Nonalcoholic fatty liver disease and liver transplantation



Nonalcoholic fatty liver disease affects a substantial proportion of the general population worldwide. This high prevalence of nonalcoholic fatty liver disease has important consequences in the donor selection process for liver transplantation, and in the posttransplant period given the high recurrence rate of disease. This paper reviews the prevalence of nonalcoholic fatty liver disease, its progressive potential, and the implications of this liver condition in both the pre- and post-liver transplantation setting. Liver Transpl 12:523–534, 2006. © 2006 AASLD.

With the increasing prevalence of obesity, diabetes mellitus, and the metabolic syndrome in the general population,1–3 nonalcoholic fatty liver disease (NAFLD) has become a household diagnosis in clinical practice of several medical specialties. As little as 10 yr ago, there was skepticism as to whether NAFLD was actually a clinical condition. Interestingly, it is now recognized by hepatologists as one of the most important areas of liver disease. NAFLD is the commonest cause of elevated liver enzymes4, 5 and possibly the most common cause of chronic liver disease. Histologically, NAFLD resembles, albeit is not identical to, alcohol-induced liver injury, but this liver condition is, by definition, not alcohol-induced.6 The disease includes a spectrum of hepatic pathology that ranges from simple, bland steatosis to steatosis plus features of cellular injury, including necrosis, hepatocyte ballooning, and inflammatory infiltrate (i.e., nonalcoholic steatohepatitis [NASH]), which may or may not be associated with increased liver fibrosis. In some patients, NASH progresses to advanced fibrosis and cirrhosis with their consequent complications of liver failure and hepatocellular carcinoma.6 Thus, most liver related morbidity and mortality in patients with NAFLD occurs in those whose liver disease has progressed to advanced fibrosis or cirrhosis. A substantial proportion of patients with cryptogenic cirrhosis suffer from the metabolic syndrome or some of its components (i.e., central obesity, raised blood pressure, increased triglyceride levels, hyperglycemia, and decreased high-density lipoprotein-cholesterol)7 suggesting that their “cryptogenic” cirrhosis is unrecognized NASH.8, 9 Indeed, as cirrhosis develops, the degree of steatosis and inflammatory features improve or disappear, and thus the liver biopsy may show cirrhotic stage disease without other features of NAFLD.10 Over the last decade, a wealth of information has been generated, indicating that NAFLD is a cause of end-stage liver disease leading to transplantation.

Insulin resistance is almost a universal finding in patients with NAFLD. Thus, NAFLD has been considered the hepatic manifestation of the metabolic syndrome, a cluster of abnormalities related to insulin resistance, of which central obesity is one of the major components.7 Excessive adiposity contributes to the tissue damage occurring in patients with the metabolic syndrome because fat-derived factors regulate the inflammatory response. At least 3 of these factors, adiponectin, tumor necrosis factor (TNF) alpha, and free fatty acids (FFA) promote NAFLD by modulating the hepatic inflammatory response.11–14 Adiponectin, through its interaction with its receptor in liver cells, inhibits FFA uptake while stimulating FFA oxidation and lipid export. These actions of adiponectin inhibit fat accumulation in hepatocytes and enhance hepatic insulin sensitivity. Thus, adiponectin has a protective effect in the liver against NAFLD development. Conversely, TNF alpha is a proapoptotic cytokine with a key role in recruiting inflammatory cells to injured tissues, and in promoting insulin resistance. TNF alpha downregulates proteins mediating the effects of insulin, such as the insulin-responsive-glucose-transporter-4 and the peroxisome-proliferator-activated-receptor gamma. Adiponectin and TNF alpha are mutually antagonistic, inhibiting each other's production and activity. Visceral adipose tissue is more resistant to insulin, exhibits greater lipolysis, and provides an important source of FFA for the liver. Increased hepatic FFA activate signals that increase hepatocyte production of TNF alpha. This inhibits hepatic adiponectin activity, promoting NAFLD and liver insulin resistance.11–14

Hepatic lipid accumulation (i.e., steatosis) does not always result in hepatocellular injury, indicating that additional insults are necessary.15 A variety of factors have been implicated in the production of an additional “insult,” including oxidative stress and gut-derived bacterial endotoxins. Increased hepatic FFA oxidation increases the generation of reactive oxygen species. Increased hepatocyte exposure to reactive oxygen species generates a state of oxidant stress and mitochondrial dysfunction. Also, fatty liver clears intestinally-derived endotoxin lipopolysaccharide poorly, permitting this potent cytokine-inducing substance to escape the liver.16 Increased exposure of peripheral fat tissue to lipopolysaccharide stimulates release of TNF alpha and FFA, which go to the liver, where they lead to further inhibition of adiponectin.


