Excellent posttransplant survival for patients with nonalcoholic steatohepatitis in the United States

Authors

  • Anita Afzali,

    Corresponding author
    1. Department of Medicine, Division of Gastroenterology, Veterans Affairs Puget Sound Health Care System and University of Washington, Seattle, WA, University of Washington, Seattle, WA
    2. Research Enhancement Award Program, Veterans Affairs Puget Sound Health Care System, Seattle, WA
    • Division of Gastroenterology, Department of Medicine, University of Washington, 1959 Northeast Pacific Street, Suite AA103, Box 356424, Seattle, WA 98195-6424
    Search for more papers by this author
    • Telephone: 206-685-3195; FAX: 206-685-8684

  • Kristin Berry,

    1. Department of Biostatistics, University of Washington, Seattle, WA
    Search for more papers by this author
  • George N. Ioannou

    1. Department of Medicine, Division of Gastroenterology, Veterans Affairs Puget Sound Health Care System and University of Washington, Seattle, WA, University of Washington, Seattle, WA
    2. Research Enhancement Award Program, Veterans Affairs Puget Sound Health Care System, Seattle, WA
    Search for more papers by this author

  • Anita Afzali M.D., M.P.H., contributed to the study concept and design, the acquisition of data, the statistical analysis and interpretation of data, the drafting of the manuscript, and a critical revision of the manuscript for important intellectual content. Kristin Berry P.H.D., contributed to the study concept and design, the acquisition of data, and the statistical analysis and interpretation of data. George N. Ioannou M.D., M.S., contributed to the study concept and design, the analysis and interpretation of data, and a critical revision of the manuscript for important intellectual content.

  • This work was supported in part by the Health Resources and Services Administration (contract 234-2005-370011C). The content is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.

Abstract

Because of the ongoing epidemics of obesity and diabetes, nonalcoholic steatohepatitis (NASH) may become a leading indication for liver transplantation. There are concerns about the posttransplant survival of patients with NASH because of associated cardiovascular and metabolic risk factors. We aimed to determine recent trends in the proportion of patients undergoing transplantation for NASH-related cirrhosis in the United States and to estimate their posttransplant survival. We used data provided by the United Network for Organ Sharing for first-time adult cadaveric liver transplants performed in the United States between January 1, 1997 and October 31, 2010 (n = 53,738). The proportion of liver transplants performed for NASH-related cirrhosis increased dramatically from 1.2% in 1997-2003 to 7.4% in 2010 when NASH was the fourth most common indication for transplantation. The posttransplant survival of patients with NASH (n = 1810) at 1 (87.6%), 3 (82.2%), and 5 years (76.7%) was superior to the survival of patients with hepatocellular carcinoma, hepatitis C virus, alcoholic liver disease, acute hepatic necrosis, hemochromatosis, or cryptogenic liver disease and was inferior to the survival of only 4 groups of patients (those with primary biliary cirrhosis, primary sclerosing cholangitis, autoimmune hepatitis, or hepatitis B virus). In conclusion, NASH-related cirrhosis is increasing rapidly as an indication for liver transplantation in the United States and is associated with excellent posttransplant survival. Liver Transpl 18:29–37, 2012. © 2011 AASLD.

Nonalcoholic fatty liver disease (NAFLD), which is defined as the excessive deposition of lipids within hepatocytes in the absence of alcohol consumption, is the most common liver disease in the United States and affects nearly 30% of the general population.1, 2 It has been estimated that 15% to 20% of patients with NAFLD have nonalcoholic steatohepatitis (NASH), which is defined histologically by the presence of hepatic inflammation, ballooning degeneration, and potentially also fibrosis.3, 4

Central obesity and insulin resistance are the most important risk factors for NAFLD/NASH.5 It is thought that with the increasing prevalence of obesity, insulin resistance, and diabetes, the prevalence of NAFLD, NASH, and NASH-related cirrhosis will continue to increase and affect the need for liver transplantation. By the year 2025, it is estimated that more than 25 million Americans may have NASH,6 and NASH may progress to cirrhosis, hepatocellular carcinoma (HCC), and liver failure in approximately 20% of these patients.3, 4, 7-9 Some investigators predict that NASH may become the leading indication for liver transplantation in the next 10 to 20 years and surpass the currently most common indication for transplantation: hepatitis C virus (HCV) infection.10 However, no national data are available for recent trends in the proportion of patients undergoing transplantation in the United States who have NASH-related cirrhosis as the underlying etiology.

