Trekking new ground: Overcoming medical and social impediments for extended criteria liver transplant recipients


  • Geoffrey W. McCaughan

    Corresponding author
    1. Australian National Liver Transplant Unit, A. W. Morrow Gastroenterology and Liver Center, Centenary Research Institute, Royal Prince Alfred Hospital, University of Sydney, Newtown, New South Wales, Australia
    • Australian National Liver Transplant Unit, A. W. Morrow Gastroenterology and Liver Center, Centenary Research Institute, Royal Prince Alfred Hospital, University of Sydney, Locked Bag No. 6, Newtown, New South Wales, Australia 2042
    Search for more papers by this author
    • Telephone: 61 2 9565 6100; FAX: 61 2 9565 6101

  • Potential conflict of interest: Nothing to report.


Key Points

1. There is an increasing recognition that previously marginal candidates for liver transplantation can receive therapies that allow transplant to take place.

2. Coronary artery disease is an increasing co-morbidity in liver transplant candidates.

3. Physio-social issues require written guidelines and patient advocates.

4. Methadone maintenance therapy per se is not a contra-indication to liver transplantation. Liver Transpl 18:S-S, 2012. © 2012 AASLD.

The terms extended criteria and marginal could be exchanged with terms such as borderline and unprofitable, and in the setting of liver transplantation, they imply that the outcomes for such patients will be significantly worse than expected. The use of the terms extended criteria and marginal implies that there are well-established criteria for the selection of liver transplant recipients in the first place. Indeed, such criteria do exist. Organizations such as the American Association for the Study of Liver Diseases and the Transplantation Society of Australia and New Zealand1, 2 have recipient criteria for liver transplantation that range from the general to the specific; for example, patients with chronic liver failure due to cirrhosis generally qualify for liver transplantation on the basis of Model for End-Stage Liver Disease (MELD) scores and/or specific complications resistant to current therapies (eg, a MELD score > 15 and/or resistant ascites/encephalopathy). Specific criteria are applied to patients with hepatocellular cancer (HCC), portopulmonary hypertension (PPH), hepatopulmonary syndrome (HPS), alcoholic liver disease (ALD), and human immunodeficiency virus (HIV) coinfections. In addition to such criteria, patients should not have significant extrahepatic diseases that cannot be reversed by transplantation and that consequently may impair optimal outcomes. Finally, patients are expected to have stable psychosocial situations and be able to cope with the stresses and strains of the transplant process for many years.


ALD, alcoholic liver disease; CAD, coronary artery disease; CAG, coronary artery graft; HCC, hepatocellular cancer; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HPS, hepatopulmonary syndrome; MELD, Model for End-Stage Liver Disease; PaO2, partial pressure of arterial oxygen; PAP, pulmonary artery pressure; PPH, portopulmonary hypertension.

Transplant patients meeting these criteria have well-accepted outcomes after liver transplantation with 1-, 5-, and 10-year survival rates of 90%, 75%, and 60%, respectively.3-5 This article discusses where and when recipient criteria can be extended with acceptable outcomes still being maintained.


Extended criteria in this setting include the consideration of liver transplantation for patients with MELD scores < 15 and for patients with very high MELD scores. First, let us consider patients with low MELD scores. The data suggest that the short-term (12-month) survival benefit of liver transplantation does not come into effect until the MELD score reaches approximately 16.6 However, this drops if the 5-year survival benefit is taken into account.7, 8 The major limitation of trekking this ground is the lack of deceased donor organs: even if such patients were offered liver transplantation, they would be competing with patients with higher MELD scores and thus would not be allocated donor organs. There has been much discussion about whether such patients should be listed and be allocated marginal donor organs not suitable for other recipients. An analysis of this practice has led to the conclusion that this is not a good idea, and survival benefits will not be seen.9 The alternative of steering these patients to available living donors has been proposed. Indeed, a recent analysis has suggested survival benefits for non-HCC patients with MELD scores < 15.10

When the MELD system was introduced, it was expected that patients with very high MELD scores (eg, >40) would in fact have worse outcomes. However, in the United States, this generally has not been the case, and high MELD scores do not limit the survival benefits of liver transplantation.7 However, poor outcomes have been observed, particularly at European centers.11 The transplant community continues to struggle with this, but there are currently no established criteria that define futility on the basis of MELD scores; however, a recent abstract suggests that patients who are ventilator-dependent, require dialysis, and are on inotropes do very poorly.12

A well-recognized consideration for the uncomplicated patient is malnutrition. This is associated with worse outcomes after transplantation.13, 14 Studies have addressed the issue of reduced muscle mass and sarcopenia in patients with end-stage cirrhosis.15, 16 A recent review recommends muscle strength training via exercise as a means of improving muscle mass with the potential to improve posttransplant outcomes.16


Most centers do not restrict transplantation on the basis of age alone, although it seems that an age beyond 70 years may be considered an extended criterion. Acceptable results for the transplantation of carefully selected patients older than 70 years have been well reported, although the overall survival rate is lower than that for younger patients.17, 18 Some authors have used an older recipient prognostic score to stratify patients; this score includes factors such as the ventilation status, the diabetes mellitus status, a serum creatinine level > 160 μmol/L, and a combined donor and recipient age > 120 years.19 When this score was >2, the 5-year survival rate was only 50%. Recently, the outcomes of patients undergoing transplantation for HCC at ages > 70 years have been reported.20, 21 Survival was lower for these patients versus younger patients, but unlike patients < 70 years old, the survival of older patients was not affected by the existence of HCC.20 The authors concluded that patients older than 70 years should not be denied liver transplantation. This approach, however, was questioned in an accompanying editorial.21 Perkins21 pointed out that survival was superior in the elderly with HCC who underwent local ablation rather than liver transplantation, although it is likely that tumor-free survival would be lower. It seems reasonable for patients older than 70 years to undergo transplantation, but careful selection is warranted with respect to nonhepatic comorbidities and small HCCs that are amendable to ablation.


A recent global consensus conference on liver transplantation for HCC concluded that “the Milan criteria are currently the benchmark for the selection of non HCC patients for liver transplantation, and the basis for comparison with other suggested criteria.”22 Moreover, “a modest expansion of the number of potential candidates may be considered on the basis of several studies showing comparable survival for patients outside the Milan criteria.”22 Thus, it seems that individual centers in different parts of the world do perform transplantation outside the standard Milan criteria. There is much literature on this topic, which has been summarized by Germani et al.23 It seems that the tumor size may be more important than the tumor number in predicting recurrence along with high alpha-fetoprotein levels (>1000 ng/mL).

Extended criteria also include down-staging. The consensus conference also concluded that “transplantation may be considered after successful downstaging … but … should achieve a 5-year survival comparable to that of HCC patients who meet the criteria for liver transplantation without requiring downstaging.”22

Thus, the trekking of new ground in HCC seems to be very fluid, geographically dependent, and under constant scrutiny with respect to outcomes. Also, the issue of immunosuppression for HCC remains under regular review with great interest in the use of mammalian target of rapamycin inhibitors.24


The 2 main areas of discussion are neuroendocrine tumors that have metastasized to the liver and cholangiocarcinoma. An analysis of the United Network for Organ Sharing database indicates that acceptable 5-year survival in the modern era can be achieved in patients with neuroendocrine tumors as long as patients are carefully selected.25

Worse survival was seen in patients with worse liver function (according to serum bilirubin and albumin levels). Also, elevated serum creatinine levels in donors seemed to affect outcomes.

The only trekking of new ground in cholangiocarcinoma has been the development more than 10 years ago of a pretransplant protocol that includes radiation/local chemotherapy and is designed to select patients less likely to experience posttransplant recurrence.26 More recent reports are available,27, 28 and experience from other centers supports the idea that acceptable results can be achieved with the careful selection of patients.29 The use of such protocols has led to established criteria that are accepted by the United Network for Organ Sharing and allow these patients to undergo liver transplantation.


It is generally agreed that patients with a mean pulmonary artery pressure (PAP) > 40 mm Hg should not undergo liver transplantation.30 The trekking of new ground in this situation is confined to transplantation for patients with severe PPH who successfully undergo pressure-lowering therapies to decrease PAP to <35 mm Hg. This seems reasonable, although the numbers reported so far are relatively limited.31

There does not seem to be an absolute partial pressure of arterial oxygen (PaO2) that should deny the HPS patient liver transplantation.32 However, performing transplantation in patients with a PaO2 < 50 mm Hg and an estimated shunt > 20% seems hazardous.33 Such patients tend to remain hypoxic after transplantation, not come off the ventilator, and die within weeks of transplantation.


It is generally accepted that the main criteria for ALD revolve around abstinence from alcohol in the short and long terms. The standard rule is 6 months of abstinence before listing, although this has been strongly criticized.34 The criticism is largely based on the idea that abstinence for 6 months does not predict long-term survival for patients with ALD, although it does predict a return to heavy drinking. As a result of this dilemma, carefully selected patients with alcoholic hepatitis were selected for transplantation in a recent French study.35 Extended criteria transplantation was restricted to patients with an initial presentation of ALD and a Maddrey score > 32 who failed to respond to a 7-day course of corticosteroid therapy and whose psychosocial issues the care team universally agreed could be addressed. The results indicated 12-month posttransplant survival rates of 78% for these patients and 28% for historical matched controls. It should be recognized that only 2% of the patients with alcoholic hepatitis fulfilled the extended criteria.


The current criteria for patients living with HIV who require liver transplantation for liver failure [usually hepatitis C virus (HCV)–related] include controlled HIV viremia, CD4 counts > 100 × 109/L, and no current opportunist infections.1 Extended criteria may lead to the inclusion of patients who have HIV viremia because of an intolerance of antiviral therapy rather than virological failure. Also, patients with acute HIV and liver failure may be considered candidates when there is no time to control the HIV infection. In both situations, it would be expected that alternative or newly instigated highly acute retroviral therapy would control HIV viremia after transplantation. It should be noted that in some centers, the outcomes of HCV/HIV-coinfected patients after transplantation have been poor despite the performance of transplantation within established criteria; this has provoked some debate about ways to improve outcomes for these patients.36, 37


It is generally accepted (Table 1) that patients with advanced respiratory or neurological diseases or established and untreatable infections that affect liver transplant outcomes are not considered candidates for transplantation. Exceptions include patients with respiratory diseases fulfilling the criteria for lung transplantation (eg, cystic fibrosis), for whom combined lung and liver transplantation has been successfully performed.38 The issues of diabetes mellitus and obesity are covered elsewhere in this issue of the journal,39 whereas the criteria for combined liver and kidney transplantation have been extensively discussed.40

However, an increasingly difficult topic is established CAD. In the early days of liver transplantation, it was thought that patients with advanced liver disease were relatively protected from CAD because of the associated hyperdynamic hemodynamics with low systemic resistance and relatively low blood pressures.41 However, because of the increasing frequency of obesity and nonalcoholic steatohepatitis and the association of HCV with insulin resistance, patients with established chronic liver failure are increasingly being found to have varying degrees of CAD. The prevalence of significant CAD (>70% stenosis) has been estimated to be 20% to 30% in several studies.42 In 1 study,43 such patients had a 50% mortality rate in the first 35 days after transplantation. Furthermore, in the long term, CAD is a major cause of nonhepatic mortality.

There is no doubt that the trekking of new ground for patients with CAD is currently ongoing, although no formal guidelines currently exist. Patients with mild CAD are not an issue, and those with advanced CAD who cannot undergo any revascularization therapies are clearly not candidates unless combined heart and liver transplantation is undertaken (there are no published series on combined transplantation for CAD and chronic liver failure). The main issue is inbetweeners: patients with single left anterior descending lesions or double/triple vessel disease for whom revascularization is possible.

Before these issues are discussed, there is much contention about how potential liver transplant recipients should be screened in the first place.44-46 Negative dopamine stress echocardiography findings are reasonably predictive of a low risk for significant CAD, but the test has poor positive predictive values. Thus, some authors have recommended immediately performing coronary angiography in patients with more than 2 known high-risk factors (eg, diabetes, obesity, a heavy smoking history, and an age > 50 years), although current cardiac society guidelines do not recommend this.47

Once significant CAD (>70% stenosis) is discovered in patients for whom revascularization is possible, there is an increasing view that stents should be inserted before transplantation, although there is no real evidence that this is beneficial for asymptomatic patients with respect to survival outcomes.47 Bare metal stents without dual platelet therapy (aspirin rather than clopidogrel) seem to be the current preference.47 If there is considerable ventricular dysfunction at this stage, such procedures will need to be monitored to establish reversibility. These practices seem reasonable, although only a relatively small number of patients have been reported in the literature.47 Such approaches should enable extended criteria patients to successfully undergo transplantation. The use of anticoagulation before and immediately after transplantation to maximize stent patency remains a challenge.

The management of the patient with established CAD who requires coronary artery graft (CAG) surgery is more complicated.48 There seem to be 4 options: (1) preoperative surgery, (2) a combined procedure at the time of liver transplantation, (3) postoperative CAG, or (4) combined heart and liver transplantation. Option 1 seems hazardous with unacceptable mortality (>50%) in patients with child pugh score B, whereas option 3 still puts the patient at risk of intraoperative and early postoperative mortality and morbidity. Option 4 would usually be reserved for patients with CAD and overt cardiac failure, but as mentioned previously, there do not seem to be reports of combined liver and cardiac surgery for these patients. Success with option 2 has been reported.48 Thus, the trekking of new ground for patients with CAD could include preoperative stenting and, if that is not possible, combined CAG surgery and liver transplantation. Of course, long-term postoperative management of cardiac risk is essential (Table 1).

Table 1. Options in Patients With CAD
  1. NOTE: This table is based on Carey et al.,42 Plotkin et al.,43 Findlay,44 Raval et al.,45 Ehtisham et al.,46 Azarbal et al.,47 Raichlin et al.,48 Gardner and Gibbs,49 Torregrosa et al.,50 Bravo and Hage,51 and Hennessey et al.52

1. How should patients be screened?
 a. Every recipient should undergo resting echocardiography.
 b. Patients with risk factors should undergo dobutamine stress echocardiography.
 c. Patients with more than 2 risk factors should perhaps be directed to coronary angiography.
2. What should be done?
 a. For mild disease, optimize medical therapies.
 b. For stentable disease, use metallic stents with an aspirin cover.
 c. For severe, nonstentable disease, perform simultaneous CAG surgery and liver transplantation or combined heart and liver transplantation.

Although issues with CAD have received the most recent attention, other heart-related conditions may require management in liver transplant recipients. Very small numbers of patients with severe aortic valve disease have undergone simultaneous valve replacement and liver transplantation, although the increasing use of percutaneous procedures may make such procedures unnecessary.49 Patients who present with clinical cirrhotic cardiomyopathy usually are not considered transplant candidates; however, aggressive medical management, including beta-blockers and angiotensin-converting enzyme pathway inhibition, may improve left ventricular dysfunction.50 Likewise, aggressive medical management combined with the use of alcohol atrial septal ablation in patients with hypertrophic cardiomyopathy or a left ventricular outflow tract obstruction also may allow liver transplantation to proceed.51 As mentioned earlier, combined heart and liver transplantation has been reported, but it has been performed mainly for amyloid (not CAD).52


There is obvious (Table 2) concern within liver transplant centers that certain patients, not defined by their medical status, may not be able to cope with the stresses and strains of the liver transplant process. Mental health issues, inadequate social support, and compliance may reflect this. Socioeconomic status is defined in the literature and is potentially measurable. It may be measured by scores based on levels of education, mean household incomes, and geographic locations.53, 54 However, actual data showing that socioeconomic status per se affects outcomes are equivocal. At least 1 study has suggested no effect, whereas others have indicated that a low socioeconomic status is associated with worse outcomes.55-57 It should be noted that 1 study found worse outcomes for patients who did not return to employment after transplantation, and socioeconomic status per se did not have a negative impact.56 However, that study indicated that returning to work after transplantation was linked to the employment status before transplantation (often a marker for a low socioeconomic status). In a recent study,57 a low mean household income was independently linked to mean fibrosis scores after transplantation in HCV-infected subjects. Thus, some data support the concept that socioeconomic status/psychosocial issues should be part of the assessment of an individual for liver transplantation because they may directly or indirectly affect outcomes.

This then begs the question whether socioeconomic status actually influences decisions about listing for liver transplantation. There is evidence that patients with a lower socioeconomic status (eg, African Americans) are listed with higher MELD scores than other patients.53, 54 Also, private health insurance was an independent predictor of liver transplant listing in 1 study.53

Socioeconomic status is also strongly linked to psychosocial barriers to transplantation, and it seems clear that both play roles in the process of listing patients for liver transplantation. Recent studies have tried to examine how such decisions are made by transplant centers.58, 59 Two-thirds of the studied centers used social barriers as a hindrance to listing for liver transplantation, but 50% of the centers had no formal evaluation tool for evaluating such barriers. These centers found that psychosocial barriers often led to “protracted and contentious discussion” within teams. The authors of the study recommended 2 strategies for minimizing prejudice and bias in such decision making: (1) the appointment of an independent patient advocate within the team and (2) documented written policies about exclusion/inclusion criteria for listing (Table 2).

Table 2. Psychosocial Issues
  1. NOTE: This table is based on Kemmer et al.,53 Kemmer and Neff,54 Yoo et al.,55 Hunt et al.,56 Verna et al.,57 Volk et al.,58 and Hogan et al.59

1. There is a correlation between psychosocial issues and low socioeconomic status.
2. There is evidence for discrimination in listing based on socioeconomic issues.
3. Socioeconomic status may not affect posttransplant outcomes.
4. The majority of centers make decisions on the basis of psychosocial considerations.
5. Psychosocial issues result in very prolonged discussions within transplant programs.
6. Definite transparent criteria and patient advocates should be introduced.

In conclusion, if socioeconomic/psychosocial criteria are used to exclude patients, then perhaps such decisions should be universal within the team and should be transparently communicated to patients and families alike (perhaps in a manner similar to that for accepting patients for transplantation with alcoholic hepatitis). The widespread adoption of such procedures would certainly trek new ground and potentially overcome some of the barriers that currently seem to be in place.


Most centers (Table 3) would exclude patients currently using intravenous drugs from liver transplantation.1 This probably reflects concerns about the chaotic nature of the lives of many patients in this situation and the likelihood of noncompliance with medical therapies and ongoing care. Although occasional intravenous drug use may be tolerated (though discouraged), the main discussion within the liver transplant community concerns patients on methadone maintenance programs. The majority of these patients have HCV as the cause of liver failure.

It is well established that, on average, patients with narcotic addictions who enter methadone maintenance programs have less chaotic lives, more stable social lives, and higher employment levels and are involved in fewer crime-related activities.60 It is also clear that when a patient is on such a program for a longer time, the likelihood of illicit opiate use decreases quite dramatically: by 10 years, fewer than 20% of patients use illicit opiates.60 Indeed, Koch and Banys60 stated that “the efficacy of methadone maintenance is one of the most over studied and over proven questions in all of addiction medicine.”

Thus, there is little rationale for routinely excluding patients in methadone maintenance programs. However, this seems not to be the case. In the past, methadone use alone has been used to exclude patients from liver transplant lists.61 Indeed, a survey performed 10 years ago indicated that only 56% of North American liver transplant units accepted patients on methadone, and 32% required the cessation of methadone use.61 When this became apparent, there was much discussion of the inappropriateness of such decisions.61 This coincided with the publication of several articles showing no survival differences between HCV-infected patients on methadone maintenance and patients not on methadone maintenance.52-64 It is recognized, however, that such patients may require increased use of pain-relieving therapies, particularly in the early postoperative period.64 It is important to use pain-relief protocols in conjunction with drug and alcohol specialists in the management of these patients in the early postoperative phase.

It is also clear that careful patient selection is required because returns to intravenous drug use on methadone maintenance have been reported after transplantation and are associated with increased noncompliance.

Nevertheless, the trekking of new ground for methadone patients really means not using methadone alone as a criterion for performing or not performing transplantation.65 Such patients may have increased medical and psychosocial issues, but these should be analyzed on their own merit, and methadone itself should not be used to determine inclusion (Table 3). This is in accordance with American Association for the Study of Liver Diseases practice guidelines.1

Table 3. Recommendations for Methadone Maintenance Patients
  1. NOTE: This table is based on Koch and Banys,60 Jiao et al.,61 Kanchana et al.,62 Liu et al.,63 Weinrieb et al.,64 and Hancock et al.65

1. It should be recognized that methadone is a therapy (not an addiction) and can stabilize patients' lives.
2. Psychosocial issues should be analyzed in their own right.
3. Patients should not be asked to come off methadone before transplantation.
4. Patients may require increased narcotic use after transplantation.
5. Patients should not be refused transplantation on the basis of methadone use alone.


An attempt has been made in Table 4 to do this. Clearly, we can extend the benefits of liver transplantation to recipients who more than 10 years ago would have had established contraindications to transplantation. In order to achieve this in an era of decreasing availability of deceased donor organs and increasing individual entitlement associated with increasing accountability and transparency, we need to establish processes by which the current extended criteria become the new current criteria. However, posttransplant outcomes for patients with very high MELD scores and patients with HCV/HIV coinfections are increasingly under scrutiny, and there have been suggestions of restricting transplantation in these groups.

Table 4. Can We Trek New Ground?
CategoryTrekking New Ground?
General indicationsPatients with low MELD scores (Berg et al.10)
Perhaps the exclusion of patients with very high MELD scores (Weismüller et al.11 and Quante et al.12)
Improvements in nutrition/muscle strength
Age>70 years (perhaps non-HCC patients; Keswani et al.17)
HCCHow far beyond the Milan criteria should we go (Clavien et al.,22 Germani et al.23)?
Non-HCC malignancyNew selection criteria for cholangiocarcinoma and neuroendocrine tumors (Nguyen et al.,25 Rosen et al.,26 Murad et al.,27 Panjala et al.,28 Hong et al.29)
PPHTreatment to reduce PAP to <40 mm Hg (Ramsay,30 Hollatz et al.31)
HPSPaO2 < 50 mm Hg but shunt < 20% (Arguedas et al.33)
ALDAlcoholic hepatitis (Mathurin et al.35)
HIVHIV viremic control only after transplantation
What should be done about HCV/HIV-coinfected patients
Structural nonhepatic diseaseCAD, stents, concurrent surgery, and combined transplantation (Carey et al.,42 Plotkin et al.,43 Findlay,44 Raval et al.,45 Ehtisham et al.,46 Azarbal et al.,47 Raichlin et al.,48 Gardner and Gibbs,49 Torregrosa et al.,50 Bravo and Hage,51 and Hennessey et al.52)
Socioeconomic/psychosocialWritten policies and patient advocates (Volk et al.58)
Narcotic addictionNonexclusion of methadone maintenance patients (Koch et al.60)