SEARCH

SEARCH BY CITATION

555

  1. Top of page

Significance of serum and hepatic microRNA-122 levels in patients with non-alcoholic fatty liver disease

Hisamistu Miyaaki, Tatsuki Ichikawa, Naoto Taura, Satoshi Miuma, Hidetaka Shibata, Takuya Honda, Kazuhika Nakao
The First Department of Internal Medicine, Nagasaki University School of Medicine, Nagasaki, Japan

Aims: Non-alcoholic fatty liver disease (NAFLD) is thought to be a type of metabolic syndrome. MicroRNA-122 (miR-122) is the most abundant microRNA in the liver and is an important factor for the metabolism of glucose and lipids. In this study, we examined the correlation of hepatic and serum miR-122 expression levels with the clinicopathological factors of patients with NAFLD. Methods: Total RNA with preserved miRNAs was extracted from liver biopsy samples of 67 patients with NAFLD. In 52 of these 67 patients, total RNA was extracted from serum. miR-122 obtained by quantitative reverse transcription-polymerase chain reaction was quantified using TaqMan MicroRNA assays. MiR-122 expression was calculated by the relative standard curve method and normalized to RNU6 in the liver and cell-miR39 expression in the serum. Results: A significantly correlation was detected between serum and hepatic miR-122 expression (correlation coefficient, 0.461; p = 0.005). Patients with mild steatosis (<33%) showed significantly lower levels of hepatic miR-122 than patients with severe steatosis (>33%) (hepatic miR-122: mild: severe = 2.158 ± 1.786: 4.836 ± 7.506, p = 0.0473; serum miR-122: mild: severe = 0.002 ± 0.005: 0.007 ± 0.001, p = 0.0491). There was no significant correlation of hepatic and serum miR-122 and NAFLD Activity score (NAS). However, hepatic and serum miR-122 levels in liver showed significantly inverse correlation of fibrosis stage [Hepatic miR-122: correlation coefficient −0.292, p =0.0161; Serum miR-122: correlation coefficient −0.316, p =0.021 8]. Moreover, hepatic and serum miR-122 levels were significantly higher in patients with mild fibrosis than in those with severe fibrosis (hepatic miR-122: mild: severe = 5.201 ± 7.275: 2.394 ± 1.547, p = 0.0087; serum miR122: mild: severe = 0.008 ± 0.011: 0.002 ± 0.004, p = 0.0191). Conclusions: Hepatic and serum miR-122 levels were associated with hepatic steatosis and fibrosis. Serum miR-122 level was well correlated with hepatic miR-122 and could be a useful predictive marker of liver fibrosis.

Disclosures:

The following people have nothing to disclose: Hisamistu Miyaaki, Īatsuki Ichikawa, Naota Taura, Satoshi Miuma, Hidetaka Shibata, Takuya Honda, Kazuhiko Nakao

556

  1. Top of page

Poor performance and diagnostic accuracy of preoperative transient elastography and ARFI imaging in morbidly obese bariatric patients

Thomas Karlas1,2Veronika Peter3, Arne Dietrich4, Nikita Garnov5, Harald Busse6, Christiane Prettin7, Volker Keim2, Tatjana Schütz3, Johannes Wiegand2, 1
1IFB AdiposityDiseases, Leipzig University Medical Center, Leipzig, Germany; 2Department of Internal Medicine, Neurology and Dermatology, Division of Gastroenterology and Rheumatology, University Hospital Leipzig, Leipzig, Germany; 3IFB AdiposityDiseases, Research Area Bariatric Surgery, Leipzig University Medical Center, Leipzig, Germany; 4Department of Surgery, University Hospital Leipzig, Leipzig, Germany; 5IFB AdiposityDiseases, Core Unit Imaging in Obesity, Leipzig University Medical Center, Leipzig, Germany; 6Department of Diagnostic and Interventional Radiology, Leipzig University Hospital, Leipzig, Germany; 7Coordination Centre for Clinical Trials, University of Leipzig, Leipzig, Germany

Background: Advanced liver fibrosis in morbidly obese patients with non-alcoholic fatty liver disease (NAFLD) is associated with a higher risk of surgical and anaesthesiological complications. A reliable non-invasive assessment of hepatic fibrosis before surgery might therefore be beneficial. We determined the reliability and accuracy of transient elastography (TE) and acoustic radiation force impulse (ARFI) imaging in bariatric patients. Patients and methods: Patients randomly underwent one of two 14-day low-energy diets meant to reduce visceral adipose tissue (VAT) and liver volumes prior to bariatric surgery (day 0). TE (M and XL probes) and ARFI were performed by experienced examiners on days −14 and −1. Valid liver stiffness measurements were assumed for 10 successful samples (success rate >60%) with a relative interquartile range (IQR) of less than 30% of the median value. Results were compared with those from intraoperative liver biopsies. VAT and liver volumes were assessed by MRI. Results: Our 41 patients (28 female; median age 46.4, range 28-64 years) had a median (range) BMI of 47 (34-60) kg/m^2 at baseline and 47 (32-57) kg/m^2 after dietary intervention. Weight loss (mean −5.2±2.1 kg) was associated with reductions in VAT and liver volumes (p<0.001, respectively). Valid TE results could be obtained in 22% (M probe) and 37% (XL probe) of all cases at day −14 as well as in 15% (M) and 54% (XL) at day −1. BMI (p<0.03) and skin-toliver-capsule distance (p<0.01) correlated significantly with invalid TE samples. ARFI data could be obtained in 88% and 90% of all cases at both time points. However, in 70% of these cases, the relative IQR was greater than 30%. In contrast to the biopsy results (Ishak F1 fibrosis: 39, cirrhosis: 1, no signs of NAFLD: 1), variations (median/range at respective time points) of TE M (6.1/3.9-11.7 and 6.3/5.1-16.8 kPa), TE XL (4.6/2.6-17.3 and 6.2/2.9-21.3 kPa) and ARFI (2.1/0.7-3.7 and 2.0/0.7-3.8 m/s) were high. Both methods failed to identify the cirrhotic patient. Valid TE measurements at both time points were only available in 10% (M) and 34% (XL). Serial ARFI data were available in 83% of the cases (less than 10% fulfilling the IQR criterion) and did not vary significantly (mean change −0.05±0.80 m/s, p=0.73) although 17 cases showed changes larger than 25% of their baseline values. Conclusion: Performance, accuracy, and reproducibility of TE and ARFI in morbidly obese patients were poor. Neither the use of an XL probe nor a moderate weight loss improved TE performance.

Disclosures:

Thomas Karlas - Grant/Research Support: Echosens, France

The following people have nothing to disclose: Veronika Peter, Arne Dietrich, Nikita Garnov, Harald Busse, Christiane Prettin, Volker Keim, Tatjana Schutz, Johannes Wiegand

557

  1. Top of page

Improvement of liver PPAR alpha expression correlates with histological improvement in a large non-bariatric surgery 1 year follow-up series of NAFLD patients

Sven M. Francque1, An Verrijken2, Ilse Mertens2, Sandrine Caron3, 4, Janne Prawit3, 4, Réjane Paumelle3, 4, Bruno Derudas3, 4, Peter P Michielsen1, Eric A. Van Marck5, Guy Hubens6, Bart Staels3, 4, Luc F. Van Gaal2; 1Gastroenterology Hepatology, Antwerp University Hospital' Edegem' Belgium; SEndocrinology' Diabetes and Metabolism, Antwerp University Hospital, Edegem, Belgium; 3Institut Pasteur, Lille, France; 4INSERM U1011, Universite de Lille, Lille, France; 5Pathology, Antwerp University Hospital, Edegem, Belgium; 6Abdominal Surgery, Antwerp University Hospital, Edegem, Belgium

Background: The pathophysiology of NASH remains poorly understood and currently no targeted treatment is approved. We recently demonstrated that human liver PPARα (but not β/δ or y) expression is negatively correlated with NASH severity in cross-sectional analysis. Aims: To study the expression of the different PPAR isotypes in NAFLD patients in 1 year follow-up biopsies and to study the relation between changes in PPAR expression and histological improvement. Methods: Patients presenting to the obesity clinic for a problem of overweight were assessed for indicators of NAFLD. If NAFLD was suspected, a liver biopsy was proposed. Patients were subsequently included in a weight management program. A control biopsy was proposed after 1 year. Histology was scored using the NASH CRN scoring system. PPAR expression was studied by quantification of mRNA levels by real time RT-PCR after total RNA extraction from liver homogenates and reverse transcription into cDNA. Results: Fifty-two patients with a complete baseline and follow-up dataset were consecutively included between 2008 and 2012, 42% male, mean BMI 37.6±6.5 kg/m2. BMI significantly improved (mean 33.7±6.7 kg/m2, p<0.001 compared to baseline). Liver histology also significantly improved on treatment (no NASH/borderline NASH/definite NASH from 21.2/36.5/42.3% to 55.8/23.1/21.2%, p<0.001; mean NAS from 4.1 ±1.9 to 2.3±2.3, p<0.001). PPARα expression was significantly different between no NASH/borderline NASH/definite NASH both at baseline (p<0.001) and followup (p=0.005) (Kruskal-Wallis) with the lowest expression in patients with definite NASH. PPARα expression also significantly correlated with the NASH activity score (NAS) (baseline: r=0.405, p=0.008, follow-up r=0.305, p=0.011). Furthermore histological improvement as expressed by reduction in the NAS significantly correlated (r=0.322, p=0.016) with a concomitant increase in PPARα expression. PPARβ/ and PPAR۷ expression did not correlate with liver histology both at baseline and follow-up. Conclusion: In this large, non-bariatric surgery, series of NAFLD-patients with paired liver biopsies, PPARα expression inversely correlated with NASH severity both at baseline and at follow-up. Furthermore, improvement in liver histology and in PPARα expression were strongly associated, providing further rationale for PPARα targeted treatment.

Disclosures:

Bart Staels - Advisory Committees or Review Panels: MSD, Roche; Consulting: Genfit; Speaking and Teaching: Abbott

The following people have nothing to disclose: Sven M. Francque, An Verrijken, Ilse Mertens, Sandrine Caron, Janne Prawitt, Rejane Paumelle, Bruno Derudas, Peter P. Michielsen, Eric A. Van Marck, Guy Hubens, Luc F. Van Gaal

558

  1. Top of page

Subcutaneous adipose tissue insulin resistance and lipolysis in patients with nonalcoholic steatohepatitis

Mattnew J. Armstrong1, Jonathan M. Hazlehurst2, Diana Hull1, Sarah Borrows3, Kathy Guo1, Jinglei Yu5, Darren Barton1,Piers Gaunt1,Stephen Gough4, Jeremy W Tomlinson2, Philip N. Newsome1
1NIHR Liver BRU and Centre for Liver Research, University of Birmingham, Birmingham, United Kingdom; 2CEDAM, University of Birmingham, Birmingham, United Kingdom; 3Wellcome Trust Clinical Research Facility, University of Birmingham, Birmingham, United Kingdom; 4OCDEM, Churchill Hospital, Oxford, United Kingdom; 5School of sports and exercise sciences, University of Birmingham, Birmingham, United Kingdom

Background and aims: Systemic insulin resistance is a primary feature in non-alcoholic steatohepatitis (NASH), however there remains limited data on subcutaneous adipose tissue insulin sensitivity and lipolysis. We examined tissue-specific insulin resistance and inflammation in patients with NASH. Methods: 16 European Caucasian patients with biopsy-confirmed NASH (n=6 fibrosis stage 0-2; n=10 stage 3-4) and 15 healthy volunteers (no medical conditions, normal blood liver enzymes, fibrosis markers, HOMA-IR) were studied. Tissue-specific (adipose tissue, muscle, liver) actions of insulin were determined using 2step hyperinsulinemic euglycemic clamps incorporating stable isotopes with concomitant subcutaneous adipose tissue microdialysis. Fasting serum adipose-derived and inflammatory cytokines were measured using Fluorokine® MAP multiplex kits. All values are means+/-SE (unpaired Student T-tests). Results: Patients with NASH (age 54.4+/-2.1 years; 69% male; BMI 34.2+/-1.0 kg/m2] had significantly higher fasting serum glucose, insulin and HOMA-IR than healthy controls (age 33.1+/- 2.2 years; 60% male; BMI 26.7+/-1.0 kg/m2). NASH patients had significant hepatic and muscle insulin resistance compared to controls, as demonstrated by lower % suppression of hepatic glucose production rate (41+/-4.3 vs.70+/-9.5 %; p<0.05) and lower glucose disposal (0.85+/-0.1 vs.1.76+/-0.39 mg/kg/min; P<0.05). NASH patients have significant adipose insulin resistance, as demonstrated by higher insulin concentration required to 1/2-maximaIIy suppress circulating free fatty acids (227+/-35.2 vs.65.2+/-14.0 pmol/L; p<0.001). Furthermore, in patients with NASH, interstitial fluid glycerol release from subcutaneous adipose tissue in response to both low-dose (379+/-42.6 vs.143+/-18.1 AUC μmol/l. h; p<0.0001) and high-dose insulin (261+/-30.7 vs.65.8+/-13.6 AUC μmol/l. h; p<0.0001) was significantly higher. NASH patients had significantly higher fasting circulating leptin: adiponectin ratio (3.22+/-0.49 vs.1.27+/-0.35 ng/μg p=0.003), TNFα (4.89+/-0.60 vs.1.32+/-0.14 pg/ml P<0.0001), hs-CRP (4.31+/-0.89 vs.1.47+/-0.47 μg/ml p<0.05), IL-6 (4.44+/-0.56 vs.2.59+/-0.51 pg/ml; p<0.05) and CCL-2 (224+/-14.4 vs.170+/-14.5 pg/ml; p<0.05) than controls. Conclusions: NASH patients have profound insulin resistance in the liver, muscle and adipose tissues. This study represents the first in-vivo description of dysfunctional subcutaneous adipose tissue in patients with NASH.

Disclosures:

Matthew J. Armstrong - Grant/Research Support: Novo Nordisk Ltd

Stephen Gough - Advisory Committees or Review Panels: Novo Nordisk, Eli Lilly, Sanofi Aventis, Takeda, GSK; Grant/Research Support: Novo Nordisk, Eli Lilly, Takeda; Speaking and Teaching: Novo Nordisk, Eli Lilly, Sanofi Aventis, GSK

Philip N. Newsome - Grant/Research Support: Novo Nordisk

The following people have nothing to disclose: Jonathan M. Hazlehurst, Diana Hull, Sarah Borrows, Kathy Guo, Jinglei Yu, Darren Barton, Piers Gaunt, Jeremy W. Tomlinson

559

  1. Top of page

Usefulness of Monitoring Serum CyfokeraHn-18 levels for Evaluating Long-term Prognosis in Patients with Non-Alcoholic Steatohepatitis/Non-Alcoholic Fatty Liver Disease

Miwa Kawanaka, Ken Nishino, Jun Nakamura, Takahito Oka, Noriyo Urata, Daisuke Goto, Mitsuhiko Suehiro, Hirofumi Kawamoto, Gotaro Yamada
Kawasaki hostital, Kawasaki Medical school, Jeneral internal Medicine2, Okayama, Japan

Background: Cytokeratin 18 (CK18) is a major intermediate filament protein in liver cells. Serum/plasma CK18 levels have been investigated as potential biomarkers for steatohepatitis, in patients with non-alcoholic fatty liver disease (NAFLD) /nonalcoholic steatohepatitis (NASH). The current study was performed to determine the usefulness of serum CK18 levels for evaluating long-term prognosis in NASH patients. Methods: The M30 enzyme-linked immunosorbent assay kit (Peviva) was used for estimating serum CK18 levels in 147 patients with NAFLD/NASH diagnosed by liver biopsies. A second liver biopsy (4.3±2.6 y) was required in 71 patients. We determined the relationship of CK18 levels and liver histology findings (fibrosis stage, lobular and portal inflammations, hepatocellular ballooning, and steatosis) with age; sex; BMI; biological markers (ALT, AST, y-GTP, total bilirubin, total cholesterol, and cholinesterase levels and platelet count); and disease markers such as HOMA-IR, and serum iron, ferritin, leptin, adiponectin, and high-sensitivity CRP, procollagen III peptide, type 4 collagen 7S, hyaluronic acid, and TIMP-1 levels. Results: CK18 levels were higher in theNA^LD activity score (NAS)>5 than NAS<4 (675.1 U/L vs 348.7 U/L; p<0.0001). The receiver operating characteristic curve indicated a cutoff value of 375 U/L, with specificity, 81.5%; sensitivity, 65%; and positive and negative predictive values, 80.8% and 43.1%, respectively, for the diagnosis NAS>5. The serum CK18 levels correlated with those of ALT (r=0.49, p<0.0001), AST (r=0.47, p<0.0001), ferritin (r=0.42, p<0.0001), TIMP-1 (r=0.44, p<0.0001), and procollagen III peptide (r = 0.31, p<0.0001) and with HOMA-IR (r=0.33, p<0.0001). In addition, the serum CK18 levels were associated with lobular and portal inflammations, hepatocellular ballooning, and Mallory body formation, but not with steatosis. Among the 71 patients who required a second liver biopsy, 19 with fibrosis progression had significantly high mean CK1 8 levels (from 539 to 907 U/L, p<0.01) and the NAS score increased from 5.2 to 6.1 (p<0.05). In contrast, a significant decrease was noted in the mean CK18 levels (from 637.3 to 468.7 U/L, p<0.001) and NAS score (from 5.8 to 4.1, p<0.0001) in 52 patients with static or improved fibrosis. Conclusion: CK18 levels were significantly elevated in NASH/NAFLD patients with fibrosis progression, but not in patients with static or improved fibrosis. Careful monitoring of serum CK18 levels may be useful in predicting fibrosis progression in NASH.

Disclosures:

The following people have nothing to disclose: Miwa Kawanaka, Ken Nishino, Jun Nakamura, Takahito Oka, Noriyo Urata, Daisuke Goto, Mitsuhiko Suehiro, Hircfumi Kawamoto, Gotaro Yamada

560

  1. Top of page

A novel method using gene set enrichment based on human genome-wide association study data implicates FXR/RXR activation as a common pathway affecting blood lipids levels and nonalcoholic fatty liver disease

Yindra M. Puentes1 , 2Corey C. Powell1 , 2Laura M. Yerges-Armstrong3, Mary F. Feitosa4, Lawrence F. Bielak5, Albert V. Smith6, Tamara B. Harris7, Jiankang Liu8, Solomon K. Musani8, Ingrid B. Borecki4, Patricia A. Peyser5, Elizabeth K. Speliotes1 , 2
1 Gastroenterology, University of Michigan, Department of Internal Medicine and Division of Gastroenterology, Ann Arbor, MI; 2Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI; 3Division of Endocrinology, Diabetes and Nutrition, Department of Medicine, University of Maryland School of Medicine, Baltimore' MD; 4Division of Statistical Genomics, Department of Genetics, Washington University, Saint Louis, MO; 5Department of Epidemiology, University of Michigan School of Public Health, Universify of Michigan, Ann Arbor, MI; 6Icelondic Heart Association, Kopavogur IS-201, Iceland; 7Laboratory of Epidemiology, Demography, and Biometry, Intramural Research Program, National Instifufe on Aging, National Institutes of Health, Bethesda, MD; 8University of Mississippi Medical Center, Jackson, MS

Nonalcoholic fatty liver disease (NAFLD) is rapidly increasing in prevalence and will become the number one cause of liver disease worldwide by 2020. NAFLD is associated with high triglycerides (TG), high low-density lipoprotein cholesterol (LDLC) levels, and low high-density lipoprotein cholesterol (HDL-C) levels. Both NAFLD and blood lipid levels are genetically influenced and may share a common genetic etiology. Purpose: We aimed to identify genes and pathways enriched for genetic associations with both blood lipids and NAFLD using human genome wide association study (GWAS) data. Methods: We examined whether genome wide, significantly associated lipid SNP sets from publicly available HDL-C, LDL-C and TG GWAS analyses (N=99, 000) were enriched for associations with hepatic steatosis, the hallmark of NAFLD, measured using computed tomography(CT) (N=7, 126, Speliotes, PLoS Gen, 2011). We then used gene set enrichment analysis (GSEA) as implemented in MAGENTA to identify pathways enriched in HDL-C, LDL-C, and TG GWAS analyses. These lipid pathways were then tested for enrichment in the NAFLD GWAS (N=7, 126) and these enriched pathways were also tested for replication in an independent CT measured hepatic steatosis GWAS sample (N=3, 124, Palmer, Hepatology, 2013). Results: We found that TG (P=0.004) and LDL-C (P=0.02) but not HDL-C GWA SNP sets were enriched in NAFLD. We identified 58 pathways that were enriched in lipid GWAS data. Three of these were also enriched in the NAFLD GWAS (N=7, 126) and one, FXR/RXR activation, also showed significant enrichment in the second independent NAFLD GWAS (N=3, 124). None of the three original NAFLD enriched pathways were enriched for associations in control publically available GWAS analyses of diastolic and systolic blood pressure (N=275, 000), body mass index (N=249, 796), and waist to hip ratio (N=77, 167) suggesting that the enrichment was specific to NAFLD. Genes associated with NAFLD (P<0.05) in FXR/RXR activation fell into three functional categories: (1)VLDL Assembly: MTTP, APOB, APOC3, (2) Nuclear Related Processes: PPAR-a, HNF1 a, NR0B2/SHP and (3) Hepatic Transport: MRP2, AE2, ABCG8, ABCG5, OAT2. Conclusions: Using a novel approach, we found that human genetic variation in or near genes involved in FXR/RXR activation affects both blood lipids and NAFLD in humans. These results suggest that genes that play a role in lipoprotein assembly, nuclear receptor biology, and hepatic transport when altered may affect NAFLD and thus could provide possible therapeutic targets for NAFLD prevention or treatment.

Disclosures:

The following people have nothing to disclose: Yindra M. Puentes, Corey C. Powell, Laura M. Yerges-Armstrong, Mary F. Feitosa, Lawrence F. Bielak, Albert V. Smith, Tamara B. Harris, Jiankang Liu, Solomon K. Musani, Ingrid B. Borecki, Patricia A. Peyser, Elizabeth K. Speliotes

561

  1. Top of page

Increased circulating soluble CD36 is associated with advanced steatosis in patients with nonalcoholic fatty liver disease

Carmelo García-Monzón6, Oreste Lo Iacono1, Javier Crespo2, Manuel Romero-Gomez3, Javier Garcia-Samaniego4, Miguel Fernondez-Bermejo5
1Gastroenterology Service, Hospital del Tajo, Aranjuez, Spain; 2Gastroenterology and Hepatology Service, University Hospital Marqués de Valdecillo, Santander, Spain; 3UG MQ Enfermedades Digestivas, Hospital Universitario Nuestra Señoro de Volme, CIBERehd, Sevilla, Spain; 4Hepatology Unit, Hospital Carlos III, CIBERehd, Madrid, Spain; 5Gastroenterology Service/ Hospifol Son Pedro de Alcántara, Cáceres, Spain; 6Liver Research Unit, University Hospital Santa Cristina, Instituto de Investigacion Sanitaria Princesa, CIBERehd, Madrid, Spain

Increased circulating soluble CD36 (sCD36), a cell-free form of fatty acid translocase CD36, clusters with insulin resistance and surrogate markers of fatty liver in population studies but no evidence exists on its relationship with hepatic fat content. The aim of the present study was to elucidate whether circulating sCD36 is linked to the amount of lipids within the liver in nonalcoholic fatty liver disease (NAFLD) and chronic hepatitis C (CHC) patients. This study comprised an overall population of 399 patients (227 with NAFLD, 87 with CHC, and 85 with histologically normal liver [NL]) who underwent a liver biopsy either by a percutaneous route for diagnostic purposes or during programmed abdominal surgery for gastroplasty or cholecystectomy. Steatosis was graded by Kleiner histological scoring system. Serum sCD36 levels and hepatic CD36 expression were assessed by immunoassay and immunohistochemistry, respectively. In NAFLD patients, serum sCD36 levels were significantly higher in those with simple steatosis than in NL subjects (361.4 ± 286.4 versus 173.9 ± 137.4 pg/mL, respectively; P < 0.001) but not in patients with steatohepatitis (229.6 ± 202.5 pg/mL; P = 0.153). In CHC patients, serum sCD36 levels were similar regardless of the absence (428.7 ± 260.3 pg/mL) or presence of steatosis (387.2 ± 283.6 pg/ml; P = 0.173). A progressive increase of serum sCD36 values was found in NAFLD patients depending on the histological grade of steatosis (P < 0.001) but not in those with CHC (P = 0.151). Serum sCD36 concentrations were independently associated with advanced steatosis in NAFLD patients when adjusted by demographic and anthropometric features [odds ratio (OR), 1.001; 95% confidence interval (CI), 1.000 to 1.002; P = 0.021] and by metabolic variables (OR, 1.002; 95% CI, 1.000 to 1.003; P = 0.001). Interestingly, in the overall cohort, a significant correlation was observed between serum sCD36 levels and hepatic CD36 expression (rho = 0.499, P <0.001). In conclusion, circulating level of sCD36 correlates with the histological grade of steatosis and is an independent factor associated with advanced steatosis in patients with NAFLD but not in CHC patients. Performance of serum sCD36 as a direct marker of steatosis, however, must be validated in further clinical studies including distinct cohorts of biopsy-proven NAFLD patients.

Disclosures:

Javier Crespo - Board Membership: MSD, Roche, Janssen, Gilead

Manuel Romero-Gomez - Advisory Committees or Review Panels: Roche Farma, SA., MSD, S. A., Janssen, S. A., Abbott, S. A.; Grant/Research Support: Ferrer, S. A.

Javier Garda-Samaniego - Consulting: Boehringer-Ingelheim

The following people have nothing to disclose: Carmelo García-Monzón, Oreste Lo Iacono, Miguel Fernandez-Bermejo

562

  1. Top of page

Glucagon like peptide 1 analogue, Liraglutide, reduces adipose insulin resistance and hepatic de novo lipogenesis in Nonalcoholic Steatohepatitis: Sub-study results of a phase II randomised-control clinical trial

Matthew J. Armstrong1, Diana Hull1, Kathy Guo1, Darren Barton1, Jonathan M. Hazlehurst2, Maryam Nasiri2, Laura L. Gathercole2, Jinglei Yu4, Piers Gaunt1, Stephen Gough3, Jeremy W. Tomlinson2, Philip N. Newsome1
1NHR Liver BRU and Centre for Liver Research, University of Birmingham, Birmingham, United Kingdom; 2CEDAM, University of Birmingham, Birmingham, United Kingdom; 3OCDEM, Churchill Hospital, Oxford, United Kingdom; 4School of sports and exercise sciences, University of Birmingham, Birmingham, United Kingdom

Background and aims: Adipose tissue insulin resistance and lipotoxicity are key pathognomonic features in nonalcoholic steatohepatitis (NASH). Liraglutide is a once-daily, long-acting glucagon-like peptide 1(GLP-1) analogue that significantly improves glycaemic control, weight and hepatic steatosis. The aim of this phase II mechanistic study was to determine the effect of Liraglutide on insulin sensitivity (hepatic, muscle, adipose), hepatic lipogenesis and markers of adipose inflammation. Methods: 14 patients with biopsy-proven NASH were randomly assigned to 1.8mg Liraglutide or placebo (once-daily, SC injections) for 12-weeks as part of the metabolic sub-study of the double-blind, randomised, placebo-controlled LEAN trial (clinicaltrials. gov. NCT01237119). At baseline and 12-weeks, patients underwent paired 2-step hyperinsulinaemic euglycaemic clamps incorporating stable isotopes with concomitant adipose tissue microdialysis. Serum adipocyfokines were quantified with Fluorokine® MAP multiplex kits. In-vitro isotope experiments were performed with Huh-7 and primary human hepatocytes from non-diabetic, male donors with normal BMI. Results: 1.8mg Liraglutide significantly decreased weight, waist circumference, HbA1c, fasting glucose, LDL and liver enzymes versus placebo. Liraglutide significantly increased the suppression of hepatic glucose production with low-dose insulin. Liraglutide significantly decreased circulating NEFA in the fasting state, low-dose and high-dose insulin states. Liraglutide significantly reduced the insulin concentration required to V2-maximally suppress circulating NEFA. Liraglutide significantly decreased adipose tissue lipolysis, as demonstrated by a reduction in interstitial fluid glycerol concentrations. Furthermore, Liraglutide significantly improved serum markers of adipose inflammation, namely leptin, adiponectin, and CCL-2. In addition, Liraglutide decreased de novo lipogenesis in-vivo, as measured by incorporation of deuterated 2H20 into palmitate, versus placebo. Endorsing our clinical observations, in-vitro experiments using both the Huh-7 human hepatoma cell line and primary cultures of human hepatocytes showed decreased de novo lipogenesis, measured by 14C-acetate incorporation into cellular lipid following treatment with GLP-1 (exendin-4 10nM, 100nM). Conclusions: Liraglutide significantly reduces metabolic dysfunction, hepatic lipogenesis, hepatic/adipose insulin resistance and adipose inflammation in patients with NASH. It is possible that GLP-1 analogue therapy may represent a novel treatment for patients with NASH, although the safety and histological efficacy await the completion of the 48week LEAN trial.

Disclosures:

Matthew J. Armstrong - Grant/Research Support: Novo Nordisk Ltd

Stephen Gough - Advisory Committees or Review Panels: Novo Nordisk, Eli Lilly, Sanofi Aventis, Takeda, GSK; Grant/Research Support: Novo Nordisk, Eli Lilly, Takeda; Speaking and Teaching: Novo Nordisk, Eli Lilly, Sanofi Aventis, GSK

Philip N. Newsome - Grant/Research Support: Novo Nordisk

The following people have nothing to disclose: Diana Hull, Kathy Guo, Darren Barton, Jonathan M. Hazlehurst, Maryam Nasiri, Laura L. Gathercole, Jinglei Yu, Piers Gaunt, Jeremy W. Tomlinson

563

  1. Top of page

Evaluation of gallbladder kinetics in patients with nonalcoholic fatty liver disease

Yasar Colak1, Gulcin Bozbey2, Ebubekir Senates3, Levent Doganay4, Ender Coskunpinar5, Oguzhan Ozturk1, Banu Mesci6, Ihsan Kuru2, Guralp Tasan1,Yusuf Yilmaz7, Ilyas Tuncer1
1Gastroenterology, Istanbul Medeniyet University' Goztepe Teaching and Research Hospital, Istanbul, Turkey; 2Radiology, Istanbul Medeniyet University, Goztepe Teaching and Research Hospital, Istanbul' Turkey; 3Gastroenterology, Dicle University, Diyarbakir, Turkey; 4Gastroenterology, Umraniye Education and Research Hospital, Istanbul, Turkey; 5Molecular Medicine, Istanbul University, Experimental Medicine Research Center, Istanbul, Turkey; 6Internal Medicine, Istanbul Medeniyet University, Goztepe Teaching and Research Hospital, Istanbul, Turkey; 7Gastroenterology, Marmara University, Institute of Gastroenterology, Istanbul, Turkey

Background&Aims: Nonalcoholic fatty liver disease (NAFLD) and cholelithiasis are among the most frequent diseases of gastrointestinal diseases. Recently, it was reported that cholelithiasis is approximately two times more often (20%) than normal in patients with NAFLD. However, to date no study showed a casual relationship and gallbladder kinetics in patients with NAFLD. In this study we aimed to evaluate the gallbladder kinetics and probable relationship with NAFLD. Material and Methods: A total of 50 patients diagnosed histopathologically with NAFLD and 38 healthy controls were included in the study. After an 8 hour overnight fasting the measurements of gallbladder were performed and after standard food ingestion the measurements were repeated at 45. minutes. Results: Fasting gallbladder wall thickness (1.12±0.38 and 1.48±0.46 mm respectively, p<0.001), volume (21.73±11.1 and 27.86±8.4 ml respectively, p=0.006) and postprandial residual volume (10.73±5.7 and 1 7.84±8.1 8 respectively, p<0.001)of patients with NAFLD was significantly higher than controls, whereas gallbladder ejection fraction (48±19.15 and 36.8±19.3% respectively, p= 0.008) was significantly lower than controls. When all individuals divided as control, simple steatosis and nonalcoholic steatohepatitis (NASH) groups, fasting gallbladder wall thickness (1.12±0.38, 1.31±0.37, 1.52±0.48 mm respectively, p=0.001), fasting gallbladder volume (21.73±11.1 27.36±11.1 27.94±8.55 ml respectively, p=0.019,) postprandial residual volume ((10.73±5.7, 16.8±7.6, 18.1 ±8.5 ml respectively, p<0.001) and gallbladder ejection fraction (48±19.15, 38.4±16.6, 36.1 ±20.2% respectively, p=0.023) were significantly different between three groups. In linear regression analysis, we found that severity of histological steatosis in patients with NAFLD was independent predictor of gallbladder ejection fraction (p = -0.414, t = -2.275, P=0.028). Conclusions: Increased gallbladder fasting volume and disturbed gallbladder emptying may be related by increased gallstone formation in patient with NAFLD. Additionally increased gallbladder wall thickness in patients with NAFLD may be associated with steatosis of gallbladder wall and this association could also affect the ejection fraction. However, our data should be supported with more comprehensive studies.

Disclosures:

The following people have nothing to disclose: Yasar Colak, Gulcin Bozbey, Ebubekir Senates, Levent Doganay, Ender Coskunpinar, Oguzhan Ozturk, Banu Mesci, Ihsan Kuru, Guralp Tasan, Yusuf Yilmaz, Ilyas Tuncer

564

  1. Top of page

Controlled Attenuation parameter (CAP) for the diagnosis of steatosis: a prospective study of 5, 323 examinations

Victor de Ledinghen1 , 2Julien Vegniol1, Faiza Chermak1, Wassil Merroucne1, Juliet Foucher1, Maylis Capdepont1, Brigitte Le Bail2 , 3
1Hepatology Unit, Hopital HautLeveque, Pessac, France; 2INSERM U1053, Université Bordeaux Segalen, Bordeaux, France; 3PathoIogy Unit, CHU Bordeaux, Bordeaux, France

Background & Aims: Controlled attenuation parameter (CAP) evaluated with transient elastography (FibroScan) is a recent and performant method for non invasive assessment of steatosis. Values range from 100 to 400 dB/m. However, its usefulness in clinical practice is unknown. We prospectively investigated factors associated with CAP failure and elevated CAP values in a large cohort of consecutive patients. Methods: CAP failure was defined as zero valid shot. The following factors were analyzed for their influence on CAP failure measurement and elevated CAP value (> 300 dB/m, cutoff value for moderate to severe steatosis): age, gender, body mass index (BMI), waist circumference, hypertension, diabetes, metabolic syndrome, alcohol use, liver stiffness measurement. Some patients had CAP measurement and liver biopsy the same day, and we evaluated the performance of CAP for the diagnosis of steatosis. Results. From April 2009 through November 2012, 4451 patients were included (mean age 55 years, male gender 54%, BMI 27 kg/m2). CAP failure occurred in 7.7% of 5, 323 examinations. By multivariate analysis, clinical factors independently associated with CAP measurement failure were age > 55 years (OR 1.56, 95%CI 1.24-1.97, p<0.001), male gender (OR 0.76, 0.62-0.95, p=0.05), BMI > 30 kg/m2 (OR 25.55, 16.44-39.70, p<0.001), and metabolic syndrome (OR 1.48, 1.17-1.87, p=0.001). By multivariate analysis, factors independently associated with CAP > 300 dB/m were BMI > 30 kg/m2 (OR 11.68, 95%CI 9.15-14.90, p<0.001), metabolic syndrome (OR 2, 35, 1.97-2.80, p<0.001), alcohol abuse (OR 1.74, 1.38-2.19, p<0.001) and liver stiffness > 6 kPa (OR 1.80, 1.52-2.14, p<0.001). A liver biopsy was available in 423 cases. Steatosis was < 10% in 205 cases (48.5%), between 11 and 33% in 85 cases (20.1%), between 34 and 66% in 71 cases (16.8%), and > 66% in 62 cases (14.7%). CAP was correlated with steatosis (r=0, 456, p<0.0001) but not with fibrosis (r=0, 095, p=0, 01). For the diagnosis of steatosis > 10%, steatosis > 33%, and steatosis > 66%, AUROCs of CAP were 0.79 (95%CI 0.74-0.84, p<0.0001), 0.84 (95%CI 0.80-0.88, p<0.0001), 0.84 (95%CI 0.80-0.88, p<0.0001), respectively. Conclusion. CAP is a promising tool for the noninvasive quantification of steatosis. Moreover, its strong association with alcohol use and metabolic syndrome could be useful for the follow-up of NAFLD and alcoholic patients in clinical routine and trials.

Disclosures:

Victor de Ledinghen - Advisory Committees or Review Panels: Merck, Janssen, Gilead, Echosens, Boehringer Ingelheim, Abbvie; Grant/Research Support: Roche, Gilead, Janssen; Speaking and Teaching: Roche, Echosens

Juliette Foucher - Board Membership: roche; Speaking and Teaching: BMS, MSD, Gilead

The following people have nothing to disclose: Julien Vergniol, Faiza Chermak, Wassil Merrouche, Maylis Capdepont, Brigitte Le Bail

565

  1. Top of page

MR-Spectroscopy at 3.0- and 7.0-Tesla in NAFLD as invivo tool for monitoring changes in fat and energy metabolism with potential identification of NASH and advanced fibrosis

Stefan Traussnigg1, Christian Kienbacher1, Emina Halilbasic1, Ma「tin Gajdosik2, Ladislav ValkoviČ2' Marek Chmelk2' Judith Stift3, Harald Hofer1, Petra E. Steindl-Munda1, Lili Kazemi-Shiiazi1, Markus Peck-Radosavljevic1, Peter Ferenci1, Ahmed Ba-Ssalamah4' Friz Wrba3, Michael Krebs5, Siegfried Trattnig2, Martin KrŠŠák2,5, Michael Tauner1
1Internal Medicine III - Division of Gastroenterology and Hepatology, Medical University Vienna, Vienna, Austria; 2Radiodiagnostics - MR Centre-of-Excellence, Medical University Vienna, Vienna, Austria; 3Clinical Pathology, Medical University Vienna, Vienna, Austria; 4Radiodiagnostics, Medical University Vienna' Vienna' Austria; 5Internal Medicine III - Division of Endocrinology and Metabolism, Medical University Vienna, Vienna, Austria

Background and Aims: With the rising incidence of non-alcoholic fatty liver disease (NAFLD) with potentially progressive steatohepatitis (NASH) non-invasive tools for risk stratification and follow-up are urgently needed. We therefore used high field magnetic-resonance spectroscopy (MRS) as non-invasive tool to assess NASH and fibrosis profiles and obtain novel mechanistic and pathogenetic insights into alterations of hepatic metabolism in NAFLD. Methods: MRS and liver biopsy were performed back-to-back on the same day in suspected NAFLD/NASH patients and data were correlated with histology (Kleiner/SAF). Hepatocellular lipid content (HCL) was measured by 3.0-T 1H-MRS and 7.0-T 31P-MRS was applied to determine phosphomonoester (PME), phosphodiester (PDE), phosphocreatine (PCr), NADPH, inorganic phosphate (Pi), a-, p- and y-ATP as well as total phosphate (TP). Results 24 patients (10 female, 14 male) were included. Median age was 51 years (24-70). Histological diagnosis was simple steatosis (SS; n=5) and NASH (n=19; 4 w/ cirrhosis). No difference in BMI or waist-to-hip ratio was observed between both groups. Steatosis assessed by 1H-MRS correlated well with histological data which was improved using a non-linear logarithmic equation [Y=-2.282+25.021*LOG(FAT)] resulting in a better r2 value (r2=0.727; p<0.001) leading to a more accurate measurement of lower grades of HCL. All NASH patients showed significantly higher values of steatosis in histology and MRS compared to SS (r=0.69, p<0.001). A distinction between SS and NASH was feasible by MRS (Cut-off: 0.41; AUROC 0.795, PPV 0.93, NPV 0.5). PME+PDE/TP ratio as marker of cell membrane alterations increased with risk groups (r=0.401; p=0.05) defined as SS, NASH with low (0-2) and high stages of fibrosis (3+4). Conversely, NADPH/PME+PDE reflecting mitochondrial function decreased (r=-0.385, p=0.06) and PCr/TP increased (r=0.537, p=0.007) in higher stages of fibrosis whereas no changes in overall ATP levels were detected. Conclusion: High field 1H-MRS signals strongly correlate with histological grades of steatosis which improved by logarithmic translation showing also differences between simple steatosis and NASH. In vivo 7.0 T 31P-MRS shows promising results indicating changes in hepatic cell membrane and energy metabolism in inflammation and fibrosis associated with NASH. Non-invasive risk profiling in NAFLD appears feasible but further validation and follow-up is required.

Disclosures:

Harald Hofer - Speaking and Teaching: Janssen, Roche, MSD

Markus Peck-Radosavljevic - Advisory Committees or Review Panels: Bayer, Gilead, Janssen, BMS; Consulting: Bayer, Boehringer-Ingelheim, Jennerex, Eli Lilly; Grant/Research Support: Bayer, Roche, Gilead, MSD; Speaking and Teaching: Bayer, Roche, Gilead, MSD, Eli Lilly

Peter Ferenci - Advisory Committees or Review Panels: Roche, Idenix, Roche, MSD, Vertex, Salix, Madaus Rottapharm, Tibotec, B√δhringer Ingelheim, Achilleon, GSK; Grant/Research Support: MSD, Vertex, Madaus Rottapharm; Patent Held/Filed: Madaus Rottapharm; Speaking and Teaching: Roche, Gilead, Roche, Gilead, Salix

Michael Trauner - Advisory Committees or Review Panels: MSD, Janssen; Consulting: Falk Pharma, Phenex, Amgen; Grant/Research Support: Intercept; Patent Held/Filed: Med Uni Graz (norUDCA); Speaking and Teaching: Falk Foundation, Roche, Gilead

The following people have nothing to disclose: Stefan Traussnigg, Christian Kienbacher, Emina Halilbasic, Martin Gajdosik, Ladislav Valkovic, Marek Chmelik, Judith Stift, Petra E. Steindl-Munda, Lili Kazemi-Shirazi, Ahmed Ba-Ssalamah, Fritz Wrba, Michael Krebs, Siegfried Trattnig, Martin KrŠŠák

566

  1. Top of page

Small dense low-density lipoprotein as a useful biomarker for predicting arteriosclerotic diseases in nonalcoholic steatohepatitis

Kento Imajo1,Masato Yaneda1, Takaomi Kessoku1, Wataru Tomeno1, Yuji Ogawa1, Hironori Mawotari1, Hiroyuki Kirikoshi1, Yoshio Sumida2, Hideyuki Hyogo3, Yuichiro Eguchi4, Masatumi Onc5, Satoru Saito1,Koichiro Wada6, Atsushi Nakajima1
1Division of Gastroenterology, Division of Gastroenterology, Yokohama City University School of Medicine, Yokohama city, Japan; 2Center for Digestive ond Liver Diseases, Nara, City Hospital, Nara, Japan; 3Department of Medicine and Molecular Science, Hiroshima University, Hiroshima, Japan; 4Division of Hepatology Saga Medical School, Saga, Japan; 5Department of Gastroenterology and Hepatology, Kochi Medical School, Kochi, Japan; 6Department of Pharmacology, Osaka University Graduate School of Dentistry, Suita, Japan

Introduction: Nonalcoholic fatty liver disease (NAFLD) is closely correlated with insulin resistance and several metabolic syndrome features. Although some investigations have shown a relationship between NAFLD and arteriosclerotic diseases, there are few studies elucidating the mechanisms. We recently showed that very low-density lipoprotein 1(VLDL1), a TG-rich lipoprotein, is increased in patients with nonalcoholic steatohepatitis (NASH) when compared with those with NAFL (nonalcoholic fatty liver) (Hepatology 2009). VLDL1 is known to be predominantly metabolized to small dense low-density lipoprotein (sdLDL), a strong risk factor for arteriosclerotic diseases. The aim of this study was to investigate the relationship between NAFLD and the risk factors for development of arteriosclerotic diseases, especially small dense LDL. Methods: Our cohort consisted of 124 patients with ultrasonography-diagnosed NAFLD including 35 patients with steatohepatitis and 41 patients with nonalcoholic fatty liver (NAFL) proven by liver biopsy, 102 agematched patients with other liver diseases including viral hepatitis, autimmune hepatitis and primary biliary cirrhosis, and 30 age-matched healthy controls. Serum total cholesterol (TC), triglyceride (TG), high-density lipoprotein-cholesterol (HDL-C) and direct LDL-C were assayed enzymatically. In addition, sdLDL, together with very low-density and intermediate-density lipoproteins (VLDL and IDL) was measured by polyacrylamide gradient gel electrophoresis (PAGE). Results: Compared to patients with other liver diseases and controls, patients with NAFLD showed significantly increased TC, TG, normal-size LDLC, non-HDL-C (cholesterol level subtracted HDL-C from TC) and sdLDL levels (Figure 1) with decreased HDL-C (all P < 0.01). In addition, patients with NASH showed significantly increased non-HDL-C and sdLDL levels (Figure 1) with decreased VLDL when compared with those with NAFL. An association between sdLDL and NAFLD activity score (NAS) was reflected in the significantly positive correlation present between these two variables in the patients with NAFLD (r=0.38, P<0.005). Conclusion: NASH is previously reported to be associated with significantly higher non-HDL-C, an independently risk factor for cardiovascular disease, when compared to NAFL. This is a first report that patients with NASH had higher sdLDL levels than those with NAFL, suggesting that the risk of arteriosclerotic diseases may be higher in NASH than NAFL. In addition, this biomarker may aid in distinguishing patients with a high risk of arteriosclerotic disease in NASH.

Disclosures:

The following people have nothing to disclose: Kento Imajo, Masato Yoneda, Takaomi Kessoku, Wataru Tomeno, Yuji Ogawa, Hironori Mawatari, Hiroyuki Kirikoshi, Yoshio Sumida, Hideyuki Hyogo, Yuichiro Eguchi, Masafumi Ono, Satoru Saito, Koichiro Wada, Atsushi Nakajima

567

  1. Top of page

Prevalence and determinants of nonalcoholic steatohepatitis (NASH) in a multicenter cohort of severely obese adolescents undergoing weight loss surgery

Stavra A. Xanthakos1, Tawny P. Wilson2, David E. Kleiner3, Todd M. Jenkins2, Reena Mourya1,Mary L. Brandt4, Carroll M. Hormon5, Michael A. Helmroth2, Marc Michalsky6, Anita P Courcoulos7, Meg H. Zeller8, Thomas H. Inge2
1Division of Gastroenterology, Hepatology, & Nutrition, Cincinnati Children's Hospital Medical Center, Cincinnati, OH; 2Division of Pediatric Surgery, Cincinnati Children's Hospital Medical Center, University of Cincinnati College of Medicine, Cincinnati, OH; 3Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, MD; 4Division of Pediatric Surgery, Texas Children's Hospital, Baylor College of Medicine, Houston, TX; 5Division of Pediofric Surgery, Children' Hospital of Alabama, University of Alabama, Birmingham, AL; 6Division of Pediatric Surgery, Nationwide Children's Hospital, The Ohio State College of Medicine, Columbus, OH; 7Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA; 8Division of Behavioral Medicine, Cincinnofi Children's Hospital Medical Center, Cincinnota, OH

Aim: While adolescent weight loss surgery (WLS) practice guidelines include severe NASH as an indication for surgery, a major nonalcoholic fatty liver disease (NAFLD) practice guideline does not endorse WLS as a treatment for NASH due to lack of controlled outcome data. The current prevalence of NASH among adolescents undergoing WLS is unknown. The aim of our prospective study was to determine the prevalence and determinants of NASH in severely obese adolescents at time of WLS. Methods: A total of 242 adolescents, ages 13-19 yrs, were enrolled in the multicenter observational Teen Longitudinal Assessment of Bariatric Surgery study from March 2007 to December 2011; 157 had routine intra-operative core liver biopsies. Exclusion criteria were insufficient tissue (n=3), medications that may cause or treat NASH (n=1 3) and alcohol intake >20 gm/day (n=0). After exclusions, 141 remained. An experienced hepatopathologist (DEK) graded liver biopsies using the NASH Clinical Research Network scoring system. Results: Mean age of the 141 subjects was 16.8 ± 1.6 yrs; 28% were male, 70% white, and 8% Hispanic. Median BMI was 52 kg/m2 (Interquartile range (IQR): 47, 60); 47% had hypertension, 14% diabetes and 78% dyslipidemia. NAFLD was present in 59.6%, borderline NASH in 14.2% and definite NASH in 6.4%. Stage 1 fibrosis was present in 15.6% and stage 2 in 2.8%; 1 subject had stage 3 and none had cirrhosis. Compared to subjects with Not-NAFLD and with NAFLDNot NASH, more subjects with borderline/definite NASH were male and had diabetes and hypertension. Serum ALT, AST, HOMa-IR, and triglyceride levels were also higher. Pre-operative weight loss >5% occurred in 11% of subjects but did not vary by disease grade. Conclusions: The presence of NASH in a multicenter cohort of severely obese adolescents undergoing WLS was associated with male gender and higher cardiometabolic risk. While NAFLD was common, prevalence of advanced fibrotic NASH was low. This may reflect younger age, demographic or referral patterns, or biological factors specific to severe obesity and merits further study.

Characteristics and significant determinants of liver histology

Not-NAFLD (n=57)NAFLD-Not NASH (n=55)Borderline/Definite NASH (n=29)
  1. NAS=NAFLD activity score

NAS score, mean (SD)0.60 (0.53)2.1(0.74)3.6 (1.7)<0.0001
Fibrosis score, mean (SD)0.07 (0.26)0.15 (0.40)0.72 (0.80)<0.0001
Male, n (%)11(19.3%)15 (27.3%)14 (48.3%)0.007
Diabetes, n (%)4 (7.0%)5 (9.1%)10(34.5%)0.001
Hypertension, n (%)21(36.8%)26 (47.3%)19 (65.5%)0.01
ALT U/L, median (IQR)20(17, 31)27 (22, 38)32 (25, 54)<0.0001
AST U/L, median (IQR)29 (22, 41)32 (23,40)40 (27, 61)0.0002
HOMA-IR, median (IQR5.7 (3.6, 7.5)7.2 (4.2, 10.2)7.5 (4.9,14.1)0.002
Triglycerides, mg/dL, median (IQR)103 (77, 133)121(90, 173)139 (93, 215)0.004

Disclosures:

Marc Michalsky - Grant/Research Support: Allergan Health, Irvine CA Thomas H. Inge - Grant/Research Support: Ethicon

The following people have nothing to disclose: Stavra A. Xanthakos, Tawny P. Wilson, David E. Kleiner, Todd M. Jenkins, Reena Mourya, Mary L. Brandt, Carroll M. Harmon, Michael A. Helmrath, Anita P. Courcoulas, Meg H. Zeller

568

  1. Top of page

Nonalcoholic fatty liver disease is a negative risk factor for biochemical recurrence of prostate cancer after radical prostatectomy

Won-Mook Choi1, Jeong-Hoon Lee1,Young Ju Lee2,Young Youn Cho1, Yuri Cho1,Dong Hyeon Lee1, Su Jong Yu1, Yoon Jun Kim1, Jung-Hwan Yoon1, Cheol Kwak2, Hyo-Suk Lee1
1Department of Internal Medicine and Liver Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea; 2Department of Urology and Clinical Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea

Background & Aims: Nonalcoholic fatty liver disease (NAFLD) is closely related to metabolic syndrome and obesity which are associated with an increased risk of various malignancies. In this study, we investigated the association between NAFLD and prostate cancer biochemical recurrence (BCR) after radical prostatectomy (RP). Methods: Consecutive prostate cancer patients who underwent RP between 2005 and 2008 at a single tertiary hospital in Korea were included in this study. The presence of NAFLD, body mass index (BMI), pre-diagnostic prostate-specific antigen (PSA), and histological findings including Gleason score were analyzed with regard to their associations with BCR. NAFLD was diagnosed based on clinical information and ultrasonography or unenhanced CT images. BCR-free survival rates were calculated using the Kaplan-Meier method. Results: A total of 222 patients were analyzed. During a median follow-up period of 54 (inter-quartile range, 44-65) months, 45 (20.3%) patients developed BCR. The presence of NAFLD was significantly associated to longer time-to-BCR (P=0.001 by log rank test), while BMI failed to show statistical significance (P=0.861). In multivariate analysis, the presence of NAFLD (hazard ratio [HR], 0.26; 95% confidence interval [CI], 0.11-0.61; P=0.002), pathological Gleason score (compared to <7, 7: HR, 2.92; 95% CI, 1.12-7.64; P=0.029, >7: HR, 6.64; 95% CI, 2.26-19.52; P=0.001), and positive surgical margin (HR, 2.17; 95% CI, 1.18-3.99; P=0.013) were independent predictive factors of BCR. Conclusions: Our study demonstrated that NAFLD may play a protective role against BCR after RP for prostate cancer. It is warranted to elucidate the mechanism of protective effect in patients with NAFLD.

Figure 1. Time-to-BCR according to the presence of NAFLD.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Won-Mook Choi, Jeong-Hoon Lee, Young Ju Lee, Young Youn Cho, Yuri Cho, Dong Hyeon Lee, Su Jong Yu, Yoon Jun Kim, Jung-Hwan Yoon, Cheol Kwak, Hyo-Suk Lee

569

  1. Top of page

Serum Metabolomic Profile and Potential Biomarkers for Activity and Steatosis in Nonalcoholic Fatty Liver Disease

Katsutoshi Tokushige, Etsuko Hashimoto, Kazunisa Kodama, Maki Tobari, Norko Matsushita, Tomomi Kogiso, Makiko Taniai' Nobuyuki Tom, Keiko Shiratori
Tokyo Women's Medical University, Tokyo, Japan

(Introduction) Useful biomarkers for diagnosing nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH) are anticipated. In order to discover novel biomarkers, especially for activity and steatosis in NAFLD, we performed metabolomic screening. (Patients) This study included 105 NAFLD patients and 48 healthy controls. Using capillary electrophoresis and liquid chromatography with mass spectrometry, we analyzed low molecular-weight metabolites. Both activityrelated and steatosis-related metabolites were detected. Predictive calculation systems of activity and steatosis were established, and area under the curve (AUC), sensitivity and specificity for diagnosis were investigated. (Results) 1. Twelve significant metabolites for activity were detected. Five were amino acids or amino acid-related molecules, 1 bile acid, and 1 nitric oxide-related molecule, as well as others. Pro is the best metabolite for activity detection (5.84 in mild activity vs 6.85 in moderate-severe activity, p=0.006). Predictive calculation system for moderate-severe activity was established with 7 metabolites. AUC was 0.66, sensitivity 70% and specificity 64%.2. Sixteen significant metabolites for steatosis were detected. Ten were amino acids or amino acid-related molecules, 2 bile acids, and 1 fatty acid-related molecule, as well as others. Gly is the best metabolite for steatosis detection (2.66 in mild steatosis vs 2.23 in moderate-severe activity, p=0.0016). Predictive calculation system of moderate-severe steatosis was established with 15 metabolites. AUC was 0.81, sensitivity 79% and specificity 77%. (Conclusion) Several metabolic products were found as biomarkers of activity and steatosis in NAFLD, and they could also be useful for diagnosis of these conditions. The diagnostic ability of these metabolites and the predictive calculation system will be confirmed by validation study.

Disclosures:

The following people have nothing to disclose: Katsutoshi Tokushige, Etsuko Hashimoto, Kazuhisa Kodama, Maki Tobari, Noriko Matsushita, Tomomi Kogiso, Makiko Taniai, Nobuyuki Torii, Keiko Shiratori

570

  1. Top of page

Fibromyalgia, Inflammation, and Steatohepatitis

Shari Rogal1, Klous Bielefeldt1, Brian M. Davis1, Francis Lotrich2, Eva Szigethy2, Andrea DiMartini2
1Division of Gastroenterology, Hepatology, and Nutrition, University of Pittsburgh, Pittsburgh, PA; 2Department of Psychiatry, University of Pittsburgh, Pittsburgh, PA

BACKGROUND: Hepatitis C virus (HCV) has been strongly associated with fibromyalgia, but fibromyalgia in patients with non-alcoholic steatohepatitis (NASH) has not been previously assessed. METHODS: We prospectively recruited patients in an outpatient hepatology clinic with cirrhosis due NASH, alcohol, and HCV. Patients with known inflammatory conditions and cancers were excluded. The American College of Rheumatology Criteria from 2010 were used to evaluate presence of fibromyalgia using the previously validated Fibromyalgia Survey Questionnaire. Psychiatric symptoms were assessed using the Hospital Anxiety and Depression Scale and the Pittsburgh Sleep Quality Inventory. Levels of inflammatory cytokines including IL-1 p, IL-6, TNF-a, and C-reactive protein were measured using standardized luminex assays at the time of the outpatient visit. Chi-square, Student's T-tests, and Wilcoxon Rank Sum tests were used for univariate comparisons. Multivariate logistic regression models with automated AIC optimization were used in order to determine the factors that were significantly associated with fibromyalgia in this population. Models were checked for multicolinearity using a prespecified VIF>5. RESULTS: The cohort consisted of 193 patients with cirrhosis. The mean age of the population was 5/.9±9.0, and 40% were women. The mean MELD score was 12.3±5.7. Cirrhosis was due to HCV in 78 (40%), NASH in 66 (34%), and alcohol in 49 (25%) patients. Fibromyalgia criteria were met by 27% of patients with NASH cirrhosis compared to 19% of patients with HCV cirrhosis, and 9% of patients with alcohol-related cirrhosis (p=0.04). Pro-inflammatory cytokine levels were not significantly different for patients by fibromyalgia status. In contrast, in multivariate analysis, fibromyalgia seemed to be most strongly related to depressive symptoms (OR=1.30, 95% CI = 1.10, 1.56), and sleep disturbance (OR=1.27, 95% CI =1.11, 1.48) in this population. Meeting criteria for fibromyalgia was significantly associated with increased self-reported disability (mean score 6 vs.3 on scales from 0-10, p<0.01) and increased pain scores (mean score of 2.4 vs.1.2 on a scale of 0-5, p<0.01). CONCLUSIONS: Though fibromyalgia has received attention as being related to HCV infection and is thought to be part of the sequelae of this disease, fibromyalgia was found more commonly in patients with NASH cirrhosis than HCV. Fibromyalgia was strongly related to psychiatric and symptoms rather than to inflammation.

Disclosures:

Eva Szigethy - Grant/Research Support: NIH; Speaking and Teaching: Merck

The following people have nothing to disclose: Shari Rogal, Klaus Bielefeldt, Brian M. Davis, Francis Lotrich, Andrea DiMartini

571

  1. Top of page

Genetic variant I148M in PNPLA3 is associated with acoustic radiation force impulse imaging in patients with NAFLD

Kentaro Yoshioka, Hiroaki Shimazaki, Naoto Kawabe, Masao Harata, Yoshifumi Nitta, Michihito Muroo, Takuji Nakano, Yuko Arima, Toshiki Kan, Masashi Ohki, Kazunori Nakaoka, Takagawa Yuko, Toru Nishikawa, Keisuke Osakabe, Naohiro Ichino, Senju Hashimoto
Department of Liver, Biliary Tract and Pancreas Diseases, Fujita Health University, Toyoake, Japan

Purpose: Nonalcoholic fatty liver disease (NAFLD) is an escalating medical problem worldwide. Anonsynonymous single nucleotide polymorphism rs738409 (I148M) in patatin-like phospholipase domain-containing protein 3 (PNPLA3) have been reported to associate with progression of non-alcoholic steatohepatitis (NASH). However the precise mechanism has not been elucidated. Acoustic radiation force impulse (ARFI) imaging is a noninvasive method for assessing liver fibrosis. The aim of the present study is to evaluate the correlation of PNPLA3 variants with velocity of shear wave (Vs) by ARFI, serum high molecular weight (HMW) adiponectin, leptin, TNFa, CK18, and various blood markers, and to elucidate the mechanism by which PNPLA3 variants play roles in NAFLD. The correlation of Vs with other parameters was also evaluated. Methods: Vs measurement by ARFI was performed with a Siemens ACUSON S2000 in 43 patients with NAFLD diagnosed with ultrasonography. PNPLA3 variant rs738409 was genotyped using a ĪaqMan assay. Written informed consent was obtained using a protocol approved by the ethics committee of Fujita Health University. Results: Vs was significantly higher in patients with PNPLA3 GG (2.31±0.92 m/s) than in those with CG/CC (1.46±0.74 m/s)(p=0.001 8). ALP levels (p=0.0166), total bilirubin levels (p=0.0123), platelet count (p=0.0271), hyaluronic levels (p=0.0030), andy-globulin levels (p=0.0455) also significantly associated with PNPLA3 variants. HMW adiponectin levels, Leptin levels, TNFa level, or CK18 levels were not significantly correlated with PNPLA3 variants. Vs significantly correlated with CK18 levels (r=0.4681, p=0.0027), albumin levels (r=-0.502, p=0.0013), platelet count (r=-0.560, p=0.0003), AST levels (r=0.502, p=0.0013), ALP levels (r=0.344, p=0.0343), prothrombin time (r=-0.655, p<0.01), hyaluronic levels (r=0.734, p<0.001), and γ-globulin levels (r=0.774, p<0.01). Vs tended to correlate with leptin level (r=0.299, p=0.0574) and TNFα levels (r=0.294, p=0.0623). Conclusions: The findings that Vs is associated with ugges assessing fibrosis stage in NAFLD. The association of PNPLA3 variants with Vs and other various fibrosis markers indicates that PNPLA3 variants affect fibrogenesis in NAFLD.

Disclosures:

Kentaro Yoshioka - Grant/Research Support: Chugai, Schering-Plough, Bristol Myers Squibb, Tanabe Mitsubishi, Taiho, Otsuka, Ajinomoto, Tore Medical, Torii, Boston,ÄÄScientific

The following people have nothing to disclose: Hiroaki Shimazaki, Naoto Kawabe, Masao Harata, Yoshifumi Nitta, Michihito Murao, Takuji Nakano, Yuko Arima, Toshiki Kan, Masashi Ohki, Kazunori Nakaoka, Takagawa Yuka, Toru Nishikawa, Keisuke Osakabe, Naohiro Ichino, Senju Hashimoto

572

  1. Top of page

The Combination of Liver Stiffness Measurement and NAFLD Fibrosis Score Improves the Nonivasive Diagnostic Accuracy for Severe Liver Fibrosis in Patients with Nonalcoholic Fatty Liver Disease

Salvatore Petta1, Ester Vanni2, Elisabetta Bugianesi2, Vito Di Marco1, Calogero Camma1, Maria Rosa Barcellona1, Daniela Cabibi3, Lavinia Mezzabota2, Antonio Caxi1
1Sezione di Gastroenterologia, DiBiMIS, University of Palermo, Italy, Palermo, Italy; 2Division of Gastro-Hepatology, Department of Internal Medicine, San Giovanni Battista Hospital, University of Torino, Torino, Italy, Torino, Italy; 3Sezione di Gastroenterologia, DiBiMIS, University of Palermo, Italy, Palermo, Italy

Background and Aims: The accuracy of nonivasive tools for the diagnosis of severe fibrosis in patients with nonalcoholic fatty liver disease (NAFLD) in clinical practice is still limited. We aimed at assessing the diagnostic performance of combined noninvasive tools in two independent cohorts of Italian NAFLD patients. Methods: We analyzed data from 321 Italian consecutive patients (179 Sicilian-training cohort, and 142 Northern Italy-validation cohort) with an histological diagnosis of NAFLD. Severe fibrosis was defined as fibrosis >F3 according to Kleiner classification. The APRI, AST/ALT, BARD, FIB-4, and NFS scores were calculated according to published algorithms. Liver stiffness measurement (LSM) was performed by FibroScan. Cut-off points of LSM, NFS and FIB-4 for rule-in or rule-out F3-F4 fibrosis were calculated by the reported formulas. Results: In the Sicilian cohort AUCs of LSM, NFS, FIB-4, LSM plus NFS, LSM plus FIB-4, and NFS plus FIB-4 were 0.857, 0.803, 0.790, 0.878, 0.888 and 0.807, respectively, while in the Northern Italy cohort the corresponding AUCs were 0.848, 0.730, 0.703, 0.844, 0.850, and 0.733 respectively. In the training cohort the combination of LSM plus NFS was the best performing strategy, providing false positive, false negative and uncertainty area rates of 0%, 1.1% and 48%, respectively. Similar results were obtained in the validation cohort with false positive, false negative and uncertainty area rates of 0%, 7.3% and 40.8%. Conclusions: The combination of LSM with NFS, two complementary, easy to perform, and widely available tools, is able to accurately diagnose or exclude the presence of severe liver fibrosis, also reducing of about 50%-60% the number of needed diagnostic liver biopsies.

Disclosures:

The following people have nothing to disclose: Salvatore Petta, Ester Vanni, Elisabetta Bugianesi, Vito Di Marco, Calogero Cammá, Maria Rosa Barcellona, Daniela Cabibi, Lavinia Mezzabotta, Antonio Craxi

573

  1. Top of page

Glucokinase regulatory protein (GCKR) Gene Polymorphism Affects Liver Fibrosis in Non-alcoholic Fatty Liver Disease

Salvatore Petta1, Luca Mieie2, Elisabetta Bugianesi3, Calogero Camma1, Chiara Rosso3, Daniela Cabibi4, Vito Di Marco1, Stefania Grimaudo1, Antonio Greco2, Rosaria Maria Pipitone1, Giulio Marchesini5, Antonio Craxi1
1Sezione di Gastroenterologia, DiBiMIS, University of Palermo, Italy, Palermo, Italy; 2Institute of Internal Medicine, School of Medicine, Catholic University of the Sacred Heart, Rome ITALY, Roma, Italy; 3Division of Gastro-Hepatology, Department of Internal Medicine, San Giovanni Battista Hospital, University of Torino, Torino, Italy, Torino, Italy; 4Cattedra di Anatomia Patologica, University of Palermo, Italy, Palermo, Italy; 5Dipartimento di Medicina e Gastroenterologia' "Alma Mater Studiorum, '' Universitá di Bologna, Italy, Bologna, Italy

Background and Aims: Variant in glucokinase regulatory protein (GCKR), associated with lipid and glycemic traits, has been suggested to affect the fatty liver infiltration. We aimed to assess whether GCKR rs780094 C->T SNPs influence the expression of steatosis, lobular inflammation and fibrosis in NAFLD patients. Methods: In 366 consecutive NAFLD patients (197 from Sicily, and 169 from center/northern Italy), we assessed anthropometric, biochemical and metabolic features; liver biopsy was scored according to Kleiner. PNPLA3 rs738409 C>G and GCKR rs780094 C>T single nucleotide polymorphisms were also assessed. Results: at multivariate logistic regression analysis in the entire cohort of NAFLD patients, the presence of significant liver fibrosis was independently linked to high BMI (OR 1.063, 95% CI 1.001-1.129, p= 0.04), high HOMA (OOR 1.155, 95% CI 1.040-1.283, p= 0.007), severity of liver steatosis (OR 1.425, 95% CI 1.028-1.976, p= 0.03), moderate-severe lobular inflammation (OR 3.727, 95% CI 2.147-6.471, p< 0.001), and GCKR C>T SNP (OR 1.885, 95% CI 1.311-2.710, p= 0.001). Similar results were observed considering separately the different NAFLD cohorts. GCKR C>T SNP was also associated with higher serum triglycerides levels (ANOVA, p=0.02) in the entire cohort. Conclusions: In patients with NAFLD, GCKR rs780094 C>T is associated with the severity of liver fibrosis and with higher triglycerides serum levels.

Disclosures:

Giulio Marchesini - Advisory Committees or Review Panels: Sanofi-Synthelabo; Grant/Research Support: Merck Sharp & Dome; Speaking and Teaching: Novo Nordisk, Merck Sharp & Dome, Boerhinger Ingelheim, Lilly

The following people have nothing to disclose: Salvatore Petta, Luca Miele, Elisabetta Bugianesi, Calogero Cammá, Chiara Rosso, Daniela Cabibi, Vito Di Marco, Stefania Grimaudo, Antonio Grieco, Rosaria Maria Pipitone, Antonio Craxi

574

  1. Top of page

Sex Hormone Influence on Pediatric NAFLD Histology

Joel E. Lavine1, Elana B. Mitchel1, Kathleen Viveiros2, Katherine P Yates3, Janis Durelle4, Jeffrey B. Schwimmer4, Cynthia A. Behling4, Aynur Unalp-Arida3
1Pediatric Gastroenterology Hepatology and Nutrition, Columbia University, New York, NY; 2Medicine, Tufts University, Boston, MA; 3Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD; 4Pediatrics, University of California, San Diego, San Diego, CA

Background: NAFLD is the most common chronic liver disease in American children. NAFLD is more prevalent in boys, who are older and later in puberty. We sought to determine the relation of sex hormones with histologic pattern and severity in pediatric NAFLD. Methods: 73 children (<18 y) with biopsied NAFLD were evaluated in a cross-sectional prospective studyfrom the NASH CRN. Clinical data, sex hormone levels, and histologic features were compared between sexes. Hormones were assayed using ELISA (androstenedione, DHEA-S, estradiol, estrone, leptin, SHBG, testosterone and adiponectin). Scored histology features included steatosis, inflammation and fibrosis as well as NAFLD patterns assessed by NASH CRN pathologists. Odds ratios and 95% CI's were calculated between each sex hormone and histologic feature adjusted for sex, ethnicity and Tanner stage. Independent predictors of NASH diagnosis were determined from a stepwise multinomial logistic regression. Results: Subjects were 12.9 ± 2.6 yrs, 68% male, 52.1% Hispanic, average BMI Z-score 2.3 ± 0.4 and Tanner stage 2.6 ±1.5. Overall, 37% had “NAFLD, No NASH, ” 14% had Zone 3 NASH, 22% had Zone 1 NASH and 27% had “Definite NASH”.32% had mild, 27% had moderate and 41% had severe steatosis. Sex hormones influenced NAFLD histology and patterning. Estradiol and estrone were significantly decreased in Zone 1 NASH vs. “NAFLD, No NASH” (Estradiol= 52.9土 15.5 vs.73.3土 30.8pg/mL, OR=0.62, p=0.01; Estrone= 40.38± 13.3 vs.61.9± 27.6 pg/mL, OR=0.58, p=0.007). Estrogen precursors were higher in mild steatosis as compared to moderate/severe (Estradiol= 66.3± 23.9 vs.64.5± 22.7pg/mL, OR=1.28, p=0.11; Estrone= 54.1± 24.9 vs.50.8± 23.7pg/mL, OR=1.24, p=0.14). Androstene-dione was decreased significantly in Zone 1 NASH vs. “NAFLD, No NASH” (1.01± 0.8 vs.1.2± 0.9ng/mL, OR=0.28, p=0.03). Adiponectin was significantly decreased in subjects with Definite NASH vs. “NAFLD, No NASH” (6467土 2834 vs.8676土 4461ng/mL; OR=0.80, p=0.04). Adiponectin and SHBG were decreased in moderate/severe vs. mild steatosis (8049±3354 vs.9862± 4630ng/mL, OR=0.85, p=0.02; 18.9± 9.4 vs.35, 4± 50.1nmol/L, OR 0.91, p = 0.004). Independent predictors of NASH diagnosis were gender, Tanner stage, DHEA-S, estrone, AST and GGT. Conclusions: Sex hormones associate with various aspects of NAFLD histology and patterning and may explain gender differences in NAFLD prevalence, histologic pattern and severity. We speculate these hormones signal through trans-activating mediators of lipid metabolism and inflammation. These clinically significant results warrant validation in larger cohorts.

Disclosures:

Joel E. Lavine - Consulting: Merck, Crosscare; Grant/Research Support: Janssen

Cynthia A. Behling - Grant/Research Support: NASH CRN

The following people have nothing to disclose: Elana B. Mitchel, Kathleen Viveiros, Katherine P. Yates, Janis Durelle, Jeffrey B. Schwimmer, Aynur Unalp Arida

575

  1. Top of page

Association of PNPLA3 genotype and clinical and molecular features of NAFLD patients

Yuji Hodo, Masao Honda, Hajime Sunagozaka, Tetsuro Shimakami, Kuniaki Arai, Tofsuyo Yamashita, Eishiro Mizukoshi, Toshinori Takamura, Shuichi Kaneko
Disease Control ond Homeostasis, Kanazawa University, Kanazawa, Japan

Background and Aims: Nonalcoholic fatty liver disease (NAFLD) is recognized as an important health concern. NAFLD includes a broad range of liver pathologies from simple steatosis, necrosis, and nonalcoholic steatohepatitis (NASH) to cirrhosis. Recently, genome wide association studies identified genetic variation rs738409 in PNPLA3 was associated with the development of NAFLD and progression of NASH. However, the mechanism of such association is unknown. Here, we investigated the association between the PNPLA3 genotype and clinical and molecular features of NAFLD patients. Methods: Genotyping of 58 histologically diagnosed NAFLD patients for the PNPLA3 rs738409 and the IL28B rs12979870 was carried out. The results were analyzed to determine the association between the PNPLA3 genotype (rs738409) and clinical and histological traits. Gene expression profiles in the liver of 58 patients were also analyzed using Affymetrix gene chip (U133 Plus 2.0). Pathway analysis was performed by gene set enrichment analysis. Results: The CC, CG and GG proportions of the rs738409 were 27.6% (16/58), 46.5% (27/58) and 25.9% (15/58), respectively. There was a significant increase in the frequency of the G allele from Matteoni's classification type1 (35.7%) to type4 (63.2%) (p= 0.028). Brunt grade and stage were also associated with the frequency of the G allele, whereas steatosis grade was not associated. Rs12979870 was not associated with clinical and histological traits. Gene expression analysis of the liver demonstrated 550 genes were differentially expressed between the PNPLA3 CC genotype and CG/GG genotype with the filtering criteria of p-values are smaller than 0.001. Some of the differentially expressed genes were reported to be associated with the pathogenesis of NASH. Gene set enrichment analysis demonstrated the Gene Ontology (GO) gene sets associated with cell cycle were significantly up-regulated in CG/GG genotype than CC genotype, while GO gene sets associated with defense response were significantly down-regulated in CG/GG genotype than CC genotype. Conclusion: The PNPLA3 genotype is associated with histological traits of NAFLD patients. Gene expression profile demonstrated the difference of molecular signature between the PNPLA3 CC genotype and CG/GG genotype.

Disclosures:

Shuichi Kaneko - Grant/Research Support: MDS, Co., Inc, Chugai Pharma., Co., Inc, Toray Co., Inc, Daiichi Sankyo., Co., Inc, Dainippon Sumitomo, Co., Inc, Ajinomoto Co., Inc, MDS, Co., Inc, Chugai Pharma., Co., Inc, Toray Co., Inc, Daiichi Sankyo., Co., Inc, Dainippon Sumitomo, Co., Inc, Ajinomoto Co., Inc, Bayer Japan

The following people have nothing to disclose: Yuji Hodo, Masao Honda, Hajime Sunagozaka, Tetsuro Shimakami, Kuniaki Arai, Īatsuya Yamashita, Eishiro Mizukoshi, Toshinari Takamura

576

  1. Top of page

PNPLA3 polymorphism vs. other risk factors of fibrosis progression in patients with fatty liver studied longitudinally with repeat transient elastography examinations

Cristina Rigamonti1, Michela E. Burlane1, Miriam Gravellone1, Andrea Magri1, Rosalba Minisini1, Cosimo Colletta2, Mario Pirisi1
1DiMT, Universitá del Piemonte Orientale, Novara, Italy; 2Medicine, Omegna Hospital, Omegna, Italy

Background: rs734809 C/G allelic variants of the adiponutrin gene (patatin-like phospholipase domain-containing protein 3, PNPLA3) modulate the risk of hepatic steatosis and liver fibrosis. Liver stiffness measurement (LSM) by transient elastography (TE) has been shown to be a useful screening tool to exclude significant and advanced fibrosis in patients with fatty liver. The aim of this study was to verify whether genetic polymorphism of PNPLA3 and/or other clinical features could identify patients with fatty liver disease, initially free of significant fibrosis at TE, at low risk of developing it during the follow-up. Materials and Methods: レ4 consecutive patients with a diagnosis of fatty liver disease who underwent a baseline TE examination between 2007 and 2010 at our Centre were genotyped for PNPLA3 rs734809 C/G, and investigated with a second TE examination at least 24 months apart. TE was considered reliable if >10 valid measurements were obtained, with a success rate >65% and an interquartile range <30% of the result. Significant fibrosis was excluded by LSM <5.8 kPa as previously reported (Wong et al, Hepatology 2010). Results: Eleven (4%) patients had unsuccessful TE, 163 (108 men, 55 females) were included, 72% nonalcoholic fatty liver disease, 18% alcoholic liver disease, 19% diabetics, 34% with arterial hypertension, median age 52 yrs (range 18-73), BMI 27.5 kq/mq (range 1941), baseline LSM 6.3 kPa (range 2.5-63.9) and follow-up LSM 5.9 kPa (range 3.3-53.2), and interval between LSMs 41 months (range 24-76). At baseline 66 (41%) patients had LSM <5.8 kPa, 55% of them being PNPLA3 wild type compared to 34% of those with LSM >5.8 kPa (p=0.01).48 (29%) patients had both baseline and follow-up LSM <5.8 kPa, showing significantly lower baseline BMI (p<0.0001), higher baseline serum HDL (p=0.019) and lower triglycerides (p<0.0001) with respect to all other patients. By multivariate logistic regression analysis PNPLA3 wild type (p=0.002, OR 3.1, 95%CI 2.1-11.5-6.4) and baseline triglycerides <120 mg/dl (p<0.0001, OR 4.4, 95%CI 2.1-9.1) independently predicted baseline LSM <5.8 kPa, conversely, BMI <29 kg/mq (p=0.003, OR 5.1, 95%CI 1.7-14.9) and triglycerides <120 mg/dl (p<0.0001, OR 4.6, 95%CI 2.1-10) independently predicted both baseline and follow-up LSM <5.8 kPa. Conclusions: Among patients with fatty liver, wild type PNPLA3 homozygosity and low serum triglycerides are independently associated with normal liver stiffness at presentation, but only the latter and being non-obese independently predict that they will maintain it during the follow-up. These results emphasize the role of modifiable risk factors in the progression of fatty liver.

Disclosures:

Mario Pirisi - Advisory Committees or Review Panels: Merck; Speaking and Teaching: Gilead, Bristol-Myers-Squibb

The following people have nothing to disclose: Cristina Rigamonti, Michela E. Burlone, Miriam Gravellone, Andrea Magri, Rosalba Minisini, Cosimo Colletta

577

  1. Top of page

Natural History of Non-alcoholic Fatty Liver Disease in Adults: A Paired Biopsy Study from the NASH CRN

David E. Kleiner1, Patricia H. Belt2, Laura Wilson2, Elizabeth M. Brunt3, Cynthia D. Guy4, Matthew M. Yeh5, Kris V. Kowdley6, Arun J. Sanyal7, Brent A. Neuschwander-Tetri8
1Laboratory of Pathology National Cancer Institute, Bethesda, MD; 2Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; 3Washington University, St. Louis, MO; 4Duke University, Durham, NC; 5University of Washington, Seattle, WA; 6Virginia Mason Medical Center, Seattle, WA; 7Virginia Commonwealth University, Richmond, VA; 8Saint Louis University, St. Louis, MO

Non-alcoholic fatty liver disease (NAFLD) has a complex natural history; only a fraction of patients progress to cirrhosis and its complications. Using paired biopsies one may detect gradual changes in fibrosis and other features prior to clinical cirrhosis. The aim of this study was to create a model using baseline and longitudinal changes that would show progression of as little as one stage of fibrosis across the full spectrum of NAFLD. Methods: Adult patients enrolled in one of the NASH CRN studies with 2 or more biopsies (excluding active treatment arms of the PIVENS study) at least a year apart were included. Laboratory and anthropometric data was included if available within 6 months of biopsy. All biopsies underwent blinded consensus review. Fibrosis progression/regression was defined as any worsening/improvement in fibrosis stage. Univariate and multivariate logistic regression models were used to assess association with fibrosis progression. Results: 359 patients (mean age 47 years, 64% female) had at least 2 biopsies, with a mean time between biopsies of 4.4 years (range 1 to 17.3).128 showed fibrosis progression and 103 showed regression.181 patients had laboratory data available at baseline. Using any degree of fibrosis progression as the outcome, only ballooning (O. R.0.67), Mallory-Denk bodies (O. R.2.4), and Caucasian race (O. R.3.4) showed significant associations (p<0.05) in multivariate models. We therefore examined the changes in laboratory data and BMI from first to last biopsy (Table) adjusting for baseline values in 169 patients with paired clinical data. While changes in BMI and transaminases were significant in the univariate model, only worsening transaminase levels were associated with fibrosis progression in the multivariate model. Conclusion: The natural history of NAFLD is complex, with both improvement and worsening observed in a mean four year interval. Baseline histologic, demographic, anthropometric and laboratory data were of little value in predicting fibrosis progression, but increased transaminases and BMI from first to last biopsy were associated with fibrosis progression.

UnivariateMultivariate
CharacteristicO. R.95% CIPO. R.95% CIP
A BMI (kg/m2)1.191.03-1.360.021.140.97-1.340.11
AALT(U/L/10)1.181.05-1.330.0041.201.04-1.380.01
A AST (U/L/10)1.351.14-1.59<0.001
A Aik phos (U/L/10)1.170.98-1.400.081.090.89-1.330.43
A Triglycerides (mg/dL/10)1.010.99-1.040.381.010.99-1.040.27
A Glucose (mg/dL/10)1.080.98-1.180.10
A Insulin (μU/mL/10)1.040.91-1.190.54
A HOMA-IR (log)1.020.98-1.060.361.000.96-1.050.86

Disclosures:

Elizabeth M. Brunt - Speaking and Teaching: Geneva Foundation

Kris V. Kowdley - Advisory Committees or Review Panels: Abbott, Gilead, Merck, Novartis, Vertex; Grant/Research Support: Abbott, Beckman, Boeringer Ingelheim, BMS, Gilead Sciences, Ikaria, Janssen, Merck, Mochida, Vertex

Arun J. Sanyal - Advisory Committees or Review Panels: Gore, Gilead, Abbott, Ikaria; Consulting: Salix, Immuron, Exhalenz, Bayer-Onyx, Genentech, Norgine, GalMed, Novartis, Echosens, Takeda; Grant/Research Support: Salix, Genentech, Genfit, Intercept, Ikaria, Takeda, Gilead; Independent Contractor: UpToDate

Brent A. Neuschwander-Tetri - Advisory Committees or Review Panels: Genentech, Nimbus Discovery

The following people have nothing to disclose: David E. Kleiner, Patricia H. Belt, Laura Wilson, Cynthia D. Guy, Matthew M. Yeh

578

  1. Top of page

Liver fibrosis progression and cardiovascular-related complications in type-2 diabetes: establishment of a significant relationship in a 7-years prospective study

Hugo Perazzo1 , 2', Yen Ngo3' Mona Munteanu3, Noemi Seurat1, Fanny Rutka4, Marion Couteau4, Sophie Jacqueminet5, Denis Monneret4, Francoise Imbert-Bismut4, Vlad Ratziu1, Agnes HartemannHeurtier5, Thierry Poynard1
1Hepatology, APHP UPMC Groupe Hopitalier Pitie-Salpetriere, Paris, France; 2Inserm U 938, Paris, France; 3BioPredictive, Paris, France; 4Biochemistry APHP Groupe Hopialier Pitie-Salpetriere,Paris, France; 5DiabetoIogy, APHP Groupe Hopitalier Pitie-Salpetriere, Paris, France

Background: FibroTest(FT), a non-invasive serum marker of liver fibrosis, has a significant prognostic value for the 5-years survival without CRC in T2D patients(1). However, no studies have evaluated the association between liver fibrosis progression and onset of new CRC. Aim: To evaluate the relationship between liver fibrosis progression and cardiovascular-related complications in T2D patients followed during 7 years with repeated evaluation of liver fibrosis by FibroTest. Methods: 627 T2D-patients with minimal fibrosis(FT<0.48 -F0F1 METAVIR) were prospectively foIIowed[2004-2013] for CRC[myocardiaI infarction, unstable angina, coronary-bypass, ischemic stroke]. Liver fibrosis progression was evaluated by repeated FT during follow-up. Progression to advanced fibrosis(AF-F2F3F4) wasdefined by FT≥0.48 at the end of follow-up. Framingham risk score(FRS) was calculated at baseline to predict CRC risk. Results: During the follow-up 46(7%) patients died. Among 581 alive T2D-patients with minimal fibrosis at baseline, 133(23%) had a re-evaluation of liver fibrosis and were included [56% males, age 57 yrs, BMI(range) 28.7(20.2-50.8)Kg/m2]. During a median follow-up of 6.8 yrs 16(12%) patients progressed to AF and 17(13%) patients developed CRC(26 coronary diseases; 1 stroke). The survival without CRC(Kaplan-Meier mean 95%CI) was 69%(41-86) in patients who progressed to AF vs 91%(84-95) in those who did not progress(Logrank p<0.01). Progression to AF increased the risk of cRc[RR=3.8(95%CI 1.3-10.7); p<0.01). In a multivariate Cox model progression to AF remained significant after adjustment on FRS for the prediction of CRC[HR=3.8(1.3-11.1); p=0.013]. Conclusion: In type-2 diabetics, progression from minimal to advanced fibrosis, estimated by FibroTest, was independently associated to higher incidence of cardiovascular-related complications. References: 1 Perazzo H et al. Hepatology 2012; 56(Sup S1): 40A-40A

Thumbnail image of

Disclosures:

Yen Ngo - Employment: BioPredictive Mona Munteanu - Employment: Biopredictive

Vlad Ratziu - Advisory Committees or Review Panels: GalMed, Abbott, Genfit, Enterome, Gilead; Consulting: Astellas, Axcan, Pfizer, Sanofi-Synthelabo, Genentech, Nycomed

Agnes Hartemann-Heurtier - Consulting: Sanofi-Aventis, Pfizer; Grant/Research Support: Lilly

Thierry Poynard - Advisory Committees or Review Panels: MSD; Speaking and Teaching: BMS; Stock Shareholder: BioPredictive

The following people have nothing to disclose: Hugo Perazzo, Noemi Seurat, Fanny Rutka, Marion Couteau, Sophie Jacqueminet, Denis Monneret, Francoise Imbert-Bismut

579

  1. Top of page

Nonalcoholic Fatty Liver Disease Associates with Cardiovascular Disease Outcomes: The Framingham Heart Study

Jessica Mellinger1, Karol M. Pencina2, Joseph M. Massaro3, Udo Hoffmann4, Sudha Seshadri5, 6, Caroline S. Fox6, 7, Christopher J. O'Donnell6,7, Elizabeth K. Speliotes1,8
1Gastroenterology University of Michigan, Ann Arbor, MI; 2Statistics and Consulting Unit, Boston University, Boston, MA; 3Statistics, Boston University School of Public Health, Boston, MA; 4Cardiology and Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA; 5Neurology, Boston University, Boston, MA; 6Framingham Heart Study, National Heart, Lung, and Blood Institute, Framingham, MA; 7Division of Intramural Research, National Heart, Lung, and Blood Institute, Framingham, MA; 8Computational Medicine and Bioinformatics, University of Michigan Hospitals, Ann Arbor, MI

Nonalcoholic fatty liver disease (NAFLD) is highly prevalent and associates with development of metabolic disease including cardiovascular disease. Purpose: to examine the association of NAFLD with prevalent clinical and subclinical cardiovascular disease (CVD) outcomes in a large communitybased sample, the Framingham Heart Study (FHS). Methods: Hepatic steatosis was measured in 3, 529 participants using multidetector computed tomography (CT) scanning. Multivariable logistic regression was used to determine whether hepatic steatosis associates with prevalent CVD adjusted for covariates (age, age2, gender, alcoholic drinks, menopause, and hormone replacement therapy). We also tested whether these effects were independent of other metabolic diseases/traits (diabetes, hypertension, as well as adiposity and lipid traits). Primary outcome was composite prevalent clinical CVD, including nonfatal MI, stroke, TIA, heart failure, and peripheral arterial disease. Secondary outcomes were subclinical CVD including coronary artery calcium (CAC) and abdominal artery calcium (AAC). Results: 3014 participants were included (50.5% women). Hepatic steatosis trended towards being statistically significantly associated with clinical CVD (OR 1.14 [P=0.07])). Hepatic steatosis was associated with both CAC and AAC (OR 1.20 [P=<.001] and OR 1.16[P=<.001], respectively). Associations persisted for CAC even when controlling for other metabolic diseases/traits, but for AAC, the associations became nonsignificant after adjustment for visceral adipose tissue. The effect of hepatic steatosis on AAC was stronger in men than in women (p gender interaction=0.022). Conclusions: There was a significant association of NAFLD with subclinical CVD and a trend towards association with clinical CVD independent of many metabolic diseases/traits. Effects on AAC were stronger in men than in women. This work begins to dissect the links between hepatic fat, metabolic disease risk factors, and CVD.

Effect of NAFLD on CVD Outcomes

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Jessica Mellinger, Karol M. Pencina, Joseph M. Massaro, Udo Hoffmann, Sudha Seshadri, Caroline S. Fox, Christopher J. O'Donnell, Elizabeth K. Speliotes

580

  1. Top of page

Incretin based medicine rapidly improved serum ALT level in non-alcoholic fatty liver disease patients with type 2 diabetes compared to conventional treatments

Takamasa Ohki1, Isogawa Akihiro2, Mori Yamagami1, Tomohoru Yomodo1, Koki Sato1, Yasuhide Yamamoto1, Michharu Seki1, Nobuo Toda1,Kazumi Tagawa1
1Gastroenterology, Mitsui Memorial Hospital, Tokyo, Japan; 2Diabetes and Metabolism, Mitsui Memorial Hospital, Tokyo, Japan

Background: Incretin based medicine, such as GLP-1 analogues or DPP-4 inhibitors, leading to improve not only glycaemic control but also liver inflammation in non-alcoholic fatty liver disease (NAFLD) patients with type 2 diabetes mellitus (DM). Aims: The aim of this study is to elucidate the effectiveness of incretin based medicine in NAFLD patients with type 2 DM compared to conventional treatments such as diet therapy, exercise therapy, and other pharmacological treatments including pioglitazone. Methods: We enrolled 155 Japanese NAFLD patients with type 2 DM and divided these patients into two groups. We compared the base line characteristics and the changes of laboratory data and body weight between the two groups at the end of follow-up. We also assessed the significant factors which contributed to rapid normalization of serum ALT level using multivariate Cox proportional hazard models. Results: There were 102 patients treated with incretin based medicine and 53 patients treated with conventional therapies. At the end of follow-up, serum ALT level, fast blood glucose level, and HbA1c level significantly improved between the two groups. Although the body weight significantly decreased in incretin based medicine group (80.0 kg to 78.1 kg, P < 0.01), the body weight did not change in conventional treatments group (76.4 kg to 77.0 kg, P = 0.68). The cumulative normalization rates of serum ALT level significantly differed between the two groups (P = 0.04); 20.5%, and 35.1% at 250, and 500 days, respectively, in the incretin based medicine group and 15.8%, and 26.7% in the conventional treatments group. Multivariate analysis indicated that administration of incretin based medicine (OR 0.44, P < 0.01), existence of hypertension (OR 1.60, P = 0.048), and comorbidity with dyslipidemia (OR 1.64, P = 0.04) as independent factors which contributed to normalization of serum ALT level. Conclusions: Administration of incretin based medicine led not only good control of type 2 DM but also reduction of body weight, and rapid improvement of liver inflammation.

Disclosures:

The following people have nothing to disclose: Takamasa Ohki, Isogawa Akihiro, Mari Yamagami, Tomoharu Yamada, Koki Sato, Yasuhide Yamamoto, Michiharu Seki, Nobuo Toda, Kazumi Tagawa

581

  1. Top of page

Prevalence and Outcome of Non-alcoholic steatohepatitis in Recipients of Liver Transplantation in a Steroid-free Post-transplant Protocol

Eric C. Fonfenot2, Richard Goldberg1, Jason Vanatta1, Oleksandra Dryn1, Nader Dbouk1, Satheesh Nair1, James Eason1, Sanjaya K. Satapathy1
1Methodist Transplant Institute, University of Tennessee Health Science Center, Memphis, TN; 2Internal Medicine, University of Tennessee Health Science Center, Memphis, TN

BACKGROUND: Recurrence of non-alcoholic steatohepatitis (NASH) in up to 33% of the liver transplant (LT) recipients has been reported (Bhagat et al 2009). It is not known whether steroid free immunosuppression reduces the risk of NASH after liver transplantation. AIM: We aimed to determine the prevalence of NASH post- transplant in a cohort of patients with NASH cirrhosis and alcoholic cirrhosis (ETOH) who received a steroid free immunosuppression regimen based on one year protocol liver biopsy, and determine risk factors for recurrence and outcomes for these patients. METHODS: We performed a retrospective review for all patients who underwent LT for NASH or ETOH who had protocol liver biopsy at one year follow up at our center from April 2006 to April 2012. Comparison was made between ETOH and NASH groups as well as those who developed steatohepatitis (SH group) and those who did not develop steatohepatitis (non-SH group). RESULTS: The study included 40 recipients (M/F: 20/20) in the NASH group and 47(M/F: 40/7) in the ETOH group. Recurrence/development of NASH in recipients of LT with cirrhosis secondary to NASH is significantly higher compared to recipients secondary to ETOH [16 of 40(40%) vs.8 of 47(17%); P= 0.017]. Multivariate analysis failed to identify any single predictive variable for predicting recurrence/development of steatohepatitis in both NASH and ETOH groups. Atherosclerotic cardiovascular events, defined as acute myocardial infraction, cerebrovascular accident, need for percutaneous coronary intervention and coronary artery bypass graft was noted in six recipients, 5(12.5%) from the NASH group, and 1(2.1%) from the ETOH group (P=0.09). Presence of steatohepatitis did not predict cardiovascular events (R=-0.68, P=0.54, OR 0.50) or death (R=0.73, P=0.37, OR 0.48) in the entire cohort of LT recipients (median follow up of 30 mos.). However, significantly higher stage of fibrosis by METAVIR scoring system was noted in the subjects with SH group compared to non-SH group (0.96±0.90 vs.0.56±0.81, P=0.02). Three and five year survival in patients surviving beyond 1 year was comparable with and without steatohepatitis by Kaplan-Meier analysis (92%, 92% in the SH group vs.89%, 57% in the non-SH group at 3 yrs., and 5yrs. respectively; P= 0.50, Log Rank). None of the recipients died of cardiovascular events. CONCLUSION: Recurrence/development of NASH post LT was still common amongst NASH patients despite the use of steroid free immunosuppression. NASH recurrence was not associated with an increased risk of cardiovascular events or worse long term survival on short-term follow up.

Disclosures:

Satheesh Nair - Speaking and Teaching: Vertex, Genetech

The following people have nothing to disclose: Eric C. Fontenot, Richard Goldberg, Jason Vanatta, Oleksandra Dryn, Nader Dbouk, James Eason, Sanjaya K. Satapathy

582

  1. Top of page

Low LDL-C and high HDL-C levels are associated with elevated serum transaminases amongst adults in the United States

Zhenghui G. Jiang1,Kenneth Mukamal3, Elliot B. Tapper2, Simon C. Robson2, Yusuke Tsugawa3
1Medicine, Beh Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; 2Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; 3General Medicine and Primary Care, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA

BACKGROUND: Dyslipidemia, typically recognized as high serum triglyceride, high low-density lipoprotein cholesterol (LDLC) or low high-density lipoprotein cholesterol (HDL-C) levels, are associated with nonalcoholic fatty liver disease (NAFLD). However, low LDL-C levels could result from defects in lipoprotein metabolism or impaired liver synthetic function, and may serve as ab initio markers for unrecognized liver diseases. Whether low LDL-C levels indicate liver diseases in the general population has not been investigated to date. METHODS: We examined the associations between alanine aminotransferase (ALT), aspartate aminotransferase (AST) and major components of serum lipid profiles in a nationally representative sample of 1.5 individuals from the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2010. RESULTS: We found that ALT and AST exhibited non-linear U-shaped associations with LDL-C and HDL-C, but not with triglyceride. After adjusting for potential confounders, individuals with LDL-C less than 40 and 41-70 mg/dL were associated with 4.2 (95% CI 1.5 11.7, p=0.007) and 1.6 (95% CI 1.1-2.5, p=0.03) times higher odds of abnormal liver enzymes respectively, when compared with those with those with LDL-C values 71-100 mg/dL. Similarly, those with HDL-C levels above 100 mg/dL was associated with 3.2 (95% CI 2.1-5.0, p<0.001) times higher odds of abnormal liver enzymes, compared with HDL-C values of 61-80 mg/dL. CONCLUSION: Both low LDL-C and high HDLC, often viewed as desirable, were associated with significantly higher odds of elevated transaminases in the general U. S. adult population. Our findings raise concerns about unrecognized hepatic dysfunction among people with particularly low LDL-C or high HDL-C.

Disclosures:

Simon C. Robson - Grant/Research Support: Pfizer, NIH; Independent Contractor: eBioscience, Biolegend, EMD Millipore, Mersana; Speaking and Teaching: ACP, Elsevier, ATC; Stock Shareholder: Nanopharma, Puretech

The following people have nothing to disclose: Zhenghui G. Jiang, Kenneth Mukamal, Elliot B. Tapper, Yusuke Tsugawa

583

  1. Top of page

Frequent alcohol consumption is associated with the remission of fatty liver in Japanese men: a longitudinal study

Akio Moriya 1, Yosniaki Iwasaki2, Souhei Ohguchi3, Eizo Kayashima3, Tadahiko Mitsumune3, Hideaki Taniguchi4, Fusao Ikeda5, Masaharu Ando1
1Department of Medicine' Mioyo Geneial Hospital' Kanonji, Japan; 2HeaIh Service Center' Okayama University' Okayama' Japan; 3Junpukai Health Cen-ter' Okayama' Japan; 4Department of Medicine' Tottori Municipal Hospital' Tottori Japan; 5Depaiimern of Gastroenterology and Hepatology, Okayama University, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences' Okayama' Japan

Background: A number of cross-sectional studies have demonstrated an inverse association between light to moderate alcohol consumption and presence of fatty liver; we have also reported an inverse correlation between drinking frequency and the prevalence of fatty liver in men. However, the influence of alcohol consumption on the development or remission of fatty liver is still controversial. Methods: We obtained clinical and laboratory data from 10,054 Japanese subjects who voluntarily underwent a baseline health checkup and once or more of follow-up studies from 2006 to 2011. The development or remission of fatty liver was assessed by ultrasonography. The time of fatty liver development or remission was assumed to be the midpoint between the checkup at which the change of fatty liver status was observed for the first time and that of the one before it. Using Cox proportional hazard model, we performed multivariable analyses to evaluate the influence of alcohol consumption on the development or remission of fatty liver with following factors: overweight or obesity (BMI > 25 kg/m2), dyslipidemia, hypertension, glucose intolerance, hyperuricemia, smoking, exercise, and age. Results: After excluding cases with concurrent liver disease and/or missing component of data, we analyzed 8,879 cases (median age, 47 years old). The total follow-up period was 17,999.8 person-years. At baseline, 2,309 of 5,488 men (42.1%) and 461 of 3,391 women (13.6%) had fatty liver. BMI (mean ± SD) in cases with and without fatty liver was 25.9 ± 3.1 kg/m2 and 22.4 ± 2.4 kg/m2 in men, and 25.9 ± 3.4 kg/m2 and 20.9 ± 2.5 kg/m2 in women. In men, the amount of alcohol consumed (mean ± SD) in each drinking frequency category (drinking 1-3 days/week, drinking 4-6 days/week, and daily drinking) was 62 ± 60 g/week (n =1,347),189 ± 112 g/week (n = 976), and 272 ± 132 g/week (n =1,764), respectively. During the follow-up, 491 cases of development and 418 cases of remission of fatty liver were observed. The remission of fatty liver was directly associated with drinking on 4-6 days/week (hazard ratio [HR], 1.50; 95% confidence intervals [CI], 1.12-2.01) and daily drinking (HR, 1.38; 95% CI, 1.05-1.81) after adjustment for other confounders. The association between alcohol consumption and the development of fatty liver was not significant. In women, 245 cases of development and 99 cases of remission of fatty liver were observed. The change of fatty liver status was not associated with alcohol consumption. Conclusions: In the longitudinal study, the protective effect of frequent alcohol consumption against fatty liver appeared to be mainly therapeutic rather than preventive in men.

Disclosures:

The following people have nothing to disclose: Akio Moriya, Yoshiaki Iwasaki, Souhei Ohguchi, Eizo Kayashima, Tadahiko Mitsumune, Hideaki Taniguchi, Fusao Ikeda, Masaharu Ando

584

  1. Top of page

Magnetic Resonance Detection of Hepatic Cholesteryl Ester Accumulation for Non-invasive Identification and Treatment Monitoring in Lysosomal Acid Lipase Deficiency

Peter E. Thelwall2'3' Fiona E. Smith2'3' Mark Leavitt,'Wei Hu1, Kieren G. Hollingsworth23' Christian Thoma2, Michael I. Trenell2' Roy Taylor23' Joseph V. Rutkowski1,Andrew M. Blamire2 , 3 ,Anthony G. Quinn1
1Synageva BioPharma Corp' Lexington' MA; 2Institute of Cellular Medicine' Newcastle University' Newcastle upon Tyne, United Kingdom; 3Newcastle Magnetic Resonance Centre, Newcastle University, Newcastle upon Tyne, United Kingdom

Purpose: The rare metabolic storage disorder Lysosomal Acid Lipase (LAL) Deficiency is caused by a marked reduction in the activity of LAL, and leads to accumulation of cholesteryl esters (CE) and triglycerides (TG) in lysosomes in many tissues, particularly the liver. Monitoring of liver lipid accumulation typically involves analysis of biopsy samples, carrying an associated risk to the patient. Magnetic resonance (MR) techniques allow non-invasive, safe and repeatable measurement of hepatic lipid content and composition. We investigated the potential of MR spectroscopy for monitoring in LAL deficient patients and in ex vivo LAL deficient rat liver tissue. We assessed the effects of enzyme replacement therapy with sebelipase alfa (a recombinant human LAL) on hepatic lipid content and composition in the preclinical model using MR spectroscopy. Methods: Two patient cohorts comprising LAL-deficient (n=3) and NAFLD (n=5) were studied. Preclinical studies comprised ex vivo liver samples from wild type, NAFLD, LALdeficient, and LAL-deficient rats receiving 4 weeks of sebelipase alfa treatment. Hepatic 1H MR spectroscopy was performed using 3T (human) and 7Ī (preclinical) MRI scanners to quantify hepatic cholesterol and triglyceride content. Magnitude of signal originating from CH3 and CH2 resonances in lipid species was determined from MR data, and a spectral model fitted to the data to determine concentrations and ratios of cholesterol and fatty acid chain moieties. Results: Hepatic cholesterol ester accumulation was identified in both human and preclinical studies. LAL deficient patients had ratios of cholesterol: fatty acid moiety of 0.39 ± 0.13, compared to <0.01 in the NALFD group. In preclinical studies a good correlation was observed between biochemical and MR assay of hepatic cholesterol content (R2 = 0.86), and marked reduction of cholesterol content was observed in LAL deficient animals treated with sebelipase alfa. Conclusions: We demonstrate an entirely non-invasive method to identify and quantify the hepatic lipid signature. The approach provides a more favorable alternative to repeated biopsy sampling for diagnosis and disease progression of patients with LAL deficiency and other disorders characterised by increased free cholesterol and/or cholesteryl esters.

1H MR Spectra from LAL Deficient & NAFLD Patients

Disclosures:

Mark Leavitt - Employment: Synageva Corp; Stock Shareholder: Synageva Corp

Thumbnail image of

Wei Hu - Employment: Synageva BioPharma Corp.

Joseph V. Rutkowski - Employment: Synageva BioPharma

Andrew M. Blamire - Grant/Research Support: Synageva

Anthony G. Quinn - Employment: Synageva BioPharma; Management Position: Synageav BioPharma; Stock Shareholder: Synageva BioPharma

The following people have nothing to disclose: Peter E. Thelwall, Fiona E. Smith, Kieren G. Hollingsworth, Christian Thoma, Michael I. Trenell, Roy Taylor

585

  1. Top of page

Role of plasma FGF21 as a biomarker for nonalcoholic fatty liver disease (NAFLD) and steatohepatitis (NASH)

Fernando Bril1, Romina Lomonaco1, Beverly Orsak2, Zhi Chang2, Carolina Ortiz-Lopez2, Joan Hecht2, 3, Fermin Tio4, 3, Kenneth Cusi1, 2
1Endocrinology, Diabetes and Metabolism, University of Florida, Gainesville, FL; 2Diabetes, University of Texas Health Science Center of San Antonio, San Antonio, TX; 3Diabetes, Audie L. Murphy Veterans Administration Medical Center, San Antonio, TX; 4Pathology, University of Texas Health Science Center at San Antonio, San Antonio, TX

Background and Aim: Fibroblast growth factor 21(FGF2 1) is believed to play a key role in the regulation of glucose and fat metabolism, promoting glucose uptake in adipose tissue and fat oxidation in the liver. Several studies have shown that elevated plasma FGF21 levels are found in subjects with disorders related to obesity and insulin resistance. We aimed to assess the role of FGF21 as a potential biomarker for the diagnosis of NAFLD and/or NASH. Methods: We recruited 204 patients with and 24 without NAFLD (51 ±1 vs.50±3 years [p=0.69], male: 72% vs.58% [p=0.24], 34.1 ±0.3 vs.29.3±1.0 kg/m2 [p<0.01]) and measured: 1)plasma FGF21 and cytokeratin-18 fragments (CK-18) levels, 2) liver fat by magnetic resonance imaging and spectroscopy (MRS), 3) hepatic insulin resistance index (HIRi=fasting plasma insulin x endogenous glucose production) and adipose tissue insulin resistance index (ATIR= fasting plasma insulin x free fatty acids), 4) muscle insulin sensitivity (Rd) during a high-dose insulin euglycemic clamp, and 5) liver histology (biopsy) (n=159). Results: Plasma levels of FGF21 were significantly increased in patients with NAFLD (337 [217-526] vs.153 [92-323] pg/ml, p<0.001). However, FGF21 only showed moderate correlations with liver fat (r=0.26, p<0.001), HIRi (r=0.26, p=0.002), ATIR (r=0.23, p<0.001) and Rd (r=-0.35, p<0.0001). As a stand-alone test for the diagnosis of NAFLD, FGF21 had rather disappointing results. With the optimal cut-off point of 205 pg/ml we obtained a sensitivity of 78% (71-84%) and specificity of 63% (41一81%). Positive and negative predictive values were 94% (89-97%) and 28% (17-42%), respectively. Patients with NASH had higher levels of FGF21 when compared to patients with simple steatosis on liver biopsy (386 [252-581] vs.328 [170—451] pg/ml, p=0.03). However, correlations between FGF21 and NAFLD activity score (r=0.22, p<0.01) and fibrosis stage (r=0.30, p<0.001) were weak. As a tool for the diagnosis of NASH (optimal cut-off point: 376 pg/ml), FGF21 also showed unsatisfactory results, with low sensitivity (53% [43 62%]) and specificity (67% [50 一 81%]). With the combined use of CK-18 and FGF21 for the diagnosis of NASH, sensitivity improved slightly to 57% (49-66%) and specificity to 85% (70-94%). However, these results were similar to the ones of CK-18 alone (sensitivity 46% [38-53%] and specificity 86% [73-95%]). Conclusions: Plasma FGF21 levels were only moderately correlated with different measures of insulin resistance, hepatic steatosis and liver histology. Based on these findings, FGF21 is not a useful stand-alone test (or combined with CK-18) for the diagnosis of NAFLD or NASH.

Disclosures:

Beverly Orsak - Employment: UTHSCSA

Kenneth Cusi - Consulting: Merck, Daichi-Sankyo, Roche, Janssen; Grant/Research Support: Takeda, Novartis, Mannkind

The following people have nothing to disclose: Fernando Bril, Romina Lomonaco, Zhi Chang, Carolina Ortiz-Lopez, Joan Hecht, Fermin Tio

586

  1. Top of page

Impact of a 4 week oral high fructose challenge (FC) on intrahepatic lipid composition and ATP homeostasis in healthy volunteers assessed by high-field 3.0-Tesla 1H and 3lP-MR spectroscopy (MRS) and its potential link to Lipoprotein A (LPA)

Christian Kienbacher1, Martin Gajdosik3, Michael Krebs2, Stefan Traussnigg1, Werner Dolak1, Petra E. Steindl-Munda1, Siegfried Trattnig3, Martin Krššák2, 3, Michael Tauner1
1Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria; 2Division of Endocrinology and Metabolism, Department of Internal Medicine III Medical University of Vienna, Vienna, Austria; 3MR Centre-ofExcellence, Department of Radiodiagnostics, Medical University of Vienna, Vienna, Austria

Background and Aims: Fructose overconsumption in Western diet is linked to non-alcoholic fatty liver disease (NAFLD) which is associated with increased risk for cardiovascular disease (CVD). LPA levels above 30mg/dL are considered an important risk factor for CVD. Conversely, LPA levels inversely correlate with the incidence of diabetes and HOMA-IR (Cardiovasc Diabetol, 2012). We therefore aimed to assess the impact of an oral 4 week 150g/day FC on hepatic lipid metabolism reflected by total hepatic lipid content (HLC), fatty acid saturation and phosphorous metabolites (PM; indicating hepatic energy metabolism), and glucose homeostasis. Moreover, we aimed to explore the role of LPA as potential biomarker predicting hepatic sensibility to fructose induced (lipo)toxicity. Methods: Ten healthy volunteers were enrolled in a pilot study (m: f=5: 5; median age 24.5 (21-37)). HLC (CH2 fraction of total signal), unsaturation- (UI), saturation- (SI=1-UI) and polyunsaturation indices (PUI) were determined by single voxel 3.0-T 1H-, PM by 31P-MRS. BMI was assessed clinically. Blood was collected at days 1, 14 and 28 for routine laboratory analysis, LPA at baseline and fasted glucose. Results: Mean BMI was 21.11 ± 2.68 (SD) kg/m^2 and increased after FC (21.51 ± 2.77; p<0.001). UIs increased from mean 0.175 ± 0.087 to 0.226 ± 0.066 (p=0.034 one-tailed; large effect size r=0.568). Similarly, fasting glucose levels increased (mean 83.2mg/dL ± 8.37 to 88.4 ± 5.48, p=0.035). Notably, total HLC, PUIs, PM, ALT, AST and yGt remained unchanged (p>0.05). Analysis of responders to FC (FCR; n=6) as reflected by increased total HLC (as potential indicator for lipid partitioning of potentially toxic lipid intermediates as neutral TG) versus non-responders (FCNR; n=4) revealed αATP depletion in FCNR (mean 3.003 ± 2.07 to 1.67 ± 0.413; p=0.034 onetailed). In FCR reflected by increased UI (n=7) γATP depletion was observed (mean 2.18 ± 0.71 to 1.51 ± 0.4; p=0.036). Notably, baseline LPA levels inversely correlated with total HLC following FC (r=-0.801; p=0.004; r^2=0.652; non-linear fit: r^2=0.594) and delta-αATP (r=-0.652; p=0.039). Conclusions: MRS is feasible to detect slight changes in intrahepatic lipid composition after FC in healthy young volunteers. Changes in fatty acid saturation indices may occur earlier compared to other well established markers of liver damage. Moreover, serum LPA may also serve as a novel biomarker to differentiate between individuals at increased risk for NAFLD, particularly under fructose challenge. Approaches aimed at lowering LPA to counteract CVD risk should also consider potential interactions with hepatic lipid content and partitioning.

Disclosures:

Michael Trauner - Advisory Committees or Review Panels: MSD, Janssen; Consulting: Falk Pharma, Phenex, Amgen; Grant/Research Support: Intercept; Patent Held/Filed: Med Uni Graz (norUDCA); Speaking and Teaching: Falk Foundation, Roche, Gilead

The following people have nothing to disclose: Christian Kienbacher, Martin Gajdosik, Michael Krebs, Stefan Traussnigg, Werner Dolak, Petra E. Steindl-Munda, Siegfried Trattnig, Martin Krššák

587

  1. Top of page

Effect of resistance training on non-alcoholic fatty liver disease

Atsushi Takahashi, Hiromichi Imaizumi, Manabu Hayashi, Ken Okai, Yukiko Kanno, Kazumichi Abe, Hiromasa Ohira
Gastroenterology and Rheumatology, Fukushima Medical University School of Medicine, Fukushima, Japan

Background and aim: Aerobic exercise is recommended for non-alcoholic fatty liver disease (NAFLD), but evidence supporting the effect of resistance training (RT) in NAFLD is currently lacking. We evaluated the effect of RT on the metabolic parameters of patients with NAFLD. Methods: RT was performed three times per week on non-consecutive days for 12 weeks. The RT program consisted of only two exercises: pushups and squats. One set of 10 push-ups and 10 squats was performed a total of three times. Including a one-minute interval between push-ups and squats, the program required 20 to 30 minutes to complete. Biochemical blood parameters, hepatic steatosis and body composition were assessed before and after RT. Hepatic steatosis and body composition were evaluated by ultrasound and bioelectrical impedance analysis, respectively. Patients self-recorded their RT compliance. Results: The study included 27 patients with NAFLD (mean age 55.2±13.6 y; mean body mass index (BMI), 28.4±3.3, 69% female). The rate of compliance with the RT program was 66.9%. Compliance was not significantly associated with the RT program or with features of the patients such as age, sex and BMI. Adverse events did not develop in any of the patients. Body composition determined as BMI, fat content and lean body mass, did not significantly change although fat tended to decrease, whereas lean body mass tended to increase after RT. However, mean levels of ALT (78.3±62.8 vs.56.7±50.4 IU/L, p<0.0001), insulin (13.3±8.4 vs.11.1±4.4 μU/mL, p=0.031), ferritin (199.8±179.1 vs.160.5±121.5 ng/mL, p=0.031) and homeostasis model assessment of insulin resistance (HOMA-IR) value (3.9±2.9 vs.3.1 ±1.5, p=0.026) significantly decreased after RT. Moreover, hepatic steatosis grade (1.85±0.77 vs.1.57±0.60 grade, p=0.003) was also significantly decreased after RT. Conclusion: RT was an effective treatment for NAFLD. Push-ups and squats are simple and safe types of RT and thus, our RT program might serve as a complement to treatment for patients who have difficulties with aerobic exercise.

Disclosures:

The following people have nothing to disclose: Atsushi Takahashi, Hiromichi Imaizumi, Manabu Hayashi, Ken Okai, Yukiko Kanno, Kazumichi Abe, Hiromasa Ohira

588

  1. Top of page

Genome-wide analysis identifies loci associated with total bilirubin levels, steatosis, and mild fibrosis in nonalcoholic fatty liver disease

Johanna DiStefano1, Christopher Kingsley1, Christopher D. Still3, Stefania Cotta Doné1, Glenn Gerhard2,3, David E. Kleiner4
1Translational Genomics Research Institute, Phoenix, AZ; 2PennsyIvania State University College of Medicine, Hershey, PA; 3Geisinger Health System, Danville, PA; 4National Institutes of Health, Bethesda, MD

Genetic factors mediate susceptibility to non-alcoholic fatty liver disease (NAFLD) and related traits, including non-alcoholic steatohepatitis (NASH). Although several markers have been associated with hepatic fat and inflammation in NAFLD through genome-wide association studies (GWAS), there has been limited investigation of the genetic determinants of clinically severe forms of NASH, including fibrosis. Therefore, the goals of this study were to identify common genetic variants associated with various phenotypes related to normal liver function, NAFLD, and NASH. We genotyped approximately 2300 individuals with extreme obesity enrolled in a bariatric surgery program using the Illumina Human omniExpress (OmniExpress) BeadChip assay. Following QC and identity by descent analysis to exclude related individuals, approximately 2084 Caucasian patients were available for the GWAS of NAFLD, NASH and related traits. We first performed linear regression of total bilirubin in 2046 patients, adjusting for multiple testing using the Bonferroni method. We observed the strongest evidence for association with rs4148325 in the UDP glucuronosyltransferase 1 family, polypeptide A complex locus (UGT1a) gene (P<1.0 x 10-111), which confirms previously reported findings. We next performed linear regression for iron binding capacity and identified rs3811658 (P<1.0 x 10-35) in the transferrin (TF) gene, also confirming previous findings. We then used linear regression for steatosis grades 0, 1,2 and 3 and observed evidence for association with markers in the neurocan gene (NCAN) on chromosome 19p12 (P<1.0 x 10-7). Using logistic regression of fibrosis stage 1a (n=337) vs. non-fibrotic ^=1747) patients and adjusting for multiple testing using the Bonferroni method, we also observed association with rs2501843, located on chr 1(P<1.0 x 10-7). Logistic regression analysis of bridging fibrosis (stage 3) using 97 cases and 1986 non-fibrosis controls identified association (P<1.0 x 10-7) with a cluster of SNPs on chromosome 6. Our results replicate several loci for liver-related phenotypes and present evidence for new genetic loci that may play a role in the pathophysiology of NAFLD and NASH.

Disclosures:

The following people have nothing to disclose: Johanna DiStefano, Christopher Kingsley, Christopher D. Still, Stefania Cotta Done, Glenn Gerhard, David E. Kleiner

589

  1. Top of page

Serum levels of PNPLA3 are related to the development of steatohepatitis in non alcoholic fatty liver disease patients

Maria Teresa Arias-Loste1, Paula Iruzubieta1, Angela Puente1, Susana LLerena1, Marcos López-Hoyos2, Maria Teresa GarcίaUnzueta3, Rocίo Gallego-Durán7, Isidora Ranchal4, Javier Abad4, Jose Luis Calleja-5, Carmelo Garcίa-Monzón6, Jose Luis Olcoz7, Manuel Romero-Gomez1, Javier Crespo1
1Digestive, Morques de Valdecilla University hospital, Santander, Spain; 2Immunology, Marques de Valdecilla University hospital, Santander, Spain; 3Clinical Analyzes, Marques de Valdecilla University hospital, Santander, Spain; 4Digestive, Puerto de Hierro hospital, Madrid, Spain; 5Digestive, Santa Cristina University hospital, Madrid, Spain; 6Digestive, Complejo Asisfenciol Universitario de Leon' Leon, Spain; 7Digestive, Volme University hospital, Sevilla, Spain

Background and aim Non-alcoholic fatty liver disease (NAFLD) is a growing clinical condition whose increase over past decades mirrors the outbreak of obesity-related disorders. Recently, a European genome-wide association study identified genetic variants in the PNPLA3 gene associated with NAFLD. Nevertheless, the role of the encoded protein, PNPLA3 or adiponutrin, in the development of the disease is not completely understood, and the usefulness of serum levels of PNPLA3 as biomarkers in NAFLD has not been analyzed yet. Therefore, we aimed to assess the basal levels of PNPLA3 in NAFLD patients and healthy controls, and to correlate these levels with the severity of the disease according to liver histology. Patients and methods We performed a multicenter cohort study that included 146 patients (55 men; mean ± SD: age 47 ± 11, BMI 35.96 ± 1 0.59) with biopsy-proven NAFLD diagnosis (78.2% simple steatosis, 21.8% NASH) and 10 healthy controls (6 men; mean ± SD: age 36±10, BMI 22.7±2.4). PNPLA3 levels were determined in fasting serum of patients and controls using a commercialized ELISA kit (Uscn, Life science Inc., Wuhan, China). NAFLD patients were classified according to Brunt's classification. Additionally, resistin' adiponectin and leptin levels were measured using commercialized ELISA kits. Results NAFLD patients displayed a statistically significant higher serum levels of PNPLA3 than healthy controls, whose levels were all below the limit of detection (mean [95% CI]; NAFLD patients 5.00 ng/ml [3.65-6.35]; p=0.007). PNPLA3 levels correlated to BMI (r=0.382; p<0.005), leptin levels (r=0.681; p<0.0005), and inversely to resistin (r=-0.278; p<0.05) and AST levels (r=0.168; p<0.05). Patients with biopsy-proven NASH showed lower serum level of PNPLA3 in comparison with simple steatosis (mean [95% CI]; 4.38 ng/ml [2.47-6.29] in NASH versus 9.20 ng/ml [4.15-14.23] in simple steatosis; p=0.006). A serum level > 10.7 ng/ml showed 32% sensitivity and 82% specificity to predict a simple steatosis according to ROC curve analysis (AUROC 0.68 [95% CI: 0.55-0.81]; p=0.01). Conclusions Serum levels of PNPLA3 correlated with steatosis degree but not with steatohepatitis. As previously reported about the variant I148M, adiponutrin seems to play a critical role in fat deposition but not in steatohepatitis progression. Further studies are warranted to demonstrate if previous association with fibrosis and NASH in NAFLD are pathogenic or consequence of the impact of confounding factor linked to the degree of fat infiltration.

Disclosures:

Manuel Romero-Gomez - Advisory Committees or Review Panels: Roche Farma, SA., MSD, S. A., Janssen, S. A., Abbott, S. A.; Grant/Research Support: Ferrer, S. A.

Javier Crespo - Board Membership: MSD, Roche, Janssen, Gilead

The following people have nothing to disclose: Maria Teresa Arias-Loste, Paula Iruzubieta, Angela Puente, Susana LLerena, Marcos López-Hoyos, Maria Teresa Garcίa-Unzueta, Rocίo Gallego-Durán, Isidora Ranchal, Javier Abad, Jose Luis Calleja, Carmelo Garcla-Monzon, Jose Luis Olcoz

590

  1. Top of page

Differing Rates of Suspected Non-alcoholic Fatty Liver Disease in a Diverse Group of Hispanic/Latino Persons; Population Based Data from the Hispanic Community Health Study/Study of Latinos

Eric R. Kallwitz1, Martha L. Daviglus2, Matthew Allison4, Jinsong Chen2, Kristen T. Emory4, Natalia A. Gouskova5, Robert C. Kaplan6, Amber Pirzada3, Gregory A. Talavera7, Marston E. Youngblood5, Lihui Zhao3, Scott Cotler1
1Loyolo University Medical Center, Maywood, IL; 2University of Illinois, Chicago, IL; 3Northwestern, Chicago, IL; 4University of California, San Diego, CA; 5University of North Carolina, North Carolina, NC; 6Einstein College of Medicine, New York, NY; 7San Diego State University, San Diego, CA

Non-alcoholic Fatty Liver Disease (NAFLD) disproportionally affects Hispanics compared to other racial/ethnic groups; however, prior studies have focused primarily on those of Mexican heritage. This study aimed to evaluate prevalence of suspected NAFLD among the diverse Hispanic/Latino participants of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). Methods: Participants were 16, 415 adult men and women. Suspected NAFLD was defined as either AST >37 IU/ml or ALT >40 IU/ml for men, and either AST or ALT >31 IU/ml for women in absence of another known cause of liver disease. Those with missing variables of interest, positive HBV/HCV serology, excessive alcohol consumption, or transferrin saturation >50% were excluded. Information on components of metabolic syndrome, acculturation, health care use, sleep quality, diet, physical activity, education and income was obtained. Results: 11, 753 participants were included. The Table shows prevalence of suspected NAFLD. It was most common with Mexican and Central American background and in men (23.1% vs women 15.6%, p<0.001). Suspected NAFLD was positively associated with age <40, and each component of metabolic syndrome. No associations with acculturation, health care use, sleep disturbance, physical activity or income were observed. In multivariable analysis adjusting for factors listed above, suspected NAFLD was observed less in persons of Cuban (OR 0.72, 95%CI 0.58, 0.90), Puerto Rican (OR 0.75, 95%CI 0.57, 1.00) and Dominican (OR 0.75, 95%CI 0.57, 1.00) background, and in females (OR 0.48, 95%CI 0.41, 0.57) and observed more with age <40, elevated waist circumference (OR 1.83, 95%CI 1.54, 2.18), elevated triglyceride (OR 1.58, 95% CI 1.32, 1.88), low HDL (OR 1.26, 95%CI 1.08, 1.47), elevated blood pressure (OR 1.25, 95%CI 1.01, 1.55) and impaired fasting glucose (OR 1.42, 95%CI 1.17, 1.73). Conclusions: Suspected NAFLD is most common in individuals of Mexican, Central and South American background. The lack of association of suspected NAFLD with cultural and behavioral measures suggest that genetic differences might contribute to differences in suspected NAFLD among diverse Hispanics/ Latinos.

AllMexican n=4891Cuban n=2327Puerto Rican n=1827Dominican n=1153Central American n=913South American n=642
All(n= 11753)
19.022.116.715.915.021.017.7
Pairwise PvalueRef<0.001<0.001<0.0010.560.03
Male(n=5377)
23.126.721.516.821.724.222.1
Pairwise PvalueRef0.03<0.0010.140.400.17
Female(n=6377)
15.618.511.315.110.918.214.2
Pairwise PvalueRef<0.0010.14<0.0010.890.05

Disclosures:

Scott Cotler - Speaking and Teaching: Genentech, Vertex, Brystal Myers, Gilead

The following people have nothing to disclose: Eric R. Kallwitz, Martha L. Daviglus, Matthew Allison, Jinsong Chen, Kristen T. Emory, Natalia A. Gouskova, Robert C. Kaplan, Amber Pirzada, Gregory A. Talavera, Marston E. Youngblood, Lihui Zhao

591

  1. Top of page

Variants in the Il6 and Ill p Genes either Alone or in Combination with C282Y HFE Mutations Modify the Hepatic Iron Phenotype of Patients with Nonalcoholic Fatty Liver Disease

James E. Nelson1,2David E. Kleiner3, Bradley E. Aouizerat4, Kris V. Kowdley1, 2
1Benaroya Research Institute, Seattle, WA; 2Liver Center of Excellence, Digestive Disease Institute, Virginia Mason Medical Center, Seattle, WA; 3Laboratory of Pathology, National Cancer Institute, Bethesda, MD; 4Nursing, UCSF, San Francisco, CA

Background: The iron regulatory hormone hepcidin is regulated by both iron and inflammatory signals including IL6 and IL1 p cytokines. Aim: The goal of this study was to investigate if IL6 and IL1p cytokine SNPs, alone or in combination with HFE gene mutations, can affect the grade and pattern of hepatic iron deposition and serum iron markers in the well characterized NASH CRN cohort. Methods: Serum hepcidin levels were determined by ELISA immunoassay (Intrinsic LifeSciences). Genotyping for the two common HFE mutations C282Y (rs1800562) and H63D (rs1799945) and the following SNPs in the IL6 and IL1 p genes was performed by RT-PCR in 787 adult (>18 yrs) subjects with biopsy proven NAFLD and hepatic iron staining results: IL6; +4272C>T (rs2069849), −174G>C (rs1800795), −6331T>C (rs10499563); IL1p; −31T>C (rs1143627), −511C>T (rs16944), +3953C>T (rs1143634). Chi2 and ordinal regression adjusting for sex was used to determine the association of each genotype with nominal and ordinal variables. Continuous variables were analyzed using regression analysis adjusting for sex. The effects of HFE mutations and the IL6 and IL1β SNPs were investigated using regression analysis with interaction terms. Results: Subjects with the IL1β −31 CT, IL1β −511 CT and IL1β +3953 CC genotype had significantly increased hepatocellular (HC) iron stain grade (p<0.04). Subjects with the IL1 p −511 TT and IL1 p +3953 CT genotype had decreased HC iron stain grade (p<0.047) Subjects with the IL1 p −31 CT genotype had significantly increased serum hepcidin levels (p=0.032), whereas the IL1 p −31 CC genotype was associated with decreased serum hepcidin levels (p=0.035). There were significant interactive effects of the C282Y mutation in subjects with the IL6 −6331 CT and TT genotypes. C282Y+ subjects with the IL6 −6331 CT genotype had significantly increased serum hepcidin levels (p=0.014), increased HC iron stain grade (p=0.020) and increased % transferrin saturation levels (p=0.048). In contrast, C282Y+ subjects with the IL6-6331 TT genotype had significantly lower serum hepcidin levels (p=0.034) and decreased HC iron stain grade (p=0.020). Conclusions: The IL1 p −31T>C, −511C>T and +3953 polymorphisms impact iron regulation in NAFLD subjects. The IL6 −6331T>C loci modifies the effect of HFE C282Y mutations upon iron regulation in NAFLD subjects. In all genotypes there was a positive association between serum hepcidin levels and markers of iron, including iron stain grade suggesting serum hepcidin levels in patients with NAFLD reflect the physiologic response to body iron stores.

Disclosures:

Kris V. Kowdley - Advisory Committees or Review Panels: Abbott, Gilead, Merck, Novartis, Vertex; Grant/Research Support: Abbott, Beckman, Boeringer Ingelheim, BMS, Gilead Sciences, Ikaria, Janssen, Merck, Mochida, Vertex

The following people have nothing to disclose: James E. Nelson, David E. Kleiner, Bradley E. Aouizerat

592

  1. Top of page

The Urine Metabolomics Uncovers Key Biomarkers of Nonalcoholic Fatty Liver Disease

Shu Dong1 , 2' Yanqin Bian1 , 2, Ping Liu1 , 2, Mingyu Sun1 , 2
1Shuguang Hospital, Shanghai University of Traditional Chinese Medicine, Key Laboratory of Liver and Kidney Diseases (Ministry of Education), Institute of Liver Diseases, Shanghai, China; 2Shuguang Hospital, Shanghai University of Traditional Chinese Medicine, Institute of Liver Diseases, shanghai, China

The urine with non-alcoholic fatty liver disease (NAFLD), including steatosis and steatohepatitis (NASH), was examined using metabolomics analysis in order to identify potential non-invasive biomarkers. Blood (separated serum for liver function or serum lipids assay) and urine sample were obtained after an overnight fast from confirmed non-diabetic subjects with NAFLD (n=84), and compared with healthy, age and sex-matched controls (n=30). The metabolic profile changes were analyzed by GC/MS with principal component analysis (PCA), partial least squares-discriminate analysis (PLS-DA) and orthogonal partial least squares-discriminate analysis (OPLS-DA). Furthermore, biochemical examinations were also carried out to compare healthy controls with NAFLD patients. Compared with the healthy controls, patients with NAFLD have abnormal liver function and high level serum lipids. Through urinary metabonomics, 31 metabolites are found between these two groups including Hypoxanthine, 6-Azathymine, Inosine, 2, 5-Furandicarboxylic acid, D-Pinitol, Galactonic acid, etc. These metabolites can be classified into nucleic acid and amino acid. Conclusion: Statistical analysis identified a panel of biomarkers that could effectively separate healthy controls from NAFLD. These biomarkers can potentially be used to follow response to clinical diagnosis and therapeutic interventions.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Shu Dong, Yanqin Bian, Ping Liu, Mingyu Sun

593

  1. Top of page

In patients with nonalcoholic fatty liver disease, altered hepatic expression of genes related to Toll-like receptor signaling is associated with intestinal microbiota

Bianca M. Arendt1, Elena M. Comelli2, Marialena Mouzaki3, Ian McGilvroy1,4,Scoff Fung1,4, Sandra E. Fischer1,4, Johane P Allord1,4
1Toronto General Hospital, University Health Network, Toronto, ON, Canada; 2Department of Nutritional Sciences, Universify of Toronto, Toronto, ON, Canada; 3Department of Pediatrics, Hospital for Sick Children, Toronto, ON, Canada; 4Department of Medicine, University of Toronto, Toronto, ON, Canada

Background: Activation of toll-like receptors (TLR) leads to inflammatory cytokine production which contributes to insulin resistance in nonalcoholic fatty liver disease (NAFLD). Since patients with NAFLD have altered intestinal microbiota (IM), there is the potential for altered signaling of the chief TLR4 ligand, gram-negative endotoxin and other IM derived TLR ligands. Aims: 1. To compare hepatic expression of genes involved in TLR signaling between patients with simple steatosis (SS) or nonalcoholic steatohepatitis (NASH) and healthy controls (HC); 2. To determine whether these genes correlate with IM. Methods: In a cross-sectional study, gene expression (Illumina Whole-Genome DASL HT-assay) was measured in liver tissue obtained from biopsies of 20 patients with SS and 19 with NASH or during live donor hepatectomy (HC: n=24). In a subgroup (6 SS, 9 NASH, 8 HC), the relative amounts of fecal Bifidobacteria, E. coli, Bacteroides/Prevotella, and Firmicutes were assessed (qPCR). ANOVA with Tukey's HSD test, Wilcoxon test, and Spearman correlations were applied. Results: Body mass index (BMI) and insulin resistance increased significantly from HC to SS and NASH. Bacteroides/Prevotella were lower in NASH compared to HC [median (IQR)] [0.90 (1.66) vs.3.45 (3.84) % of total bacteria; p=0.012) but not different from SS [3.03 (4.97) %]. Seven genes involved in TLR signaling were differentially expressed (>2.0-fold change, adjusted p<0.05) between NAFLD patients and HC. There were no differences between SS and NASH. Bacteroides/Prevotella (%) were correlated with TLR3, TLR7, and PELI1. When controlling for BMI, this remained significant: TLR3 (r=-0.85, p<0.0001), TLR7 (r=-0.599, p=0.0041), PELI1 (r=0.59, p=0.021). Conclusion: TLR signaling related genes are differentially expressed in NAFLD vs. HC, but expression profiles are similar in SS and NASH. The proportion of Gram-negative Bacteroides/Prevotella in stool was higher in NASH vs. HC and was correlated with the expression of important regulators of TLR signaling. The role of IM in TLR-induced inflammation and insulin resistance in NAFLD warrants further study.

Disclosures:

Scott Fung - Advisory Committees or Review Panels: Merck, Vertex; Grant/Research Support: Gilead Sciences, Roche; Speaking and Teaching: Gilead Sciences, BMS

The following Comelli, Marialena Mouzaki, Ian McGilvray, Sandra E. Fischer, Johane P. Allard

594

  1. Top of page

Modest Alcohol Consumption and Risk of Cardiovascular Diseases among Men with Non-Alcoholic Fatty Liver Disease

Dong Hyun Sinn1,2Geum Youn Gwak1, Yong Han Paik1, Moon Seok Choi1, Joon Hyeok Lee1, Kwang Cheol Koh1, Seung Woon Paik1, Byung Chui Yoo1
1Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea; 2Sanggye Paik Hospital, Inje University School of Medicine, Seoul, Republic of Korea

Objective To evaluate cardiovascular benefits of modest alcohol consumption among men with nonalcoholic fatty liver disease (NAFLD). Design Cross-sectional study of 10, 581 consecutive male participants aged 40-69 years undergoing abdominal ultrasonography and carotid artery ultrasonography at the Center for Health Promotion of the Samsung Medical Center in Korea from January 2009 to December 2009. Results There were total 2, 129 men diagnosed with NAFLD, and the mean age of the participants was 52.6 years old. Among them, 1, 641 were modest alcohol drinkers, and NAFLD fibrosis score was not different between the modest drinkers and nondrinkers. Modest alcohol intake had the inverse association with carotid plaque [odd ratio (OR): 0.74, 95% confidence interval (CI): 0.59 - 0.91] and carotid artery stenosis (CAS) (OR: 0.58, 95% CI: 0.40 - 0.84) after adjusting age, smoking and metabolic syndrome. In addition, inverse association between modest alcohol consumption and carotid plaque and CAS was well observed in men with a low NAFLD fibrosis score, but not with men with an intermediate or a high NAFLD fibrosis score. Conclusion Modest alcohol consumption has a favorable association with carotid plaque or CAS, especially in individuals with a low NAFLD fibrosis score.