Hepatic steatosis often regresses after the development of NASH-related cirrhosis.5, 6 Therefore, some patients with underlying NASH may be misdiagnosed with cryptogenic cirrhosis (CC),6, 11, 12 and this could potentially lead to an underestimation of the number of liver transplants performed for NASH-related cirrhosis.

There is a high incidence of cardiovascular morbidity and mortality in patients with NAFLD/NASH because of the associated high prevalence of adverse cardiovascular and metabolic risk factors in this group.13-15 These known risk factors may worsen posttransplant survival. On the other hand, the posttransplant survival of patients with NASH may be expected to be good because there is a lower likelihood of disease recurrence, and this may mean lower rates of graft failure in comparison with other liver conditions such as HCV infection. Furthermore, adverse psychosocial factors may be less common in patients with NASH versus patients with other liver diseases such as HCV and alcoholic liver disease.

There are limited data on the long-term posttransplant survival of patients with NASH-related cirrhosis. Using data reported to the United Network for Organ Sharing (UNOS) for all liver transplants performed in the United States from January 1, 1997 to October 31, 2010, we compared the posttransplant survival of patients with NASH-related cirrhosis or CC to the survival of patients with other indications for liver transplantation. Additionally, we assessed the absolute number of orthotopic liver transplants performed for NASH and CC over time and determined whether the proportion of transplants for NASH and CC has increased with time.

PATIENTS AND METHODS

Data Collection

All transplantation centers and organ procurement organizations in the United States are required to electronically submit standardized collection forms to UNOS. These forms include the Transplant Candidate Registration Form, which contains patient information at the time of listing for liver transplantation; the Deceased Donor Registration Form, which includes information on donors; the Transplant Recipient Registration Form, which includes clinical information about the patient before and after transplantation, treatment information, and the patient's status at hospital discharge; and the Transplant Recipient Follow-Up Form, which includes information about the patient on each subsequent transplantation anniversary as well as clinical and treatment information and the patient's status. All this information is entered into a single Standard Transplant Analysis and Research file, which includes 1 record per transplant and the most recent follow-up information for the patient's status as of the date on which the file was created. We obtained a Standard Transplant Analysis and Research file from UNOS that included data up to October 31, 2010.

This study was approved by the Veteran Affairs Puget Sound Health Care System Institutional Review Committee.

Study Population

In all, 69,962 liver transplants were performed for recipients older than 18 years of age in the United States from January 1, 1997 to October 31, 2010. Patients with donors less than 10 years old or more than 75 years old (n = 1689), patients with living donors (n = 2881), patients with split liver donors (n = 1047), patients with non–heart-beating donors (n = 2128), patients with multiple simultaneous organ transplants (n = 3634), and patients with previous liver transplants (n = 4845) were excluded; this left 53,738 subjects for our univariate analyses.

For our assessment of the posttransplant survival of patients with NASH or CC versus patients with all other causes of liver disease, we also explored whether differences in either donor or recipient characteristics could contribute to survival. For these multivariate analyses, we additionally excluded patients with missing information for the cold ischemia time (n = 5607), donor body mass index (BMI; calculated as kilograms per meter squared; n = 64), recipient BMI (n = 961), serum albumin level (n = 1060), total bilirubin level (n = 1633), serum creatinine level (n = 473), and recipient diabetes status (n = 1594). We also excluded patients with very abnormal values (possibly erroneous recordings) or implausible values (almost certainly erroneous recordings) for the cold ischemia time (n = 456), recipient or donor BMI (n = 536), serum albumin level (n = 39), total bilirubin level (n = 331), and serum creatinine level (n = 10). Very abnormal values were defined as follows: a cold ischemia time < 1 or > 24 hours, a BMI < 15 or > 55 kg/m2, a serum albumin level < 0.5 or > 6 g/dL, a total bilirubin level < 0.1 or > 50 mg/dL, and a creatinine level < 0.1 or > 15 mg/dL. Implausible values were considered even more extreme than the abnormal values. The international normalized ratio (INR) was missing for a very large number of recipients (n = 12,713). Therefore, we performed an adjusted analysis with the INR (n = 33,995) and without the INR (n = 43,255). The inclusion of the INR had little effect on the results, so we present the results without the INR in multivariate models to allow the use of a much larger and more representative sample of liver transplant recipients.

