SEARCH

SEARCH BY CITATION

2079

  1. Top of page

Prospective study of the relationship between progression pattern and survival in patients with hepatocellular carcinoma (HCC). The BCLC rationale for second line trial design and analysis

Maria Elisa Reig1, Jordi Rimola2, Ferran Torres3, Anna Darnell2, Carlos Rodriguez de Lope1, Alejandro Forner1, Neus LLarch1, José Ríos3, Carmen Ayuso2, Jordi Bruix3;
1 Barcelona Clinic Liver Cancer (BCLC) Group, Liver Unit, Hospital Clínic Barcelona, IDIBAPS, University of Barcelona, Spain .Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digest, Hospital Clinic, Barcelona, Spain; 2Barcelona Clinic Liver Cancer (BCLC) Group, Radiology Department, Hospital Clínic Barcelona, IDIBAPS, University of Barcelona, Spain., Hospital Clinic, Barcelona, Spain; 3Biostatistics and Data Management Platform, IDIBAPS, Hospital Clínic; Biostatistics Unit, School of Medicine, Universitat Autònoma de Barcelona; Barcelona, Hospital Clinic, Barcelona, Spain

Background: Sorafenib delays time to tumor progression (TTP) and improves survival (OS) in HCC patients. However, despite the use of TTP as a surrogate of OS, this has never been proven. Indeed, OS predictors upon radiologic progression are not established and it is needed to characterize post-progression survival (PPS) and assess with proper time-dependent covariate analysis if PPS is influenced by progression pattern, as well as by simultaneous impairment of liver function and performance status (PS). Our aim was to prospectively validate tumor progression as a surrogate of OS and evaluate the impact of progression pattern in OS and PPS of sorafenib treated patients. Methods: HCC patients under sorafenib were followed clinically and biochemically every 4 weeks. Radiologic assessment of progression (RECIST1.1) was done at week 4 and then, every 8 weeks. Progression pattern was divided into: [intra-hepatic (IHG)/extrahepatic(EHG) increase in tumor size, new intra-hepatic lesion (NIH), new extra-hepatic lesion/vascular invasion (NEH)]. To define the predictors of OS we registered baseline parameters [PS (0/1), Child Pugh score (A/B7points), BCLC (B/C), extra-hepatic spread (yes/no), total bilirubin, albumin, AFP and prior treatment] as well as the impact of the transition from Child-Pugh A (used as reference) into Child-Pugh B or C, sorafenib dose modification, change in PS, presentation of encephalopathy or untreatable ascites, decrease in pro-thrombin time <50%, albumin <2.8 mg/dl, and AFP. Inferential analysis for time to event data used the Cox uni- and multi-vari-ate regression model with time-dependent covariates to estimate hazard ratios (HR) and 95%CI. Results: We included 147 patients (HCV 57%, PS 0 84%, Child-Pugh A 82% and BCLC-C 47%). Median OS is 12.7 months and its independent predictors (HR [95%CI]) are: baseline BCLC 2.49[1.66–3.73], PS 1.86[1.12–3.10], registration during follow-up of Child-Pugh B orC scores (2.36[1.51–3.69] and 2.89[1.62–5.15], respectively) definitive sorafenib interruption 2.48[1.54–4.01] and tumor progression 3.39[1.89–6.1]. Follow-up development of NEH 2.42[1.32–4.44] emerged as independent predictor of OS and PPS. Conclusion: Our results establish for 1st time the correlation between tumor progression and OS. This supports the use of time to progression as an early signal of activity for new agents. However, the negative impact of progression in PPS is restricted to development of new extra-hepatic sites or appearance of vascular invasion, even if BCLC-C at baseline. Thus, prognostic prediction and 2nd line trial design and analysis should take into account progression pattern to avoid vulnerable results.

Disclosures:

Maria Elisa Reig - Grant/Research Support: Bayer, Bayer, Bayer, Bayer

Carlos Rodriguez de Lope - Speaking and Teaching: Bayer

Alejandro Forner - Advisory Committees or Review Panels: Bayer HealthCare, Bayer HealthCare, Bayer HealthCare, Bayer HealthCare; Speaking and Teaching: Bayer HealthCare, Bayer HealthCare, Bayer HealthCare, Bayer HealthCare

Carmen Ayuso - Speaking and Teaching: Bayer

The following people have nothing to disclose: Jordi Rimola, Ferran Torres, Anna Darnell, Neus LLarch, José Ríos, Jordi Bruix

2080

  1. Top of page

The impact of pegylated interferon based therapy after radiofrequency ablation in hepatocellular carcinoma patients with hepatitis C virus

Takamasa Ohki1, Shinpei Sato2, Koki Sato1, Eriko Goto6, Jun Ima-mura4, Hiroyoshi Taniguchi5, Masatoshi Akamatsu3, Hideo Yoshida5, Shuntaro Obi2, Nobuo Toda1, Yukihiro Koike6;
1Gas-troenterology, Mitsui Memorial Hospital, Tokyo, Japan; 2Hepatol-ogy, Kyoundo Hospital, Tokyo, Japan; 3Gastroenterology, JR Tokyo General Hospital, Tokyo, Japan; 4Gastroenterology, Tokyo Metropolitan Cancer and Infectious diseases Center Komagome Hospital, Tokyo, Japan; 5Gastroenterology, Japanese Red Cross Medical Center, Tokyo, Japan; 6Gastroenterology, Kanto Central Hospital, Tokyo, Japan

Aim: Hepatocellular carcinoma (HCC) frequently recurs even after curative treatment with radiofrequency ablation (RFA). The aim of this study is to investigate the association between antiviral pegylated interferon (PEG-IFN) based therapy after curative treatment with RFA and risk of HCC recurrence in patients with hepatitis C virus (HCV). Method: Between January 2000 and December 2011, a total of 1143 HCV related HCC patients were curatively treated with RFA in our institutes. Among them, PEG-IFN was only administered in 27 patients. We used propensity score-matching to compare patients who received PEG-IFN therapy after curative RFA vs. who did not receive. Their clinical data, recurrence rates, and survival rates, were compared. We also assessed the prognostic values of recurrence and survival rates using multivariate Cox proportional hazard models. Result: PEG-IFN was only administered in 27/1143 (2.4%) patients and 15 patients (55.6%) achieved sustained virological response. Finally, according to the propensity matched analysis, a pair of 24 patients were chosen; PEG-IFN treated HCC group (n = 24) and the controls (n = 24). There were no significant differences in liver function-related factors, and tumor-related ones between the two groups. The cumulative recurrence rates significantly differ between the two groups (P = 0.03); 16.7%, 52.5%, and 62.0% at 1, 3, and 5 years, respectively, in the PEG-IFN treated HCC group and 46.2%, 75.3%, and 100% in the controls. The cumulative survival rates also significantly differ between the two groups (P < 0.01); 100%, 95.5%, and 90.4% at 1, 3, and 5 years, respectively, in the PEG-IFN treated HCC group and 91.7%, 68.7%, and 51.5% in the controls. Multivariate analysis indicated administration of PEG-IFN (HR 0.39, P < 0.01), and DCP > 40 mAU/mL (HR 2.73, P < 0.001) as independent risk factors of HCC recurrence, and indicated only administration of PEG-IFN (HR 0.21, P < 0.01) as an independent risk factor of survival. Conclusion: PEG-IFN based HCV eradication therapy after curative treatment with RFA dramatically reduce HCC recurrence and improve survival time.

Disclosures:

The following people have nothing to disclose: Takamasa Ohki, Shinpei Sato, Koki Sato, Eriko Goto, Jun Imamura, Hiroyoshi Taniguchi, Masatoshi Akamatsu, Hideo Yoshida, Shuntaro Obi, Nobuo Toda, Yukihiro Koike

2081

  1. Top of page

Microvascular invasion in hepatocellular carcinoma is defined by tumor characteristics and aberrant angio-genesis in peritumoral tissue

Hironori Kusano1,2, Jing Han2, Marian Bulthuis2, Peter J. Zwiers3, Koert P. de Jong4, Hirohisa Yano1, Grietje Molema3, Annette S. Gouw2;
1Department of Pathology, Kurume University School of Medicine, Kurume, Japan; 2Department of Pathology and Medical Biology, Pathology section, University Medical Center Groningen, University of Groningen, Groningen, Netherlands; 3Department of Pathology and Medical Biology, Medical Biology section, University Medical Center Groningen, University of Groningen, Groningen, Netherlands; 4Division of Hepato-Pancreato-Biliary Surgery and Liver Transplantation, Department of Surgery, University Medical Center Groningen, University of Groningen, Groningen, Netherlands

Background/aim: Several recent studies have established the pivotal role of microvascular invasion (micro-VI) in hepatocellular carcinoma (HCC) as predictor for poor prognosis and tumor recurrence even after curative resection or liver transplantation. We investigated the molecular and cellular background of micro-VI to establish the potential profile of micro-VI positive or negative HCC. Methods: We studied HCC of 43 transplanted patients from the perspective of a three-step process of vascular invasion: tumor characteristics, epithelial-mesenchymal transition (EMT) and angiogenesis. Gross and microscopic evaluation included tumor size, number of tumors, histological grade according to Edmondson and Steiner classification, and microvessel density. Fresh-frozen samples of 35 HCC and 25 adjacent non-cancerous tissue samples were subjected to quantitative reverse transcription PCR assay to examine the expression levels of messenger RNA (mRNA). Gene expression of factors relevant to the 3-step process mentioned above were investigated: cytokeratin 19 and EpCAM as progenitor cell markers; EMT-associated markers E-cadherin, Twist, Snail, S100A4, Matrix metalloproteinase-9, and angiogenesis markers consisting of Hypoxia inducible factor 1-α, vascular endothelial growth factor (VEGF)-A, Placental growth factor (PlGF), VEGF receptor (VEGFR)-1, VEGFR-2, Angiopoietin (Ang)-1, Ang-2, and Tie-2. Results: Our results showed that higher tumor grade, decreased tumor E-cadherin gene and protein expression and a higher expression of PlGF and VEGFR-1 in the adjacent non-cancerous tissue were associated with micro-VI. Progenitor cell characteristics in the tumor were found in a relatively low frequency and there was no correlation with micro-VI. Conclusion: The profile of the micro-VI positive group is determined by higher tumor grade which represents more aggressive tumor cells and higher EMT activity as expressed by more prominent loss of E-cadherin. As PlGF/VEGFR-1 is known to induce pathologic angiogenesis in which abnormal vessels are formed, higher PlGF/VEGFR-1 expression in the adjacent tissue may potentiate vascular invasion. Grading of HCC can contribute to predict micro-VI and the PlGF/VEGFR-1 activity may provide a possible target of anti-angiogenic treatment to inhibit the generation of abnormal vessels or to normalize the abnormal vessels, in parallel with results of experimental studies.

Disclosures:

The following people have nothing to disclose: Hironori Kusano, Jing Han, Marian Bulthuis, Peter J. Zwiers, Koert P. de Jong, Hirohisa Yano, Grietje Molema, Annette S. Gouw

2082

  1. Top of page

Role played by tumor markers in surveillance for hepatocellular carcinoma depends on the reliability of ultra-sonography

Shintaro MIkami1, Ryosuke Tateishi1, Shu Hagiwara2, Naoto Fuji-wara1, Masaya Sato2, Tatsuya Minami1, Koji Uchino1, Kenichiro Enooku1, Hayato Nakagawa1, Yoshinari Asaoka1, Yuji Kondo1, Shuichiro Shiina3, Hitoshi Ikeda2, Haruhiko Yoshida1, Kazuhiko Koike1;
1Department of Gastroenterology, Graduate School of Medicine, University of Tokyo, Tokyo, Japan; 2Department of Clinical Laboratory Medicine, University of Tokyo, Tokyo, Japan; 3Department of Gastroenterology, School of Medicine, Juntendo University, Tokyo, Japan

Background/Aims:The role of tumor markers in surveillance for hepatocellular carcinoma (HCC) is still controversial whereas ultrasonography is considered as indispensable. However the accuracy of ultrasonography is significantly affected by the various factors including operator's skill, performance of the equipment, liver parenchymal atrophy and coarseness and patient's obesity. The likelihood ratios for positive or negative results in ultrasound examination should be different according to patient's characteristics. We conducted this retrospective study to evaluate the role of tumor markers in surveillance for HCC with consideration for the reliability of ultrasound examination. METHODS:We enrolled 313 patients who were first diagnosed as HCC from February 2000 to December 2010 through a surveillance program using ultrasonography combined with alpha-fetoprotein, des-gamma-carboxy prothrombin and lens culinaris agglutinin A-reactive fraction of alpha-fetoprotein at the authors' institution. We divided these patients into two groups based on the triggering event: US group in which ultrasonography first detected the tumor and TM group in which tumor marker elevation led to the diagnosis of the tumor not detected by ultrasonography. We also retrospectively reviewed the image of ultrasonography and scored the reliability on 4-point scale based on 3 items (atrophy, obesity and parenchymal pattern). We also assess the survival of the patients with Kaplan-Meier method and log-rank test. RESULTS:The patients consisted of 185 males and 128 females with median age (IQR) of 68.3 (62.9–74.5) years. Of them 253 (80.3%) were anti-HCVAb positive and 31 (9.9%) were HBsAg positive. Mean ± SD of the tumor size was 2.1 ± 0.8 cm. Number of tumor was 1 in 190 (60.7%), 2–3 in 96 (30.7%), and >3 in 27 (9%). Patients were divided intoTM group (n =32) and US group (n =281). The reliability of ultrasonography was evaluated as Good in 208 (66.5%), Fine in 80 (8.0%), Poor in 21 (6.7%) and Waste in 4 (1.2%). The proportion of patients in US group decreased significantly according to the score from 93% to 75% (P=0.01). The survival rates of patients at 1, 3, and 5 years were 95.7%, 79.3%, 59.4% in the US group, and 96.8%, 83.7%, 57.2% in the TM group, respectively (P=0.98). CONCLUSION:Tumor makers may play a role in patients with unreliable results on ultrasonography. The survival of patients diagnosed by elevated tumor markers was not worse than those diagnosed by ultrasonography.

Disclosures:

Ryosuke Tateishi - Grant/Research Support: Eisai Co. Ltd.

Kazuhiko Koike - Speaking and Teaching: Bristol-Myers Squibb

The following people have nothing to disclose: Shintaro MIkami, Shu Hagiwara, Naoto Fujiwara, Masaya Sato, Tatsuya Minami, Koji Uchino, Kenichiro Enooku, Hayato Nakagawa, Yoshinari Asaoka, Yuji Kondo, Shuichiro Shiina, Hitoshi Ikeda, Haruhiko Yoshida

2083

  1. Top of page

Personalized Therapy in a Multidisciplinary Team Improves Survival in Patient with HCC: The Hepatocatt Experience

Emanuele Rinninella1, Francesca Romana Ponziani1, Marco Bio-lato2, Massimo Siciliano1, Francesca D'Aversa1, Lucrezia Laterza1, Gianluca Ianiro1, Martina Montagna1, Annalisa Tortora1, Valentina Cesario1, Mariachiara Campanale1, Nicoletta de Matthaeis2, Luca Miele2, Laura Riccardi2, Maria Assunta Zocco1, Antonio Saviano2, Valeria Abbate2, Alfonso W. Avolio3, Maria Vellone4, Michele Basso5, Roberto Iezzi6, Anna Maria De Gae-tano6, Lorenzo Bonomo6, Salvatore Agnes3, Felice Giuliante4, Fabio Maria Vecchio7, Carlo Barone5, Gian Ludovico Rapaccini2, Maurizio Pompili2, Antonio Grieco2, Antonio Gasbarrini1;
1Internal Medicine and Gastroenterology, Catholic University of the Sacred Hearth, Rome, Italy; 2Internal Medicine, Catholic University of the Sacred Hearth, Rome, Italy; 3Transplant Unit, Catholic University of the Sacred Heart, Rome, Italy; 4Hepatobiliary Surgery, Catholic University of the Sacred Heart, Rome, Italy; 5Oncology, Catholic University of the Sacred Heart, Rome, Italy; 6Radiology, Catholic University of the Sacred Heart, Rome, Italy; 7Pathologic Anatomy, Catholic University of the Sacred Heart, Rome, Italy

INTRODUCTION Hepatocellular carcinoma (HCC) is the sixth most common and the is the third cause of cancer-related death worldwide. Median survival in West countries population is less than 2 years, being lower in the advanced stages. HCC treatment needs to be multidisciplinary, as it involves several specific skills and teams for a personalized approach to patient. AIMS AND METHODS To analyze the impact of a multidisciplinary approach on patients' survival, comparing patient survival of our centre with data published in literature. We evaluated data from the prospective database of the multidisciplinary team "Hepatocatt". From September 2008 to March 2013 we collected 712 patients affected by primitive lesions of the liver (including also benign tumors and dysplastic nodules). We included in the following analysis 385 patients affected by HCC for whom an adequate follow up was available. According to the stage of disease and clinical conditions patients were referred to a personalized therapy or just to follow up, generally according to BCLC international guidelines. Survival statistical analysis was carried out with Kaplan-Meier and log-rank test. RESULTS At the stage of enrollment 53% of patients was in early stage (BCLC A), 30% in intermediate stage (BCLC B), 12% in advanced (BCLC C), and only 5% in terminal stage (BCLC D). Of 385 patients evaluable at March 2013 228 (59%) resulted still alive. The overall median survival was 32±1 months (95% CI:29,35). According to BCLC the median survival was 37±2 (95% CI: 34, 41) in BCLC A; 30 ±2 in BCLC B (95% CI: 25,35); 20 ±3 (95% CI 3, 13) in BCLC C; 12±3 (95% CI: 3, 6) The 1 and 3-year overall survival rates were 80% and 50%, respectively. Patient survival according to BCLC was 80% at 1 and 60% at 3 years in BCLC A, 70% at 1 and 50% at 3 years in BCLC B, 50% at 1 years in BCLC C, 30% at 1 years for BCLC D, p<0.001. CONCLUSIONS In HCC patients multidisciplinary approach and personalized therapy can obtain survival results similar to those reported in literature, and more favorable in advanced stages

Thumbnail image of

Disclosures:

Carlo Barone - Advisory Committees or Review Panels: Merck, Amgen; Grant/Research Support: Pfizer; Speaking and Teaching: Roche, Bayer

The following people have nothing to disclose: Emanuele Rinninella, Francesca Romana Ponziani, Marco Biolato, Massimo Siciliano, Francesca D'Aversa, Lucrezia Laterza, Gianluca Ianiro, Martina Montagna, Annalisa Tortora, Valentina Cesario, Mariachiara Campanale, Nicoletta de Matthaeis, Luca Miele, Laura Riccardi, Maria Assunta Zocco, Antonio Saviano, Valeria Abbate, Alfonso W. Avolio, Maria Vellone, Michele Basso, Roberto Iezzi, Anna Maria De Gae-tano, Lorenzo Bonomo, Salvatore Agnes, Felice Giuliante, Fabio Maria Vecchio, Gian Ludovico Rapaccini, Maurizio Pompili, Antonio Grieco, Antonio Gasbarrini

2084

  1. Top of page

Genome-wide methylation profiling identifies a 36-sig-nature that predicts survival in surgically resected hepa-tocellular carcinoma

Augusto Villanueva1, Anna Portela2, Sergi Sayols2, Virginia Her-nandez-Gea1, Carlo Battiston3, Yujin Hoshida4, Jesus Mendez-Gon-zalez2, Loreto Boix1, Sandrine Imbeaud5, Eric Letouzé6, Josep Fuster1, Jordi Bruix1, Jessica Zucman-Rossi5, Vincenzo Mazza-ferro3, Manel Esteller2, Josep M. Llovet1,4;
1BCLC, Liver Unit, CIBERehd, Hospital Clinic, Barcelona, Spain; 2Cancer Epigenetics and Biology Programme, Bellvitge Biomedical Research Institute (IDIBELL), Barcelona, Spain; 3Gastrointestinal Surgery and Liver Transplantation Unit, National Cancer Institute, Milan, Italy; 4Icahn School of Medicine, Mount Sinai, New York, NY;5INSERM-U647, Paris, France; 6Ligue contre le cancer, Paris, France

INTRODUCTION: Aberrant methylation is a frequent event in human malignancies, however, its potential role in hepatocel-lular carcinoma (HCC) is under-explored. AIMS: (1) To develop a DNA methylation-based prognostic signature in surgically resected HCC and (2) to evaluate the prognostic performance of integrated transcriptomic and epigenetic data. METHODS: Tumors from 304 HCC resected patients were analyzed in training (Heptromic cohort, n=221) and validation sets (French cohort, n=83). Methylome profiling was done with Illumina HumanMethylation450. Transcriptome profiling was performed with Affymetrix Human Genome U219 Plate and miRNA Chip 2.0. Random Survival Forest was used to generate the prognostic signature. Gene ontology and enrichment analyses were assessed with PANTHER, INTERPRO and KEGG. Kaplan-Meier plots and Cox regression were used to assess associations with outcome. RESULTS: Training set were mostly males (172, 78%), median of 66 years/old, viral-related liver damage (HCV: 101,47%, HBV: 44, 20%), and a median tumor size of 3.5 cm. Patients had predominantly uninodular disease (166, 75%), and presented without micro-vascular invasion (142, 65%) and no satellites (158, 71%). A methylation signature based on 36 probes accurately discriminate patient's survival, as defined by a mortality index. Median survival for high MI (MI>230, 20% of patients) was 13.3 mo vs 79.7 mo for non-high MI (p<0.001). MI significantly correlated with known predictors of poor outcome such as satellites, multinodularity, vascular invasion, BCLC stage, AFP, bilirubin, platelet count and albumin. Cox modeling confirmed MI as an independent predictor of survival (HR: 19.5, 95% CI 9.7–39.2, P<0.001) along with multinodularity (HR: 1.85, 95% CI 1.3–2.7, P<0.001) and platelet count (HR: 1.58, 95% CI 1.03–2.43, P=0.03) in the training set. MI was also an independent predictor of recurrence (P<0.001) along with multinodularity (P=0.007). In the validation set, patients with MI>230 had lower cancer-related survival (P=0.01). In Heptromic cohort, patients with high MI were enriched in the molecular subclass of proliferation with progenitor cell origin [EpCAM (P=0.009) and S2 (P=0.006)], and a differential non-coding mRNA expression pattern (5 miRNAs and 1 snoRNA). One of them, miR-27b, has been characterized as a regulatory hub in PPAR signaling. CONCLUSIONS: A 36 CpG island methylation-based signature correlates with survival in patients with resected HCC, and correlates with mRNA-based signatures of progenitor cell origin. In silico analyses suggest a possible involvement of non-coding RNA involved in PPAR signaling.

Disclosures:

Jessica Zucman-Rossi - Consulting: pfizer; Grant/Research Support: Integragen; Speaking and Teaching: bayer, lilly

Vincenzo Mazzaferro - Advisory Committees or Review Panels: Bayer; Grant/Research Support: Nordion; Speaking and Teaching: Merck Serono S.p.A.