NAFLD affects a substantial proportion of the general population from several countries. Using proton magnetic resonance spectroscopy, the Dallas Heart Study (a population-based cohort study performed in an ethically diverse community in the United States) reported that 1 in 3 adult Americans have steatosis (as defined by hepatic triglyceride content greater than 5.5%) mostly unrelated to alcohol abuse.17 Although there was no histological correlation to determine that these subjects indeed had steatosis, the finding, if true, indicates that over 60 million adult Americans and an unknown proportion of children in this country suffer from NAFLD. The prevalence of NAFLD in the United States seems to be different among different ethnic groups with a significantly higher prevalence among Hispanics and white men as compared to African Americans or white women.17 Differences in body fat distribution and body composition among the different ethnic groups may partially explain the racial differences in prevalence. For instance, populations with lower heights (i.e., Hispanics) have a higher proportion of body fat and higher waist-to-hip ratio than their taller counterparts.18 Similarly, Asians have a higher proportion of visceral fat and a lower proportion of lean body mass than white subjects with the same body-mass index (BMI).19 The central (or upper body) obesity phenotype is associated with increased intraabdominal (or visceral) fat. Visceral adipose tissue has greater lipolytic potential than subcutaneous adipose tissue and the release of FFA from visceral fat directly into the portal circulation creates a “first pass” effect.20 Increased FFA concentrations, in turn, are considered a major mediator of insulin resistance. The increased delivery of FFA to the liver may be responsible for hepatic insulin resistance, which is the driving pathogenic force behind NAFLD.21 In contrast, FFA flux and concentrations in individuals with predominantly lower-body obesity tend to be normal, regardless of BMI.22 Therefore, patients with central obesity are characteristically insulin-resistant, and more commonly present with elevated aminotransferases due to NAFLD as compared to patients with lower-body obesity.23

The remarkably high prevalence of NAFLD is not exclusive to the United States population. Using liver ultrasonography, a recent population-based cohort study performed in Italy found that 1 in 4 or 5 adults in that country suffer from NAFLD.24 NAFLD also has reached epidemic proportions among populations typically considered at “low-risk” for this liver condition. For instance, population-based cohort studies in China25 and Japan26 have reported a prevalence of NAFLD of 15% and 14%, respectively, among adults. The clinical implications of this alarming prevalence of NAFLD are derived from the fact that this liver condition may progress to end-stage liver disease and liver cancer.

All these population-based studies17, 24–26 provide a more realistic scenario of the real prevalence of NAFLD in the general population as compared to autopsy studies, hospital series, or studies performed exclusively in obese populations. The prevalence of NAFLD, however, is expected to increase worldwide, as we are encountering a global obesity epidemic and the trend in developing countries toward the Western lifestyle.


NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; HCV, hepatitis C virus; BMI, body mass index; FFA, free fatty acids; TNF, tumor necrosis factor; PNF, primary nonfunction; OLT, orthotopic liver transplantation.


The natural history of NAFLD has been evaluated based on studies of sequential liver biopsies,10, 27, 28 retrospective studies of patients seen in tertiary referral centers,29–35 and in a community-based study.36

Studies of Sequential Liver Biopsies

The histological changes occurring over time in the absence of effective treatment have been described in 3 studies with sequential liver biopsies.10, 27, 28 The average time interval between biopsies was 5.7 yr (range 1.4-15.7 yr) in 22 patients,27 4.3 yr (range 3.0-14.3 yr) in another 22 patients,28 and 3.2 yr (range 0.7-21.3 yr) in the largest series of 103 patients.10 Progression in fibrosis stage occurred in 1 of 3 patients in each 1 of these 3 series, but improvement in fibrosis stage was seen in 18% in 1 series27 and in 37% in another series.10 Interestingly, as fibrosis develops and progresses over time, other features of NAFLD including steatosis, inflammation, and ballooning of hepatocytes significantly improve or disappear,10 and thus, liver biopsy features other than fibrosis severity may not be useful to predict the long-term prognosis in an individual patient with NAFLD. Further, the histological features of NAFLD that create the basis for the histological diagnosis of NASH (i.e., inflammation and hepatocyte ballooning) are unequally distributed throughout the liver parenchyma, with liver biopsy resulting in misdiagnosis in some patients.37 Also, aminotransferases are within the normal range in almost 80% of patients with NAFLD in the general population,17 and when elevated, they improve or normalize spontaneously overtime despite fibrosis progression.10

Studies on Long-Term Mortality

Few studies have evaluated the long-term prognosis of patients with NAFLD; they are summarized in Table 1.29–36 Overall, the disease progresses slowly over many years or decades, but the prognosis is not the same across the different stages of NAFLD. On one end of the spectrum, patients with biopsy-demonstrated bland steatosis and no histologic evidence of NASH or fibrosis appear to have a more benign prognosis.29, 30 For instance, a Danish study of a cohort of 109 predominantly morbidly obese subjects followed for nearly 17 yr found the incidence of cirrhosis to be <1%.30 During follow-up, 1 in 4 patients died but the survival curve of the general population fell within the 95% confidence interval of the survival curve of patients with bland steatosis. In that study,30 the patient who developed cirrhosis was the only one who died from liver-related causes. On the other end of the spectrum, patients with cirrhotic stage NASH have a worse prognosis. For instance, in a recent study35 1 in 4 patients with cirrhotic stage NASH died within 5 yr of follow-up with most causes of death related to end-stage liver disease.

Table 1. Studies on Long-Term Prognosis of NAFLD
ReferencesDiagnosisNCirrhosis prevalenceLiver-related deaths n (%)Overall deaths n (%)Average F/U (yr)
  1. NOTE. Cirrhosis prevalence includes all patients diagnosed with cirrhosis at both baseline and during follow-up.