Definitions of NASH, CC, and Other Liver Diseases

The primary diagnosis leading to liver transplantation was identified by UNOS with the variable LI_DGN and was determined at hospital discharge after liver transplantation had been performed. On the basis of this primary diagnosis, the recipients were divided into the following categories: NASH, CC, HCV [including HCV with alcohol-related cirrhosis and HCV with hepatitis B virus (HBV)], HBV, alcohol-related cirrhosis, primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), hemochromatosis, autoimmune hepatitis, acute hepatic necrosis, HCC, and other. Before 2004, a diagnostic code was unavailable for NASH, and it was documented in the UNOS database as a text code only. Therefore, in addition to the diagnostic codes, we also looked for text codes for NASH in order to capture all NASH transplant recipients during the time period from 1997 to 2003.

Because patients with HCC have particularly poor posttransplant survival, we listed patients in the HCC category not only if they had a primary diagnosis of HCC but also if they had a secondary diagnosis of HCC or received a Model for End-Stage Liver Disease (MELD) exception on account of HCC, regardless of their primary diagnosis.

Predictors of Posttransplant Survival

The posttransplant survival of patients with CC or NASH was compared to the survival of patients with other causes of liver disease with or without adjustments for donor and recipient characteristics that could be predictors of posttransplant survival.16 These characteristics included the following: the donor age, race, sex, and BMI; the cold ischemia time; and the recipient age, race, sex, BMI, serum albumin level, total bilirubin level, serum creatinine level, and diabetes status.

Causes of Death

The primary cause of posttransplant death was identified in the UNOS database with the variable COD. This variable was modeled as a categorical variable and was divided into the following categories: cardiovascular complications, graft failure, infection, malignancy, multiorgan failure, and other.

Statistical Analysis

Statistical analysis was performed with Stata/SE 11 software (Stata Corp., College Station, TX). A chi-square test was used to compare categorical variables between groups (NASH versus all other causes of liver diseases, CC versus all others, and CC versus NASH). A t test was used to compare continuous variables between these groups. A P value < 0.05 was considered statistically significant. Patient survival 1, 3, and 5 years after transplantation was estimated with the Kaplan-Meier method. A Cox proportional hazards regression analysis was used to compare patients with NASH, CC, or other liver diseases with respect to patient and graft survival after transplantation with and without adjustments for the predictors of posttransplant survival discussed earlier.

Patient death was determined with the variable PSTATUS. Patients were censored when they were last determined to be alive. Graft failure was defined as liver failure, with or without the need for retransplantation, or death from any cause. This was identified with the variable GSTATUS. Patients who remained alive and did not have liver failure were censored when they were last determined to be alive.

Abbreviations:

BMI, body mass index; CC, cryptogenic cirrhosis; CI, confidence interval; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NS, not significant; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; UNOS, United Network for Organ Sharing.

RESULTS

Only a very small number of transplants (n = 279 or 1.2% of all transplants) were undertaken for NASH-related cirrhosis in the 6-year period of 1997-2003 (Fig. 1). However, the number and proportion of liver transplants performed for NASH steadily increased annually since 2003, and the proportion reached 7.4% in 2010. Before 2004, a diagnostic code was unavailable for NASH, and it was documented in the UNOS database as a text code only. Even when we assessed all text codes for NASH during the time period of 1997-2003, there was a slow linear upward trend in the percentage of liver transplants performed for NASH-related cirrhosis. The proportion of transplants performed annually for CC decreased slightly from 2003 (6.6%) to 2010 (5.2%). Since 2008, liver transplants for NASH exceeded those performed for CC. The proportion of liver transplants performed for NASH and CC combined also increased steadily from 2003 onward, and this suggests that the increase in NASH-related transplantation was not simply due to a diagnostic transfer from CC to NASH. In 2010, NASH was the fourth most common cause of liver transplantation after HCC (34.3%), HCV (22.1%), and alcohol-related liver disease (11.1%).