Josep M. Llovet - Consulting: Bayer Pharmaceutical, Bristol Myers Squibb, Imclone, Biocompatibles, Novartis; Grant/Research Support: Bayer Pharmaceutical, Bristol Myers Squibb, Boehringer-Ingelheim

The following people have nothing to disclose: Augusto Villanueva, Anna Portela, Sergi Sayols, Virginia Hernandez-Gea, Carlo Battiston, Yujin Hoshida, Jesus Mendez-Gonzalez, Loreto Boix, Sandrine Imbeaud, Eric Letouzé, Josep Fuster, Jordi Bruix, Manel Esteller

2085

  1. Top of page

Statins Are Associated With a Reduced Risk of Liver Cancer: Data from a Large U.S. Prospective Cohort Study

Lindsay Y. King1, Hamed Khalili1, Edward Huang3, Raymond T. Chung1, Andrew T. Chan1,2;
1Medicine, Gastroenterology, Massachusetts General Hospital, Boston, MA; 2Channing Division of Network Medicine, Brigham and Women's Hospital, Boston, MA; 3Medicine, Gastroenterology, Good Samaritan Hospital, San Jose, CA

Background/Aims: While experimental studies have shown that statins could have chemopreventive effects on hepatocellular carcinoma (HCC), data from large prospective cohort studies in the United States are lacking. We therefore examined the association between statin use and HCC in two large US population-based cohorts with over 20 years of follow-up. Methods: We conducted a prospective study of 39,634 men enrolled in the Health Professionals Follow-up Study since 1990 and 96,544 women enrolled in the Nurses' Health Study since 1994 without a prior history of cancer. Biennially, with greater than 90% follow-up, we collected updated data on statin use, other lifestyle risk factors, and diagnoses of cancer and other chronic diseases. We documented cases of HCC (ICD9 155) identified through participant reports or follow-up of deaths through review of medical records. We used Cox proportional hazards models to estimate hazard ratios (HR) and 95% confidence intervals (CI) for HCC adjusting for known and putative risk factors for HCC Results: Over 20 years of follow-up, we documented 125 incident cases of HCC over 1,918,414 person years. Compared with non-users, regular users of statins had a multivariate HR=0.57, 95%CI 0.35–0.94, p=0.03 for HCC after adjustment for age, body mass index, diabetes, aspirin use, smoking status, and alcohol intake. The inverse association of statins and HCC appeared similar in women compared with men. Conclusions: In this large US prospective cohort study, statin use was associated with a lower risk of

HCC. These data support a potential role for statins in the chemoprevention of HCC.

Disclosures:

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

AndrewT. Chan -Consulting: Pfizer Inc, Bayer Healthcare, Pozen Inc, Millennium Pharmaceuticals

The following people have nothing to disclose: Lindsay Y. King, Hamed Khalili, Edward Huang

2086

  1. Top of page

Validation of immunohistochemistry-based classification of hepatic adenomas: Different subtype distribution in a large US experience

Renumathy Dhanasekaran1, Yoo Na Kang2, Rory Smoot2, Catherine D. Moser2, Gregory J. Gores1, Tsung-Teh Wu2, Taofic Mounaj-jed2, Lewis R. Roberts1;
1 Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN; 2Mayo Clinic, Rochester, MN

Aim: Hepatic adenomas (HA) are benign liver tumors and a new phenotypic classification associated with distinct genotypes has been proposed recently from French studies. This has not been sufficiently validated in the US population. Methods: We analyzed clinical, histopathologic and immunohistochem-istry data from patients with a diagnosis of HA who underwent surgical resection at a single center. Immunohistochemical staining for liver-fatty acid binding protein (L-FABP), serum amyloid A (SAA), glutamine synthetase (GS) and β-catenin (BC) were used to classify adenomas. Results: Total of 188 nodules were found in the 105 patients (89% female). Oral contraceptive use was reported in 86.4% of the women with available data. Mean age at diagnosis was 38.4 (SD 12.8) and mean BMI was 28.4 (SD 7.2). Abdominal pain was present in 48.6% and bleeding in 18.1% of patients. The nodules were sub-classified into 4 groups based on immunohistochemical staining: Inflammatory (SAA+, BC neg); Steatotic (LFABP-ve, SAA-ve); Beta catenin (BC pos and GS+) subtype and Unclassified. Table 1 shows the differences in clinical and histopathologic characteristics between the two largest subgroups, steatotic and inflammatory. Abnormal beta catenin staining was present in three nodules (2%) from three separate patients (3%). None of these patients had abnormal LFABP staining and all three had abnormal GS staining. Mean age at diagnosis of 23 yrs and they had large single tumors (mean 6.3cms). None of them showed significant steatosis or inflammation and one patient showed cellular atypia and rosette formation. We had a large group (35%) of patients who could not be classified by the above IHC stains. There were two distinct IHC patterns in this group: (a) isolated abnormal GS staining (11%) and (b) absence of all four IHC markers (24%). Further genotype analysis will clarify whether nodules with isolated GS staining have a beta catenin activating mutation without nuclear beta catenin staining or alternatively have activation of the Wnt/beta catenin pathway in the absence of beta catenin mutations. Conclusion: To our knowledge, this is the largest US experience of patients with HA. In this study, the prevalence of inflammatory and beta catenin subtypes was less than previously reported in the European experience. Genetic analysis for beta catenin activating mutations may be warranted in patients with isolated GS staining, as this may affect prognosis.

TYpeIHCMaleObesityBleedingMultipleInflammation.DysplasiaCytologic atypiasteatosisBallooning changesGS
SteatoticLFABP-ve12%50%44%84% 15%9%32%59%10%
InflammatorySAA+vc, BC-ve29%63%12%81%68%10%17%4%.7%36%
p value 0.0390.29800530.598<0.0010.5510.222<0.001<0.0010.004

Disclosures:

Gregory J. Gores - Advisory Committees or Review Panels: Bayer, Chugia, Dai-ichi, Generon, Conatus, IntegraGen

Lewis R. Roberts - Advisory Committees or Review Panels: Inova; Grant/Research Support: Bristol Myers Squibb, Bayer, Nordion; Speaking and Teaching: Nordion

The following people have nothing to disclose: Renumathy Dhanasekaran, Yoo Na Kang, Rory Smoot, Catherine D. Moser, Tsung-Teh Wu, Taofic Mounajjed

2087

  1. Top of page

Ascites and Alfa-Feto-Protein Improve the BCLC Staging Classification for Prediction of Prognosis in Hepatocellu-lar Carcinoma

Asmaa Gomaa1, Wael Abdel-Razek1, Mohammed Saad Hashem1, Zahraa ElKhateeb2, Imam Waked1;
1Hepatology, National Liver Institute, Menoufiya, Egypt; 2Oncology, National Liver Institute, Shebin El Kom, Egypt

The BCLC staging classification is widely accepted as the best staging system for HCC to predict prognosis and allocate treatment options. Whether adding ascites & AFP to BCLC improves prediction of survival in early stages of HCC is assessed in this analysis. Methods: 2,000 Patients with HCC presenting to a tertiary referral center were included. Patients were categorized according to BCLC stage. The effect of the presence of ascites & AFP level on survival was evaluated. Results: Mean age 56±9 years, 84% males, 90% had cirrhosis, in 87% HCV related, 40% had ascites, AFP >200ng/ml in 42%. At presentation, 25% were BCLC stage A, 30.4% stage B, 20.3% stage C, 24.3% stage D. Of 500 BCLC stage A patients, 18% had ascites, 33% had AFP>200ng/ml and 40 had both ascites and AFP>200ng/ml. Overall median survival was 15 months. Mean survival for BCLC stage A patients: 31 months, stage B: 22 months, stage C: 10 months, stage D: 8 months (p<0.0001). Ascites and AFP further discriminated patient survival within BCLC stages A and B. In BCLC stage A patients: median survival in presence of ascites was shorter if AFP was >200ng/ml (19 vs 24 months), and in the absence of ascites, patients with AFP >200 ng/ml had shorter survival (27 vs 34 months) (figure). In BCLC stage B patients, survival for similar groups was 14, 18, 20 and 25 months. The 1, 2, & 3 year survival for stage A patients without ascites and AFP<200 ng/ml was 93%, 76% and 70%, & for patients with ascites & AFP>200 ng/ml was 83%, 24%, and 0% respectively (p<0.001). Conclusion: Adding AFP & ascites as variables to the BCLC staging classification can improve predicting patients' prognosis, especially at earlier stages.

Thumbnail image of

Disclosures:

Imam Waked - Speaking and Teaching: Hoffman L Roche, Merck, Bayer, BMS

The following people have nothing to disclose: Asmaa Gomaa, Wael Abdel-Razek, Mohammed Saad Hashem, Zahraa ElKhateeb

2088

  1. Top of page

Hepatitis C Core Mutations Are Associated With Hepatocellular Carcinoma (HCC) Risk

Kirti Shetty1, Brent Korba2, Bin Zhou3, Alexei V. Medvedev2, Pras-anth Viswanathan2, Kepher H. Makambi3, Habtom W. Ressom3, Christopher A. Loffredo3;
1Georgetown University Hospital, Washington, DC; 2Microbiology and Immunology, Georgetown University Medical Center, Washington, DC; 3Lombardi Cancer Center, Georgetown University Medical Center, Washington, DC

Background: Prior research in the Asian population suggests that specific mutations in codons 70 and 91 of the HCV genotype (GT) 1b core sequence are associated with 6 to 8-fold increased risk of HCC. Such biomarkers of HCC risk are urgently needed to improve early detection rates. Our goal was to conduct a pilot study of the association of mutations in codons 70 and 91 with HCC in a multiethnic U.S. population with HCV cirrhosis, a subgroup with HCC, who are enrolled in a longitudinal study with stored specimens at baseline and complete medical follow-up data. Since the association with HCC risk was originally identified for HCV GT1b, we wanted to determine if this association was also present for GT1a patients. Lastly, we sought to determine if other mutations in the HCV core gene were associated with HCC risk in our population. Methods: HCV core gene sequences from 79 GT1 HCV carriers with advanced cirrhosis, including 37 with HCC, were analyzed. HCV RNA was extracted from stored sera, RT and PCR amplified, and sequenced using standard, previously established procedures. Nucleotide sequence data was compared with reference sequences available from GenBank. Mutation frequency in HCV sequences from HCC and non-HCC patients were analyzed using standard statistical methods. Results: This cohort was predominantly male (74%) and Caucasian (52%) with a mean MELD score of 12.3. The majority of patients (76%) had GT1a disease. The frequency of mutations at codons 70 or 91 were not statistically different in HCC and non-HCC GT1 b carriers. For GT1 a carriers, mutations at codon 70 were also not different between these two groups. In GT1a carriers, mutations in the first two nucleotides of codon 91 were more frequent in HCC vs. non-HCC patients (4/21 vs.0/26, p=0.03, Fisher's exact test). For GT1a carriers, mutations in other codons were also observed in HCC patients, but not in any of 26 non-HCC patients: codon 24 (4/21, p=0.03), 42 (4/21, p=0.03), and 144 (7/20, p=0.001). Conclusion: Similar to other studies, specific mutations in GT1 HCV core gene sequences at different locations, which are not present in non-HCC carriers, can be observed in a subset of patients with HCC. This finding has important implications for the early detection of HCC in HCV-infected individuals.

Disclosures:

Kirti Shetty - Grant/Research Support: Ikaria, Novartis, Onyx-Bayer, Hyperion; Speaking and Teaching: Merck-Schering Plough, Salix, Gilead, Onyx

The following people have nothing to disclose: Brent Korba, Bin Zhou, Alexei V. Medvedev, Prasanth Viswanathan, Kepher H. Makambi, Habtom W. Ressom, Christopher A. Loffredo

2089

  1. Top of page

Diabetes is not associated with increased risk of hepato-cellular carcinoma in patients with hepatitis C and advanced fibrosis or cirrhosis

Ju Dong Yang2,1, Donghee Kim3, Lewis R. Roberts1, W. Ray Kim1;
1Mayo Clinic College of Medicine, Rochester, MN; 2Departmentof Internal Medicine, University of Arkansas For Medical Science, Little Rock, AR; 3Department of Internal Medicine, Healthcare Research Institute, Seoul National University Hospital Healthcare System Gangnam Center, Seoul, Republic of Korea

Background: Diabetes has been associated with hepatocellular carcinoma (HCC), presumably from non-alcoholic fatty liver disease. It is much less clear, however, whether diabetes increases the risk of HCC in patients who already have cirrhosis from other causes such as hepatitis C. Aim: We investigated whether diabetes or insulin resistance is associated with occurrence of HCC in patients with hepatitis C virus (HCV) infection with advanced fibrosis or cirrhosis. Methods: All patients (n=1050) enrolled in the Hepatitis C Antiviral Long-term Treatment against Cirrhosis (HALT-C) Trial were included in our study. Diabetes was defined by a past medical history or fasting serum glucose >126mg/dL. Insulin resistance was assessed by the HOMA2-IR index. The Cox regression analysis was used to examine the effect of diabetes and insulin resistance on the risk of HCC. Results: The mean age was 50.1 years and majority were male (71%) and non-Hispanic white (72%). Diabetes and cirrhosis were present in 24% and 41 %, respectively. HCC developed in 88 patients (18 with diabetes and 70 without) after a median follow up of 6.1 years. Diabetes was more common with cir-rhotic (47% vs 39%, p<0.02) and those with higher CTP score (5.3 vs 5.2, p<0.01). The proportion of patients with diabetes increased as the stage of liver fibrosis increased - 15%, 21%, 26%, 26%, and 30% of patients in Ishak stage 2, 3, 4, 5, and 6, respectively had diabetes (p=0.04). In the Table, neither diabetes nor HOMA2-IR was associated with an increased risk of HCC. Surprisingly, higher BMI was associated with decreased risk of HCC. Conclusions: HCC risk in patients with HCV cirrhosis is increased to the degree that hyperinsulinemia or insulin resistance may not increase it further. The negative association between obesity and HCC may warrant further investigations.

 Univariate ModelMultivariate M(del 1Multivariate Model 2
 HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P value
Diabetes0.84(0.49–1.38)0.500.72(0.41–1.22)0.23  
H0MA2-IR0.98(0.92–1.03)0.49  0.98(0.91–1.04)0.58
Age (1 year)1.04(1.01–1.07)<0.011.05(1.02–1.08)<0.011.06(1.02–1.09)<0.01
Black race1.61(0.95–2.62)0.081.61(0.95–2.62)0.081.86(1.03–3.24)0.04
Female0.73(0.47–1.27)0.350.62(0.36–1.02)0.060.46(0.23–0.85)0.01
BMI(1 unit)0.94 (0.89–0.98)<0.010.94(0.89–0.98)<0.010.92(0.87–0.98)<0.01

Model 1- effect of diabetes on HCC risk after adjusting for age, race, gender, BMI, and Ishak fibrosis stage

Model 2 - effect of HOMA2-IR on HCC risk after adjusting for age, race, gender, BMI, and Ishak fibrosis stage

Ishak fibrosis stage was significant predictor in both univariate and multivariate analysis (p<0.001) and was adjusted in multivariate models in model 1 and 2.

HR: Hazard ratio, CI:confidence interval

Disclosures:

Lewis R. Roberts -Advisory Committees or Review Panels: Inova; Grant/Research Support: Bristol Myers Squibb, Bayer, Nordion; Speaking and Teaching: Nordion

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

The following people have nothing to disclose: Ju Dong Yang, Donghee Kim

2090

  1. Top of page

CK19 and SOX9 expression in hepatocellular carcinoma predicts early postoperative recurrence

Takayuki Kawai1, Kentaro Yasuchika1, Takamichi Ishii2, Hokahiro Katayama1, Elena Y. Yoshitoshi1, Satoshi Ogiso1, Sadahiko Kita1, Katsutaro Yasuda1, Ken Fukumitsu1, Shinji Uemoto1;
1Hepatobil-iary, Pancreas and Transplant Surgery, Kyoto University Hospital, Kyoto, Japan; 2Surgery, Nishikobe Medical Center, Kobe, Japan

Background: Recent developments in stem cell biology have revealed the existence of cancer stem cells (CSCs) in various tumors. CSCs possess high malignant potential and common features of normal tissue stem cells. During the normal hepatic development, cytokeratin 19 (CK19) and SRY (sex determining region Y)-box9 (SOX9) are expressed in hepatic progenitor zone. Aim: We aimed to demonstrate the expression of CK19 and SOX9 in hepatocellular carcinoma (HCC) would relate to the feature of cancer stem cell, which predicts the prognosis after surgical resection. Patients and methods: 104 patients with HCC confirmed by pathologic analyses, who had undergone a hepatic resection at Kyoto University Hospital from January 2005 to December 2006 were included in this study. In order to examine the CK19 and/or SOX9 expression in human clinical HCC, immunohistological assays were performed. Recurrence-free survival (RFS) and overall survival (OS) were investigated by Kaplan-Meier method and statistically analyzed by Log-rank test. Results: Of 104 HCC patients, 61 were CK19-/SOX9-, 32 were CK19-/SOX9+, 6 were CK19+/SOX9- and 5 were CK19+/SOX9+. Among the four groups, CK19+/SOX9+ group had the worst RFS. Log-rank test revealed that CK19 expression (CK19+), SOX9 expression (SOX9+), low albumin concentration, portal vein invasion, and liver cirrhosis were associated with worse RFS. Multivariate analysis demonstrated that CK19+, SOX9+ and portal vein invasion were independent predictors of postoperative recurrence. Regarding OS, CK19+, low albumin concentration, portal vein invasion and liver cirrhosis were associated with OS by log-rank test. Multivariate analysis revealed that low albumin concentration, portal vein invasion and liver cirrhosis were independent predictors of OS, while CK19+ was not. Conclusion: The expression of CK19 and/or SOX9 in HCC could be a useful predictor for postoperative recurrence.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Takayuki Kawai, Kentaro Yasuchika, Takamichi Ishii, Hokahiro Katayama, Elena Y. Yoshitoshi, Satoshi Ogiso, Sadahiko Kita, Katsutaro Yasuda, Ken Fukumitsu, Shinji Uemoto

2091

  1. Top of page

Risk Prediction of Hepatocellular Carcinoma in Patients with Cirrhosis: The ADRESS-HCC Risk Model

Jennifer A. Flemming1, Eric Vittinghoff2, W. Ray Kim3, Norah Ter-rault1;
1Medicine, University of California San Francisco, San Francisco, CA; 2Epidemiology and Biostatistics, University of California San Francisco, San Francisco, CA; 3Medicine, Mayo Clinic, Rochester, MN

Surveillance for hepatocellular carcinoma (HCC) is recommended for cirrhotic patients with an HCC risk ≧ 1.5% per year. We aimed to develop an HCC risk prediction model estimating the 1-year probability of HCC in an individual patient with cirrhosis. Methods: In constructing a cohort of cirrhotic patients at risk of developing HCC in the United States, 34,932 cirrhotic adults wait-listed for transplantation between March 2002 and December 2011 were identified from the Scientific Registry of Transplant Recipients database. Incident HCC was identified by HCC-MELD exception at least six months after listing. Using Cox proportional hazards regression we selected the prediction model with optimum cross-validated C-index within a non-random development sub-cohort, then validated the model in the complementary validation sub-cohort. Calibration was determined by comparing predicted and observed cumulative incidences of HCC at 1 year in the validation cohort. Results: Incident HCC developed in 1,960 (5.6%) patients during a median follow-up of 1.3 years (interquartile range 0.47–2.83 years). Six baseline variables independently associated with the development of HCC were used to develop the ADRESS-HCC risk model (figure). C-indices in the derivation and validation cohorts were 0.704 and 0.691 respectively. In the validation cohort, the predicted cumulative incidence of HCC by the ADRESS-HCC model closely matched the observed data. An ADRESS-HCC score of > 4.67 identified patients meeting the AASLD criterion for HCC surveillance (> 1.5% per year). Conclusions: The ADRESS-HCC risk model is an accurate tool for predicting the 1 year risk of HCC among cirrhotic patients. It may be used to guide clinical decisions and patient counseling regarding HCC risk and to help inform HCC surveillance policies.

Figure: ADRESS-HCC Risk Score

Thumbnail image of

Disclosures:

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

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

The following people have nothing to disclose: Jennifer A. Flemming, Eric Vit-tinghoff

2092

  1. Top of page

Clinicopathological relevance of vasohibin-1-associated angiogenesis in hepatocellular carcinoma

Keigo Murakami1,2, Atsuko Kasajima2, Naoki Kawagishi1, Yasu-fumi Sato3, Hironobu Sasano2, Noriaki Ohuchi1;
1Department of Advanced Surgery, Tohoku University Hospital, Sendai, Japan; 2Department of Pathology, Tohoku University Hospital, Sendai, Japan; 3Department of Vascular Biology, Institute of Development, Aging, and Cancer, Tohoku University, Sendai, Japan

Vasohibin-1 is an endothelium-derived negative feedback regulator of angiogenesis which is induced by fibroblast growth factor-2 (FGF-2) and vascular endothelial growth factor-A (VEGF-A). Vasohibin-1 is a useful marker for detection of tumor specific angiogenesis and its clinicopathological relevance has been studied in variable human cancer. We investigated vasohibin-1-associated angiogenesis and its regulator protein, FGF-2 and VEGF-A, expression in 181 consecutive cases of hepatocellular carcinoma (HCC). Double immunostaining of an endothelial marker CD34 and vasohibin-1 with Ki67 was performed to assess angiogenic activity of endothelial cells in HCC. We further analyzed clinicopathological relevance of vasohibin-1-associated angiogenesis. The ratio of Ki-67-posi-tive endothelial cells in vasohibin-1-positive vessels was significantly higher than that in CD34-positive vessels (22% versus 9%, P < 0.001). Moreover, the ratio of vasohibin-1-positive vessels to CD34-positive vessels (vasohibin-1 /CD34) increased according to tumor progression. The ratio positively correlated with FGF-2 immunoreactivity (P < 0.05), but the correlation was not significant with VEGF-A immunoreactivity (P = 0.061). Patients were classified into two groups; high vasohibin-1/CD34 and low vasohibin-1 /CD34. Cutoff value of the ratio was set to be 0.459 by ROC-analysis. The 10-year overall survival and the 2-year disease-free survival rates of the low vasohibin-1 /CD34 group (vasohibin-1 /CD34 ≦0.459) were significantly higher than those of the high vasohibin-1 /CD34 group (vasohibin-1 /CD34 >0.459) (survival, 48% versus 38% and 52% versus 35%, P < 0.001 and P < 0.05, respectively). Further, vasohibin-1 /CD34 in HCC patients was an independent marker of poor prognosis as determined by multivariate analysis (risk ratio, 1.973; 95% confidence interval, 1.049–3.711, P = 0.035). These results suggest that vasohibin-1 is more specific endothelial marker for the detection of active angiogenesis compared to CD34. Furthermore, vasohibin-1/CD34 may reflect the exact ratio of proliferating endothelial cells to overall endothelial cells and is a useful prognostic marker for HCC patients.

