ORIGINAL ARTICLE

De Novo Nonalcoholic Fatty Liver Disease AfterLiver TransplantationSuk Seo,1 Kalyani Maganti,1 Manjit Khehra,1 Rajendra Ramsamooj,2 Alexander Tsodikov,3

Christopher Bowlus,1 John McVicar,4 Mark Zern,1 and Natalie Torok1

1Division of Gastroenterology and Hepatology, University of California Davis Medical Center, Sacramento,CA; 2Division of Pathology, University of California Davis Medical Center, Sacramento, CA; 3Division ofBiostatistics, University of California Davis Medical Center, Sacramento, CA; 4Department of Surgery,University of California Davis Medical Center, Sacramento, CA

Hepatic steatosis is a recognized problem in patients after orthotopic liver transplant (OLT). However, de novo developmentof nonalcoholic fatty liver disease (NAFLD) has not been well described. The aim of this study was to determine the prevalenceand predictors of de novo NAFLD after OLT. A retrospective analysis was performed on 68 OLT patients with donor liverbiopsies and posttransplantation liver biopsies. Individual medical charts were reviewed for demographics, indication for OLT,serial histology reports, genotypes for hepatitis C, comorbid conditions, and medications. Liver biopsies were reviewed blindlyand graded according to the Brunt Scoring System. Multivariate logistic regression analysis was used to study the risk factorsfor developing NAFLD. The interval time from OLT to subsequent follow-up liver biopsy was 28 � 18 months. A total of 12patients (18%) developed de novo NAFLD, and 6 (9%) developed de novo NASH. The regression model indicated that the useof angiotensin-converting enzyme inhibitors (ACE-I) was associated with a reduced risk of developing NAFLD after OLT (oddsratio, 0.09; 95% confidence interval, 0.010-0.92; P � 0.042). Increase in body mass index (BMI) of greater than 10% after OLTwas associated with a higher risk of developing NAFLD (odds ratio, 19.38; 95% confidence interval, 3.50-107.40; P � 0.001).In conclusion, de novo NAFLD is common in the post-OLT setting, with a significant association with weight gain aftertransplant. The use of an ACE-I may reduce the risk of developing post-OLT NAFLD. Liver Transpl 13:844-847, 2007.© 2007 AASLD.

Received April 27, 2006; accepted July 23, 2006.

See Editorial on Page 788

It is estimated that nonalcoholic fatty liver disease(NAFLD) and nonalcoholic steatohepatitis (NASH) affect3% of the general population, respectively,1-3 but re-markably little is known about de novo development ofNAFLD/NASH after orthotopic liver transplant (OLT).This lack of data may partly be attributed to traditionalthinking that NAFLD is a diagnosis of exclusion. How-ever, rapid weight gain, diabetes mellitus, hyperlipid-emia, hypertension, and consequent metabolic syn-drome are more common posttransplantation than inthe general population; therefore, the risk of developing

NAFLD/NASH in this population may increase.4-6

There are small case reports of NASH developing denovo or recurring in OLT recipients.7,8 These patientstypically exhibit features of the metabolic syndrome,lack of alcohol consumption, and have liver histologyconsistent with NAFLD/NASH, including a high degreeof steatosis, lobular inflammation with neutrophils,ballooning degeneration, Mallory bodies, and fibrosis.The use of immunosuppressive medications could bepartially responsible, as both corticosteroids and cal-cineurin inhibitors promote hypertension and hyper-cholesterolemia. Prednisone, tacrolimus, and cyclo-sporine A are diabetogenic, and sirolimus induceshyperlipidemia.9 NAFLD and NASH could also coexist

Abbreviations: NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; OLT, orthotopic liver transplant; HCV,hepatitis C virus; ACE-I, angiotensin converting enzyme inhibitor; BMI, body mass index; HMG coA-I, 3-hydroxy-methylglutaryl-coenzyme A reductase inhibitor.Address reprint requests to Suk Seo, MD, Division of Gastroenterology/Hepatology, UC Davis Medical Center, 4150 V Street, Suite 3500,Sacramento, CA 95817. Telephone: 916-734-3751; FAX: 916-734-7908; E-mail: suk.seo@ucdmc.ucdavis.edu

DOI 10.1002/lt.20932Published online in Wiley InterScience (www.interscience.wiley.com).