Figure 1.

Annual percentages of liver transplants performed in the United States from 1997 to 2010.

Patients with NASH or CC were older and were more likely to be white, female, obese, and diabetic in comparison with transplant recipients with other causes of liver disease (Table 1). The serum albumin, creatinine, and total bilirubin levels, INRs, and MELD scores were similar for recipients with NASH or CC and recipients without NASH or CC. Also, the donor characteristics were similar for recipients with NASH, CC, or other causes of liver disease.

Table 1. Select Donor and Recipient Characteristics of Liver Transplantation for NASH, CC, and Other Liver Diseases Between 1997 and 2010 (n = 53,738)
CharacteristicNASH (n = 1810)CC (n = 3843)Other (n = 48,085)P Value
NASH Versus All OthersCC Versus All OthersCC Versus NASH
  • NOTE: Missing, abnormal, and implausible values were excluded for the cold ischemia time; the donor and recipient BMIs; the serum albumin, total bilirubin, and serum creatinine levels; and the recipient diabetes status. P values < 0.05 were considered statistically significant.

  • *

    The data are presented as means and standard deviations.

Donors    
 Age (years)*43.8 ± 17.041.7 ± 17.240.5 ± 16.6<0.001<0.001<0.001
 Cold ischemia time (hours)*7.4 ± 2.97.7 ± 3.07.5 ± 3.0NS<0.01<0.01
 Less than 15 hours (%)98.397.597.9 
 Race/ethnicity (%)    
  White71.373.170.1NS<0.001NS
  Black/African American16.114.114.4NSNS0.048
  Hispanic9.69.812.1<0.01<0.001NS
  Asian/Pacific Islander2.02.12.4NSNSNS
  Other/unknown1.00.91.0NSNSNS
 Female (%)41.441.440.5NSNSNS
 BMI (kg/m2)*27.5 ± 6.126.4± 5.826.4 ± 5.6 
  Normal: <24 kg/m2 (%)30.237.237.3<0.001NS<0.001
  Overweight (%)41.941.141.5NSNSNS
  Obese: >30 kg/m2 (%)27.921.721.2<0.001NS<0.001
 Diabetes (%)11.68.38.4<0.001NS0.001
Recipients    
 Age (years)*57.6 ± 8.356.0 ± 9.951.8 ± 9.9<0.001<0.001<0.001
 Race/ethnicity (%)    
  White87.779.173.6<0.001<0.001<0.001
  Black/African American1.94.58.9<0.001<0.001<0.001
  Hispanic8.813.511.9<0.0010.001<0.001
  Asian/Pacific Islander1.12.34.7<0.001<0.0010.001
  Other/unknown0.50.60.9NSNSNS
 Female (%)47.143.431.5<0.001<0.0010.01
 BMI (kg/m2)*33.0 ± 5.930.0 ± 5.828.2 ± 5.5 
  Normal: <24 kg/m2 (%)5.615.322.7<0.001<0.001<0.001
  Overweight (%)25.637.845.1<0.001<0.001<0.001
  Obese: >30 kg/m2 (%)68.846.932.2<0.001<0.001<0.001
 Diabetes (%)50.235.618.1<0.001<0.001<0.001
 Serum albumin (g/dL)*2.9 ± 0.62.9 ± 0.72.9 ± 0.7NSNSNS
 Total bilirubin (mg/dL)*6.3 ± 7.96.8 ± 8.87.1 ± 9.3<0.001NS0.02
 Serum creatinine (mg/dL)*1.6 ± 1.11.4 ± 1.01.3 ± 1.0<0.001<0.001<0.001
 INR*1.8 ± 0.81.9 ± 1.41.9 ± 1.70.04NS0.01
 MELD score*21.2 ± 7.921.3 ± 8.420.1 ± 9.4<0.0010.03<0.001

There were 53,738 transplant recipients from 1997 to 2010: 13,990 of these recipients (26.0%) died, and 17,267 (32.1%) experienced graft failure during a mean follow-up of 4.1 years. In comparison with recipients with other causes of liver disease, those with NASH or CC did not have an increased risk of death [unadjusted hazard ratio for NASH = 0.94, 95% confidence interval (CI) = 0.84-1.05; unadjusted hazard ratio for CC = 0.97, 95% CI = 0.91-1.03] or graft failure (unadjusted hazard ratio for NASH = 0.89, 95% CI = 0.80-0.98; unadjusted hazard ratio for CC = 0.95, 95% CI = 0.90-1.01). Similarly, recipients with NASH and CC (combined) did not have an increased risk of death (unadjusted hazard ratio = 0.96, 95% CI = 0.91-1.01) or graft failure (unadjusted hazard ratio = 0.93, 95% CI = 0.89-0.98) in comparison with recipients with other causes of liver disease.