Disclosures:

The following people have nothing to disclose: Keigo Murakami, Atsuko Kasajima, Naoki Kawagishi, Yasufumi Sato, Hironobu Sasano, Noriaki Ohuchi

2093

  1. Top of page

Gd-EOB-DTPA-enhanced MRI and Serum AFP Predict the Prognosis of Early-stage Hepatocellular Carcinoma

Taro Yamashita1, Azusa Kitao2, Osamu Matsui2, Takehiro Hayashi1, Kouki Nio1, Mitsumasa Kondo1, Hikari Okada1, Tatsuya Yamashita1, Eishiro Mizukoshi1, Honda Masao1, Yasuni Nakanuma3, Hiroyuki Takamura4, Tetsuo Ohta4, Yasunari Nakamoto5, Masakazu Yamamoto6, Tadatoshi Takayama7, Shigeki Arii8, Xin W. Wang9, Shuichi Kaneko1;
1Department of Gastroen-terology, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan; 2Department of Radiology, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan; 3Department of Pathology, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan; 4Department of Gastroenterologic Surgery, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan; 5Second Department of Internal Medicine, Fukui University School of Medicine, Fukui, Japan; 6Department of Surgery, Tokyo Women's Medical University, Tokyo, Japan; 7Department of Digestive Surgery, Nihon University School of Medicineersity, Tokyo, Japan; 8Department of Hepato-Biliary-Pancreatic Surgery, Tokyo Medical and Dental University, Tokyo, Japan; 9Laboratory of Human Carcinogenesis, National Cancer Institute, Bethesda, MD

[Background] Survival of patients with hepatocellular carcinoma (HCC) varies on an individually basis, even after surgery for early-stage tumors. Here we evaluated the usefulness of a recently introduced technology—gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (EOB-MRI)—for diagnosing and predicting prognosis of patients with early-stage HCC. [Methods] A total of 417 patients who received radical resection for HCC were enrolled. We examined 70 patients diagnosed with EOB-MRI to evaluate Gd-EOB-DTPA uptake, gene/protein expression, and prognosis (cohort 1). We analyzed the microarray data of 238 patients to understand the molecular basis of the EOB-MRI findings. RNA interference was used to evaluate the role of identified transcription factors on tumorige-nesis and Gd-EOB-DTPA uptake in the subcutaneous xenotrans-plantation model using surgically resected primary HCCs and immune deficient mice. We evaluated the prognosis of 109 patients with independent, early-stage HCC in a multicenter cohort (cohort 2). [Results] Atypical uptake of Gd-EOB-DTPA in the hepatobiliary phase was observed in 12.9% of HCCs in cohort 1. This uptake correlated with the low serum alpha-feto-protein (AFP) levels, maintenance of hepatocyte function with the up-regulation of OATP1 B3 and HNF4A, and good prognosis. In contrast, HCC showing reduced Gd-EOB-DTPA uptake with high serum AFP presented poor prognosis with the activation of an oncogene FOXM1. Knockdown of HNF4A in cancer cells, obtained from a surgically resected HCC showing atypical uptake of Gd-EOB-DTPA, resulted in the increase of AFP and FOXM1 and the loss of OATP1B3 expression accompanied with the mesenchymal morphological changes, enhanced cell proliferation and tumorigenesis in vivo. HCC classification based on EOB-MRI and serum AFP predicted the prognosis of early-stage HCC patients in cohort 2, prospectively. [Conclusions] EOB-MRI together with serum AFP was validated as a useful diagnostic tool for predicting the prognosis of early-stage HCC. This classification system could be incorporated into current HCC staging to improve management algorithms, especially in early stage.

Disclosures:

Shuichi Kaneko - Grant/Research Support: MDS, Co., Inc, Chugai Pharma., Co., Inc, Toray Co., Inc, Daiichi Sankyo., Co., Inc, Dainippon Sumitomo, Co., Inc, Aji-nomoto 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: Taro Yamashita, Azusa Kitao, Osamu Matsui, Takehiro Hayashi, Kouki Nio, Mitsumasa Kondo, Hikari Okada, Tatsuya Yamashita, Eishiro Mizukoshi, Honda Masao, Yasuni Nakanuma, Hiroyuki Takamura, Tetsuo Ohta, Yasunari Nakamoto, Masakazu Yamamoto, Tadatoshi Takayama, Shigeki Arii, Xin W. Wang

2094

  1. Top of page

The impact of post-progression survival on overall survival of patients with advanced hepatocellular carcinoma treated by sorafenib therapy

Takeshi Terashima, Tatsuya Yamashita, Noboru Takata, Hidetoshi Nakagawa, Kuniaki Arai, Takashi Kagaya, Eishiro Mizukoshi, Masao Honda, Shuichi Kaneko;
Gastroenterology, Kanazawa University Hospital, Kanazawa, Japan

Background and Aim: In some cancers, an increasing numbers of new agents have been successfully developed and prolonged post-progression survival (PPS). Little is known about the clinical course after sorafenib therapy, although sorafenib has been established as standard care in advanced hepatocellular carcinoma (HCC). The aim of the present study was to investigate the impact of PPS on overall survival (OS) in advanced HCC patients treated by sorafenib monotherapy. Materials and Methods: We used the PubMed database until 31 October, 2012, to search for reports describing the survival data of advanced HCC patients treated by sorafenib monotherapy. We selected reports that provided data for both OS and progression-free survival (PFS) and we did not stick to their primary end-point if each report dealt with more than 20 patients. Median PPS was defined as the period obtained by subtracting median PFS from median OS for each report. We investigated the correlation of OS and either PPS or PFS using Spearman's rank correlation coefficient. Results: We analyzed 35 reports with 3396 patients who met our criteria. The median OS of all reports ranged from 4.8 to 26.1 months, and the median PFS and median PPS ranged from 2.0 to 9.0 months and from 0.5 to 18.1 months, respectively. We found that median PPS correlated highly with median OS (r = 0.879, p < 0.001). However, median PFS did not correlate with median OS as highly as PPS did (r = 0.538, p < 0.001). From the survival data stratified by Child-Pugh classification, the average proportion of median PPS to median OS was significantly higher in Child-Pugh class A than that in Child-Pugh class B (56.4 ±8.9 % vs. 30.9 ± 12.6 %, p = 0.0035). Conclusion: PPS correlated highly with OS in patients with advanced HCC treated by sorafenib monotherapy. PPS, not only PFS, should be emphasized as a treatment strategy for advanced HCC patients treated by sorafenib monotherapy, especially in Child-Pugh class A.

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: Takeshi Terashima, Tatsuya Yamashita, Noboru Takata, Hidetoshi Nakagawa, Kuniaki Arai, Takashi Kagaya, Eishiro Mizukoshi, Masao Honda

2095

  1. Top of page

The Best Response Predicts Long-Term Survival Better than the Initial Response Following Repeated Transarter-ial Chemoembolization for the Treatment of Hepatocellular Carcinoma

Jonggi Choi1, Ju Hyun Shim1, Kang Mo Kim1, Young-Suk Lim1, Han Chu Lee1, Dong Jin Suh1,2;
1Department of Gastroenterology, Asan Liver Center, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; 2Department of Internal Medicine, Vievis Namuh Hospital, Seoul, Republic of Korea

Introduction: Additional TACE procedures have recently been shown to provide favorable survival outcomes, even in patients who fail to respond to the first session. The aims of this study were to 1) elucidate the clinical implications of the best response in comparison with the initial response during repeated TACE and 2) identify the clinical factors associated with the need for TACE repetition that will ultimately achieve the best response. Methods: We evaluated 332 patients diagnosed with intermediate-stage HCC, which is considered a representative indication of TACE, and Child-Pugh class A liver disease, all of whom were initially treated with repeated TACE sessions (median number of sessions: 1; range: 1–10). All patients were diagnosed with >1 measurable lesion measuring >1 cm, and the response measured after each session was based on EASL criteria. Results: The initial response after the first TACE session was complete response (CR) in 64 patients (19.3%), partial response (PR) in 106 (31.9%), stable disease (SD) in 132 (39.8%), and progressive disease (PD) in 30 (9.0%); in terms of the best response, 135 patients (40.7%), 115 (34.6%), 52 (15.7%), and 30 (9.0%) achieved these corresponding responses following the administration of serial TACE sessions, respectively. Of the 106 patients who initially demonstrated PR, CR was achieved as the best response in 71 patients (67.0%) and PR was observed as the best response in 80 of the 132 patients (60.6%) who were initially diagnosed with SD after a median of 2 TACE sessions. Overall survival analysis revealed that there were no differences between patients who obtained CR as the initial or best response (48.0 vs 49.5 months; P=NS). Median survival time was significantly longer in patients who demonstrated responses that improved from PR to CR than in patients who demonstrated sustained PR (56.6 vs 22.1 months; P<0.05). Only marginal significance was noted in terms of the difference in median survival between patients who improved from SD to PR and in patients who maintained SD (32.3 vs 25.6 months; P=0.06). Tumor size was only positively correlated with the number of TACE sessions that were required to achieve the best response (P<0.05). Multivariate Cox analysis indicates that achieving CR or PR as the best response is the most significant predictor of survival following TACE (hazard ratio, 0.41: P<0.05). Conclusion: Our data indicate that the best response, rather than the initial response, observed during serial TACE most strongly predicts the long-term survival outcomes of patients with intermediate-stage HCC. The clinical decision to perform repeated TACE sessions should be based on tumor size.

Disclosures:

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

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

2096

  1. Top of page

Genetic risk analysis of natural killer cell receptor G2D (NKG2D) variant in a large European cohort of patients with cholangiocarcinoma

Marcin Krawczyk1, Bianca Simon1, Florentina Mihalache2, Monica Acalovschi2, Frank Lammert1, Vincent Zimmer1;
1Department of Medicine II, Saarland University Hospital, Homburg/Saar, Germany; 2Department of Medicine III, University Iuliu Hatieganu, Cluj-Napoca, Romania

Background: Although cholangiocarcinoma (CCA) is the second most common primary liver cancer, its pathogenesis has not been fully elucidated1. Genetic predisposition to this cancer has for long been suggested, and in our most recent analysis we identified an association between the MST-1 variant predisposing to primary sclerosing cholangitis (PSC) and CCA2. Previously it was shown that genetic variants within the NKG2D-MICA pathway modulate the risk to develop CCA in PSC patients3. In the current study we investigate the effects of the NKG2D variant on the prevalence of the CCA in a cohort encompassing mostly individuals with sporadic carcinoma. Patients and methods: In total, we genotyped 221 patients with diagnosed CCA (131 males; age 28 - 90 years; PSC prevalence < 2 %) from Germany (n = 164) and Romania (n = 57). The control group consisted of 297 CCA-free individuals (131 males, age 22 - 90 years). The NKG2D single nucleotide polymorphism (SNP) rs2617163 was genotyped using a PCR-based assay with 5'-nuclease and fluorescence detection. Exact tests were performed to check the consistency of genotyping results with Hardy-Weinberg equilibrium (HWE); the association of the SNP with CCA was tested in contingency tables (chi2 test for alleles; Armitage's trend test for genotypes). Results: Overall, we achieved 100% genotyping success, and no deviation from Hardy-Weinberg-equilibrium was observed, indicating robust genotyping. The association tests did not provide evidence for genetic risk modulation by the NKG2D variant either in the whole cohort (common OR = 1.207, P = 0.20) or after exclusion of individuals with PSC (common OR = 1.209, P = 0.20). Of note, in the analysis performed separately for individuals with extra- and intrahepatic tumor localization, we identified a potential association of this variant with extrahepatic tumors (n = 45, common OR = 1.631, P = 0.04). In contrast, we did not detect any association of the NKG2D polymorphism with intrahepatic CCA (n = 176, common OR = 1.108, P > 0.05). Conclusions: The NKG2D variant, previously associated with the risk of CCA in patients with PSC, may modulate the risk of extrahepatic bile duct cancer in individuals without PSC. These results, together with the previously published association3, point to a potential genotype-based screening strategy for individuals who are at-risk of cholangiocellular carcinoma due to the NKG2D polymorphism. 1. Khan et al. Gut. 2012 2. Krawczyk etal. Dig Liver Dis. 2013 3. Melum et al. Hepatology. 2008

Disclosures:

The following people have nothing to disclose: Marcin Krawczyk, Bianca Simon, Florentina Mihalache, Monica Acalovschi, Frank Lammert, Vincent Zimmer

2097

  1. Top of page

Cell-Of-Origin Based Intrahepatic Cholangiocarcinoma Classification Corresponds To Kras Mutation Status As well As Cetuximab Efficacy

Mina Komuta1, Jun Akiba2, Sachiko Ogasawara2, Sara Vander Borght1, Chris Verslype3, Vincent Vandecaveye4, Frederik Nevens3, Wim Laleman3, Werner Van Steenbergen3, Jacques Pirenne5, Raymond Aerts6, Baki Topal6, Eric Van Cutsem7, Hirohisa Yano2, Tania Roskams1;
1Translational Cell & Tissue Research, KU Leuven, Leu-ven, Belgium; 2Pathology, Kurume University School of Medicine, Kurume, Japan; 3Hepatology, University Hospitals Leuven, Leuven, Belgium; 4Radiology, University Hospitals Leuven, Leuven, Belgium; 5Abdominal Transplant Surgery, University Hospitals Leuven, Leuven, Belgium; 6Abdominal Surgery, University Hospitals Leuven, Leuven, Belgium; 7Clinical Digestive Oncology, University Hospitals Leuven, Leuven, Belgium

Introduction: Intrahepatic cholangiocarcinoma (ICC) is a lethal malignancy with no standardized chemotherapy. Tumor heterogeneity makes it difficult to characterize tumors as well as to choose an optimal treatment. Recently, we succeeded in characterizing the heterogeneity of ICCs based on their possible cell of origin (ref); mucin-producing ICCs (muc-ICCs) (possibly derived from large bile ducts), and ICCs with histological diversity (mixed-ICCs) (presumably originating from bipotential hepatic progenitor cells). We hypothesized that the cell-of-origin based CC classification may be useful to personalize treatment for CC patients because overexpression of epidermal growth factor receptor (EGFR) protein has been reported in CC and correlates with a worse prognosis. Methods: We investigated the mutation status of KRAS, BRAF, and EGFR in ICCs and compared their mutation status with the different types of ICCs. Human CC cell lines (KMC-1, KMC-2, KMBC, KMCH-1, and KMCH-2) were employed to gain functional insights into the action of cetuximab. In brief, KMC-1, -2, and KMBC correspond to the muc-CCs, and KMCH-1 and -2 correspond to the mixed-CCs. Results: Among 63 resected ICCs (29 muc-ICCs, 34 mixed-ICCs), KRAS mutations were found in 12 (19%) [p.G12D (n=7), p.G12V (n=4), p.G12R (n=1)], an EGFR mutation was seen in one (1.6%) ICC, and no BRAF mutations were detected. Importantly, KRAS mutations were only observed in the muc-ICCs, and represented 41.4% of that group. KRAS-mutated muc-ICCs showed a frequent papillary component, biliary dysplasia, and a significantly better prognosis compared with non-KRAS-mutated muc-ICCs. Of the five human CC cell lines, KRAS mutation was seen in KMC-1 and -2, but not in KMBC, KMCH-1, and KMCH-2. Western blotting confirmed EGFR and ERK1/2 activation in all cell lines, and that was clearly suppressed in KRAS-mutated cell lines, KMC-1 and KMC-2, after cetuximab treatment. Interestingly, p-AKT activation was gradually observed in KMCH-1 and -2 after cetuximab treatment, suggesting an 'escape' mechanism from cetuximab. After treatment, cell proliferation was significantly suppressed in KMBC, KMC-2, and KMCH-2 compared with others. Conclusion: KRAS mutation was only seen in muc-ICCs, and not in mixed-ICCs. This confirmed the validity of our cell-of origin based ICC classification. Results from the cell lines suggested cetuximab treatment efficacy in ICCs, especially KRAS-mutated muc-ICCs. References: Komuta M, et al. Hepatology. 2012 Jun;55(6):1876–88

Disclosures:

Chris Verslype - Advisory Committees or Review Panels: Bayer, Pfizer; Grant/Research Support: Novartis, Sirtex

Frederik Nevens - Grant/Research Support: Ipsen, Roche, MSD, Astellas, CAF

The following people have nothing to disclose: Mina Komuta, Jun Akiba, Sachiko Ogasawara, Sara Vander Borght, Vincent Vandecaveye, Wim Laleman, Werner Van Steenbergen, Jacques Pirenne, Raymond Aerts, Baki Topal, Eric Van Cutsem, Hirohisa Yano, Tania Roskams

2098

  1. Top of page

Hepatocellular carcinoma is more advanced at diagnosis in HIV/HCV co-infected patients. Result of an interco-hort case control study

Moana Gelu-Simeon1, Maite Lewin2, Rodolphe Sobesky1,3, Marita Ostos1, Elina Teicher4,5, Faroudy Boufassa6, Laurence Meyer6, Hélène Fontaine7, Dominique Salmon8, Olivier Seror9, Jean Claude Trinchet10, Jean-Charles Duclos-Vallee1,3; 1Centre Hepato-Biliaire, AP-HP Hopital Paul Brousse, Villejuif, France; 2Service de Radiologie, AP-HP, Hopital Paul Brousse, Villejuif, France; 3Unit 785, Inserm, Villejuif, France; 4UMR-S785, Univ Paris-Sud, Villejuif, France; 5Service des Maladies infectieuses, AP-HP, Hopital de Bice-tre, Le Kremlin Bicetre, France; 6Service Sante Publique, AP-HP, Hopital de Bicetre, Le Kremlin Bicetre, France; 7Pole Hepato-Gas-troentérologie Médico-Chirurgicale, AP-HP, Hopital Cochin, Paris, France; 8Service des Maladies Infectieuses, AP-HP, Hopital Cochin, Paris, France; 9Service Radiologie, AP-HP, Hopital Jean Verdier, Bondy, France;
10Service Hepato-Gastro-Enterologie, AP-HP, Hopital Jean Verdier, Bondy, France

There is significant increase in liver complications in HIV+/HCV+ co-infected patients (pts), mainly due to a faster fibrosis progression. The prognosis of hepatocellular carcinoma (HCC) in HIV/HCV co-infected pts is globally poor, without clear identification of prognostic factors. The aim of this study was to assess clinical and radiological features at diagnosis of HCC in HIV/HCV co-infected pts. Methods: The Carcinovic cohort is composed by cases of HCC in HIV+/HCV+ pts from three prospective cohorts: Prethevic cohort (n=98) of HIV/HCV pts with end stage liver disease, HepaVih cohort (n=1225) of HIV/HCV co-infected pts and CirVir cohort (n=1823) of pts with viral cirrhosis. HCC controls in HCV mono-infected pts were selected from Cirvir cohort. The diagnosis of HCC was assessed with the determination of typical imaging features, according to AASLD criteria by two experimented radiologists. Results: We diagnosed 35 cases of HCC in HIV+/HCV+ pts (Prethevic n=23, HepaVih n=11), Cirvir n=1) and compared theses patients with 35 HIV-/HCV+ pts (Cirvir). Median age was younger in HIV+/HCV+ pts (50 years [43–65]) compared to HIV-/HCV+ pts (56 years [41–83]) (p<0.0001). At diagnosis, serum AFP in HIV+/HCV+ pts and HIV-/HCV+ pts were 23 ng/ml [3.5–18,740] and 30 ng/ml [4–2,068], respectively, (ns). HIV+/HCV+ pts had good immunity with a median CD4 count at 345/mm3 [24–1121]) with an antiretroviral therapy in 34/35 (97.1%) pts. Median serum AFP was 23 [3.5–18,740] ng/ml in HIV+/HCV+ pts and 30 [4–2,068] ng/ml in HIV-/HCV+ pts (ns). HIV+/HCV+ pts had more decompensated cirrhosis at HCC diagnosis compared to HIV-/HCV+ pts (14/35 (40%) vs 5/35 (14.3%), respectively (p=0.01)). Radiological features showed infiltrative forms in 8/35 HIV+/HCV+ pts (22.9%) vs 0/35 (0%) in HIV-/HCV+ pts (p=0.002). The median diameter of the largest nodule was similar in both groups (25 mm [12–130] in HIV+/HCV+ vs 21 mm [13–90] in HIV-/HCV+, ns). Tumoral portal thrombosis was diagnosed in 10/35 HIV+/HCV+ (28.6%) vs 0/35 in HIV-/HCV+ (0%) pts (p=0.028). A curative treatment was performed in 16/35 HIV+/HCV+ (45.7%) vs 25/35 (73.5%) in HIV-/HCV+ (p=0.01), including liver transplantation in 1/35 (3%) HIV+/HCV+ vs 5/35 (14%) HIV-/HCV+ (ns). Median survival was longer in HIV-/HCV+ pts (27 months [1.5–61]) vs (14 months [0.7–77]) in HIV+/HCV+ pts (p=0.01). Death was due to tumoral progression in 15/35 (42.9%) HIV+/HCV+ pts vs 9/35 (26.5%) in HIV-/HCV+ pts (p=0.0008). Conclusion: HCC has a more advanced radiological presentation at diagnosis in HIV+/HCV+ pts compared to HIV-/HCV+ pts. Such presentation is according to clinical aggressiveness of HCC observed in HIV+/HCV+ co-infected pts.

Disclosures:

Hélène Fontaine - Independent Contractor: gilead, BMS, MSD, Roche

The following people have nothing to disclose: Moana Gelu-Simeon, Maite Lewin, Rodolphe Sobesky, Marita Ostos, Elina Teicher, Faroudy Boufassa, Laurence Meyer, Dominique Salmon, Olivier Seror, Jean Claude Trinchet, Jean-Charles Duclos-Vallee

2099

  1. Top of page

Risk of tumor seeding does not increase by liver biopsy performed during radiofrequency ablation for hepatocellular carcinoma

Soo Young Park, Se Young Jang, Jung Gil Park, Yu Rim Lee, Eun Jung Kang, Won Young Tak, Young Oh Kweon, Jun Young Choi, Sun Young Ahn;
Hepatology, Kyungpook National University Hospital, Daegu, Republic of Korea

Background: Although pathologic diagnosis of hepatocellular carcinoma (HCC) provides important information on accurate diagnosis, differentiation and prognosis of HCC, liver biopsy is not performed widely because procedure might be associated with tumor seedidng during percutaneous ablative procedure. The present study aimed to assess the incidence, risk factors, and prognosis of tumor seeding after liver biopsy during radiofrequency ablation (RFA) for HCC. Method: From January 2009 to December 2012, 955 patients underwent treatment with RFA for HCC in Kyungpook National University Hospital. Two hundred eighty seven patients underwent percutaneous liver biopsy during RFA procedure. Biopsy was performed by coaxial technique using guiding catheter through which automated biopsy needle and RFA electrodes were passed. The risk and pattern of tumor recurrence in patients who underwent biopsy were assessed and compared to the group of patients who do not have significantly difference in tumor size and number. Results: There were no significant differences in baseline characteristics between two groups. There were 2 cases of tumor seeding in RFA electrode tract in each group. The rate of tumor recurrence was 36.2% (104 patients) in biopsy group and 39.8% (96 patients) in non-biopsy group (p=0.418). The risk of local recurrence was not significantly different between two groups (15.7% in biopsy group vs. 13.1% in non-biopsy group, p=0.485). The risk factor associated with tumor recurrence were identified as the number of tumor (OR 1.736, CI 1.148–2.625, p=0.009) Conclusion: Percutaneous liver biopsy of HCC by coaxial technique does not increase the risk of tumor seeding in patients undergoing RFA.

Disclosures:

Won Young Tak - Advisory Committees or Review Panels: Gilead Korea; Grant/Research Support: SAMIL Pharma; Speaking and Teaching: BMS Korea

The following people have nothing to disclose: Soo Young Park, Se Young Jang, Jung Gil Park, Yu Rim Lee, Eun Jung Kang, Young Oh Kweon, Jun Young Choi, Sun Young Ahn

2100

  1. Top of page

Risk of seeding with biopsies of hepatocellular carcinoma: a United States community perspective

Jean-Luc Szpakowski1, Liisa L. Lyon3, Todd E. Drasin2;
1Gastroen-terology, Kaiser Permanente, Fremont, CA; 2Radiology, Kaiser Per-manente, Fremont, CA; 3Division of Research, Kaiser Permanente, Oakland, CA

PURPOSE The reported risk of seeding the needle tract during biopsies for hepatocellular carcinoma (HCC) has varied widely. A 2006 metanalysis reported a 2.29% rate in biopsy proce-dures1. The AASLD 2009 liver biopsy guidelines state "that this risk is almost certainly overstated".2 We undertook a study to determine the risk of seeding in a large community setting.