LIVER TRANSPLANTATION 13:844-847, 2007

© 2007 American Association for the Study of Liver Diseases.

or be superimposed on other liver diseases after OLTsuch as recurrent hepatitis C virus (HCV) infection.

Recognizing posttransplantation NAFLD/NASH isimportant because it may significantly affect graft sur-vival by promoting fibrosis and cirrhosis. In addition,data from the pretransplantation population indicatethat patients with NAFLD and HCV are much morelikely to be nonresponder to antiviral therapy.10-13 Theaim of this study was to determine the prevalence of denovo NAFLD post-OLT and attempt to identify risk fac-tors that might lead to the development of future treat-ment options for NAFLD in the post-OLT setting.

PATIENTS AND METHODS

Study Design

A retrospective analysis of 68 OLT patients was per-formed at the University of California Davis MedicalCenter. All patients had donor liver biopsies. The post-OLT liver biopsies were performed for abnormal livertests. All study patients were transplanted and followedat University of California Davis Medical Center. Wereviewed individual medical charts for demographics,indication for OLT, serial histology reports, specific ge-notypes for HCV patients, comorbid conditions, andmedication profiles (Table 1). All the factors were de-fined as categorical variables being present or absentduring a 6-month period prior to the post-OLT liverbiopsy. Significant weight gain was defined as at least a10% increase in BMI compared to pre-OLT weight. Di-agnosis of hypertension was made according to theguideline set by the Joint National Committee.7 Diag-

nosis of hyperlipidemia was made by the guideline setby the National Cholesterol Education Program III.

This study was approved by the University of Califor-nia Davis Medical Center Institutional Review Board.

Histology

The liver biopsy specimens were reviewed in a blindedfashion by a hepatopathologist for the presence of pre-dominant macrovesicular hepatic steatosis, ballooningdegeneration, Mallory’s hyaline, perisinusoidal fibrosis,and neutrophilic lobular inflammation. De novo NAFLDwas determined in patients who were found to have agreater than 33% increase in steatosis when comparedto their donor biopsies. Each specimen with significantsteatohepatitis was graded according to the Brunt Scor-ing System.14 A minimum grade of 2 (moderate) wasused to define de novo NASH.

Data Analysis

Data were analyzed with SPSS version 13.0 (SPSS, Chi-cago, IL). To determine the independent predictors ofdevelopment of NAFLD and NASH, multiple logistic re-gression analysis was performed for the potential fac-tors shown in Table 1. The backward variable selectionprocedure based on the likelihood ratio was used todetermine the subset of variables in the best regressionequation. A P value � 0.05 was considered as statisti-cally significant. Odds ratios were given with the 95%confidence intervals.

RESULTS

Characteristics of the Patient Population

A total of 56 (81%) of the patients were male. The age (�standard deviation) at OLT was 48 � 8 yr (Table 1). HCVwas the most common indication for OLT (84%), withthe following genotype distribution: 67% genotype 1;6% genotype 2; and 27% genotype 3. Our study pa-tients had the following metabolic conditions: 26 pa-tients had type 2 diabetes mellitus; 47 hypertension;and 17 hyperlipidemia. Among the 26 patients noted tohave diabetes mellitus post-OLT, 11 patients had pre-existing diabetes mellitus before the OLT.

The use of tacrolimus (in 61 patients), sirolimus (11),mycophenolate mofetil (12), prednisone (14), ACE-I(22), peroxisome proliferator-activated receptor-� re-ceptor agonist (3), HMG-coA reductase inhibitor (12),and ursodiol (8) were also included in the analysis.

A total of 12 patients (18%) developed de novoNAFLD, and 6 (9%) developed de novo NASH, accordingto our definitions. Median donor age (range) for thestudy population, the de novo NAFLD subgroup, andthe de novo NASH subgroup were comparable: 31 (13-79), 24 (15-55), and 22 (15-55), respectively. The donorbiopsies showed 0% steatosis in 43 (63%) patients and1 to 33% steatosis in 25 (37%). Among those with 0%donor steatosis, 22 (32%) continued to show no steato-sis at the post-OLT biopsies, 16 (24%) developed �33%increase in steatosis, and 5 (7%) developed �33% ste-