Adjustments for donor characteristics alone had little impact on these hazard ratios. However, when adjustments were made for both donor and recipient characteristics, there were decreased risks of death and graft failure for recipients with NASH (adjusted hazard ratio for death = 0.75, 95% CI = 0.66-0.85; adjusted hazard ratio for graft failure =.76, 95% CI = 0.68-0.85) or CC (adjusted hazard ratio for death = 0.86, 95% CI = 0.80-0.93; adjusted hazard ratio for graft failure = 0.90, 95% CI = 0.84-0.96) in comparison with recipients with other liver diseases. There were also decreased risks of death and graft failure in recipients with NASH and CC (combined) when adjustments were made for donor and recipient characteristics (adjusted hazard ratio for death = 0.84, 95% CI = 0.78-0.89; adjusted hazard ratio for graft failure = 0.86, 95% CI = 0.81-0.91; Table 2).

Table 2. Comparison of the Posttransplant Survival of Recipients With or Without NASH or CC From 1997 to 2010
 Patients (n)Patient Years (n)Patient Deaths (n)Mortality per 100 Patient YearsUnadjusted Hazard Ratio (95% CI) for DeathAdjusted Hazard Ratio (95% CI) for Death*Adjusted Hazard Ratio (95% CI) for Death
  • *

    Adjusted for donor characteristics (including the donor age, race, sex, and BMI and the cold ischemia time).

  • Adjusted for donor and recipient characteristics (including the donor age, race, sex, and BMI; the cold ischemia time; and the recipient age, race, sex, BMI, diabetes status, total bilirubin level, serum creatinine level, and serum albumin level).

NASH181046633297.10.94 (0.84-1.05)0.89 (0.79-1.01)0.75 (0.66-0.85)
CC384317,45510626.10.97 (0.91-1.03)0.97 (0.91-1.04)0.86 (0.80-0.93)
NASH plus CC565322,11813916.30.96 (0.91-1.01)0.95 (0.90-1.01)0.84 (0.78-0.89)
Other48,085192,70812,5996.51.01.01.0

Very few patients underwent transplantation for NASH cirrhosis from 1997 to 2003 (279/24,027 or 1.2%). Therefore, we also assessed the period from 2004 to 2010 separately because the vast majority of transplants for NASH (1531/1810) occurred in this period. Those with NASH or CC did not have an increased risk of death (unadjusted hazard ratio for NASH = 0.96, 95% CI = 0.84-1.09; unadjusted hazard ratio for CC = 0.96, 95% CI = 0.86-1.07) or graft failure (unadjusted hazard ratio for NASH = 0.91, 95% CI = 0.81-1.02; unadjusted hazard ratio for CC = 0.97, 95% CI = 0.88-1.07) in comparison with recipients with other causes of liver disease in this time period. Similarly, recipients with NASH and CC (combined) did not have an increased risk of death (unadjusted hazard ratio = 0.96, 95% CI = 0.88-1.04) or graft failure (unadjusted hazard ratio = 0.95, 95% CI = 0.88-1.02) in comparison with recipients with other causes of liver disease.

Again, there was little effect when adjustments were made for donor characteristics alone. However, when adjustments were made for both donor and recipient characteristics, there were decreased risks of death and graft failure for recipients with NASH (adjusted hazard ratio for death = 0.79, 95% CI = 0.68-0.91; adjusted hazard ratio for graft failure = 0.80, 95% CI = 0.70-0.91) or CC (adjusted hazard ratio for death = 0.79, 95% CI = 0.70-0.89; adjusted hazard ratio for graft failure = 0.85, 95% CI = 0.77-0.95). When adjustments were made for donor and recipient characteristics, recipients with NASH and CC (combined) also had decreased risks of death and graft failure (adjusted hazard ratio for death = 0.79, 95% CI = 0.72-0.87; adjusted hazard ratio for graft failure = 0.83, 95% CI = 0.76-0.91; Table 3).