METHODS A retrospective observational data only study was conducted at a Northern California integrated health care delivery system examining patients undergoing biopsies of HCC between january 1996 and december 2010. Biopsies were done by staff radiologists at one of 16 medical centers. Patients were included if they had undergone a targeted liver biopsy of a radiologically detected lesion that eventually proved to be HCC. Patients were included under diagnosis code 50.11 and procedure codes 47000, 47001, 76003, 77002, 77012. Chart reviews were done as needed to clarify whether a biopsy was targeted or random. Pathology files were reviewed to determine if there were subsequent biopsies with a diagnosis of HCC. This was supplemented by review of radiology records for a lesion compatible with seeding. Possible cases of seeding underwent detailed review by an interven-tional radiologist (TD) and by a hepatologist (JLS). RESULTS 962 patients had percutaneous biopsies, with 400 having multiple biopsies, and 26 also had percutaneous ablation procedures. One case of seeding has been identified, in a patient who also had a percutaneous ablation procedure. The risk of seeding with biopsy was 0.001 (confidence interval 0.0002-.0059). Further detailed electronic chart review is being performed to review the risk with ablations. CONCLUSION Biopsies of lesions that proved to be HCC were associated with a minimal rate of seeding (0.1%) when done within our community setting. 1. Livraghi T, Lazzaroni S, Meloni F, Solbiati L. Risk of tumour seeding after percutaneous radiofrequency ablation for hepatocellular carcinoma. British Journal of Surgery. 2005;92:856–858. 2. Rockey DC, Caldwell SH, Goodman ZD, Nelson RC, Smith AD. Liver Biopsy. Hepatology. March 2009 2009;49(3):1017–1044.

Disclosures:

The following people have nothing to disclose: Jean-Luc Szpakowski, Liisa L. Lyon, Todd E. Drasin

2101

  1. Top of page

Chemotherapy beyond first line in Cholangiocarcinoma

Nora Schweitzer1, Mareike Hoffmann1,2, Sebastian Schmidt1, Martha Kirstein1, Michael P. Manns1, Arndt Vogel1;
1Gastroen-terology, Hepatology and Endocrinology, Hannover Medical School, Hannover, Germany; 2Klinikum Hildesheim, Hildesheim, Germany

Chemotherapy beyond first line in Cholangiocarcinoma Introduction: Cholangiocarcinoma (CC) is a primary cancer of the liver with a low incidence in Western countries. Therefore, large prospective clinical trails are difficult to conduct and there are no randomized trails concerning chemotherapy (CT) and survival beyond first line. Methods: We performed a monocentric retrospective study of patients with CC who presented between 2000 and 2010. Patients with intrahepatic or extrahepatic CC, ampullary carcinoma, gallbladder carcinoma and mixed CC/ hepatocellular carcinoma were included. Results: 570 patients were analyzed. In total, palliative CT of 307 patients was revised. Metastasized and irresectable patients who were able to receive CT had a significant longer survival than patients receiving only best supportive care alone (9 and 13months (m) vs. 2m, p<0.0001). Different CT regimens were applied in 1st line. 58.4% received gemcitabine as single agent and 20.2% combined with a platinum (cisplatin 9.1%, Oxaliplatin 11.1%). In total, 81.8% were treated with gemcitabine alone or in combination with other agents, 12.8% with fluoropyrimidine-based CT and 5.5% with other agents. Survival after 1st line CT with gemcitabine and platinum was better than after gemcitabine alone, although statistically not significant (10m vs 13m, p=0.4). 2nd, 3rd and 4–6th line therapy was applied to 121, 52 and 14 patients, respectively. The median overall survival (mOS) of patients with 1st, 2nd 3rd and 4–6th line CT was 11.7m, 17.0m, 26.8 and 30.9m respectively. In 2nd line, the most frequently used treatment protocol was FOLFOX (23.1%) and other fluoropyrimidine-based therapies (64.4%). Gemc-itabine-based CT was applied in 28%. The survival after start of the 2nd line CT was 9.6m in the gemcitabine-group and 8.1 m in the 5-FU-group (ns) with an overall survival of 24.5m and 21.7m respectively (ns). Median duration of 1st line and 2nd line CT was 104 and 83 days. Patients who were treated more than 104 days in first line survived 1.76 years in median. In contrast, patients who were treated less than 104 days, only survived 7.8 months (p<0.0001). Thus, disease control in 1st line CT seems to be of prognostic importance. Conclusion: Our retrospective data indicate that further lines of chemotherapy beyond 1st line in CC are feasible. Median treatment duration in 2nd line was only slightly shorter compared to 1 st in eligible patients. mOS of patients receiving 2nd line therapy was 22.1 months. Thus, randomized trials are urgently warranted to prospectively investigate the survival benefit of CT after failure of the first line CT in CC.

Disclosures:

Michael P. Manns - Consulting: Roche, BMS, Gilead, Boehringer Ingelheim, Novartis, Idenix, Achillion, GSK, Merck/MSD, Janssen, Medgenics; Grant/Research Support: Merck/MSD, Roche, Gilead, Novartis, Boehringer Ingelheim, BMS; Speaking and Teaching: Merck/MSD, Roche, BMS, Gilead, Janssen, GSK, Novartis

The following people have nothing to disclose: Nora Schweitzer, Mareike Hoffmann, Sebastian Schmidt, Martha Kirstein, Arndt Vogel

2102

  1. Top of page

Age and Hepatocellular carcinoma in HIV-Infected Patients

Maaz B. Badshah1, Beatriz Minguez3, M. Ventura3, Nazia Qazi4,2, Mamta K. Jain5, Marianne Harris6,7, L. E. Taylor8,9, Luciana O. Kikuchi10, Norbert Brau1,11;
1Internal Medicine, JJP VA Medical Center, Bronx, NY; 2George Washington University School of Medicine, Washington, DC; 3Liver Unit, Hospital Univer-sitario Vall d'Hebron, Barcelona, Spain; 4Washington, DC VA Medical Center, Washington, DC; 5University of Texas Southwestern Medical Center, Dallas, TX; 6St. Paul's Hospital, Vancouver, BC, Canada; 7University of British Columbia, Vancouver, BC, Canada; 8Miriam Hospital, Providence, NY; 9Brown University, Providence, RI; 10Universidade de São Paulo, São Paulo, Brazil; 11Divisions of Infectious Diseases and Liver Diseases, Mount Sinai School of Medicine, New York, NY

Background: Hepatocellular carcinoma is diagnosed with increasing frequency in HIV-infected patients. It is unknown what role age plays in the course of this malignancy. Methods: HIV-infected patients with HCC were retrospectively identified from 1995–2013 in 38 centers in 8 countries through local cancer registries. Each HCC diagnosis was confirmed using the 2005 AASLD criteria. Results: Among 255 HIV-infected persons with HCC, the median age was 52.8 years, and 151 (59%) were 50 years or older and 104 (41 %) were under 50 years old. Older subjects were more frequently black (47% vs. 17%, p<0.001) than younger patients, and the etiology of their HCC was more often chronic hepatitis C (83% vs. 70%, p=0.005). They had a similar rate of excessive alcohol consumption (39% vs. 33%, p=0.32), presenting through HCC screening with AFP or imaging (55% vs. 64%, p=0.20) and had a similar mean Child-Turcotte-Pugh score (6.8 vs. 7.1, p=0.16). They presented equally often with multiple tumors 56% vs. 46%, p=0.11), portal vein thrombosis (21% vs. 20%, p=0.95), and distant metastases (14% vs. 17%, p=46), and median AFP levels were similar (198 vs. 138 ng/ml, p=0.56).

They tended to be on HIV therapy less often (81% vs. 89%, p=0.09), but the rate of HIV viral suppression to <400 copies/mL was similar (68% vs. 72%, p=0.49) as was median CD4+ cell count (336 vs. 295 per mm3, p=0.29). Even though older patients presented with Barcelona-Clínic-Liver-Cancer stages A&B more often (54% vs. 39%, p=0.013), they received effective HCC therapy at an equal rate (55% vs. 62%, p=0.33). However, they had significantly shorter median survival (7.9 vs. 16.5 months, p=0.008). In Cox multi-variable regression analysis, age <50 years was independently predictive of survival (H.R., 0.56; 95% C.I., 0.39 - 0.81, p=0.002) together with diagnosis through screening, BCLC staging, HCC therapy and AFP level. Conclusion: Compared to patients under 50, older HIV patients with HCC more often are black and have chronic hepatitis C. Even though they present more often with early BCLC stage A&B, they have a shorter survival

Disclosures:

Mamta K. Jain - Advisory Committees or Review Panels: Merck, Vertex, Boehringer Ingelheim ; Grant/Research Support: Vertex, Bristol-Myers Squibb, Pfizer, Janssen, Gilead, Boehrnger Ingelheim , Viiv Healthcare, EMD Serono

Luciana O. Kikuchi - Speaking and Teaching: Bayer, Bayer, Bayer, Bayer

Norbert Brau - Advisory Committees or Review Panels: Janssen; Grant/Research Support: BMS, Gilead, Vertex; Speaking and Teaching: Vertex, Onyx

The following people have nothing to disclose: Maaz B. Badshah, Beatriz Minguez, M. Ventura, Nazia Qazi, Marianne Harris, L. E. Taylor

2103

  1. Top of page

Predictors of HCC recurrence following liver transplantation: the significance of alpha-fetoprotein

Evangelia Fatourou, James R. Maggs, John G. O'Grady, Michael A. Heneghan, Alberto Quaglia, Varuna Aluvihare, Kosh Agarwal, Nigel Heaton, Abid Suddle;
Liver Unit, King's College Hospital, London, United Kingdom

Background: Liver transplantation (LT) is a curative option for a subset of patients with hepatocellular carcinoma (HCC). High pre-LT alpha-fetoprotein (AFP) and transplantation beyond Milan criteria have been associated with HCC recurrence. Aim: To determine predictors of recurrence following LT in patients meeting radiological Milan criteria (1997–2010) or United Kingdom Transplant (UKT) criteria (2010–2011) in a single liver transplant centre. Methods: We analysed a prospectively collected database of 382 consecutive patients transplanted for HCC from 1997–2011. Explanted livers were evaluated for the number and diameter of lesions, histological differentiation and microvascular or macrovascular invasion. Based on the above, we identified a subset of patients who were beyond Milan criteria histologically. Various epidemiological and laboratory characteristics were examined. Results: Of 382 patients that received liver transplantation, 303 (79.3%) were male, mean age 55.6±9.6 years and mean follow up 59.5±45.3 months. The most common underlying aetiology of liver disease was hepatitis C (n=140, 37%), followed by alcohol (n=82, 21.5%) and hepatitis B (n=64, 16.8%). Based on liver explant findings, 92/382 (24.3%) patients were outside Milan criteria at the time of LT. Fourty-six patients (12.0%) had evidence of HCC recurrence following LT. In univariate analysis, post-transplant recurrence was significantly associated with poor histological differentiation (p=0.04), AFP>100 ng/ml (p<0.0001), microvascular invasion (p=0.004) and transplantation beyond histological Milan criteria. Multivariate analysis showed that AFP>100 ng/ml (OR 4.470, 95%CI 1.702–11.741; P=0.002) and transplantation beyond histological Milan criteria (OR 2.843, 95%CI 1.077–7.504; P=0.035) were independent predictors of tumour recurrence following LT. Conclusions: As LT beyond the histological Milan criteria was associated with significantly increased HCC recurrence, expansion of current radiological criteria is not warranted. Pre-LT AFP levels are a valuable marker of tumour recurrence and should be further explored as part of current listing criteria.

Disclosures:

John G. O'Grady - Advisory Committees or Review Panels: Astellas, Novartis; Speaking and Teaching: Astellas, Roche

Varuna Aluvihare - Speaking and Teaching: Astellas

Kosh Agarwal - Advisory Committees or Review Panels: Gilead, BMS, Novartis, Janssen, Abbott, Gilead, BMS, Novartis, Janssen, Abbott, Gilead, BMS, Novartis, Janssen, Abbott, Gilead, BMS, Novartis, Janssen, Abbott; Grant/Research Support: Roche, MSD, Roche, MSD, Roche, MSD, Roche, MSD; Speaking and Teaching: Astellas, Astellas, Astellas, Astellas

Nigel Heaton - Advisory Committees or Review Panels: Novartis, Roche; Speaking and Teaching: Astellas

The following people have nothing to disclose: Evangelia Fatourou, James R. Maggs, Michael A. Heneghan, Alberto Quaglia, Abid Suddle

2104

  1. Top of page

US GIDEON Final Analysis: Evaluation of Patient and Disease Characteristics, and Sorafenib Safety in Hepatocellular Carcinoma (HCC) Patients with Advanced Liver Disease

Pierre M. Gholam1, Allen Cohn2, Anthony El-Khoueiry3, Jean-Francois H. Geschwind4, Alec Goldenberg5, Jorge A. Marrero6, Robert Martin7, Brendan M. McGuire8, Rebecca A. Miksad9, Bilal Piperdi10, ArunJ. Sanyal11, Alan Venook12, Joseph F. Germino13, Parvez S. Mantry14; 1 Liver Center of Excellence, University Hospitals Case Medical Center, Cleveland, OH; 2Rocky Mountain Cancer Center - US Oncology, Denver, CO; 3USC Norris Comprehensive Cancer Center, Los Angeles, CA; 4Johns Hopkins University Interventional Radiology Center, Baltimore, MD; 5New York University School of Medicine, New York, NY; 6University of Texas Southwestern Medical Center, Dallas, TX; 7University of Louisville, Louisville, KY; 8University of Alabama at Birmingham, Birmingham, AL; 9Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; 10Montefiore Medical Center, New York, NY;
11Virginia Commonwealth University Medical Center, Richmond, VA; 12University of California, San Francisco, San Francisco, CA; 13Bayer HealthCare, Wayne, NJ; 14The Liver Institute at Methodist Dallas Medical Center, Dallas, TX

GIDEON (Global Investigation of therapeutic DEcisions in hepatocellular carcinoma and Of its Treatment with sorafeNib) is a worldwide, prospective, non-interventional study to evaluate sorafenib (SOR) safety in a variety of patients under real-life practice conditions. This report presents only United States (US) data. Patients with unresectable HCC who were candidates for systemic therapy, and for whom a decision was made to treat with SOR, were eligible. Patient and tumor characteristics, SOR dosing, safety, and overall survival (OS) were analyzed by Child-Pugh (CP) status. As an observational study, all results are descriptive. In the US, 645 patients were enrolled; 553 were valid for efficacy and 563 for safety. Results are reported for the safety population except for OS, which refers to those enrolled. In contrast to the SHARP and AP trials, CP A patients comprised 34% of the study population; CP B and CP C comprised 31% and 8%, respectively, at study entry. 27% of patients were considered unevaluable, owing primarily to missing International Normalized Ratio in 22%. Gender, race, age, and body mass index were similar across CP subgroups. Percent distributions of performance status (0/1/2/3/4) were: CP A, 37/38/10/3/0.5; CP B, 20/41/23/3/0; CP C, 14/44/26/12/0. Tumors were predominantly Barcelona Clinic Liver Cancer stage C (36%). Initial and median daily SOR doses were comparable by CP (initial: 800 mg/d in 58%/59%/52%/43%/58% and 400 mg/d in 31%/31%/39%/40%/28%for A, B7, B8, B9, and C, respectively; corresponding median daily dose (mg): 555/593/509/570/515. Frequency of adverse events (AEs) was similar by CP, as were dose interruptions, modifications, and discontinuations. Serious AEs (SAEs) and liver-related AEs were more frequent with worsening liver function and duration of treatment was shorter. However, 33% of CP A, 25% of CP B, and 14% of CP C patients remained on therapy longer than 28 weeks. Median OS was 341, 177, and 118 days for CP A, B, and C patients, respectively. GIDEON is the largest global repository of data of SOR use in HCC patients with advanced CP status. SOR dosing and prescribing patterns were similar by liver function. Not unexpectedly, duration of SOR decreased and SAEs increased as CP status worsened.

%CPAn= 192CP Bn= I73CP C n = 43
  1. *All grades

  2. ‡On study and up to 30 d after last dose

AEs*9997100
Drug-related AEs767351
SAEs*436572
Drug-related SAEs7II2
AEs leading to SOR discontinuation314537
Deaths†274044

Disclosures:

Pierre M. Gholam - Advisory Committees or Review Panels: ONYX, Gilead; Grant/Research Support: ONYX, Gilead; Speaking and Teaching: ONYX, Vertex, Merck, Salix

Allen Cohn - Advisory Committees or Review Panels: genomic health, sanofi; Speaking and Teaching: bayer, pfsizer

Anthony El-Khoueiry - Advisory Committees or Review Panels: Bayer/Onyx; Grant/Research Support: Bayer/Onyx

Jean-Francois H. Geschwind - Consulting: Biocompatibles/BTG, Bayer Health-Care, Guerbet, Nordion, Delcath, Prescience Labs, Jennerex; Grant/Research Support: Biocompatibles/BTG, Bayer HealthCare, Philips Medical, Nordion, Guerbet, DOD, NCI-ECOG, NIH-RO1; Stock Shareholder: PreScience Labs -Founder/CEO

Jorge A. Marrero - Advisory Committees or Review Panels: Bayer, Onyx; Grant/Research Support: Bayer, BMS

Brendan M. McGuire - Consulting: hepahope; Grant/Research Support: roche (genentech) laboratories, bayer healthcare, vital therapies, ikaria therapeutics, cumberland pharmaceuticals, vertex pharmaceuticals

Rebecca A. Miksad - Consulting: OptumInsight, Advance Medical; Grant/Research Support: Bayer/Onyx, NewLink, Sanofi

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

Alan Venook - Grant/Research Support: Bayer Pharm, Onyx Joseph F. Germino - Employment: Bayer Healthcare

Parvez S. Mantry- Consulting: Gilead, Janssen; Grant/Research Support: Vertex, Merck, Salix, Abbvie, Gilead, Boehringer-Ingelheim, Bristiol Myers Squibb, San-taris; Speaking and Teaching: Vertex, Merck, Sailx, Genentech, Bayer-Onyx, Kadmon

The following people have nothing to disclose: Alec Goldenberg, Robert Martin, Bilal Piperdi

2105

  1. Top of page

Diabetes Increases the Risk of Hepatocellular Carcinoma in Alcohol but not Hepatitis C Related Cirrhosis

Jennifer A. Flemming1, W. Ray Kim3, Eric Vittinghoff2, Norah Ter-rault1;
1Medicine, University of California San Francisco, San Francisco, CA; 2Epidemiology and Biostatistics, University of California San Francisco, San Francisco, CA; 3Medicine, Mayo Clinic, Rochester, MN

The presence of diabetes (DM) is associated with hepatocellular carcinoma (HCC) in patients with cirrhosis. Whether this effect differs based on the cause of cirrhosis remains to be determined. Methods: We constructed a cohort of patients with hepatitis C (HCV, n= 10,607), alcohol (EtOH, n=5,452) and HCV+EtOH (n=3,204) related cirrhosis without HCC listed for liver transplantation in the US between 03/2002 and 12/2011. Incident HCC was identified by HCC-MELD exception at least six months after listing. The interaction between the presence of diabetes and etiology of cirrhosis on HCC risk was evaluated using multivariate Cox regression analysis. Results: The cohort was 70% male with a median age at cohort entry of 53 years (interquartile range (IQR): 48–58 yrs) and a median MELD of 13 (IQR: 10–16). Diabetes was present in 16.1% with HCV 12.9% with HCV+EtOH and 15.7% with EtOH (p <.001). After a median follow-up of 1.3 years (IQR: 0.5 - 2.8 years), 1,307 patients (6.8%) developed HCC for an overall incidence rate (IR) of 3.6 per 100 person-years (py). The IR in HCV, HCV+EtOH, and EtOH were 4.4 (95% CI 4.1–4.7), 4.1 (95% CI 3.6–4.7), and 1.7 (95% CI 1.5–2.0) per 100-py respectively. DM was a statistically significant risk factor for HCC among patients with EtOH-related cirrhosis, but not among those with HCV or HCV+EtOH related disease (figure). The interaction was confirmed based on the Mantzel-Hanzel test for homogeneity (p <.001). In the multivariate Cox regression analysis, a significant interaction was found between diabetes and EtOH cirrhosis on incident HCC (p <.001). Conclusion: In HCV cirrhosis patients (with or without EtOH), the risk of HCC already high and diabetes may not increase the risk further. However, in patients with EtOH cirrhosis, diabetes increases the risk of HCC by more than two fold, highlighting that this population should be targeted for HCC surveillance.