Disclosures:

The following people have nothing to disclose: Dong Hyun Sinn, Geum Youn Gwak, Yong Han Paik, Moon Seok Choi, Joon Hyeok Lee, Kwang Cheol Koh, Seung Woon Paik, Byung Chul Yoo

595

  1. Top of page

The significance of serum adipokines in non-alcoholic fatty liver disease

Akharawit Pulsombat1, Daruneewan Warodomwicnitl, Pattana Sornmayura2, Napat Angkathunyakul2, Sasivimol Rattanasiri3, Wathanee Chaiyaratana4, Piyaporn Kaewdoung1, Supanna Petraksa1, Abhasnee Sobhonslidsuk1
1Medicine, Ramathibodi hospital, Bangkok, Thailand; 2Pathology, Ramathibodi hospital, Bangkok, Thailand; 3Section for Clinical Epidemiology and Biostatistic, Ramathibodi hospital, Bangkok, Thailand; 4ClinicaI Research center, Ramathibodi hospital, Bangkok, Thailand

Background: Non-alcoholic fatty liver disease (NAFLD) is a hepatic manifestation of metabolic syndrome that has insulin resistance as an underlying cause. Non-alcoholic steatohepatitis (NASH), an advanced stage of NAFLD, can progress to cirrhosis. Chemerin, vaspin and omentin-1 are new serum adipokines released from visceral adipose tissue. Recent studies suggest that adipokines may play an important role in the pathogenesis of NAFLD. Objectives: We aimed to assess for chemerin, vaspin and omentin-1 levels in patients with NAFLD, and to identify predictive markers for NASH and advanced fibrosis. Methods: A cross-sectional study was performed. Patients with liver biopsy-confirmed NASH within 2 years prior to the study were enrolled together with age- and sex-matched controls. Study subjects underwent anthropometric measurement and laboratory tests for biochemistry, index of insulin resistance (HOMA-IR) and adipokine (: chemerin, vaspin and omentin-1) levels. Adipokine levels were assessed by enzymelinked immunosorbent assay. NAFLD activity score (NAS) and fibrosis staging were graded according to Kleiner et al. Liver histology with NAS >5 was defined as NASH and NAS 3-4 for borderline steatohepatitis. Fibrosis stages >3 were defined as advanced fibrosis. Statistical analysis was done with student ttest, non-parametric or chi-square tests as appropriate. A pvalue <0.05 was taken as statistical significance. Logistic regression analysis was performed for factors with p-value <0.20. Results: Sixty NAFLD patients and 55 controls were enrolled. Mean (SD) ages of NAFLD and control subjects were 54.7 (8.7) and 46.9 (8.1) years. Data of anthropometric measurement, liver chemistry, lipid profiles and HOMA-IR of NAFLD patients were significantly different from control group. Chemerin and omentin-1 levels in NAFLD patients were higher than control group [194.58 vs 120.51 (p-value <0.001) and 452.25 vs 372.1 ng/ml (p-value <0.006)]. Twenty-two (36.7%) and 38 (63.3%) NAFLD patients had liver pathology consistent with NASH and simple steatosis. The levels of chemerin, vaspin and omentin-1 in NASH and simple steatosis groups were not significantly different. Twenty-one (35%) NAFLD patients had advanced fibrosis. From univariate analysis, age, NAS, hemoglobin A1C, omentin-1 levels were significantly elevated in advanced fibrosis patients. Only age and NAS were independently associated with advanced fibrosis. Conclusions: Chemerin and omentin-1 levels are elevated in NAFLD patients. Age and NAFLD activity score, but not chemerin, omentin-1 and vaspin levels, are associated with NAFLD with advanced fibrosis. The role of adipokines in NAFLD requires further study

Disclosures:

The following people have nothing to disclose: Akharawit Pulsombat, Daruneewan Warodomwichit, Pattana Sornmayura, Napat Angkathunyakul, Sasivimol Rattanasiri, Wathanee Chaiyaratana, Piyaporn Kaewdoung, Supanna Petraksa, Abhasnee Sobhonslidsuk

596

  1. Top of page

Small intestinal bacterial overgrowth and microbial translocation products in peripheral blood samples and its relationship with nonalcoholic fatty liver disease, in morbid obesity patients

Francisco Domper1, Aurora Gil-Rendo2, Soledad Illescas3, Alicia Hernandez-Albujar1 ,Maria Alonso-Lablanca1,Carmen Moho Cabrera4, Alberto Jara2, Cristina Murillo5, Francisco MartinDavila5, Maria Adan1, Roberto Paton1, Jesus Martin-Fernandez2, Bartolome Lopez-Viedmo1;
1Gastroenterology and Hepatology Hospital General Universitario de Ciudad Real, Ciudad Real, Spain; 2General Surgery, Hospital General Universitario de Ciudod Real, Ciudad Real, Spain; 3MicrobioIogy, Hospital General Universitario de Ciudad Real, Ciudad Real, Spain; 4BiochemicaI, Hospital General Universitario de Ciudad Real' Ciudad Real, Spain; 5PathoIogy Hospital General Universitario de Ciudad Real, Ciudad Real, Spain

Background: Nonalcoholic fatty liver disease (NAFLD), the hepatic manifestation of metabolic syndrome, is the most common cause of liver disease in our environment. The prevalence of NAFLD in patients with morbid obesity reaches 80%. It has been proposed that the small intestinal bacterial overgrowth (SIBO) and microbial translocation through the intestinal wall (MT) are related to NAFLD, in its initial form of simple steatosis and in a more advanced stage of steatohepatitis (NASH). The aim of this study was to investigate in patients with morbid obesity and NAFLD the relationship between SIBO and MT measured by serum levels of lipopolysaccharide (LPS) and LPS binding protein (LBP), with NAFLD activity score (NAS) and the severity of steatosis. Patients and Methods: We included consecutive morbid obese patients (BMI >40 kg/m2 or >35 kg/m2 in association with comorbidities) prior to bariatric surgery intervention. Exclusion criteria were: normal liver biopsy, other causes of liver disease or duodenal mucosal atrophy. Endoscopy was performed to obtain duodenal aspirate for culture and duodenal mucosa biopsy. We studied peripheral venous blood sampling for liver enzymes, viral hepatitis, LPS and LBP. The following values were considered pathological: 10000 CFU/mL (SIBO), LPS >5 EU/mL and LBP >10 mg/mL. Liver biopsy was performed during surgery. The severity of steatosis was quantified in three grades (1: 5-33% 2: 34-66% 3: >66%) and NAS >4 was associated with increased likelihood of having NASH. Results: 71 patients were initially included, but 26 were excluded because they had normal liver biopsy. Forty-five patients had NAFLD therefore. Eighteen men, mean age 45.88 years (22-69) and mean BMI 47.81 kg/m2 (37-58).25% had SIBO measured with a sensitive and specific method as the duodenal aspirate culture. Degree of steatosis: 20/17/8. NAS >4 in almost half of patients. Statistical significance was observed between LBP levels and SIBO with the severity of steatosis (p <0.05 and p =0.077, respectively). There was no statistical relationship with NAS, although the distribution of the values had a trend towards an association. LPS levels did not keep any relationship with the severity of steatosis or NAS. Conclusions: In patients with morbid obesity there are relationship between MT phenomena measured by peripheral blood levels of LBP and SIBO, with the degree of hepatic steatosis. The relationship of these with NAS probably reach statistical significance in studies with larger numbers of patients.

Disclosures:

The following people have nothing to disclose: Francisco Domper, Aurora GilRendo, Soledad Illescas, Alicia Hernandez-Albujar, Maria Alonso-Lablanca, Carmen Maria Cabrera, Alberto Jara, Cristina Murillo, Francisco Martin-Davila, Maria Adan, Roberto Paton, Jesus Martin-Fernandez, Bartolome Lopez-Viedma

597

  1. Top of page

Disruption of autophagic flux in hepatocytes is linked to increased endoplasmic reticulum stress and apoptosis during the development of nonalcoholic fatty liver disease

Agueda Gonzalez-Rodriguez1, Rafael Mayoral2, Noelia Agra2, Virginia Pardo Marques1, Oreste Lo Iacono3, Lisardo Bosca2, Carmelo García-Monzón4, Paloma Martin-Sanz2, Angelo M. Valverde1
1Instituto de Investigaciones Biomedicos “Alberto Sols” (CSIC/UAM)/ CIBERDEM, Madrid, Spain; 2Instituto de Investigaciones Biomedicos “Aberfo Sols” (CSIC/UAM)' CIBEREHD, Madrid, Spain; 3Gastroenterology Service, Hospital del Tajo, Aronjuez, Spain; 4Liver Research Unit, University Hospital Santa Cristina, Instituto de Investigación Sanitaria Princesa' CIBEREHD' Madrid, Spain

Endoplasmic reticulum (ER) stress and impaired autophagy have been implicated in the pathogenesis of nonalcoholic fatty liver disease (NAFLD), but the molecular mechanisms involved are not fully defined. The aim of the present study was to assess the relationship between ER stress and the autophagic flux in NAFLD patients as well as in murine models of NAFLD and human hepatocytes. This study comprised 49 patients with biopsy-proven nonalcoholic steatosis (NAS) or nonalcoholic steatohepatitis (NASH) and 34 subjects with histologically normal liver (NL). Experiments of real-time PCR, Western blot, immunofluorescence and electronic microscopy were carried out to assess hepatic expression of markers of ER stress and autophagy. In addition, mice fed with high fat diet (HFD) for 30 weeks or methionine-choline-deficient (MCD) diet during 4 weeks were used to evaluate ER stress and autophagy within the liver. Human Huh7 hepatocytes loaded with palmitic acid (PA) were also studied as an in vitro model. In patients with NAS and NASH, hepatic mRNA levels of ER stress markers were elevated together with increased p62 autophagic substrate and LC3-II accumulation within liver cells. However, immunofluorescence analysis revealed hepatocellular LC3-II punctuate was less evident in patients with NASH. On the other hand, livers from mice fed with HFD or MCD diet showed increased phosphorylation of mTOR/S6K1, JNK, PERK, eIF2 along with elevated expression of ER chaperones GRP78 and CHOP. As observed in NAFLD patients, p62 and LC3-II were up-regulated in murine liver cells independently of diet interventions. Interestingly, less LC3-II punctuate and more apoptotic hepatocytes were observed in mice fed with MCD diet which displayed NASH features. In addition, in human Huh7 hepatocytes, incubation with PA for 8 hours activated the autophagic flux as assessed by decreased p62 and increased LC3-II/LC3-I ratio, and apoptosis was not observed. Noteworthy, when exposure to PA was prolonged for 24 hours, ER stress markers and apoptosis were significantly increased along with a marked accumulation of p62 and autophagosomes, as observed by electronic microscopy, reflecting the loss of autophagic flux. Of note, combined treatment with rapamycin and PA for 24 hours decreased p62 and the amount of autophagosomes, indicating the activation of the autophagic flux, and protected Huh7 cells against apoptosis. In conclusion, autophagic flux is impaired in hepatocytes from both NAFLD patients and murine models of NAFLD. Our findings also indicate that, in lipid-overloaded human hepatocytes, disruption of autophagic flux is associated with increased ER stress and apoptosis.

Disclosures:

The following people have nothing to disclose: Agueda Gonzalez-Rodriguez, Rafael Mayoral, Noelia Agra, Virginia Pardo Marques, Oreste Lo Iacono, Lisardo Bosca, Carmelo Garcla-Monzon, Paloma Martin-Sanz, Angela M. Valverde

598

  1. Top of page

Clinicopathological features of HCC derived from steatohepatitis

Tomomi Kogiso, Etsuko Hashimoto, Kazuhisa Kodama, Maki Tobari, Noriko Matsushita, Nobuyuki Torii, Makiko Taniai, Katsutoshi Tokushige, Keiko Shiratori
Gastroenterology, Tokyo Women's Medicol Univ. ' Tokyo, Jopon

Aim: Steatohepatitis is one of the important etiologies of hepatocellular carcinoma (HCC) and the incidence is increasing. Here, we analyzed the clinicopathological features of HCC derived from steatohepatitis related with non-alcohol and alcohoi consumption to characterize HCC risks in these groups. Methods: 1334 patients were diagnosed with histologicallyproven steatohepatitis at our University between 1988 and 2012. Among those, 243 patients developed HCC. We classified into two groups according to the amount of alcohol consumption (less than 20g/day (non-alcoholic fatty liver disease; NAFLD) and over 70g/day (alcohol liver disease; ALD)). The expression status of 4-hydroxy-2/-noneal (HNE) and 8-hydroxydeoxyguanosine (8-OHdG), the markers of the oxidative DNA damage and Nanog a pluripotent gene involving in the regulating self-renewal of cancer-stem cells, were determined by immunohistochemistry. This study was performed after informed consent. Results: 1)Clinical background of patients: In 241 HCC patients, 62 were NAFLD and 179 were ALD. The median age was 70/65 years old, BMI was 26/ 23 kg/m2, and the proportion of male patients was 68/96% in NAFLD/ALD. Regarding the complication, diabetes was 74/ 54%, hypertension was 66/36%, and dyslipidemia was 47/20%.32/ 8% of patients had all three lifestyle-related diseases.2) Characteristics of HCC: The median diameter of NAFLD-HCC/ALD-HCC was 30/28mm.75/50% showed solitary nodule and 17/30% revealed more than three nodules (p<0.01). Most of the HCC showed trabecular, moderately differentiated and simple nodular or simple nodular type with extranodular growth in both groups.3) Characteristics of noncancerous tissues: The patients with advanced fibrosis were 48/68% in NAFLD/ALD. More than moderate steatosis was observed in 87/22%, and iron deposition was observed in 13/56% (p<0.01).4) Immunohistochemical analysis: The HNE and 8-OHdG were stained in HCC tissues more than in noncancerous tissues in both groups.25/18% of cases strongly expressed Nanog in both HCC and non-cancerous tissue.5) Treatment and prognosis: 46/63% of patients were surgically resected.47/59% of patients died of liver-related deaths. Five year survival rate was 60/50%. Conclusions: HCCs in NAFLD cases were diagnosed in older and more obese patients with lifestyle-related diseases than in ALD cases. HCCs in ALD were observed mostly in male, with more severe fibrosis and iron deposition in the liver tissues. Positive staining of HNE, 80HdG, and Nanog, may be useful markers of HCC risks. These characteristics should be taken into consideration for the early detection of HCCs during the care of the steatohepatitis patients.

Disclosures:

The following people have nothing to disclose: Tomomi Kogiso, Etsuko Hashimoto, Kazuhisa Kodama, Maki Tobari, Noriko Matsushita, Nobuyuki Torii, Makiko Taniai, Katsutoshi Tokushige, Keiko Shiratori

599

  1. Top of page

MERTK Variant is Associated with the Severity of Liver Steatosis and of Liver fibrosis in Patients with Non-alcoholic Fatty Liver Disease

Salvatore Petta1, Stefania Grimaudo1, Calogero Comma1, Vito Di Marco1, Daniela Cabibi2, Roberto Fonte1, Emanuele Orlando1, Rosario Maria Pipitone1, Antonio Croxi1
1Sezione di Gastroenterologia, DiBiMIS, University of Palermo, Italy, Palermo, Italy; 2Cattedra di Anatomia Patologica, University of Palermo, Italy, Palermo, Itoly

Background and Aims: Nonalcoholic fatty liver disease (NAFLD), and its progressive variant 一 nonalcoholic steatohepatitis (NASH)- had a complex pathogenesis with a relevant role of both classical - obesity and insulin resistance - and new risk factors - gene polymorphisms and apoptosis-. A recent genomewide study on patients with chronic hepatitis C demonstrated that variants in MERTK, TUPL1 and RNF7, genes involved in apoptosis and phagocitosis control, were associated with liver fibrosis progression in this clinical setting. We aimed to assess whether rs4374383 MERTK, rs9380516 TULP1 and rs16851720 RNF7 single nucleotide polymorphisms (SNP) influence the expression of steatosis, lobular inflammation and fibrosis in NAFLD patients. Methods: Two hundred sixteen consecutive NAFLD patients, were assessed by liver biopsy (Kleiner score) and anthropometric, biochemical and metabolic features. rs4374383 MERTK, rs9380516 TULP1 and rs16851720 RNF7 SNP were tested. Results: Most patients were males (65%), mean age and BMI were respectively 45 years and 29.9 Kg/m2, and HOMA values were quite high (mean value 4.18). One patients on 3 had grade 3 steatosis and one patient on two had F0-F2 fibrosis. The prevalence of MERTK GG, GA and AA genotype was 40.7%, 44.4% and 14.9% respectively; of TULP-1 CC, CT and TT genotype was 66.6%, 29.6%, and 3.8%, respectively; and of RNF7 AA, AC and CC genotype was 65.7%, 30.7% and 3.7% respectively. Patients carrying the MERTK AA genotype had a significant lower prevalence of grade 3 steatosis (5/32 vs 67/184, p=0.02) and of F0-F2 fibrosis (9/32 vs 92/184, p=0.02) compared to MERTK GG /GA patients. Accordingly, by multivariate logistic regression analysis BMI (OR 1.068, 95% CI 1.142, p=0.05), ALT (OR 1.007, 95% CI 1.012, p=0.008),and MERTK AA (OR 0.288, 95% CI 0.099-0.842, p=0.02) were independently linked to severe steatosis, as well as age (OR 1.027, 95% CI 1.053, p=0.03), HOMA (OR 1.160, 95%CI 1.018-1.322, p=0.02), MERTK AA (OR 0.327, 95% CI 0.128-0.839, p=0.02) and lobular inflammation(OR 3.163, 95%CI 1.867-5.357, p< 0.001) were independently associated with significant fibrosis. No association was found between TULP1 or RNF7 genotypes and severity of liver damage. Conclusions: In patients with NAFLD, MERTK AA homozygosis is protective against severity of steatosis and of fibrosis.

Disclosures:

The following people have nothing to disclose: Salvatore Petta, Stefania Grimaudo, Calogero Camma, Vito Di Marco, Daniela Cabibi, Roberto Fonte, Emanuele Orlando, Rosaria Maria Pipitone, Antonio Craxi

600

  1. Top of page

Mean platelet volumes in non-alcoholic liver disease (NAFLD): is there a relationship to cardiovascular events?

Lisa Alvarez1, Daisha Cipher2, Rick A. Weideman2, Geri Brown1, 2
1Division of Digestive ond Liver Diseases' University of Texas Southwestern Medical Center, Dallas, TX; 2VA North Texas Health Core System, Dallas, TX

Background: Increased morbidity and mortality from CV disease is now recognized in patients with NAFLD. Mean platelet volume (MPV) is a biomarker of platelet activity and elevations in MPV have been seen in the setting of acute CV events. The aim of this study is to assess whether increases in MPV during acute CV events is observed in patients with NAFLD. Methods: A retrospective case control study of 104 patients with NAFLD who had cardiovascular events (CE) and 104 NAFLD patients (matched by age, gender and BMI) without cardiovascular events (non-CE) was performed. Demographics, CV risk factors, laboratory data, and medications were collected. Included CV events were myocardial infarction/unstable angina (n=59), coronary artery bypass graft (n=36), stroke/transient ischemic attack/peripheral vascular disease (n = 12) and congestive heart failure (n=9). MPV's were collected at the initial time of the study (To),at the time of the CV event (TCE)(or equivalent time period in the non-CE group) and at the end of follow up (Tf). To assess liver severity, the AST platelet ratio index (APRI) defined by (AST/upper limit normal AST/platelets X 100) was calculated. The Framingham risk score (FRS) was calculated to assess CV risk prior to the events. Statistical analysis was performed using student's T test, Pearson's chi squared and Mann Whitney U tests. Results: Demographics included a mean age of 56 ± 8 years (yr) with a majority of white ethnicity (CE, n= 85 versus non-CE, n=76), an average BMI of 31.8 ± 5.1 and an average time to follow up 9.5 ± 2.7 yr. No difference in liver severity as assessed by APRI was noted between the groups (0.34 ± 0.19 in CE and 0.36 ± 0.17 in non-CE group, p= 0.47). The CE group had higher CV risk calculated by the FRS (24 ± 11.6%) compared to the non-CE group (18 ±12%) (p=0.0002). More patients in the CE group were exposed to aspirin and clopidogrel, p<0.0001. Average time from Tq to Tce was 4.9 ± 2.8 yr. Importantly, the absolute changes of MPV from T。to Tce and from T。to Tf [(MPV Tce -MPV T。) and (MPV Tf- MPV T。)] were statistically higher in the CE group than in the non-CE group (table 1). Conclusion: The absolute change in the MPV level at the time of the CV event was higher in the CE group when compared to the non-CE group (0.54± 1.1 versus 0.21 ± 0.9, p=0.023). In addition, the increase in the MPV at study end was also higher in the CE group when compared to the non-CE group.

Table 1

Change in MPV (T- T0)Change in MPV (TF-T0)
CE group (Std dev)0.54(1.1)0.63(1.2)
Non-CE group (Std dev)0.21 (0.9)0.26 (1.2)
P value0.0230.02

Disclosures:

The following people have nothing to disclose: Lisa Alvarez, Daisha Cipher, Rick A. Weideman, Geri Brown

601

  1. Top of page

Normal Weight Nonalcoholic Steatohepatitis (NASH)

Niharika Samala1, Kevin p. May2, David E. Kleiner3, Rohit Loomba4, Norah Terroult5, Brent A. Neuschwander-Tetri6, Joy H. Hoofnagle7
1Liver Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD; 2Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; '^Loborofory of Pathology, National Cancer Institute/National Institutes of Health, Bethesda, MD; 4Department of Medicine, University of California San Diego, San Diego, CA; 5Department of Medicine, University of California San Francisco, San Francisco, CA; 6Department of Internal Medicine, Saint Louis University, St. Louis, MO; 7Liver Disease Research Branch, National Institutes of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD

Background: Nonalcoholic fatty liver disease (NAFLD) typically occurs in persons who are overweight or obese and have metabolic risk factors such as diabetes, hypertension and hyperlipidemia. Yet a proportion of patients do not have these risk factors. Using the NASH CRN database, patients with NAFLD despite normal weight were identified and further investigated for their specific clinical features, metabolic risk factors, response to therapy and PNPLA3 genotype. Methods: The NASH CRN is an NIH-funded, multicenter network whose aim is to help elucidate the pathogenesis, natural history and therapy of NAFLD. For this study, 1259 adult subjects enrolled in NASH CRN database who had undergone liver biopsy and had accompanying laboratory results were selected. Patients were categorized as normal weight (BMI <25), overweight (BMI 25-<30) or obese (BMI >30) and compared in regards to clinical, laboratory and histological features of NAFLD. Results: Of the 1259 subjects, 50 (4%) were normal weight (BMI range 18.73 to 24.96), 286 (23%) overweight and 923 (73%) obese. Normal weight patients with NAFLD were more likely to be Asian (24% vs 13% and 2%: p<0.001) than the overweight and obese cohorts but were similar in regards to age and sex. Patients who were normal in weight also had significantly lower mean fasting blood glucose and insulin levels than the overweight and obese patients. Importantly, normal weight patients tended to have similar elevation of mean ALT, AST and GGT levels but had less severe steatosis, ballooning degeneration and fibrosis on liver biopsy. Definite NASH as judged histologically was less common among normal weight (38%) than among overweight (44%) or obese (58%) subjects. Distributions of PNPLA3 genotypes (Rs738409 G vs C) were similar among the 3 weight groups. Frequency of overall response to therapy with vitamin E, pioglitazone and placebo was higher among NASH patients with normal (overall 57%) than those with excessive weight (34% and 29%) but the numbers were too small in the individual treatment groups to achieve statistical significance. Conclusions: Adults with NAFLD who are normal weight are more likely to be Asian and to have on average milder disease with less steatosis, ballooning and fibrosis. Responses to therapy may be greater in patients with normal weight suggesting that early intervention may be appropriate. Similar distribution of PNPLA3 genotypes in the three weight groups suggests the likelihood of other genetic factors that might contribute to development of NASH.

Disclosures:

Rohit Loomba - Consulting: Gilead Inc, Corgenix Inc; Grant/Research Support: Daiichi Sankyo Inc, AGA

Norah Terrault - Advisory Committees or Review Panels: Eisai, Biotest; Consulting: BMS; Grant/Research Support: Eisai, Biotest, Vertex, Gilead, AbbVie, Novartis

Brent A. Neuschwander-Tetri - Advisory Committees or Review Panels: Genentech, Nimbus Discovery

The following people have nothing to disclose: Niharika Samala, Kevin P. May, David E. Kleiner, Jay H. Hoofnagle

602

  1. Top of page

Progression to bridging fibrosis in non-alcoholic fatty liver disease over 4 years in the NASH CRN

Elizabeth M. Brunt1, Patricia H. Belt2, Laura Wilson2, Cynthia D. Guy3, Matthew M. Yeh4, Kris V. Kowdley5, Arun J. Sanyal6, Brent A. Neuschwander-Tetri7' David E. Kleiner8
1Pathology and Immunology, Washington University School of Medicine, St Louis, MO; 2Epidemiology, Johns Hopkins, Baltimore, MD; 3Pathology, Duke University, Durham, NC; 4Pathology, University of Washington, Seattle, WA; 5Digestive Diseases Institute, Virginia Mason Medical Center, Seattle, WA; 6Internal Medicine, Virginia Commonwealth, Richmond, VA; 7Internal Medicine, Saint Louis University, St Louis, MO; 8Pathology Laboratories, National Cancer Institute, Bethesda, MD

The natural histories of non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH) are poorly understood. For optimal patient management, it is important to understand which patients with early stage disease are at greatest risk for progression to advanced stage. We used paired biopsies to study patients who progressed to bridging fibrosis (BF) or cirrhosis with patients who remained at early stages. Methods: Adult patients enrolled in one of the NASH CRN studies with 2 or more biopsies (excluding active treatment arms of the PIVENS study) at least a year apart and in which the first biopsy had a fibrosis stage less than 3 were included. Laboratory and anthropometric data were included if available within 6 months of biopsy. All biopsies underwent blinded consensus review. The endpoint was progression to BF or cirrhosis from first to last biopsy. Chi-square, ANOVA, Kruskal-Wallis, and CochraneArmitage tests were used to assess difference between progressers and non-progressers at baseline. Multivariate logistic regression models were used to assess association with fibrosis progression. Results: 270 patients (mean age 46 years, 62% female) had at least 2 biopsies, with a mean time between first and last biopsies of 4.4 years (range 1 to 17.3).43 (16%) showed progression to BF or cirrhosis.149 patients had laboratory data available at baseline. Patients who progressed were older, had higher ALT, AST and glucose, and were more often diabetic or had metabolic syndrome at baseline (all p<0.02). Initial biopsies of progressors had more ballooning, portal inflammation, Mallory Denk bodies, higher NAFLD Activity Scores, and more often showed steatohepatitis (all p≤0.02). The table shows the results of separate multivariate logistic regression models for the histological and clinical/demographic factors. Only features with p<0.05 are shown. Conclusion: Progression of NAFLD and NASH from early to late fibrosis stage is associated mainly with histological features of NASH, as well as age, higher transaminase levels and the presence of diabetes and metabolic syndrome at baseline. These data suggest that clinical models can be developed to identify patients with early stages of fibrosis at risk for progression to advanced fibrosis.

Baseline FindingsOR95% CIP
Histological Model
Portal Inflammation2.141.01-4.530.047
Acidophil Bodies2.301.03-5.160.04
Mallory Denk Bodies4.911.68-14.370.004
Clinical Model
Metabolic Syndrome6.460.98-42.530.05
ALT (log U/L)5.241.78-15.400.003

Disclosures:

Elizabeth M. Brunt - Speaking and Teaching: Geneva Foundation

Kris V. Kowdley - Advisory Committees or Review Panels: Abbott, Gilead, Merck, Novartis, Vertex; Grant/Research Support: Abbott, Beckman, Boeringer Ingelheim, BMS, Gilead Sciences, Ikaria, Janssen, Merck, Mochida, Vertex

Arun J. Sanyal - Advisory Committees or Review Panels: Gore, Gilead, Abbott, Ikaria; Consulting: Salix, Immuron, Exhalenz, Bayer-Onyx, Genentech, Norgine, GalMed, Novartis, Echosens, Takeda; Grant/Research Support: Salix, Genentech, Genfit, Intercept, Ikaria, Takeda, Gilead; Independent Contractor: UpToDate

Brent A. Neuschwander-Tetri - Advisory Committees or Review Panels: Genentech, Nimbus Discovery

The following people have nothing to disclose: Patricia H. Belt, Laura Wilson, Cynthia D. Guy, Matthew M. Yeh, David E. Kleiner

603

  1. Top of page

Individuals with the PNPLA3 (adiponutrin) prosteatotic variant present with unexpected body fat composition: a study in one thousand healthy subjects

Marcin Krawczyk1, Agnieszka Kempinska2, Marta Klak2, Malgorzoto Blatkiewicz2, Piotr Milkiewicz3,4, Frank Lammert1, Malgorzata Milkiewicz2
1Department of Medicine II, Saarland University Medical Center, Homburg/Saar, Germany; 2Medical Biology Laboratory, Pomeranian Medical University, Szczecin, Poland; 3Liver Research Laboratories, Pomeranian Medical University, Szczecin, Poland; 4Department of General, Transplant and Liver Surgery of the Medical University of Warsaw, Warsaw, Poland

Introduction: The common PNPLA3 (adiponutrin) variant p.1 148M represents a major genetic driver of progression in fatty liver disease (NAFLD).1 Fatty liver, which in patients with non-alcoholic steatohepatitis (NASH) may progress to liver cirrhosis, is commonly associated with traits of the metabolic syndrome.2 Hence NAFLD is mostly suspected in obese individuals. Here we investigate the association between the PNPLA3variant and anthropometric traits, including body mass index (BMI) and whole body fat distribution, in a cohort of healthy blood donors. Patients and methods: We recruited a total of 1.000 (females n = 500; median age 24 years, range 18 - 66 years) blood donors from the Regional Blood Donor Center in Szczecin (Poland). All subjects had a medical checkup, and good state of health was a prerequisite to qualify for blood donation. The PNPLA3variant was genotyped using PCR-based assays with 5'-nucIease and fluorescence detection. All individuals were phenotyped with respect to anthropometric characteristics (body weight, height, BMI, hip, waist, chest, shin and forearm circumferences). We also determined the percentage of total fat (F%) and active tissue (TA%) of body weight. Results: Overall, we determined the following frequencies of the PNPLA3 genotypes: [II] - 61.0%, [IM] - 33.3%, [MM] - 5.7%. [IM] and [MM] carriers, although not differing in height from individuals with the genotype [II], displayed significantly lower body weight (72.5 ± 14.9 vs.75.4 ± 16.1, P = 0.005) and lower BMI (24.2 ± 3.8 vs.24.9 ± 4.2 kg/m2, P = 0.009), higher TA% (68.4 ± 6.3 vs.67.6 ± 5.5%, P = 0.03) but lower F% (31.6 ± 6.3 vs.32.4 ± 5.5%, P = 0.03) and smaller waist, chest and shin circumferences (all P < 0.05). Separate analysis for males and females demonstrated an association between the [IM] and [MM] genotypes and lower TA% but higher F% (both P = 0.04) in females. In males, in turn, only shin circumference was significantly associated with the PNPLA3 variant (P =0.02). Discussion: Several loci modulating whole body fat distribution, also in gender-specific manner, have been identified so far.3 Here we demonstrate for the first time that individuals carrying the prosteatotic PNPLA3 allele p.148M might be leaner and have lower amounts of whole body fat as compared to the carriers of the common allele. Hence in clinical practice, carriers of variant PNPLA3 polymorphism may be easily overseen since they do not necessarily present with the anthropometric characteristics commonly associated with severe hepatic steatosis.1. Sookoian et al, Hepatology.2011 2. Krawczyk et al, Best Pract Res Clin Gastroenterol.2010 3. Heidel et al, Nat Genet.2010

Disclosures:

The following people have nothing to disclose: Marcin Krawczyk, Agnieszka Kempinska, Marta Klak, Malgorzata Blatkiewicz, Piotr Milkiewicz, Frank Lammert, Malgorzata Milkiewicz

604

  1. Top of page

Signaling Pathway Discovery in Nonalcoholic Fatty Liver Disease by Phosphoproteomic Profiling

Julia Wattacheril2,1,Kristie Rose1,Christian P Lanciault1,Clark R. Murray1, Noji N. Abumrad1, Robb Flynn1
1Surgery, Vanderbilt University, Nashville, TN; 2Medicine, Columbia University, New York, NY

Molecular signaling events associated with the distinct necroinflammatory changes in nonalcoholic steatohepatitis (NASH) as opposed to the steatotic changes in nonalcoholic fatty liver disease (NAFLD) are not completely understood. The former is strongly associated with the eventual development of advanced fibrosis alongside metabolic changes; the latter is clearly associated with the development of insulin resistance and the metabolic syndrome with less impact on the liver parenchyma. In order to understand these signaling events, we used an unbiased strategy to compare the global differences in liver protein reversible phosphorylation across 30 Class III obese subjects (10 obese normal, 10 simple steatotic and 10 NASH subjects) biopsied during bariatric surgery. Complex phosphopeptide mixtures were enriched by titanium dioxide and subject to multidimensional protein identification technology proteomic profiling utilizing on-line strong cation exchange and reversed phase nano-flow LC. In total, 4, 122 phosphorylation sites (3, 403 phosphoserine, 654 phosphothreonine, 215 phosphotyrosine) were detected and mapped to 2, 033 phosphoproteins. Manual and bioinformatics-based comparisons of phosphorylation abundance revealed specialized signaling pathways unique to each NAFLD cohort. Analyses of proteins identified differences among several signaling pathways, such as insulin signaling, TCA cycle, and lipid metabolism. When combined with spectral abundance measurements of detected kinases and their substrates, our findings shed new light on pathways not previously emphasized in the pathogenesis of NASH. Bioinformatics analyses suggest progressive shifts in the activity of at least 20 different kinases associated with the more deleterious and clinically relevant variant of NAFLD, NASH. Furthermore, consistent with the importance of Wnt/catenin signaling in maintaining zonation and mediating tight junction functionality, several phosphorylation sites on a, p and 5-catenin as well as other downstream targets were differentially expressed across NAFLD severity. This largest ever repository of site-specific phosphorylation data specific to human NAFLD opens the door to a better understanding of protein signaling in the liver and provides unprecedented insight into the etiopathogenesis of NASH.

Disclosures:

The following people have nothing to disclose: Julia Wattacheril, Kristie Rose, Christian P. Lanciault, Clark R. Murray, Naji N. Abumrad, Robb Flynn

605

  1. Top of page

Elevated HIF-1a in Pediatric Non-Alcoholic Fatty Liver Disease

Shikha Sundaram1/, Sean Colgan2, Zhaoxing Pan2, Kristen N. Robbins1, Ann Halbower1, Kelley E. Capocelli1, Amanda Bayless1, Ronald J. Sokol1
1 Children's Hospital Colorado, University of Colorado School of Medicine, Aurora, CO; 2University of Colorodo School of Medicine, Aurora, CO

Emerging evidence suggests that obstructive sleep apnea (OSA), mediated by intermittent hypoxemia, may play a role in Non-Alcoholic Fatty Liver Disease (NAFLD) pathogenesis. Objective: To evaluate the relationship between OSA and hypoxia inducible factor (HIF) in pediatric NAFLD. Methods: Adolescents (10-18 yrs) with biopsy proven NAFLD and lean age-matched controls (BMI <85%; normal AST/ALT) participated. Demographic and clinical data and fasting blood tests were collected. Circulating HIF-1 a concentrations from white blood cell (WBC) buffy coat were assessed by electrochemiluminescent ELISA. NAFLD subjects underwent standard sleep study. Results: We studied 24 NAFLD subjects (mean age 12.2 yrs; mean BMI z score 2.3, 67% male, 88% Hispanic) and 8 lean controls (mean age 13.4 yrs; mean BMI z score 0.03, 38% male, 25% Hispanic). Mean ALT (117 ± 86 vs 33土 3 IU/L), AST (71 ± 42 vs 41 ± 10 IU/L), triglyceride (145 ± 71 vs 83 ± 4), cholesterol (146 ± 45 vs 119 ± 11), HOMA-IR (8.8 ± 7.5 vs 3.0 ± 1.9), CRP (2.5 ± 2.9 vs 0.6 ± 1.1) and leptin (31.2 ±11. / vs 11.43 ± 11.6) were higher, and adiponectin (6.9 ± 3.3 vs 13.7 ± 1.4) lower, in NAFLD vs lean controls, respectively, p<0.02. Circulating WBC cell HIF-1 a levels in NAFLD subjects were increased 1.9 fold compared to lean controls (1.7 ±1.4 vs.0.9 ± 0.6 per mg protein; p=0.03). OSA was present in 54% of NAFLD subjects. The Apnea Hypopnea Index (AHI) was significantly higher in NAFLD with OSA (10.7 ± 7.9) than without OSA (1.1 ± 0.7; p=0.0009). NAFLD subjects with OSA had lower oxygen nadirs (81.2 ± 5.3 vs 88.1 ± 3.2) and % time with saturations <90% (3.7 ± 4.5 vs 0.33 ± 0.78) than those without OSA, p<0.02. Although circulating WBC HIF-1 a levels were similar between NAFLD with and without OSA (1.37 ± 1.53 vs.2.16 ± 1.24 per mg protein, p=0.3), a significant inverse correlation was noted between HIF-1 a levels and AHI (r= −0.48, p=0.02). While histologic grade was similar between groups, subjects with OSA had significantly worse fibrosis (Stage 0: 21%, Stage 1: 29%, Stage 2: 14% Stage 3: 36%) compared to those without OSA (Stage 0: 9%, Stage 1: 73%, stage 2: 18%; p=0.03). Moreover, WBC HIF1a levels inversely related to worsening fibrosis stage (r= −0.50, p=0.01). Conclusions: Pediatric NAFLD patients frequently have moderate OSA and hypoxia, with increased circulating WBC HIF-1 a levels compared to lean children. Elevated circulating WBC HIF-1 a levels are present in subjects with the mildest OSA and fibrosis. These findings suggest a potential protective role of HIF in disease progression of NAFLD that requires further study.

Disclosures:

Ronald J. Sokol - Advisory Committees or Review Panels: Yasoo Health, Inc., Ikaria, Yasoo Health, Inc., Ikaria; Consulting: Roche, Roche; Grant/Research Support: Lumena

The following people have nothing to disclose: Shikha Sundaram, Sean Colgan, Zhaoxing Pan, Kristen N. Robbins, Ann Halbower, Kelley E. Capocelli, Amanda Bayless

606

  1. Top of page

Weight Loss And Resolution Of Steatohepatitis Is Associated With Improved Pancreatic Beta-Cell (p-Cell) Function In Patients With Nonalcoholic Steatohepatitis (NASH)

Mohammad S. Siddiqui1, Velimir A. Luketic1, Puneet Puri1, Sherry L. Boyett1, Carol Sargeant1, Katherine P Yates5, Naga P Chalasani3, Cynthia D. Guy2, Rohit Kohli4, Aynur Unalp-Arida5, Arun J. Sanyal1
1 Internal Medicine, VCU, Richmond, VA; 2Duke University, Durham, NC; 3Indiana University, Indianapolis, IN; 4Cincinnati Children's Hospital, Cleveland, OH; 5John Hopkins University, Baltimore, MD

BACKGROUND: Nonalcoholic steatohepatitis (NASH) is associated with increased risk of type 2 diabetes, insulin resistance and pancreatic p cell failure. It is not known if resolution of steatohepatitis is associated with improved p cell function. AIM: To test the hypothesis that resolution of steatohepatitis is associated with improvement in p-cell function in non-diabetic patients with NASH. METHODS: Analysis of data from the PIVENS trial where subjects with NASH received either pioglitazone (PIO), Vitamin E (VitE) or placebo (PL) for 96 weeks was performed (Sanyal et al. NEJM 2010). Individual clinical, laboratory and histological outcomes at baseline and end of treatment (EOT) were available. Pancreatic -cell function was calculated via insulin sensitivity index (ISIest). The ISIest measures insulin response at 2 hours to plasma glucose at 90 min in an oral glucose tolerance test (oGīT). RESULTS: At study entry, the p cell function was comparable across groups. PIO improved p cell function (entry to EOT mean ISIest: 0.007 vs 0.034, p< 0.05) whereas VitE or PL had no significant effect.47.1% vs.36.2% vs.20.8% subjects (PIO vs vit E vs PL) did not have steatohepatitis at EOT. The mean ISIest was higher (better p cell function) in those without steatohepatitis (vs those with) at EOT in the PIO (0.05 vs 0.016, p< 0.004) and PL arms (0.045 vs 0.004, p< 0.01); these differences did not reach significance for VitE (0.025 vs 0.016, p=ns). For the entire cohort, regression analysis demonstrated that absence of steatohepatitis at EOT, pioglitazone therapy and weight loss were independently associated with improvement in p cell function. Due to co-linearity, insulin sensitivity could not be included in the model and the impact of improved insulin sensitivity could not be assessed. The correlation coefficients of changes in individual histological features of steatohepatitis and p cell function were: Steatosis (ISIest: 0.188, p < 0.001), ballooning (ISI est: −0.098 p=ns) and inflammation (ISIest: −0.043, p ns). CONCLUSION: Improved pancreatic p cell function is associated with resolution of steatohepatitis, improvement in steatosis, weight loss and PIO therapy.

Disclosures:

Velimir A. Luketic - Grant/Research Support: Intercept, Merck, Idenix, Vertex, Gilead, BMS, Novartis, abbvie, Genfit, Takeda

Puneet Puri - Advisory Committees or Review Panels: Health Diagnostic Laboratory Inc.; Consulting: NPS Pharmaceuticals Inc.

Naga P. Chalasani - Consulting: Salix, Abbott, Merck, Lilly, Enterome, Aegerion; Grant/Research Support: Intercept, Lilly, GenFit, Gilead, Enterome, Cumberland, Galectin

Rohit Kohli - Grant/Research Support: Johnson and Johnson, Johnson and John-

Arun J. Sanyal - Advisory Committees or Review Panels: Gore, Gilead, Abbott, Ikaria; Consulting: Salix, Immuron, Exhalenz, Bayer-Onyx, Genentech, Norgine, GalMed, Novartis, Echosens, Takeda; Grant/Research Support: Salix, Genentech, Genfit, Intercept, Ikaria, Takeda, Gilead; Independent Contractor: UpToDate

The following people have nothing to disclose: Mohammad S. Siddiqui, Sherry L. Boyett, Carol Sargeant, Katherine P. Yates, Cynthia D. Guy, Aynur Unalp-Arida

607

  1. Top of page

Novel Quantification of Liver Fat by Ultrasound Backscatter Technique & its Correlation with MRI-estimated Proton Density Fat Fraction in NAFLD vs Healthy Controls

Rohit Loombo1' Abdullah Alturki2, A. Han2, Jessica Lam2, Brandon Ang1, Archana Bhatt1, Jonathan Hooker2, A. Shah2, K. Zand2, Michael S. Middleton2, Tanya Wolfson3, William D. O'Brien4, Claude B. Sirlin2, Michael P. Andre2
1Division of Gastroenterology, University of California at San Diego, La Jolla, CA; 2Department of Radiology University of California at San Diego, La Jolla, CA; 3Department of Mathematics, University of California at San Diego' Lo Joilo, CA; "Departments of Bioengineering, and Medical Information Sciences, University of Illinois, Urbana-Champaign, IL

Background: Conventional ultrasound (US) lacks sensitivity & specificity in diagnosing hepatic steatosis (HS) & is unable to quantify liver fat content (LFC). MRI proton density fat fraction (PDFF) can accurately diagnose & quantify HS but is expensive & impractical for population-based screening of NAFLD. We have developed a novel quantitative US (QUS) method that in animal studies shows promise for detection & quantification of LFC. Aim: To correlate QUS backscatter coefficient (BSC) with MRI PDFF as indicators of HS in a cohort of adults pts with NAFLD & normal controls. Methods: We conducted a prospective study derived from a cohort of consecutive pts with NAFLD (MRI PDFF > 5%) & normal controls (NC) (MRI PDFF < 5%). Both the NAFLD & the NC group underwent a detailed clinical research visit. Liver MRI & US was performed on the same day. MRI-PDFF was estimated & QUS measurements were made. Performance of QUS-derived BSC to diagnose HS, using MRIPDFF > 5% as the reference standard, was evaluated by ROC analysis. Results: Among the total of 68 (67% M) pts, 29 had NAFLD (MRI PDFF range; 5.7-35.3%) & 39 were NC (MRI PDFF range; 1.1-4.6%). The mean (± SD) age & BMI of NAFLD pts vs NC was 47.8 (±13.8) vs 37.7 (±20.5) yrs, & 32.5 (±4.5) vs 25.9 ((±5.8) kg/m2, respectively. QUS BSC at 3.0 MHz ranged over two orders of magnitude from 0.0002 to 0.087 1/(cm-sr) & correlated well with MRI PDFF (R2=0.76, figure). A BSC threshold of 0.002/cm-sr provided a raw sensitivity of 97%, & specificity of 92% with an AUROC of 99% (95% CI, 95.9-99.9) for the diagnosis of HS. Conclusion: In this proof of concept study, the QUS BSC, measureable with a simple & inexpensive US technique, can accurately detect the presence of NAFLD & quantify HS in human subjects. If validated in a larger cohort, these results may have significant implications for screening NAFLD at the level of population.

Thumbnail image of

Disclosures:

Rohit Loomba - Consulting: Gilead Inc, Corgenix Inc; Grant/Research Support: Daiichi Sankyo Inc, AGA

Michael S. Middleton - Consulting: Gilead, Pfizer, Synageva, Merck, Bracco; Grant/Research Support: Isis, Genzyme, Siemens, Bayer; Stock Shareholder: General Electric

Claude B. Sirlin - Advisory Committees or Review Panels: Bayer, ISIS, Bayer, ISIS; Consulting: Genzyme, Gilead, Siemens; Grant/Research Support: GE, Bayer, GE, Bayer, Pfizer; Speaking and Teaching: Bayer, Bayer

The following people have nothing to disclose: Abdullah Alturki, A. Han, Jessica Lam, Brandon Ang, Archana Bhatt, Jonathan Hooker, A. Shah, K. Zand, Tanya Wolfson, William D. O'Brien, Michael P. Andre

608

  1. Top of page

The impact of phlebotomy in non-alcoholic fatty liver disease: interim results of a randomized controlled trial

Leon Adams1,2 Michael J. House3, Timothy G. St Pierre3, Darrell Crawford4, Katherine A. Stuart5, Helena Ching2, Jenny Kava7, Malcolm Webb6, John K. Olynyk7
1Medicine and Pharmacology, University of Western Australia, Nedlands, WA, Australia; 2Gastroenterology and Hepatology, Sir Charles Gairdner Hospital, Perth, WA, Australia; 3School of Physics, University of Western Australia, Nedlands, WA, Australia; 4School of Medicine, University of Queensland, Brisbane, QLD, Australia; 5Gastroenterology, Princess Alexandria Hospital, Brisbane, QLD, Australia; 6Hematology, Fremantle Hospital, Perth, WA, Australia; 7Gastroenterology, Fremantle Hospital, Perth, WA, Australia

Iron is implicated in the pathogenesis of liver injury and insulin resistance. Consequently iron removal by phlebotomy has been proposed as a treatment strategy for patients with non-alcoholic fatty liver disease (NAFLD). We wished to examine the impact of iron reduction by phlebotomy on liver injury, hepatic steatosis and insulin resistance in patients with NAFLD by performing a prospective 6-month randomized controlled trial. Interim results of the initial 53 completed subjects are presented. Methods: Subjects with imaging confirmed NAFLD were randomly allocated to phlebotomy plus lifestyle advice or lifestyle advice only. Phlebotomy was performed every 2-3 weeks as tolerated, aiming for a ferritin<100. Subjects were assessed at baseline and end of study for liver injury (aminotransaminases, Hepascore) and hepatic steatosis and iron content using magnetic resonance imaging. Insulin resistance was assessed using HOMA (fasting glucose and insulin) and the insulin sensitivity index (ISI) based upon the frequently sampled oral glucose tolerance test. Results: 63 of a planned 66 subjects have been screened and randomized with 53 subjects completed. The mean (±SD) age was 52 (±11) years with 33 (62%) being male. The baseline median (IQR) serum ferritin was 392 (201-685) mcgm/l, transferrin saturation 29% (23-35%), liver iron concentration 1.0 (0.6-1.5) mg/gm and hepatic fat index 0.17 (0.10-0.30). Phlebotomy (n=26) and control (n=27) groups had similar anthropometric, biochemical and metabolic parameters apart from serum cholesterol, which was significantly higher in the controls [232 (35)mg/dl vs.186 (35) mg/dl, p<0.001]. Subjects in the phlebotomy group underwent a median of 6 (IQR 3-8) venesections which were tolerated well without complications. Subjects in the phlebotomy group had a significantly greater reduction in serum ferritin over the study period compared to controls [284 (114-510) mcgm/l vs.64 (25-156) mcgm/l, p=0.002). After 6 months, there was no difference in liver aminotransaminases, Hepascore values, hepatic steatosis, hepatic iron concentration, HOMA or ISI (p>0.2 for all). No significant differences between groups were noted at end of study after stratification by baseline serum ferritin, number of venesections, hepatic iron concentration or hepatic steatosis content. Conclusions: Interim results do not support a role of phlebotomy to improve liver enzymes, hepatic fat or insulin resistance in subjects with NAFLD.

Disclosures:

Michael J. House - Consulting: Resonance Health; Patent Held/Filed: Resonance

Timothy G. St. Pierre - Board Membership: Resonance Health Ltd; Consulting: Resonance Health Ltd; Patent Held/Filed: Resonance Health Ltd; Stock Shareholder: Resonance Health Ltd

Darrell H. Crawford - Advisory Committees or Review Panels: Roche Products Pty Ltd, Bristol Myers Squibb, Gilead Sciences, Novartis, MSD, Abbvie; Consulting: Roche Products Pty Ltd; Grant/Research Support: Roche Products Pty Ltd; Speaking and Teaching: Roche Products Pty Ltd, Bristol Myers Squibb, Gilead Sciences,

Katherine A. Stuart - Grant/Research Support: Gilead, Bayer, Roche

The following people have nothing to disclose: Leon Adams, Helena Ching, Jenny Kava, Malcolm Webb, John K. Olynyk

609

  1. Top of page

RBC Lipid Composition Is Favorably Altered in NASH Patients Treated with Omega-3 Fish Oil versus Placebo

Glenn Moulder, Elliot Z. Smith, Stephen H. Caldwell, Curtis K. Argo; Internal Medicine, Gastroenterology and Hepatology, Universify of Virginia' Charlottesville' VA

Background: Dietary polyunsaturated fatty acids (PUFAs) mediate hepatocyte inflammation. The ratio of pro-inflammatory omega-6 fatty acids, primarily arachidonic acid (AA), to antiinflammatory omega-3 fatty acids, primarily eicosapentaenoic acid (EPA), is elevated in NASH patients. We aimed to evaluate the effects of treatment with omega-3 fatty acid supplementation on RBC fatty acid levels in patients with biopsy proven NASH. Methods: Thirteen patients with biopsy proven NASH received a daily omega-3 (n=6) or placebo (n=7) for one year as a subset of a larger NASH trial. Treatment efficacy was determined via comparisons of baseline and one year serum erythrocyte fatty acid levels of AA: EPA (omega-6: omega-3) ratios. Clinical response was measured via 1) serologic analysis of lipid metabolism, glucose metabolism, and NASH biomarkers; 2) magnetic resonance imaging (MRI) to determine hepatic steatosis; and 3) liver biopsy samples graded by Dixon fat scoring system and NAFLD activity score (NAS). Results: After one year of treatment, the omega-3 group displayed a significant decrease in AA: EPA levels (p=0.02) as well as an decreased ratio compared to patients who received placebo (p=0.003). The omega-3 group displayed no significant change in other lipid or glucose metabolism parameters, while serum biomarkers of hepatocyte damage suggestive of NASH (ALT, M30/M65E cytokeratin antigens) all decreased but not to significant levels. On imaging, the omega-3 group displayed a decrease in MRI fat score (p=0.009). Histologic analysis demonstrated the omega-3 group had decreases in Dixon fat scores (p=0.04) and NAS (p=0.01). Furthermore, based on NAS histology, four patients (80%) showed histologic resolution of NASH after a year of omega-3 compared to two (29%) patients taking placebo. Conclusions: Daily omega-3 supplementation decreased the ratio of pro-inflammatory to antiinflammatory PUFA levels based on AA: EPA (omega-6: omega-3) measurement, which we speculate reflects shifting of hepatic fat metabolism to a less lipotoxic and inflammatory form. The treatment effects of omega-3 fatty acids on NASH need further study with larger randomized placebo controlled trials.

Placebo (n=7) Omega-3 (n=6)
TOT12p valueTOT12p value
Body Mass Index35.1 ±9.135.4 ± 10.30.6632.1 ±3.630.6 ±4.30.04
AA: EPA ratio81.9 ±33.446.9 ± 19.70.08854.3 ± 30.613.3 ±8.50.02
MRI fat score238 ± 147186 ±1330.26295 ±139127 ±1050.009
Dixon fat score11.9 ±7.59.2 ±5.10.2318.3 ± 118.33 ±6.30.04
NAFLD activity score4.86 ± 0.854.36 ±1.50.135.92 士 0.864.1 土 1.10.01

Disclosures:

Stephen H. Caldwell - Advisory Committees or Review Panels: Vital Therapy; Consulting: Wellstat diagnostics; Grant/Research Support: Hemosonics, Gilead Sciences

Curtis K. Argo - Consulting: Wellstat Diagnostics; Independent Contractor: Genentech/Roche

The following people have nothing to disclose: Glenn Moulder, Elliot Z. Smith

610

  1. Top of page

NAFLD as a contributor to early atherosclerotic lesions: a longitudinal study in 2169 asymptomatic patients at high cardiovascular risk

Raluca Pais, Philippe Giral, Hugo Perazzo, Jean-Francois Khan, Lory so Fedchuk, Pascal Lebray, David Rosenbaum, Mono Munteonu, Thierry Poynard, Vlod Rofziu
Universife Pierre ef Marie Curie, Hopital Pitie Salpetriere, Assistance Publique Hopitaux de Paris, Paris, France

Background. NAFLD increases the risk of cardiovascular (CV) events independent of the presence of traditional CV risk factors. Whether NAFLD is associated with early atherosclerotic lesions and their progression is unknown. Aim: to evaluate the impact of NAFLD and significant hepatic fibrosis on the presence and progression of carotid intima-media thickness (CIMT) and early carotid plaques (CP), in patients (pts) at high CV risk. Methods: Pts with >2 CV risk factors (hypertension, diabetes, dyslipidemia, obesity, active smoking), but no CV events, chronic liver disease and alcohol<50g/d underwent baseline and follow-up (f/u) carotid ultrasonography and Framingham risk score (FRS) calculation. CIMT was graded 0: normal; 1: <1 mm wall thickening without CP; 2: moderate CP (≥1-≤2.5 mm), 3: CP>2.5 mm). Progression was defined as transition to the next, higher class. Steatosis and fibrosis were assessed by the Fatty Liver Index (FLI, NAFLD present when >60), and by FibroTest (FT). Results.2169 patients were enrolled: 56% males, 52 year-old, BMI 25.4 kg/m2; mean FRS 12±9%, mean CIMT 0.62±0.14 mm; 40.5% had CP and 24% had a FLI60. Pts with NAFLD had higher CIMT (0.64±0.15 vs.0.61 ±0.14 mm, p=0.001), higher prevalence of CP (30% vs.24%, p=0.005) and higher FRS (17±9% vs.10±8%, p<0.001). FLI was associated with baseline CIMT (p=0.002) and FLI≥60 with CP at baseline (OR=1.27, p=0.04), both independent of age, sex, smoking, diabetes and hypertension. The median f/u was 8 yrs, 6.4 yrs in NAFLD pts and 8.5 yrs in non-NAFLD pts (p<0.001). During f/u, CIMT increased from 0.62 to 0.65 mm, p<0.001, prevalence of CP increased from 41% to 58%, p<0.001 while 38% of pts developed CP. NAFLD at baseline was associated with the progression and occurrence of CP independent of age, sex or the FRS (HR 1.30 and 1.34, respectively both p<0.01). Among non-NAFLD pts at baseline, those who developed NAFLD during f/u had a larger increase in CIMT than those who stayed NAFLD-free.455 pts were evaluated by FT, they were not different from the 1714 untested pts for age, BMI, CIMT and CP prevalence.2% of these patients had a FT>0.48 compatible with bridging fibrosis. Unexpectedly, bridging fibrosis was associated with the presence of CP at baseline and with progression of CP at f/u (HR 3.83, p<0.01), both independent of FRS and FLI>60. Conclusion: In patients at high CV risk, NAFLD and in particular bridging fibrosis contribute to early atherosclerosis and progression thereof, independent of traditional CV risk factors.

Disclosures:

Pascal Lebray - Grant/Research Support: Schering Plough; Speaking and Teaching: Janssen, MSD, Gilead

Mona Munteanu - Employment: Biopredictive

Thierry Poynard - Advisory Committees or Review Panels: Merck; Grant/Research Support: BMS, Gilead; Stock Shareholder: Biopredictive

Vlad Ratziu - Advisory Committees or Review Panels: GalMed, Abbott, Genfit, Enterome, Gilead; Consulting: Astellas, Axcan, Pfizer, Sanofi-Synthelabo, Genentech, Nycomed

The following people have nothing to disclose: Raluca Pais, Philippe Giral, Hugo Perazzo, Jean-Francois Khan, Larysa Fedchuk, David Rosenbaum

611

  1. Top of page

Epicardial fat thickness (EAT) in non alcoholic fatty liver disease: a candidate marker of steatosis and vascular damage

Anno Ludovico Froconzoni1, Giuseppino Pisano1,Rosa Lombardi1, Luca Valenti1, Cristina Bertelli1, Tiziana Tonella2, Ferdinando Massari2, Andrea Baragetti3, Liliana Grigore3, Alberico Catapano3, Silvio Fargion1
1 Department of clinical, surgical pathophisiology and organ transplant, University of Milan, Maggiore Policlinico Hospital,IRCCS CA Granda Foundation, Millan, Italy; 2Dipartimenfo di Torocopolmonore e Cardiocircolatorio' University of Milan' Maggiore Policlinico Hospital,RCCS CA Granda Foundotion, Milan, Italy; 3Centro Studi Aterosclerosi, Dipartimento di Scienze Farmacologiche, University of Milan, Milan, Italy

Background & Aims: Epicardial adipose tissue (EAT) has been implicated in the pathogenesis of coronary atherosclerosis, and its measurement during echocardiography proposed as a new index of cardiac and visceral adiposity. EAT was found increased in patients with metabolic syndrome, of which nonalcoholic fatty liver disease (NAFLD) represents the hepatic manifestation. Aim of this study was to evaluate 1) whether EAT thickness is increased in subjects with NAFLD vs. healthy controls 2) clinical factors associated with EAT and the association with the severity of NAFLD 3) whether EAT predicts early atherosclerotic vascular damage, evaluated by common carotid arteries intima-media thickness (CC-IMT), and subclinical cardiac dysfunction. Methods: EAT thickness, i. e. echofree space between the outer wall of the myocardium and the visceral layer of pericardium, was evaluated by transthoracic echocardiogram in 155 consecutive patients, 43 with biopsy proven NAFLD (mean age 51 ±11 years), and in 87 healthy controls (mean age 52± 11 years), in whom hepatic steatosis was excluded by abdominal ultrasonography. In all patients complete anthropometric, clinical and biochemical data were obtained, and CC-IMT evaluated. Results: Patients with NAFLD had higher EAT thickness than controls (5.1 ±2.6 vs.3.0±2.0 mm, p=0.001). At univariate analysis, BMI (OR 1.16, 95% CI1.33), fasting glucose >100 (OR 4.2, 95% CI 1.14-20), HbA1c (OR 2.6, 95%CI 1.19-8.3), diabetes (OR 1.28, 95% CI1.73), and metabolic syndrome (OR 1.34, 95% CI 1.151.57) were significantly associated with increased EAT thickness (above median value). In the 43 patients with liver histology (15 simple steatosis, 25 nonalcoholic steatohepatitis without cirrhosis, and 3 NAFLD with cirrhosis), EAT thickness increased with the severity of steatosis (p=0.01). Increased EAT thickeness was associated with CC-IMT >0.65 mm (median value of controls: OR 4.6, 95% CI 1.6-15.9). Furthermore, EAT thickness was inversely correlated with cardiac early diastolic dysfunction, as detected by the early/atrial peak flow ratio (E/A ratio) (OR 0.07, 95% CI 0.01-0.38). Conclusions: EAT thickness is higher in NAFLD patients than in healthy controls, is associated with adiposity and insulin resistance, and with early markers of cardiovascular risk. Further studies in large cohorts are required to define whether EAT thickness represents an easily assessable diagnostic tool to predict cardiovascular risk.

Disclosures:

The following people have nothing to disclose: Anna Ludovica Fracanzani, Giuseppina Pisano, Rosa Lombardi, Luca Valenti, Cristina Bertelli, Tiziana Tonella, Ferdinando Massari, Andrea Baragetti, Liliana Grigore, Alberico Catapano, Silvia Fargion

612

  1. Top of page

Metabolomics in NAFLD: Evidence of a protein catabolic state and a basis for mTOR activation

Moureen M. Guichelaar2' Anurodho Krishnan1,Edith M. Koehler3, X. Moi Persson4, Sreekumoron K. Noir5, Michael R. Charlton1
1Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN; 2Gastroenterology and Hepatology, Medical Spectrum Twente, Enschede, Netherlands; 3Department of Gastroenterology and Hepatology, Erasmus MC, Rotterdam, Netherlands; 4Metabolomics Core, Mayo Clinic, Rochester, MN; 5Endocrinology Mayo Clinic, Rochester, MN

Introduction: Insulin resistance and inflammation are hallmarks of NASH. In addition to the well-known effects on carbohydrate metabolism, insulin resistance and inflammatory cytokines have profound effects on protein and amino acid metabolism. The aim of this study was to develop detailed amino acid metabolism signatures across the histologic stages of NAFLD. Methods: Five groups of participants were studied: 1) lean controls (n=20), 2) obese with normal histology (n=10), 3) obese with simple steatosis (n=10), 4) obese with NASH (FS 0-3)(n=20). Obese patients were matched for BMI, age and gender. Al participants were matched for age and gender. Serum was analyzed for 36 amino acids and their metabolites on a Thermo Quantum Ultra triple quadrupole mass spectrometer coupled with a Waters Acquity ultra high pressure liquid chromatograph (LC MS/MS). Results: Levels of 20 amino acids and metabolites were significantly higher in study participants with obesity when compared to lean participants (p=0.03-0.001). Patients with NASH (versus simple steatosis and normal histology) had higher levels of alanine (chi square 6.32, p = 0.01), b-aminobutyric acid (chi square 4.99, p= 0.03), 2 of the 3 aromatic amino acids; tyrosine (chi square 4.99, p = 0.03), and tryptophan (chi square 7.55, p = 0.006). Histological liver injury parameters correlated negatively with branched chain amino acids and aromatic amino acids, alanine and several other amino acids. Conclusion: This study illustrates that obesity, despite chronic overnutrition, is associated with an amino acid profile that is profoundly proteolytic / catabolic in nature. Amino acid abnormalities correlated with histological features of NASH. This proteolytic profile seen in this study has been shown to cause distinct metabolic effcts; i. e. the increase in serum amino acids and increased BCAA oxidation has been associated with increased mTOR activity, contributing to insulin resistance, mitochondrial dysfunction, and increased cancer and cardiovascular risk. Further research is necessary to determine the effects of increased amino acids, mTOR and liver injury in the context of NASH.

Disclosures:

The following people have nothing to disclose: Maureen M. Guichelaar, Anuradha Krishnan, Edith M. Koehler, X. Mai Persson, Sreekumaran K. Nair, Michael R. Charlton

613

  1. Top of page

Optimal Modality and Method for the Evaluation of Donor Steatosis for Living Donor Liver Transplantation

Mi-Jung Jun1, Ju Hyun Shim2, Kang Mo Kim2, Young-Suk Lim2, Han Chu Lee2, Dong Jin Suh3
1Asan Medical Center, Seoul, Republic of Korea; 2Department of Gastroenterology, Asan Medical Center, Seoul, Republic of Korea; 3Department of Internal Medicine, Vievis Namuh Hospital, Seoul, Republic of Korea

Background: Living donor liver transplantation (LDLT) has been increasing due to the critical shortage of cadaver livers and the increasing number of patients waiting for LT. Hepatic steatosis is an important factor associated with primary non-function of the recipient allograft as well as an increase in the risk of donor complications caused by a reduced functional hepatic mass. To date, there is no consensus regarding the pre-donation workup strategy for selecting appropriate donor livers. The purpose of this study is to assess the optimal tools for evaluating the donor liver fat content. Method: A total of 1, 766 living donors underwent abdominal ultrasonography (USG) and computed tomography (CT) as donor selection procedures, along with preoperative right liver biopsy and intraoperative paired right and left liver biopsies. The agreement of the steatosis grade (<5%; 5-15%; 15-30%; and >30%) between imaging and biopsy and between biopsies was assessed using the k statistic, and the clinico-metabolic factors related to sampling variability were identified using logistic regression analysis. Result: The sensitivities of USG and CT were 58.9% and 46.4%, respectively, for detecting >30% steatosis with positive predictive values of 6.9% and 16.9%. The combination of the two imaging modalities improved the sensitivity result, but still remained unsatisfactory (71.3% for sensitivity and 7.7% for the positive predictive value). For the total steatosis grade between preoperative right biopsy versus intraoperative right and left biopsies, moderate agreement was seen as reflected by weighted kappa values of 0.44 and 0.40. The weighted kappa value between intraoperative right and left biopsies was 0.77, and thus indicating substantial agreement. Similar results were noted in macrovesicular and microvesicular steatosis subtypes. Multivariate analysis indicated that independent factors affecting the sampling variability of the total steatosis in preoperative and intraoperative biopsies, included greater systolic blood pressure (odds ratio [OR], 1.01), body mass index (OR, 1.08) and serum alanine aminotransferase (OR, 1.02), and less high-density lipoprotein cholesterol (OR, 0.98; P <0.05 for all). Conclusions: Our data suggest that radiological studies were considerably limited for detecting donor hepatic steatosis in LDLT and that further substantial sampling variability exists between preoperative and intraoperative liver biopsies, according to the clinical and metabolic parameters. Therefore, preoperative and selective intraoperative liver biopsies should be performed to assess donor steatosis in LDLT.

Disclosures:

Han Chu Lee - Grant/Research Support: Medigen Biotechnology Co., Novartis, Roche, Bayer HealthCare, Bristol-Myers Squibb, INC research, Boehringer Ingelheim, Taiho Pharmaceutical Co., Yuhan Co.

The following people have nothing to disclose: Mi-Jung Jun, Ju Hyun Shim, Kang Mo Kim, Young-Suk Lim, Dong Jin Suh

614

  1. Top of page

A Large Independent Validation Study of Nonivasive Scores to Predict Advanced Fibrosis in NAFLD

Charles D. Rice1, Elisabetta Bugianesi6, Jacob George7, Christopher P. Day2, Einar Bjornsson3, Phunchai Charatcharoenwitthaya4, Peter R. Mills5, Paul Angulo1
1Department of Medicine, Division of Digestive Diseases & Nutrition, University of Kentucky Medical Center, Lexington, KY; 2Institute of Cellular Medicine, Faculty of Medical Sciences, Newcastle University, Newcastle, United Kingdom; 3Section of Gastroenterology and Hepatology, National University Hospital, Reykjavik, Iceland; 4FacuIty of Medicine, Siriroj Hospital, Mahidol University, Bangkok, Thailand; 5GartnaveI General Hospital, Glasgow, United Kingdom; 6Division of Gastroenterology and Hepatology, Department of Medical Sciences, University of Torino, Torino, Italy; 7Storr Liver Unit, Westmead Millennium Institute, University of Sydney, Department of Gastroenterology and Hepatology, Westmead Hospital, Sydney, and Westmead, NSW, Australia

Several noninvasive scoring systems had been developed in patients with NAFLD aimed at distinguishing between those with and without advanced liver fibrosis. They are the NAFLD fibrosis score (NFS), the aspartate aminotransferase (AST)/platelet ratio index (APRI), the FIB-4 score, the NIKEI score, and the BARD score. Validation studies, however, had included small number of patients and most came from single centers. The AIM of our study was to perform a large, independent validation of those scoring systems. METHODS: A total of 672 patients with NAFLD were included. Patients came from several institutions around the world and none of these patients had been included in the original prior studies that created the scoring systems. Fibrosis was staged on the scale 0 to 4 proposed by Kleiner et al. with stage 3 and 4 meaning adavnced fibrosis. The five scores mentioned above were calculated using the original published formulas. The same two cut-points described in the original publications were used to group patients; they were −1.455 and 0.676 for the NAFLD-FS; 0.5 and 1.5 for the APRI; 1.30 and 2.67 for the FIB-4; and 0. 。535 and 0.2294 for the NIKEI score. For the BARD score, a value >2 was used. We calculated the area under the receiver operating characteristics curve (AUROC), the positive predictive value (PPV) to identify advanced fibrosis, the negative predictive value (NPV) to rule out advanced fibrosis, and the proportion of patients with indeterminate values. RESULTS: The accuracy of the scores is described in the Table below. Based on the 95%CI, the NFS was equally accurate as the FIB-4 score, but significantly more accurate than the APRI, NIKEI and BARD scores. The FIB-4 was similar to the NIKEI and BARD scores, but significantly better than the APRI score. There was no a significant difference among the NIKEI, APRI and BARD scores. The NIKEI had the lowest indeterminate score value, but also the lowest NPV and PPV. CONCLUSIONS: This large independent validation analysis demonstrates the high accuracy of noninvasive scores to distinguish between patients with and without advanced fibrosis. Combining the NPV and the PPV, the NFS seems the most accurate followed by the FIB-4, the APRI, the NIKEI and the BARD scores. Accuracy of the socre to distinguish between patients with and without advanced (stage 3/4) fibrosis

AUROC (95%CI)Indeterminate scoreNPV (low cut-point)PPV (high cut-point)
NFS•825 (.789,.861)30%89%84%
APRI.729 (.686,.771)49%86%62%
FIB-4.814 (.777,.851)29%89%76%
NIKEI.749 (.711,.788)1%78%55%
BARD.744 (.702,.786)NA78%55%

Disclosures:

Charles D. Rice - Employment: Sanofi-Spouse; Management Position: SanofiSpouse; Stock Shareholder: Sanofi-Spouse

Jacob George - Advisory Committees or Review Panels: Roche, BMS, MSD, Gilead, Janssen

Christopher P. Day - Advisory Committees or Review Panels: Abbott; Board Membership: Abbott

Paul Angulo - Grant/Research Support: NIDDK, Mochida, Genfit

The following people have nothing to disclose: Elisabetta Bugianesi, Einar Bjornsson, Phunchai Charatcharoenwitthaya, Peter R. Mills

615

  1. Top of page

Coffee and liver disease: evidence of beneficial effects in experimental non-alcoholic steatohepatitis and of direct antifibrotic properties of caffeine in vitro

Juan E. Oyorzun1, Marjolein Tiebosch2, Pablo Quintero1, Margarita Pizarro1, Nancy Solis1, Klaas Nico Faber2, Silvana Zanlungo1, Han Masnhoge2, Marco Arrese1
1 Department of Gastroenterology, Catholic University of Chile, Santiago, Chile; 2Department of Gastroenterology and Hepatology, University Medical Center Groningen, University of Groningen, Groningen, Netherlands

Background: Recent evidence suggest that coffee consumption could be of benefit for those In non-alcoholic fatty liver disease (NAFLD) patients at risk of developing hepatic fibrosis. The underlying mechanisms of hepatic benefits of coffee intake remain unclear. Aims: a) to assess if coffee administration influences hepatic inflammation and fibrosis or mitochondrial respiration in a dietary model of non-alcoholic steatohepatitis (NASH), b) to test the effect of caffeine on hepatic stellate cells (HSC). Methods: C57bl6 mice were fed a choline-deficient amino acid-defined (CDAA) diet for 22 weeks to induce NASH. Unfiltered coffee was added to drinking water during diet administration (CDAA-C and CSAA-C groups). Hepatic steatosis, inflammation, and fibrosis were scored histologically (hematoxylin-eosin and Sirius red staining). Hepatic triglyceride content (HTG) and mRNA expression of inflammatory markers such as TNF-α and MCP-1 as well as mitochondrial respiration were assessed. In addition, primary rat HSC were culture-activated and treated with caffeine for 96h (5 to 20mM) or its main metabolite 1, 7-dimethylxanthine (1, 7-DMX) (for 72h, 1mM). Markers of hepatic stellate cell activation and fibrogenesis (alpha-smooth muscle actin (α-SMA), and collagen1 a 1) were measured using qPCR and western blot, as well as proliferation of stellate cells. Results: CDAA-elicited a typical histological picture of advanced NASH. Coffee administration significantly reduced HTG (117.46±23.59 in CDAA vs.81.74±28.5 in CDAA-C group p<0.05) and mRNA levels of both TNF-a and MCP-1. Also coffee administration was associated to lower scores of fibrosis (2.38±0.48 in CDAA group vs.1.5±0.58 in CDAA-C group, p<0.05) and partially corrected CDAA diet-induced mitochondrial dysfunction. In addition, HSC treated with caffeine exhibited a decrease (−60%) of a-SMA and collagen1 a 1 mRNA levels in a time- and dosedependent manner. Protein levels of a-SMA were also reduced after 72h of caffeine treatment (20mM). Treatment with 1, 7DMX did not result in a reduction of α-SMA and/or collagen1 α 1. Caffeine (20mM) also reduced proliferative capacity of HSC by 50% after 96. CYP1A2 (the enzyme that metabolizes caffeine) was not detected in stellate cells by qPCR. Conclusion: Coffee administration has beneficial effects in a mouse model of NASH. This may be related to reduction in HTG and improvement in mitochondrial function. In addition, caffeine directly reduced HSC activation and proliferation in vitro, independent of its metabolites. These results may explain the protective effects of caffeine on NASH and hepatic fibrogenesis. (FONDECYT 1110455, Conicyt, project ACT79/CARE Basal Project).

Disclosures:

The following people have nothing to disclose: Juan E. Oyarzun, Marjolein Tiebosch, Pablo Quintero, Margarita Pizarro, Nancy Solis, Klaas Nico Faber, Silvana Zanlungo, Han Moshage, Marco Arrese

616

  1. Top of page

CC Genotype of PNPLA3 Polymorphism is Protecti Nonalcoholic Steatohepatitis (NASH)

Julie Guider, Stephen H. Caldwell, Curtis K. Argo
Division of Gastroenterology and Hepatology, University of Virginia, Charlottesville, VA

Background: Prior data suggests that GG genotype of the PNPLA3 SNP confers a higher likelihood of liver fat, inflammation, and ballooning in NASH patients. Conversely, it is unclear whether CC genotype is protective regarding histological disease in NASH. Methods: 33 patients (31 Caucasians, 1 East Indian, and 1 East Asian) with NASH underwent testing for PNPLA3 genotyping, lipids, BMI, and HOMA-IR. MRI imaging was performed to quantify steatosis. Histological data included NAS score, inflammation, ballooning, fibrosis, and steatosis, which was measured both by standard histopathological evaluation and computer-aided image analysis of photomicrographs. Unpaired t-tests compared baseline data from subjects with CC genotype vs. GC+GG genotype. Seventeen patients were treated with fish oil and 16 received placebo for 1 year. We also performed paired t-tests to assess whether genotype predicted response to fish oil therapy. Results: Baseline HOMA scores were higher in the CC group compared with GC+GG (8.3 v.5.4, P = 0.07). Despite this finding, baseline histology tended to be significantly less severe in the CC group with lower fat on biopsy (1.7 v.2.2, P = 0.05), less ballooning (0.9 v.1.3, P = 0.04), less fibrosis (1.6 v.2.0, P = 0.33), and significantly lower NAS scores (4.5 v.5.5, P = 0.0027). Among the patients treated with fish oil, paired t-tests demonstrated no difference in response to therapy in CC genotype subjects. Conversely, the GC+GG genotype subjects showed significant improvements in TGs, image fat, MR fat, and NAS score, although this analysis was not adjusted for change in weight over the study period. Conclusions: At baseline, subjects with CC genotype were more insulin resistant by HOMA-IR. Despite these findings, the CC genotype appears protective histologically regarding liver fat, ballooning, inflammation, and ultimately NAS score. While the GG genotype may predispose to more liver fat deposition, the CC genotype appears to generate a phenotype that is protected from inflammatory and fibrotic changes related to NASH but may be less likely to respond to fish oil.

BMIHOMAFatBallooningFibrosis scoreNAS Score
CC (n=ll)30.58.31.70.91.64.5
GC+GG (n=22)33.65.42.21.32.05.5
P value0.260.070.050.040.330.0027

Disclosures:

Stephen H. Caldwell - Advisory Committees or Review Panels: Vital Therapy; Consuiting: Wellstat diagnostics; Grant/Research Support: Hemosonics, Gilead Sciences

Curtis K. Argo - Consulting: Wellstat Diagnostics; Independent Contractor: Genentech/Roche

The following people have nothing to disclose: Julie Guider

617

  1. Top of page

Is It True That Tamoxifen Can Induce Hepatotoxicity in Clinical Practice?

Myung Jin Oh, Heon Ju Lee, Si Hyung Lee, Sung Bum Kim, Yeoun Su Jung, Jung Woo Lee
Department of Internal Medicine, Yeungnam University College of Medicine, Daegu, Republic of Korea

Purpose: Tamoxifen, an anti-estrogen agent used for treatment of estrogen receptor-positive breast cancer, is widely known to cause hepatotoxicity or non-alcoholic steatohepatitis (NASH). However, a recent study reported that tamoxifen plays a hepatoprotective role against steatosis or NASH. This study aimed to investigate that tamoxifen can induce hepatotoxicity or protect liver injury in clinical practice, and assess risk factors associated with tamoxifen-induced steatosis. Methods: Between Jan.2010 and Dec.2011, of 660 patients who received tamoxifen therapy, a total of 511 patients above 18 years in age were selected.149 patients were excluded due to acute or chronic liver diseases by virus and autoimmune, heavy alcoholic intake, NASH at baseline, and no baseline or follow-up liver function tests. We retrospectively evaluated frequency of tamoxifen-induced hepatotoxicity, analyzed risk factors related to hepatic injury by tamoxifen, and assessed occurrence of hepatic protection by tamoxifen. Results: Mean age of selected patiens was 49.5±9.2 years. Female patients were predominant (93.2%). Mean administration day of tamoxifen was 722.1 ±304.0 days. The hepatotoxicity or elevation of aminotransferase above upper limit of normal was occurred in 80 patients (15.7%). The hepatotoxicity related to tamoxifen developed at 360.6±249.3 days after administration of tamoxifen. Mean elevation level of alanine aminotransferase and aspartate aminotransferase was 42.9 IU/L, and 36.0 IU/L, respectively. Between hepatotoxicity group (n=80) and non-hepatotoxicity group (n=431), body mass index (BMI, 24.9 vs.22.7 kg/m2), baseline alanine aminotransferase (25.3 vs.18.7 IU/L), aspartate aminotransferase (24.9 vs. 21.7 IU/L), Y-glutamyl transferase (39.0 vs.17.5 IU/L), alkaline phosphatase (165.0 vs.146.8 IU/L), total-cholesterol (200.0 vs.186.2 mg/dL), and LDL-cholesterol (155.3 vs.80.5 mg/dL) were significantly different (P<0.05). However, baseline glucose level and diabetes were not statistically significant for tamoxifen-induced steatosis (P>0.05). BMI was the only independent risk factor of tamoxifen-induced steatosis (Hazard ratio: 1.227, 95% confidence interval: 1.039-1.448; P=0.016). Furthermore, of excluded 36 patients with NASH at baseline, the levels of aminotransferase of 19 patients were normalized after tamoxifen therapy (52.8%). The normalization of aminotransferases took 108.6±66.8 days after administration of tamoxifen. Conclusions: Our study showed that tamoxifeninduced hepatotoxicity might be associated with BMI, and tamoxifen could cause hepatoprotective effect as well as liver injury.

Disclosures:

The following people have nothing to disclose: Myung Jin Oh, Heon Ju Lee, Si Hyung Lee, Sung Bum Kim, Yeoun Su Jung, Jung Woo Lee

618

  1. Top of page

Metabolic syndrome, but not IBD severity, increases risk of NAFLD severity in patients with dual diagnosis of IBD and NAFLD

Rotonya M. Carr1, Arpan A. Patel2, Caroline Kerner1,Ann Tierney3, Kimberly A. Forde1,Gary R. Lichtenstein1
1GI, University of Pennsylvania, Philadelphia, PA; 2Medicine, University of Pennsylvania, Philadelphia, PA; 3Biostatistics Analysis Center, University of Pennsylvania, Philadelphia, PA

BACKGROUND: Nonalcoholic fatty liver disease (NAFLD) is a common cause of hepatic steatosis in patients with inflammatory bowel disease (IBD). Both metabolic syndrome (MetS) and intestinal inflammation are implicated in NAFLD pathogenesis. In our hospital population of patients with dual diagnosis of NAFLD and IBD, the prevalence of MetS parallels national trends. We examined whether MetS and IBD severity increase the risk of NAFLD severity in patients with NAFLD and IBD. METHODS: A retrospective medical record review was performed for all patients entered into the electronic medical record of a tertiary care university hospital from January 1997December 2011. Patients with a dual diagnosis of IBD and NAFLD were included. We excluded patients with viral, alcoholic, autoimmune or genetic etiology of hepatic steatosis. Patients were grouped according to “MetS” (>3 of the following: hypertension, hyperlipidemia, BMI>30, and diabetes/insulin resistance) or “non-MetS. ” BARD score or liver pathology was used to grade NAFLD severity. Physician global assessment and Montreal classification were used to determine IBD severity/pattern. Patient demographics, medications, and serology were analyzed. Statistical analysis was performed using Fisher Exact test or Mann-Whitney U test. RESULTS: 84 pts were included in our analysis (25 UC, 59 Crohn's). Pts were predominantly female (57%) and Caucasian (85%). The majority of UC pts had ulcerative proctitis; most Crohn's pts had ileocolonic disease. Pts were diagnosed with NAFLD a mean of 12 years after the time of IBD diagnosis.25% of pts had MetS. Pts with IBD and MetS were diagnosed with NAFLD at older age than IBD and non-MetS pts (54.8 vs 45.8, p=0.015). Mean BMI at NAFLD diagnosis was 34.4 and 28.8 kg/m2 in MetS and non-MetS pts (p<0.001).30% of pts were referred to hepatology independent of MetS diagnosis. Compared with nonMetS, MetS pts had higher ALT (57 vs 38, p=0.007); AST (56 vs 36, p=0.002) and higher prevalence of advanced fibrosis by liver pathology or BARD score (95% vs 63.9%, p=0.009). MetS pts had lower bilirubin (0.6 vs 0.8, p=0.024) and CRP (2.4 vs 30.5, p=0.097) compared with non-MetS. MetS and non-MetS pts had similar IBD medication patterns. Statin use was more common in MetS pts. TZD and Vitamin E use was rare. IBD severity did not correlate with NAFLD severity (p=0.2). CONCLUSIONS: NAFLD is increasingly recognized as a cause of hepatic steatosis in IBD pts. Unexpectedly, IBD disease severity was not associated with advanced NAFLD. MetS appears to be a risk factor for advanced liver fibrosis as in the general population and should prompt hepatology referral.

Disclosures:

Gary R. Lichtenstein - Consulting: Abbvie, Abbott, Alaven, Janssen Orthobiotech, Elan, Ferring, Millenium Pharmaceuticals, Ono Pharmaceuticals, Pfizer Pharmaceuticals, Prometheus, Salix Pharmaceuticals, Santarus, Schering - Plough, Shire, Takeda, UCB, Warner Chilcotte; Grant/Research Support: Alaven, Bristol Myers Squibb, Jansen Orthobiotech, Ferring, Hospira, Prometheus, Salix Pharmaceuticals, Shire, UCB, Warner Chilcotte

The following people have nothing to disclose: Rotonya M. Carr, Arpan A. Patel, Caroline Kerner, Ann Tierney, Kimberly A. Forde

619

  1. Top of page

Nonalcoholic steatohepatitis (NASH) in adults with features of metabolic syndrome (MS)

Victor Aguilor-Urbano1 5, Teresa Pereda2, Juana Gonzalo-Marin1 , 5, Pedro Moreno-Mejίas1, Julio Bercedo4, Jose Verdugo4, Francisco Moyo-Donoso4, Francisco Fernández-Cono1, Francisco RivasRuiz3, Norberto Gándara-Adán4, J. M. Navarro1 , 3
1Service of Gastroenterology, Health Agency Costa del Sol Marbella, Malaga, Spain; 2Integrated Area Clinical Laboratories, Pathology Unit, Health Agency Costa del Sol Marbella, Malaga, Spain; 3Clinical Research Unit, Health Agency Costa del Sol Marbella, Malaga, Spain; 4Service of General Surgery and Digestive, Health Agency Costa del Sol Marbella, Malaga, Spain; 5Service of Gastroenterology, Hospital Quiron Marbella, Malaga, Spain

NASH is hepatic expression of the MS. Prognosis is unknown because the liver biopsy (gold standard for diagnosis), is done in rare cases. The presentation of features MS is common and in this the prevalence and severity is unknown. OBJECTIVES: Determine prevalence of NASH histopathological criteria in adult >40 years with features MS without previous known or suspected liver disease. Describe what features MS are associated with increased risk of NASH. Determine what parameters increased liver damage. METHODS: Adults >40 years with some features of the MS (hypertension, dyslipidemia, diabetes mellitus, obesity, hyperuricemia), which were to undergo a scheduled abdominal surgery. We excluded patients with known previous liver disease, use of hepatotoxic drugs or alcohoi. NASH score was defined according to the NASH-CIinicalResearch-Network, classifying in: NASH (definite and borderline NASH) and Non-NASH. RESULTS: We included 75 patients, between 40 - 80 years, 33 males (44%). MS traits that presented were: hypertension 61.3%, dyslipidemia 40%, diabetes 22.7%, obesity 62.7% and 14.7% hyperuricemia. They presented a single trait of MS 38.7%, 28% two, three 28%, four 2.7% and five features 2.7%. Non-NASH was observed in 27 cases (36%) and NASH in 48 (64% - borderline 21 and definite 27). In 89% the biopsy have some degree of ballooning. Regarding fibrosis in 73.33% had some degree of fibrosis being 60% > grado1C, and 3 patients had cirrhosis. The fibrosis is associated with the number of features MS (p <0.05). Transaminase were lower in NASH (p <0, 05). NASH was more common in younger cases and sooner after onset of obesity (p <0, 05). Predictive of NASH were dyslipidemia (odds ratio 5.30) and age (odds ratio 0.950). Regarding fibrosis in 73.33% had some degree of fibrosis being 60% > grado1C, and 3 patients had cirrhosis. The fibrosis is associated with the number of features MS (p <0.05). CONCLUSIONS: NASH is common in adults with features of MS, and the risk increases with the number of traits that are suffering. Dyslipidemia is the only feature of MS predictor of NASH. Age factor has been a protector of NASH in our population. In addition, fibrosis is a common finding. Therefore, we suspect NASH in any adult with MS traits even when no prior evidence of liver.

Tabla 1. -differences between NASH and non-NASH

Non-NASHNASHP
Age (years)69/ 1760, 50/ 15, 500, 02
Hypertension17 (37, 0)29 (63, 0)1
Dyslipemia5 (16, 7)25 (83, 3)0, 01
Diabetes6 (35, 3)11(64, 7)1
Obesity17 (36, 2)30 (63, 8)1
Number of features MS One Two Three Four Five13 (44, 8) 8 (38, 1)6 (28, 6) 0 (0, 0) 0 (0, 0)16(55, 2) 13(61, 9) 15(71, 4) 2 (100, 0) 2 (100, 0)0, 07

Disclosure:

The following people have nothing to disclose: Victor Aguilar-Urbano, Teresa Pereda, Juana Gonzalo-Marfn, Pedro Moreno-Mejlas, Julio Bercedo, Jose Verdugo, Francisco Moya-Donoso, Francisco Femandez-Cano, Francisco Rivas-Ruiz, Norberto Gandara-Adan, J. M. Navarro

620

  1. Top of page

The association of insomnia with gastroesophage reflux symptoms in biopsy-proven nonalcoholic fa liver disease

Hiroyoshi Taketani1, Yoshio Sumida2, Saiyu Tanaka1, Kazuyuki Konemoso1, Masato Yonedo3, Kento Imajo3, Atsushi Nakajima3, Hideyuki Hyogo4, Kazuaki Chayama4, Masafumi Ono5, Toshiji Soiboro5, Hideki Fujii6, Yuichiro Eguchi7, Yoshito Itoh2
1 Center for Digestive ond Liver Diseases, Nara City Hospital, Nara, Japan; 2Department of Gastroenterology and Hepatology, Kyoto Prefectural University of Medicine, Kyoto, Japan; 3Division of Gastroenterology, Yokohama City University Graduate School of Medicine, Yokohama, Japan; 4Departmern of Medicine ond Molecular Science, Graduate School of Biomedical Sciences, Hiroshima University' Hiroshima, Japan; 5Department of Gastroenterology and Hepatology, Kochi Medical School, Kochi, Japan; 6Department of Hepatology, Graduate School of Medicine, Osaka City University, Osaka, Japan; 7Department of Internal Medicine' Saga Medicoi School, Saga, Japan

BACKGROUND/AIMS: Nonalcoholic fatty liver disease (NAFLD) is associated with a higher prevalence of insomnia and gastroesophageal reflux disease (GERD). NAFLD encompasses nonalcoholic fatty liver (NAFL) and nonalcoholic steatohepatitis (NASH). The association of insomnia with GERD in NAFLD remains unknown. The present study investigated the relationships between GERD symptoms and insomnia in subjects with biopsy-proven NAFLD. Methods: One hundred twenty three patients with biopsy-proven NAFLD (median age: 59) were enrolled in this study. Insomnia was assessed by the Athens Insomnia Scale (AIS), a self-assessment psychometric instrument designed for quantifying sleep difficulty based on the ICD-10 criteria. Patients with AIS of more than 6 were considered as insomniacs. GERD symptoms were evaluated by using a frequency scale for the symptoms of GERD (FSSG). Patients with FSSG scores of more than 8 were considered as positive. Logistic regression models were used to evaluate the association of insomnia with GERD, after adjusting for potential confounders. Thirteen NAFLD patients with GERD symptoms were administrated a proton pump inhibitor (PPI), rabeprazole (RPZ) (10mg/day) for 12 weeks to investigate the effect on insomnia. Results: Overall, 62% were female, and 71% were obese. The prevalence of NAFLD with AIS> 6 and FSSG score > 8 was 28% and 25%, respectively. Liver biopsy revealed 40 NAFL and 83 NASH. There were no differences between the two groups in FSSG and AIS. Overall, AIS was positively correlated only with FSSG among clinical parameters. AIS did not correlate with histological steatosis, grade, and fibrosis. The levels of yGT, HOMA-IR, and FSSG were significantly higher in insomniacs compared to noninsomniacs. GERD symptoms were more prevalent in insomniacs (56%) than in noninsomniacs (13%, p<0.0001). Logistic regression model demonstrated that FSSG (odds ratio [OR] 1.232, 95% confidence interval [CI]: 1.1221.352, p<0.0001) and 丫GT (OR: 1.011, 95%CI: 1.0031.019, p=0.00o) were independently associated with insomniacs. In treated-patients, AIS was significantly decreased with significant reductions of FSSG after the treatment with RPZ. Four patients of eleven insomnias (44%) were relieved after RPZ. Conclusions: In biopsy-proven NAFLD patients, insomnia was found in nearly thirty percent of cases, related to ۷GT and GERD symptoms, and can be relieved by RPZ.