Table 3. Comparison of the Posttransplant Survival of Recipients With or Without NASH or CC From 2004 to 2010
 Patients (n)Patient Years (n)Patient Deaths (n)Mortality per 100 Patient YearsUnadjusted Hazard Ratio (95% CI) for DeathAdjusted Hazard Ratio (95% CI) for Death*Adjusted Hazard Ratio (95% CI) For Death
  • *

    Adjusted for donor characteristics (including the donor age, race, sex, and BMI and the cold ischemia time).

  • Adjusted for donor and recipient characteristics (including the donor age, race, sex, and BMI; the cold ischemia time; and the recipient age, race, sex, BMI, diabetes status, total bilirubin level, serum creatinine level, and serum albumin level).

NASH153129652428.20.96 (0.84-1.09)0.93 (0.81-1.07)0.79 (0.68-0.91)
CC181145133437.60.96 (0.86-1.07)0.94 (0.83-1.05)0.79 (0.70-0.89)
NASH plus CC334274785857.80.96 (0.88-1.04)0.93 (0.85-1.02)0.79 (0.72-0.87)
Other26,36961,53649908.11.01.01.0

A separate analysis was also performed to assess patient and graft survival when recipients with HCC were excluded from the other category because patients with HCC have the worst posttransplant survival and may have a variety of underlying liver diseases (including NASH). Again, in this analysis, recipients with NASH or CC did not have a higher risk of death (unadjusted hazard ratio for NASH = 0.96, 95% CI = 0.86-1.08; unadjusted hazard ratio for CC = 1.02, 95% CI = 0.95-1.08) or graft failure (unadjusted hazard ratio for NASH = 0.90, 95% CI = 0.81-1.00; unadjusted hazard ratio for CC = 0.98, 95% CI = 0.93-1.04). After adjustments for both donor and recipient characteristics, there were decreased risks of death and graft failure for recipients with NASH (adjusted hazard ratio for death = 0.76, 95% CI = 0.67-0.87; adjusted hazard ratio for graft failure = 0.78, 95% CI = 0.69-0.87) or CC (adjusted hazard ratio for death = 0.90, 95% CI = 0.84-0.9; adjusted hazard ratio for graft failure = 0.93, 95% CI = 0.87-0.99) in comparison with recipients with other causes of liver disease. Similarly, recipients with NASH and CC (combined) had decreased risks of death and graft failure after adjustments for both donor and recipient characteristics (adjusted hazard ratio for death = 0.87, 95% CI = 0.81-0.93; adjusted hazard ratio for graft failure = 0.89, 95% CI = 0.84-0.94).

The posttransplant survival of recipients with NASH and CC (combined) from 1997 to 2010 at 1 (87.1%), 3 (81.1%), and 5 years (75.2%) was comparable to the posttransplant survival of recipients with all other causes of liver disease (88.2% at 1 year, 79.5% at 3 years, and 73.1% at 5 years). In Table 4, the causes of liver disease are arranged in the order of descending 5-year survival. Recipients with NASH had a 5-year posttransplant survival rate that was superior to the rates of transplant recipients with HCC (66.7%), HCV (70.2%), acute hepatic necrosis (74.0%), CC (74.6%), alcoholic liver disease (75.6%), or hemochromatosis (75.7%; Fig. 2).

Table 4. Patient Survival After Transplantation for Select Liver Diseases From 1997 to 2010
Liver Disease1-Year Survival (%)3-Year Survival (%)5-Year Survival (%)
  1. NOTE: The diseases are listed in the order of decreasing 5-year survival.

PSC93.688.084.2
PBC90.586.382.6
HBV89.284.281.5
Autoimmune hepatitis88.183.878.9
NASH87.682.276.7
Hemochromatosis84.578.875.7
Alcohol-related88.981.775.6
Other87.181.275.3
CC86.880.674.6
Acute hepatic necrosis82.478.274.0
HCV87.377.670.2
HCC88.475.266.7
Figure 2.