Thumbnail image of

Disclosures:

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

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

The following people have nothing to disclose: Jennifer A. Flemming, Eric Vit-tinghoff

2106

  1. Top of page

A novel and simple real world predictive risk score (RWS-HCC) for the development of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B (CHB) infection

Zhongxian Poh1, Liang Shen2, Boon Bee George Goh1, Jason Chang1, Chee-Kiat Tan1;
1Gastroenterology & Hepatology, Singapore General Hospital, Singapore, Singapore; 2Biostatistics Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore

Background/Aims: The development of HCC is a sentinel event in the natural history of CHB infection. Current HCC predictive risk scores in CHB patients include HBV DNA quantification which is a costly test not available in all parts of the world. Globally, the majority of CHB patients are seen in healthcare settings where only simple liver biochemistries are available, thus limiting the applicability of such scores. Serum transami-nases have shown to have high discriminant values as independent predictors of HCC.1 We aim to identify independent risk factors, and create an easy and practical RWS-HCC for HCC development based on baseline parameters of 673 CHB patients in a 10-year prospective HCC surveillance programme. Methods: Outpatients with CHB seen at our Department of Gastroenterology and Hepatology between 1 st March 2003–1 st March 2004 were enrolled in a physician driven HCC surveillance program comprising 3–6 monthly liver biochemistries and serum AFP, and 6–12 monthly imaging. Baseline liver biochemistries and AFP were done. Cirrhosis was diagnosed on histology or imaging with supportive clinical evidence. HCC was diagnosed on dynamic CT/MRI scan or histology. Census was done 10 years later on 1st March 2013. Results: 673 patients were enrolled with 545(81%) still on follow-up after 10 years. 62.6% were males. Mean age was 56.4years (±12.7). The rates of cirrhosis and HCC development were 1.8%/year and 0.8%/year respectively over 10 years. The rate of HCC development in cirrhotics was 3.0%/year. Male gender, age, AST and AFP were independent risk factors for HCC development. A risk score was derived with AUC 0.873 (95% CI: 0.823 - 0.923). (See table) 85.0% sensitivity and 76.4% specificity was achieved with RWS-HCC > 7.5 for prediction of risk of HCC development within 10 years. Conclusion: We have formulated a simple and practical RWS-HCC utilizing only gender, age and basic liver biochemistry (serum AST and AFP) without HBV DNA quantification that can predict the risk of HCC development within 10 years. We plan to validate this score in various patient populations. This can potentially be used to identify high risk CHB patients for enhanced HCC surveillance. Reference: 1. CP Wen et al. J Natl Cancer Inst 2012;104:1599–1611

Real World Predictive Risk Score (RWS-HCC)

Risk FactorORp-valueScore assigned
Male5.940.0013
Age>55 years3.330.0072
AST 27–33U/L AST>33U/L3.85 5.560.028 0.00123
AFP 4.1 -20ug/L AFP>20ug/L5.15 14.2<0.0001 <0.00013 4.5

RWS-HCC > 7.5 predicts 10 year risk of HCC development. AUC 0.873 (95% CI:0.823 - 0.923) 85.0% sensitivity and 76.4% specificity

Disclosures:

The following people have nothing to disclose: Zhongxian Poh, Liang Shen, Boon Bee George Goh, Jason Chang, Chee-Kiat Tan

2107

  1. Top of page

Efficacy of continued sorafenib treatment after radiological confirmation of disease progression in patients with advanced hepatocellular carcinoma

Yoshiyuki Wada1, Yuko Takami1, Tomoki Ryu1, Masaki Tateishi1, Hideki Saitsu1,2;
1 Department of Hepato-biliary-pancreas surgery, National Kyushu Medical Center, Fukuoka, Japan; 2Clinical Research Center, National Kyushu Medical Center, Fukuoka, Japan

Background: The survival benefits of sorafenib treatment for advanced hepatocellular carcinoma (HCC) have been established; however, the effects of continued sorafenib treatment for progressive disease remain unclear. Although some studies reported good responses to sorafenib treatment that was continued even after radiological confirmation of disease progression, the effectiveness of the therapy remains unknown. Here we investigated the clinical outcomes of sorafenib treatment that was continued after radiological confirmation of disease progression to determine the feasibility of this treatment in patients with advanced HCC. Materials and methods: From June 2009 to March 2013, 110 advanced HCC patients received sorafenib treatment, of which 96 underwent assessment for estimation of radiological response. A total of 79 patients with radiologically confirmed disease progression were enrolled to investigate the effectiveness of continuous sorafenib treatment. The median survival time (MST) was compared between patients who continued treatment after confirmation of disease progression and those who did not, and it was also evaluated in the former group after continued treatment. Moreover, time-to-progression (TTP; >4 and <4 months) was evaluated before and after continued treatment to determine criteria for assessing patients eligible to receive continued treatment. Radiological tumor response was estimated using the modified RECIST criteria. Results: Of the 79 patients with radiologically confirmed disease progression, 43 continued to receive sorafenib treatment while 36 discontinued treatment. MST in the former group (14.3 months) was greater than that in the latter group (10.7 months; p = 0.02). Among the 43 patients who continued treatment, time-to-progression (TTP) was >4 months in 20. Of these, 15 (75%) exhibited stable disease and 5 (25%) exhibited disease progression after continued sorafenib treatment. TTP and MST after continued treatment in this group were 7.1 months and 12.9 months, respectively. TTP in the remaining 23 patients was <4 months. Of these, 3 (13%) exhibited SD and 20 (87%) exhibited disease progression after continued sorafenib treatment. TTP and MST after continued treatment in this group were 1.8 months (p < 0.01) and 8.9 months, respectively (p = 0.11). Conclusion: Continued sorafenib treatment after radiological confirmation of disease progression improved survival in patients with advanced HCC, particularly those with an initial TTP of >4 months.

Disclosures:

The following people have nothing to disclose: Yoshiyuki Wada, Yuko Takami, Tomoki Ryu, Masaki Tateishi, Hideki Saitsu

2108

  1. Top of page

Systemic Metabolic Derangements or Histologic Features on Liver Biopsy are not Associated with Increased Risk of HCC in NASH

Achuthan Sourianarayanane1, Patrick Varley1, Allan Tsung2, Jaideep Behari1;
1Department of Gastroenterology, Hepatology and Nutrition, University of Pittsburgh Medical Center, Pittsburgh, PA; 2Department of Surgery, University of Pittsburgh Meredical Cent, Pittsburgh, PA

Background and Aim: Epidemiological studies have suggested an association between nonalcoholic steatohepatitis (NASH) and increased risk of developing hepatocellular carcinoma (HCC). The aim of our study was to elucidate whether NASH patients with or without HCC exhibit differences in liver histology or systemic metabolic derangements (diabetes, hypertension, dyslipidemia, and obesity), which could be used to identify NASH patients at higher risk of developing HCC. Methods: We utilized a retrospective case-control study design. We queried electronic medical records at our institution for patients with biopsy-proven NASH treated between 1997 and 2011. We identified 101 cases with HCC and intraoperative liver biopsy demonstrating background NASH. For controls, we identified 80 patients with biopsy-proven NASH with no evidence of HCC on imaging studies. We excluded patients with serologic or biochemical evidence of other co-existing liver diseases. We recorded patient demographics, clinical and laboratory results and liver histology findings. We performed univariate and multivariate analysis for factors associated with HCC in NASH. Results: Patients with NASH and HCC were older compared with patients with NASH without HCC (69 vs 54 years; p<0.0001), had lower body mass index (30.8 vs 33.7; p<0.05), and metabolic derangements (2 vs 3; p<0.01). On liver histology, NASH/HCC patients compared with patients without HCC demonstrated lower steatosis scores (1.2 vs 1.6; p<0.001), and NASH activity scores (NAS; 2 vs 2.9; p<0.01). However, there were significantly more cirrhotics (55 vs 45%; p<0.01) among HCC patients. There was no difference in Model for Endstage Liver Disease (MELD) scores among NASH-cirrhosis patients with or without HCC. Multivariate analysis including all patients demonstrated that age was independently associated with increased risk of HCC (OR 1.08, 95% CI: 1.02, 1.13) while the number of metabolic derangement present (OR 0.68, 95% CI:0.47, 0.97) and presence of steatosis (OR 0.4, 95%CI:0.16, 0.96) were associated with lower risk of HCC. In the subgroup of patients with biopsy-proven cirrhosis, on multivariate analysis only age was independently associated with increased risk of HCC (OR 1.1, 95%CI:1.05, 1.18) for increased risk of HCC. Conclusion: HCC occurred in NASH patients in the absence of cirrhosis although NASH patients with HCC were older and more likely to exhibit advanced fibrosis/cirrhosis compared with NASH patients without HCC. However, greater number of metabolic derangements or higher fibrosis and NAS scores on liver biopsy were not associated with increased risk of HCC in NASH patients.

Disclosures:

The following people have nothing to disclose: Achuthan Sourianarayanane, Patrick Varley, Allan Tsung, Jaideep Behari

2109

  1. Top of page

Primary Care Physician Attitudes and Practices Regarding Screening for Hepatocellular Carcinoma

Christopher E. McGowan, Mai-Uyen T. Luong, Paul H. Hayashi;
UNC Liver Center, University of North Carolina, Chapel Hill, NC

Background: Only a minority of cirrhotic patients receive screening for hepatocellular carcinoma (HCC). Primary care physicians (PCP's) are often the most accessible providers for cirrhotic patients. Community effective screening will need PCP participation, but their attitudes and knowledge of HCC screening are not well described. Aim: We aimed to assess PCP screening practices, and knowledge of screening recommendations & HCC therapies. Methods: A written survey study was conducted among North Carolina primary care physicians (Internal Medicine and Family Practice). A random sample of 1000 physicians was generated from the North Carolina Medical Board Roster Report. Physicians received an introductory letter, followed by a 12-item questionnaire addressing physician demographics, screening practices, and knowledge of HCC screening guidelines and treatment options. Results: 391 (39%) PCP's completed the survey. Mean years in practice was 22 (SD: 9.9), 60% were male, 83% in private practice, 12% in academics, and 3.4% in the Veterans Administration. A high percentage (88%) cared for cirrhotic patients in their practice, but only 45% screen for HCC. There was no association between gender, practice type, or years in practice and screening. Among PCP's who screen, the most common modality was ultrasound and alpha feta protein at 12-month intervals (66%). Reasons for screening included supported by available evidence (72%), recommended by medical societies (42%), and malpractice liability for not screening (26%). Perceived barriers to screening included poor patient adherence (56%), out-of-pocket cost (56%), lack of insurance (51%), and insurance restrictions on coverage (32%). Of PCP's who don't screen, 91% refer patients to subspecialists, 26% were unaware of recommendations, 8% uncertain of benefit, and 8% concerned over cost. Only hepatic resection was identified by a majority (59%) of PCPs as an effective therapy. Other therapies were identified as effective by less than half each: liver transplanta-tion-49%, transarterial chemoembolization-37%, radiofre-quency ablation-31%, and sorafenib-23%. Summary: PCP's routinely see cirrhotic patients, but only a minority screen for HCC. Those who do not screen rely heavily on subspecialists to consider screening. PCP knowledge of effective HCC therapy options is poor. Conclusions: Efforts to enlist PCP's in HCC screening should focus on increasing their awareness of effective therapies. Such awareness may lead to increased belief in screening benefit and willingness to order ultrasounds themselves, as opposed to relying on subspecialty referral that adds an unnecessary step to community effectiveness.

Disclosures:

The following people have nothing to disclose: Christopher E. McGowan, Mai-Uyen T. Luong, Paul H. Hayashi

2110

  1. Top of page

Hepatocellular carcinoma: ART score can predict response after TACE : Validation in a large HCC cohort mostly associated with HCV and HBV

Adhoute Xavier1, Guillaume Penaranda3, Paul Castellani1, Herve Perrier1, Bernard L. Pol1, Olivier Monnet1, Gaelle Le Folgoc1, Beau-rain Patrick1, Bayle Olivier1, Manuela Campanile1, Jean-Luc Raoul2, Marc Bourlière1;
1Saint-joseph Hospital, Marseille, France; 2Paoli-Calmettes Institute, Marseille, France; 3Alpha-Bio Laboratory, Marseille, France

TACE is recommended for intermediate BCLC B HCC stage and could be an option for inoperable limited HCC. The ART score allows to selects which HCC patients should receive a second TACE. It is calculated before the second TACE and based on increased transaminases and Child-Pugh score and the radiological response. Based on a cohort of 107 pts with predominantly alcoholic cirrhosis, it distinguished two different prognostic groups 0–1.5 points and ≧ 2.5 points with respective survival of 23.7 vs 6.6 months (p <0.001). Sieghart W, Hepa-tology. 2013. Aim of study : to assess the prognostic value of ART score in a French cohort of HCC developed mostly on viral cirrhosis. Material and method: During the period 01/2007 -12/2012, 373 HCC consecutive pts were admitted in our unit. 52% were treated by TACE. We exclud TACE before a LT and pts with additional treatment after TACE. The population therefore included 139 pts, mostly male, mean age 66 years. Cirrhosis was present in 96% of cases. Esophageal varices grade 2 or 3 were present in 37%. Underlying liver disease was viral in 47%, mixed in 5% (HCV, alcohol), alcohol-related 34%, a steatopathy 10%. 45% of pts were within Milan criteria. These pts for age or medical reasons could not get a LT. 45% of HCC were BCLC A, 32% BCLC B, 19% BCLC C. Segmental portal vein thrombosis was present in 16% of HCC. Pts were treated on average by two TACE. We distinguished two groups: 0 to 1.5 points A and 2.5 points or more B. Results: Before the second TACE, 43 pts (31%) had a score ART > 2.5 and 96 pts (69%) a score between 0 and 1.5. The median overall survival of our population was 28 months (22, 34). There was a significant difference between the two groups regarding median survival (p <0.0001), with a median of 34 months A (28–38) and 13 months B (10–16). There was a significant difference between the two groups regarding the value of AFP, Child-Pugh score, BCLC classification, encapsulated character or infiltrating tumor, the presence of a segmental portal vein thrombosis, in radiologic response and with the median time to progression. There was no difference regarding patient age, BMI, underlying liver disease, the presence of diabetes, the circumstances of discovery, the presence of significant VO, the existence of a previous treatment. In univariate analysis, Child-Pugh score and ASAT increase, response to TACE were prognostic factors. Conclusion: This study confirms the prognostic value of ART score calculated before the second TACE in a different population with more often viral disease. Elevated transaminases frequently observed in viral diseases do not interfere on the reliability of the ART score.

Disclosures:

Jean-Luc Raoul - Advisory Committees or Review Panels: Bayer SP, BMS, Nordion, Arqule; Speaking and Teaching: Bayer SP

Marc Bourlière - Advisory Committees or Review Panels: Schering-Plough, Bohringer inghelmein, Merck, Schering-Plough, Bohringer inghelmein, Merck ; Board Membership: Bristol-Myers Squibb, Gilead, Bristol-Myers Squibb, Gilead; Consulting: Roche, Novartis, Tibotec, Abott, glaxo smith kline, Roche, Novartis, Tibotec, Abott, glaxo smith kline

The following people have nothing to disclose: Adhoute Xavier, Guillaume Penaranda, Paul Castellani, Herve Perrier, Bernard L. Pol, Olivier Monnet, Gaelle Le Folgoc, Beaurain Patrick, Bayle Olivier, Manuela Campanile

2111

  1. Top of page

Epithelial to mesenchymal transition affects the stem cell-like features of primary liver cancers

Naoki Oishi1,2, Juling Ji2, Taro Yamashita1, Qinghai Ye3, Shuichi Kaneko1, Xin W. Wang2;
1Department of Gastroenterology, Kanazawa University Hospital, Kanazawa, Japan; 2Laboratory of Human Carcinogenesis, National Cancer Institute, Bethesda, MD; 3Liver Cancer Institute, Fudan University, Shanghai, China

Background: The epithelial to mesenchymal transition (EMT) is a physiological cell-reprogramming event utilized in tissue remodeling during embryonic development. The presence of EMT-like cells in tumors has been linked to their increased inva-siveness and metastasis. While it is hypothesized that cancer progression is driven by the presence of cancer stem cells (CSCs), which are also responsible for treatment resistance and tumor relapse, it is unclear how EMT affects stem cell-like features. In this study, we characterized two different primary liver cancer (PLC) subgroups, epithelial-like and mesenchymal-like PLC, and identified the relationship between the CSC features of these subgroups. Materials and Methods: Three cohorts were used for analysis. Cohort 1 consisted of 242 asian hepatocellular carcinoma (HCC) samples, cohort 2 consisted of 139 caucasian HCC samples, and cohort 3 contained 43 asian intrahepatic cholangiocarcinoma and asian combined hepato-cellular-cholangiocarcinoma samples. Unsupervised hierarchical clustering analysis and class comparison analysis were performed using BRB-Array Tools software. Transcription factor analysis and gene network analyses were performed using Ingenuity Pathway Analysis. Results: We classified samples as epithelial-like HCC or mesenchymal-like HCC based on the expression ratio of E-cadherin to vimentin. In the mesenchymal-like HCC subgroup, stem cell markers, TGF-β1, ZEB2, and EMT markers, were highly expressed, and metastasis occurred at a high frequency. These data indicated that mesenchymal-like HCC had stemness features with more aggressive tumors, resulting in poor outcome. EpCAM+ and NCAM1+ cells were observed in PLC. The expression levels of EpCAM were significantly associated with a poor outcome in epithelial-like PLC, while high levels of NCAM1 expression were associated with poor prognosis in mesenchymal-like PLC. Next, we identified 1117 EpCAM-associated genes in epithelial-like HCC and NCAM1-associated genes in mesenchymal-like HCC by class comparison. The functional connectivity of the gene network showed further interactions with MYCN and MYC in epithelial-like HCC cases or with CTNNB1 and SMAD4 in mesenchymal-like HCC cases. These results indicate that the activation of these transcription factors may be a key genetic or epigenetic event associated with poor outcome. Conclusion: EMT affects reprogramming factors including transcription factors in PLCs, and the mechanisms of CSCs in the two subgroups were different. The classification of epithelial-like or mesenchymal-like subgroups may be useful for personalized CSC-targeted therapy of PLCs.

Disclosures:

Shuichi Kaneko - Grant/Research Support: MDS, Co., Inc, Chugai Pharma., Co., Inc, Toray Co., Inc, Daiichi Sankyo., Co., Inc, Dainippon Sumitomo, Co., Inc, Aji-nomoto 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: Naoki Oishi, Juling Ji, Taro Yamashita, Qinghai Ye, Xin W. Wang

2112

  1. Top of page

Diagnosis Of Hepatocellular Carcinoma Using A GALAD Model By Objective Clinical And Serological Factors

Hidenori Toyoda1, Takashi Kumada1, Chiaki Kagebayashi2, Shinji Satomura2, Mabel Teng3, Phillip Johnson3;
1 Ogaki Municipal Hospital, Ogaki, Japan; 2Wako Life Sciences, Inc., Mountain View, CA; 3School of Cancer Sciences, University of Birmingham, Birmingham, United Kingdom

Late diagnosis of hepatocellular carcinoma (HCC) frequently results in poor patient outcome. Therefore, routine surveillance is recommended to detect early stage HCC to be able to apply curative treatments as early as possible. The most common tests used for surveillance are alpha-fetoprotein (AFP) tests and ultrasound (US). However, interpretation of US can be challenging without comparison to previous imaging results. Performance of US can be limited in patients who are obese or have severe background liver cirrhosis. US quality is user dependent which directly affects quality of, US imaging and its ability to be used to detect HCC lesions early. Therefore, reliable serological bio-markers are needed. Lens culinaris agglutinin-reactive fraction of alpha-fetoprotein (AFP-L3) and des-gamma-carboxy pro-thrombin (DCP) are biomarkers widely used for surveillance in Japan. These biomarkers are complementary and their simultaneous measurement is recommended. In this study, we propose a newly developed and validated statistical model using the biomarkers and objective factors for HCC diagnosis. The model, which is called GALAD (G as Gender, A as Age, L as AFP-L3, A as AFP and D as DCP) was built by combining these gender and age with the three high risk of developing HCC, by Johnson et al. of University of Birmingham. The GALAD model was developed using the clinical data from Ogaki Municipal Hospital, Japan, and University of Birmingham, UK. For the study, sera from 1045 patients with either chronic liver disease or HCC were collected from Ogaki Municipal Hospital and 658 sera were assayed from the University of Birmingham. The two different datasets were evaluated by the GALAD model using Receiver Operator Characteristic analysis and area under the curve analysis (AUC). The GALAD model gave consistently high figures for the AUC in the datasets from both Ogaki and Birmingham (0.82 and 0.97, respectively). The AUC values of the subgroup of patients who had a tumor sizes over 2 cm were higher than that of patients with tumors less than 2 cm (0.79 and 0.90 for Ogaki and Birmingham, respectively). In conclusion, the GALAD model may increase the specificity of diagnosis of HCC on the grounds of objective clinical and serological factors.

Disclosures:

Shinji Satomura - Management Position: Wako Life Sciences, Inc

Phillip Johnson - Advisory Committees or Review Panels: Novartis, MedImmune, Astellas; Consulting: BMS, ISIS; Speaking and Teaching: Bayer

The following people have nothing to disclose: Hidenori Toyoda, Takashi Kumada, Chiaki Kagebayashi, Mabel Teng

2113

  1. Top of page

Acquisition of aldo-keto reductase family 1 member B10 expression in the early stages of human hepatocarcinogenesis

Hironori Tsuzura, Takuya Genda, Shunsuke Sato, Ayato Murata, Yoshio Kanemitsu, Yutaka Narita, Sachiko Ishikawa, Tetsu Kikuchi, Masashi Mori, Katsuharu Hirano, Katsuyori Iijima, Takafumi Ichida;
gastroenterology and hepatology, Juntendo university shizuoka hospital, Izunokuni-shi, Japan

Background and Aims: Aldo-keto reductase family 1 member B10 (AKR1 B10) is an enzyme that converts retinals into retinols.

Its upregulation reduces intracellular retinoic acid, resulting in the inhibition of cell differentiation. As AKR1B10 gene expression is increased in human hepatocellular carcinoma (HCC), its involvement in hepatocarcinogenesis is intriguing. AKR1B10 expression increases in the early stages of well-differentiated HCC but is minimal in normal liver tissue, similar to the expression of the heat shock protein 70 (HSP70) and glypican3 (GPC3). We previously analyzed the expression profiles of approximately 41,000 genes in patients with chronic hepatitis C (CHC) and found that AKR1B10 was upregulated in the livers of CHC patients who were at high risk of HCC (Liver Int 2012). To further study the significance of AKR1 B10 alterations in the early stages of hepatocarcinogenesis, we evaluated AKR1B10 expression in HCCs and corresponding nontumorous liver tissues (NTs) showing chronic hepatitis and cirrhosis, which are considered precancerous conditions. Methods: Sixty-one paired samples of primary HCCs and corresponding NTs and 8 control samples of histologically normal liver tissues (NLs) were obtained from surgically resected tissue. The expressions of AKR1B10, HSP70, and GPC3 were examined by using immunohistochemical analyses and were quantified as percentages of positive staining area by using image analysis software. Univariate and multivariate regression analyses were used to estimate the relationship of AKR1B10 expression with demographic, histological, and biochemical variables in NTs. Results: Median AKR1B10 expression levels were 54.8%, 2.1%, and 0.3% in HCCs, NTs, and NLs, respectively. AKR1 B10 levels were higher in HCCs than in NTs (P < 0.001) and higher in NTs than in NLs (P < 0.001). HSP70 and GPC3 were expressed in HCCs, but minimally in NTs and NLs. Both HSP70 and GPC3 expression were not significantly different between NTs and NLs. AKR1B10 expression in NTs was not associated with fibrosis stage, etiology of liver disease, or biochemical variables. Multivariate analysis identified that only hepatic steatosis was associated with AKR1 B10 expression in NTs (P = 0.020). There was a significant positive correlation between AKR1 B10 expression and hepatic steatosis in NTs (r = 0.40, P = 0.001). Conclusions: Of the 3 early molecular markers of HCC, only AKR1 B10 was upregulated in precancerous conditions. The stepwise upregulation of AKR1B10 suggests its involvement in the early stages of human hepatocarcinogenesis. Hepatic steatosis-related metabolic disorders might influence AKR1 B10 expression in precancerous conditions.

Disclosures:

The following people have nothing to disclose: Hironori Tsuzura, Takuya Genda, Shunsuke Sato, Ayato Murata, Yoshio Kanemitsu, Yutaka Narita, Sachiko Ishikawa, Tetsu Kikuchi, Masashi Mori, Katsuharu Hirano, Katsuyori Iijima, Takafumi Ichida

2114

  1. Top of page

The role of microRNA in serum exosomes on recurrence of human hepatocellular carcinoma after liver transplantation

Keishi Sugimachi, Tae Matsumura, Ryutaro Uchi, Hiroki Ueo, Hidetoshi Eguchi, Tomoya Sudo, Koshi Mimori;
Surgery, Kyushu University Beppu Hospital, Beppu, Japan

Background: The predictive biomarker for the recurrence of hepatocellular carcinoma (HCC) has great benefit on the selection of treatment options including liver transplantation (LT) for HCC. The objective of this study is to identify specific miRs in exosomes of HCC recurrence from the serum of the patients and validate these molecules as novel biomarkers for HCC recurrence. Methods: Serum exosome was purified by ultracentrifuge method. We employed microarray-based profiling of miR expression from exosome in serum and bone marrow of the patient with HCC to define a biomarker that distinguishes between patients with and without HCC recurrence after liver transplantation or hepatic resection. This was followed by a real-time semi-quantitative PCR validation in a separate cohort of 59 HCC patients who underwent living related LT. Result: We found that 2 miRs (miR-718 and miR-1246) showed significantly different expression in the serum exosome with HCC cases that had recurrence after LT compared to those without recurrence. Decreased expression of miR-718 was associated with tumor aggressiveness of HCC in the validated cohort series, and precursor miR-718 transfected HCC cells showed suppressed cell proliferation in vitro. The altered expression of a miR-718 target gene was associated with aggressive behavior of HCC. Conclusions: Circulating miR in serum exosome has potential as a novel biomarker to predict recurrence of HCC.

Disclosures:

The following people have nothing to disclose: Keishi Sugimachi, Tae Matsumura, Ryutaro Uchi, Hiroki Ueo, Hidetoshi Eguchi, Tomoya Sudo, Koshi Mimori

2115

  1. Top of page

Is Detection of Hypervascular Foci Present at Angiogra-phy but not Visible on Cross-Sectional Imaging Associated with a Removal from Transplant List due to Hepatocellular Carcinoma Progression or Post-transplant Tumor Recurrence?

Kellie Young1, Nicholas Fidelman1, Francis Y. Yao2, Robert Ker-lan1;
1Department of Radiology, University of California San Francisco, San Francisco, CA; 2Division of Gastroenterology and Hepatology, University of California San Francisco, San Francisco, CA

PURPOSE. To determine whether detection of more or fewer hypervascular hepatic foci suspicious for hepatocellular carcinoma (HCC) at angiography performed prior to transarterial chemoembolization (TACE) than expected based on preceding contrast-enhanced CT/MRI led to adverse clinical outcomes. MATERIALS AND METHODS. 221 adult liver transplant candidates with hypervascular HCC (diagnosed by AASLD imaging guidelines) who underwent TACE at our institution between 2006 and 2010 were included. Patients were grouped based on the agreement between angiography and contrast-enhanced CT or MRI into three categories: (1) number of nodules at CT/MRI was concordant to the number of hypervascular foci detected at angiography (n=139); (2) Number of nodules at CT/MRI was less than the number of hypervascular foci at angiography (n=37); (3) Number of nodules at CT/MRI was greater than the number of hypervascular foci at angiography (n=45). Study outcomes were removal from transplant list for HCC progression or any other reason, HCC recurrence following transplant, overall survival, and extent of viable HCC in explanted organs (based on HCC TNM staging by Marsh, et al., Cancer 2000;88:538). Odds ratios were calculated between the patient groups in univariate and multivariate logistic regression models with p<0.05 considered statistically significant. RESULTS. Total number of evaluable matched angiography and CT/MRI imaging sets was 376 (median 1,1–6 per patient). Discordance between number of hypervascular foci at angiography and the findings at preceding CT/MRI in either direction were not significantly associated with any adverse clinical or pathologic outcomes tested. However, trends were seen towards worse overall survival if 3 or more discrepant foci were present at angiography than at CT/MRI (p=0.07) and towards detection of higher active disease burden (T3 or T4) at explant when 1 or 2 more hypervascular foci were detected at angiography than at CT/MRI (p=0.06). CONCLUSION. Discordance between the number of hypervascular foci detected at angiography and the number of suspicious nodules at cross-sectional imaging may not have a significant impact on patients' odds of de-listing due to HCC progression while on transplant list, HCC recurrence after transplant, overall survival, or high viable tumor burden in the explants

Disclosures:

The following people have nothing to disclose: Kellie Young, Nicholas Fidelman, Francis Y. Yao, Robert Kerlan

2116

  1. Top of page

Hepatocellular Carcinoma (HCC) in The Setting of Chronic Hepatitis B Virus Infection (HBV): Recurrence and Survival Rates Following Liver Transplantation (LT)

Mohannad Dugum1, Ibrahim A. Hanouneh2, Rocio Lopez4, Fed-erico N. Aucejo3, Bijan Eghtesad3, NizarN. Zein2;
1 Internal Medicine, Cleveland Clinic Foundation, Cleveland, OH; 2Gastroenterology and Hepatology, Cleveland Clinic Foundation, Cleveland, OH; 3Transplantation Center, Cleveland Clinic Foundation, Cleveland, OH; 4Quantitative Health Sciences, Cleveland Clinic Foundation, Cleveland, OH

Data from non-US patients (pts) suggested a higher incidence of tumor recurrence of HCC in the setting of chronic HBV infection following LT. These findings may have important implications on pts selection and post-LT monitoring. AIM: To assess the rate of post-LT HCC recurrence in HBV and non-HBV pts using the United Network for Organ Sharing (UNOS) database. METHODS: All pts who underwent LT for HCC in USA (1987–2013) were included. Kaplan-Meier curves were generated to test for differences in recurrence of HCC and survival following LT. RESULTS: 17,074 pts with HCC were included of which 144 had HBV. Average age was 57 ± 8 years and 77% were males. 98% of LT were cadaveric and only 5% of pts had tumors beyond Milan criteria at the time of LT. Pts were followed for 39 ± 36 months after LT during which 6% had HCC recurrence. Among pts within Milan criteria those with HBV did not have a significantly increased risk of HCC recurrence following LT compared to all other etiologies of liver disease combined (HR=1.3, 95% CI 0.74–2.3, p-value 0.35). However, compared to each group individually; pts with HBV had increased risk of recurrent HCC post-LT than those with alcoholic liver disease (HR=2.3, 95% CI 1.2–4.3), HCV with alcoholic liver disease (HR=2.1, 95% CI 1.06–4.1) and A1A-deficiency (HR=3.4, 95% CI 0.77–16.7). There was no difference in survival outcome post-LT between pts with HBV compared to any of the other etiologies. CONCLUSION: Pts with HCC secondary to HBV infection being considered for LT may need further stringent evaluation even if the tumor is within Milan criteria as they are more likely to experience tumor recurrence than other specific etiologies of liver disease.

Thumbnail image of

Disclosures:

Bijan Eghtesad - Grant/Research Support: Genzyme (Sanofi)

The following people have nothing to disclose: Mohannad Dugum, Ibrahim A. Hanouneh, Rocio Lopez, Federico N. Aucejo, Nizar N. Zein

2117

  1. Top of page

Multiple adenoma and liver adenomatosis: incidence, patient characteristics and management

Mirelle Bröker, Robert A. de Man, Bettina E. Hansen, Pavel Taimr, Susanna van Aalten, Jan IJzermans;
Erasmus Medical Univercity Center, Rotterdam, Netherlands

Previously it has been described patients with hepatocellular adenoma usually have a solitary nodule and only 10–24% present with multiple adenomas or liver adenomatosis. There are no evidence-based guidelines how to treat patients with multiple adenomas or liver adenomatosis. In this study we investigated the incidence and presentation of patients with multiple lesions as compared to patients with a solitary adenoma. All patients with adenoma admitted to our institute between 2001 and 2012 were included in this study. In a prospective database information was collected about patients' and lesion characteristics, management and outcome of follow up. Data were compared stratifying patients in three groups; patients with a single adenoma, multiple adenoma's and liver adenomatosis. 331 patients were diagnosed with hepatocellular adenoma, including 87(26%) with a single adenoma, 162(49%) with multiple adenoma and 82(25%) with liveradenomatosis. Body Mass Index was significantly different between groups (<0.05). In follow up no malignant degeneration was observed in any group; regression of adenoma size was seen with reduction of body weight. A high incidence of multiple adenoma and liver adenomatosis was found, which is strongly related to an increased BMI. During the follow up of the unique series from a single center multiple liver adenomas did not show any signs of excessive growth, bleeding or malignant degeneration. These data suggest that a conservative management of patients with multiple liver adenomas may be warranted and must be balanced against the danger of high-risk liver surgery in patients with overweight. Reduction of body weight should be advised in patients with more than one adenoma.

Figure 1, BMI and number of adenoma's.

This figure shows BMI of patients with a liver adenomatosis (LA), multiple liver adenomas (MA) and a single hepatocellular adenoma (HCA).

Thumbnail image of

Disclosures:

Robert A. de Man - Advisory Committees or Review Panels: crucell; Grant/Research Support: biotest, gilead

The following people have nothing to disclose: Mirelle Bröker, Bettina E. Hansen, Pavel Taimr, Susanna van Aalten, Jan IJzermans

2118

  1. Top of page

Significantly Higher Long-Term Survival in Hepatocellular Carcinoma (HCC) Patients in the Post-MELD Era: A Population-Based Cohort Study

Robert Wong1, Pardha Devaki2, Long H. Nguyen3, Ramsey Cheung4,1, Cheryl Cho5, Mindie H. Nguyen1;
1Division of Gastroen-terology and Hepatology, Stanford University Medical Center, Palo Alto, CA; 2Department of Internal Medicine, Detroit Medical Center/Wayne State University, Detroit, MI; 3Department of Medicine, Stanford University School of Medicine, Stanford, CA; 4Division of Gastroenterology and Hepatology, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA; 5Division of Medical Oncology, Stanford University Medical Center, Stanford, CA

Background: Recent U.S. population-based studies report poor HCC 5-year survival of 16%. Beginning in 2002, the allocation of MELD exception points for HCC patients within Milan criteria improved liver transplantation waitlist priority for this cohort. However, the impact of these changes on HCC survival in the post-MELD exception era is unclear. The current study evaluates long-term survival of HCC patients in pre-MELD and post-MELD eras. Methods: HCC patients were identified from the Surveillance, Epidemiology, and End Results 1998–2010 registry. Overall survival, stratified by pre-MELD (1998–2003) and post-MELD (2004–2010) eras, was analyzed using Kaplan Meier methods and log-rank testing. Analysis of post-MELD era used a 2003 cut-off to account for lag-time in effect of MELD implementation. Results: Overall HCC 5-year survival was significantly higher in the post-MELD era than pre-MELD era (21.9% vs. 13.0%, p<0.001), especially among those within Milan criteria (Figure). Higher 5-year survival in the post-MELD era was seen among different treatment groups (no therapy: 15.2% vs. 10.2%, p<0.001; local tumor destruction:37.6% vs. 22.1%, p<0.001; resection:55.5% vs. 39.2%, p<0.001) with the exception of transplantation (77.2% vs. 73.1%, p=0.12). Mul-tivariate Cox proportional hazards models, inclusive of sex, age, ethnicity, Milan criteria, tumor stage, tumor size and number, treatment, and MELD era, demonstrated improved survival in the post-MELD era (HR 0.87, 95% CI 0.83–0.91, p<0.001 Males had poorer survival (HR 1.13, 95% CI 1.07–1.IE p<0.001).Asians (HR 0.81, 95% CI 0.77–0.86, p<0.001) an Hispanics (HR 0.89, 95% CI 0.84–0.95, p<0.001) had bette survival, whereas blacks had a trend towards poorer survival (HR 1.05, 95% CI 0.98–1.13, p=0.16). Conclusion: HCC patients in post-MELD era had significantly better survival tha pre-MELD era overall and across genders and ethnicities, though blacks continued to have the lowest survival rates.

Thumbnail image of

Disclosures:

Mindie H. Nguyen - Consulting: Gilead Sciences, Inc., Bristol-Myers Squibb, Bayer AG; Grant/Research Support: Gilead Sciences, Inc., Bristol-Myers Squibb, Novartis Pharmaceuticals, Roche Pharma AG

The following people have nothing to disclose: Robert Wong, Pardha Devaki, Long H. Nguyen, Ramsey Cheung, Cheryl Cho

2119

  1. Top of page

Multidisciplinary management of hepatocellular carcinoma improves utilization of therapy and patient survival

Parul D. Agarwal, Paulina K. Czapiga, Luke C. Hillman, Adnan Said;
Medicine, Division of Gastroenterology and Hepatology, University of Wisconsin School of Medicine and Public Health, Madison, WI

BACKGROUND: Hepatocellular carcinoma (HCC) is a major health problem and a leading cause of cancer mortality worldwide. Given the complexity of patients with HCC and the plethora of potential treatment options available, it is widely accepted that a multidisciplinary team approach (or 'tumor boards') be utilized to optimize therapy for the individual patient. It is estimated that less than half of physicians adopt this approach, and utilization of potential therapies for HCC is sub-optimal. There is a paucity of evidence on therapy utilization and patient outcomes following this multidisciplinary approach for HCC management. The aim of this study was to determine utilization of potential therapies and outcomes for patients with HCC, comparing those who were managed through our institutional multidisciplinary tumor board (MTB) to those who were not. METHODS: Only patients with a diagnosis of HCC were included in this study. A database analysis of all patients with HCC managed through our MTB, since its inception in 2007, was performed. A database of all patients with HCC in the MELD era, not managed through the MTB at our institution, was similarly created and reviewed. Our institutional tumor board comprises of hepatologists, medical oncologists, transplant surgeons, pathologists and interventional radiologists. The gamut of potential therapies offered includes surgical resection, liver transplantation, thermal ablation, chemo- and/or radioembolization, systemic chemotherapy and stereotactic radiation. RESULTS: 304 patients with HCC, from 2007 to 2011 were managed through our MTB, in comparison to 351 patients, from 2002 to 2011 who were not. There were no significant differences in baseline demographic data (age, sex, or MELD at presentation). Patients managed through MTB were more likely to present at an earlier tumor stage (T2, unilobar disease, less often multifocal disease or extrahepatic spread), and with lower serum AFP (P=0.007). The odds of receiving any treatment for HCC was higher in patients managed through MTB (OR 2.80, 95% CI 1.71–4.59, P<0.0001) independent of MELD score, serum AFP, total bilirubin, and tumor stage. There was significantly greater survival of patients managed through MTB (19.1 ± 2.5 mos. vs. 7.6± 0.9 mos., P<0.0001). In a mul-tivariable survival analysis, independent predictors for better survival included management through MTB, lower serum AFP, receipt of liver transplant and T2 tumor stage (compared to > T2). CONCLUSION: Patients with HCC managed through an institutional multidisciplinary tumor board had significantly higher rates of receipt of therapy and improved survival compared to those who were not.

Disclosures:

The following people have nothing to disclose: Parul D. Agarwal, Paulina K. Czapiga, Luke C. Hillman, Adnan Said

2120

  1. Top of page

Presence of Portal Vein Encasement Predicts Efficacy of Radiation Therapy in the Liver Transplantation for Peri-hilar Cholangiocarcinoma protocol

Mamatha Bhat, Matthew Hathcock, Walter K. Kremers, Sarwa Dar-wish Murad, James Martenson, Steven Alberts, Charles B. Rosen, Gregory J. Gores, Julie Heimbach;
Mayo Clinic, Rochester, MN

Survival and recurrence of cancer after liver transplant for per-ihilar cholangiocarcinoma (CCA) following neoadjuvant chemoradiotherapy is strongly correlated with the presence of residual CCA in the liver explant. Aim: To determine the factors predicting response to neoadjuvant therapy in patients who undergo liver transplantation following completion of chemoradiotherapy for perihilar CCA, using presence of residual CCA in the explant as a surrogate marker. Methods: Only those patients selected for the CCA protocol with hilar mass or ERCP brushings/biopsy positive for CCA were included. Demographic and clinical characteristics were abstracted from the CCA liver transplant database at the Mayo Clinic, Rochester, Minnesota, with particular attention to parameters thought to potentially influence radiation therapy efficacy (presence of hilar mass, hemoglobin, body mass index (BMI), presence of cirrhosis, CA-19–9 level prior to and after external beam radiation therapy, brachytherapy dosage, and duration of chemotherapy administration). Factors predictive of the presence of macroscopic and microscopic CCA on explant pathology were determined by ANOVA statistical analysis. Results: Our patient population was 75% male and 27% female, with mean age of 51.6 +/- 10.6 years, and mean BMI of 25.6 +/-4.3 kg/m2. 63% had primary sclerosing cholangitis (PSC), 70% had cirrhosis, with a MELD score of 11.2 +/- 5.3. 53% had a hilar mass, 81% had either cytology or biopsy positive for adenocarcinoma, and 49% had polysomy on fluorescence in situ hybridization (FISH). The median pre-treatment CA-19–9 was 67.5 (IQR: 16, 233.5) IU/mL and median post-treatment CA-19–9 was 64 (24, 163) IU/mL. In univariate logistic regression, log CA-19–9 pre-and post-treatment, low bilirubin, presence of mass, portal vein encasement and hepatic artery encasement were all associated with higher risk of macroscopic disease on explant. In the multivariable logistic model, it was found that after accounting for age and gender, the presence of portal vein encasement (Odds ratio: 11.8 ; 95% CI: 2.43 -57.21; p=0.002) as well as MELD score at presentation (Odds ratio: 1.13; 95% CI: 1.02 - 1.26; p=0.017) were predictive of the presence of residual macroscopic disease on explant (c-sta-tistics 0.78). Conclusion: Portal vein encasement was strongly predictive of decreased response to radiation therapy, as reflected by residual macroscopic cholangiocarcinoma in the liver explant. Taking the presence of portal vein encasement into account may optimize patient selection for liver transplantation.

Disclosures:

Gregory J. Gores - Advisory Committees or Review Panels: Bayer, Chugia, Dai-ichi, Generon, Conatus, IntegraGen

The following people have nothing to disclose: Mamatha Bhat, Matthew Hath-cock, Walter K. Kremers, Sarwa Darwish Murad, James Martenson, Steven Alberts, Charles B. Rosen, Julie Heimbach

2121

  1. Top of page

PNPLA3 in end-stage liver disease: alcohol consumption, HCC development and liver transplantation

Kilian Friedrich1, Andreas Wannhoff1, Christian Rupp1, Karl Heinz Weiss1, Peter Schemmer2, Wolfgang Stremmel1, Daniel Gotthardt1;
1 Department of Internal Medicine IV, University Hospital of Heidelberg, Heidelberg, Germany; 2Department of General Surgery, University Hospital of Heidelberg, Heidelberg, Germany

Background & Aims: End-stage liver disease represents a final path for hepatic disorders and can only be cured by liver transplantation. The rs738409 variant (I148M) of the PNPLA3 gene is associated with a variety of liver malfunctions but has not been addressed in an end-stage liver disease setting. Methods: The I148M polymorphism was genotyped in a well characterized cohort of 421 Caucasian patients with end-stage liver disease, characterized as liver transplantation or current enrollment at Eurotransplant. We obtained a detailed history of alcohol consumption in patients with alcoholic liver disease (ALD) Results: The G allele of the I148M variant was significantly overrepresented in patients with ALD (p < 0.001), chronic viral hepatitis (p=0.004) and cryptogenic liver cirrhosis (p=0.002) compared to healthy controls. In alcoholic liver disease patients, the G allele was closely associated with hepato-cellular carcinoma development (odds-ratio = 2.399; 95% CI: 1.292–4.455; p =.008). Transplantation-free survival was significantly decreased when carrying either one or two mutated G allele (CC= 36.8±10.2 months, 95% CI: 16.7–56.9; GC= 19.6±4.3 months, 95% CI: 11.0–28.2; GG= 15.2±8.0 months, 95% CI: 0.0–31.1; p=.041) from time of enrolment at Eurotransplant. Analysis of alcohol consumption revealed that wildtype patients consumed significantly more alcohol than heterozygous or homozygous I148M patients (CC = 731.0 ± 216.5; CG = 601.4 ± 172.2; GG = 523.3 ±190.5 [g EtOH//week]; p = 0.001). Although alcohol intake of I148M homozygous patients (GG) was boarderline to the cut-off value of 60g alcohol/day, they still shared the same features of end-stage liver disease than wildtype patients. Conclusions: In a cohort of end-stage liver disease patients we identified ALD to be predominantly affected by the I148M polymorphism causing a significantly increased risk of HCC development with reduced transplantation-free survival. According to public perception ALD is a stigmatized hepatic disorder and one of the reasons for liver donation aversion. However, we were able to show that the I148M polymorphism highly impairs hepatic malfunction in ALD patients despite reduced alcohol consumption compared to wildtype patients.

Disclosures:

The following people have nothing to disclose: Kilian Friedrich, Andreas Wannhoff, Christian Rupp, Karl Heinz Weiss, Peter Schemmer, Wolfgang Stremmel, Daniel Gotthardt

2122

  1. Top of page

Macrophage colony-stimulating factor induces angiogenic factors and play a pivotal role in hepatocarcinogenesis

Hiroshi Kono, Shinji Furuya, Michio Hara, Kazuyoshi Hirayama, Hideki Fujii;
First Department of Surgry, University of Yamanashi, Chuo, Japan

Aim; The specific purpose of this study was to investigate the role of macrophage colony-stimulating factor (M-CSF) in angio-genesis in hepatocellular carcinoma (HCC). Materials and Method; In clinical analysis, the expression of M-CSF (anti-M-CSF antibodies), population of Mφs (anti-CD68 antibodies), and angiogenesis (anti-CD31 antibodies) were assessed by immunohistochemistry in liver tissues containing paired tumor and peritumoral liver tissue from 55 HCV-infected patients who had undergone hepatectomy for histologically proven HCC in University of Yamanashi Hospital. Prognostic values of these and other clinicopathologic factors were evaluated. To test the hypothesis that M-CSF may activate the hepatic macrophages and induce the angiogenesis, further experiments were performed in vitro. The hepatic macrophages (Mφs) or the mono-cytes were isolated from C57BL/6 mice and they were cultured with media added the different dose of M-CSF. Production of vascular endothelial growth factor (VEGF) was assessed by ELISA. In another set of experiments, vascular endothelial cells (VEC) isolated from the aorta were co-cultured with or without Kupffer cells in presense with M-CSF and their cell proliferations were assessed. Result; In clinical studies, neither intratumoral M-CSF nor density of hepatic Mφs was associated with disease-free survival (DFS). On the other hand, the expression of peritumoral M-CSF, Mφ density, and CD31 which correlated with tumor size, presence of intrahepatic metastasis, and high TNM stage, were independent prognostic factors for DFS. These factors affected incidence of early recurrence after hepatic resection. Furthermore, the combination of peritumoral M-CSF and hepatic Mφ had a better power to predict the disease recurrence. Production of VEGF increased both in the hepatic Mφs and the monocytes incubated with M-CSF in dose dependent manner. Furthermore, this production was significantly greater in the hepatic Mφs compared with that in the monocytes. Importantly, under isolated VEC co-cultured with hepatic Mφs in present or absent with M-CSF, the proliferation of VEC was greatest in VECs cultured with the KC and M-CSF. Conclusions; Thus, M-CSF is involved in hepatocarcinogenesis by inducing an angiogenic factor via the hepatic Mφs. Therefore, M-CSF could be targets of postoperative adjuvant therapy in HCC.

Disclosures:

The following people have nothing to disclose: Hiroshi Kono, Shinji Furuya, Michio Hara, Kazuyoshi Hirayama, Hideki Fujii

2123

  1. Top of page

Inducible Nitric Oxide Synthase expression in Hepato-cellular Carcinoma as a marker for Carcinogenesis and Tumor Progression

Robert McGuffey2, Bryan Rea1, Miguel H. Malespin2, Roshan S. Patel3, Rohini Chennuri1, Grace Guzman1;
1 Pathology, University of Illinois at Chicago Medical Center, Chicago, IL; 2Medicine, University of Illinois at Chicago Medical Center, Chicago, IL; 3Bio-chemistry and Molecular Genetics, University of Illinois at Chicago Medical Center, Chicago, IL

Chronic liver disease and the interval risk of hepatocellular carcinoma remains a major health concern in the United States and worldwide. Inducible nitric oxide synthase (iNOS) expression has been linked to chronic liver disease, and animal models have shown correlation with iNOS expression and liver cancer. In this study, we postulate that iNOS expression can be correlated to HCC and areas of dysplasia as a marker of carcinogenesis. In addition, we theorize that iNOS expression may also be related to etiology of liver disease, tumor type and cancer staging. An array of liver tissue, derived from 158 subjects with cirrhosis (48 with HCC and 100 without HCC) and 8 normal liver samples were analyzed. Standard immunohisto-chemistry was employed and quantified by Aperio Image Analysis to determine levels of iNOS expression in normal, cirrhosis, dysplasia, and HCC liver samples. Nonparametric statistical tests were employed using SPS statistics 20. iNOS expression was found to be increased in patients with HCC when compared to areas of cirrhosis or dysplasia (p<0.038). Additionally, iNOS expression was increased across areas of poorly differentiated tissue when compared to areas of well-differentiated HCC (p<0.03). It was also increased in pleomor-phic tumor types (p<0.05). There was no significant difference in iNOS expression across areas of dysplasia or cirrhosis. When comparing across etiology of liver disease (Viral/EtOH/NASH/Cryptogenic), there were no significant differences in iNOS expression in HCC samples(p<0.684), but a significant difference in iNOS expression was found across areas of dysplasia (p<0.021). Finally, there were no correlations of iNOS expression when comparing types of HCC as well as between histological types(p<0.544) or cancer staging (p<0.272). Conclusion: iNOS was significantly increased in HCC when compared to dysplasia or cirrhosis. iNOS expression was also elevated in poorly differentiated HCC as well as pleomorphic variant type as compared to well differentiated HCC. However, no correlation was found between staging, etiology or histological types of HCC. Our findings show that iNOS expression is increased in areas of poor differentiation and pleomorphic type indicating potential prognostic and clinical significance. However, our results cannot distinguish iNOS expression as an effect of carcinogenesis versus a true causal relation.

Disclosures:

The following people have nothing to disclose: Robert McGuffey, Bryan Rea, Miguel H. Malespin, Roshan S. Patel, Rohini Chennuri, Grace Guzman

2124

  1. Top of page

Liver cancer in Texas: recent trends and risk factors for primary liver cancer. A comparison between the state of Texas and the United States

Daria Zorzi1, Marco Sequi2, Cristiana Rastellini1, Luca Cicalese1;
1Surgery, University of Texas Medical Branch, Galveston, TX; 2Public Health, Mario Negri" Pharmacological Research Institute, Milan, Italy

BACKGROUND: Cancer is the leading cause of death in Texas (TX), with primary liver cancer (LK), ranked number 5 in males and number 8 in females. LK, in the majority of cases hepatocellular carcinoma, has increased dramatically over the last two decades and TX is the second continental state with the highest incidence of LK in the U.S. The reason why incidence of LK in TX is so high compared to the national population remains unclear. This study aims to evaluate the distribution of known risk factors associated with LK and to correlate their geographical distribution in TX. METHODS: Data were provided by the Texas Cancer Registry, National Program of Cancer Registries, Centers for Disease Control and Prevention, and National Health and Nutrition Evaluation Survey. RESULTS: Incidence rate for liver cancer in TX is rising and above US rate, with annual percentage change (APC) of 5.7, significantly different from zero and significantly higher than the national APC of 4. Age-adjusted annual incidence rate per 100,000 is equal to 9.1 in TX (rate period 2005–2009), reaching up to 25 in Walker County, compared to US rate of 6.6. Death rate for liver cancer in TX is reported as rising (APC of 1.8 significant) and above US rate, especially for the city counties, reaching up to 28 (Anderson County). Death rate for liver cancer in TX is higher than other parts of the country with a Age-adjusted annual death rate of 7.2/100,000, compared to the national average rate of 5.5 (rate period 2005–2009). The overall prevalence of HCV infection in TX matches that of general US population (1.8%), with county prevalence varied from 1.04% to 2.18%. Both TX and US have a consistent markedly higher incidence of LK in Hispanic (15.5) followed by African American (13) and Caucasian (8.7). On the contrary both in TX and US the prevalence of HCV is higher in African American (2.73%) followed by Caucasian (1.44%) and Hispanic (1.25%). Comparison and analysis of the geographic distribution of HCV prevalence and LK incidence in TX counties showed no statistically significant correlation (p=0.678). When analyzing other known risk factors for LK, no statistically significant correlation was observed between LK incidence and Hepatitis B infection (p=0.3431), diabetes (p=0.2518), and obesity (p=0.4089) in the TX counties. CONCLUSIONS: Our study showed that the TX population has a significantly higher, and rising, incidence of LK than the general US population. The distribution of HCV infection and other known risk factors does not correlate with this trend. These findings likely reflect the high concentration of other LK-related risk factors in TX that warrant further investigation.

Disclosures:

The following people have nothing to disclose: Daria Zorzi, Marco Sequi, Cristiana Rastellini, Luca Cicalese

2125

  1. Top of page

Impact of Screening on Staging and Survival of Hepatocellular Carcinoma (HCC) in HIV/HCV-Coinfected Patients

Rena K. Fox1, Sonja Marcus2, Nicolás Merchante3, Luciana Kikuchi4, Beatriz Minguez5, Eugenia Vispo6, Marianne Harris7, Maaz B. Badshah2, Mamta K. Jain8, Emma E. Page9, David E. Kaplan10, Mark Nelson9, Pablo Barrero6, Juan A. Pineda3, Nor-bert Brau2''; 1 University of California San Francicso, San Francisco, CA; 2Bronx VA Medical Center, Bronx, NY; 3Hospital Universitario de Valme, Sevilla, Spain; 4Universidade de São Paulo, São Paulo, Brazil; 5Hospital Universitario Vall d'Hebron, Barcelona, Spain; 6Hospital Carlos III, Madrid, Spain; 7St. Paul's Hospital & University of British Columbia, Vancouver, BC, Canada; 8University of Texas Southwestern Medical Center, Dallas, TX; 9Chelsea and Westminster Hospital & Imperial College, London, United Kingdom; 10Philadelphia VA Medical Center & University of Pennsylvania, Philadelphia, PA;
1 1 Mount Sinai School of Medicine, New York, NY

Background: Current recommendations for HCC screening in HCV-infected patients with cirrhosis (liver sonography every 6 months) are based on expert opinions. No data are available for the effectiveness of HCC screening in HIV/HCV-coinfected patients. Methods: HIV/HCV-coinfected patients with HCC (AASLD criteria) were retrospectively identified from 1995–2013 in 39 centers in North and South America, Europe, and Australia. Patients were considered screened if they initially presented with an abnormal AFP level or imaging study and not screened if they presented with symptoms. Results: Among 198 HIV/HCV-coinfected patients with HCC, 117 (59%) were screened and 81 (41+%) were unscreened. Screened patients abused alcohol less often (30% vs. 51%, p=0.003), and had a lower mean Child-Turcotte-Pugh score (6.6 vs. 7.8, p<0.001). They more often had HIV RNA < 400 copies/mL (79% vs. 54%, p<0.001) and had a higher median CD4+ cell count (344 vs. 274/mm3, p=0.027). Screened patients also had a smaller median tumor size (3.0 vs. 5.15 cm, p< 0.001), more commonly met Milan criteria for liver transplantation (64% vs. 29%, p< 0.001), and less often had multiple tumors (45% vs. 61%, p=0.035), portal vein thrombosis (12% vs. 31%, p=0.001) or extrahepatic metastases (9% vs. 29%).They also presented less frequently with advanced Barcelona-Clinic-Liver-Cancer (BCLC) stages C+D (39% vs. 73%, p< 0.001) and more often received effective HCC therapy (76% vs. 33%, p< 0.001). With adjustment for lead time of 8.6 months, screened patients had a longer median survival (19.1 vs. 3.5 months, p< 0.001). In multi-variable Cox analysis, screening was independently associated with survival (H.R., 2.66; 95% C.I., 1.7 -4.1; p<0.001) together with HCC therapy and BCLC staging. In recent years, the rate of patients diagnosed through screening increased, it was 52% in 1995–2004, 49% in 2005–2008, and 77% in 2009–2013. Conclusion: Many HIV/HCV-coinfected patients with HCC were not diagnosed through screening. Yet, screening was associated with earlier HCC stages, more HCC therapy, and independently predicted survival. In recent years, more patients with HCC were diagnosed through screening

Disclosures:

Mamta K. Jain - Advisory Committees or Review Panels: Merck, Vertex, Boehringer Ingelheim ; Grant/Research Support: Vertex, Bristol-Myers Squibb, Pfizer, Janssen, Gilead, Boehrnger Ingelheim , Viiv Healthcare, EMD Serono

David E. Kaplan - Grant/Research Support: Merck, Bayer

Mark Nelson - Advisory Committees or Review Panels: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead; Consulting: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead; Grant/Research Support: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead, Roche; Speaking and Teaching: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead

Norbert Brau - Advisory Committees or Review Panels: Janssen; Grant/Research Support: BMS, Gilead, Vertex; Speaking and Teaching: Vertex, Onyx

The following people have nothing to disclose: Rena K. Fox, Sonja Marcus, Nicolás Merchante, Luciana Kikuchi, Beatriz Minguez, Eugenia Vispo, Marianne Harris, Maaz B. Badshah, Emma E. Page, Pablo Barrero, Juan A. Pineda

2126

  1. Top of page

Characteristics of Hepatocellular Carcinoma (HCC) Patients Who Received Transarterial Chemoemboliza-tion (TACE) Prior to Sorafenib: US GIDEON Experience

Jean-Francois H. Geschwind1, Allen Cohn2, Anthony El-Khoueiry3, Pierre M. Gholam4, Alec Goldenberg5, ParvezS. Mantry6, Robert Martin7, Brendan M. McGuire8, Rebecca A. Miksad9, Bilal Piperdi10, Arun J. Sanyal11, Alan Venook12, Joseph F. Germino13, Jorge A. Marrero14;
1Interventional Radiology, Johns Hopkins University, Baltimore, MD; 2Rocky Mountain Cancer Center - US Oncology, Denver, CO; 3USC Norris Comprehensive Cancer Center, Los Angeles, CA; 4Liver Center of Excellence, University Hospitals Case Medical Center, Cleveland, OH; 5New York University School of Medicine, New York, NY; 6The Liver Institute at Methodist Dallas Medical Center, Dallas, TX; 7University of Louisville, Louisville, KY; 8University of Alabama at Birmingham, Birmingham, AL; 9Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA; 10Montefiore Medical Center, New York, NY; 11Virginia Commonwealth University Medical Center, Richmond, VA; 12University of California, San Francisco, San Francisco, CA; 13Bayer HealthCare, Wayne, NJ; 14University of Texas Southwestern Medical Center, Dallas, TX

GIDEON (Global Investigation of therapeutic DEcisions in hepatocellular carcinoma and Of its Treatment with sorafeNib) is a worldwide, prospective, non-interventional study to evaluate sorafenib (SOR) safety in a variety of patient subsets under real-life practice conditions. Patients with unresectable HCC who were candidates for systemic therapy and for whom a decision was made to treat with SOR were eligible for inclusion. Owing to the observational nature of this study, all results are descriptive. In the US, 645 patients were enrolled; 553 were valid for efficacy and 563 for safety. Results are reported for the safety population except for overall survival (OS), which refers to the enrolled population. 209 patients received prior TACE (prT); 354 did not (NprT). Demographics, HCC etiology, and performance status were similar in patients who received prT compared to NprT. Baseline tumor characteristics and liver function differed between the prT and NprT groups (see table). Initial SOR dose selection did not vary by number of TACE; 800 mg/d was initiated in 56%, 52%, 47%, and 53% of patients with 0, 1,2, and >3 prior TACE, respectively, and 400 mg/d in 35%, 36%, 36%, and 27%. However, differences were noted in median daily dose (484 mg prT; 579 mg NprT) and duration of treatment (18 weeks prT; 11 weeks NprT). Incidences of the most common drug-related adverse events (DRAEs) were similar between prT and NprT patients except for hand-foot skin reaction (grade 3: 10% prT; 4% NprT). Drug-related serious adverse events (DRSAEs) were also similar in frequency (see table). Adverse events (AEs) leading to permanent discontinuation of SOR occurred in 37% of prT and 36% of NprT patients. Median overall survival was 317 days for prT and 212 days for NprT patients. Conclusions: DRAEs were comparable in nature and frequency to those seen in the SHARP trial, with no new safety signals observed in a real-life practice setting. Dissimilarities in liver function, tumor characteristics, and duration of SOR treatment may have played a role in the OS differences reported. Given the observational nature of this study, selection bias cannot be eliminated.

%prT n = 209NprT 0 = 354
  1. *Barcelona Clinic Liver Cancer

  2. ‡Unknown

  3. **Not evaluable

  4. ‡fPortal vein thrombosis

BCLC* A, B, CD, UK†27, 14, 15,4,3311, 10,33,7,33
Child-Pugh A, B, C, NE**38,21,5,3734,31,9,26
PVT‡1029
AEs9998
SAEs4362
DRAEs7869
DRSAEs78

Disclosures:

Jean-Francois H. Geschwind - Consulting: Biocompatibles/BTG, Bayer Health-Care, Guerbet, Nordion, Delcath, Prescience Labs, Jennerex; Grant/Research Support: Biocompatibles/BTG, Bayer HealthCare, Philips Medical, Nordion, Guerbet, DOD, NCI-ECOG, NIH-RO1; Stock Shareholder: PreScience Labs -Founder/CEO

Allen Cohn - Advisory Committees or Review Panels: genomic health, sanofi; Speaking and Teaching: bayer, pfsizer

Anthony El-Khoueiry - Advisory Committees or Review Panels: Bayer/Onyx; Grant/Research Support: Bayer/Onyx

Pierre M. Gholam -Advisory Committees or Review Panels: Onyx, Gilead, Vertex; Consulting: Onyx; Grant/Research Support: Onyx, Gilead; Speaking and Teaching: Onyx, Salix, Vertex, Merck, Otsuka

Parvez S. Mantry- Consulting: Gilead, Janssen; Grant/Research Support: Vertex, Merck, Salix, Abbvie, Gilead, Boehringer-Ingelheim, Bristiol Myers Squibb, San-taris; Speaking and Teaching: Vertex, Merck, Sailx, Genentech, Bayer-Onyx, Kadmon

Brendan M. McGuire - Consulting: hepahope; Grant/Research Support: roche (genentech) laboratories, bayer healthcare, vital therapies, ikaria therapeutics, cumberland pharmaceuticals, vertex pharmaceuticals

Rebecca A. Miksad - Consulting: OptumInsight, Advance Medical; Grant/Research Support: Bayer/Onyx, NewLink, Sanofi

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

Alan Venook - Grant/Research Support: Bayer Pharm, Onyx Joseph F. Germino - Employment: Bayer Healthcare

Jorge A. Marrero - Advisory Committees or Review Panels: Bayer, Onyx; Grant/Research Support: Bayer, BMS

The following people have nothing to disclose: Alec Goldenberg, Robert Martin, Bilal Piperdi

2127

  1. Top of page

Tumor imaging enhancement by measuring attenuation in intrahepatic cholangiocarcinoma

Atsushi Nanashima, Takafumi Abo, Junichi Arai, Takeshi Nagayasu;
Department of Surgical Oncology, Nagasaki University, Nagasaki, Japan

Purpose: Arterial enhancement of intrahepatic cholangiocarcinoma (ICC) has been noted. To precisely identify the characteristics of tumor enhancement patterns, we examined the relationship between CT attenuation in the tumor and clinico-pathological parameters or prognosis. Methods: Subjects were 42 ICC patients who had undergone hepatectomy. Microvessel density (MVD) determined by CD34 staining was compared with imaging. Attenuation was calculated in images from multi-detector CT of tumor and non-tumorous regions. Enhancement patterns were divided into two groups: arterial enhancement with higher attenuation (>16 HU; Hyper group, n=12); and arterial enhancement with lower attenuation (Hypo group, n=30). Results: Univariate analysis identified high tumor marker level, increased size, less-differentiation, incomplete resection, increased bleeding and lower MVD as significantly associated with poor survival (p<0.05). Increased attenuation throughout the whole ICC correlated significantly with radiological findings and MVD. Concomitant hepatitis, well-differentiation, and smaller tumor were more significantly frequent in the Hyper group than in the Hypo group (p<0.05). Postoperative early recurrence was significantly less frequent in the Hyper group, and overall survival was significantly better in the Hyper group (p<0.05). Conclusions: Increased CT attenuation correlated with ICC tumor vascularity. Increased tumor enhancement in the arterial phase was associated with chronic hepatitis, lower malignancy and better survival.

Disclosures:

The following people have nothing to disclose: Atsushi Nanashima, Takafumi Abo, Junichi Arai, Takeshi Nagayasu

2128

  1. Top of page

Vitamin K dosing could improve anticancer outcome of sorafenib treatment for hepatocellular carcinoma

Yoshimichi Haruna, Atsuo Inoue;
Department of Gastroenterology and Hepatology, Osaka General Medical Center, Osaka, Japan

Backgrounds and Aims. Sorafenib, a multikinase inhibitor, is the only oral anticancer agent for advanced hepatocellular carcinoma (HCC). However, the SHARP study showed that the disease control rate was 53% and that the median overall survival period was prolonged 2.3 months, which were not necessarily satisfying. On the other hand, it was reported that combination of vitamin K and sorafenib had a synergistic inhibitory effect on Raf kinase and that antitumor action of sorafenib was enhanced by vitamin K in vitro and in vivo. In this study, we retrospectively investigated the influence of vitamin K dosing in sorafenib treatment for HCC. Patients and Methods. We examined twenty-one patients with the Barcelona Clinic Liver Cancer (BCLC) stage B (intermediate), who were contraindicated for surgical resection, radio frequency ablation, percutaneous ethanol infusion treatment or transarterial chemoembolization (male/female 12/9, median age 77 yrs (range 58–84 yrs)). Nine patients were orally given vitamin K2 (45mg daily). We compared the time to radiologic progression and overall survival in the vitamin K dosing and not dosing groups. The radiologic assessment was performed according to modified RECIST. Results. In the group of sorafenib + vitamin K, the anticancer outcome was CR: 0 case, PR: 5 cases, stable disease (SD): 4 cases, and progression disease (PD): 0 case. In the sorafenib monotherapy group, that was CR: 1 case, PR: 1 case, SD: 4 cases, and PD: 6 cases. The disease control rate of the vitamin K dosing group was higher than that of the other one (100% vs. 50%, respectively, p=0.019). In sorafenib + vitamin K group, progression-free survival rate was 87.5% (6 mo.), 18.8% (12 mo.) and 18.8% (24 mo.), whereas that was 20% (6 mo.), 10% (12 mo.), 10% (24 mo.) in sorafenib monotherapy group (Log-rank test, p=0.020). The overall survival rate was 100% (6 mo.), 87.5% (12 mo.) and 65.6% (24 mo.) in sorafenib + vitamin K group while that was 72.7% (6 mo.), 72.7% (12 mo.) and 20.2% (24 mo.) in sorafenib monotherapy group (Log-rank test, p=0.070). The median overall survival period was 31 months in sorafenib + vitamin K group and 16 months in sorafenib monotherapy group. Fifteen months extension of the median survival time was obtained by the vitamin K dosing. No add-on adverse event appeared by the vitamin K dosing. Conclusion. It was shown that the nontoxic agent, vitamin K, dosing could markedly improve the anticancer outcome of sorafenib treatment for HCC with BCLC-B.

Disclosures:

The following people have nothing to disclose: Yoshimichi Haruna, Atsuo Inoue

2129

  1. Top of page

Prediction of aggressiveness in early-stage hepatocellular carcinoma for selection of surgical treatment

Suk-won Suh, Kwang-Woong Lee, Jeong-moo Lee, Tae Yoo, Youn-gRok Choi, Nam-Joon Yi, Kyung-Suk Suh;
Surgery, Seoul National University College of Medicine, Seoul, Republic of Korea

Background: In early-stage hepatocellular carcinoma(eHCC), radiofrequency ablation(RFA) has been reported to show comparable outcomes with surgery. However, microsatellites can be present in eHCC, and it results in tumor recurrence after local ablation. Biologic components of tumor as poor tumor grade and microvascular invasion were known as the risk factors for microsatellite, but it was not considered in current selection criteria for RFA. Therefore we investigated the predictors of aggressiveness in eHCC to select high risk patients of recurrence after RFA who needs surgical resection. Methods: Between July 2007 and December 2011, a total of 370 patients with newly diagnosed eHCC(3 or fewer in number and less than 3cm in size) who had no history of previous treatment were evaluated. Patients were divided into 2 groups by the type of procedure, as surgical resection group, 126 patients and RFA group, 244 patients. At first step, in surgical resection group, preoperative risk factors for tumor aggressiveness in eHCC, which represents for the risk of microsatellites, poor tumor grade or microvascular invasion were investigated. At second step, in RFA group, identified factors were validated to reveal its influence on tumor recurrence. Results: At, first step, in surgical resection group, tumor size, PIVKA-II > 40 nAU/mL, and multiplied AFP and PIVKA-II (A*P) > 1,600 were significant in univariate analysis and the only significant factor in multi-variate analysis was A*P > 1,600 (HR 3.837. 95% CI 1.493 - 9.861, p = 0.005). At second step, in RFA group, identified risk factor, A*P > 1,600 was validated and the significant risk factors for tumor recurrence in RFA group were albumin (HR 0.456, 95% CI 0.313 - 0.666, p = 0.000) and A*P > 1,600 (HR 1.835, 95% CI 1.222 - 2.755, p = 0.003).Recurrence free survivals were analyzed after adjusting several factors and it shows significant difference in RFA group by the level of combined AFP and PIVKA (HR = 1.835, 95% CI = 1.222–2.755, p = 0.003), but there was no significant difference in surgical resection group.(p = 0.214 ) Among the patients of single tumor in eHCC with A*P > 1,600, surgical resection group shows superior outcomes than RFA group in terms of HCC recurrence within 1 year. (HR = 2.340, 95% CI = 1.049 - 5.219, p = 0.038) But there was no significant difference by type of procedure in patients, A*P < 1,600. (p = 0.872) Conclusions: Combined AFP and PIVKA is useful predictor of aggressiveness in early-stage HCC. It predicts tumor recurrence after RFA, as well. Therefore, if the level of this predictor is high in early-stage HCC surgical resection should be considered for 1st line therapy.

Disclosures:

The following people have nothing to disclose: Suk-won Suh, Kwang-Woong Lee, Jeong-moo Lee, Tae Yoo, YoungRok Choi, Nam-Joon Yi, Kyung-Suk Suh

2130

  1. Top of page

GNAS and KRAS mutations are common in intraductal papillary neoplasm of the bile duct

Motoko Sasaki1, Takashi Matsubara2, Yasunori Sato1, Norihide Yoneda2, Yasuni Nakanuma1;
1Human Pathology, Kanazawa University Graduate School of Medicine, Kanazawa, Japan; 2Radiology, Kanazawa University Graduate School of Medicine, Kanazawa, Japan

Backgrounds/Aims. Intraductal papillary neoplasms of the bile duct (IPNB) shows favorable prognosis and is regarded as a biliary counterpart of intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Although activating point mutations of GNAS at codon 201 have been detected in approximately two thirds of IPMNs of the pancreas, there have been few studies on GNAS mutations in IPNBs. This study investigates the status of GNAS and KRAS mutations and their association with clinico-pathological factors (age, gender, location, subtype, grade, maximum size, presence of invasive carcinoma, lymph node metastasis, mucin production and MUC mucin profiles) in IPNBs. Methods. We examined the status of GNAS mutation at codon 201 and KRAS mutation at codon 12&13, degree of mucin production and immunohistochemical expressions of MUC mucin core proteins (MUC1, MUC2, MUC5AC, MUC5B and MUC6) in 29 patients (M/F=15/14) with IPNB in intra-hepatic and perihilar bile ducts (perihilar IPNB) and 6 patients (M/F=5/1) with IPNB in distal bile ducts (distal IPNB). GNAS and KRAS mutation were analyzed using genomic DNA extracted from IPNBs. Results. GNAS mutations at codon 201 were detected in 15 of 30 IPNBs (50%). KRAS mutations at codon 12&13 were detected in 12 of 26 IPNBs (46.2%). There was no significant association between the status of GNAS and KRAS mutations. There was no significant correlation between the status of GNAS mutation and clinicopathological factors in IPNBs. The status of KRAS mutation was significantly inversely correlated with the degree of MUC2 expression in IPNB (p< 0.05). All IPNBs showed mucin production to various degrees. Twenty and 15 patients were divided into IPNBs with high- and low- mucin production, respectively. Degree of mucin production was significantly high in patients with perihilar IPNB, compared to patients with distal IPNBs (p<0.05). MUC2 and MUC5AC expression was siginificantly high in IPNBs with high-mucin production, compared to those with low-mucin production (p<0.01 and p<0.05, respectively). IPNB with high-mucin production is characterized by perihilar location and high expression of MUC2 and MUC5AC. Invasion tended to be associated with low-mucin production, although there was no significance. The expression of MUC1 was significantly correlated with invasion, but it was not significantly associated with degree of mucin production. Conclusions. This study firstly highlights an involvement of GNAS mutations in IPNBs, similarly to IPMNs, and this may suggest a common pathogenesis of IPNBs and IPMNs.