TABLE 1. Univariate Analysis of the Potential Factors

Associated With Post-OLT De Novo NAFLD

Variable n (%) P

GenderMale 55 (81) 0.753

Indication(s) for OLTHCV 57 (84) 0.641Genotype 1 38 (67) 0.823Genotype 2 3 (6) 0.331Genotype 3 16 (27) 0.355

Increase in BMI � 10%after OLT

24 (35) 0.000

ComorbiditiesHypertension 47 (69) 0.096Diabetes mellitus 26 (38) 0.514Hyperlipidemia 17 (25) 0.594

MedicationsTacrolimus 61 (90) 0.693Prednisone 14 (21) 0.910Mycophenolate 12 (18) 0.556Sirolimus 11 (16) 0.453ACE-I 22 (32) 0.014HMG CoA-I 12 (17) 0.386Ursodiol 8 (12) 0.829PPAR-� receptor agonist 3 (4) 0.497

DE NOVO NONALCOHOLIC FATTY LIVER DISEASE 845

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

atosis. Among OLT patients with the presence of mini-mal donor steatosis, 8 (12%) showed no significantchange in steatosis and 17 (25%) showed increasedsteatosis at the post-OLT biopsies. The median lengthof time from OLT to the biopsy (range) for the studypopulation, the de novo NAFLD subgroup, and the denovo NASH subgroup were comparable: 23 (9-95), 17(11-51), and 16 (13-51) months, respectively.

The indication for OLT for all 6 patients with de novoNASH was HCV-related cirrhosis. Among these pa-tients, 3 had genotype 3, 2 had genotype 1, and 1 hadgenotype 2 infection. All these patients were HCV ribo-nucleic acid positive. Median HCV ribonucleic acid(range) for de novo NASH vs. non-NASH patients wascomparable at 1.6 � 106 (5.7 � 105 to 2.7 � 106) vs.1.6 � 106 (2.5 � 104 to 3.9 � 107) copies/mL.

Analysis of the Potential Predictors

Table 2 shows the distribution of the study patientswith and without the development of de novo NAFLDwith respect to their ACE-I use and weight gain afterOLT. Among patients who developed de novo NAFLDpost-OLT, 83% (10/12) had also gained weight greaterthan 10% of pretransplantation BMI. Among patientswith no significant post-OLT NAFLD, only 25% (14/56)gained significant weight. We found that among pa-tients with NAFLD, only 1 of 12 was receiving ACE-I forat least 6 months prior to biopsy. Among non-NAFLD,38% (21/56) were receiving ACE-I. The indication forthe use of ACE-I was hypertension in all patients in ourstudy group.

Table 3 represents the results of multivariate logisticregression analysis performed on potential predictorsin Table 1. The best regression equation with significanteffects showed that the increase in BMI of greater than

10% post-OLT was associated with increased risk ofdeveloping de novo NAFLD (odds ratio, 19.38; 95% con-fidence interval, 3.50-107.40; P � 0.001). In contrast,the use of ACE-I was associated with a reduced risk ofdeveloping de novo NAFLD after OLT (odds ratio, 0.09;95% confidence interval, 0.01-0.92; P � 0.042).

DISCUSSION

In patient with chronic hepatitis C, most investigatorsrequire the presence of at least 33% steatosis to war-rant a diagnosis of concomitant NAFLD because thesignificance of minimal steatosis is uncertain.6,10,15

Based on this rationale, we have performed the firststudy that addressed the prevalence of de novo NAFLDin OLT patients.

Our multivariate regression identified an increase inBMI and the use of ACE-I as a positive and a negativerisk factor for development of de novo post-OLT NAFLD,respectively. Increase in the BMI is known to be asso-ciated with NAFLD in the nontransplantation setting.16

In the first few months after a successful OLT, weightgain can be regarded as one of the positive effects oftransplantation. However, within 2 yr of transplanta-tion, excess body weight is recorded in up to 60 to 70%of patients, and 20% of the previously nonobese trans-plant recipients become obese.4,5 The exact mecha-nisms leading to excessive weight gain in such a largeproportion of patients are incompletely understood, al-though a major role is attributed to the development ofpost-OLT insulin resistance, diabetes mellitus, and themetabolic effects of immunosuppressive medications(corticosteroids, calcineurin inhibitors).

