order to avoid performing the arterial anastomosis at the site often damaged by atherosclerotic changes, like the aorta patch surrounding the ostium of the celiac trunk, we usually performed the anastomosis at the level of the first tract that was spared from atherosclerosis. This more often corresponded to the common hepatic artery of the graft.
A safe use of the hepatic artery is also reported in the description of other different surgical procedures, such as the attempts to revascularize the renal artery when made critically stenotic by atherosclerotic plaques (20). The common hepatic artery seems to be relatively spared from atherosclerotic processes, even in the event of massive involvement of the aorta and splanchnic arteries (20).
In spite of these technical considerations, we have observed an unusually high incidence of vascular defects, almost absent in the population of patients transplanted with grafts obtained from younger donors.
This observation eventually led us to change our policy of accepting grafts from very aged donors. Much more attention is now placed on the evaluation of the status of the arterial supply of the grafted livers rather than their functional reserve. The condition of the hepatic artery is careful evaluated on the back-table before prepping the recipient and we have introduced the possibility of discarding the organ even at this stage.
In the series of Marino and colleagues (11), a higher incidence of ischemic injuries was reported among patients transplanted with livers from donors over 60 years old compared with those receiving organs from younger donors. Even though the authors did not specify which vascular site was involved, an ischemic damage as cause of graft failure was the only variable approaching statistical significance between the two groups examined.
Our study reveals that postoperative mortality, incidence of PGNF and retransplantation were not statistically different between the two groups of patients.
Besides lower values of PT, aPTT, AT III, total plasma proteins and albumin during the first 14 p.o. days, we registered a slight increase in cholestasis.
A progressive growth of cholestatic indexes, accompanied by simultaneously decreasing cytolitic activity, is a common and already described finding (4,16,19,21), with a significant influence on survival rates (21). Although a similar course was observed in some of our cases, this did not seem to determine differences in p.o. stay nor an increased rate of complications. Number and severity of rejection episodes were also comparable in the two groups. No definitive explanation has been proposed for this phenomenon.
In conclusion, donors aged 70 yr and over should not be excluded a priori from evaluation for liver transplantation in an elective setting. Their use requires the careful assessment of hemodynamic conditions and the normality of liver function tests before procurement; the ischemia period should be maintained as short as possible, while the absence of morphological alterations of liver parenchyma should always be verified.
The frequency of atherosclerotic damage in elderly individuals can determine a significant risk of vascular complications. The surgeon should therefore carefully evaluate this when carrying out the arterial anastomosis, also in order to better identify the safest site to perform it.