Disclosures:

Kazuaki Chayama - Consulting: Abbvie; Grant/Research Support: Dainippon Sumitomo, Chugai, Mitsubishi Īanabe, DAIICHI SANKYO, Toray, BMS, MSD; Speaking and Teaching: Chugai, Mitsubishi Īanabe, DAIIcHi SANKYO, KYORIN, Nihon Medi-Physics, BMS, Dainippon Sumitomo, MSD, ASKA, Astellas, AstraZeneca, Eisai, Olympus, GlaxoSmithKline, ZERIA, Bayer, Minophagen, JANSSEN, JIMRO, TSUMURA, Otsuka, Taiho, Nippon Kayaku, Nippon Shinyaku, Takeda, AJINOMOTO, Meiji Seika, Toray

The following people have nothing to disclose: Hiroyoshi Taketani, Yoshio Sumida, Saiyu Tanaka, Kazuyuki Kanemasa, Masato Yoneda, Kento Imajo, Atsushi Nakajima, Hideyuki Hyogo, Masafumi Ono, Toshiji Saibara, Hideki Fujii, Yuichiro Eguchi, Yoshito Itoh

621

  1. Top of page

Glycemic variability is an independent predictive factor for development of h hepatic fibrosis in nonalcoholic fatty liver disease

Tsunehiro Ochi1,Motoi Hashiba1,Masafumi Ono1, Hideyuki Hyogo2, Yukio Ikedo3, Kensuke Munekoge1, Nobufo Okomofo1, Shinji Iwasaki1, Kazuaki Chayama2, Yuichiro Eguchi4, Toshiji Saibara1
1Department of Gastroenterology ond Hepatology, Kochi Medical School, Nankoku, Japan; 2Department of Medicine ond Molecular Science,Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, Japan; 3Diabetes Center, Kochi Memorial Hospital, Nangoku, Japan; 4Liver Disease Control ond Assistance Task Force, Saga Medical School, Saga, Japan

Background & Aims: Patients with nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH) often have metabolic disorders including insulin resistance and type diabetes mellitus (T2DM). We clarified the predictive factors in glucose metabolism for progression of hepatic fibrosis in patients with NAFLD by the 75-g oral glucose tolerance test (75gOGTT) and a continuous glucose monitoring system (CGMS). Methods: One hundred sixty-nine patients (68 female and 101 male patients) with biopsy-proven NAFLD with performance with 75gOGTT were enrolled and divided into four groups according to the stage of hepatic fibrosis (F0-3). Results: The proportion of patients with T2DM significantly gradually increased, HbA1c and the homeostasis model assessment of insulin resistance were significantly elevated, and 1, 5-anhydroglucitol (1, 5-AG) was remarkably decreased with the progression of fibrosis. In the 75gOGTT, both plasma glucose and insulin secretion were remarkably increased with the progression of fibrosis. The only factor significantly associated with advanced fibrosis was 1, 5-AG (P = 0.008) as determined by multivariate logistic regression analysis. We next evaluated the changes in blood glucose during 24 hours by monitoring with the CGMS to confirm the relationship between glycemic variability and progression of fibrosis. Variability of median glucose, standard deviation of median glucose (P = 0.0022), maximum blood glucose (P = 0.0019), and AMin-max blood glucose (P = 0.0029) were remarkably higher in severe fibrosis than in mild fibrosis. Conclusions: Hyperinsulinemia and hyperglycemia, especially glycemic variability, are important predictive factors in glucose impairment for the progression of hepatic fibrosis in NAFLD.

Disclosures:

Kazuaki Chayama - Consulting: Abbvie; Grant/Research Support: Dainippon Sumitomo, Chugai, Mitsubishi Īanabe, DAIICHI SANKYO, Toray, BMS, MSD; Speaking and Teaching: Chugai, Mitsubishi Īanabe, DAIIcHl SANKYO, KYORIN, Nihon Medi-Physics, BMS, Dainippon Sumitomo, MSD, ASKA, Astellas, AstraZeneca, Eisai, Olympus, GlaxoSmithKline, ZERIA, Bayer, Minophagen,

Shinyaku, Takeda, AJINOMOTO, Meiji Seika, Toray

The following people have nothing to disclose: Tsunehiro Ochi, Motoi Hashiba, Masafumi Ono, Hideyuki Hyogo, Yukio Ikeda, Kensuke Munekage, Nobuto Okamoto, Shinji Iwasaki, Yuichiro Eguchi, Toshiji Saibara

622

  1. Top of page

Association between Gallstone Disease and Fatty Liver in a Large Apparently Healthy Population

Donghee Kim1, Min-Sun Kwak1, Goh Eun Chung1, Won Kim3' Yoon Jun Kim2, Jung-Hwon Yoon3
1Department of Internal Medicine, Healthcare Research Institute, Seoul National University Hospital Gangnam Center, Seoul, Republic of Korea; 2Department of Internal Medicine, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Republic of Korea; 3Department of Internal Medicine, Liver Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea

Background/Aims: Gallstone disease and fatty liver are both prevalent diseases in the general populations and share the same risk factors such as obesity and insulin resistance. However, association between gallstone disease and ultrasonographically diagnosed fatty liver has not been completely established. The aim of this study was to characterize the relationship between gallstone disease and fatty liver in large population. Methods: A cross-sectional study with 24, 050 health check-up subjects was conducted. Gallstone disease was defined as the presence of gallstones on abdominal sonography or previous history of cholecystectomy. Fatty liver was diagnosed on the basis of typical ultrasonographic findings. Subjects positive for hepatitis B or C virus or with a history of other forms of hepatitis were excluded. Results: The mean age of the subjects was 48.7 ± 11.1 years and 54.5% were male. The prevalence of gallstone disease was 5.3% (n=1, 280). The prevalence of fatty liver increased with presence of gallstone disease (43.0% vs.31.3%, p <0.001). In the same manner, the prevalence of gallstone disease increased with presence of fatty liver (7.2% vs.4.4%, p <0.001). The gallstone disease was significantly associated with fatty liver after adjusted for age and sex [odds ratio (OR) 1.50 95% confidence interval (Cl) 1.331.69]. Multivariate regression analysis after adjustment for body mass index, waist circumference, total cholesterol, triglycerides, HDL cholesterol, HbA1 c, and systolic blood pressure showed that gallstone disease was statistically significantly associated with fatty liver (OR 1.23, 95% CI 1.06-1.42, p=0.007). These association was attenuated, however still statistically significant after adjusting for insulin resistance (OR 1.27 95% Cl 1.04-1.55, p=0.018). Conclusions: Patients with fatty liver have a high prevalence of gallstone disease. Gallstone disease is associated with fatty liver independently of known metabolic risk factors, especially insulin resistance.

Disclosures:

The following people have nothing to disclose: Donghee Kim, Min-Sun Kwak, Goh Eun Chung, Won Kim, Yoon Jun Kim, Jung-Hwan Yoon

623

  1. Top of page

Inverse association between hepatitis B virus infection and fatty liver disease: A large-scale study in populations seeking for check-up

Chien-Wei Su1,2 Yuan-Jen Wang3, 2, Jaw-Ching Wu4, 5, Teh-Ia Huo1, 6, Yi-Hsiang Huang1, 4, Han-Chieh Lin1,2 Fo-Youh Lee1,2 Shou-Dong Lee2, 7
1 Division of Gastroenterology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan; 2Faculty of Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan; 3Division of Healthcare and Services, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan; 4Institute of Clinical Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan; 5Department of Medical Research and Education, Taipei Veterans General Hospital, Taipei, Taiwan; 6Department and Institute of Pharmacology, National Yang-Ming University, Taipei, Taiwan; 7Division of Gastroenterology, Department of Medicine, Cheng Hsin General Hospital, Taipei, Taiwan

Background: Although many studies have attempted to clarify the association between hepatitis B virus (HBV) infection and fatty liver disease, no prior studies have emphasized the relationship of HBV and fatty liver regarding different demographics of age and body mass index (BMI). Aim: To investigate the correlation of HBV and fatty liver in the different demographics of age and BMI. Methods: We enrolled consecutive subjects who had received health check-up services at the Taipei Veterans General Hospital from 2002 to 2009 and ultrasonography was used to diagnose fatty liver according to the practice guidelines of the American Gastroenterological Association. Results: Among the 33, 439 subjects enrolled in this study, fatty liver was diagnosed in 43.9% of the population and 38.9% of patients with chronic HBV infection. Multivariate analysis showed that BMI, age, waist circumference, systolic blood pressure, fasting glucose, cholesterol, alanine aminotransferase (ALT) levels, and platelet counts were positively associated, while hepatitis B surface antigen (HBsAg) positivity was inversely associated with fatty liver, especially for subjects with BMI> 22.4 kg/m2 and age>50 years. On the contrary, HBV infection was positively correlated with the presence of elevated serum ALT levels in subjects with fatty liver disease regardless of their age and BMI. Conclusions: Metabolic factors are important determinants for the prevalence of fatty liver. Patients with HBV infection were inversely associated with fatty liver disease than the general population, especially in older and obese patients. Furthermore, metabolic factors and HBV infection were associated with elevated serum ALT levels in fatty liver disease.

Age distribution of fatty liver stratified by HBV status.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Chien-Wei Su, Yuan-Jen Wang, Jaw-Ching Wu, Teh-la Huo, Yi-Hsiang Huang, Han-Chieh Lin, Fa-Yauh Lee, ShouDong Lee

624

  1. Top of page

PNPLA3 GG Genotype and Carotid Atherosclerosis in Patients with Non-alcoholic Fatty Liver Disease

Salvatore Petto1, Luca Valenti2, Giulio Marchesini3, Vito Di Marco1, Anna Licata1, Calogero Camma1, Maria Rosa Borcellono1, Daniela Cabibi4, Benedetta Donati2, Anna Ludovica Fracanzani2, Stefania Grimaudo1, Gaspare Parrinello5, Rosario Maria Pipitone1, Daniele Torres5, Silvia Fargion2, Giuseppe Licata5, Antonio Croxi1
1Sezione di Gastroenterologia' Di. Bi. M. I. S/ Universifa di Palermo, Italia, Palermo, Italy; 2Department of Pathophysiology and Transplantation, Section Internal Medicine, Universitá degli Sfudi/ Fondazione Cá, Granda IRCCS Ospedole Maggiore Policlinico, Milano, Italy, Milano, Italy; 3Dipartimento Biomedico di Medici no Inferno e Speciolisfico (Di. Bi. M. I. S)/ Universifa di Palermo, Italia, Palermo, Italy; 4Dipartimento di Patologia umana, Universifa di Palermo, Italia, Palermo,Italy; 5Diportimento di Medicino e Gastroenterologia' “Alma Mater Studiorum'” Universifa di Bologna, Italy, Bologna, Italy

Background and aim: To test the association of carotid atherosclerosis with gene variants influencing hepatic fat accumulation and the severity of liver damage in patients with NAFLD. Methods: We assessed anthropometric, metabolic and histological data(Kleiner score) in 162 consecutive, biopsy-proven Sicilian NAFLD patients. Intima-media thickness(IMT), IMT thickening(IMT>1 mm) and carotid plaques(focal thickening of >1.3mm at the level of common carotid artery) were evaluated using ultrasonography. IL28B rs12979860 C>T, PNPLA3 rs738409 C>G, GCKR rs780094 C>T, LYPLAL1 rs12137855 C>T, and NCAN rs2228603 C>T single nucleotide polymorphisms were also assessed. The results were validated in a cohort of 267 subjects with clinical or histological diagnosis of NAFLD from Northern Italy, 63 of whom had follow-up examinations. Results: Carotid plaques, IMT thickening and mean maximum IMT were similar in the two cohorts, whereas the prevalence of diabetes, obesity, NASH, and PNPLA3 GG polymorphism(21 %vs.1 3%, p=0.02) were significantly higher in the Sicilian cohort. In this cohort, the prevalence of carotid plaques and IMT thickening was higher in PNPLA3 GG compared to CC/CG genotype(53%vs.32%, p=0.02; 62%vs.28%, p<0.001, respectively), and these associations were confirmed at multivariate analyses(OR2.94; 95%C. I.1.12-7.71, p=0.02, and OR4.11; 95%C. I.1.69-9.96, p=0.002, respectively), but were only observed in patients <50years. Also in the validation cohort, PNPLA3 GG genotype was independently associated with iMī thickening in younger patients only(OR6.00, 95%C. I. 1.36-29, p=0.01), and to IMT progression(p=0.05) in patients with follow-up examinations. Conclusion: PNPLA3 GG genotype is associated with higher severity of carotid atherosclerosis in younger patients with NAFLD. Mechanisms underlying this association, and its clinical relevance need further investigations.

Disclosures:

Giulio Marchesini - Advisory Committees or Review Panels: Sanofi-Synthelabo; Grant/Research Support: Merck Sharp & Dome; Speaking and Teaching: Novo Nordisk, Merck Sharp & Dome, Boerhinger Ingelheim, Lilly

The following people have nothing to disclose: Salvatore Petta, Luca Valenti, Vito Di Marco, Anna Licata, Calogero Camma, Maria Rosa Barcellona, Daniela Cabibi, Benedetta Donati, Anna Ludovica Fracanzani, Stefania Grimaudo, Gaspare Parrinello, Rosaria Maria Pipitone, Daniele Torres, Silvia Fargion, Giuseppe Licata, Antonio Craxi

625

  1. Top of page

Normal controlled attenuation parameter values: A prospective study of healthy subjects undergoing health check-ups and living liver donors in South Korea

Younq Eun Chon1, Kwang Joon Kim2, Seung Up Kim1, Dong Jin Joo3, Beom Kyung Kim1, Jun Yong Park1, Do Young Kim1, Sang Hoon Ahn1, Kwang-Hyub Han1; 1Department of Internal Medicine, Institue of Gastroenterology Yonsei university college of medicine, Seoul, Republic of Korea; 2Institue of Endocrinology, Yonsei university college of medicine, Seoul, Republic of Korea; 3Department of Surgery, Yonsei university college of medicine, Seoul, Republic of Korea

Background/Aims: The controlled attenuation parameter (CAP) is a noninvasive method of measuring hepatic steatosis. We aimed to define the normal range of CAP values and evaluate factors influencing these values in healthy subjects. Methods: CAP values were measured in a cohort of healthy subjects who were screened for service as living liver transplantation donors and underwent health check-ups. Subjects with chronic liver disease, abnormalities on liver-related laboratory tests, or fatty liver on ultrasonography or biopsy were excluded. Results: The mean age of the 264 recruited subjects (131 men and 133 women; 76 potential liver donors and 188 subjects who had undergone health check-ups) was 49.2 years. The mean CAP value was 224.8 ± 38.7 dB/m (range, 100.0-308.0 dB/m), and the range of normal CAP values from the 5th to 95th percentile was 156.0-287.8 dB/m. The mean CAP value was significantly higher in subjects who had undergone health check-ups than in potential liver donors (227.5 ± 42.0 vs.218.2 ± 28.3 dB/m, P = 0.040). CAP values did not differ significantly according to sex or age in potential liver donors or subjects who had undergone health check-ups (all P > 0.05). In a multivariate linear regression analysis, body mass index (p = 0.271, P = 0.024) and triglyceride levels (p = 0.348, P = 0.008) were independent factors influencing CAP values. Conclusion: We defined the normal range of CAP values and found that body mass index and triglyceride levels can influence CAP values among healthy subjects.

Disclosures:

The following people have nothing to disclose: Young Eun Chon, Kwang Joon Kim, Seung Up Kim, Dong Jin Joo, Beom Kyung Kim, Jun Yong Park, Do Young Kim, Sang Hoon Ahn, Kwang-Hyub Han

Thumbnail image of

626

  1. Top of page

Hyaline Arteriosclerosis: A Diabetic Complication of the Liver

Maya Balakrishnan1, Guadalupe Garcia-Tsao2, Yannonq Deng3, Maria Ciarleglio3, Dhanpat Jain4
1Digestive Diseases Section, Yale University School of Medicine, New Haven, CT; 2Digestive Diseases Section, VA-CT Healthcare System, West Haven, CT; 3Yale Center for Analytical Sciences, Yale University School of Public Health, New Haven, CT; 々Department of Surgical Pathology Yale University School of Public Health, New Haven, CT

Background&Aim: Hyaline arteriosclerosis (microangiopathy) is a well-known complication of diabetes. Diabetic nephropathy and retinopathy are two well-established manifestations of small vessel hyaline arteriosclerosis in diabetes. Based on our observation of isolated cases of patients with diabetes and liver enzyme abnormalities in whom the only finding on liver biopsy was hyaline arteriosclerosis, we decided to undertake a crosssectional blinded study assessing this and other histological findings in diabetics. While nonalcoholic steatohepatitis is a well-recognized hepatic complication of diabetes, microangiopathy of the liver has never previously been reported in diabetic patients. The aim of this study was to evaluate the association between hyaline arteriosclerosis and diabetes. Methods: Liver biopsy findings from 62 patients with diabetes that met inclusion criteria between January 2006 and December 2009 were compared to those of 62 patients without diabetes matched by age and gender. Patients with cirrhosis, liver mass, right heart failure, significant alcohol use, or insufficient available clinical information were excluded. Medical records were reviewed for the presence of diabetes, BMI, diabetes treatment, and comorbidities at time of biopsy (e. g. underlying liver disease, hypertension, dyslipidemia). An experienced pathologist (DJ) blinded to all clinical data (including presence or absence of diabetes) reviewed all biopsies. Results: Diabetic and control groups had the same average age (50y) and proportion of females (42%). Prevalence of hepatitis C was not different between the groups (68% vs.53%, p=0.12). Diabetics had a higher average BMI (35m/kg2 vs.29m/kg2, p=0.001), prevalence of hypertension (76% vs.34%, p <0.001) and prevalence of dyslipidemia (47% vs.23%, p=0.0025). Among diabetics, 87% had type 2 diabetes and 48% used insulin. Histologically, prevalence of steatosis (23% vs.15%, p=0.25) and steatohepatitis (34% vs.29%, p=0.52) were not different between diabetics and controls. Hyaline arteriosclerosis was significantly more prevalent among diabetics compared with controls: 47% vs.26% (p=0.012). A subgroup analysis among diabetics showed that age was significantly greater in patients with hyaline arteriosclerosis (54y vs.46y, p=0.021). Sex, BMI, insulin use, hypertension, and dyslipidemia were not associated with hyaline arteriosclerosis among diabetics. Conclusions: Hyaline arteriosclerosis of hepatic arterioles is a small vessel hepatic complication of diabetes described for the first time. The clinical and prognostic implications of this finding, particularly regarding liver injury, remain to be investigated.

Disclosures:

The following people have nothing to disclose: Maya Balakrishnan, Guadalupe Garcia-Tsao, Yanhong Deng, Maria Ciarleglio, Dhanpat Jain

627

  1. Top of page

Baseline liver enzymes and incident diabetes mellitus and vascular events: first results from a large real-life population study

Antonio De Vincentis, Umberto Vespasiani Gentilucci, Gaetano Pic-cinocchi, Roberto Piccinocchi, Giovanni Galati, Paolo Gallo, Chiara Dell'Unto, Antonio Picardi;
Clinical Medicine and Hepatol-ogy, Campus Bio-Medico University, Roma, Italy

Background and aims. Many large population-based studies have shown the association between baseline levels of liver enzymes, mainly ALT and GGT, and the medium-term incidence of diabetes mellitus (DM) and vascular events. Nonetheless, the predictive role of liver enzymes has never been confirmed in a real-life context, where patients are not tested per-protocol and results are obtained from labs with different analyzers. We aimed to verify the association between baseline ALT, GGT and AST/ALT ratio, the latter as a proxy of liver disease evolution, and the incidence of DM, stroke and coronary heart disease (CHD), in a large real-life population. Patients and Methods. Subjects who underwent routine blood tests including AST, ALT and GGT between 2000 and 2005 were extracted from a validated software employed by 120 general practitioners in the area of Naples (Italy), in charge of about 170.000 subjects. Incident DM, stroke and CHD were registered after a median follow-up time of 102 months (8.5 years). After exclusion criteria (known liver disease, HBsAg+, HCVAb+, age<20), data from 16.689 subjects were analyzed. Results. Mean age of the study population was 62.3 +/- 17.7, male/female 43.8/56.2%. Cumulative incident DM, stroke and CHD were respectively 5.1%, 1.2% and 4.6%. In multivariate-adjusted analysis, ALT was associated with incident DM (OR 1.17; CI 1.06-1.29; p=0.002), but not with stroke and CHD. GGT was associated with incident DM (OR 1.32; CI 1.19-1.46; p<0.001), and stroke (OR 1.25; CI 1.05-1.49; p=0.009), but not with CHD, while AST/ALT ratio was not associated with any outcome. DM was diagnosed in 3.2%, 5.2% and 6.9% of subjects with baseline GGT in the lower, medium and upper tertile, respectively (p=0.02). Conclusion. Except for GGT and incident stroke, our study, the first carried out in a real-life setting, does not support an association between baseline liver enzymes and the occurrence of vascular events, while confirms an independent predictive role of both ALT and GGT levels for incident DM. These results add to the accumulating evidence that the liver is a strong contributor to insulin resistance rather than a simple target of dysmetabolism.

Disclosures:

Antonio Picardi - Grant/Research Support: Schering-Plough, Roche, Schering-Plough, Roche, Schering-Plough, Roche, Schering-Plough, Roche; Speaking and Teaching: Bayer, Bayer, Bayer, Bayer

The following people have nothing to disclose: Antonio De Vincentis, Umberto Vespasiani Gentilucci, Gaetano Piccinocchi, Roberto Piccinocchi, Giovanni Galati, Paolo Gallo, Chiara Dell'Unto

628

  1. Top of page

Blood polychlorinated biphenyls congeners are associated with nonalcoholic steatohepatitis

Vasily Isakov1, Vladimir Bessonov2, Elena Khromchenkova2, Natalia Topilskaya1, Ksenia Selezneva1, Victor Tutelyan3;
1Department of Gastroenterology & Hepatology, Institute of Nutrition, Moscow, Russian Federation; 2Laboratory of food chemistry, Institute of Nutrition, Moscow, Russian Federation; 3Laboratory of nutritional enzymology, Institute of Nutrition, Moscow, Russian Federation

Polychlorinated biphenyls (PCBs) are organic environmental pollutants. In experimental studies, addition of PCB congener 153 to high-fat diet produced liver inflammation distinctive for nonalcoholic steatohepatitis (NASH), however no data were published concerning the distribution of PCB congeners in the blood of healthy controls and patients with NASH. Methods: Eighteen NASH patients were selected using US liver pattern correspondent to fatty liver, blood chemistry results (chronic elevation of ALT > 2N), absence of chronic intake of any medication and alcohol. Age and sex matched control group (18 volunteers) was also included into the study. PCBs were determined in blood samples by capillary gas chromatography with the electron capture detector (column SE-54, internal standard PCB119). Identification of individual congeners was carried out by a relative retention times, quantitative calculations were performed using relative response factors. Results: Mean serum total PCBs concentration was significantly higher in NASH patients in compare to healthy controls (1,46 ± 1,39 vs 0,66 ± 0,29 ng/g, p=0,02). There were no differences between groups in serum concentrations of congener 118 (0,22 ±0,14 vs 0,20 ±0,16 ng/g p=0,2). Congeners 183 and 185 were found only in 41% of controls, but not in patients with NASH. Mean serum concentration of congeners 99, 101, 138 and 153 were lower in control in compare to patients with NASH (0,12 ± 0.05 vs 0.06 ± 0.04, p=0.007; 0.72 ± 0.57 vs 0.05 ± 0.07, p=0.0002; 0.48 ± 0.78 vs 0.08 ± 0.06, p=0.04; 0.33 ± 0.33 vs 0.06 ± 0.07, p=0.007). In conclusion, total serum PCBs concentration and concentration of congeners 99, 101, 138 and 153 were lower in healthy control than in patients with NASH. Congeners 183 and 185 were found only in healthy controls. It means that NASH patients in compare to healthy controls experienced long-term exposure for toxic lipophilic environmental pollutants, which can be additional factors facilitating development and progression of NAFLD. Further studies are needed to clarify the role of different congeners and its transporters in liver for development of the disease.

Disclosures:

Vasily Isakov - Advisory Committees or Review Panels: Abbvie, Bristol-Myers Squibb, Gilead, Janssen, Merck, Vertex; Consulting: Bristol-Myers Squibb, Merck; Speaking and Teaching: Bristol-Myers Squibb, Janssen, Merck

The following people have nothing to disclose: Vladimir Bessonov, Elena Khromchenkova, Natalia Topilskaya, Ksenia Selezneva, Victor Tutelyan

629

  1. Top of page

AgRP and FASN Expression In Vitro and In Vivo Implies Concomitant Lipophagy and Fatty Acid Synthesis May Play a Role in the Pathogenesis of Non-Alcoholic Steatohepatitis (NASH)

J. Michael Estep1, David Van Natta1,2, Thomas Jeffers1, Alyssa C. Hosey1, Zobair M. Younossi1,3;
1Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA; 2Cen-ter for the Study of Chronic Metabolic Diseases,, George Mason University, Fairfax, VA; 3Center for Liver Diseases, Department of Medicine, Inova Fairfax Hospital, Falls Church, VA

Background: AgRP is an orexigenic peptide directly regulated by fatty acid uptake; FASN is involved in fatty acid synthesis and its expression is sensitive to glucose concentration. Aim: The aim of this pilot study is to measure the expression of AgRP and FASN in vitro under elevated glucose and lipid concentrations and compare findings to a set of NAFLD patients. Methods: HepG2 cells were challenged with 10mM of oleic acid (OA), 50mM glucose (GLU), or both and expression of AgRP and FASN was measured by qPCR and compared to untreated controls. Visceral adipose tissue was collected and flash frozen for preconcented patients with biopsy-proven NAFLD and AgRP and FASN gene expression was measured by qPCR. Additionally, expression of both genes was measured in formalin fixed paraffin embedded hepatic tissue from a subset of patients. Results: In three separate experiments using HepG2 cells, GLU treatment decreased AgRP expression, while OA and combination treatment resulted in an increase (FD= -2.2, 4.3, 1.6, respectively). FASN expression increased with glucose, decreased in OA, but remained neutral to the control in combination treatment (FD = 2, -2.6, 1). Visceral adipose from 46 NAFLD patients (NASH = 26, and non-NASH NAFLD=21) was tested for expression AgRP and FASN. AgRP showed a significant decrease in patients with NASH as compared to Non-NASH NAFLD (FD -4.9, P=0.02) while FASN showed no significant change. Additionally AgRP expression showed a modest but significant correlation with presence of histologic NASH (r= -0.38, P<0.01). In the hepatic tissue (N=10), the expression of AgRP, and FASN tended to show an increase in patients with NASH, although only FASN was statistically significant (AgRP FD 1.8, P=0.1, FASN FD=3.09, P=0.05). Additionally hepatic FASN expression shows moderate but significant correlation with presence of NASH (r=0.66, P=0.03). Conclusion: These findings are consistent with a model of NASH pathogenesis in which both lipogenesis and lipophagy are concomitant. Significant decrease in AgRP production in the visceral adipose implies a decrease in adipose lipophagy in NASH patients. More research is necessary to confirm these hypotheses.

Disclosures:

Zobair M. Younossi - Advisory Committees or Review Panels: Merck, Vertex, Tibotec/J and J; Consulting: Gilead Sciences

The following people have nothing to disclose: J. Michael Estep, David Van Natta, Thomas Jeffers, Alyssa C. Hosey

630

  1. Top of page

Controlled Attenuation Parameter for detection and stadification of hepatic steatosis in alcoholic and non alcoholic fatty liver disease

Antonia Lepida1, Francesco Puleo1, Delphine Degre1, Laurine Ver-set2, Pieter Demetter2, Thierry Gustot1, Massimo Bocci1, Jonas Schreiber1, Michael Adler1, Eric Trépo1, Christophe Moreno1;
1 Hepatopancreatology and digestive Oncology, Erasme Hospital, Brussels, Belgium; 2Pathology, Erasme Hospital, Brussels, Belgium

Aim: To determine the accuracy of Controlled Attenuation Parameter (CAP), a new non-invasive tool for the evaluation of liver fat content in an alcoholic (ALD) and non-alcoholic fatty liver disease (NAFLD) population and to identify specific cut-offs which predict the severity of steatosis. Methods: 78 consecutive ALD or NAFLD patients candidate for a liver biopsy, were also evaluated for the amount of steatosis with CAP. The time interval between the liver biopsy and the CAP measurement was less than 3 months. Patients with other cause of liver disease were excluded from the study. The percentage of steatosis among total hepatocytes was assessed histologically as follow: S0: <5%, S1: 5-33%, S2: 34-66%, S3: 67-100%. The Fatty Liver Index (FLI), a composite serum marker of steatosis was also calculated. Areas under receiver operating characteristic curves (AUROC) were used to evaluate performance of CAP for diagnosing steatosis compared with histology. Results: Characteristics of the patients included were: median age 51 years, median BMI 27 kg/m2, ALD 49%, NAFLD 36%, mixed aetiology 15%. The prevalence of steatosis was: S0 28%, S1 37%, S2 18%, S3 17%. CAP correlates significantly with the percentage of histological steatosis (p < 0.001) and tends to be associated with steatosis grade (p=0.054) and FLI (p=0.052). The median CAP values for each steatosis grade (SG) were: for S0: 235 dB m-1 (IQR: 193-266); S1: 286 dB m-1 (IQR: 234.5-349); S2: 342 dB m-1 (IQR: 274.3-363.5); and S3: 315 dB m-1 (IQR: 292.5-340). The optimal cut-offs were 262 dB m-1 for SG >S1 (AUROC 0.82, Se 75%, Sp 74%) and 302 dB m-1 for both SG >S2 (AUROC 0.76, Se 74%, Sp 77%) and S3 (AUROC 0.78, Se 77%, Sp 67%). The AUROC using FLI to detect SG >S1 was 0.67 with an optimal cut-off of 68 (Se 77%, Sp 50%), for SG >S2 it was 0.645 and for SG = S3 it was 0.66. In univariate analysis, variables associated with steatosis >5% were: CAP (p<0.001), diabetes (p=0.026) and GGT (p=0.047). In multivariate analysis only CAP (p<0.001) and GGT (p=0.047) remained significantly linked to liver fat content. Conclusions: CAP is a new non-invasive technique that can adequately predict the presence of steatosis (>5%) in a mix population of ALD and NAFLD patients and was more reliable than FLI. CAP had also a good accuracy to detect moderate steatosis (>33%). However, it failed to distinguish moderate (>33%) from severe steatosis (>66%). Further studies in independent cohorts are warranted to confirm our results.

Disclosures:

The following people have nothing to disclose: Antonia Lepida, Francesco Puleo, Delphine Degre, Laurine Verset, Pieter Demetter, Thierry Gustot, Massimo Bocci, Jonas Schreiber, Michael Adler, Eric Trépo, Christophe Moreno

631

  1. Top of page

Diagnostic value of the Sonographic Hepato-Renal Index to quantification of hepatic steatosis

José Luis Martín-Rodríguez, Juan P. Arrebola, Nicolás Olea, Javier Fernández-Mena, Jorge L. González-Calvin;
Radiology, San Cecilio Universitary Hospital, Granada, Spain

Context: Non-alcoholic fatty liver disease is the most frequent hepatic disorder in the developed world. Currently, liver biopsy and proton magnetic resonance spectroscopy (1H-MRS) are considered the gold standard methods for the quantification of liver fat deposits. Objective: To determine whether a computerized Sonographic Hepato-Renal Index (SHRI) calculated using a standard workstation, without specifically-designed software, is an adequate alternative to 1H-MRS for the quantification of fat liver content and diagnosis of steatosis in the general population. Methods: One hundred twenty-one subjects volunteers (mean age=46 yrs, range=21-77 yrs) were recruited from three medical centers in Granada, Southern Spain, among those attending to routine general checkups. All subjects were examined by ultrasound and by 1H-MRS 3T, which served as reference for the diagnosis of steatosis. The computerized SHRI was calculated as the ratio between the echogenicity of the liver and that of the right renal parenchyma. The validity of the methodology was assessed with receiver operating characteristic curves and correlation tests. Results: The quantitative SHRI showed a strong correlation (Spearman coefficient = 0.89, p< 0.001) with the 1H-MRS 3T. The optimal SHRI cut-off points for the prediction of steatosis >5%, >25%, and >50% were 1.28, 1.75, and 2.29, respectively. Cut-off points of 1.21, 1.28, and 2.15 yielded 100% sensitivity for the diagnoses of steatosis >5%, >25%, and >50%, respectively, with a specificity >70%. Conclusion: This study demonstrates that the SHRI is a valid, simple, reliable, and cost-effective screening tool for identifying, assessment and quantification of hepatic steatosis in the general population.

Validity of SHRI at different levels of steatosis

Table 1. 
Steatosis>5% (n=75)>25% (n=53)>50% (n=21)
  1. SHRI: Sonographic Hepato-Renal Index ; CI: Confidence interval

SHR1 (cut-off value)1.281.752.29
Area under curve99.10%96.60%95.90%
95% CI(98-100)%(93.9-99.3)%(92.6-99.2)%
Sensitivity94.67% (71 /75)90.57% (48 / 53)95.24% (20/21)
Specificity95.65% (44 / 46)91.18% (62/68)84% (84/100)
Positive predictive value97.26% (71/73)88.89% (48 / 54)55.56% (20/36)
Negative predictive value91.67% (44/48)92.54% (62/67)98.82% (84 / 85)
Positive likelihood ratio21.7710.265.95

Disclosures:

The following people have nothing to disclose: José Luis Martín-Rodríguez, Juan P. Arrebola, Nicolás Olea, Javier Fernández-Mena, Jorge L. González-Calvin

632

  1. Top of page

Hepatic Gene Expression Profiling and Gene Network Analyses in Patients Undergoing Bariatric Surgery Reveals Distinct Molecular Pathways in Normal Histology, Bland Steatosis, and NASH

Kiran Bambha, Jonathan A. Schoen, Kevin Rothchild, Susan C. Hartley, Linling Cheng, Lucy Golden-Mason, Ivana Yang, Hugo R. Rosen;
University of Colorado Denver, Aurora, CO

Understanding the genetics of NASH will aid in unraveling its pathogenesis. Aim: To identify gene networks and pathways in NASH. Methods: Hepatic gene expression was performed using Affymetrix Human Gene 1.0 Array in 10 women undergoing bariatric surgery (4 NASH; 4 Bland Steatosis [BS]; 2 normal liver). Expression profiles were compared with ANOVA. Genes with > 1.5-fold change between NASH vs BS; and NASH vs Normal, were analyzed by Ingenuity Pathways. Results: Mean age was 37±10 years and median BMI: 52 kg/m2[IQR, 46-58]. All were non-diabetic. Median NAS was 3 [IQR, 2-4]. Most patients had no fibrosis (70%); 30% had Stage 1. No single gene reached statistical significance after genomewide adjustment. There were 95 genes with >1.5-fold change in expression between NASH vs BS and these were analyzed with canonical pathway analysis yielding the following pathways: LXR/RXR Activation (p 0.008); PXR/RXR Activation (p 0.008); LPS/IL-1 Mediated Inhibition of RXR Function (p 0.002); Glutathione-Mediated Detoxification (p 0.009); Valine Degradation (p 0.005). In gene network analyses, the significant cellular/molecular biological functions associated with the NASH genes were: Lipid Metabolism (p 0.0001-0.04); Molecular Transport (p 0.0001-0.04); Small Molecule Biochemistry (p 0.0001-0.04); Amino Acid Metabolism (p 0.001-0.03); Cellular Development (p 0.002-0.04). The top scoring gene network in the comparison of NASH vs BS is outlined in the Table. We also identified 448 genes with >1.5-fold change in expression between NASH vs Normal liver and the top scoring gene network in this comparison is also demonstrated in the Table. Conclusion: These data reveal canonical pathways, gene networks, and biological functions associated with NASH in patients undergoing bariatric surgery, demonstrating the utility of gene pathway analyses to facilitate identification of higher priority target candidates in NASH.

Table 2. 
COMPARISON GROUPSMOLECULES IN NETWORK
Genes UpregulatedGenes DownregulatedOther Molecules
NASH vs Bland SteatosisAPOD, CXCL10, DUSP6, EGR1, FOXK2, FRZB, JUN, MGP, NR4A2, 0RM1, PTGES, SERPINE1, TRIB1ABCB11, ALOX12, APOA4, CYP7A1, DMTFldelta, HERC1, IL7R, MAOB, MAZ, RAD50, SAA1, SAA2deltaERK1/2, HDL, Hsp70, NFkB, Nrlh, RAR, RXR, Thyroid hormone receptor
NASH vs Normal LiverCXCL9, CXCL10, EPCAM, HLA-DQA1, HLA-DQB1, IGHG1, IGJ, IGK, IGKC, SCDAVPR1A, BCL6, CCL3, CCL4, EGR2, GPR133, IGF1, IGFALS, IL1B, LRG1, MYH11, MYOM1, NPBWR1, RPL13A, TSPAN8Chemokine, Gpcr, Histoneh4, MHC Class II (complex), Myosin, NFkB, RNApolymerase II.TGF beta

Disclosures:

The following people have nothing to disclose: Kiran Bambha, Jonathan A. Schoen, Kevin Rothchild, Susan C. Hartley, Linling Cheng, Lucy Golden-Mason, Ivana Yang, Hugo R. Rosen

633

  1. Top of page

Usefulness in clinical practice of NASHMRi for the detection of steatohepatitis in patients with NAFLD

Pablo Cerro-Salido2, Rocío Gallego-Durán1, María J. Pareja1, Emilio Gómez-González2, Maria Carmen Rico1, Rafael Aznar Méndez3, Sandra Macho3, Elisabetta Bugianesi4, Javier Crespo5, Maria Teresa Arias-Loste5, Javier Abad6, Susana Soto Fernandez7, Reyes Aparcero López1, Inmaculada Moreno-Herrera8, Raul J. Andrade8, Jose Luis Calleja6, Oreste Lo Iacono7, Manuel Romero-Gomez1;
1Valme University Hospital, Sevilla, Spain; 2Interdisciplinary Physics Group, University of Sevilla, Sevilla, Spain; 3Radiology Unit, Valme University Hospital, Sevilla, Spain; 4Uni-versity of Turin, Turin, Italy; 5Marqués de Valdecilla University Hospital, Santander, Spain; 6Puerta de Hierro University Hospital, Madrid, Spain; 7Tajo Hospital, Madrid, Spain; 8Virgen de la Victoria University Hospital, Málaga, Spain

AIM: To validate in clinical practice the potential of NASHMRi as a non-invasive method for the diagnosis of steatohepatitis in patients with non-alcoholic fatty liver disease (NAFLD). METHODS: Seventy-seven consecutive patients suffering from biopsy-proven NAFLD were included, mean age 51 + 14 years, 62% male, 39 patients showed steatohepatitis. One non-contrast-enhanced MRI protocol was performed using 1.5 Tesla General Electric MRI (n=41) and Philips MRI system (n=36). Optical analysis and architectural neural networks was used to define NASHMRi [Estimators E28, E42, E48, 74] as previously reported (Gallego-Durán et al. J Hepatol 2013;58:S8). RESULTS: Eighty-one patients were recruited for the study. Of them, two patients refused to participate because of claustrophobia, and two due to higher abdominal perimeter that did not fit in the gantry. NASHMRI output value was between 0 and 1. Cut off point selected was 0.50 for detecting steatohepatitis. 39/77 (51%) patients presented a NASHMRI higher than 0.50, of them 32/39 (82%) showed steatohepatitis in liver biopsy. 38/77 (49%) showed a NASHMRI output below 0.50, and 31/38 (82%) showed simple steatosis. Sensitivity of this method was 82%, specificity 82%, PPV 82%, NPV 82% and diagnostic accuracy of 82%. CONCLUSIONS: NASHMRi showed a high potential as a steatohepatitis predictor. It is a safe method, independent of the MR manufacturer, uses MRI protocols applied in clinical practice and explores the whole liver, and does not need to be supplemented with other non-invasive diagnostic method to accurately predict steatohepatitis. ACKNOWLEDGEMENTS: “The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° HEALTH-F2-2009-241762 for the project FLIP.”

Thumbnail image of

Disclosures:

Javier Crespo - Board Membership: MSD, Roche, Janssen, Gilead

Manuel Romero-Gomez - Advisory Committees or Review Panels: Roche Farma,SA., MSD, S.A., Janssen, S.A., Abbott, S.A.; Grant/Research Support: Ferrer, S.A.

The following people have nothing to disclose: Pablo Cerro-Salido, Rocío Gal-lego-Durán, María J. Pareja, Emilio Gómez-González, Maria Carmen Rico, Rafael Aznar Méndez, Sandra Macho, Elisabetta Bugianesi, Maria Teresa Arias-Loste, Javier Abad, Susana Soto Fernandez, Reyes Aparcero López, Inmaculada Moreno-Herrera, Raul J. Andrade, Jose Luis Calleja, Oreste Lo Iacono

634

  1. Top of page

Non-Alcoholic Fatty Liver Disease Fibrosis Score (NFS) is an Independent Predictor of Mortality in Patients with NAFLD

Zobair M. Younossi1,2, Maria Stepanova1,2, Linda Henry1, James N. Cooper3,1, Shirley K. Kalwaney2, Chapy Venkatesan2, Alita Mishra2;
1Betty and Guy Beatty Center for Integrated Research, Inova Health System, Falls Church, VA; 2Department of Medicine, Inova Fairfax Hospital, Falls Church, VA; 3College of Science, George Mason University, Fairfax, VA

BACKGROUND AND AIM: Presence of hepatic fibrosis in NAFLD has been shown to be independently associated with mortality. However, staging of fibrosis requires a liver biopsy which is invasive with associated risks and costs. The NAFLD fibrosis score (NFS) is a non-invasive tests that has been shown to correlate well with hepatic fibrosis in patients with NAFLD. However, the ability of NFS to predict long-term mortality has not been validated. The aim of this study was to assess the performance of NFS in predicting long-term mortality in patients with NAFLD. METHODS: We used the third National Health and Nutrition Examination Survey with National Death Index-linked Mortality Files (NHANES III-NDI). NAFLD diagnosis was established by the presence of moderate to severe hepatic steatosis on the hepatic ultrasound without any other causes of chronic liver disease (alcohol consumption<20gr/day, negative HBs-antigen and anti-HCV, transferrin saturation<50%). NFS score was calculated for each eligible participant based on previously published formula using age, BMI, diabetes status, AST/ALT ratio, serum albumin and platelet count. Association of NFS with mortality was validated using Cox proportional hazard model with adjustment for confounders not accounted for by NFS. RESULTS: The initial NHANES III population included 20,500 adults. Of those, 2,334 had NAFLD. After 179 months of follow-up, the overall mortality was 20% (N=467). Five most common causes of death in NAFLD were cardiovascular-respiratory diseases (41.5%), solid organ malignancies excluding liver (21.0%), diabetes mellitus 32 (5.3%), and chronic liver diseases (3.17%). NFS was found to be independently associated with mortality: adjusted hazard ratio (aHR) = 1.361 (p<0.0001). After additional adjustment for age, the association of NFS with overall mortality remained significant (aHR=1.146, p<0.0001). Furthermore, in a series of survival analyses with different thresholds for NFS ranging from -5 to 5, we determined the best possible thresholds for NFS for the association with overall mortality. These survival analyses showed that NFS remained associated with mortality between a score of -1.80 (aHR=1.276, p=0.0498), and 1.25 (aHR=1.607, p=0.0203). The most significant association of NFS with mortality was found for a NAFLD Fibrosis score of 0.75 (aHR=1.775, p=0.0004). CONCLUSIONS: The well-established NAFLD Fibrosis Score is associated with overall mortality even after adjustment for age. The threshold for NFS that would return the outcome with best association with mortality in NAFLD cohort is 0.75 which is close to 0.68 suggested by the authors for ruling-in fibrosis in NAFLD patients.

Disclosures:

Zobair M. Younossi - Advisory Committees or Review Panels: Merck, Vertex, Tibotec/J and J; Consulting: Gilead Sciences

The following people have nothing to disclose: Maria Stepanova, Linda Henry, James N. Cooper, Shirley K. Kalwaney, Chapy Venkatesan, Alita Mishra

635

  1. Top of page

Excessive alcohol consumption enhances hepatocarcino-genesis in Japanese fatty liver disease patients without viral hepatitis

Yusuke Kawamura1,3, Yasuji Arase1,2, Kenji Ikeda1,3, Taito Fukushima1,3, Tasuku Hara1,3, Tetsuya Hosaka1,3, Masahiro Kobayashi1,3, Satoshi Saitoh1,3, Hitomi Sezaki1,3, Norio Akuta1,3, Fumitaka Suzuki1,3, Yoshiyuki Suzuki1,3, Yuki Ohmoto2,3, Kazuhisa Amakawa2,3, Hiroshi Tsuji2,3, Hiromitsu Kumada1,3;
1Hepatology, Toranomon Hospital, Tokyo, Japan; 2Health Management Center, Toranomon Hospital, Tokyo, Japan; 3Okinaka Memorial Institute for Medical Research, Toranomon Hospital, Tokyo, Japan

Purpose: The aims of this study were to determine the influence of excessive alcohol consumption on hepatocarcinogenesis and the risk factors for hepatocellular carcinoma (HCC) and to elucidate the utility of non-invasive predictive procedures for liver fibrosis, such as the FIB4-index, in the prediction of HCC in a large population of Japanese fatty liver patients without viral hepatitis. Methods: This was a retrospective cohort study conducted at a public hospital. Study subjects included 6,621 patients with non-alcoholic fatty liver disease (NAFLD) and 946 patients with alcoholic fatty liver disease (AFLD; daily alcohol consumption >70 g) diagnosed by ultrasonography. The median follow-up period was 5.9 years. The primary endpoint was onset of HCC. Evaluation was performed using Kaplan-Meier methodology and Cox proportional hazards analysis. Results: In NAFLD patients, there were 19 (0.29%) cases of newly diagnosed HCC, and the cumulative rates of NAFLD-related HCC were 0.05% at year 4, 0.21% at year 8, and 0.63% at year 12. In contrast, in AFLD patients, there were 18 (1.9%) cases diagnosed with HCC, and the cumulative rates of AFLD-related HCC were 0.67%, 1.89%, and 2.6%, respectively. The annual incidence of HCC was 0.05% and 0.22%, respectively. Multivariate analysis identified daily alcohol consumption >70 g (hazard ratio [HR]: 11.64; p<0.001), platelet count <150x103/μL (HR: 7.69; p<0.001), age >60 years (HR: 4.28; p<0.001), diabetes (HR: 3.96; p<0.001), serum AST level >40 IU/L (HR: 3.79; p<0.001), and serum albumin level <4.0 g/dL (HR: 2.56; P=0.008) as independent risk factors for HCC. Regarding the FIB4-index, 130 NAFLD patients (1.96%) and 24 (2.54%) AFLD patients were considered to have advanced fibrosis (presence of bridging fibrosis equivalent to NASH stage 3-4), and these estimated advanced fibrotic patients had a significantly higher incidence of HCC than estimated non-advanced fibrotic patients in each group. Conclusions: Excessive alcohol consumption has a considerable effect on hepatocarcinogenesis in fatty liver disease compared with NAFLD. And, non-invasive predictive procedures of liver fibrosis the FIB4-index possibly useful for prediction of high risk group of HCC in fatty liver patients with or without excessive alcohol consumption.

