Outcome of Liver Transplantation in Septuagenarians: A Single-Center Experience | Gastrointestinal Surgery | JAMA Surgery | JAMA Network
[Skip to Navigation]
Sign In
Figure 1. 
Age distribution of recipients in group 1.

Age distribution of recipients in group 1.

Figure 2. 
Comparison of survival between groups 1 and 2. A Kaplan-Meier survival curve was calculated for both groups for the duration of the study. Overall survival through 10 years demonstrated no statistically significant difference between group 1 and group 2 (P = .14, log-rank test).

Comparison of survival between groups 1 and 2. A Kaplan-Meier survival curve was calculated for both groups for the duration of the study. Overall survival through 10 years demonstrated no statistically significant difference between group 1 and group 2 (P = .14, log-rank test).

Table 1. 
Causes of Liver Failure
Causes of Liver Failure
Table 2. 
Patient Survival, Unadjusted
Patient Survival, Unadjusted
Table 3. 
Causes of Death in Group 1 Patients
Causes of Death in Group 1 Patients
Table 4. 
Comparison of Donor, Recipient, and Surgical Variables Between Groups
Comparison of Donor, Recipient, and Surgical Variables Between Groups
Table 5. 
Bivariate Mortality After Controlling for Age Group
Bivariate Mortality After Controlling for Age Group
Table 6. 
Multivariate Cox Model for Mortality
Multivariate Cox Model for Mortality
Table 7. 
Cox Model of Adjusted and Unadjusted Survival a
Cox Model of Adjusted and Unadjusted Survival a
August 1, 2007

Outcome of Liver Transplantation in Septuagenarians: A Single-Center Experience

Author Affiliations

Author Affiliations: Dumont-UCLA Transplant Center, Department of Surgery (Drs Lipshutz, Hiatt, Ghobrial, Farmer, Yersiz, and Busuttil and Ms Martinez), and Department of Biomathematics (Dr Gornbein), David Geffen School of Medicine at UCLA, Los Angeles, California.

Arch Surg. 2007;142(8):775-784. doi:10.1001/archsurg.142.8.775

Hypothesis  We hypothesized that selected septuagenarians may do as well after transplantation as those of a younger group of older recipients. This work compares post–liver transplant survival in septuagenarians with that of patients aged 50 to 59 years.

Design  Review of a prospectively maintained database.

Setting  University transplant center.

Patients  First-time liver transplant recipients treated from January 1, 1988, to December 31, 2005. Group 1 consisted of liver transplant recipients aged 70 years or older at the time of transplant. Group 2 was a younger cohort of patients aged 50 to 59 years.

Interventions  Liver transplantation.

Main Outcome Measures  Patient survival. Survival data were stratified, Kaplan-Meier survival was calculated, and a multivariate analysis was performed.

Results  Group 1 included 62 patients aged 70 years or older (average, 71.9 ± 2.1 years). Group 2 included 864 patients aged 50 to 59 years (average, 54.3 ± 2.9 years). Unadjusted patient survival of group 1 at 1, 3, 5, and 10 years was 73.3%, 65.8%, 47.1%, and 39.7%, respectively. Unadjusted patient survival of group 2 at 1, 3, 5, and 10 years was 79.4%, 71.5%, 65.3%, and 45.2%, respectively. The difference was not statistically significant (P = .14). Multivariate analysis for factors affecting survival demonstrated preoperative hospitalization, cold ischemia time, and hepatitis C/ethanol as risk factors for death. Age 70 years or more was not a strong risk factor (mortality ratio, 1.28; P = .27).

Conclusions  When other risk factors for mortality are controlled in older recipients, risk of death due to age is reduced in well-selected recipients. Age by itself should not be used to limit liver transplantation.

Since orthotopic liver transplantation (OLT) was first performed by Starzl et al1 in 1963, advances in case selection, operative techniques, anesthesia, and immunosuppression have significantly expanded the application and success of the procedure. Orthotopic liver transplantation is now a standard therapy for most patients with end-stage liver disease and for patients who have acute liver failure.

While the need for cadaveric organ grafts has increased, the supply of organs has remained static in recent years, with a growing disparity between donor supply and organ demand,2-4 which has catalyzed substantial debate5 regarding current allocation policies. One of the most controversial areas in patient selection and donor allocation is the high-risk patient, including patients of advanced age. Numerous centers consider older patients to be high-risk candidates because of comorbid illnesses, including cardiovascular disease.6

Because many diseases causing liver failure are age specific, and some are increasingly common in patients of advancing age, the number of older transplant candidates is substantial. The median age for primary biliary cirrhosis, a disease in which outcomes after transplantation are generally excellent, is 60 years.7-9 Age of presentation in patients with alcoholic liver disease is greater than 60 years in 28% and greater than 70 years in 6%.10 Greater age at diagnosis is an independent factor associated with progression of hepatocellular carcinoma, a common cause of death among patients with compensated cirrhosis.11,12 Other studies have demonstrated that age may be associated with worse survival outcomes.13,14

As OLT is a potential and sometimes a sole therapy for liver diseases in older patients, it is important to analyze outcomes after transplantation for such patients and to evaluate factors that affect survival. The present study was undertaken to evaluate patient survival among liver transplant recipients 70 years or older at the time of OLT and to compare these results with those of a younger group of older recipients from the same center. We also examined risk factors for poor survival and causes of death among patients of advancing age.

Patients and data collection

This study was approved by the Office for the Protection of Research Subjects at UCLA, Los Angeles, California. Using a regularly updated database of liver transplants, we conducted a review of all primary cases of adult liver transplants and patients from January 1, 1988, through December 31, 2005.