Currently, NAFLD has no known definitive medicaltherapy, but our finding showed that ACE-I may inde-pendently reduce the risk of de novo development ofNAFLD in the post-OLT setting. While the true impact ofACE-I on de novo NAFLD must be validated with futurestudies, experimental data shows that angiotensin re-ceptor inhibition improves insulin sensitivity,17 re-duces hepatic triglycerides with a decrease in the ma-crovesicular steatosis score,18,19 decreases oxidativestress,20 and attenuates the progression of hepatic fi-brosis in rodents.21 In addition, angiotensin receptorinhibition also improved histological scores in patientswith NASH.22 While the molecular basis for these ob-servations is not well-described, there is considerableinterest in a potential role of angiotensin inhibition onperoxisome proliferator-activated receptor-�. Peroxi-some proliferator-activated receptor-� has an impor-tant role in adipocyte differentiation as well as regula-tory function in both hepatic fatty acid synthesis andoxidation, and its activation seems to be beneficial inNAFLD patients.23-27

Commonly used immunosuppressive medicationssuch as prednisone, calcineurin inhibitors, and siroli-mus in the OLT setting can increase insulin resistance;but we did not find any statistically significant associ-ation between these variables and development of post-OLT NAFLD and NASH. Hyperlipidemia and diabetesare often associated with weight gain and have been

TABLE 2. Distribution of the Patients With and

Without the Development of de novo NAFLD With

Respect to Their ACE-I Use and Weight Gain After OLT

ACE-I

n (%)

Weight gain

(�BMI � 10%)

n (%)

de novo NAFLD; n � 12 1 (8) 11 (92)No NAFLD; n � 56 21 (38) 14 (25)

TABLE 3. Multivariate Analysis of the Factors

Independently Associated With De Novo

NAFLD After OLT

Variable

Odds

ratio

95%

confidence

interval P

Weight gain(�BMI � 10%)

19.38 3.50–107.40 0.001

ACE-I 0.09 0.01–0.92 0.042

846 SEO ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

implicated in NASH in nontransplantation patients;however, these variables also did not reach statisticalsignificance in our multivariate study. The statisticalanalysis of these subgroups may be limited by the smallsample size, and in the future these findings will have tobe verified with larger studies and longer follow-up.

The finding that the use of ACE-I could potentially bebeneficial in reducing risk of the post-OLT NAFLD andNASH, is both novel and significant. Larger, prospectivetrials which include its effect on graft survival areneeded to validate the use of ACE-I in the post-OLTpopulation.

REFERENCES

1. Wanless IR, Lentz JS. Fatty liver hepatitis (steatohepatitis)and obesity: an autopsy study with analysis of risk factors.Hepatology 1990;12:1106-1110.

2. Hilden M, Christoffersen P, Juhl E, Dalgaard JB. Liverhistology in a ‘normal’ population—examinations of 503consecutive fatal traffic casualties. Scand J Gastroenterol1977;12:593-597.

3. Neuschwander-Tetri B, Caldwell S. Nonalcoholic steato-hepatitis: summary of an AASLD single topic conference.Hepatology 2003;37:1202-1219.

4. Everhart JE, Lombardero M, Lake JR, Wiesner RH, Zetter-man RK, Hoofnagle JH. Weight change and obesity afterliver transplantation: incidence and risk factors. LiverTranspl Surg 1998;4:285.

5. Palmer M, Schaffner F, Thung SN. Excessive weight gainafter liver transplantation. Transplantation 1991;51:797.

6. Burke A, Lucey MR. Non-alcoholic fatty liver disease, non-alcoholic steatohepatitis and orthotopic liver transplanta-tion. Am J Transplant 2004;4:686-693.

7. Angelico F, Del Ben M, Conti R, Francioso S, Feole K,Maccioni D, et al. Non-alcoholic fatty liver syndrome: ahepatic consequence of common metabolic diseases. Eur JGastroenterol Hepatol 2003;15:99-102.

8. Cauble MS, Gilroy R, Sorrell MF, Mailliard ME, Sudan DL,Anderson JC, et al. Lipoatrophic diabetes and end-stageliver disease secondary to nonalcoholic steatohepatitiswith recurrence after liver transplantation. Transplanta-tion 2001;71:892-895.

