Liver transplantation is the only chance of cure for patients with end-stage liver disease and results in a 5-year survival rate of 70% to 85% in most centers. Although demand for the operation has increased, rates of deceased organ donation have not kept pace, and this had led to a marked shortage of organs available for transplantation.1 One response to this shortage has been to accept increasingly older donors. In 1988, only 28% of all donors were older than 50 years, and 0.05% were older than 65. By 1995, these percentages had increased to 42% and 5%, and by 2007, they had increased even further to 65% and 10%, respectively.1
Aging affects nearly all the cell types that play roles in liver transplantation outcomes. These include both the donor-derived hepatocytes2–5 and the recipient cells: inflammatory cells and those cells that gradually populate the graft and are derived from recipient bone marrow.6, 7 Aging profoundly affects the ability of the liver to respond to stress.4, 8–11 This response to stress, whether ischemia-reperfusion injury,4, 8 inflammation,9 or hepatitis C virus (HCV) recurrence,10, 11 triggers a complex cascade involving both donor-derived hepatocytes and endothelial cells and recipient-derived nonparenchymal cells, such as T-cells, platelets, and macrophages. These processes involve both donor- and recipient-derived cells; therefore, the effects of donor and recipient age on short- and long-term outcome should not be considered independently but rather together. It has recently been recognized that increased donor age is associated with more severe HCV recurrence.12–14 In contrast, the effect of donor age on reperfusion injury and non–HCV-related long-term outcomes remains unclear. Although some series have suggested an increased risk for primary nonfunction and a higher incidence of biliary complications in recipients receiving an older graft,15, 16 others have failed to show any significant differences in any of the outcome parameters examined.17 The effect of recipient age on short- and long-term outcomes in patients receiving an older or younger graft has not been studied to date.
In this study, we sought to determine the impact of donor and recipient age on transplantation/ischemia-reperfusion injury (TIRI) and long-term clinical outcomes following liver transplantation in the present era of transplantation, using a large population of patients from a single North American center. A particular focus was to determine whether recipient age affects outcomes of liver transplantation when livers from older donors have been used.
ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CI, confidence interval; CIT, cold ischemia time; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HR, hazard ratio; ICU, intensive care unit; INR, international normalized ratio; LOS, length of hospital stay; MELD, Model for End-Stage Liver Disease; TIRI, transplantation/ischemia-reperfusion injury; WIT, warm ischemia time.
PATIENTS AND METHODS
The study was approved by the Research Ethics Board of the University Health Network (Toronto, Canada). Between August 1998 and March 2008, 822 patients underwent whole-organ deceased donor liver transplantation at the Toronto General Hospital, University of Toronto. Partial grafts, such as living donor grafts and split grafts, were not included in the series. Donor and recipient information was prospectively collected and retrospectively analyzed for donor and recipient variables associated with immediate and long-term postoperative outcomes. Maximum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels within 48 hours after transplantation were assumed to reflect the extent of hepatocellular reperfusion injury. Early graft function was assessed with the international normalized ratio (INR) and serum bilirubin levels at 7 day after surgery. Kidney function was assessed by serum creatinine and requirements for dialysis at 1 week after surgery. All complications within the same hospitalization were recorded and graded according to the Clavien score.18 The severity of HCV recurrence was determined with the Metavir scoring system.19 Graft survival and patient survival (at 3 and 5 years) were used as markers for long-term outcome.
Donors were divided into those less than 60 years old and those 60 years old or older, as described by others.20 There is no age limit at our institution for cadaveric liver donors, and this categorization defined 20% of our donor population as “older.” Similarly, recipients were categorized into those less than 50 years old and those 50 years old or older. The median age of liver transplant recipients at our institution is 55 years. The recipient age limit was defined to describe a clinically relevant age group, which contained about 60% of our recipient population. Additional donor characteristics, such as donor gender, length of intensive care unit (ICU) stay, and cause of death, were analyzed. Liver biopsy at the time of transplantation is performed in our program only if gross steatosis is suspected on macroscopic evaluation at the time of organ retrieval. Recipient characteristics such as recipient age and gender, type of liver disease, acute liver failure versus chronic liver failure, medical Model for End-Stage Liver Disease (MELD) status at the time of transplantation (disregarding tumor MELD points), year of transplantation, cold ischemia time (CIT), warm ischemia time (WIT), and presence or absence of hepatocellular carcinoma (HCC) were analyzed as well.
Fisher's exact test and an analysis of variance test were used to compare categorical and continuous parametric variables, respectively. The Mann-Whitney U test was performed for nonparametric continuous endpoints. Graft survival and patient survival were determined with the Kaplan-Meier method and survival curves and were compared with the log rank test (SPSS for Windows 11.0, SPSS, Chicago, IL). Multiple linear regression analysis was used to assess variables that were independently associated with graft injury and outcome. A Cox proportional hazard model was used to determine risk factors for graft and patient survival. Bonferroni post hoc correction was applied in all multiple variable analyses. A P value of <0.05 was taken to indicate statistical significance.
Impact of Donor Age on Short-Term Graft Injury and Long-Term Outcomes
Donors were grouped into those less than 60 years old (n = 625) and those 60 years old or older (n = 197), as described previously. The young and old donor groups had similar characteristics in terms of donor gender, ICU stay, cause of death, recipient age, recipient gender, recipient medical MELD score (without tumor MELD points), CIT and WIT, presence of HCC, chronic liver failure versus fulminant liver failure, and type of immunosuppression (Table 1).
Table 1. Recipient Characteristics of the Old and Young Donor Groups
Donor < 60 Years Old
Donor ≥ 60 Years Old
Abbreviations: BMI, body mass index; HCC, hepatocellular carcinoma; ICU, intensive care unit; MELD, Model for End-Stage Liver Disease.
Although baseline demographics were similar in the young and old donor groups, hepatocellular injury and function—but not graft or patient survival—differed significantly in the early postoperative period. As shown in Table 2, donor age ≥ 60 was associated with higher postoperative maximum AST and ALT. In addition, livers from older donors had an increased INR on days 3 and 7 and a trend toward increased bilirubin on day 3 without statistical significance being reached. In contrast, no difference was observed in the postoperative creatinine, alkaline phosphatase (ALP), overall complication score, or length of hospital stay (LOS; Table 2). The 30-day patient survival (97% versus 97%) and graft survival (95% versus 95%) were identical for recipients of both older and younger grafts.
Table 2. Outcomes of Liver Transplantation from Younger Grafts (<60 Years Old) Versus Older Grafts (>60 Years Old)
Donor < 60 Years Old
Donor ≥ 60 Years Old
Abbreviations: ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; INR, international normalized ratio; LOS, length of hospital stay.
Multivariate regression analysis was performed to determine the significant risk factors underlying TIRI. In the initial analysis, donor age, length of ICU stay, recipient age, CIT and WIT, recipient diagnosis, medical MELD score, and acute liver failure versus chronic liver failure were entered into a univariate analysis to determine factors associated with increased AST release following reperfusion. WIT, CIT, and donor age were identified as the only factors significantly associated with AST elevations, whereas donor age and CIT were significantly associated with posttransplant ALT elevations. In contrast, recipient age and gender, MELD score, presence of HCC, transfusion requirements, and acute liver failure versus chronic liver failure did not affect the severity of immediate graft injury after transplantation in the population as a whole.
A similar univariate analysis was performed to determine risk factors for poorer immediate graft function and short-term outcomes. In this analysis, the MELD score was the only independent risk factor associated with 1-week peak bilirubin and creatinine elevation. The MELD score and WIT were independent risk factors for maximum postoperative INR within the first week, whereas no independent risk factors could be identified for ALP elevation, LOS, or the incidence of postoperative complications.
Although early recipient and graft survival rates were similar, we found that long-term outcomes were different in patients who received older grafts. Graft survival was significantly decreased with old grafts versus young grafts with 1-, 3-, and 5-year survival of 86% versus 90%, 73% versus 85%, and 72% versus 81%, respectively (P < 0.0002; Fig. 1A). Using Cox proportional regression, we analyzed the impact of donor age on graft survival. Donor age as a continuous variable [hazard ratio (HR): 1.016, confidence interval (CI): 1.0–1.026, P < 0.002] and as a categorical variable (<60 versus ≥60 years, HR: 2, CI: 1.95–2.3, P < 0.0001) significantly increased the risk of graft loss. Patient survival was lower with older grafts versus younger grafts with 1-, 3-, and 5- year survival of 88% versus 92%, 77% versus 88%, and 77% versus 82%, respectively (P < 0.003; Fig. 1B).
Having demonstrated that donor age is significantly associated with graft outcomes, we performed a Cox proportional regression analysis to determine independent risk factors for graft survival in patients receiving an older graft. Donor variables (donor gender, donor body mass index, and length of donor ICU stay), intraoperative parameters (CIT and WIT), and recipient characteristics (recipient diagnosis, recipient age, recipient gender, presence of HCC, medical MELD, and intubation prior to transplant) were entered into the analysis. The presence of HCV infection (P < 0.02, HR: 3, CI: 1.3–6.8) and recipient age ≥ 50 (P < 0.006, HR: 5.8, CI: 1.7–19.8) were identified as the only independent risk factors for the survival of grafts ≥ 60 years old.
Does Recipient Age Affect Graft Injury or the Long-Term Outcome of Older Grafts?
In the previous analysis, we identified recipient age ≥ 50 years as an independent risk factor for graft survival from donors ≥ 60 years of age. Therefore, we analyzed the impact of recipient age on graft injury and long-term outcome in the overall population and in the subgroup of patients who received an older graft. As noted previously, recipients were divided into those younger than 50 years (n = 302) and those 50 years old or older (n = 520). The 2 recipient groups had similar donor (gender, ICU stay, and central venous pressures) and recipient characteristics (gender, medical MELD score, CIT, WIT, presence of HCC, and chronic liver failure versus fulminant liver failure; Table 3).
Table 3. Characteristics of the Old and Young Recipient Groups
Recipient < 50 Years Old
Recipient ≥ 50 Years Old
NOTE: The MELD score was calculated on the basis of creatinine, bilirubin, and INR; MELD points for HCC were disregarded. Abbreviations: BMI, body mass index; HCC, hepatocellular carcinoma; ICU, intensive care unit; MELD, Model for End-Stage Liver Disease.
When we analyzed the effect of recipient age in the population as a whole, no difference was observed in postoperative AST, ALT, ALP, bilirubin, creatinine, LOS, or severity of complications between young and old recipients. Similarly, old and young recipients had comparable 1-, 3-, and 5-year graft survival (89% versus 92%, 83% versus 85%, and 79% versus 83%, P = 0.14; Fig. 2A) and patient survival (91% versus 93%, 85% versus 86%, and 80% versus 85%, P = 0.08).
However, there was a significant association between recipient age and short-term graft injury and long-term outcomes in patients who received a graft from an older donor. One hundred twenty-two grafts from donors 60 years old or older were transplanted into recipients 50 years old or older, whereas 58 grafts from older donors were transplanted into recipients younger than 50. Parameters of immediate and long-term outcomes were analyzed as described previously. Transplanting an older liver into a younger recipient versus an older recipient was associated with less elevation in postoperative AST release (1517 ± 946 versus 2732 ± 1240 (IU/L), young versus old, P < 0.03) and ALT release (806 ± 519 versus 1310 ± 931 (IU/L), P < 0.02), whereas no effect was observed on parameters of short-term graft function, such as bilirubin or INR on day 3 or day 7, creatinine, ALP, LOS, or postoperative complications.
A similar analysis was performed to search for an effect of recipient age on long-term clinical outcomes. Graft survival and patient survival were compared for old grafts (≥60 years) implanted into old recipients (≥50 years) and young recipients (<50 years). Old grafts implanted into older recipients versus younger recipients led to significantly decreased 1-, 3-, and 5-year graft survival (84% versus 96%, 71% versus 83%, and 69% versus 83%, P < 0.03; Fig. 2B) and decreased patient survival (86% versus 95%, 72% versus 87%, and 70% versus 87%, P < 0.02).
Impact of Graft Age on Short- and Long-Term Outcomes in HCV(+) Recipients
In our series, HCV infection was the second independent risk factor (with recipient age) for graft survival in older grafts. Thus, we determined the association of HCV infection with short-term graft injury and long-term outcomes in all recipients and in the subgroup that received grafts from older donors (≥60 years). In the total study population (822 patients), the presence of HCV infection did not alter parameters of reperfusion injury, graft function, or clinical short-term outcome (Table 4). When compared to HCV(−) patients, patients with HCV had similar 1-, 3-, and 5-year graft survival (91% versus 90%, 82% versus 85%, and 79% versus 82%, P = 0.9) and patient survival (92% versus 92%, 83% versus 87%, and 80% versus 84%, P = 0.77).
Table 4. Outcomes of Liver Transplantation for Younger HCV(+) Recipients Versus Older HCV(−) Recipients
Abbreviations: ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; INR, international normalized ratio; LOS, length of hospital stay.
However, in the subgroup of recipients receiving an older graft from a donor ≥ 60 years of age, HCV infection status was highly associated with graft survival. In HCV(−) patients, there was a small but significant difference between old and young grafts in 1-, 3-, and 5-year graft survival (87% versus 90%, 79% versus 86%, and 78% versus 82%, P = 0.04) and patient survival (87% versus 92%, 81% versus 88%, and 79% versus 83%, P = 0.045; Fig. 3A). In contrast, HCV(+) patients exhibited more dramatically worsened long-term graft survival when receiving older grafts, with 1-, 3-, and 5-year graft survival of 86% versus 90%, 68% versus 83%, and 64% versus 83% (old versus young, P < 0.009), respectively (Fig. 3B). Similarly, patient survival in the HCV(+) group was significantly reduced in recipients of old grafts versus young grafts at 1, 3, and 5 years (90% versus 91%, 72% versus 84%, and 64% versus 81%, P < 0.05).
Does Recipient Age Affect the Outcome of Hepatitis C Recurrence in Older Grafts?
The previous analysis suggests that in the entire population of HCV(+) recipients, donor age but not recipient age affects graft and patient survival. We next evaluated whether recipient age affects graft and patient survival in the subset of HCV(+) recipients who received an older graft. Twenty-three young HCV(+) patients (<50 years) received an older graft, whereas 45 older HCV(+) patients (≥ 50 years) received an older liver. For HCV(+) patients who received a graft from an older donor (≥60 years), recipient age was associated with increased graft loss (HR: 2.1, CI: 1.06–3.9, P < 0.04) and patient death (HR: 2.3, CI: 1.2–4.8, P < 0.02). Long-term outcomes in HCV(+) patients receiving an older graft were decreased in old recipients versus young recipients, with 1-, 3-, and 5-year graft survival of 87% versus 94%, 63% versus 84%, and 63% versus 84% (P = 0.052) and with 1-, 3-, and 5-year patient survival of 93% versus 95%, 69% versus 84%, and 66% versus 84% (P = 0.053; Figs. 4 and 5).
In order to gain insight into the consequences of HCV recurrence in old grafts, we determined the progression of fibrosis in older grafts (≥60 years) transplanted into young HCV(+) recipients (<50 years, n = 23) versus old HCV(+) recipients (≥50 years, n = 45) with the Metavir score. Two years post-transplantation, the mean Metavir score was 0.6 ± 0.7 in young recipients versus 1.18 ± 1.2 in the old recipient group (P < 0.03; Fig. 6).
This is the first study in the modern era of liver transplantation to investigate the combined effect of donor and recipient age on immediate and long-term outcomes after liver transplantation. Our study demonstrates that advanced donor age is associated with increased hepatocellular injury without affecting short-term outcomes. The use of older donors is associated with decreased long-term graft and patient survival in the overall patient population, which is mainly due to the effect of donor age on HCV recurrence. Importantly, the negative effect of advanced donor age on graft and patient survival can be prevented by the transplantation of older livers into younger recipients, particularly among HCV(+) patients.