The patients were divided into 2 groups for comparison according to age at time of transplantation: group 1, patients 70 years or older, and group 2, patients aged 50 to 59 years. Patient survival was measured until last known follow-up date or confirmation of death, or until the date of retransplantation. Group 2 patients were time-matched to form a comparison group based on immunosuppressive regimens and other characteristics that were variable over the duration of the study. Multiorgan recipients were excluded.

Donor liver allografts were accepted according to standard criteria. Absolute contraindications to donor acceptance included serologic positivity for human immunodeficiency virus, human lymphotrophic virus, or hepatitis B surface antigen; macrovesicular steatosis of greater than 40% on liver biopsy results; and presence of extracranial malignancies.

All prospective OLT recipients received cardiovascular evaluation including history and physical examination, electrocardiogram, surface echocardiography, and nuclear medicine stress imaging. Patients who demonstrated echocardiographic evidence of pulmonary hypertension underwent right-sided cardiac catheterization to measure pulmonary artery pressures. Left-sided heart catheterization was performed in selected patients with abnormalities on nuclear medicine imaging or risk factors for left ventricular and/or coronary artery disease such as long-standing diabetes mellitus.


Maintenance immunosuppression regimens consisted of a triple cyclosporine-based drug regimen including azathioprine (Imuran; GlaxoSmithKline, Triangle Park, North Carolina) and prednisone from 1988 to 1991, or dual immunosuppression that used tacrolimus (Prograf; Astellas, Deerfield, Illinois) and prednisone. In 1996, another cyclosporine preparation (Neoral; Novartis, Basel, Switzerland) was routinely substituted for the cyclosporine preparation used initially (Sandimmune; Novartis). Routine use of tacrolimus was initiated at UCLA in 1994. Supplemental immunosuppression, when required, has included mycophenolate mofetil (CellCept; Hoffman-LaRoche, Inc, Nutley, New Jersey). On the day of transplantation, a rapid corticosteroid taper was begun from 1 g of methylprednisolone sodium succinate to 20 mg/d over 1 week. Oral prednisone was started on day 8 and tapered within 2 months to 5 mg/d. Beginning in 1995, corticosteroids were discontinued at 3 to 6 months in patients with hepatitis C virus who did not exhibit rejection episodes.

Statistical analyses

Kaplan-Meier curves and corresponding log-rank tests were computed to estimate and compare patient survival by age group. Wilcoxon rank sum tests and χ2 tests were used to compare other continuous or discrete covariates, respectively, between the 2 age groups. The log-rank test was also used to compute P values for the bivariate effect on patient survival of each covariate after controlling for age group. In this bivariate analysis, continuous covariates were dichotomized at their median. A stepdown backward multivariate Cox model was used to assess the simultaneous effect of age group and all covariates on patient survival. Interactions between important covariates (model for end-stage liver disease [MELD] score) and age group were also assessed in this Cox model. The simultaneous adjusted P values for 1, 3, 5, and 10 years were computed according to the methods of Klein and Moeschberger15 for multiple adjustment. Computations were carried out with SPSS version 14 (SPSS Inc, Chicago, Illinois) and JMP version 5 (SAS Institute Inc, Cary, North Carolina) statistical software.

Recipient characteristics

During the 18-year period of the study, 62 patients in group 1 (aged 70 years or older) and 864 patients in group 2 (aged 50 to 59 years) underwent OLT. The mean ± SD age was 71.9 ± 2.1 years (range, 70-78 years) in group 1 (Figure 1) and 54.3 ± 2.9 years (range, 50-59 years) in group 2 (P<.001). The male to female ratio was 1.2:1 in group 1 and 1.6:1 in group 2 (P = .32). In group 1, no transplants to as many as 7 transplants were performed per year.

Causes of liver failure are compared in Table 1. The most common cause of liver failure in group 1 was hepatitis C (48%), followed by cryptogenic cirrhosis (18%). In group 2, hepatitis C (36.5%) was most common, followed by Laënnec cirrhosis (15.3%).

Survival and cause of death

Short- and long-term survival was compared by age group (Table 2 and Figure 2). In group 1, the death rate per 1000 follow-up months was 11.4, with a median follow-up time of 31.8 months (range, 0.2-183.9 months). In group 2, the death rate per 1000 follow-up months was 8.4, with a median follow-up of 29.5 months (range, 0.0-207.8 months). Survival at 1, 3, 5, and 10 years was 73.3%, 65.8%, 47.1%, and 39.7%, respectively, in group 1 and 79.4%, 71.5%, 65.3%, and 45.2%, respectively, in group 2. There was no statistical difference in survival at 1, 3, 5, and 10 years (P = .49, P = .54, P = .10, and P = .65, respectively) or overall survival (P = .14). There was no statistical difference in death rate per 1000 follow-up months (P = .15) or median follow-up (P = .15).

Of 31 deaths in group 1, 16 occurred during the first year after transplant and 15 occurred later (Table 3). Early deaths were primarily due to infection (sepsis, multisystem organ failure, or pneumonia), followed by cardiac causes. Later deaths were caused more commonly by cardiac or vascular disease and malignant neoplasms.

Bivariate analysis of group variables

Donor characteristics are compared in Table 4. There were no statistically significant differences in quality of donor organs between the groups when donor age, cold ischemia time, length of hospital stay, serum sodium level, use of vasopressors, and the requirement for cardiopulmonary resuscitation were analyzed. Group 1 patients received whole deceased donor livers in 57 cases and in situ split deceased donor grafts in 5. No recipients in this group had living donor candidates. Group 2 patients received whole deceased donor grafts (n = 837), split deceased donor grafts (n = 10), and living donor transplants (n = 17). Graft type had no effect on survival.