Kaplan-Meier curves of posttransplant survival for recipients with NASH or other select causes of liver disease from 1997 to 2010.

In comparison with recipients with other causes of liver disease, recipients with NASH-related cirrhosis were more likely to die of cardiovascular complications (19.0% versus 12.3% for other liver diseases) and were less likely to die of graft failure (8.6% versus 16.6% for other liver diseases). However, the primary cause of death was missing (n = 1499) or unknown (n = 2020) for approximately 25% of the total deaths (n = 13,990) in our analysis.

DISCUSSION

The proportion of liver transplants performed for NASH increased dramatically from 1.2% in 1997-2003 to 7.4% in 2010. NASH is currently the fourth most common indication for liver transplantation in the United States, and the proportion of liver transplant recipients with NASH appears to be continuing to increase. The posttransplant survival rates of patients with NASH at 1 (87.6%), 3 (82.2%), and 5 years (76.7%) are excellent. As an indication for liver transplantation, NASH cirrhosis ranks fifth in posttransplant survival behind PSC, PBC, HBV, and autoimmune hepatitis and is in fact superior to the most common indications for transplantation, including HCC, HCV, acute hepatic necrosis, alcoholic liver disease, hemochromatosis, and CC.

Before our study, only single-center studies had reported the posttransplant survival of patients with NASH, and the numbers were relatively small. The survival of patients with NASH (n = 18) or CC (n = 239) was found to be comparable to the survival of recipients with other causes of liver disease (n = 1795) in a study from Baylor University (1986-2004).17

In another single-center study, NASH recipients constituted 4.8% (98/2021) of the liver transplants performed from 1997 to 2008 at the University of Pittsburgh.7 Although the posttransplant survival rates were similar, there were higher rates of early mortality in the transplant recipients with NASH (30-day mortality = 6.1%, 1-year mortality = 21.4%) versus the recipients with other liver diseases. The recipients with NASH who died within the first year were more likely to be older (age ≥ 60 years) and obese (BMI ≥ 30 kg/m2) and to have pretransplant diabetes and hypertension.

In our unadjusted analyses, recipients with NASH or CC had survival comparable to that of recipients with other causes of liver disease in the United States from 1997 to 2010 and during a more recent period (2004-2010), which included the vast majority of transplants performed for NASH. We adjusted for donor characteristics in order to account for potential differences in the quality of donor livers allocated to patients with NASH or CC versus patients with other liver diseases. For example, if livers from younger donors had been preferentially allocated to patients with NASH, this would have led to improved survival for patients with NASH (and masked otherwise worse expected survival). However, there was little difference in donor characteristics, so the effect of adjustments for donor characteristics was small. After additional adjustments for recipient characteristics, patients with NASH (and, to a lesser extent, patients with CC) actually had better posttransplant survival than other patients. This means that a patient with NASH is expected to have better survival than a patient without NASH who is the same with respect to age, race, sex, BMI, diabetes status, serum bilirubin, serum creatinine, serum albumin, and cold ischemia time.

Why do patients with NASH have posttransplant survival similar to that of patients without NASH, even though they are older, are more obese, and are more likely to be diabetic? One explanation suggested by our data is that they are less likely to die of graft failure. Only 8.6% of deaths in patients with NASH were caused by graft failure, whereas 16.6% of deaths in patients without NASH were caused by graft failure. This was likely due to the lower rates of NASH and cirrhosis recurrence in transplanted livers versus the recurrence of other diseases such as HCV and HBV.

Another possible explanation is that transplant candidates undergo very extensive screening for cardiovascular disease, which includes repeated cardiac echocardiography, stress testing, and frequently cardiac catheterization. Thus, patients with substantial cardiovascular disease, who represent a greater proportion of patients with NASH versus patients without NASH, are identified and excluded from transplantation. Still, we have found that cardiovascular deaths constitute a higher proportion of deaths in patients with NASH versus patients with other diseases (19.0% versus 12.3%). However, this is likely counterbalanced by a lower proportion of deaths due to other causes such as graft failure, so the overall posttransplant survival is similar. The careful selection of patients with NASH before transplantation (in order to identify and potentially exclude those with substantial cardiovascular disease) and the aggressive management of cardiovascular risk factors after transplantation should continue to be emphasized.