Disclosures:

Kenji Ikeda - Speaking and Teaching: Dainippon Sumitomo Pharmaceutical Company

Norio Akuta - Patent Held/Filed: SRL. Inc.

Hiromitsu Kumada - Speaking and Teaching: Bristol-Myers Squibb,Pharma International

The following people have nothing to disclose: Yusuke Kawamura, Yasuji Arase, Taito Fukushima, Tasuku Hara, Tetsuya Hosaka, Masahiro Kobayashi, Satoshi Saitoh, Hitomi Sezaki, Fumitaka Suzuki, Yoshiyuki Suzuki, Yuki Ohmoto, Kazuhisa Amakawa, Hiroshi Tsuji

636

  1. Top of page

The effect of smoking on progression and carcinogene- sis in fatty liver disease

Kazuhisa Kodama, Katsutoshi Tokushige, Etsuko Hashimoto, Maki Tobari, Noriko Matsushita, Tomomi Kogiso, Makiko Taniai, Nobuyuki Torii, Keiko Shiratori;
Tokyo Women's Medical University, Tokyo, Japan

Introduction) Smoking increases the risk of cardiovascular diseases and lung cancer. However, the effect of smoking on progression and carcinogenesis in liver diseases such as nonalcoholic fatty liver disease (NAFLD) and alcoholic liver diseases (ALD) has not been clear. In this study, we investigated the relationship between smoking and the clinical features in NAFLD, the rates of hepatocellular carcinoma (HCC) and extra-hepatic malignancies. Patients and Methods) 1) Three hundred forty-six NAFLD patients who underwent liver biopsy were divided into three groups: a non-smoking group (212 patients; mean age 52, male 63%), a past-smoking group (65 patients; mean age 54, male 66%) and a present-smoking group (69 patients; mean age 52, male 65%). Among the three groups, lifestyle-related diseases prevalence, blood test results and liver histological findings were compared. 2) Seventy-two patients with NAFLD liver cirrhosis (NAFLD-LC) and 85 patients with ALD liver cirrhosis (ALD-LC) were enrolled. The occurrence rate of HCC and extrahepatic malignancies were investigated. Results)1. Age and gender were almost the same among the three groups of NAFLD. Serum liver function test results (albumin, total bilirubin, AST, ALT, g-GTP, Platelet counts, prothrom-bin time) were not significantly different. However, HbA1C in the present-smoking groups was significantly higher (mean HbA1C<%>: present-smoking 6.6; past-smoking 5.9; nonsmoking 5.9). The prevalence of obesity, diabetes, hypertension and dyslipidemia were not different among the three groups, nor were the fibrosis grade, inflammatory grade and steatosis grade of liver biopsy (mean fibrosis grade: present-2.4; past- 2.25; non- 2.3). The Brinkman index was not correlated with fibrosis grade. 2. The HCC occurrence rate was not different between the smoking groups and non-smoking group for either ALD-LC or NAFLD-LC. The rate of extrahepatic malignancies in ALD-LC with smoking was higher than that without smoking (5-year extrahepatic malignancy rate: 19.6% in smoking vs. 0% in non-smoking). Regarding NAFLD-LC, the rate of extrahepatic malignancies was not influenced by smoking. Conclusion) Smoking worsened the control of diabetes, but did not influence the clinical and liver histological changes in NAFLD. In addition, smoking did not increase the HCC occurrence rate in either ALD-LC or NAFLD-LC. However, it increased the extrahepatic malignancies in ALD-LC, suggesting the synergic effect of alcohol and smoking on extrahepatic malignancies.

Disclosures:

The following people have nothing to disclose: Kazuhisa Kodama, Katsutoshi Tokushige, Etsuko Hashimoto, Maki Tobari, Noriko Matsushita, Tomomi Kogiso, Makiko Taniai, Nobuyuki Torii, Keiko Shiratori

637

  1. Top of page

Liver steatosis, presumed by SteatoTest or Controlled Attenuation Parameter (CAP), is associated to a risk of false-positive liver stiffness measurement by transient elastography in type-2 diabetic patients

Hugo Perazzo1,2, Yen Ngo3, Mona Munteanu3, Noemi Seurat1, Elena Luckina1, Fanny Rutka4, Marion Couteau4, Sophie Jacquem-inet5, Denis Monneret4, Françoise Imbert-Bismut4, Vlad Ratziu1, Agnes Hartemann-Heurtier5, Thierry Poynard1;
1Hepatology, APHP UPMC Liver center, Pitié -Salpêtrière Hospital, Paris, France; 2Inserm U938, Paris, France; 3BioPredictive, Paris, France; 4Biochemistry, Groupe Hopitalier Pitie-Salpetriere, Paris, France; 5Dia-betology, Groupe Hopitalier Pitie-Salpetriere, Paris, France

Background:Transient elastography(TE) with controlled attenuation parameter(CAP), based on liver stiffness measurement(LSM); FibroTest(FT), ActiTest(AT) and SteatoTest(ST) are validated non-invasive alternative to assess liver injury in NAFLD-risk patients as type-2 diabetics(T2D). Necro-inflammatory activity and steatosis might influence LSM leading to overestimation fibrosis stages. Aims:To evaluate the impact of i steatosis (SS)[>32%] on LSM in T2D patients. Methods: 142 T2D, without liver disease history, screened for fibro sis with FT were reinvestigated by FT and LSM(M and XL probes) after a median delay of 7 years. Patients with minimal fibrosis(FT<0.48-F0F1 METAVIR) at baseline and without progression during follow-up were included. Exclusion criteria were presence of advanced fibrosis(AF)[FT≥0.48] or activity[AT≥0.27] at the reinvestigation. Patients without AF as per FT(<0.48), but with AF LSM≥7.1kPa, at the reinvestigation,were supposed as false-positive of LSM(FP-LSM). SS(>32%) was defined as per ST≥0.69 or CAP≥283 dB/m. Results: 106 T2D patients with minimal fibrosis in the last 7 yrs and without necro-inflammatory activity were pre-included[54% males, age 63yrs, median BMI 27.6(20.8-52.8)Kg/m2,ALT 23(10-59)U/L].After exclusion of non-applicable LSM by both probes(6.6%), 99 patients were analyzed. Patients supposed to be a LSM-FP (26%) had no liver-related complications. In uni-variate analysis, patients considered as FP-LSM versus non-FP-LSM, had higher: BMI[32.3(21.3-49.5)vs26.5(1 9.6-35.2)],ST(0.64±0.17vs0.46±0.19); waist circumference(115±18vs100±11cm), thoracic fold(25±1 0vs19±6mm) and higher rates of SS(58%vs19%), all p<0.001. SS patients as per ST, had higher median LSM(range)[7.7(5-75)vs 5.5(3-64),p=0.02]. In logistic regression, the presence of SS, by ST[OR=6.9(95%CI 1.7-28.4);p=0.007], remained significantly associated to FP-LSM in a multivariate model adjusted for age, gender, thoracic fold, waist circumference and metabolic factors. Among 59 patients with an applicable CAP simultaneous to ST, Spearman's correlation coefficient was r=0.37, p=0.03.sSupposed FP-LSM patients had also higher rates of SS by CAP(40%vs15%,p=0.04) compared to non- FP-LSM. Patients with SS as per CAP, had higher median LSM[6.5(4.4-13.6)vs 5.7(3.2-8.7)kPa,p=0.01].The high failure rate (35%) of the M probe that measures concomitantly the CAP, limited the multi-variate analysis for CAP in this population. Conclusion: In type-2 diabetic patients, the presence of severe steatosis presumed by SteatoTest was independently associated to the overestima-tion of liver fibrosis by liver stiffness measurement.

Disclosures:

Yen Ngo - Employment: BioPredictive

Mona Munteanu - Employment: Biopredictive

Vlad Ratziu - Advisory Committees or Review Panels: GalMed, Abbott, Genfit, Enterome, Gilead; Consulting: Astellas, Axcan, Pfizer, Sanofi-Synthelabo, Genen-tech, Nycomed

Agnes Hartemann-Heurtier - Consulting: Sanofi-Aventis, Pfizer; Grant/Research Support: Lilly

Thierry Poynard - Advisory Committees or Review Panels: MSD; Speaking and Teaching: BMS; Stock Shareholder: BioPredictive

The following people have nothing to disclose: Hugo Perazzo, Noemi Seurat, Elena Luckina, Fanny Rutka, Marion Couteau, Sophie Jacqueminet, Denis Mon-neret, Françoise Imbert-Bismut

638

  1. Top of page

Factors associated with diabetes and prediabetes in HBV-infected patients residing in North America: Results from the Adult Cohort Study of the NIDDK-Sponsored Hepatitis B Research Network (HBRN)

Mandana Khalili1, Arun J. Sanyal2, Yona K. Cloonan3, Marc G. Ghany4, W. Ray Kim5, Raymond T. Chung6, Norah Terrault1, Elizabeth M. Brunt7, Daryl Lau8, Mauricio Lisker-Melman7, Anna S. Lok9;
1University of California San Francisco, San Francisco, CA; 2Virginia Commonwealth University, Richmond, VA; 3University of Pittsburgh, Pittsburgh, PA; 4National Institute of Health, Bethesda, MD; 5Mayo clinic, Bethesda, MN; 6Massachusetts General Hospital, Boston, MA; 7Washington University, St. Louis, MO; 8Beth Israel Deaconess Medical Center, Harvard Medical School,, Boston, MA; 9University of Michigan, Ann Arbor, MI

Background/Aims: Diabetes is associated with increased risk of hepatocellular carcinoma (HCC) and disease progression in chronic liver disease but factors associated with diabetes in patients infected with hepatitis B virus (HBV) living in North America (NA) have not been previously studied. Therefore, we set out to determine the rates and factors predictive of diabetes and impaired fasting glucose (IFG) in a large multi-ethnic NA cohort of patients infected with HBV. Methods: HBsAg+ adults without liver decompensation, HCC, liver transplant, HIV or current antiviral therapy from 21 US/Canada centers were enrolled. Those with known diabetes or fasting glucose data were included in the analysis. Diabetes was defined by history/medications or fasting glucose >126 mg/dL and IFG as fasting glucose 110-125 mg/dL. Overweight/obese was defined by race-adjusted BMI. Results: 900 patients were included in the analysis: median age 43 years, 51% male, 71% Asian (14% black, 11% white), 81% born outside US/Canada (30% migrated >20 yrs ago). Most whites were born in NA, while 66% of blacks, and 92% of Asians were born in Africa and Asia, respectively. Median BMI (Kg/m2) was 27 in whites, 28 in blacks, 23 in Asians. 27% were HBeAg+, 20% had HBV DNA >10Λ7 IU/mL and 4% had cirrhosis. 70 (8%) patients had IFG and 112 (12%) had diabetes. Diabetes rate by racial category and birth place were: 24% in blacks, 18% in whites, 9% in Asians, and 13% in other races; 21% in NA-born, 15% in foreign-born (FB) migrated >20 yrs, and 6% in FB migrated>20 yrs. On univariable analysis, compared to those with normal glucose (N=71 8), diabetes was associated with older age (OR 1.08), race (black OR 1.49; Asian OR 0.43), overweight/obese (OR 4.04), cirrhosis (OR 2.99), diabetes family history (OR 3.9), NA birth (OR 2.03), FB migrated ≤ 20 yrs ago (OR 0.40), moderate alcohol intake (OR 0.55), and HBeAg+ status (OR 0.50). On multivariable analysis, independent predictors of diabetes were age (OR 1.08, 95%CI (CI) 1.05-1.12), black (vs white) race (OR 3.13, CI 1.22-8.02), overweight/obese (3.29, CI 1.76-6.14), diabetes family history (OR 3.55, CI 1.94-6.47), and ALT (>2x ULN, OR 1.91, CI 1.07-3.41). Independent predictors of IFG were age (OR 1.04, CI 1.02-1.07) and overweight/obese (OR 2.17, CI 1.03-4.51). Conclusion: In this large NA multi-ethnic and mostly foreign-born HBV-infected cohort, diabetes and prediabetes were associated with known risks and the main modifiable factor is obesity. These results highlight the opportunities for interventions to prevent diabetes especially among the at-risk ethnic groups with HBV infection.

Disclosures:

Mandana Khalili - Advisory Committees or Review Panels: Gilead Inc.; Grant/Research Support: Gilead Inc., BMS Inc, BMS Inc

Arun J. Sanyal - Advisory Committees or Review Panels: Gore, Gilead, Abbott, Ikaria; Consulting: Salix, Immuron, Exhalenz, Bayer-Onyx, Genentech, Norgine, GalMed, Novartis, Echosens, Takeda; Grant/Research Support: Salix, Genentech, Genfit, Intercept, Ikaria, Takeda, Gilead; Independent Contractor: UpToDate

W. Ray Kim - Advisory Committees or Review Panels: Salix; Consulting: Bristol Myers Squibb, Gilead

Raymond T. Chung - Advisory Committees or Review Panels: Idenix; Consulting: Enanta; Grant/Research Support: Gilead, Merck, Mass Biologic, Gilead

Norah Terrault - Advisory Committees or Review Panels: Eisai, Biotest; Consulting: BMS; Grant/Research Support: Eisai, Biotest, Vertex, Gilead, AbbVie, Novartis

Elizabeth M. Brunt - Speaking and Teaching: Geneva Foundation

Daryl Lau - Advisory Committees or Review Panels: Gilead, BMS; Consulting: Roche; Grant/Research Support: Gilead, Merck

Mauricio Lisker-Melman - Speaking and Teaching: Gilead, GlaxoSmithKline, Inter-mune, Roche, Valeant, Schering-Plough, Novartis, Gilead, GlaxoSmithKline, Inter-mune, Roche, Valeant, Schering-Plough, Novartis, Gilead, GlaxoSmithKline, Intermune, Roche, Valeant, Schering-Plough, Novartis, Gilead, GlaxoSmithKline, Intermune, Roche, Valeant, Schering-Plough, Novartis

Anna S. Lok - Advisory Committees or Review Panels: Gilead, Immune Targeting System, MedImmune, Arrowhead, Bayer, GSK, Janssen, Novartis; Grant/Research Support: Abbott, BMS, Gilead, Merck, Roche

The following people have nothing to disclose: Yona K. Cloonan, Marc G. Ghany

639

  1. Top of page

Predictors of response after extended pioglitazone treatment in patients with prediabetes or T2DM and nonalcoholic steatohepatitis (NASH)

Fernando Bril1, Romina Lomonaco1, Beverly Orsak2, Joan Hecht2,3, Carolina Ortiz-Lopez2, Jean Hardies1, Fermin Tio3,4, Kenneth Cusi1,2;
1Endocrinology, Diabetes and Metabolism, University of Florida, Gainesville, FL; 2Diabetes, University of Texas Health Science Center at San Antonio, San Antonio, TX; 3Diabetes, Audie L. Murphy Veterans Administration Medical Center, San Antonio, TX; 4Pathology, University of Texas Health Science Center at San Antonio, San Antonio, TX

Background and Aim: An early short-term RCT suggested that pioglitazone may improve insulin resistance and liver histology in patients with prediabetes or type 2 diabetes (T2DM) and nonalcoholic steatohepatitis (NASH) (Belfort et al, NEJM 2006). However, this small (n=55), short-term (6 months) study did not allow to clearly establish predictors of response. We aimed to determine predictors of histological improvement after PIO treatment in a recent 18-month RCT in this population (unpublished, main results reported elsewhere at this meeting). Methods: Patients with biopsy-proven NASH were randomized to PIO (n=51) or placebo (n=50) and followed for 1 8 months. We measured before and after treatment: 1) Liver histology by biopsy; 2) Liver fat by magnetic resonance imaging and spectroscopy (MRS); 3) Total body fat (TBF) by DXA; 4) Liver (suppression of hepatic glucose production), adipose tissue (suppression of plasma free fatty acids [SupprFFA]) and muscle (Rd) insulin sensitivity during an euglycemic insulin clamp; and 5) Hepatic insulin resistance index (HIRi = hepatic glucose production x fasting plasma insulin). Results: Both groups were well-matched at baseline forage, gender, BMI/TBF, prevalence ofT2DM, insulin resistance and liver histology. PIO significantly improved insulin resistance and liver histology vs. placebo (steatosis, ballooning, inflammation [all p<0.001] and fibrosis [p=0.03]). As a group, changes in AST and ALT correlated moderately with changes in liver histology (both r=0.43, p<0.001). Changes in several inflammatory biomarkers (TGF-β, TNF-α, IL-6, E/P-selectin, VCAM, hsCRP, other) had a low or no correlation with changes in liver histology (assessed as the NAFLD activity score [NAS]). Only fasting plasma adiponectin (r= -0.57, p=0.02) and insulin (r= 0.33, p=0.004) concentrations showed a strong correlation. When the subgroup of patients treated with PIO was examined, improvement in the NAS with PIO did not correlate with changes in BMI or TBF, AST/ALT, fasting glucose, A1c or liver fat (MRS). Of note, improvement in the NAS with PIO treatment correlated most strongly with amelioration of liver (HIRi: r=0.54, p<0.01) as well as adipose tissue and muscle insulin resistance (both r= -0.39, p=0.02-04). Conclusions: These results suggest that the beneficial effect of extended PIO therapy on liver histology in patients with prediabetes or T2DM is directly related to an improvement in insulin sensitivity. This emphasizes the role of insulin resistance in the pathogenesis of NASH, and suggests that insulin sensitizers may have a key role in the treatment of NASH in prediabetes or type 2 diabetes (T2DM).

Disclosures:

Beverly Orsak - Employment: UTHSCSA

Kenneth Cusi - Consulting: Merck, Daichi-Sankyo, Roche, Janssen; Grant/Research Support: Takeda, Novartis, Mannkind

The following people have nothing to disclose: Fernando Bril, Romina Lomonaco, Joan Hecht, Carolina Ortiz-Lopez, Jean Hardies, Fermin Tio

640

  1. Top of page

Serum Vitamin D is Inversely Associated with Nonalcoholic Fatty Liver Disease and Advanced Fibrosis in the General Population

Donghee Kim1, Won Kim2, Hwa Jung Kim3;
1Department of Internal Medicine, Healthcare Research Institute, Seoul National University Hospital Gangnam Center, Seoul, Republic of Korea; 2Department of Internal Medicine, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Republic of Korea; 3Department of Clinical Epidemiology and Biostatis-tics, san Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea

Background and Aims: Nonalcoholic fatty liver disease (NAFLD) is a clinicopathological liver disease that encompasses simple steatosis, nonalcoholic steatohepatitis, advanced fibrosis, and cirrhosis. NAFLD is related to dyslipidemia, diabetes, and metabolic syndrome. Recently, subjects with low vitamin D levels are associated with metabolic syndrome and diabetes. However, associations between low vitamin levels, and ultra-sonographically diagnosed NAFLD and advanced fibrosis in patients with NAFLD have not been completely established. The aim of this study was to characterize the relationship between vitamin D levels and NAFLD in the large, national, general population. Methods: The Third National Health and Nutrition Examination Survey (NHANES) from 1988 to 1994 were utilized in this study. A total of 11,808 participants without known liver disease were finally selected and analyzed. NAFLD was defined as ultrasonography diagnosed hepatic steatosis without other known liver diseases. The presence and absence of advanced fibrosis in NAFLD was determined by the NAFLD fibrosis score. Results: The prevalence of ultrasonography diagnosed NAFLD was 22.9%. NAFLD was significantly inversely associated with vitamin levels after adjusted for age and sex [odds ratio (OR) 0.85 95% confidence interval (CI) 0.75-0.96]. The age, sex-adjusted prevalence of NAFLD decreased steadily with increasing levels of vitamin D [OR 0.53 95% CI 0.40-0.70, lowest quintile vs. highest quintile, p for trend <0.001]. Multivariate regression analysis after adjustment for known risk factors showed that NAFLD was statistically significantly inversely associated with vitamin levels (>20 ng/ml) [OR 0.79, 95% CI, 0.65-0.95] and the grade of vitamin levels in a dose-dependent manner [OR 0.76, 95% CI, 0.58-1.00 in 5th quintile vs. lowest quintile, p for trend=0.001]. With regards to advanced fibrosis, age, sex-adjusted advanced fibrosis in patient with NAFLD decreased steadily with increasing levels of vitamin D [OR 0.23 95% CI 0.08-0.66, highest tertile vs. lowest tertile, p for trend <0.001]. Multivariate regression analysis after adjustment for known risk factors showed that NAFLD was statistically significantly inversely associated with the grade of vitamin levels in a dose-dependent manner [OR 0.17, 95% CI, 0.24-0.74 in highest tertile vs. lowest tertile]. Conclusions: Serum vitamin D, even in the range of normal levels, was found to be inversely related to NAFLD in a dose-dependent manner. Vitamin D is inversely associated with advanced fibrosis in patients with NAFLD independently of known metabolic risk factors. Vitamin D might have protective effect for NAFLD and advanced fibrosis.

Disclosures:

The following people have nothing to disclose: Donghee Kim, Won Kim, Hwa Jung Kim

641

  1. Top of page

Prediction of Liver Fibrosis by Non-invasive Markers

Vipin Verma1, Shiv K. Sarin2, Ravi Kant3, Archana Rastogi4, Chhagan Bihari4, Ashish Kumar5;
1Internal Medicine, MedStar Georgetown University Hospital/MedStar Washington Hospital Center, Washington, DC; 2Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India; 3Endocrinology, University of Maryland Medical Center, Baltimore, MD; 4Pathology, Institute of Liver and Biliary Sciences, New Delhi, India; 5Hepatology, Ganga Ram Institute for Postgraduate Education & Research (GRIPMER), Sir Ganga Ram Hospital, New Delhi, India

Background: There is an urgent need to validate non-invasive markers to quantify fibrosis because of invasive nature and complications of Liver Biopsy. Aim: To compare various non invasive markers available for liver fibrosis including Lok score, Fibro Index(FI)score, King score, Fibro Q score, AST to Platelet Ratio Index(APRI), FIB 4 score and AST to ALT ratio(AAR)with Ishak stage(IS). Patients and Methods: Consecutive patients of chronic hepatitis, of various etiologies, attending our hospital, were evaluated for fibrosis by liver biopsy and various non-invasive markers. Blood for non-invasive markers was drawn on same day as of liver biopsy. Ishak stage was used to grade fibrosis histologically. Spearman's correlation was used to compare non-invasive markers with Ishak stage. Each non-invasive marker was also evaluated by ROC curve to predict significant fibrosis(IS >2). Youden's index was used to find out best cut-off value of each score in predicting significant fibrosis. Sensitivity, specificity, PPV, NPV and accuracy were calculated for each score. Results: 80%(96/120) enrolled patients had viral etiology and 20% had autoimmune, alcohol, or NASH etiology.Their mean age was 36.7±12.5 years and 78.3% were males. The median ISHAK stage was 2(range 0-6)and 45% patients had significant fibrosis(IS >2). All non-invasive scores showed significant correlation with Ishak stages(Table-1). The highestaccuracy to predict significant fibrosis(IS >2)was obtained by King score[sensitivity=77.8%;specificity=78.8%;area under receiver operating characteristic(AUROC)=0.84](Table-1). Conclusion: Among various non-invasive markers available to predict liver fibrosis, king score[age(yrs)xAST(U/L)xlNR/Platelets(per nL)]showed the highest accuracy(78.3%)but not good enough to replace liver biopsy clinically. However, these markers can be used in combinations to identify the hepatic fibrosis patient, when liver biopsy is not feasible or available, until a better marker is identified. Our study shows that the currently available non-invasive markers can be useful in predicting hepatic fibrosis in certain clinical scenarios but due to lack of enough diagnostic accuracy cannot replace gold standard liver biopsy yet.

Table 3. Statistical evaluation of various non-invasive markers.
ScoreSpearman's rho correlationP valueAUROCBest Cut-offSensitivitySpecificityPPVNPVAccuracy
Lok.48<0.01.721.144.494.085.767.467.4
Fibro Index.48<0.01.722.263.077.370.072.971.7
King.66<0.01.849.277.878.875.081.378.3
Fibro Q.57<0.01.805.670.478.873.176.575.0
APRI.55<0.01.780.6270.480.374.576.875.8
FIB 4.48<0.01.750.5057.483.373.870.571.7
AAR.190.036.561.1242.675.859.061.760.8

Disclosures:

Ashish Kumar - Consulting: Abbott India Limited, Ranbaxy India Limited

The following people have nothing to disclose: Vipin Verma, Shiv K. Sarin, Ravi Kant, Archana Rastogi, Chhagan Bihari

642

  1. Top of page

Non-invasive measurement of steatosis by controlled attenuation parameter can detect very early fatty change in health checkup subjects

Jung Ran Choi1, Ja Kyung Kim1, Jung Il Lee1, Ah Ran Choi1, Kyung Ah Lee1, HyunJung Chung2, Da Hyun Jung1, Kwan Sik Lee1;
1Yonsei University College of Medicine, Seoul, Republic of Korea; 2Inha University, Inchon, Republic of Korea

Background/Aims: Steatosis may facilitate the progression of several chronic liver diseases that can result in fibrosis and cirrhosis. Until now, the most practically used non-invasive means of detecting steatosis is ultrasonography (US), although it can only detect steatosis of greater than 30%. Recently, controlled attenuation parameter (CAP) being implemented on FibroScan(®) (Echosens, Paris, France), can evaluate both steatosis and fibrosis simultaneously, and is reported to be efficient in detecting even low grade steatosis (>10% steatosis) noninvasively. We analyzed the CAP value in health checkup subjects and investigated the correlation between CAP value and US finding along with other clinical parameters. Methods: CAP results were retrospectively collected with other data including demographics, blood test results, and finding of abdominal ultrasonography from database of health checkup center. Steatosis grade was decided by cut-offs of CAP according to a previous report (Sasso M et al., 2010): 238 dB/m for S ≥ 1, 260 dB/m for S ≥ 2, and 293 dB/m for S3. Results: Total 280 subjects (M:F=157:123) were included. Mean age was 51 ± 11 years. Fatty liver was detected in 119 (42.5%) patients by US, when it was detected in 160 (57.1 %) patients by CAP. According to the CAP value, S0:S1:S2:S3 patients were 120:59:58:43, respectively. Mean CAP values were 203.34 ± 28.39 dB/m for S0, 248.83 ±6.14 dB/m for S1, 274.33 ± 8.53 dB/m for S2, and 322.35 ± 22.20 dB/m for S3. CAP was significantly correlated with the body weight (r = 0.404, p < 0.001), BMI (r = 0.445, p < 0.001), and the fatty liver grade by US (r = 0.472, p < 0.001). Among 161 patients whose fatty changes were not detected by US, steatosis was detected in 65 (40.4%) patients by CAP. Patients with steatosis that was recognized by CAP only had significantly low stiffness, heavy weight, high height, BMI, body fat rate, visceral fat, systolic blood pressure, triglyceride, low density lipoprotein, and low high density lipoprotein than patients without steatosis that was confirmed by the both measurements. Conclusions: In health check-up subjects, CAP can be more sensitive in detecting very early steatosis. Even with normal US finding, patients with very early steatosis that could only be detected by CAP had worse metabolic parameters. Histologic validation is warranted to use cut-offs of CAP for steatosis grade in non-chronic live disease subjects.

Disclosures:

The following people have nothing to disclose: Jung Ran Choi, Ja Kyung Kim, Jung Il Lee, Ah Ran Choi, Kyung Ah Lee, Hyun Jung Chung, Da Hyun Jung, Kwan Sik Lee

643

  1. Top of page

Can ‘healthy’ normal alanine aminotransferase levels identify metabolically obese phenotype regardless of body mass index?

Hee Yeon Kim, Jong Young Choi, Chang Wook Kim, Chang Don Lee, Chung-Hwa Park, Young Sok Lee, Nam Ik Han;
Internal Medicine, Catholic University Medical College, Seoul, Republic of Korea

Background: The aim of the study was to revise the upper limit of normal (ULN) of serum alanine aminotransferase (ALT) and to investigate the predictive value of updated sex-specific ALT level for metabolic derangement stratified according to body mass index (BMI) in a large sample representative of the Korean population. Methods: We analyzed nationally representative Fourth Korea National Health and Nutrition Examination Survey (KNHANAES IV) data. This cross-sectional study included 2416 healthy cohort aged 33.9 ± 0.3 years. Upper threshold of healthy ALT level was set to the 95th percentile of the reference healthy population. A binary logistic regression analysis was performed to assess the relative risk for metabolic syndrome according to the healthy normal ALT level. Results: The revised ULN of serum ALT level in low-risk healthy participants were 30 IU/L and 22 IU/L for men and women, respectively. Serum ALT level was higher in individuals with metabolic syndrome compared to those without metabolic syndrome in both genders stratified according to BMI. After adjusting for age, smoking, alcohol drinking and regular physical activity, unhealthy normal ALT level (males; 30-40 IU/L, females; 22-40 IU/L) was a reliable marker predictors for the metabolic syndrome stratified according to BMI and sex. Conclusions: Newly revised threshold for ALT could be proposed as a simple clinical metabolic parameter whicn can identify metabolic abnormality. We suggest that people with unhealthy normal ALT levels may need further investigation for the presence of metabolic syndrome.

Comparisions of serum ALT level in individuals with and without metabolic syndrome stratified by BMI and sex.

Table 4. 
 normal/overweightobese
 metabolic syndromeno metabolic syndromep valuemetabolic syndromeno metabolic syndromeP value
Males
age, years49.3±0.836.9±0.4<.000l46.9±0.638±0.6<.0001
ALT, IU/L22.8±10.618.9±0.2<.000126.510.423.510.4<.0001
% of unhealthy normal ALT19.1(2.9)10.4(0.8)0.000336.0(2.8)21.0(2.1)<.0001
Females
age, years52.410.538±10.2<.000148.9±0.541.1±0.5<.0001
ALT, IU/L18.4±0.513.5±0.1<.000120.6±0.316.7±0.3<.0001
% of unhealthy normal ALT29.1(2.7)7.9(0.5)<.000140.6(2.3)19.7(1.9)<.0001

Disclosures:

Chang Wook Kim - Consulting: Gilead; Grant/Research Support: BMS, Boehringer Ingelheim, Pharmicell; Speaking and Teaching: BMS, GSK, Dae-woong

The following people have nothing to disclose: Hee Yeon Kim, Jong Young Choi, Chang Don Lee, Chung-Hwa Park, Young Sok Lee, Nam Ik Han

644

  1. Top of page

Reticuloendothelial Cell System Iron Staining is a Predictor of Progression to Borderline or Definite Steatohepatitis in Patients without Fibrosis

Kris V. Kowdley1,2, David E. Kleiner3, Laura Wilson4, Patricia H. Belt4, James E. Nelson1,2;
1Liver Center of Excellence, Digestive Disease Institute, Virginia Mason Medical Center, Seattle, WA; 2Benaroya Research Institute, Seattle, WA; 3Laboratory of Pathology, National Cancer Institute, Bethesda, MD; 4Bloomberg School of Public Health, Johns Hopkins, Bethesda, MD

Background: The presence of reticuloendothelial cell system (RES) iron staining has been associated with several histological features of disease including advanced fibrosis, increased apoptosis, increased ballooning and a definitive diagnosis of nonalcoholic steatohepatitis (NASH) in cross-sectional studies in subjects with nonalcoholic fatty liver disease (NAFLD). Aim: The aim of this study was to investigate if RES iron would be associated with NAFLD progression including fibrosis development in a longitudinal analysis of paired biopsies. Methods: Adult patients enrolled in one of the NASH CRN studies with 2 or more biopsies at least a year apart, (excluding treatment arms of the PIVENS study) in which the first biopsy had iron staining results, were included. All biopsies underwent blinded consensus review. Univariate and stepwise forward multivariable logistic regression models (cutoff p<0.15) adjusting for sex, age at biopsy, time between biopsies and 14 histologic variables were used to assess association with A) the development of any fibrosis in patients with no fibrosis on initial biopsy or B) the development of borderline or definite steatohepatitis in patients with a diagnosis of “not NASH“ or “not NAFL“ on initial biopsy. Results: 310 patients (age 47±11 years) with multiple biopsies and iron staining results were studied. The mean time between biopsies was 4.4 ± 2.4 years and the majority of patients were female (65%) and Caucasian (81%). 34/72 patients without fibrosis on an initial biopsy developed fibrosis on a second biopsy, ranging from stage 1-3. In univariate logistic regression on the development of fibrosis, there was a trend for grade of both hepatocellular (OR=2.5, p=0.071) and RES iron (OR 3.0, p=0.054). However, only RES iron grade remained significant using multivariable analysis (OR 4.9, 95%CI 1.2-20, p=0.02). 24/44 patients without fibrosis and an initial diagnosis of “not NASH“ or “not NAFL“ developed borderline or definite steato-hepatitis on a second biopsy. In multivariable logistic regression analysis only RES iron grade predicted progression to borderline or definite steatohepatitis (OR 25, 95%CI 1.4-443, p=0.027). Conclusions: Only RES iron grade, independent of any other histologic feature, as well as age, sex and duration between biopsies, predicted both the initial development of any fibrosis in all patients and progression to NASH/borderline NASH in patients without fibrosis and having a diagnosis of “not NASH“ or “not NAFL“. These data suggest RES iron may be detrimental in the early stages of NAFLD and may contribute to the initiation of fibrogenesis, but this finding should be confirmed in a larger cohort.

Disclosures:

Kris V. Kowdley - Advisory Committees or Review Panels: Abbott, Gilead, Merck, Novartis, Vertex; Grant/Research Support: Abbott, Beckman, Boeringer Ingelheim, BMS, Gilead Sciences, Ikaria, Janssen, Merck, Mochida, Vertex

The following people have nothing to disclose: David E. Kleiner, Laura Wilson, Patricia H. Belt, James E. Nelson

645

  1. Top of page

A new role of Na+/taurocholate cotransporter pump in progression of nonalcoholic fatty liver disease

Varenka J. Barbero-Becerra1, Jorge A. López-Velázquez1, Vicente Sánchez- Valle1, Luis D. Carrillo-Córdova1, Nancy E. Aguilar-Olivos1, Norberto C. Chavez-Tapia1, Fredy Chablé-Montero2, José M. Ramírez-Jaramillo2, Misael N. Uribe-Esquivel1, Nahum Mén-dez-Sanchéz1;
1 Biomedical Research Department, Medica Sur Clinic & Foundation, Mexico, Mexico; 2Pathology Research Unit, Medica Sur Clinic & Foundation, Mexico City, Mexico

Background: Nonalcoholic fatty liver disease (NAFLD) presents a high prevalence and diverse outcomes. Little is known about inflammatory and apoptosis mechanisms underlying the progression of NAFLD and its association with critical features as Kupffer cells and hepatocellular ballooning. Recently, PNPLA3 and bile acids have been associated in chronic liver disease progression. Aim: Investigate the expression of Na+/tauro-cholate cotransporter pump (NTCP) and PNPLA3 in NAFLD progression and the association of inflammatory and apoptosis markers in liver biopsies of patients with NAFLD. Methods: 117 human liver biopsies of NAFLD patients were collected from 2009 to 2012 and classified according to disease spectrum (66 steatosis (S), 40 steatohepatitis (NASH) and 11 cirrhosis (C)). Biopsies were analyzed by immunohistochemistry in a tissue microarray for expression of CD68 as a kupffer cell bio-marker, CD163 as a marker of monocyte macrophage; therefore inflammatory and apoptosis markers were evaluated. RT-PCR real time was carried out for NTCP and PNPLA3 gene expression. H&E staining was employee to identify hepatocellular ballooning. Results. 117 patients, 51/49% (women/men) with a mean age of 52.4±14 years. 40% of the population had metabolic syndrome, 65% were overweight or obese, 24% diabetes, 73% hypertriglyceridemia, 38% hypercholesterolemia and 38% hypertension, and elevated AST, ALT and GGT in 54%, 56% and 65 %, respectively. NTCP expression was significantly up-regulated in C vs. S (P <0.01). A strong up-regu-lation was observed in PNPL3 vs. NTCP expression (p<0.05) according to liver injury degree. Regarding to inflammatory and apoptosis markers, an increase in CD68 expression between S vs. NASH (77% vs 100%), S vs. C (77% vs 36%) and NASH vs. C (100% vs 36%) (P <0.01) was shown. In NASH stage CD68+, CD1 63+ and positive hepatocellular ballooning were significantly associated with inflamatory and apoptosis biomarkers (Table 1). Conclusion. NTCP as an important regulator of bile acids transport seems to play a major role in the progression of NAFLD and presents an important asociation with PNPLA3 in terms of NAFLD stages. Kupffer cells represent a central component where pro-inflammatory and early apoptosis markers are associated with the presence of hepatocyte ballooning promoting hepatocellular damage through liver damage progression.

Table 1

BiomarkerIL-1(%)IL-6 (%)IL-18 (%)TNFa (%)Caspase-3 (%)
  1. Data were expressed as percentage of positive samples vs. percentage of negative samples.* p< 0.05; **p< 0.01.

CD68+82 vs. 18**72 vs. 18**66 vs. 3478 vs. 22*78 vs. 22**
CDI63+22 vs. 7876 vs. 2464 vs. 3676 vs.24*76 vs. 27*
Hepatocellular Ballooning+79 vs. 2182 vs. 1864 vs. 3686 vs.l4**82 vs. 18

Disclosures:

The following people have nothing to disclose: Varenka J. Barbero-Becerra, Jorge A. López-Velázquez, Vicente Sánchez-Valle, Luis D. Carrillo-Córdova, Nancy E. Aguilar-Olivos, Norberto C. Chavez-Tapia, Fredy Chablé-Montero, José M. Ramírez-Jaramillo, Misael N. Uribe-Esquivel, Nahum Méndez-Sanchéz

646

  1. Top of page

Differences in hepatic gene expression between simple steatosis and nonalcoholic steatohepatitis are correlated with polyunsaturated fatty acids and lipid peroxidation in the liver

Bianca M. Arendt1, Elena M. Comelli2, Ian McGilvray1,4, David W. Ma3, Wendy Lou5, Scott Fung1,4, Johane P. Allard1,4;
1Toronto General Hospital, University Health Network, Toronto, ON, Canada; 2Department of Nutritional Sciences, University of Toronto, Toronto, ON, Canada; 3Department of Human Health and Nutritional Sciences, University of Guelph, Guelph, ON, Canada; 4Department of Medicine, University of Toronto, Toronto, ON, Canada; 5Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada

Low levels of polyunsaturated fatty acids (PUFA) in the liver and increased oxidative stress may contribute to nonalcoholic fatty liver disease, and ultimately cancer, by modulating hepatic gene expression. Aims: 1. To compare hepatic gene expression among patients with simple steatosis (SS) or nonalcoholic steatohepatitis (NASH) and healthy controls (HC); 2. To determine whether altered gene expression correlates with hepatic PUFA and oxidative stress parameters. Methods: In a cross-sectional study, gene expression (whole genome microarray), omega-6 and omega-3 PUFA in % of total lipids (gas chro-matography), lipid peroxidation, and antioxidant power (test kits) were measured in liver tissue from 20 patients with SS, 19 with NASH, and 24 live liver donors (HC). Differentially expressed genes were selected by ANOVA with Tukey's HSD (p<0.05) and filtered for ≥2-fold difference. Spearman correlations were calculated for SS + NASH combined. Values are mean±SD or median (IQR). Results: Omega-3 PUFA (%) were lower in NASH than in HC [2.3 (0.3) vs. 4.2 (1.9); p<0.05] and omega-6 PUFA (%) were lower in both SS (20.7±2.7) and NASH (17.1 ±3.9) than in HC (24.2±3.7, p<0.01). 732 non-redundant genes were differentially expressed among HC, SS, and NASH, including 21 genes that were different between SS and NASH. Of these, 4 and 7 were positively or negatively correlated with n-6 PUFA, respectively. ANXA2 and PEG10 were negatively, and MAG was positively correlated with n-3 PUFA. At least 7 of the genes that are differentially expressed between SS and NASH and correlated with omega-3 PUFA, are linked to hepatocellular carcinoma. Lipid peroxidation was correlated with ACOT1 (r=-0.369, p=0.025) and DPPIV (r=0.333, p=0.044), which regulate fatty acid overload, diabetes, and hepatic steatosis, but no associations were seen between gene expression and antioxidant power. In conclusion, hepatic PUFA content and lipid peroxidation correlate with gene expression. These data support a role of PUFA depletion and lipid peroxidation in the pathogenesis of NASH and its complications.

Disclosures:

David W. Ma - Advisory Committees or Review Panels: Heinz; Consulting: Vegetable Oils Industry of Canada, PepsiCo; Speaking and Teaching: Unilever

Scott Fung - Advisory Committees or Review Panels: Merck, Vertex; Grant/Research Support: Gilead Sciences, Roche; Speaking and Teaching: Gilead Sciences, BMS

The following people have nothing to disclose: Bianca M. Arendt, Elena M. Comelli, Ian McGilvray, Wendy Lou, Johane P. Allard

647

  1. Top of page

Peripheral but not hepatic insulin resistance correlates with severity of non alcoholic steatohepatitis

Jaividhya Dasarathy2, Ciaran E. Fealy3, Carol A. Hawkins1, Patricia T. Brandt1, Arthur J. McCullough1,3, John P. Kirwan3,1, Srinivasan Dasarathy1,3;
1 Gastroenterology and hepatology, Cleveland Clinic Foundation, Cleveland, OH;2Family Medicine, MetroHealth Medical Center, Cleveland, OH; 3Pathobiology, Cleveland Clinic, Cleveland, OH

Background. Hepatic insulin resistance (HIR) is believed to be a primary pathogenic mechanism for the development of NASH. Recently, non invasive measures of HIR using an oral glucose tolerance test have been suggested as diagnostic markers of NAFLD. Methods. A prospective study was performed in 33 (14M, 19F) well characterized non-diabetic patients with NASH and 23 healthy, weight matched controls (8M, 15F) to assess the relationship between indices of insulin resistance and histological severity of NASH. All subjects had a 2 hour oral glucose tolerance test. The homeostasis model assessment (HOMA) index was calculated as fasting insulin xfasting glucose/22.5. HIR index was calculated in 2 ways: AUC0-30min glucose × AUC0-30min insulin and the formula: -0.091 + (log insulin AUC0-120min × 0.400) + (log fat mass % × 0.346) -(log HDL Cholesterol × 0.408) + (log BMI × 0.435). AUC90-120 glucose and insulin were also calculated. The trapezoidal method was used to calculate glucose and insulin AUC during an OGTT. Chi square test, analysis of variance and area under the curve were determined for comparing NASH with controls and for histological severity of NASH. Results. Patients with NASH had significantly higher (p<0.05) transaminases (ALT 69.8±6.9 vs. 21.7±2.1U/l) and lower HDL (45.1 ±2.0 vs. 52.1 ±2.5 U/l) compared to controls. Fasting and 2 h plasma insulin concentrations were significantly (p<0.05) higher in NASH (31.0±3.4 μIU/dl and 213.7±24.7μIU/l) than controls (18.1 ±1.7 μIU/dl and 135.7±16.3 μIU/dl). Glucose AUC0-30 and Glucose AUC90-120 were not significantly different between NASH and control subjects. HOMA IR (7.61 ±0.8 vs. 4.37±0.4), Matsuda ISI (1.47±0.1 vs. 2.15±0.2), and QUICKI (0.293±0.003 vs. 0.313±0.004) were significantly different (p<0.01) between NASH and controls. HIR measured by either method was not significantly different between NASH and controls. Insulin AUC90-120 was significantly higher (p<0.05) in patients with advanced NASH (defined using the ballooning and NAS scores). Conclusions. Compared to healthy controls, hyperinsulinemia and measures of peripheral IR rather than hepatic IR are increased in NASH. These data suggest that the skeletal muscle is a potential therapeutic target in NASH.

Disclosures:

The following people have nothing to disclose: Jaividhya Dasarathy, Ciaran E. Fealy, Carol A. Hawkins, Patricia T. Brandt, Arthur J. McCullough, John P. Kirwan, Srinivasan Dasarathy

648

  1. Top of page

Randomized controlled trial of omega 3 fatty acids in the treatment of non alcoholic steatohepatitis in type 2 diabetes mellitus

Srinivasan Dasarathy1, Jaividhya Dasarathy2, Amer Khiyami2, Lisa M. Yerian1, Ruth Sargent1, Carol A. Hawkins2, ArthurJ. McCullough1;
1Gastroenterology, Cleveland Clinic, Cleveland, OH; 2Gastroenterology, MetroHealth Medical Center, Cleveland, OH

Background. Type 2 diabetes mellitus (T2DM) occurs in 30% of nonalcoholic fatty liver disease (NAFLD) and is the major independent risk factor for advanced fibrosis and nonalcoholic steatohepatitis (NASH- the severe type of NAFLD). However, there is no established effective therapy for NASH patients with T2DM. N-3 polyunsaturated fatty acids (PUFA) are dietary supplements that have been shown in animal and human studies to have a beneficial effect on many of the comorbidities associated with NASH. Aim: To study the efficacy of PUFA on NASH patient with T2DM. Methods: A randomized double blind placebo controlled trial was performed in NASH patients with T2DM. 40 patients with well controlled diabetes (HbA1C<8.5%) were randomized to receive either polyunsaturated fatty acids (PUFA) containing eicosapentaenoic acid (EPA) 2160 mg and docosahexaenoic acid (DHA) 1440 mg daily or an isocaloric,identical placebo containing corn oil for 48 weeks. Clinical characteristics, biochemical labs, body composition using DEXA® and liver biopsy were done at randomization and at the end of treatment. The primary endpoint was a change of at least 2 points in the NASH CRN criteria. Liver biopsy was scored by a liver pathologist.. An intention to treat analysis was used to determine the response to treatment. Results. At inclusion, gender, age, liver biochemistries, HgbA1c, HOMA,lipids, BMI, waist circumference and each histologic component of the NAFLD activity score were similar in the 2 treatment groups. Thirty seven patients (18 PUFA and 19 placebo) completed the study. At the end of treatment, no significant differences were observed in the primary endpoint but a number of secondary endpoints (NAS score and insulin Conclusions. Despite strong animal and preliminary reports that PUFA may be beneficial in diseases associated with insulin resistance, no beneficial effects and possible adverse effects were observed in the present double blind, randomized controlled study in NASH patients with diabetes.

Disclosures:

The following people have nothing to disclose: Srinivasan Dasarathy, Jaividhya Dasarathy, Amer Khiyami, Lisa M. Yerian, Ruth Sargent, Carol A. Hawkins, Arthur J. McCullough

649

  1. Top of page

Effects of Lifestyle, Serum Adiponectin and Radiation on Prevalence and Progression of Nonalcoholic Fatty Liver Disease (NAFLD)

Waka Ohishi1,3, Keiko Ueda1,3, Yoshimi Tatsukawa1, Eiji Nakashima2, Michiko Yamada1, Ikuno Takahashi1, Masataka Tsuge3, Kazuaki Chayama3;
1Department of Clinical Studies, Radiation Effects Research Foundation, Hiroshima, Japan; 2Department of Statistics, Radiation Effects Research Foundation, Hiroshima, Japan; 3Department of Gastroenterology and Metabolism, Hiroshima University Hospital, Hiroshima, Japan

Background/Aims: An increase of non-B, non-C hepatocellular carcinoma (HCC) has been observed recently in Japan, and it is pointed out that NAFLD plays a role in the etiology of non-B, non-C HCC. In the previous study, we reported that alcohol consumption, smoking, obesity and radiation exposure were associated with increased risk of non-B, non-C HCC (Hepatology 53, 2011). In the present study, we conducted a cross-sectional study to investigate factors, which are associated with prevalence and progression of NAFLD in the longitudinal follow-up cohort of atomic-bomb survivors. Methods: The subjects of this study included 1,072 individuals (333 males and 739 females) after excluding, from among the Adult Health Study subjects who underwent health examinations during the period from 2008 to 2010, those with liver diseases (type B and C chronic liver disease, autoimmune liver disease, and HCC) and habitual drinkers (≥20 g/day in males, ≥10 g/day in females). NAFLD cases were diagnosed based on abdominal ultrasound findings and exclusion criteria. We analyzed association of prevalence and liver fibrosis severity (serum levels of hyaluronic acid and type IV collagen) of NAFLD with gender, age, lifestyle-related factors, serum levels of total adiponectin and radiation dose. Results: 1) Of the subject population, 367 people (34.2%) were diagnosed with NAFLD (225 females, or 61.3%), with the average age of 68.7 years. 2) Odds ratios (95% confidence interval (CI)) of factors independently associated with prevalence of NAFLD were 1.27 (1.21-1.33) for BMI per+1kg/m2, 1.7 (1.1 8-2.46) for diabetes, 1.78 (1.40-2.25) for ALT per +10 U/L, and 1.5 (1.08-2.09) for radiation dose at 1 Gy. Those estimates changed little when further adjustment was made for total adiponectin, and the odds ratio (95% CI) for ln (total adiponectin) per +1 unit was 0.23 (0.16-0.33). 3) Among the NAFLD cases, hyaluronic acid levels showed significant positive association with female, age, BMI, past smoking, and diabetes, and type IV collagen levels showed significant positive association with age, BMI, past and current smoking, diabetes, and levels of total adiponectin. Conclusions: Increased prevalence of NAFLD was associated independently with obesity, diabetes, elevated ALT levels, declined levels of total adiponectin, and radiation dose. Severity of liver fibrosis in NAFLD was also associated independently with aging, obesity, smoking habit, diabetes, and elevated levels of total adiponectin. These results suggest that factors such as aging, obesity, smoking habit, diabetes, and change of adiponectin level may be useful as indicators associated with progression of NAFLD.

Disclosures:

Kazuaki Chayama - Consulting: Abbvie; Grant/Research Support: Dainippon Sumitomo, Chugai, Mitsubishi Tanabe, DAIICHI SANKYO, Toray, BMS, MSD; Speaking and Teaching: Chugai, Mitsubishi Tanabe, DAIICHI SANKYO, KYORIN, Nihon Medi-Physics, BMS, Dainippon Sumitomo, MSD, ASKA, Astellas, AstraZeneca, Eisai, Olympus, GlaxoSmithKline, ZERIA, Bayer, Minophagen, JANSSEN, JIMRO, TSUMURA, Otsuka, Taiho, Nippon Kayaku, Nippon Shinyaku, Takeda, AJINOMOTO, Meiji Seika, Toray

The following people have nothing to disclose: Waka Ohishi, Keiko Ueda, Yoshimi Tatsukawa, Eiji Nakashima, Michiko Yamada, Ikuno Takahashi, Masataka Tsuge

650

  1. Top of page

Usefulness of combination of platelet count and AST/ALT ratio (PAAR index) for excluding advanced fibrosis in nonalcoholic fatty liver disease

Yoshio Sumida1, Saiyu Tanaka2, Hiroyoshi Taketani2, Kazuyuki Kanemasa2, Takeshi Nishimura1, Kanji Yamaguchi1, Hironori Mitsuyoshi1, Kohichiroh Yasui1, Masahito Minami1, Yoshito Itoh1;
1Department of Gastroenterology and Hepatology, Kyoto Prefectural University of Medicine, Kyoto, Japan; 2Center for Digestive and Liver Disease, Nara City Hospital, Nara, Japan

Background: A reliable and inexpensive noninvasive marker of hepatic fibrosis is required in patients with nonalcoholic fatty liver disease (NAFLD). AST/ALT ration (AAR) and platelet (PLT) count have been expected to detect or exclude advanced fibrosis in chronic liver diseases. The aim is to examine whether the combination of AAR and PLT is useful to detect or exclude advanced stages of NAFLD. Methods: A total of 259 patients (male: female= 134: 125, age 54.0±15.9 yr) with biopsy-proven NAFLD were involved in the present study. Advanced fibrosis was defined as stage 3 or 4 fibrosis according to Brunt's criteria. The areas under the receiver operating characteristic curves (AUROC) were compared.The Youden index was used to identify the optimal cutoff points. Results: A total of 139 subjects had steatohepatitis, of whom 64 subjects (11 %) had advanced fibrosis.The distribution of fibrosis stages included stage 0 (n = 120), stage 1 (n = 71), stage 2 (n = 39), stage 3 (n = 21), and stage 4 (n = 8). Patients with advanced fibrosis were significantly older, predominantly female, more likely to have lower levels of PLT and albumin, and higer levels of AST and AAR compared to those with no or mild fibrosis. The AUROC was greatest for PLT (0.784), followed by AAR (0.765), age (0.747), prothrombin time (0.712), albumin (0.687), AST (0.648), body mass index (0.560), and ALT (0.540). ROC analysis revealed that the optimal cutoff values of PLT and AAR to differentiate advanced stage from no or mild fibrosis were 195,000/μL (sensitivity 72%, specificity 75%) and 0.8 (sensitivity 72%,specificity 74%), respectively. Logistic regression analysis revealed AAR> 0.80 (odds ratio [OR]: 3.33, 95% confidence interval [CI]: 1.23-8.99, p=0.01 8) and PLT >195,000/ μL (OR: 2.08, 95% CI: 1.28-3.38, p=0.003) as independent parameters for predicting advanced stage of fibrosis. Negative predictive value was 98% for excluding advanced fibrosis in 143 patients with AAR<0.80 and PLT>195,000/ μL who can avoid liver biopsies. Positive predictive value was 37% for detecting advanced fibrosis in patients with AAR>0.80 and PLT<195,000/ μL (p<0.0001). Conclusions: NAFLD patients with AAR<0.80 and PLT>195,000/ μL, who are unlikely to have advanced fibrosis, can avoid liver biopsies. Although validation studies are essential in a larger population in multicenter institutions, the combination of PLT and AAPR, we call PAAR index, is useful and easily determined even in routine clinical practice or health checkups.

Disclosures:

Yoshito Itoh - Grant/Research Support: MSD KK, Bristol-Meyers Squibb, Dainippon Sumitomo Pharm. Co., Ltd., GlaxoSmithkline, Chugai Pharm Co., Ltd, Mitsubish iTanabe Pharm. Co.,Ltd.

The following people have nothing to disclose: Yoshio Sumida, Saiyu Tanaka, Hiroyoshi Taketani, Kazuyuki Kanemasa, Takeshi Nishimura, Kanji Yamaguchi, Hironori Mitsuyoshi, Kohichiroh Yasui, Masahito Minami

651

  1. Top of page

Association between Helicobacter pylori and Nonalcoholic Fatty Liver Disease in the General Population

Donghee Kim1, Seung Joo Kang1, Hwa Jung Kim2, Won Kim3, Yoon Jun Kim4, Jung-Hwan Yoon4;
1Department of Internal Medicine, Healthcare Research Institute, Seoul National University Hospital Gangnam Center, Seoul, Republic of Korea; 2Department of Clinical Epidemiology and Biostatistics, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; 3Department of Internal Medicine, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Republic of Korea; 4Department of Internal Medicine, Liver Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea

INTRODUCTION: Helicobacter pylori infection has been implicated in the pathogenesis of various gastrointestinal, hematologic, cardiovascular, and systemic diseases. Association between Helicobacter pylori infection and nonalcoholic fatty liver disease (NAFLD) is poorly characterized. The aim of this study was to investigate the association between H. pylori positivity with cagA status and NAFLD in the large, national, general population. METHODS: The Third National Health and Nutrition Examination Survey (NHANES) from 1988 to 1994 was utilized in this study. NAFLD was defined by ultrasonographic detection of hepatic steatosis without other known liver diseases. Antibodies to H. pylori and cagA of participants 20 years and older were measured using the H. pylori IgG and anti-cagA IgG ELISA. RESULTS: Among total of 11,808 participants who had result of both ultrasonography and H. pylori serology, the prevalence of NAFLD was 22.9%. The prevalence of NAFLD was higher in H. pylori positive subjects (33.5±1.79%) than in negative subjects (26.1 ±1.65%, p<0.001). Compared with cagA positive group, participants with negative cagA had higher prevalence of NAFLD (31.1 ±2.30% vs. 36.4±2.37%, p<0.001). Overall participants with NAFLD had higher prevalence of H. pylori positivity in multivariable analysis (Odds ratio [OR]: 1.17; 95% confidence interval [CI]: 0.95-1.43) with marginal significance. With regard to presence of cagA protein, H. pylori and cagA positivity was not associated with NAFLD (OR: 1.05; 95% CI: 0.81-1.37) but, cagA negative H. pylori positivity was significantly associated with NAFLD in multivariable analysis (OR: 1.30; 95% CI: 1.01-1.67). CONCLUSIONS: The prevalence of NAFLD was higher in H. pylori positive subjects than in negative subjects. Especially, cagA negative H. pylori positivity was significantly associated with NAFLD, independent of other known factors in the general population.

Disclosures:

The following people have nothing to disclose: Donghee Kim, Seung Joo Kang, Hwa Jung Kim, Won Kim, Yoon Jun Kim, Jung-Hwan Yoon

652

  1. Top of page

Gender-specific non-invasive diagnosis of liver fibrosis in nonalcoholic steatohepatitis

Rasha Alshaalan1, Marc Deschenes1, Peter Ghali1, Philip Wong1, Mazen Hassanain2,3, Ayat Salman4, Peter Metrakos4, Giada Sebastiani1;
1Department of Medicine, McGill University Health Centre, Montreal, QC, Canada;2Department of Oncology, McGill University Health Centre, Montreal, QC, Canada; 3Department of Surgery, King Saud University, Riyadh, Saudi Arabia; 4Department of Surgery, McGill University Health Centre, Montreal, QC, Canada

Staging of hepatic fibrosis and steatosis is vital for prognosis and interventions in non-alcoholic steatohepatitis (NASH). Liver biopsy, the gold standard, is invasive, costly and prone to error. Non-invasive methods for hepatic fibrosis and steatosis have been proposed but their validation in NASH is unsatisfactory. We conducted a retrospective study of consecutive patients with biopsy-proven NASH seen between 2007 and 2012 in our Unit. APRI, FIB-4 and NAFLD fibrosis score were used to diagnose liver fibrosis (>F2) and cirrhosis (F4). Ultrasound, Xenon-133 scan and hepatic steatosis index (HSI) were used to diagnose severe hepatic steatosis (>66%, S3). The cut-off values of the original reports were applied. Non-invasive tests were done within 6 months from liver biopsy, used as gold standard. Variables associated with each outcome were determined by multivariate logistic regression. The performance of non-invasive methods was expressed as sensitivity, specificity, positive and negative predictive values (PPV, NPV), area under the curve (AUC). We also modelled the best combination algorithm able to increase the accuracy of the single methods. Overall, 114 (mean age 49.6, 69.5% males) patients were included. Biopsy length range was 0.5-3.3cm, 57% of cases being >1.5cm. Fibrosis stages by Brunt were as follows: F0-F1=50%, F2=16.8%, F3 = 19.2%, F4=14%. Steatosis grades were as follows: S0-1=16%, S2=53.3%, S3=30.7%. The following variables were associated with the outcome measures: age (p<0.0001), diabetes (p=0.01) and steatosis (p=0.02) for >F2; female gender (p<0.05) and triglycerides (p=0.04) for F4; diabetes (p<0.05) and fibrosis (p=0.01) for S3. The performance of the non-invasive methods is depicted in the Table. Overall, the best method for detection of >F2 and F4 was FIB-4. Xenon scan outperformed the other methods but its AUC for S3 was <0.70. Notably, an algorithm combining gender and FIB-4 showed an AUC of 0.90, with 100% NPV to exclude cirrhosis. Conclusions: A gender-specific FIB-4 is highly predictive of cirrhosis in NASH. Severe steatosis cannot be reliably diagnosed by non-invasive methods. A gender-adjustment for more complex non-invasive fibrosis methods may be considered in future studies.

Table 5. 
 Sensitivity (%)Specificity (%)PPV (%)NPV (%)AUC
Fibrosis / cirrhosis (>F2 / F4)
APRI20/1266/9885/5056/870.78/0.86
FIB-421/5097/9385/5356/910.80/0.89
NAFLD fibrosis score27/6198/9392/5755/930.75/0.82
Steatosis (S3)
Ultrasound10014.632.91000.54
Xenon scan2785.346790.67
HSI4064.532.271.60.53

Disclosures:

Philip Wong - Advisory Committees or Review Panels: gilead, gilead, gilead, gilead; Grant/Research Support: merck, roche, merck, roche, merck, roche, merck, roche

The following people have nothing to disclose: Rasha Alshaalan, Marc Deschenes, Peter Ghali, Mazen Hassanain, Ayat Salman, Peter Metrakos, Giada Sebastiani

653

  1. Top of page

Efficacy and compliance of low carbohydrate diet education on nonalcoholic fatty liver disease in Korea

Dae Won Jun1, Ho Hyun Nam1, Jin-Hwa Moon2, Joo Hyun Sohn1, Tae Yeob Kim1;
1 Internal Medicine, Hanyang University, Seoul, Republic of Korea; 2Pediatrics, Hanyang University, Seoul, Republic of Korea

Background: Until now there is no specialized diet education program in nonalcoholic fatty liver disease (NAFLD). So, diet education program for obesity or dyslipidemia have been used to NAFLD patients in Korea. Both conventional diet programs mainly stressed on reducing fat consumption. However fat energy percent is less than 20% in Korea. We would like to investigate the efficacy and compliance of low carbohydrate diet in Korean NAFLD patients. Methods: One hundred and six NAFLD patients were enrolled from five hospitals. The patients were randomly selected to the conventional obesity diet program and low carbohydrate program. Liver chemistry, liver/spleen ratio, visceral fat CT scan, and serum CK-18 were measured at baseline and after 8 weeks. All participants completed five-day diet diary survey twice before and after diet education. Diagnosis of NAFLD was based on sonographic fat infiltration with elevated aminotransferase activity. Results: Both conventional diet program and low carbohydrate diet program decreased body weight and waist circumference. However, only low carbohydrate group showed significant decrease in ALT, AST, LDL-cholesterol, and blood pressure level compared to baseline. The ALT normalization at 8 weeks was 38.5% for the low carbohydrate and 16.7% for the low fat group (p=0.016). More than 80% of low carbohydrate group decreased serum ALT activity, while only 57% of conventional low fat group decreased ALT level compare to base line (p=0.012). Total abdominal fat area (401.3 ± 184.3 vs. 378.0±1 66.3, p=0.0001) and liver/spleen HU ratio (0.88±0.25 vs. 0.92±0.24, p=0.015) were decreased from the baseline in only low carbohydrate group. Not only carbohydrate consumption level but also total energy intake and fat consumption levels decreased more in low carbohydrate group than conventional anti-obesity program. Compliance of both two programs and physical activities during follow up period were not difference. Conclusions: Low carbohydrate diet program is more effective in reducing total energy intake and ALT normalization in NAFLD patients in Korea.

Disclosures:

The following people have nothing to disclose: Dae Won Jun, Ho Hyun Nam, Jin-Hwa Moon, Joo Hyun Sohn, Tae Yeob Kim

654

  1. Top of page

Study of hepatic leptin and insulin signaling during nonalcoholic fatty liver disease (NAFLD)progression in morbid obesity

Angel Carazo1, Laura Sanjuan3, Luis Miguel Alcázar1, Trinidad Caballero2, Ana Gila1,4, Jose Antonio Muñoz-Gámez1, Jesus Garcia-Rubio5, Antonio Cozar5, Manuel De la Mata6,4, Paloma Muñoz-de-Rueda1,4, Javier Salmeron1,4;
1Clinical Management of Digestive Diseases, UNAI, Hospital Universitario San Cecilio, Granada, Spain; 2Pathologic Anatomy, Hospital Universitario San Cecilio, Granada, Spain; 3Medicine Department, Granada University, Granada, Spain; 4CIBER Enfermedades Hepáticas y Digestivas, CIBERehd, Instituto de Salud Carlos III, Granada, Spain; 5Surgical Unit, Hospital Universitario San Cecilio, Granada, Spain; 6Clinical Management of Digestive Diseases, Hospital Universitario Reina Sofía, Córdoba, Spain

Introduction NAFLD is considered the hepatic exponent of metabolic syndrome, in which insulin resistance is the most important factor. Accordingly, hepatic and systemic insulin resistance have been reported in patients with NAFLD. Moreover, many arguments suggest that obese patients with NAFLD also develop hepatic leptin resistance. However, the molecular mechanisms that lead to liver leptin resistance have yet to be described. The main objective of this study is to analyze the relation between NAFLD progression during morbid obesity and the expression level of genes related to hepatic leptin and insulin signaling. Method Patient cohort included 87 morbidly obese subjects who underwent bariatric surgery. Liver biopsies were obtained at the moment of surgery. NAFLD was diagnosed by anatomopathological evaluation of liver biopsies using the Kleiner score. Hepatic gene expression level was estimated by measuring mRNA concentration using a methodology based on Real Time PCR (including normalization of mRNA concentration by three housekeeping genes). Seven expression levels were analyzed: long and short leptin receptor isoforms (OB-Rb and OB-Ra), insulin receptor (INS-R), ISR-1,IRS-2 (Insulin Receptor Substrate 1 and 2), SOCS-1 and SOCS-3 (Suppressor of Cytokine Signaling 1 and 3). Results The patients were classified into three groups: 10 without NAFLD (group 1); 33 with liver steatosis but without steatohepatitis (group 2, Kleiner score<3) and 44 with probable or confirmed steatohepatitis (group 3, Kleiner score≥3). Comparison of the average gene expression levels in obese patients without NAFLD (group 1) and in those with NAFLD (groups 2 and 3) revealed a non-significant tendency toward a decrease in leptin and insulin receptors, IRS-1 and SOCS-3 (p<0.1), and a non-significant tendency toward an increase in SOCS-1 (p<0.1). Moreover, the patients without NAFLD presented a marked degree of correlation between the expression of leptin and insulin signaling-related genes (above 0.8 in all combinations and a maximum of 0.986 for both leptin receptor isoforms). Interestingly, all these genetic correlation levels decreased or disappeared as NAFLD progressed. Conclusions In morbidly obese patients without NAFLD, high levels of correlation between leptin and insulin signaling-related genes suggest that hepatic leptin and insulin signaling pathways share key expression factors. Moreover, the reduction in these correlations as NAFLD progresses suggests that liver endocrine homeostasis is affected by NAFLD development. This study presents new perspectives on the mechanism that gives rise to leptin and insulin hepatic resistance during NAFLD origin and progression.

Disclosures:

The following people have nothing to disclose: Angel Carazo, Laura Sanjuan, Luis Miguel Alcázar, Trinidad Caballero, Ana Gila, Jose Antonio Muñoz-Gámez, Jesus Garcia-Rubio, Antonio Cozar, Manuel De la Mata, Paloma Muñoz-de-Rueda, Javier Salmeron

655

  1. Top of page

Combined serum biomarker in non-invasive diagnosis of non-alcoholic steatohepatitis

Jingmin Zhao1, Mei Yang1, Fangli Zhang1, Wenshu Li2, Jin Li3, Jiyun Lv3;
1 Department of Pathology and Hepatology, Beijing 302 Hospital, Beijing, China;2Department of Clinical Trial, Beijing 302 Hospital, Beijing, China; 3Beijing 302 Hospital, Beijing, China

Background & Aims: Non-alcoholic Fatty Liver Diseases (NAFLD), especially its critical stage of non-alcoholic steatohepatitis (NASH), has become one of the most important public health issues worldwide. Although liver biopsy remains the Golden Standard for the diagnosis of nonalcoholic steatohepatitis (NASH), non-invasive methods are eagerly needed. Our study aims at testing the accuracy of cytokeratin-18 (CK-18) fragment, fibroblast growth factor 21 (FGF21), interleukin 1 receptor antagonist (IL-1 Ra), Pigment epithelium-derived factor (PEDF) and osteoprotegerin (OPG) for the diagnosis of NAFLD and NASH. Methods: 98 patients with biopsy-proven NAFLD and 43 age- and gender-matched healthy controls without evidence of liver disease were included. Serum CK-18, FGF21, IL-1 Ra, PEDF and OPG levels were measured by enzyme-linked immunosorbent assay. Results: The levels of Serum CK-18, FGF21, and IL-1 Ra increased in a stepwise fashion control subjects (median 7.52 ng/L, 12.65 ng/L, 39.68 ng/L), non-NASH NAFLD patients (13.12 ng/L, 32.03 ng/L, 99.36 ng/L) and NASH patients (29.11 ng/L, 62.92 ng/L, 239.36 ng/L) (p<0.01). The concentration of PEDF and OPG decreased successively from NASH patients (median 16.78 ug/L, 139.57 ng/L), to non-NASH NAFLD patients (38.42 ug/L, 373.90 ng/L) and to controls (49.50 ug/L, 516.98 ng/L, respectively) (P < 0.01). In the diagnosis of NASH, the 5 serum biomarkers all displayed statistical significance(p<0.01). Among them, only OPG could differentiate simple fatty liver from control group (P=0.011), and only PEDF can differentiate borderline NASH from other groups (p<0.05). The area under receiver-operating characteristics curve to diagnose NASH was 0.86 for CK-18, 0.81 for FGF21, 0.92 for IL-1 Ra, 0.90 for PEDF and 0.94 for OPG. At cut-off of 171.79ng/L, the negative predictive value (NPV) for NASH of IL-1 Ra was 82.16% and positive predictive value (PPV) was 83.22%. The use of a cut-off level was defined as 28.15ug/L for serum PEDF levels yielded sensitivity and specificity values of 87.9% and 94.9%, respectively. When the cut-off was defined as 242.96 ng/L for OPG, the sensitivity, specificity, PPV and NPV were 93.26%, 97.4%, 92.35% and 97.96%, respectively. We also obtained a predictive model of NASH using regression analysis: logitP=22.1 178-0.0413OPG+0.0159CK-18+0.0212FGF21-0.053PEDF (P < 0.05). Conclusion: CK-18, FGF21, IL-1 Ra, PEDF and OPG may serve as noninvasive biomarkers combination to identify NASH patients. PEDF and OPG may be the indicators of the development of NAFLD. The NASH prediction model can be an innovation in the prediction and diagnosis of NASH.

Disclosures:

The following people have nothing to disclose: Jingmin Zhao, Mei Yang, Fangli Zhang, Wenshu Li, Jin Li, Jiyun Lv

656

  1. Top of page

Is there a link between hypothyroidism and nonalcoholic fatty liver disease?

Fernando Bril1, Romina Lomonaco1, Beverly Orsak2, Sreevidya Subbarayan1, Sushma Kadiyala1, Carolina Ortiz-Lopez2, Amy Webb4,3, Kenneth Cusi1,2;
1Endocrinology, Diabetes and Metabolism, University of Florida, Gainesville, FL; 2Diabetes, University of Texas Health Science Center at San Antonio, San Antonio, TX; 3Diabetes, Audie L. Murphy Veterans Administration Medical Center, San Antonio, TX; 4Gastroenterology and Nutrition, University of Texas Health Science Center at San Antonio, San Antonio, TX

Background and Aim: Previous studies have reported that hypothyroidism may play a role in the development of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH). However, data have frequently been inconsistent due to small sample size and/or lack of liver histopathology. The aim of this study was to assess the role of hypothyroidism in NAFLD. Methods: We measured in 253 patients (51±1 years, 66% male, 33.2±0.3 kg/m2, 49% with type 2 diabetes mellitus [T2DM]) plasma TSH and free T4 together with measurement of: 1) liver fat content by magnetic resonance imaging and spectroscopy (MRS), 2) insulin resistance during an euglycemic insulin clamp with glucose turnover determinations, and 3) a liver biopsy (n=145) to assess liver histology. Hypothyroidism was defined as having a previous diagnosis of the disease and receiving LT4 replacement, or newly diagnosed patients with hypothyroidism according to TSH and free T4 levels during the initial screening. Results: Patients were divided according to the presence (n=209) or absence (n=44) of NAFLD by MRS. Patients with a fatty liver showed a worse metabolic profile with higher BMI, more frequent T2DM and more severe insulin resistance. They also had a higher prevalence of hypothyroidism (18 [6.7%] vs. 0 [0%], p<0.05). When patients with biopsy-proven NAFLD were divided into those with and those without NASH, no difference in the prevalence of hypothyroidism was observed (14 [10.3%] vs. 2 [4.3%], p=0.37). After excluding patients on levothyroxine replacement, plasma TSH and free T4 levels were similar between patients with and without NAFLD, and those with and without NASH. Patients were also divided into those with and without hypothyroidism to assess if it had any impact in the development of NAFLD or NASH. Patients with hypothyroidism showed a similar amount of liver fat (21±2% vs. 20±1%, p=0.79), NAFLD activity score (NAS) (3.8±0.3 vs. 4.0±0.1, p=0.74) and liver fibrosis (0.8±0.3 vs. 0.8±0.1, p=0.92). Finally, TSH and free T4 plasma levels were not correlated to insulin sensitivity, liver fat by MRS or the severity of histological damage. Conclusions: We found a slight increase in the prevalence of hypothyroidism in patients with NAFLD. However, in this cohort of middle-aged predominantly obese patients we did not find any suggestion that hypothyroidism (or a TSH elevation within the normal range) worsens liver disease (NAS or fibrosis) in patients with NASH. This suggests that the association of hypothyroidism with NAFLD may be more closely linked to obesity and the associated metabolic abnormalities that lead to steatosis rather than the severity of liver disease.

Disclosures:

Beverly Orsak - Employment: UTHSCSA

Kenneth Cusi - Consulting: Merck, Daichi-Sankyo, Roche, Janssen; Grant/Research Support: Takeda, Novartis, Mannkind

The following people have nothing to disclose: Fernando Bril, Romina Lomonaco, Sreevidya Subbarayan, Sushma Kadiyala, Carolina Ortiz-Lopez, Amy Webb

657

  1. Top of page

Vitamin D deficiency and development of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH)

Fernando Bril1, Romina Lomonaco1, Beverly Orsak2, Carolina Ortiz-Lopez2, Diane Biernacki1,3, Ashley Klaczak1, Zhi Chang2, Jean Hardies2,4, Kenneth Cusi1,2;
1 Endocrinology, Diabetes and Metabolism, University of Florida, Gainesville, FL; 2Diabetes, University of Texas Health Science Center at San Antonio, San Antonio, TX; 3Diabetes, Malcom Randall Veterans Administration Medical Center, Gainesville, FL; 4Diabetes, Audie L. Murphy Veterans Administration Medical Center, San Antonio, TX

Background and Aim: The true role of vitamin D deficiency in the development of non-alcoholic fatty liver (NAFLD) and steatohepatitis (NASH) remains poorly understood. Previous studies have been overall small, retrospective or relied on surrogate markers for the diagnosis of NAFLD and NASH. We aimed to assess the relationship between vitamin D deficiency and NAFLD. Methods: We recruited 235 patients (52±1 years, 67% male, 33.4±0.3kg/m2, 57% with type 2 diabetes mellitus) and assessed: 1) plasma vitamin D levels and other measures of bone metabolism; 2) fasting plasma glucose and insulin (HOMA index); 3) tissue-specific measurements of insulin resistance during an euglycemic insulin clamp with 3-3H-glucose; 4) total body fat (TBF) by DXA scan; 5) liver fat by magnetic resonance imaging and spectroscopy (MRS); and 6) liver histology (biopsy) (n=151). Results: Vitamin D deficiency (<20ng/ml) was frequent in our cohort (n=167; 70%). Patients with vitamin D deficiency showed higher BMI (34.0±0.4 vs. 32.1 ±0.6 kg/m2, p=0.01) and liver fat (21.4±1.1 vs. 16.7±2.1%, p=0.04). However, they had a similar degree of insulin resistance (HOMA index: 4.2±0.3 vs. 4.0±0.4 mg/dl . μU/ml, p=0.66), ALT/AST levels (54±3 vs. 48±4 and 40±2 vs. 38±2 U/L, p=0.21 and 0.58, respectively), liver histology (NAFLD activity score: 3.7±0.2 vs. 4.2±0.2, p=0.09) and fibrosis (0.8±0.1 vs. 1.2±0.2, p=0.13). When patients were divided according to their NAFLD or NASH status, vitamin D levels were similar between patients with and without NAFLD (16.5±0.5 vs. 18.6±1.5 ng/ml, p=0.13) and with and without NASH (18.4±0.8 vs. 17.9±1.3 ng/ml, p=0.76). To further assess the possible link between vitamin D and liver disease, we assessed the correlations between plasma vitamin D concentration and BMI, TBF, liver fat by MRS, and liver histology. There was no significant correlation between vitamin D levels and BMI (r=-0.11), TBF (r=0.05), liver fat by MRS (r=-0.09), steatosis (r=-0.02), inflammation (r=-0.13), ballooning (r=-0.02) or fibrosis (r=0.01). Vitamin D levels did not correlate with any specific measure of insulin resistance in patients with NAFLD. Conclusions: Vitamin D levels are not associated with liver fat accumulation or the histological severity of NASH. Moreover, vitamin D did not show any association with measures of insulin sensitivity, which are thought to play an important role in NAFLD development. The link between vitamin D and obesity (and secondarily to the metabolic syndrome and NAFLD) may be due to the common presence of sedentarism that promotes both obesity and vitamin D deficiency, rather than to a pathophysiologic role of vitamin D.

Disclosures:

Beverly Orsak - Employment: UTHSCSA

Kenneth Cusi - Consulting: Merck, Daichi-Sankyo, Roche, Janssen; Grant/Research Support: Takeda, Novartis, Mannkind

The following people have nothing to disclose: Fernando Bril, Romina Lomonaco, Carolina Ortiz-Lopez, Diane Biernacki, Ashley Klaczak, Zhi Chang, Jean Hardies

658

  1. Top of page

Pilot Study on the Role of Multidisciplinary Approach in Risk Reduction of Risk of Non- alcoholic Fatty Liver in Subjects with Eating Disorders

Consuelo Cefalo1, Annamaria Strangio2, Luca Miele1,3, Lucio Rinaldi2, Giuseppe Marrone1, Simona Racco1, Andrea Zanché1, Gian Ludovico Rapaccini1,3, Pietro Bria2, Antonio Gasbarrini1, Antonio Grieco1;
1 Internal Medicine, Università Cattolica del Sacro Cuore, Rome, Italy; 2Psychiatric Medicine, Università Cattolica Del Sacro Cuore, Rome, Italy; 3Association Columbus, Rome, Italy

INTRODUCTION: Excessive caloric intake in patients with eating disorders is possible nutritional causes of nonalcoholic fatty liver disease (NAFLD). Therapeutic approach in patients with NAFLD includes weight loss, physical activity, drugs, surgery and control of cardiovascular risks. Some data show an improvement in metabolic and hepatic parameters of NAFLD patients subjected to a multidisciplinary approach. AIMS&METHODS: Purpose of our pilot study was to evaluate consecutive patients enrolled in a tertiary referral centre for eating disorders from a metabolic point of view, define the risk of NAFLD and assess the effect of an integrated multidisciplinary approach on NAFLD risk, hepatic and metabolic parameters and psychiatric disorder. We recorded anthropometric, anamnestic and hematochemical data for all 39 patients enrolled, at the beginning and at the end of the twelve months of treatment. Each patient underwent two psychiatric tests at baseline and endpoint of the study: E.D.I. (Eating Disorders Inventory) and Q.E.B. (Questionnaire of Eating Behaviours). NAFLD risk was evaluated using the Fatty Liver Index (FLI) based on assessment of γGT, Body Mass Index (BMI), triglycerides and waist circumference at the beginning and at the end of study. RESULTS: At the end of treatment an improvement in BMI (35.16 ± 10.13vs. 33.94 ±8.99; p= 0.05) and median waist circumference (97 vs. 103 cm) was evident, and Metabolic Syndrome (MS) prevalence was reduced. Liver parameters improved (γGT value 31.85± 36.80 vs. 22.68 ± 13.83; p=0.006) and the FLI score (66.00±36.82 vs. 64.29±37.55); p= 0.003) decreased significantly. Four of the eight EDI sub-scales (Drive for thinness (p:0.008), Interoceptive awareness (p<0.001), Bulimia (p:0.001), Ineffectiveness (p:0.014)), and two of the three QEB subscales (Binge Eating (p:0.001) and Food Restriction (p:0.016)) improved. CONCLUSION: In conclusion the pilot study showed an increased risk of NAFLD in eating disorder patients and a significant benefit in using multidisciplinary approach to improve metabolic, hepatological and psychiatric outcome.

Disclosures:

The following people have nothing to disclose: Consuelo Cefalo, Annamaria Strangio, Luca Miele, Lucio Rinaldi, Giuseppe Marrone, Simona Racco, Andrea Zanché, Gian Ludovico Rapaccini, Pietro Bria, Antonio Gasbarrini, Antonio Grieco

659

  1. Top of page

Plasma cathepsin D: a novel marker for non-invasive diagnosis of non-alcoholic steatohepatitis

Sofie Walenbergh1, Sander Rensen1, Veerle Bieghs1, Tim Hendrikx1, Patrick van Gorp1, Mike Jeurissen1, Wim Buurman1, Anita Vreugdenhil2, Jan-Willem Greve3, Jogchum Plat1, Marten H. Hofker4, Patrick Lindsey1, Ger H. Koek2, Ronit Shiri-Sverdlov1;
1Maastricht University, Maastricht, Netherlands; 2Maastricht University Medical Center, Maastricht, Netherlands; 3Atrium Medical Center Parkstad, Heerlen, Netherlands;4University Medical Center Groningen, Groningen, Netherlands

Background & Aims: Non-alcoholic steatohepatitis (NASH) is characterized by liver lipid accumulation and inflammation. Currently there are no specific non-invasive markers to detect NASH. Laboratory abnormalities reflect mostly liver cell damage but not early inflammation, leaving an invasive liver biopsy as the only gold standard to diagnose NASH. We previously demonstrated a clear association between hepatic inflammation and increased lysosomal cholesterol accumulation inside Kupffer cells of hyperlipidemic mice. Lysosomal cholesterol accumulation is often associated with disturbances in trafficking of lysosomal enzymes and consequently, with elevated levels of lysosomal enzymes in plasma. We hypothesized that NASH patients have increased levels of lysosomal enzymes in plasma compared to subjects with a healthy or steatotic liver. Methods: Liver biopsies from morbidly obese subjects were histologically evaluated for NASH. Cathepsin D (CatD), the most abundant lysosomal protease, and alanine aminotransferase (ALT) were measured in plasma and tested statistically by using receiver operating characteristic analysis. Results: In contrast to ALT, plasma CatD was significantly increased in NASH patients compared to subjects with either steatosis or a normal liver. Whereas ALT demonstrated to be a late marker for NASH grade (grade 2 and 3), CatD was elevated at early inflammation (grade 1). The sensitivity and specificity of ALT for detecting hepatic inflammation improved markedly through addition of CatD. Conclusions: The combination of CatD and ALT in plasma is a potential, specific non-invasive marker to assess NASH and to monitor disease progression.

Disclosures:

Jan-Willem Greve - Consulting: GI Dynamics; Grant/Research Support: GI Dynamics

The following people have nothing to disclose: Sofie Walenbergh, Sander Rensen, Veerle Bieghs, Tim Hendrikx, Patrick van Gorp, Mike Jeurissen, Wim Buurman, Anita Vreugdenhil, Jogchum Plat, Marten H. Hofker, Patrick Lindsey, Ger H. Koek, Ronit Shiri-Sverdlov

660

  1. Top of page

Comparison between Results of Hepatic Transient Elastography (fibroscan®) and Controlled Attenuation Parameter (cap™) versus Liver Biopsy in NAFLD Patients

Denise S. Vanni, Claudia P. Oliveira, Daniel F. Mazo, Fabiola Rabelo, José Tadeu Stefano, Flair J. Carrilho;
Gastroenterology, University of São Paulo School of Medicine, São Paulo, Brazil

BACKGROUND AND AIMS: The aim of this study was to compare the results of Fibroscan® and CAP™ versus liver biopsy in patients with Non Alcoholic Fatty Liver Disease (NAFLD). METHODS: We enrolled patients with NAFLD diagnosed by liver biopsy between May of 2011 and January of 2013 at Sao Paulo University Hospital. They underwent liver stiffness measurements to assess fibrosis by Fibroscan® using median and extra large probes according to their skin-liver distance. CAP™ was also used to assess steatosis when Fibroscan® measures were made with the median probe. The Fibroscan® was operated by 2 experts in the procedure. The time frame between liver biopsy and Fibroscan® plus CAP™ was of sixty days at most. We considered failure of Fibroscan® and CAP™ when: we couldn't have ten valid measures; the total success rate was below 60% and/or the interquartile range (IQR) was above 30%. The results of these noninvasive methods were compared with liver histology (BRUNT criteria), used as the reference standard. The corresponding values of Fibroscan®(kPa) to fibrosis stages and of CAP™ (dBm-1) to steatosis grades considered were based in previous studies of these methods in NAFLD patients. The gamma distribution function was used to compare the results of Fibroscan® and CAP™ versus liver biopsy. RESULTS: A total of 65 patients were enrolled, 71 % female and 29% male with mean age of 56 years old (1 3-71 years). Mean body mass index (BMI) and abdominal circumference were 31.29Kg/m2 (19.6-47.7Kg/m2) and 102.3cm (77-135cm), respectively. Mean distance between skin surface and liver was 2.06cm (0.98-4.26cm). Patient's comorbidities were: 46% diabetes; 73% dyslipidemia; 60% systemic arterial hypertension. The Fibroscan® was feasible in 83 %(95%CI: 0.7193 -0.9039) of a total of 65 patients and CAP™ was feasible in 74% (95%CI: 0.603 - 0.848) of a total of 47 patients, respectively. The results of comparison between Fibroscan®, CAP™ and liver biopsy (noninvasive methods evaluated separately) using gamma distribution function were: Fibroscan® gamma= 0.38(95%CI 0.09-0.66) and CAP™ gamma= 0.46(95%CI 0.06-0.85). CONCLUSIONS: Fibroscan® and CAP™ were feasible in most of NAFLD patients studied. However, the agreement of these noninvasive methods when compared to liver histology was unsatisfactory. Besides the prevalence of high BMI, perhaps the heterogeneous fibrosis and steatosis of NAFLD liver histology were to blame.

Disclosures:

The following people have nothing to disclose: Denise S. Vanni, Claudia P. Oliveira, Daniel F. Mazo, Fabiola Rabelo, José Tadeu Stefano, Flair J. Carrilho

661

  1. Top of page

Hepatic microRNAs are differentially expressed in nonalcoholic fatty liver disease patients compared to healthy controls and expression levels correlate with hepatic omega-3 polyunsaturated fatty acids

Elena M. Comelli1, Laura Bosco1, TaeHyung Kim2,3, Wendy Lou4, Zhaolei Zhang2,3, David W. Ma5, Bianca M. Arendt6, Scott Fung6,7, Johane P. Allard6,7;
1Nutritional Sciences, University of Toronto, Toronto, ON, Canada; 2Computer Sciences, University of Toronto, Toronto, ON, Canada; 3Molecular Genetics, University of Toronto, Toronto, ON, Canada; 4Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada; 5Human Health and Nutritional Sciences, University of Guelph, Guelph, ON, Canada; 6Toronto General Hospital, University Health Network, Toronto, ON, Canada; 7Medicine, University of Toronto, Toronto, ON, Canada

Introduction: Nonalcoholic steatohepatitis (NASH) has been associated with low levels of hepatic polyunsaturated fatty acids (PUFA), particularly omega-3 PUFA. There is no data on whether omega-3 PUFA modulate microRNA (miRNA) expression in liver. Some studies have reported altered miRNA expression in obese patients with NASH but there are no studies on whether there is an association with PUFA. The aim of this study was 1) to compare miRNA expression between patients with simple steatosis (SS), NASH, and healthy controls (HC), and 2) to examine correlations between hepatic miRNA expression and omega-3 PUFA. Methods: miRNA expression was measured by NanoString technology in liver tissue from 24 living liver donors as healthy controls (HC; n= 24) and patients with biopsy proven SS (n= 25) or NASH (n= 22). Groups were compared by t-test ( p<0.001). Polyunsaturated fatty acids in hepatic total lipids were measured by gas chromatography. Results are given in % of total fatty acids. Spearman correlations were used to identify potential associations. Results: Twenty-six miRNAs were differentially expressed between HC, SS and NASH, including miR1 0b which was upregulated in HC vs NASH (p=0.00001). Total omega-3 PUFA were lower in NASH (mean± SD) (2.35±0.65 %) and SS (3.28±1.23 %) compared with HC (4.44±1.61 %) (p<0.05). Docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), the biologically active long-chain omega-3 PUFA, were also lower in NASH and SS than in HC (p<0.05). Twenty of the differentially expressed miRNAs were significantly correlated with at least one omega-3 PUFA, including miR1 0b which was positively correlated with DHA (r=0.417, p=0.001) and total omega-3 PUFA (r=0.343, p=0.008). Conclusion: The expression of miR1 0b was higher in HC than NASH and positively correlated with hepatic omega-3 PUFA. A potential target of miR1 0b is peroxisome proliferator-activated receptor-α, which can contribute to steatogenesis and inflammation in NAFLD. These results support the concept of associations between PUFA, epi-genetic mechanisms, and NAFLD-related gene expression. Further studies are required to establish cause-effect relationships and examine the potential of omega-3 PUFA supplementation to regulate miRNA in NAFLD.

Disclosures:

David W. Ma - Advisory Committees or Review Panels: Heinz; Consulting: Vegetable Oils Industry of Canada, PepsiCo; Speaking and Teaching: Unilever

Scott Fung - Advisory Committees or Review Panels: Gilead Sciences, Vertex; Grant/Research Support: Merck, Roche; Speaking and Teaching: BMS, Gilead

The following people have nothing to disclose: Elena M. Comelli, Laura Bosco, TaeHyung Kim, Wendy Lou, Zhaolei Zhang, Bianca M. Arendt, Johane P. Allard

662

  1. Top of page

Are HCV + transplant recipients disproportionately affected by donor graft steatosis?

Neil G. Kumar, M. Katherine Dokus, Randeep Kashyap, Mark S. Orloff;
Solid Organ Transplant, University of Rochester Medical Center, Rochester, NY

Background: Hepatitis C (HCV) is the most frequent indication for liver transplantation. Several donor factors have been identified influencing post-transplant outcomes; however the impact of donor graft steatosis is debated. The aim of this study is to assess the impact of donor graft steatosis on patient and graft survival in HCV+ recipients after transplantation Methods: We reviewed the clinical course of all adult primary liver transplants from 2002 - 2010. 448 patients were included in the final analysis. Patients were grouped according to their HCV status (+/-) and level of donor steatosis (>30% or ≤30%); group 1: HCV-/DSteatosis≤30%; 2: HCV+/DSteatosis≤30%; 3: HCV-/DSteatosis>30%; 4: HCV+/DSteatosis>30%. Survival was analyzed with univariate statistics and regression models and correlated with donor and recipient characteristics; associations were included in the final multivariate model. Results: Patients were followed up for a median of 60 months. Overall patient and graft survival was significantly different across the 4 groups: graft 78.7%, 70.3%, 71.8%, 36% (p=0.01); patient: 87.2, 79.7, 79.7, 45.6% for group 1, 2, 3 and 4 respectively (p=0.02). HCV positive patients who received a graft with more than 30% steatosis demonstrated the worst overall graft and those with non-HCV diagnosis and ≤30% steatosis had superior outcomes to all other groups. This held true after multivariate adjustment graft (p=0.02) patient survival (p= 0.03) (figure 1). Conclusions: Donor graft steatosis adversely affects patient and graft survival after liver transplantation. Survival is further diminished in HCV positive recipients when steatosis is greater than 30%. Liver transplantation with >30% steatotic grafts should be carefully considered in all recipients, especially those with HCV cirrhosis.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Neil G. Kumar, M. Katherine Dokus, Randeep Kashyap, Mark S. Orloff