  2. Abbreviations: NAFLD, includes both patients with simple steatosis and patients with NASH; F/u, follow-up.

Teli et al.29Bland steatosis400%014 (35)9.6
Dam-Larsen et al.30Bland steatosis1091%1 (0.9%)27 (24.8)16.7
Matteoni et al.31NAFLD9820%9 (9%)48 (49)8.3
Adams et al.36NAFLD4205%7 (1.7%)53 (12.6)7.6
Lee et al.32NASH3916.3%1 (3%)10 (26)3.8
Powell et al.33NASH427%1 (2%)2 (5)4.5
Evans et al.34NASH264%04 (15)8.7
Hui et al.35Cirrhotic-stage NASH23100%5 (21%)6 (26)5.0

Overall, a diagnosis of NAFLD is associated with a shorter than expected survival. For instance, in a recent community-based study of 420 patients with NAFLD the observed mortality was significantly higher among NAFLD patients as compared to the expected mortality of the general population of the same age and gender.36 In that study, the most common causes of death among patients with NAFLD were cancer and cardiovascular disease, just as observed in the general population; however, liver-related complications constituted the third most common cause of death among NAFLD patients as compared to the 13th most common cause of death in the general population.36 This indicates that complications of end-stage liver disease contribute importantly to mortality in patients with NAFLD. In that community-based study,36 patients dying from liver-related causes were those with more advanced NAFLD confirming prior observations of smaller studies.30, 35

NASH as the Progressive Form of NAFLD

Typically, NASH has been considered to be the progressive form of NAFLD, but it remains uncertain what proportion of NASH patients will progress to end-stage liver disease. Further, there is no consensus on the best definition of NASH, and it remains uncertain what prognostic implication can be derived from grading and staging the different histological features necessary to make the diagnosis of NASH, including inflammation and hepatocyte ballooning. Undoubtedly, liver fibrosis per se is a worrisome feature on liver biopsy and may provide the most prognostic information as compared to other histological features. Liver fibrosis suggests a more severe and progressive liver injury in patients with NAFLD just as occurs in liver diseases of different etiologies. Increased fibrosis is common in patients with NAFLD undergoing liver biopsy. For instance, based on an analysis of 12 cross-sectional studies including a total of 673 patients with liver biopsy-proven NAFLD, 14% had cirrhotic stage disease, 25% had either cirrhosis or septal fibrosis, and 66% of the patients had some degree of fibrosis.6


NAFLD as an Unrecognized Cause of Cryptogenic Cirrhosis

The frequency with which NAFLD progresses to end-stage liver disease requiring liver transplantation remains uncertain. In a large, single-center experience, only 2.6% of patients with end-stage liver disease evaluated for liver transplantation had underlying end-stage NASH.38 Although this indicates that only a minority of patients with NAFLD will progress to end-stage liver disease, data from liver transplant centers provide only a partial estimate of the problem as many patients, due to important comorbidities-related complications are not evaluated for liver transplantation as discussed further in this review.

Perhaps the potential for NAFLD to result in end-stage liver disease is better highlighted by some evidence suggesting that NAFLD underlies a substantial proportion of cases of cryptogenic cirrhosis.8, 9, 39 In a large study,8 73% of patients with cryptogenic cirrhosis had a BMI in the obese category or suffered from diabetes, whereas another series9 reported a prevalence of obesity or diabetes in about one-half of the cryptogenic cirrhotic patients. The prevalence of NAFLD as an unrecognized cause of cryptogenic cirrhosis in both series8, 9 is most likely underestimated, as some nondiabetic, nonobese (i.e., BMI < 30) patients may suffer from central obesity and/or dyslipidemia, which were not consistently measured in these series. Central obesity and dyslipidemia may be the only risk factors for NAFLD that are in some cases labeled as cryptogenic cirrhosis. In another series,39 33% of patients undergoing liver transplantation due to cryptogenic cirrhosis had histological evidence of NAFLD on a more detailed histological review of their explanted livers. This 33% is a clear underestimate of NAFLD as a cause of cirrhosis, as the histological features of NAFLD including steatosis, inflammation, and hepatocyte ballooning often resolve as NAFLD progresses to cirrhosis. Thus, the prevalence of NAFLD as an unrecognized cause of cryptogenic cirrhosis is most likely higher than reported in these 3 series.8, 9, 39

NAFLD Increases Disease Severity/Progression in Other Liver Diseases

NAFLD often coexists with, and most likely contributes to the development of, cirrhosis in patients with liver disease of various etiologies. For instance, hepatic steatosis is commonly found in patients with chronic hepatitis C virus (HCV) infection, occurring in about one-half (range 30-70%) of the liver biopsy samples. Although HCV has a direct role in the development of steatosis,40 the coexistence of NAFLD and the risk factors for NAFLD development (i.e., obesity and type 2 diabetes) influences the progression of liver fibrosis41–43 and possibly the development of hepatocellular carcinoma.44 Also, excess body weight has been recognized as a risk factor for the development of cirrhosis in alcoholic liver disease. In subjects with heavy alcohol consumption, excess body weight markedly increases the presence of steatosis45 and is a risk factor for the development of acute alcoholic hepatitis and cirrhosis.46, 47 Similarly, in patients with hemochromatosis, the odds of having liver fibrosis increases 3.9-fold when steatosis is present.48 All this evidence clearly indicates that NAFLD is an important cofactor, contributing significantly to disease severity and progression in patients with liver disease of various etiologies.