Recipient and operative variables were similarly compared (Table 4). The 8 recipient variables that were considered included age, bilirubin level, creatinine level, sex, prothrombin time, MELD score, preoperative hospitalization, and surgical period of transplantation. The single operative variable was warm ischemia time. These variables have been previously demonstrated to be important to posttransplant survival.13

By bivariate comparison, there were no differences between the 2 groups for 4 of the recipient variables and the single operative variable. Differences were found for 3 of the 7 recipient variables, including age (P<.001), preoperative bilirubin level (P = .006), and average MELD score (17.9 in group 1 and 21.2 in group 2; P = .02). In addition, a larger percentage of patients underwent OLT after 1990.

Bivariate survival analysis after controlling for age

The impact of each potential covariate was assessed on patient survival one at a time after controlling for age (Table 5). Prolongation of cold ischemia time, preoperative bilirubin level, preoperative creatinine level, MELD score, preoperative hospitalization, and the combination of hepatitis C and Laënnec cirrhosis as the cause of liver failure were statistically significant negative predictors of posttransplant survival for both groups

Multivariate analysis of survival

Of the 26 variables or causes of liver failure considered for adult patient mortality, 4 were simultaneously statistically significant by Cox multivariate regression analysis (Table 6). The impact of hepatitis C and Laënnec cirrhosis had the greatest effect, with a relative risk of 2.52, followed by pretransplant recipient hospitalization (relative risk, 1.54), MELD score (relative risk, 1.014 per point), and a borderline effect of increasing cold ischemia time (relative risk, 1.03 per hour). The effect of age group (age ≥70 years) was in the expected direction of increased risk (relative risk, 1.28) but was not statistically significant by the P<.05 criteria (P = .27). We also controlled for surgical period because surgical methods and advances in critical care have occurred over time. As expected, mortality decreased in more recent surgical periods (compared with the reference period of 1988-1995), although the decrease was not statistically significant when all other variables were controlled for. We also considered an interaction of MELD score by age group and determined that this was not statistically significant (P = .40).

Hospitalization and retransplantation

Mean length of posttransplant hospitalization was 29.8 days (range, 9-162 days) in group 1 and 27.4 days (range, 1-536 days) in group 2 (P = .14). Retransplantation was performed in 4 patients in group 1. Of the 4 patients, 1 survived for more than 3 years. The others died within 2 months of retransplantation, 2 of cardiac causes and 1 of intracranial bleeding. Retransplantation was performed in 109 patients in group 2.


Life expectancy has increased in recent decades, with persons older than 70 years representing an important segment of the general population. From 1950 to 2005 in the United States, the growth rate for people aged 65 to 75 years was double the overall annual growth rate; for the group older than 75 years, the growth rate was nearly triple.16 With appropriate medical care and nutrition, a healthy 70-year-old adult with minimal comorbidities living in a developed country may expect an additional 10 to 20 years of life.

As longevity has increased, the burden of liver disease in patients of advancing age has also increased and is associated with a higher mortality than in younger adults.3 In the 1980s, the death rate from chronic liver disease was highest in patients 65 to 74 years of age.17 This has led to more older patients undergoing liver transplantation.

The supply and demand disparity in organ transplantation has led to controversy regarding organ allocation and a more critical examination of the process. The debate has focused on the competing variables of justice and medical utility.3 Some authors have argued that liver transplantation should be performed in the young and less ill, citing better survival rates and decreased financial costs,18,19 while others believe that factors related to efficiency dictate that organs should be allocated to patients at a time that is most beneficial to them, not addressing who should receive an organ.3 While age is not part of the MELD score, it has also entered into this debate and raises the issue of outcome after transplantation in older patients.20

It has been previously demonstrated that the youngest of all adult patients tend to survive the longest.13,14 A review of the entire experience at our center demonstrated that survival was longest in those aged 1 to 18 years, followed by those aged 18 to 55 years. Those older than 55 years have the worst survival when these groups are compared. However, a close evaluation of these older groups and a determination of factors affecting their outcome are essential as more elderly patients proceed to transplantation.

The issue of liver transplantation in septuagenarians has rarely been addressed. However, transplantation in patients in their 60s has been studied in several reports from other centers,4,21-28 of which all have not demonstrated favorable outcomes. Survival at 3 years after transplantation has ranged from as low as 35%4 to as high as 83%.24 At the present time, there is no established chronological age limit for organ transplantation, and individual centers set their own rules based on the concept of “physiologic age,” which is not completely defined.

Herein we report the largest series to date of liver transplantation in recipients 70 years or older and the finding that age as a sole criterion does not affect survival when compared with a cohort of patients in their sixth decade of life. We found no statistically significant difference in survival in the first 10 years after transplantation for a group of 62 patients 70 years or older when compared with a younger cohort of 864 recipients aged 50 to 59 years with similar characteristics. The longest-surviving patient was 88 years old at 15 years after transplantation. One-year unadjusted survival of septuagenarians in the most recent surgical period, 2001 to 2005, was 94.4%.

Recipient, donor, and operative variables were studied for their effect on patient survival after primary transplantation in adult recipients; such variables have been used in our previous studies of recipient outcome.13 We previously found that characteristics such as donor age, need for retransplantation, and extended cold and warm ischemia times were significant independent risk factors for mortality.13 Of the 26 variables analyzed as predictors of adult patient mortality, 4 (preoperative hospitalization, prolonged cold ischemia time, increasing MELD score, and cirrhosis due to hepatitis C and alcohol) were simultaneously significant by Cox multivariate regression analysis, while age 70 years or older was not. We tested for an age × MELD interaction in our model and found that this was not statistically significant (P = .40). This lack of an interaction implies that the effect of MELD on mortality is about the same for both age groups.