One of the most important findings of our study is the dramatic increase in the number and proportion of patients undergoing transplantation annually with NASH-related cirrhosis (Fig. 1). Although NASH-related cirrhosis was an extremely rare indication for transplantation in the United States up to 2002, it has been increasing in a linear fashion annually as a proportion of all transplants since 2003, and the proportion reached 7.4% of all liver transplants in 2010. Some diagnostic transfer from CC to NASH might have occurred since 2003 because of the greater awareness of NASH and the potential for NASH patients to present with little steatosis in the setting of advanced cirrhosis. However, the fact that the number of transplants performed for NASH and CC combined has also been increasing steadily since 2003 suggests that there has been a real increase in liver transplants performed for NASH since 2003. This is consistent with the ongoing epidemics of known risk factors for NASH in the United States, including obesity and diabetes.18 The proportion of patients undergoing transplantation for NASH is expected to continue to rise for many years to come and, at the rate shown in Fig. 1, might exceed the proportion of patients undergoing transplantation for alcohol-related liver disease in approximately 3 to 5 years.

Caldwell et al.19 first proposed that CC might represent advanced NASH-related cirrhosis, and they described many similar characteristics shared by the 2 conditions. Specifically, patients with NASH or CC are older and more likely to be female, obese, and diabetic in comparison with patients with other causes of cirrhosis. Subsequent studies have further supported the hypothesis that NASH is an underrecognized cause of CC.11, 12, 20, 21

In our analysis, the mean age and the rates of obesity, diabetes, and female sex were highest for recipients with NASH and lowest for patients with other causes of liver disease. Recipients with CC appeared to fall between these 2 groups. This suggests that some patients with CC have a phenotype similar to that of NASH patients [older, female, obese (BMI > 30 kg/m2), and diabetic] and likely have underlying NASH-related cirrhosis, whereas others may still have idiopathic cirrhosis.

Our analysis of the UNOS database is limited by the fact that the diagnosis of liver disease (specifically NASH or CC) is based on what is reported by each transplantation center to UNOS, and we cannot confirm the accuracy of the data. However, these diagnoses were all made by experienced hepatologists and surgeons in transplant centers, and an incorrect diagnosis is unlikely to have occurred in a significant number of cases.

Similarly, the reported causes of death may be inaccurate when a large database is involved. The primary cause of death was missing or unknown for approximately a quarter of all posttransplant deaths (3519 of 13,990 deaths). This may be due to the limited data available to the transplant centers concerning causes of recipient deaths, particularly because most deaths occur away from the transplant center or hospital and are later reported by a family member or a transplant coordinator. The accuracy of the causes of death, is therefore, poor and requires further validation before we can determine what cause of death is more common for patients with particular liver diseases.

The mean posttransplant follow-up was 4.1 years. It is possible that as patients who have undergone transplantation for NASH (for the most part after 2003) accrue more years of follow-up, differences after transplantation may arise (eg, because of the development of cardiovascular disease in patients with NASH). Therefore, our analyses should be repeated in 1 to 2 years with an updated UNOS database. A significant strength of our study is that it represents the entire US transplant experience from 1997 to 2010.

In conclusion, our study has demonstrated that the posttransplant survival of recipients with NASH is excellent and comparable to the survival of patients with all other liver diseases combined. The posttransplant survival of patients with NASH is in fact superior to the posttransplant survival of patients with other common indications for transplantation, including HCC, HCV infection, acute hepatic necrosis, alcoholic liver disease, hemochromatosis, and CC. NASH-related cirrhosis is rising dramatically as an indication for transplantation. Because of the liver donor shortage, transplant programs are often faced with difficult decisions with respect to the allocation of relatively few organs to a much larger pool of patients who need liver transplantation. Our study suggests that patients with NASH-related cirrhosis should continue to be considered potentially good candidates for liver transplantation and will likely account for an increasingly prominent proportion of liver transplants in the United States. It is important, however, to practice careful pretransplant screening for cardiovascular disease and aggressive posttransplant care of underlying cardiac and metabolic risk factors in patients with NASH.

Acknowledgements

The authors thank UNOS for providing the database. They also thank Dr. Thomas Vaughan for his critical review of this article and helpful suggestions.

Ancillary