Data from the United Network for Organ Sharing

Table 2 summarizes the number and proportion of liver transplants performed from 1995 to May 2005 in cirrhotic stage NASH and cryptogenic cirrhosis in the United States, using data from the United Network for Organ Sharing.49 The first cases of cirrhotic-stage NASH undergoing liver transplantation were recorded in 1996. The proportion of patients with cirrhotic-stage NASH undergoing liver transplantation has increased from 0.1% or less in the period 1995-2000 to 3.5% in 2005. This is a more than 35-fold increment in the overall proportion of NASH patients undergoing liver transplantation. However, as shown in Table 2, the overall proportion of cryptogenic cirrhosis undergoing liver transplantation had decreased from 9.6% in 1995 to 6.6% in 2005; thus, it is possible that as physician awareness of the disease increases, more cases are now called NASH instead of cryptogenic. If one assumes that all cases of cryptogenic cirrhosis are indeed unrecognized NAFLD, then the proportion of patients with cirrhotic-stage NAFLD undergoing liver transplantation has not changed dramatically, from 9.6 in 1995 to 10.2 in 2005; or from 4.8% to 6.9% if only one-half of the patients with cryptogenic cirrhosis are counted as unrecognized NAFLD.

Table 2. Number of Patients with Cryptogenic Cirrhosis or Cirrhotic-Stage NASH Undergoing Liver Transplantation from 1995 to May 2005
  • NOTE. Data from the United Network for Organ Sharing.49

  • *

    Includes data from January 1, 2005 to May 31, 2005.

  • NAFLD denotes NASH + 50% of patients with cryptogenic cirrhosis.

Cryptogenic cirrhosis378379347350357405352340376397180
Fatty liver or NASH04334413142612896
Total liver transplant procedures39334082418645164750499251885331567161692715
Transplants due to cryptogenic cirrhosis (%)
Transplants due to cirrhotic-stage NASH (%)
Transplants due to cryptogenic cirrhosis + NASH (%)
Transplants due to NAFLD (%)

Figure 1 summarizes the proportion of liver transplants divided by cause of cirrhosis, performed in the United States from 1992 to May 2005 using data from the United Network for Organ Sharing.49 The proportion of patients with cirrhotic-stage NAFLD (i.e., cirrhotic NASH plus one-half of cryptogenic cirrhosis) undergoing liver transplantation is still lower than the proportion of patients with cirrhosis due to HCV infection or alcohol abuse, but similar (or even higher in the last few years) to primary biliary cirrhosis, primary sclerosing cholangitis, and autoimmune hepatitis. However, patients with NAFLD are underrepresented in any liver transplant database, as many patients with end-stage NAFLD suffer from severe obesity and long-term complications of diabetes and/or dyslipidemia such as cardiovascular and pulmonary diseases and extrahepatic malignancies, precluding them from liver transplantation. Many patients with end-stage NAFLD are not even included in the waiting list for transplantation and thus liver transplant databases provide only a partial estimate of the potential of NAFLD to progress to end-stage liver disease.

Figure 1.

Number of liver transplant procedures performed in the United States from 1992 to May 2005 and divided by cause of cirrhosis (data from the United Network for Organ Sharing).49 NAFLD includes all cases of NASH plus one-half of cases of cryptogenic cirrhosis.


Over the last 2 decades, orthotopic liver transplantation (OLT) has become a routinely applied and successful treatment for end-stage liver disease from a variety of causes. Over the same period of time, however, organ availability has increased at a much slower rate and has plateaued between 4,000 and 6,000 annual liver grafts in the last 10 yr.49 This disproportion between supply and demand for donor livers has lead to a substantial expansion in the United Network for Organ Sharing liver transplant waiting list and a sizeable increase in the number of deaths (over 1,000 every year) while waiting for liver transplantation.49 The disparity between supply and demand for donor livers has lead to pursue some strategies aimed at increasing the organ pool. One of these strategies includes the use of “extended criteria donors.”50 Based on the extended criteria donors concept, the graft from such a donor is at increased risk of early failure due to primary nonfunction (PNF) or initial poor function of the graft, or such a graft predisposes to decreased graft or patient survival. One of the factors predisposing to this increased risk or poor initial graft function includes hepatic steatosis, among other factors such as older donor age, female gender, donor hypernatremia, positive hepatitis serologies, pressor requirement, and the use of non-heart beating donors.50 One or more of these factors have been the reason to exclude a donor for consideration. However, these factors are now being more closely examined for the actual degree of risk they impart to a recipient and for the extent to which they can be matched to a recipient to maintain acceptable graft and patient outcome.