This study was limited by the small number of recipients 70 years or older, but it does confirm the results of previous studies by our group13 and others29 that have demonstrated the increase in risk of death with pretransplant hospitalization and cold ischemia beyond 10 hours. This study did not demonstrate multiple donor organ variables that increase the risk of mortality. Donor variables in both groups indicated that the groups were well matched and therefore may not have affected survival. Cold ischemia times on average were less than 8 hours in both groups, and warm ischemia time, an operative variable, was less than 45 minutes. In our previous study, warm ischemia time beyond 55 minutes more than doubled the risk of mortality, and a cold ischemia time greater than 10 hours increased the risk nearly 1.5 times.13

We found that, while not statistically significant, survival of both groups began to diminish after 3 years, and this effect was more pronounced in the older group (P = .14 for overall survival). Data from this study demonstrated an unadjusted 5- and 10-year survival in septuagenarians of 47.1% and 39.7%, compared with 65.3% and 45.2% in the younger group (P = .10 and P = .65, respectively). Data from the European Liver Transplant Registry demonstrated a similar finding for patients in the seventh decade of life. While 5-year survival rates were similar for patients aged 45 to 60 years and patients older than 60 years (68% vs 63%, respectively), survival at 10 years was significantly lower (58% vs 48%; P < .001).30 However, we used our multivariate Cox model to adjust for differences in survival between groups in our study. After adjustment we saw that the differences in survival were less at 1 year (4.3% vs 6.0% unadjusted) and 5 years (7.3% vs 18.7% unadjusted) between the septuagenarians and the 50- to 59-year-old group. These differences remained not statistically significant (P = .14 for unadjusted, P = .27 for adjusted) (Table 7).

Advancing age is a risk factor for the development of coronary artery disease, a significant cause of death for elderly patients. Mortality of patients with coronary artery disease who undergo liver transplantation may be as high as 50%.31 Even with rigorous pretransplant cardiac evaluation, including stress testing, 25% of deaths in the first posttransplant year were due to cardiac causes; infection and sepsis, however, remained the most important early causes.

Death after the first year was generally due to chronic conditions including cardiovascular disease and recurrence of malignant neoplasm. Given that greater age is a strong risk factor for coronary heart disease, death from a cardiac or vascular cause may be preventable with more aggressive pretransplant screening and postoperative risk factor modification. Our present policy is to perform cardiac catheterization in all patients of advanced age and other significant cardiovascular risk factors to determine candidacy.

However, the groups were not identical: the MELD score and creatinine level were lower and there was less preoperative hospitalization in group 1 patients than in those in group 2. These data suggest that these patients should have transplants at lower MELD scores than younger cohorts. Patients with a combination of hepatitis C and alcoholic liver disease and those who were hospitalized preoperatively likely have poorer survival due to debility and colonization or infection, respectively. Therefore, to have good survival outcomes in septuagenarians, it is important that standard criteria livers be used to decrease the likelihood of need for early retransplantation. In addition, short warm and cold ischemia times are essential. The MELD scores should be lower and patients should not be in the hospital secondary to end-stage liver disease at the time of transplantation. It may be prudent to perform cardiac catheterization in all candidates 70 years or older regardless of stress test results. Moreover, patients should be screened for hypertension, hyperlipidemia, and diabetes mellitus.32 Finally, patients who have liver failure from the combination of hepatitis C virus and Laënnec cirrhosis should likely undergo transplantation with caution.

In conclusion, biological and physiological variables may play a more important role than advanced age in predicting poor survival after liver transplantation. Measures of physiological age and risk of complications should be used in the evaluation process of elderly transplant candidates. Chronological age by itself is not a sole predictor of outcome.

Correspondence: Ronald W. Busuttil, MD, PhD, Department of Surgery, David Geffen School of Medicine at UCLA, Room 77-120 CHS, Los Angeles, CA 90095 (Rbusuttil@mednet.ucla.edu).

Accepted for Publication: April 6, 2007.

Author Contributions:Study concept and design: Lipshutz, Ghobrial, Yersiz, and Busuttil. Acquisition of data: Lipshutz and Martinez. Analysis and interpretation of data: Lipshutz, Hiatt, Ghobrial, Farmer, Gornbein, and Busuttil. Drafting of the manuscript: Lipshutz, Hiatt, and Martinez. Critical revision of the manuscript for important intellectual content: Lipshutz, Hiatt, Ghobrial, Farmer, Yersiz, Gornbein, and Busuttil. Statistical analysis: Gornbein. Obtained funding: Lipshutz. Administrative, technical, and material support: Ghobrial, Martinez, Yersiz, and Busuttil. Study supervision: Lipshutz, Ghobrial, Farmer, and Busuttil.

Financial Disclosure: None reported.

Previous Presentation: This paper was presented at the 78th Annual Meeting of the Pacific Coast Surgical Association; February 18, 2007; Kohala Coast, Hawaii; and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.