9. Marchetti P, Navalesi R. The metabolic effects of cyclo-sporin and tacrolimus. J Endocrinol Invest 2000;23:482-490.

10. Sanyal A. Non-alcoholic fatty liver disease and hepatitisC-risk factors and clinical implications, Aliment Pharma-col Ther 2005;22(Suppl 2):48-51.

11. Yaginuma R, Ikejima K, Okumura K, Kon K, Suzuki S,Takei Y, et al. Hepatic steatosis is a predictor of poorresponse to interferon alpha-2b and ribavirin combinationtherapy in Japanese patients with chronic hepatitis C.Hepatol Res 2006;35:19-25.

12. Lonardo A, Loria P, Adinolfi LE, Carulli N, Ruggiero G.

Hepatitis C and steatosis: a reappraisal. J Viral Hepat2006;13:73-80.

13. Asselah T, Rubbia-Brandt L, Marcellinn P, Negro F. Ste-atosis in chronic hepatitis C: why does it really matter?Gut 2006;55:123-130.

14. Brunt EM, Janney CG, Di Bisceglie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: A pro-posal for grading and staging the histologic lesions. Am JGastroenterol 1999;94;2467-2474.

15. Contos MJ, Cales W, Sterling RK, Luketic VA, ShiffmanML, Mills AS, et al. Development of nonalcoholic fatty liverdisease after orthotopic liver transplantation for crypto-genic cirrhosis. Liver Transpl 2001;7:363-373.

16. Ong JP, Elariny H, Collantes R, Younoszai A, ChandhokeV, Reines HD, et al. Predictors of nonalcoholic steatohepa-titis and advanced fibrosis in morbidly obese patients.Obes Surg 2005;15:310-315.

17. Moises RS, Carvalho CR, Shiota D, Saad MJ. Evidence fora direct effect of captopril on early steps of insulin actionin BC3H-1 myocytes. Metabolism 2003;52:273-278.

18. Ackerman Z, Oron-Herman M, Grozovski M, Rosenthal T,Pappo O, Link G, Sela BA. Fructose-induced fatty liverdisease: hepatic effects of blood pressure and plasma tri-glyceride reduction. Hypertension 2005;45:1012-1018.

19. Toblli JE, Ferder L, Stella I, Angerosa M, Inserra F. Ena-lapril prevents fatty liver in nephrotic rats. J Nephrol2002;15:358-367.

20. Bataller R, Brenner DA. Liver fibrosis [Review]. J ClinInvest 2005;115:209-218. [Erratum: J Clin Invest 2005;115:1100].

21. Jonsson JR, Clouston AD, Ando Y, Kelemen LI, Horn MJ,Adamson MD, et al. Angiotensin converting enzyme inhi-bition attenuates the progression of rat hepatic fibrosis.Gastroenterology 2001;121:148-155.

22. Yokohama S, Yoneda M, Haneda M, Okamoto S, Okada M,Aso K, et al. Therapeutic efficacy of an angiotensin II re-ceptor antagonist in patients with nonalcoholic steato-hepatitis. Hepatology 2004;40:1222-1225.

23. Benson SC, Pershadsingh HA, Ho CI, Chittiboyina A, De-sai P, Pravenec M, et al. Identification of telmisartan as aunique angiotensin II receptor antagonist with selectivePPAR�-modulating activity. Hypertension 2004;43:993-1002.

24. Schupp M, Janke J, Clasen R, Unger T, Kintscher U.Angiotensin type 1 receptor blockers induce peroxisomeproliferator-activated receptor-� activity. Circulation2004;109:2054-2057.

25. Berger JP, Petro AE, Macnaul KL, Kelly LJ, Zhang BB,Richards K, et al. Distinct properties and advantages of anovel peroxisome proliferator- activated protein {�} selec-tive modulator. Mol Endocrinol 2003;17:662-676.

26. Sharma AM, Janke J, Gorzelniak K, Engeli S, Luft FC.Angiotensin blockade prevents type 2 diabetes by forma-tion of fat cells. Hypertension 2002;40:609-611.

27. Sugimoto K, Qi NR, Kazdova L, Pravenec M, Ogihara T,Kurtz TW. Telmisartan but not valsartan increases caloricexpenditure and protects against weight gain and hepaticsteatosis. Hypertension 2006;47:1003-1009.

DE NOVO NONALCOHOLIC FATTY LIVER DISEASE 847

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases