Outcomes of liver transplantation in patients with cirrhosis due to nonalcoholic steatohepatitis versus patients with cirrhosis due to alcoholic liver disease

Authors


Abstract

Nonalcoholic steatohepatitis (NASH) is becoming a common cause of liver cirrhosis requiring liver transplantation (LT). Cardiovascular complications related to metabolic syndrome and NASH recurrence in the transplanted liver may affect the outcome of LT in these patients. We compared the outcomes of LT for NASH cirrhosis and alcoholic cirrhosis (ETOH) in a large transplant center. A retrospective chart review was performed for all patients who underwent LT for cryptogenic cirrhosis with the NASH phenotype (the NASH group) or ETOH (the ETOH group) at the University of Miami from January 1997 to January 2007. There was no significant difference in survival between the NASH and ETOH groups, despite a trend toward lower survival in the former (P = 0.1699). Sepsis was the leading cause of posttransplant death in both groups, and it was followed by cardiovascular causes in the NASH group (26% versus 7% in the ETOH group, P = 0.21) and malignancies in the ETOH group (29% versus 0% in the NASH group, P = 0.024). Recurrent steatohepatitis (33% versus 0%, P < 0.0001) and acute rejection (41% versus 23%, P < 0.023) were significantly more frequent in the NASH group than in the ETOH group. There was no difference in graft failure between the groups (24% in the NASH group versus 18% in the ETOH group, P = 0.3973). In conclusion, despite a numerical trend favoring the ETOH group, there were no statistically significant differences in posttransplant survival and cardiovascular mortality between the NASH and ETOH groups. Acute rejection and recurrent steatohepatitis were significantly more frequent in the NASH group but did not lead to higher rates of retransplantation. Liver Transpl 15:1814–1820, 2009. © 2009 AASLD.

Nonalcoholic fatty liver disease (NAFLD) is the most common form of liver disease in Western countries, affecting 20% to 30% of the general population in the United States.1–4 Some of these patients develop nonalcoholic steatohepatitis (NASH), with an estimated prevalence of 3% to 5% of the general population.5, 6 NASH can progress to cirrhosis (in 22% of cases in 1 study)3 and liver failure requiring liver transplantation (LT).7, 8 Hepatic steatosis may in fact disappear after the development of cirrhosis, which may mask the diagnosis of NASH in some patients with NASH-related end-stage liver disease.9 Cryptogenic cirrhosis accounts for 7% to 14% of LT procedures in the United States,10 and there is strong evidence suggesting that NASH is the leading cause of cryptogenic cirrhosis.11, 12 The long-term (>5 years) posttransplant outcomes (graft survival and patient survival) of NASH cirrhosis have not been assessed. Recurrence of the primary disease and/or cardiovascular complications related to the underlying metabolic syndrome may affect the outcome of NASH patients after LT. Alcoholic cirrhosis (ETOH) is a well-established indication for LT and is associated with an excellent outcome if the patient remains abstinent from further alcohol use.13 In this study, using LT for ETOH as a comparator, we assessed the outcomes of LT in patients with cryptogenic cirrhosis with the NASH phenotype.

Abbreviations

BMI, body mass index; DM, diabetes mellitus; ETOH, alcoholic cirrhosis; HTN, hypertension; IGT, impaired glucose tolerance; LT, liver transplantation; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; SD, standard deviation.

PATIENTS AND METHODS

After obtaining institutional review board approval, a retrospective chart review was performed for all patients who underwent LT for cryptogenic cirrhosis or ETOH at the University of Miami from January 1997 through January 2007.

NASH Group

The NASH group included patients with cryptogenic cirrhosis with the NASH phenotype (as defined later). Cryptogenic cirrhosis was defined by the absence of significant alcohol use (>20 g/day), negative tests for viral hepatitis (including hepatitis B surface antigen and hepatitis B DNA by polymerase chain reaction and hepatitis C antibody and hepatitis C RNA by polymerase chain reaction), negative autoimmune markers such as anti-nuclear antibody and anti-mitochondrial antibody, and negative markers for hemochromatosis, Wilson disease, and alpha-1 antitrypsin deficiency. Of the patients with cryptogenic cirrhosis, the following 2 subgroups were included in the NASH group. Subgroup 1 included patients who showed cirrhosis on a liver histological specimen and had the clinical NASH phenotype as defined by the presence of 3 or more of the following components of metabolic syndrome (based on Adult Treatment Panel III of the National Cholesterol Education Program in 200514 with minor changes mentioned later) but did not have objective (biopsy or ultrasound) evidence of NASH:

  • 1Hypercholesterolemia: defined as a fasting serum total cholesterol level ≥ 200 mg % or drug treatment for hypercholesterolemia instead of a high density lipoprotein level < 40 mg % for men and < 50 mg % for women.
  • 2Hypertriglyceridemia: defined as a fasting serum triglyceride level ≥ 150 mg % or drug treatment for hypertriglyceridemia.
  • 3Obesity: defined as a body mass index (BMI) ≥ 30 instead of a waist circumference ≥ 102 cm for men and ≥ 88 cm for women.
  • 4Hypertension (HTN): defined as a blood pressure ≥ 135/85 mm Hg or drug treatment for HTN.
  • 5Impaired glucose tolerance (IGT)/diabetes mellitus (DM): defined as a fasting blood glucose level ≥ 100 mg % or drug treatment for IGT/DM.

Subgroup 2 included patients with objective evidence of NASH based on histology (a liver biopsy specimen showing moderate to severe steatosis) or imaging (ultrasound showing a fatty liver) before undergoing LT. Patients with hepatocellular carcinoma and cryptogenic cirrhosis without the NASH phenotype were excluded from the NASH group.

ETOH Group

The ETOH group included patients who had a history of significant alcohol use as defined previously, had no biochemical, serological, or histological evidence of other known causes of cirrhosis as mentioned previously, and underwent LT during the same time period from January 1997 through January 2007. Patients with hepatocellular carcinoma and ETOH with the NASH phenotype were excluded from the ETOH group.

Study Design and Follow-Up

The following baseline data were recorded prior to LT and 6 months post-LT: significant alcohol use, BMI, blood pressure, and fasting serum levels of glucose, total cholesterol, and triglycerides when available. There was no upper age limit or upper BMI limit for including patients in the 2 groups. Body weight was not corrected for the presence of ascites. Patients with previous cardiovascular disease were not excluded from the analysis, but all patients had a negative dobutamine stress echocardiogram or similar stress test before undergoing LT. Patients with significant alcohol use after LT were excluded from the study. The following characteristics were also recorded: recipient age, recipient sex, recipient ethnicity, donor age, cold ischemia time, biopsy-proven acute and/or chronic rejection, biopsy-proven recurrence of moderate to severe steatohepatitis in the transplanted liver, causes of retransplantation, graft survival, causes of death, and patient survival.

Statistical Analysis

Statistical analysis was performed with SAS 9.2. The chi-square test was used to compare proportional data. The Student t test was used to compare means between groups. A P value of <0.05 was considered statistically significant. Kaplan-Meier curves were constructed to evaluate patient survival probabilities. Survival probabilities were compared between the NASH and ETOH groups by the log-rank test. Multiple regression analysis was used to evaluate predictors of poor graft survival and patient survival. Patients in the NASH group (subgroups 1 and 2) were compared with patients in the ETOH group in analysis 1. Patients with biopsy/ultrasound-proven NASH (subgroup 2) were compared with the ETOH group in analysis 2.

RESULTS

Of a total of 1705 LT procedures performed from January 1997 through January 2007 at the University of Miami, 114 patients underwent LT for cryptogenic cirrhosis, and 166 patients underwent LT for ETOH. One hundred twenty-six patients were excluded after the application of the exclusion criteria. The remaining 71 patients in the NASH group were compared with the remaining 83 patients in the ETOH group in analysis 1 (Fig. 1). Of the 71 patients in the NASH group, only 32 had biopsy-proven (n = 25) or ultrasound-proven (n = 7) fatty liver disease (subgroup 2). This subgroup of the NASH group (n = 32) was compared with the ETOH group in analysis 2 (Fig. 1). Table 1 shows the baseline (pre-LT) characteristics of the patients in the NASH and ETOH groups. There was no difference in the mean age or median follow-up in the 2 groups. The ETOH group had more male patients. As determined by the study design, all components of metabolic syndrome were significantly more prevalent in the NASH group compared to the ETOH group. Table 2 shows the rate of new development of components of metabolic syndrome in both groups 6 months post-LT. There was no difference in the newly developed components of metabolic syndrome between the 2 groups, except for the fact that patients in the ETOH group developed HTN more often than patients in the NASH group.

Figure 1.

Flow chart of the study design and selection of the NASH and ETOH groups. Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

Table 1. Baseline Characteristics of the NASH and ETOH Groups
 NASH GroupETOH GroupP Value
  1. NOTE: Hypercholesterolemia and hypertriglyceridemia are defined in the Patients and Methods section.

  2. Abbreviations: BMI, body mass index; DM, diabetes mellitus; ETOH, alcoholic cirrhosis; HTN, hypertension; IGT, impaired glucose tolerance; NASH, nonalcoholic steatohepatitis; SD, standard deviation.

Number of patients7183 
Median follow-up (days)1517 (range: 4–3611)1686 (range: 1–3640)0.8
Recipient age (years; mean ± 1 SD)56 ± 9.5 (range: 18–74)54 ± 8.2 (range: 34–68)0.21
Recipient sex (male)46/71 (65%)67/83 (81%)0.03
Recipient ethnicity (white)48/71 (68%)63/83 (76%)0.28
Obesity42/71 (59%; BMI range: 22–40)17/83 (20%; BMI range: 20–37)0.000001
HTN41/69 (59%)13/83 (16%)<0.000001
IGT/DM35/71 (49%)9/83 (11%)<0.000001
Hypercholesterolemia25/63 (40%)0/82 (0%)<0.000001
Hypertriglyceridemia26/62 (42%)0/82 (0%)<0.000001
Donor age (years; mean ± 1 SD)43.8 ± 17.643 ± 16.60.77
Cold ischemia time (hours and minutes; mean ± 1 SD)7:06 ± 1:467:23 ± 2:020.42
Table 2. New Development of Components of Metabolic Syndrome in the NASH Group (n = 71) and the ETOH Group (n = 83) 6 Months Post-LT
ComponentNASH GroupETOH GroupP Value
  1. NOTE: Hypercholesterolemia and hypertriglyceridemia are defined in the Patients and Methods section.

  2. Abbreviations: DM, diabetes mellitus; ETOH, alcoholic cirrhosis; HTN, hypertension; IGT, impaired glucose tolerance; NASH, nonalcoholic steatohepatitis.

Obesity4/24 (17%)11/63 (18%)1.0
HTN9/26 (35%)39/64 (61%)0.04
IGT/DM5/32 (16%)20/69 (29%)0.22
Hypercholesterolemia6/34 (18%)14/72 (19%)1.0
Hypertriglyceridemia6/32 (19%)19/72 (26%)0.46

Table 3 shows the numbers and causes of deaths in the NASH and ETOH groups. Sepsis with or without multiorgan system failure was the most common cause of death post-LT, and it was followed by cardiovascular causes in the NASH group (26% versus 7% in the ETOH group, P = 0.21) and malignancies in the ETOH group (29% versus 0% in the NASH group, P = 0.024). Four patients in the NASH group had previous cardiovascular diseases versus 1 patient in the ETOH group, but none of these patients died during the follow-up period. Total death rates were not statistically different between the groups (P = 0.17). Figure 2 shows Kaplan-Meier curves of patient survival in both groups. The 1-, 3-, 5-, and 9-year patient survival probabilities were 82%, 79%, 75%, and 62% in the NASH group and 92%, 86%, 86%, and 76% in the ETOH group. There was no statistically significant difference in the survival of NASH patients compared to ETOH patients, but there was a trend toward lower survival in the NASH group (P = 0.1699). Figure 3 shows Kaplan-Meier curves of patients with ultrasound/biopsy-proven NASH versus patients with ETOH. There was no statistically significant difference in patient survival between the 2 groups (P = 0.2959). Subgroup multiregression analyses were carried out with the following variables for patient survival: recipient age, recipient sex, recipient ethnicity, etiology of liver disease, obesity, HTN, IGT/DM, hypercholesterolemia, hypertriglyceridemia, donor age, cold ischemia time, and history of biopsy-proven acute and/or chronic rejection. None of these factors affected patient survival in the NASH group. A longer cold ischemia time predicted poor patient survival in the ETOH group (P = 0.0001) but not in the NASH group (P = 0.67).

Table 3. Causes of Death in the NASH Group (n = 71) and the ETOH Group (n = 83)
CauseNASHETOHP Value
  1. Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

Sepsis with/without multisystem organ failure10 (53%)6 (43%)0.73
Cardiovascular (myocardial infarction and stroke)5 (26%)1 (7%)0.21
Malignancies0 (0%)4 (29%)0.024
Operative complications0 (0%)2 (14%)0.17
Hospice2 (11%)0 (0%)0.5
Rejection and graft failure0 (0%)1 (7%)0.42
Acute renal failure1 (5%)0 (0%)1
Suicide1 (5%)0 (0%)1
Total19/71 (26.8%)14/83 (16.9%)0.17
Figure 2.

Comparison of the patient survival probabilities for the NASH group (n = 71) and the ETOH group (n = 83). Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

Figure 3.

Comparison of the patient survival probabilities for the biopsy/ultrasound-proven NASH group (subgroup 2, n = 32) and the ETOH group (n = 83). Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

Twenty-one of the 64 patients (33%) in the NASH group showed moderate to severe recurrence of steatohepatitis in liver biopsy specimens obtained at any time after 6 months post-LT, whereas none of the 77 patients (0%) in the ETOH group did (P < 0.000001). None of the patients in the NASH group developed cirrhosis or underwent retransplantation because of recurrent disease during the follow-up period. In the NASH group, a significantly higher number of patients had biopsy-proven acute rejection, as shown in Table 4 (P = 0.023). Four patients in the NASH group and 3 patients in the ETOH group underwent re-LT, as shown in Table 5. Seventeen patients (24%) in the NASH group and 15 (18%) in the ETOH group were diagnosed with graft failure over the period of follow-up, but there was no statistically significant difference in graft survival between the groups (P = 0.3973). Subgroup multiregression analyses were carried out with the following variables for graft survival: recipient age, recipient sex, recipient ethnicity, etiology of liver disease, obesity, HTN, IGT/DM, hypercholesterolemia, hypertriglyceridemia, donor age, cold ischemia time, and history of biopsy-proven acute and/or chronic rejection. A longer cold ischemia time (P = 0.0001) and a history of biopsy-proven rejection (P = 0.03) predicted poor graft survival in the ETOH group. A history of biopsy-proven rejection predicted poor graft survival in the NASH group (P = 0.0024).

Table 4. Acute and Chronic Rejection in the NASH Group (n = 71) and the ETOH Group (n = 83)
GroupAcuteChronicTotal
  1. Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

NASH29/71 (41%)3/71 (4%)32/71 (45%)
ETOH19/83 (23%)2/83 (2%)21/83 (25%)
P value0.0230.660.011
Table 5. Causes of Retransplantation in the NASH Group (n = 71) and the ETOH Group (n = 83)
CauseNASHETOH
  1. Abbreviations: ETOH, alcoholic cirrhosis; NASH, nonalcoholic steatohepatitis.

Hepatic artery thrombosis10
Primary nonfunction23
Chronic rejection10
Total (P = 0.023)4/71 (5.6%)3/83 (3.6%)

DISCUSSION

NAFLD is an increasingly important cause of end-stage liver disease and is likely the leading cause of cryptogenic cirrhosis in Western countries.11 In the absence of a diagnostic test for NAFLD, the diagnosis is often based on a clinical background of metabolic syndrome and the presence of hepatic steatosis or steatohepatitis in a patient with cryptogenic cirrhosis.15 There is compelling evidence to suggest that insulin resistance has a central role in the pathogenesis of NAFLD, which is closely associated with metabolic syndrome.16–18 The components of metabolic syndrome strongly correlate with the development of cardiovascular diseases, which may account for the poor outcomes of LT in patients with NASH cirrhosis.10 Moreover, the liver disease often recurs in the transplanted livers of NAFLD patients, and some develop NASH and consequently cirrhosis requiring retransplantation.12, 19, 20 The prognosis and post-LT outcomes of NASH cirrhosis have not been well defined; hence, we performed a retrospective analysis of NASH patients who underwent LT at our center from January 1997 through January 2007. ETOH is a common indication for LT and is associated with very good post-LT outcomes if the patient remains abstinent from further alcohol use.13 One recent large study showed that survival rates following LT in patients with ETOH are similar to the rates for patients with other causes of cirrhosis (89% versus 84% and 84% versus 81% at 3 and 5 years, respectively).21

Our results show that although there was a trend toward lower patient survival in the NASH group compared to the ETOH group, the difference was not statistically significant. A relatively small number of patients may have accounted for this finding. These results are similar to those published previously. Five-year survival rates post-LT in adults with cryptogenic cirrhosis range from 72% to 81%.10, 22 Sanjeevi et al.20 and Heneghan et al.22 showed that short-term and 5-year post-LT graft survival and patient survival for patients with cryptogenic cirrhosis were comparable to those for patients with other causes of cirrhosis. Contos et al.19 showed that post-LT graft survival and patient survival for patients with cryptogenic cirrhosis and the NASH phenotype were similar to those for patients with ETOH and primary biliary cirrhosis.19 Obesity and DM have been shown to be independent predictors of poor outcomes following LT.23, 24 However, in our study, these factors did not affect patient survival in the NASH group.

In both the NASH and ETOH groups, sepsis (with or without multisystem organ failure) was the most common cause of death post-LT, and this is consistent with previously published data25; sepsis was followed by cardiovascular causes in the NASH group and malignancies in the ETOH group. Although a higher number of patients died from cardiovascular causes in the NASH group (26%) compared with the ETOH group (7%), the difference did not reach statistical significance (P = 0.21). Nevertheless, cardiovascular mortality in our NASH group was much higher than that based on published autopsy results for LT patients.25 In a recently published study, major risk factors for atherosclerosis were significantly more prevalent in subjects with NASH cirrhosis than in subjects with cirrhosis of other etiologies.26 These findings stress the importance of controlling metabolic syndrome post-LT in patients transplanted for NASH cirrhosis.

Thirty-three percent of the patients in the NASH group showed the development of steatohepatitis in biopsy specimens taken at any time after 6 months post-LT. However, none of these biopsy specimens showed advanced fibrosis (stage 3) or cirrhosis. This finding could be an underestimation because we do not perform protocol biopsy in our center. Moreover, the small number of patients precluded predictive factor analysis for recurrent NASH. None of the NASH patients underwent retransplantation for recurrent disease during the follow-up period. However, it is conceivable that recurrence of the underlying disease may affect the long-term outcomes of these patients. This finding again stresses the importance of aggressive control of metabolic syndrome post-LT. A recent report suggested that the use of an angiotensin-converting enzyme inhibitor may reduce the risk of developing posttransplant NAFLD.27

Acute rejection episodes were significantly more frequent in the NASH group compared to the ETOH group. Previous studies have reported similar acute rejection rates in patients transplanted for cryptogenic cirrhosis (but not NASH) and ETOH.22 Despite higher acute rejection rates, graft survival and retransplantation rates were not different between the 2 groups in our study. It is unclear whether the difference in the acute rejection incidence was related to the presence of metabolic syndrome or to differences in the type and dose of immunosuppressant therapy, as the study was not designed to collect information on immunosuppressant therapy. It is important to keep a balance between optimal immunosuppression and control of metabolic syndrome because most immunosuppressants (especially steroids and calcineurin inhibitors) are known to worsen metabolic syndrome. Moreover, published data suggest that the cumulative steroid dose correlates with the time to development of allograft steatosis in patients transplanted for NASH cirrhosis.19

Our study presents the largest patient population with the longest follow-up published so far with respect to the survival of patients with NASH cirrhosis. With the use of a control group (ETOH), we have shown that NASH cirrhosis is an acceptable indication for LT, with graft and patient survival rates comparable to those for other causes of cirrhosis. Although there were higher rates of acute rejections and recurrent steatohepatitis in the NASH group, graft survival and retransplantation rates were not different in the NASH and ETOH groups. There are some limitations to our study. First, it is a retrospective chart review with a relatively small number of patients. We considered those patients with 3 components of metabolic syndrome to have a clinical phenotype of NASH in our database, but the presence of this phenotype is not diagnostic of NASH cirrhosis. Moreover, we used BMI (without correcting the body weight for the presence of ascites) instead of waist circumference to define obesity and total cholesterol instead of high density lipoprotein in our definition of metabolic syndrome, and we did not have data on insulin resistance in our patients. We did not perform protocol biopsy after LT in our patients, and hence it is difficult to comment on recurrent disease in the transplanted liver. We did not collect information about immunosuppressant therapy. Differences in immunosuppression may have affected rejection rates, graft survival, and patient survival.

In summary, cryptogenic cirrhosis with the NASH phenotype is an acceptable indication for LT with posttransplant survival rates similar to those for other causes of cirrhosis. Aggressive control of metabolic syndrome may have some role in the post-LT management of these patients not only to prevent disease recurrence in the transplanted liver but also to prevent cardiovascular mortality. Optimal immunosuppression is also important for preventing rejection of the transplanted liver while controlling metabolic syndrome in these patients. Large, randomized, multicenter, prospective studies are needed to improve our understanding of the natural history of patients undergoing LT for NASH.

Ancillary