Starzl  TEMarchiore  TLVonkaulla  KNHermann  GBrittain  RSWaddell  WR Homotransplantaion of the liver in humans.  Surg Gynecol Obstet 1963;117659- 676PubMedGoogle Scholar
Marino  IR Are we ready to match donor and recipient in liver transplantation?  Liver Transpl 2006;12 (11) 1574- 1576PubMedGoogle ScholarCrossref
Keswani  RNAhmed  AKeeffe  EB Older age and liver transplantation: a review.  Liver Transpl 2004;10 (8) 957- 967PubMedGoogle ScholarCrossref
Collins  BHPirsch  JDBecker  YT  et al.  Long-term results of liver transplantation in older patients 60 years of age and older.  Transplantation 2000;70 (5) 780- 783PubMedGoogle ScholarCrossref
Lopez  PMMartin  P Update on liver transplantation: indications, organ allocation, and long-term care.  Mt Sinai J Med 2006;73 (8) 1056- 1066PubMedGoogle Scholar
Hoshida  YIkeda  KKobayashi  M  et al.  Chronic liver disease in the extremely elderly of 80 years or more: clinical characteristics, prognosis and patient survival analysis.  J Hepatol 1999;31 (5) 860- 866PubMedGoogle ScholarCrossref
Neuberger  J Liver transplantation for primary biliary cirrhosis.  Autoimmun Rev 2003;2 (1) 1- 7PubMedGoogle ScholarCrossref
Almdal  TPSorensen  TIDanish Association for the Study of the Liver, Incidence of parenchymal liver diseases in Denmark, 1981 to 1985: analysis of hospitalization registry data.  Hepatology 1991;13 (4) 650- 655PubMedGoogle ScholarCrossref
Metcalf  JJames  O The geoepidemiology of primary biliary cirrhosis.  Semin Liver Dis 1997;17 (1) 13- 22PubMedGoogle ScholarCrossref
Potter  JFJames  OF Clinical features and prognosis of alcoholic liver disease in respect of advancing age.  Gerontology 1987;33 (6) 380- 387PubMedGoogle ScholarCrossref
Fattovich  GStroffolini  TZagni  IDonato  F Hepatocellular carcinoma in cirrhosis: incidence and risk factors.  Gastroenterology 2004;127 (5) ((suppl 1)) S35- S50PubMedGoogle ScholarCrossref
Kato  YNakata  KOmagari  K  et al.  Risk of hepatocellular carcinoma in patients with cirrhosis in Japan: analysis of infectious hepatitis viruses.  Cancer 1994;74 (8) 2234- 2238PubMedGoogle ScholarCrossref
Busuttil  RWFarmer  DGYersiz  H  et al.  Analysis of long-term outcomes of 3200 liver transplantations over two decades: a single-center experience.  Ann Surg 2005;241 (6) 905- 916PubMedGoogle ScholarCrossref
Jain  AReyes  JKashyap  R  et al.  Long-term survival after liver transplantation in 4,000 consecutive patients at a single center.  Ann Surg 2000;232 (4) 490- 500PubMedGoogle ScholarCrossref
Klein  JPMoeschberger  ML Survival Analysis: Techniques for Censored and Truncated Data.  New York, NY Springer Science+Business Media Inc1997;
 Chartbook on Trends in the Health of Americans.  Hyattsville, MD National Center for Health Statistics2005;40127
Centers for Disease Control and Prevention (CDC)., Deaths and hospitalizations from chronic liver disease and cirrhosis-United States, 1980-1989.  MMWR Morb Mortal Wkly Rep 1993;41(52-53)969- 973PubMedGoogle Scholar
Keeffe  EB Liver transplantation: current status and novel approaches to liver replacement.  Gastroenterology 2001;120 (3) 749- 762PubMedGoogle ScholarCrossref
Yu  ASAhmed  AKeeffe  EB Liver transplantation: evolving patient selection criteria.  Can J Gastroenterol 2001;15 (11) 729- 738PubMedGoogle Scholar
Zarembo  A With kidney transplants, a question of how to ration life.  Los Angeles Times. November5 2006;1Google Scholar
Bromley  PNHilmi  ITan  KCWilliams  RPotter  D Orthotopic liver transplantation in patients over 60 years old.  Transplantation 1994;58 (7) 800- 803PubMedGoogle ScholarCrossref
de la Peña  AHerrero  JISangro  B  et al.  Liver transplantation in cirrhotic patients over 60 years of age.  Rev Esp Enferm Dig 1998;90 (1) 3- 14PubMedGoogle Scholar
Emre  SMor  ESchwartz  ME  et al.  Liver transplantation in patients beyond age 60.  Transplant Proc 1993;25 (1, pt 2) 1075- 1076PubMedGoogle Scholar
Filipponi  FRoncella  MBoggi  U  et al.  Liver transplantation in recipients over 60.  Transplant Proc 2001;33 (1-2) 1465- 1466PubMedGoogle ScholarCrossref
Garcia  CEGarcia  RFMayer  ADNeuberger  J Liver transplantation in patients over sixty years of age.  Transplantation 2001;72 (4) 679- 684PubMedGoogle ScholarCrossref
Pirsch  JDKalayoglu  MD’Alessandro  AM  et al.  Orthotopic liver transplantation in patients 60 years of age and older.  Transplantation 1991;51 (2) 431- 433PubMedGoogle ScholarCrossref
Stieber  ACGordon  RDTodo  S  et al.  Liver transplantation in patients over sixty years of age.  Transplantation 1991;51 (1) 271- 273PubMedGoogle ScholarCrossref
Zetterman  RKBelle  SHHoofnagle  JH  et al.  Age and liver transplantation: a report of the Liver Transplantation Database.  Transplantation 1998;66 (4) 500- 506PubMedGoogle ScholarCrossref
Ekka-Zohar  AZitser-Gurevich  YMandel  M  et al.  Graft survival and its determinants: a 3 year national experience with liver transplantation in Israel.  Isr Med Assoc J 2006;8 (6) 400- 405PubMedGoogle Scholar
Adam  RBismuth  HKaram  VDelvart  VCochet  E European Liver Transplant Registry: A Service of the European Liver Transplant Association: Data Analysis Booklet: 05/1968–06/2004.  Villejuif, France European Liver Transplant Registry2004;
Plotkin  JSScott  VLPinna  ADobsch  BPDe Wolf  AMKang  Y Morbidity and mortality in patients with coronary artery disease undergoing orthotopic liver transplantation.  Liver Transpl Surg 1996;2 (6) 426- 430PubMedGoogle ScholarCrossref
Sheiner  PAMagliocca  JFBodian  CA  et al.  Long-term medical complications in patients surviving ≥5 years after liver transplant.  Transplantation 2000;69 (5) 781- 789PubMedGoogle ScholarCrossref