Assessing the Presence and Severity of Steatosis in Donors

Given the epidemic proportion of NAFLD in the general population, it is not surprising that graft steatosis is commonly found in donor livers. Using hematoxylin and eosin-stained sections of liver biopsy specimens, some degree of steatosis is found in one-third to one-half of potential living donors (i.e., a relatively healthy nonbiased population) with 5% to 15% of cases showing more than 30% of hepatocytes affected.51, 52 In cadaveric donors, the prevalence of steatosis reported has been as high as 37%53 to 51%,54 with 21% of cases showing more than 30% of hepatocytes affected.54

Although obese donors are most likely to have steatosis,51 the BMI per se correlates weakly with presence and severity of steatosis.52 Subjects with a BMI in the nonobese category may still have risk factors for the development of hepatic steatosis, including central obesity, glucose intolerance, and/or dyslipidemia. Different imaging procedures, including ultrasonography, computed tomography scan, and magnetic resonance imaging, have been used to detect and quantify the fat content in the liver. Magnetic resonance imaging and spectroscopy allow the detection of small amounts of fat components in hepatocytes whereas liver tissue density assessed by computed tomography scan accurately reflects the presence of steatosis. In general, a positive imaging finding usually correlates positively with steatosis on biopsy, but the converse may also occur with a false-negative rate of 30% by magnetic resonance imaging and 24% by computed tomography when steatosis affects 10% or less of hepatocytes.51 The sensitivity of imaging studies in detecting steatosis increases paralleling the increment on liver fat amount. In a recent large series, imaging was only 12% sensitive to steatosis of 5% to10%, but increased to 80% to steatosis of 30% or more.52 However, even in cases of steatosis of 30%, imaging studies may be falsely negative.52 Hence, accurate quantification of hepatic fat is not affordable by BMI or imaging studies alone in all donor candidates. Furthermore, fat infiltration of the liver has a nonuniform distribution in some cases, and thus, 1 or 2 liver biopsy specimens from the entire liver may provide inaccurate and misleading information.37

In the United States almost all transplant surgeons perform a liver biopsy of the cadaveric donor, particularly under some circumstances such as adverse donor risk factors (i.e., older age, increased BMI) or when the liver appears macroscopically steatotic.55 The role of liver biopsy in the evaluation of potential living donors, however, remains under debate with practices varying widely among centers. Certainly, besides helping to confirm and quantify steatosis, a preoperative liver biopsy in potential living donors may reveal an occult liver condition leading to exclusion from donation.52, 56 Conversely, appropriate weight control with diet and physical activity in overweight/obese living donors may reduce the underlying moderate/severe steatosis, thus increasing the donor pool.57, 58

The Use of Steatotic Livers for Transplantation

Steatotic livers are associated with higher rates of initial poor function and PNF, and poorer outcome. In a large series, patients receiving livers with up to 30% of macrovesicular steatosis had a higher rates of PNF (5.1% vs. 1.8%) and worse 2-yr patient (77% vs. 91%) and graft (70% vs. 82%) survival as compared to patients receiving livers without steatosis.59 A large, retrospective, single-center study comparing donor grafts with mild (<30%), moderate (30-60%), and severe (>60%) amounts of macrovesicular steatosis showed increased initial poor graft function, increased graft loss rates, and higher 3-month patient mortality among recipients of livers with moderate and severe steatosis.53 Higher rates of HCV recurrence were seen in these categories as well.53 Steatosis greater than 30% is associated with a 3.7-fold higher risk of graft initial poor function as compared to grafts with no or <30% steatosis.60 Although different definitions of initial poor function of the graft have been used, this posttransplantation complication occurs in about 1 in 3 grafts with >30% steatosis.60 The occurrence of PNF, a complication with a high mortality rate and emergent need for retransplantation, is far more common among grafts with severe (>60%) steatosis.60, 61 Conversely, microvesicular steatosis does not appear to result in increased sensitivity of a liver to ischemia and/or reperfusion. For instance, in a cadaveric donor setting, Fishbein et al.62 reviewed 426 transplants and identified 40 cases containing >30% microvesicular steatosis and 386 livers with <30% microvesicular steatosis. In that study,62 the incidence of PNF was 5% and 5.1% in the 2 groups with a similar 1-yr patient (80% vs. 79.8%) and graft (72.5% vs. 68.4%) survival.

There is no consistent threshold of estimated macrovesicular steatosis above which liver transplantation is precluded. Among cadaveric donors, livers with >60% of macrovesicular steatosis should probably be excluded automatically, whereas liver with <30% of macrosteatosis can be used, particularly when no other donor risk factors exist and cold ischemia times are minimized. The use of grafts with moderate macrosteatosis (i.e., 30-60%) is more controversial, but they probably may be utilized in the absence of additional risk factors in both the donor and recipient, or when the recipient circumstances are judged to be sufficiently critical to justify the risk.63, 64 While the use of marginal cadaveric donor livers is an appropriate strategy to increase the number of available grafts for transplantation, the criteria are more rigorous in living donor liver transplantation. Presence of moderate/severe steatosis or any kind of hepatitis or fibrosis are absolute contraindications for living donation in most transplant centers. However, it remains controversial whether potential donors with mild macrosteatosis (up to 30% steatosis) should be denied from donation. A recent study of 55 cases undergoing living donor liver transplantation65 with grafts containing less than 30% macrosteatosis suggests the safe use of these mildly steatotic livers for living donation. In that study,65 steatosis decreased to <10% in all grafts by day 10 after liver transplantation and hepatic regeneration power was not impaired; no PNF or delayed graft function occurred in any case. These results, although encouraging, need to be reproduced in larger studies that should also evaluate the impact of fat on the donor.

Mechanisms of Injury in Steatotic Grafts

The prevention of PNF of fatty liver posttransplantation should be based on the knowledge of the mechanism(s) of injury. It has been demonstrated that the type of ischemic injury in steatotic livers is not the same as in normal livers.66–68 Steatotic livers develop a necrotic form or cell death after reperfusion whereas reperfusion of ischemic lean livers results in the activation of the apoptotic pathway.69 Steatosis is associated with decreased adenosine triphosphate generation, which is necessary as an energy source in the apoptotic pathway. In a mouse model of liver ischemia, steatotic hepatocytes (as compared to their lean counterparts) contained a significant decrease in the amount of adenosine triphosphate within 24 hours of reperfusion.70 Restoring adenosine triphosphate levels in fatty livers protects against necrosis. Also, hepatocytes of steatotic livers show a decreased ability to regenerate after major tissue loss. In a model of partial hepatectomy of steatotic livers (Zucker rats), regeneration was reduced by 60% when compared to lean controls.71 The failure to progress from the G1 to the S phase of mitosis has been linked to the inhibition of cyclin D production in steatotic hepatocytes.72 This inability of steatotic liver to regenerate is important because preservation and rewarming of the graft is associated with parenchymal and no parenchymal cell death. A recent series of 100 patients undergoing major liver resection demonstrates ischemic preconditioning (induced by 10 minutes of ischemia followed by 10 minutes of reperfusion) as a protective strategy against hepatic ischemia in humans.73 In that study,73 ischemic preconditioning was particularly effective in patients with steatotic livers and in younger individuals with longer duration of inflow occlusion (up to 60 minutes). The beneficial effects of ischemic preconditioning may be related to preservation of adenosine triphosphate content in the liver. An outstanding review on several protective strategies against ischemic injury of the liver has been published recently.74


Recurrence of NAFLD after Liver Transplantation for Cirrhotic-Stage NAFLD or Cryptogenic Cirrhosis

In 1992, Burke et al.75 reported on 2 patients undergoing OLT for advanced cirrhosis that had developed several years after jejunoileal bypass for morbid obesity. In 1 of these cases, development of NASH was documented 2 months after OLT. Although detailed information of the explanted liver was not provided, the case possibly represents 1 of the first reports suggesting that NAFLD may recur after liver transplantation. Since then, several series have been published reporting recurrence or development of NAFLD after liver transplantation for both cirrhotic stage NAFLD and cryptogenic cirrhosis. Kim et al.76 reported 8 cases undergoing OLT for advanced-stage NAFLD. Seven patients were either obese or overweight pre-OLT, whereas the only patient with normal BMI had undergone jejunoileal bypass previously. Recurrence of steatosis was documented in 6 of 8 patients from 3 weeks to 2 yr after OLT, with 3 patients meeting criteria for NASH. In 2 of the 3 patients who developed NASH, progression from simple steatosis to NASH was documented at 1 and 2 yr with development of pericentral fibrosis documented in both cases at 2 yr. The 6 patients with disease recurrence remained above their ideal body weight and 5 of them had hypertriglyceridemia; the 2 patients without disease recurrence achieved some weight loss post-OLT. Subsequently, 2 other cases of NASH recurrence were reported; in both cases NASH was documented at about 1.5 yr post-OLT;77, 78 and 1 of these cases78 was documented with cirrhotic stage NASH in as shorts as 77 weeks post-OLT.

After these initial cases, several series were reported analyzing the course of patients with cirrhotic-stage NAFLD or cryptogenic cirrhosis undergoing liver transplantation.38, 79–82 In a series of 30 patients with NASH-related cryptogenic cirrhosis, Contos et al.79 documented the development of steatosis in about 1 in 4 patients in the first 12 months post-OLT, with more than one-half of patients having steatosis or NASH 4 yr post-OLT. A striking finding of the study,69 was that in the long-term, the hazard for developing recurrent NAFLD (either steatosis alone or NASH) approached 100%, with all patients evaluated having developed NAFLD in the graft by 5 yr, as compared to only 25% of age- and gender-matched controls with primary biliary cirrhosis / primary sclerosing cholangitis, or alcoholic liver disease undergoing OLT in the same period. On multivariate analysis, only the cumulative steroid dose correlated with time to development of graft steatosis. Three patients developed NASH some time after they had been documented with steatosis alone, and 1 of these patients developed bridging fibrosis. Interestingly, both steatosis and NASH developed in some individuals without an elevation in alanine aminotransferase levels with the disease detected on protocol liver biopsy. During a mean follow-up of 3.5 ± 2.7 yr, 3 patients died (1 due to PNF and 2 due to sepsis) with a cumulative survival post-OLT similar to patients transplanted for primary biliary cirrhosis/primary sclerosing cholangitis or alcoholic liver disease during the same period of time.79

Ong et al.80 reported their experience with 51 patients with cryptogenic cirrhosis undergoing OLT. Twenty-five patients had at least 1 liver biopsy post-OLT, and 13 of them developed NAFLD, including 8 patients with steatosis alone and 5 patients with NASH. Before OLT the 5 patients who developed NASH post-OLT were more commonly diabetic and had higher triglyceride levels than the 12 patients who did not develop NAFLD post-OLT. Similarly, pre-OLT the 8 patients with steatosis post-OLT had higher triglyceride levels than the 12 patients without NAFLD post-OLT. After OLT, the prevalence of diabetes in those with post-OLT NASH, and steatosis, and in those without disease recurrence was 62.5%, 40%, and 16.7%, respectively. Overall, the mean BMI increased in those who developed steatosis or NASH post-OLT and decreased in those who did not. Interestingly, 2 of the 5 patients with NASH post-OLT progressed to advanced fibrosis, including septal fibrosis at 37 months post-OLT in 1 patient, and cirrhotic stage disease in the other patient at 80 months post-OLT.

The Mayo Clinic series38 reported a total of 546 liver transplantations performed from 1993 to 1998, 16 of which (2.6%) were for end-stage NASH. Protocol liver biopsy was performed at 4 months, 1 yr, and then annually post-OLT in the 15 patients surviving beyond 4 months. At 4 months post-OLT, 9 (60%) patients had recurrence of steatosis in the allograft, and 5 of these 9 patients met criteria for NASH (2 at 4 months, 2 at 1 yr, and 1 at 2 yr posttransplantation). The proportion of patients with steatosis posttransplantation was higher in recipients with end-stage NASH as compared to recipients with other liver disease. The rate of fibrosis development among the 15 NASH recipients was 20% at 4 months and 33% at 1 yr. Two of these patients progressed to cirrhosis in the follow-up period; 1 of them who had had NASH documented at 4 months and cirrhotic stage disease at 2 yr developed liver failure requiring retransplantation at 27 months after the first OLT. This indicates than NAFLD recurrent after liver transplantation is not a benign condition, but it may progress to end-stage disease with liver failure and the need for retransplantation.

Further evidence of NAFLD development/recurrence after liver transplantation and its potential to progress to cirrhosis was reported recently by Sutedja et al.81 Among 1,710 patients undergoing liver transplantation at the Queen Elizabeth Hospital in the UK, 39 who were transplanted for cryptogenic cirrhosis and survived beyond 1 yr posttransplantation underwent protocol liver biopsies at 1 yr and thereafter. Twelve (37.5%) of these patients developed steatosis (n = 9) or steatohepatitis (n = 3). Three patients had developed steatohepatitis at the first yr posttransplant including 1 found with increased fibrosis at 48 months and the other 2 progressing to cirrhotic stage disease at 72 and 84 months, respectively. The rate of steatosis and steatohepatitis posttransplantation among the cryptogenic recipients was significantly higher as compared to patients transplanted for no cryptogenic causes.

Another series from the Mayo Clinic reported the outcomes posttransplantation for cryptogenic cirrhosis.82 The series included 39 patients undergoing OLT for cryptogenic cirrhosis among a total of 560 nonfulminant cases of OLTs performed in that institution. Overall, the cumulative 1- and 5-yr post-OLT mortality was significantly higher in recipients receiving transplantation for cryptogenic cirrhosis than in those transplanted for other reasons. Twenty-seven of the 39 cryptogenic recipients surviving more than 6 months post-OLT underwent annual protocol liver biopsy during a mean follow-up of 5.5 yr. Among these 27 biopsied patients, 6 (22%) had persistent hepatitis including 2 (7%) with steatohepatitis. Different characteristics of the noncryptogenic control patients may explain the difference in survival found by Contos et al.79 and the Mayo Clinic experience.82

De Novo NAFLD After Liver Transplantation

De novo development of NAFLD after liver transplantation has been documented in some cases. In 2000, Morales et al.83 reported in abstract form the progression from steatosis to NASH and fibrosis within 5 yr of liver transplantation for primary biliary cirrhosis in 1 patient, and for alpha-1-antitrypsin deficiency-associated liver disease in another patient. Both cases were reported alive after a median follow-up of 5.7 yr. Although details of the clinical condition or explanted liver were not provided, the patients seemed to have suffered from risk factors for NAFLD pre- and/or posttransplantation.83

Subsequently, 4 cases of de novo NAFLD post-OLT were reported. The underlying cause of cirrhosis leading to OLT was alcohol abuse in 1 case, and HCV infection in the other 3. NAFLD developed within the first 3 months post-OLT and was essentially in the form of macrovesicular steatosis. Although HCV genotype 3a recurrence after OLT in 1 case may explain the development of steatosis, no clear-cut predisposing factors were identified in the other 3. Interestingly, the 4 cases showed moderately severe preservation injury on protocol liver biopsy performed at day 7-8 post-OLT. As discussed by the authors,84 preservation injury induces some abnormalities that may lead to NAFLD including mitochondrial dysfunction,85 oxidative stress, and lipid peroxidation.86

Factors Leading to Disease Recurrence and Therapeutic Implications

Based on the analysis of the series described above,38, 75–84 it is clear that NAFLD recurs in the allograft of patients undergoing transplantation for end-stage NASH, or develops after transplantation for cryptogenic cirrhosis. Recurrence of NAFLD is not exclusive of recipients of cadaveric grafts, but it has been documented in at least 1 living related liver transplant recipient.87 The recurrence/development of NAFLD takes place early in the posttransplantation period, and the rate of disease recurrence/development increases as the posttransplantation follow-up lengthens. NAFLD posttransplantation can be severe, with some patients progressing to cirrhotic stage disease. Nevertheless, the long-term prognosis of patients undergoing liver transplantation for end-stage NAFLD or cryptogenic cirrhosis has to be determined in studies with longer follow-up, but at least 1 large clinical experience suggests survival posttransplantation in these patients is significantly affected as compared to patients transplanted for other indications.82

Insulin Resistance in the Posttransplantation Period

Although no studies have been performed to determine the insulin sensitivity status in patients with NAFLD recurrence/development posttransplantation, insulin resistance most likely plays a central role in disease recurrence/development in most or all cases. Several of the risk factors that lead to development of end-stage NAFLD pretransplantation, including increased adiposity, glucose intolerance/diabetes, and dyslipidemia, persist (and may worsen) in the posttransplant period, or these risk factors may develop posttransplantation.76, 77, 79, 80 In fact, the prevalence of these risk factors for NAFLD is higher in the posttransplantation period among liver transplant recipients than in the general population (Table 3).88 At the time of transplantation many patients are overweight/obese and some even gain weight posttransplantation. Posttransplant diabetes mellitus is common and caused by increased resistance to insulin and by deficit in insulin secretion.89 The use of immunosuppressive agents posttransplantation promotes insulin resistance, playing a role in NAFLD recurrence/development; for instance, steroids increase insulin resistance, and the calcineurin inhibitors cyclosporin and tacrolimus inhibit the synthesis and secretion of insulin by pancreatic beta cells, resembling an insulin-resistant state. Both steroids and calcineurin inhibitors promote hypertension and hypercholesterolemia.90, 91 Furthermore, the use of high dosages of steroids early posttransplantation leads to weight gain and worsens glucose intolerance. The finding that cumulative steroid dose was independently indicative of time to disease recurrence79 gives further support to the hypothesis that insulin resistance plays a role in NAFLD recurrence/development posttransplantation.

Table 3. Prevalence of Risk Factors for the Development of NAFLD in Patients Post Liver Transplantation and in the General Population
Risk factorPrevalence posttransplantationPrevalence in the United States
  1. NOTE. Modified from reference 88.

  2. Abbreviations: BP, blood pressure; HDL, high-density lipoprotein.

Obesity (BMI > 30)39%-43%31%
Diabetes mellitus21%-32%8%
HDL-cholesterol <35 mg/dL52%12%
Hypercholesterolemia (>240 mg/dL)20%-66%14.9%
Hypertension (BP > 140/90)41%-81%15.7%

No effective medical therapy is currently available for all patients with NAFLD,92 and no studies have evaluated any treatment modalities aimed at preventing NAFLD development/recurrence or its progression to more advanced liver disease posttransplantation. In theory, however, improving insulin sensitivity would be expected to reduce the odds of disease recurrence, and to improve the disease in the posttransplantation period. The management of the associated risk factors plays an essential role. All efforts should be made to achieve and maintain appropriate weight control as well as to treat any lipid and glucose abnormalities. Lifestyle intervention with increased physical activity and diet recommendations in the posttransplantation period should constitute an essential component of patient management. Increased physical activity per se improves insulin sensitivity in muscle mass.93 Just as occurs in the nontransplant setting,94 patients with NAFLD posttransplantation who are not obese and do not suffer from components of the metabolic syndrome may still be insulin resistant, and thus increased physical activity per se is expected to improve both insulin resistance and the liver disease.

Although no drugs have proven effective for the treatment of NAFLD before liver transplantation, some small, open-label pilot studies have reported promising results with medications that improve insulin sensitivity (i.e., thiazolidinediones, metformin) and those that increase the antioxidant defenses in the liver (i.e., vitamin E, betaine).95 However, no clinical trials have been reported in the posttransplantation setting. Preliminary data suggest that the thiazolidinediones, pioglitazone and rosiglitazone, are effective and safe in the management of diabetes posttransplantation.96, 97 However, the risk-benefit of the thiazolidinediones in the treatment of NAFLD posttransplantation has to be evaluated in carefully controlled clinical trials.


NAFLD affects a substantial proportion of the general population, and in most cases, the condition follows a relatively benign course. Patients with NAFLD with advanced fibrosis or cirrhosis are at highest risk to progress to end-stage liver disease with the consequent need for liver transplantation. In patients with progressive disease, the histological features of NAFLD improve or resolve over time despite fibrosis progression, and thus most cases of cryptogenic cirrhosis are most likely unrecognized NAFLD. Cryptogenic cirrhosis and cirrhotic stage NAFLD explain a substantial proportion of the liver transplant procedures performed to date. The coexistence of NAFLD and the risk factors for NAFLD development contribute significantly to disease progression/severity in other liver diseases such as HCV infection, alcoholic liver disease, and hemochromatosis. NAFLD recurrence/development occurs in the first few months post liver transplantation in a substantial proportion of those transplanted for end-stage NAFLD or cryptogenic cirrhosis, with some patients progressing to cirrhotic stage disease. Persistence and de novo development of risk factors for NAFLD including obesity, glucose intolerance/diabetes, and dyslipidemia in the posttransplant setting suggest that insulin resistance may play a central role in disease recurrence/development posttransplantation. Further studies are necessary to determine the long-term implications of posttransplant NAFLD. Although no therapeutic trials for NAFLD posttransplantation have been reported, improving insulin resistance may improve the liver disease.