The effect of donor age on outcomes following liver transplantation has been examined in the past with conflicting results. Marino et al.20 reported decreased graft and patient survival in recipients of liver transplants from older donors (>60 years of age). This study was performed in the early 1990s, and graft and patient survival rates were inferior to what most centers report today, with 1-year patient and graft survival rates of about 80%. Long-term outcomes after 2 years were not investigated by the authors, and the study did not investigate the effect of recipient age on older grafts. In a smaller series, Nardo et al.21 compared 30 patients receiving grafts from donors 80 years old or older with 60 patients receiving grafts from donors less than 40 years old. The authors found comparable aminotransferase levels after transplantation in the 2 groups with no difference in postoperative morbidity or mortality. Long-term patient and graft survival was reduced in patients receiving grafts from older donors; this effect was most prominent in HCV(+) patients. In this study, older grafts had a significantly shorter CIT, which might explain why no elevation in postoperative aminotransferase levels was seen in the old donor group. Nardo et al. reported that older grafts were more frequently transplanted into older recipients; however, the effect of recipient age on the outcome of transplants involving older donors was not investigated.
Our data argue that grafts from donors older than 60 years of age are more susceptible to TIRI, as indicated by the increased AST, ALT, and ALP post-transplantation. However, the liver has a large functional reserve, and we did not see an increase in short-term complications, 30-day mortality, or LOS. In the study population as a whole, long-term graft and patient survival was clearly affected by donor age but not by recipient age. It has been reported by several groups that older livers have worse outcomes in HCV(+) patients.12–14 In our series, the difference in graft and patient survival was less when the survival analysis was corrected for HCV(+) patients. This finding, when considered with the accelerated fibrosis in older grafts transplanted into HCV(+) recipients, suggests that the effect of donor age on long-term outcomes is probably linked to an accelerated course of recurrent HCV in the older grafts. Interestingly, when we focused our analysis on the effect of recipient age, combining older livers with younger recipients reversed the negative effects of advanced donor age on graft and patient survival: younger HCV(+) patients with older grafts had long-term results identical to those of younger HCV(+) patients receiving younger grafts.
The mechanisms underlying why improved outcomes are seen when older livers are transplanted into younger recipients are unclear, but they may be related to infiltrating recipient nonparenchymal cells. When we consider the effects of age on liver transplantation, it is important to note that the donor hepatocytes and infiltrating recipient nonparenchymal cells originate from patients of potentially quite different ages. Both hepatocytes and nonparenchymal cells are involved in the response of the liver to TIRI, HCV recurrence and progression, and hepatic repair mechanisms such as liver regeneration. After reperfusion, recipient thrombocytes and leukocytes are directly involved in the immediate graft injury, with neutrophils infiltrating the liver parenchyma and platelets adhering to the sinusoidal endothelia cells.22 Donor Kupffer cells, the main source of tumor necrosis factor α production in the liver, are usually replaced within 48 hours by recipient macrophages.23, 24
Nonparenchymal cell signaling is an important factor for hepatocyte injury and could be critical for the response of old hepatocytes to stress. Using a murine model of hepatic ischemia/reperfusion, we recently demonstrated that older livers have increased tumor necrosis factor α release following ischemia-reperfusion injury with subsequently increased activation of intracellular mediators of the death cascade, such as caspase 3 cleavage and cytochrome C release into the cytoplasm.25 The increased reperfusion injury in this mouse model was likely the result of enhanced stimulation of nonparenchymal cells in the old liver, which was associated with an increased inflammatory response after reperfusion. An increased inflammatory response after reperfusion could favor HCV replication: there is more aggressive HCV recurrence in various settings of increased reperfusion injury, such as prolonged ischemia prior to transplantation.26–28 Strategies to prevent reperfusion injury could also be effective in reducing the outcome of HCV recurrence after liver transplantation. Indeed, if the improved survival of older livers in younger recipients is related to differences in nonparenchymal cell function, then strategies targeting nonparenchymal cell activation (eg, Kupffer cell activation, leukocyte rolling and sticking, and platelet activation) after reperfusion could be of particular interest in HCV.
In conclusion, our study suggests that recipient age and HCV status affect patient and graft survival with older livers. Our findings indicate that combining older grafts with younger recipients could be beneficial, preventing negative long-term effects of aging on grafts and patient survival. In particular, pairing an old graft with an older HCV-infected recipient should be avoided whenever possible.