Chris E. Freise, MD, San Francisco, California: This study by Lipshutz, Ghobrial, Busuttil, and colleagues describes their group's experience with liver transplantation in a cohort of patients age 70 and greater, comparing outcomes to a cohort age 50 to 59. They state their purpose very clearly: septuagenarians are a higher-risk group in terms of survival, but as our population ages, there is an increasing demand for transplant in patients with advancing age. Not too many years ago, transplant in patients even over the age of 60 was undertaken cautiously. With improvements in management, liver transplant in sexagenarians has been quite successful and is no longer as controversial. This paper attempts to extend the age limit even further.

Data available to date are from a few small center-specific reports and from the national UNOS [United Network for Organ Sharing] database. Five-year patient survival from the UNOS database is 62% for patients 65 and greater, and 70.3% for patients 50 to 64. While the age cut points are a bit different than in this UCLA report, these data show that there is a clear decrease in survival with advancing age.

The current paper found a similar trend, with a 5-year survival in septuagenarians of 47.1% compared with a 65.3% survival in the younger cohort. While this was not statistically significant, it raises the question that we are often faced with as the stewards of a limited resource—what level of survival is good enough? I do not expect the presenters to answer this question; I raise it only as food for thought. With nearly 17 000 patients on the liver waiting list, and only a stagnant 6500 livers available each year, careful thought must go into the use of this valuable resource. If we left these decisions purely to the statisticians and survival-modeling experts, we probably would not pursue transplant in higher-risk groups, whether it be age related, cancer related or comorbidity related. Conversely, if we left the decisions to the family of the loved one dying with liver disease, 5-year results would likely suffer. Therefore, as transplant professionals, we have to take all the available evidence into account and make these decisions using our best judgment for an individual patient. I sometimes find this to be the most challenging part of our profession.

Of course, this problem of resource allocation is not unique to liver transplantation, and in some ways is even more of a problem with kidney transplantation, where the waiting list is even larger, and octogenarians are often considered candidates. A recent change in the allocation system has been proposed, using the concept of net benefit. This system essentially directs organs to the patients who will achieve the longest survival, or net benefit. It is an attempt to elevate utility over justice, and when it was presented last week to the public, it was not particularly well received. It would appear that the public has a strong sense of entitlement to receive the best care possible, and the concept of an organ allocation system driven primarily by utility is unacceptable. These issues will certainly require a great amount of thought as the gap between supply and demand continues to widen.

So what of the septuagenarian in need of liver transplant? As we all know, using age alone may overestimate or underestimate physiology. The current report also attempts to define factors that may allow for a successful transplant in this population. Clearly this cohort does better when transplanted in a healthier state, with good-quality livers. This raises my first question. In the current allocation system, where the sickest patients get the first opportunity for livers, where will the less sick septuagenarian find a liver? I also noted that there were no living donor transplants in this cohort, compared with 17 in the younger cohort. Living donors represent a potential method to transplant patients who are less ill, and I was wondering about the current thinking or practice of living donor transplantation in septuagenarians at UCLA.

Death due to sepsis or cardiovascular disease would explain half of the deaths in the first year. You mentioned in your manuscript that all patients now undergo cardiac catheterization. What level of disease is acceptable—ie, is a stent okay, or do coronaries have to be free of significant disease? In terms of sepsis, do you alter your immunosuppression strategy in patients with advancing age? And lastly, do you have a sense of the number of septuagenarians who are felt to be an unacceptable risk for transplant and are turned down at your center? It appears that careful patient selection is crucial for success in this area.

I congratulate the UCLA group for pursuing transplant in this high-risk population, and I would hope the knowledge gained from this report will help all transplant centers wisely use our scarce resources.

Dr Ghobrial: This was a tremendous comment by Dr Freise talking about how we select the organs for our recipients in times of crunch when there aren't enough organs to go around. The biggest issue on our minds in this paper was how we could justify transplanting older people if the data suggest that they do not live long enough. I think one of the findings of the study is that older recipients tend to do well if all other factors were controlled, for instance, if they are transplanted at low MELD scores with reasonably young organs.

The first question is, how are we able to deliver those organs to those recipients because their MELD scores are not very high? That is a tough question. We have to say that we have not been very successful in doing so. Just looking at our paper, there have been about 62 patients transplanted over the course of 18 years. That is about 3 or 4 recipients older than 70 years every year. That is not very many, considering that we do about 200 to 250 transplants per year. The means that we go about to accomplish this include using split livers, writing exception letters to the board, and getting upgrades by the presence of cancer. About 10% to 15% of these patients did have cancer. So the answer is that getting organs to older recipients with low MELD scores is very difficult.

The other question is whether we use living donors for recipients older than 70. We have not done it yet at UCLA, but I do not see why a well-selected patient, free of cardiac disease and in good physiological status, couldn't become a candidate for a living donor, assuming that we balance the risks of a living donor for the donor vs the benefits to the recipient.

The third question is regarding death because of cardiovascular disease. What is an acceptable cardiac risk for us? We have gone through the years from doing the dobutamine stress test to performing cardiac catheterization for anybody who is above 70 years of age, and we are even moving lower, looking at populations who are 65 to 70 years of age. We have chosen cardiac catheterization to be our gold standard for cardiac evaluations in this patient population.

What level of cardiac disease do we accept? Data suggest that if any patients require coronary artery bypass grafts, they do not do well in the long-term. They have approximately 50% mortality at the time of the transplant. So patients who require cardiac bypass operations are not considered as candidates. However, if there is a defined small lesion that is stented in the absence of other complicating factors, such as hepatopulmonary syndrome in an older patient, we may accept his or her candidacy. Otherwise, we are starting to use more and more strict criteria that the coronary arteries have to be very clean.

Sepsis was another common cause of death in the postoperative period. Do we modify our immunosuppression for this particular population of patients? The answer is no. But in general, we are using less steroids and we are tending to drop the tacrolimus to lower levels; at about 3 months we are running them at about 6 to 8 [ng/mL].

The fourth question addresses who are the candidates whom we decline. If someone is not physiologically fit, he or she will be declined; at the end of the day it is all about selection, selection of the recipient and selection of the donor. Any candidate with cardiac or pulmonary issues will be declined. I do not have a good sense of how many patients we have declined. Probably a lot of patients have been presented to us, and we have declined some. My guess is we probably list 1 of every 3 candidates over 70 years of age.

Ryutaro Hirose, MD, San Francisco: I just wanted to reiterate a point and ask a question based on your last comment that liver transplantation, just like the NBA, is all about shot selection. But sometimes we shouldn't maybe be allowed to take some shots. One question is that a 70-year-old really doesn't need 30 years of liver function. Do you think we should, as transplant professionals, even be allowed to take a 20-year-old liver for a 73-year-old recipient when that liver might, more utilitarianly speaking, be useful in a 35-year-old recipient? Obviously, you are going to get better results with the 70-year-old if you take a nice young donor, but is that something that we ought to be doing?

Dr Ghobrial: I am going to try to set aside my own biases when I am answering that question because this question has been answered to us by society. When we were trying to put the MELD scores together, when we were trying to predict as a nation how we were going to assess the risk of death because of liver disease, we as a nation chose to use the MELD score. The MELD score includes the variables that influence death the most. However, age was one highly significant variable to predict death from liver disease. As a nation, we chose to exclude age from the MELD formula so we would not discriminate against older recipients.

Similarly, a practicing physician needs to comply with the will of the country. So the decision making of accepting an organ often has to be blinded to the age of the recipient. So if I get a 20-year-old liver and it is offered to a 70-year-old recipient, I will use it. Along the same line of thought, I can ask similar questions such as, why do we prioritize cancer patients for transplantation? Why don't we instead prioritize patients with primary biliary cirrhosis? Why do we transplant patients with hepatitis C who may reject fast? There are a lot of ethical obligations regarding why and how we do things in liver transplantation. We must abide by the will of our society rather than our own biases.

Dr Hirose: Right, and we can change that with policy as opposed to individual decisions when you are on call. One other point I wanted to push a little bit is in terms of your immunosuppression. Older folks probably have a lower rate of rejection, so why not consider modifying your immunosuppressive regimen for the older folks since indeed your cause of death was infection, and obviously cardiovascular effects of immunosuppression on lipid profiles, diabetes, and hypertension could really be adversely affecting your outcomes in this particular population?

Dr Ghobrial: I totally agree and more recently we have been dropping immunosuppression levels more rapidly once the liver functions are okay. The other issue is the sepsis and immunosuppression. The comment I have is that immunosuppression may not be the only cause of sepsis, especially in the initial period after transplantation. Other factors such as premorbid conditions, prehospitalizations, and renal failures and so forth must be addressed as well. However, you bring up an excellent point.

Lawrence D. Wagman, MD, Duarte, California: It is really fascinating to see the development of looking at this population, getting away from age as a bias variable and looking at it more as a pure variable in outcome. It seems to me, though, that there is a lot of age-related change in activity that occurs. We are all interested not just in mortality, the 2 survival curves, but we are very interested in morbidity. I think a lot of what surgery does as it advances is look at its morbidity. In an older population, I also wonder, not in terms of going back to work and how many productive days you spend at your desk for the rest of your life, but the more subtle features of the quality of life of the older patient who receives a liver transplant. Do they get out of their wheelchair and walk around again? Do they go back to playing tennis? Are they able to achieve the lifetime goals that patients set for themselves, like getting to somebody's graduation or wedding? This is a key feature in both making an argument for treating this population and also making an argument for who is not to be treated, patients whom you think will not get to those goals.

Dr Ghobrial: These are all excellent comments. Your question revolves around the quality of life of liver transplantations. It has been shown across the board that liver transplantation tremendously improves the quality of life of any recipient when compared with pretransplant status. The issue here is, can the quality of life itself be chosen as a selection for the recipient when they are in the older age group? That is something that is usually taken into consideration, but it is not a decisive factor because you can never tell how well that patient is going to do. However, quality-of-life studies, particularly in the older age group, are important and should be done.

Adil Haider, MD, Baltimore, Maryland: The previous question provides a good segue to mine. A colleague of ours, Martin Makary, has developed an objective assessment tool known as the “Frailty Score,” which looks at a patient's physical fitness attributes instead of the patient's age and has shown it to predict outcomes well in complex hepatobiliary surgery, like the Whipple procedure. Have you considered using this score, or would you?

Dr Ghobrial: We could, but liver transplantation is a very special procedure because it's dependent not only on recipient factors for survival but on other factors such as donor variables. In this study that we presented today, the donor variables didn't impact the survival very much because the donors were very uniform.

There are also other operative factors such as cold ischemia time and warm ischemia time, and these can affect the outcome of the operation. So it's not only a recipient-dependent procedure but an interaction between the surgeon who does the operation, the donor, and the recipient variables. That is why most of the models that are trying to predict outcomes of liver transplantation have to consider 3 sets of issues at the same time.

Carlos O. Esquivel, MD, PhD, Stanford, California: My question relates to the design of the study. Why did you pick that age bracket, 50 to 59, as a control group and not 30 to 40 or 20 to 30? How can you convince us that there was no bias in selecting that age group? Maybe you picked that age group because that was the only group where you found no differences between the control and the septuagenarian groups?

Dr Ghobrial: I was waiting for that question, actually. First, we do not want to mislead anybody about saying that an older patient would survive as well or better than one in a younger age group. We have looked at over 3500 patients who were transplanted at UCLA over the last 2 decades, and we have clearly shown that the populations under 55 survive better than the recipients above 55, and that's why we chose this cutoff. That was documented in a previous paper that was published.

In this paper we chose to focus the attention on the group of patients above 55, and we chose the 2 extremes in order to highlight differences and tried to focus on the variables that can really impact the survival of the older population, above 70 years of age.

Susan L. Orloff, MD, Portland, Oregon: This is a very nice paper and an important area of research. I have 2 questions. First, I am still not clear how, logistically, we are going to be able to get these septuagenarians (and older patients) with low MELD scores transplanted in a timely fashion, and hence have better outcomes in these patients, when you have a significant number of patients on the list who have MELD scores over 17 and will receive a liver allograft first. How are you going to be able to get these older patients transplanted with the current allocation system?

Second, how do you achieve such short cold ischemia times? The average cold ischemia time across the board in the nation is about 10 hours; yours was closer to 7. I commend you on getting the organs to your patients and getting them transplanted so quickly.

Dr Ghobrial: These are very insightful comments. The first issue is that in this paper we have pinned down that the lower the MELD score, the better the outcomes are. So these data may be used as a justification for us and for others to obtain a MELD exception for older patients so they can get transplanted at a lower MELD score.

How are we doing it? How can we transplant lower-MELD-score older patients in this allocation system? It may be difficult. Remember that a lot of those patients were treated before the MELD era, and that is a calculated MELD score. We may have to use split liver or B blood groups and so forth. If you consider only 3 to 4 patients being done who are older than 70 with a MELD score of 18 in a 250-patient population a year, this tells you that it is not easy to be able to do this. We agree. Hopefully the allocation will change in the future to allow early transplantation of older patients.

How we keep a short cold ischemia time: we put the patient in the operating room when the donor is being visualized in the donor hospital. I think one of the factors that really help us to get better outcomes is the short cold ischemia times. We have even tried to cut cold ischemia time down to about 6 hours and cut down our warm ischemia times to less than 30 minutes. In this way we are able to control and maybe use extended donor criteria successfully.

Linda Sher, MD, Los Angeles, California: I would like to congratulate you, Mark, because in this day when our programs are under close scrutiny resulting in the need to transplant lower-risk recipients, it's nice to know that as we march through life, all of us can rest assured that at 70 we may still be eligible for a liver. I have 3 questions. First, given our particular region and the difficulty that we have compared to other regions of getting our low-MELD-score patients transplanted, do you recommend to these patients that they get listed in other regions where they are likely to get a liver much faster at a lower MELD score and when they are in better medical condition?

The second question is, do you do combined liver-kidney transplants in patients over 70? The third question is, do you have a cutoff and inform these patients that, should they reach a certain severity of illness, they will be removed from the list?

Dr Ghobrial: Again, these are very important issues. Do we recommend to patients to go to other centers? No. I believe that UCLA is one of the best centers. However, we do inform the patients that, in different parts of the country, they can get transplanted at lower MELD scores and patients are informed at the time of evaluation. If they do decide to go elsewhere, we help them. Most of the time we find that the patients prefer to stay close to home. It is very hard for a patient from California to get transplanted in Alabama. Also, they may not be able to get provider coverage. So, yes, we do inform them. If they want to go there, we help them to go there. But we find that the majority stay with us.

Regarding liver-kidney transplants, in this patient population we have not done a combined liver-kidney transplant. I don't know how old the oldest patient is in whom we have done a combined liver-kidney transplant. Although liver-kidney transplants have increased in number, they are still a small portion of the transplants performed. Our kidney transplant nephrologists may not be enthusiastic to do a combined liver-kidney transplant in a 78-year-old, but it is something that we would strongly consider if the right recipient came up for it.

Regarding the cutoff for delisting our patients, we feel a moral obligation to a recipient who has been on the list with us for a long time; we believe that patients should not be abandoned when their condition deteriorates on the list. Unless the patient really reaches a state where we cannot turn him around, we will do the transplant. In other words, intubation and dialysis are not contraindications at UCLA. We will still transplant the recipient if we feel he has good hemodynamics and is free of sepsis. So our criteria for getting somebody off the list are actually very stringent. Once we put them on the list, we go for them all the way.

Financial Disclosure: None reported.