Disclosures:

The following people have nothing to disclose: Motoko Sasaki, Takashi Matsubara, Yasunori Sato, Norihide Yoneda, Yasuni Nakanuma

2131

  1. Top of page

Reverse transcription polymerase chain reaction is sensitive method for lymph node Micrometastasis in Gallbladder cancer, Cholangiocarcinoma and Pancreas cancer

Pranjal Deka1, Ritu Khosla2, Nirupma Trehanpati2, Archana Ras-togi3, Shiv K. Sarin4, Tushar K. Chattopadhyay1;
1Hepato-Pancre-ato-Biliary, Institute of Liver and Biliary Sciences, New Delhi, India; 2Research, Institute of Liver and Biliary Sciences, New Delhi, India; 3Pathology, Institute of Liver and Biliary Sciences, New Delhi, India; 4Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India

Background: Gallbladder, bile duct and pancreas cancer are a significant clinical problem and despite rapid advances in the diagnosis and treatment of these diseases, their prognosis remains grim. Cancer recurrence occurs in the regional lymph nodes in a substantial number of patients. Routine histopatho-logical examination of surgically resected lymph nodes is unable to diagnose micrometastasis and hence immunohisto-chemical analysis was initiated. Recently, genetic diagnosis of micrometastasis by reverse transcription Polymerase Chain Reaction (RT-PCR) is being considered more sensitive method. Aim: To study the prevalence of micrometastasis in Biliary and Pancreatic Cancer using Real Time Reverse Transcriptase PCR and Immunohistochemistry Patients and Methods: Total 71 lymph nodes (68 from cancers and three from non cancer) were dissected along the hepatoduodenal ligament, common hepatic artery, celiac axis, peripancreatic, suprapyloric and infrapy-loric region from Gallbladder cancer, Cholangiocarcinoma and Pancreas cancer. Histological examination for metastatic tumor cells was made by haematoxylin and eosin staining (H&E) and immunohistochemistry analysis was performed on serial sections of formalin-fixed paraffin embedded tissue sections using CAM 5.2 and K-ras monoclonal antibodies. Total RNA was isolated tissue samples tissue samples stored in RNA latter. 1μg of total RNA was used to prepare cDNA by reverse transcriptase method (Invitrogen, USA). Quantitative RT-PCR was performed in triplicate in a 7900 ABI Prism Sequence Detection system using the Syber Green and specific primers for Kras, MAGE 3, CK20 and CEA. Amplification of 18s RNA was used as the control for normalization. Relative quantification of each gene was analyzed by calculating the Log RQ of Ct value. Results: Out of 68 lymph nodes, 23 lymph nodes were positive for malignancy on histopathological examination. Of all these positive lymph nodes, 17 lymph nodes also showed more than two fold of up-regulated expression of Kras, MAGE3, CK20 or CEA gene. Most of the nodes were positive for dual expression of Kras MAGE3 or CK 20 by RT-PCR. 18 HE negative lymph nodes also showed upregulated expression of either one of these genes. However, these nodes did not reveal CAM 5.2 and Kras staining by IHC. Conclusion: RT-PCR appears to have an important role in the diagnosis of micro metastasis and more sensitive technique for detecting micro-metastasis than IHC. Long term follow up for the recurrence of micro-metastasis in these patients will further strengthen the importance of RT-PCR method in prognostic work-up.

Disclosures:

The following people have nothing to disclose: Pranjal Deka, Ritu Khosla, Nirupma Trehanpati, Archana Rastogi, Shiv K. Sarin, Tushar K. Chattopadhyay

2132

  1. Top of page

Sorafenib versus cytotoxic chemotherapy for patients with recurrent hepatocellular carcinoma after liver transplantation

Bo Hyun Kim1, Sang Myung Woo1, Seong Hoon Kim1, Seung Duk Lee1, Young-Kyu Kim1, Eun Uk Jung2, Woo Jin Lee1, Joong-Won Park1, Eun Kyung Hong1, Chang-Min Kim1;
1Center for Liver Cancer, National Cancer Center, Goyang, Republic of Korea; 2Depart-mentn of Internal Medicine, Pusan Paik Hospital, College of Medicine, Inje University, Busan, Republic of Korea

Background: As hepatocellular carcinoma (HCC) has become a major indication for liver transplantation (LT), post-LT HCC recurrence is also increasing. However, chemotherapy has been scarcely studied. The aim of this study was to evaluate the outcomes and prognostic factors of systemic chemotherapy in patients with post-LT HCC recurrence. Methods: Data of patients given palliative chemotherapy for recurrent HCC after LT between 2004 and 2012 were reviewed. Patients receiving sorafenib as a first-line systemic agent were compared with those receiving cytotoxic chemotherapy in the previous era. Results: HCC has recurred in 62 out of 409 patients who underwent LT. Twenty-eight patients (median age 53, range [37–66]; male 89.3%) received palliative chemotherapy. Histological examination of the explanted liver showed macro- and microvascular invasion in 25.0% and 78.6%, respectively. A total of 82.1% had a pathological stage T3/T4 by the modified UICC system. Median time to recurrence was 4.9 months (95% confidential interval [CI], 3.7–7.9). As a first-line systemic therapy, sorafenib (Arm S, n=18) or cytotoxic chemotherapy (Arm C, n=10) was given. The most commonly used regimen was capecitabine/cisplatin (n=9). Median overall survival (OS) was 5.5 months (95% CI, 3.5–11.1) and 14.6 months (95% CI: 0.7–26.8) in Arm S and Arm C, respectively (P=0.06); the corresponding median time to progression (TTP) was 2.5 months (95% CI, 1.9–3.6) and 2.3 months (95% CI, 0.6–6.7), respectively (P=0.24). At the initiation of chemotherapy, 33.3% of Arm S and 60.0% of Arm C had extrahepatic spread only (P=0.24). Arm S had more vascular invasion (22.2 % vs. 10.0%; P=0.40) and a serum alpha-fetoprotein (AFP) level of > 200 ng/mL (50.0% vs. 20.0%; P=0.23); however, they were less frequently given additional therapies after progression (55.6% vs. 80.0%; P=0.25) compared to Arm C. The grade 3/4 hematological toxicity was more common in Arm C (30.0% vs. 16.7%; P=0.63), and so was grade 3/4 hepatic toxicity in Arm S (44.4% vs. 10.0%; P=0.06). Multivariate analysis indicated that pathological stage T3/T4 at the time of LT (HR 4.65; 95% CI, 1.25–17.28) and a serum AFP level of > 200 ng/mL at the initiation of chemotherapy (HR 3.41; 95% CI, 1.03–11.29) were independent prognostic factors associated with the OS. Conclusions: Patients receiving cytotoxic agents as a first-line systemic therapy seemed to have better OS, but similar TTP when compared with those receiving sorafenib. Advanced stage at the time of LT and a higher AFP level at the initiation of chemotherapy were independent prognostic factors associated with the overall survival of patients with recurrent HCC after LT.

Disclosures:

Joong-Won Park - Advisory Committees or Review Panels: Bayer Heathcare Co., Taiho Pharmaceutical Co.

The following people have nothing to disclose: Bo Hyun Kim, Sang Myung Woo, Seong Hoon Kim, Seung Duk Lee, Young-Kyu Kim, Eun Uk Jung, Woo Jin Lee, Eun Kyung Hong, Chang-Min Kim

2133

  1. Top of page

Amplifications At Chromosome 8Q11.23 And 8Q12.1 Are Associated With Poor Survival In Patients With Hepatocellular Carcinoma

Ho Joong Kim, Joo An Hwang, Hyo Jung Cho, Sung Won Cho, Jae Youn Cheong;
Ajou univ. medical center, Suwon city, Republic of Korea

Introduction: Array comparative genomic hybridization (CGH) has emerged as a revolutionary platform, enabling the high resolution detection of DNA copy number aberrations. In hepatocellular carcinoma (HCC), gain of 8q has been reported to be associated with aggressive tumor behavior and lead to the overexpression of oncogenes. Methods: Primary HCC tissues from 45 patients and control liver samples from 25 patients were obtained during hepatic resection as treatment for HCC and metastatic liver tumor. A technology of array CGH using DNA chips spotted with 4041 BAC clones were applied to examine the DNA copy number aberrations in control and HCC tissues. Results: We identified 259 significant clones associated with differential aberration frequencies between normal liver tissue and HCC samples. Amplifications at 8q11.23 and 8q12.1 occurred more frequently in patients with poor survival (P<0.01). ST18 and SOX17 genes are located in chromosome 8q11.23, 12.1, a region found to be amplified in HCCs. Therefore, these alterations might contribute to a worse prognosis by triggering aggressive tumor phenotypes such as vascular invasion or intrahepatic metastasis. Conclusion: We identified probable chromosomal regions associated with the patient survival in HCC subjects. ST18 and SOX17 were potential targets within the chromosomal gain at 8q11.23 and may be associated with poor prognosis in patients with HCC.

Disclosures:

The following people have nothing to disclose: Ho Joong Kim, Joo An Hwang, Hyo Jung Cho, Sung Won Cho, Jae Youn Cheong

2134

  1. Top of page

Altered Gene Expression of Vitamin D Key Players and VDR in Human Hepatocellular Carcinoma

Evelin Horváth1, Bernadett Balla1, János Kósa1, Péter Lakatos1, Áron Lazáry1, Dániel Németh1, Hasan Jozilan1, Áron Somorácz2, Anna Korompay2, Benedek Gyöngyösi2, Katalin Borka2, András Kiss2, Péter K. Kupcsulik3, Zsuzsa Schaff2, Ferenc Szalay1;
11st Department of Medicine, Semmelweis University, Budapest, Hungary; 22nd Department of Pathology, Semmelweis University, Budapest, Hungary; 31 st Department of Surgery, Semmelweis University, Budapest, Hungary

Aim: 1,25-DihydroxyvitaminD3 (1,25(OH)2D3), the active form of vitamin D, mediates antitumor effects mostly via the vitamin D nuclear receptor (VDR) in various cancers, also in hepatocellular carcinoma (HCC). More than the tierce part of the patients with HCC had complete response or disease stabilization after systemic 1,25(OH)2D3 analogue treatment. To understand the different response of the patients, we examined mRNA and protein expression differences of 1,25(OH)2D3-inactivating CYP24A1 and mRNA expression of the activating CYP27B1 enzymes, as well as that of VDR between human non-tumorous liver and hepatocellular carcinoma (HCC) tissue samples, and studied the correlation with clinicopathological characteristics. Patients and Methods: We evaluated 49 patients who had undergone liver surgery for therapeutic purposes. mRNA expression of VDR, CYP24A1 and CYP27B1 was measured by RT-PCR in HCC tissues compared with the surrounding non-tumorous liver tissue from the same patient. CYP24A1 protein was detected by immuncytochemistry.

Results: VDR and CYP27B1 mRNA were both expressed in HCC and in surrounding liver tissues as well, although to a different extent. Expression of VDR and CYP27B1 was significantly lower in HCC compared with non-tumorous liver (p<0.05). Interestingly, neither of the non-tumorous liver expressed CYP24A1 mRNA, in contrast the majority of the HCC samples we found the expression of this gene. Immuncytochemistry study proved that the gene activation was followed by CYP24A1 protein synthesis in these samples. The highest VDR mRNA expression was found in patients with hepatitis B (HBV) infection, that was significantly higher than in patients with other etiology (p<0.05). CYP27B1 mRNA was significantly lower in female non-tumorous liver tissues compared with males (p<0.05). Conclusions: Our novel data on the presence of 1,25(OH)2D3 inactivating CYP24A1 mRNA, as well as reduced expression of VDR and CYP27B1 mRNA in human HCC samples indicate the decreased bioavailability of 1,25(OH)2D3 in HCC cells, providing an escape mechanism from the anti-tumor effect of 1,25(OH)2D3. This could be the reason for 1,25(OH)2D3 insensitivity in the majority of patients with HCC to the 1,25(OH)2D3 treatment. Based on our novel results, it could be indicated to individualize the selection of patients with HCC for 1,25(OH)2D3 treatment.

Disclosures:

The following people have nothing to disclose: Evelin Horváth, Bernadett Balla, János Kósa, Péter Lakatos, Áron Lazáry, Dániel Németh, Hasan Jozilan, Áron Somorácz, Anna Korompay, Benedek Gyöngyösi, Katalin Borka, András Kiss, Péter K. Kupcsulik, Zsuzsa Schaff, Ferenc Szalay

2135

  1. Top of page

Comparison of efficacy of hepatic arterial infusion chemotherapy and sorafenib in advanced hepatocellular carcinoma with portal vein tumor thrombosis

Do Seon Song1,2, Myeong Jun Song1,2, Sung Won Lee1,2, Si Hyun Bae1,2, Woo Jin Chung3, Jae Young Jang4, Young Seok Kim4, Jun Yong Park5, Hyung Joon Yim6, Sung Bum Cho7, Soo Young Park8, Jong Young Choi1,2, Seung Kew Yoon1,2, Jin-Mo Yang1,2;
1The Catholic University Liver Research Center, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea; 2Department of Internal Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea; 3Depart-ment of Internal Medicine, Keimyung University School of Medicine, Daegu, Republic of Korea; 4Department of Internal Medicine, Soonchunhyang University College of Medicine, Seoul, Republic of Korea;5Department of Internal Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea; 6Department of Internal Medicine, Korea University, Seoul, Republic of Korea;7Department of Internal Medicine, Hwasoon Chonnam National University Hospital, Kwangju, Republic of Korea; 8Department of Gastroenterol-ogy and Hepatology, Kyungpook National University, Daegu, Republic of Korea

Background. Patients with hepatocellular carcinoma (HCC) with portal vein tumor thrombosis (PVTT) have an extremely poor prognosis. The objective of this study was to compare the efficacy of hepatic arterial infusion chemotherapy and sorafenib in advanced HCC patients with PVTT. Methods. The study involved 11 0 advanced HCC patients with PVTT (50 in the HAIC group and 60 in the sorafenib group) observed from February 2008 to January 2013 in 7 Korean centers. All patients had good liver functions (Child-Pugh score<7). Primary outcomes were overall survival (OS) and time to progression (TTP). Secondary outcomes were response rates. Results. Median follow-up duration was 5.6 months (range, 0.7–54.7 months). There was no significant difference in median OS between HAIC group [median 7.2 months (95% CI 4.97 - 9.43 months)] and sorafenib group [median 5.7 months (95% CI 4.45 - 6.96 months)] (p=0.150). Median time to progression were 3.3 months (95% CI, 2.70 - 3.90) in HAIC group and 2.1 months (95% CI, 1.52 - 2.68) in sorafenib group (p=0.034). Objective response (complete response + partial response) rate were 24.5% and 13.3% in HAIC group and sorafenib group, respectively (p=0.146). Disease control (objective response + stable disease) rate were 91.8% and 45% in HAIC group and sorafenib group, respectively (p<0.001). Tumor size (≧10cm) (hazard ratio 2.12, 95% CI 1.29 - 3.49) and combination with locoregional therapy (hazard ratio 0.49, 95% CI 0.28 -0.86) were independent predictive survival factors. Conclusion. In advanced HCC patients with PVTT, HAIC treatment can provide results comparable to sorafenib in terms of similar OS, longer TTP and higher disease control rate.

Disclosures:

Hyung Joon Yim - Grant/Research Support: GSK Korea, Handok Pharm, Gilead Korea; Speaking and Teaching: BMS Korea

The following people have nothing to disclose: Do Seon Song, Myeong Jun Song, Sung Won Lee, Si Hyun Bae, Woo Jin Chung, Jae Young Jang, Young Seok Kim, Jun Yong Park, Sung Bum Cho, Soo Young Park, Jong Young Choi, Seung Kew Yoon, Jin-Mo Yang

2136

  1. Top of page

Comparison Among Transarterial Chemoembolization, Radiofrequency Ablation, and Hepatic Resection for Small Single Nodular Hepatocellular Carcinoma Using Inverse Probability Weighting

Hyo-Joon Yang1, Young Youn Cho1, Jeong-Hoon Lee1, Dong Hyeon Lee1, Su Jong Yu1, Yoon Jun Kim1, Jung-Hwan Yoon1, Hyo-Cheol Kim2, Jeong Min Lee2, Jin Wook Chung2, Kwang-Woong Lee3, Nam-Joon Yi3, Kyung-SukSuh3, Hyo-Suk Lee1;
1 Department of Internal Medicine and Liver Research Institute, Seoul national university hospital, Seoul, Republic of Korea; 2Department of Radiology, Seoul national university hospital, Seoul, Republic of Korea; 3Department of Surgery, Seoul national university hospital, Seoul, Republic of Korea

Background: Transarterial chemoembolization (TACE) is frequently used for small single nodular hepatocellular carcinoma (HCC) when hepatic resection (HR) and radiofrequency ablation (RFA) are not indicated. However, there are few studies comparing TACE, RFA, and HR. In this study, we aimed to compare the long-term outcomes among HR, RFA, and TACE in patients with small single nodular HCC. Methods: This retrospective cohort study included consecutive patients with single nodular HCC which is <3 cm in diameter and has no vascular invasion who initially treated with HR, RFA, or TACE from January 2005 to December 2006 at Seoul National University Hospital. The primary endpoint was death from any cause. To control selection bias, inverse probability weighting (IPW) was used. Results: A total of 218 patients were included: 57 patients received HR, 86 received RFA, and 75 received TACE as an initial treatment. There were significant differences in the baseline liver function which was best in the HR group followed by the RFA group and then, the TACE group. The cumulative probability of overall survival at 5 years of follow up was 92.4% in the HR group, 86.6% in the RFA group, and 74.5% in the TACE group (P=0.030). However, after the baseline values were balanced using IPW, overall survival became quite comparable (5 year cumulative probability of survival; 84.3% with HR, 86.4% with RFA, and 80.4% with TACE). When baseline imbalances were further adjusted by multiple Cox regression, TACE showed hazard ratio of 1.048 [95% confidence interval (CI), 0.463 - 2.372, P=0.910] compared to HR and 1.268 (95% CI, 0.628 - 2.562, P=0.507) compared to RFA. Conclusions: TACE is an effective treatment achieving long-term survival comparable to those of HR and RFA in patients with small single nodular HCC.

Thumbnail image of

Disclosures:

Jeong Min Lee - Advisory Committees or Review Panels: Bayer healthcare

The following people have nothing to disclose: Hyo-Joon Yang, Young Youn Cho, Jeong-Hoon Lee, Dong Hyeon Lee, Su Jong Yu, Yoon Jun Kim, Jung-Hwan Yoon, Hyo-Cheol Kim, Jin Wook Chung, Kwang-Woong Lee, Nam-Joon Yi, Kyung-Suk Suh, Hyo-Suk Lee

2137

  1. Top of page

Prediction of Survival in Patients with Hepatocellular Carcinoma Treated with Sorafenib by Comprehensive Serum Glycan Analysis

Kazuhiro Nouso1,2, Koji Miyahara2, Yuuki Morimoto2, Yasuto Takeuchi2, Hiroaki Hagihara2, Kenji Kuwaki2, Hideki Onishi1,2, Fusao Ikeda1,2, Yasuhiro Miyake2, Shinichiro Nakamura2, Hide-nori Shiraha2, Akinobu Takaki2, Koichi Takaguchi3, Takahisa Sato4, Shinpei Sato4, Shuntaro Obi4, Kazuko Hirose5, Maho Amano5, Shin-Ichiro Nishimura5, Kazuhide Yamamoto2;
1Molecular Hepa-tology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan; 2Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Okayama, Japan; 3Internal Medicine, Kagawa Prefectural Central Hospital, Takamatsu, Japan; 4Internal Medicine, Kyoundo Hospital, Tokyo, Japan;5Faculty of Advanced Science and Graduate School of Life Science, Hokkaido University, Sapporo, Japan

Background and aims: Sorafenib is widely used for the treatment of advanced hepatocellular carcinoma (HCC). However, prognostic factors in these patients have not been established well. Recently, Kamiyama et al. examined the serum glycan profile by comprehensive quantitative high-throughput glycome analysis and reported that glycan expressions in the serum correlated with the recurrence and the prognosis of the patients with HCC who underwent hepatectomy (Hepatology, 2013, Jan 15). The aim of this study is to elucidate the clinical utility of the N-glycan profile in advanced HCC patients treated with sorafenib using the same method used by Kamiyama et al. Methods: Eighty five advanced HCC patients treated with sorafenib in our institute and collaborative hospitals were enrolled. The serum was collected before treatment and the expression of serum N-glycans were analyzed using an automated machine for glycoblotting, "Sweetblot" prototype 7 (System Instruments Co.), and MALDI-TOF-MS (Ultraflex 3, Bruker, Germany). Obtained expression profile was compared with patients' outcome and clinical importance of each glycan was assessed. Results: Median age of the patients was 69 (range, 36–88) years and 76 (89%) patients were male. HBs-Ag and HCV-Ab were positive in 18 (21%) and 45 (53%) patients, respectively. Child -Pugh grade of 70 (82%) patients were A, and that of 15 (18%) patients were B. Vascular invasion and extra-hepatic metastasis were observed in 28 (33%) and 50 (59%) patients, respectively. Forty eight patients were progressive disease (PD) and the rest (n=37) were non-PD. Of the 61glycans detected in the serum, 15 glycans were elevated in patients with PD (P<0.05). They included 4 high mannose type glycans, 1 hybrid type, and 10 complex type glycans (4 bi-antennary, 3 tri-antennary, and 3 tetra-antennary). AUROC of these glycans for differentiate non-PD and PD were 0.62–0.69. Among the 15 glycans, G2890 (HR, 1.88; 95%CI, 1.04–3.48) in addition to Eastern Cooperative Oncology Group performance status and Child-Pugh class were independent risk factors for survival in multivariate Cox-proportional hazard model. G2890 was the glycan that showed significant correlation with the recurrence after surgical resection in the previous study. Overall survival of sorafenib-treated patients with high G2890 (median, 8.4 months) was shorter than that with low G2890 (median, 11.4 months, P=0.017) Conclusions: Comprehensive analysis of serum N-glycan profiling could determine a promising biomarker for advanced HCC patients treated with sorafenib.

Disclosures:

Kazuhide Yamamoto - Advisory Committees or Review Panels: Shionogi Pharmaceutical Co; Grant/Research Support: Tanabe Mitsubishi Co, MSD, Chugai Pharmaceutical Co, Esai Co

The following people have nothing to disclose: Kazuhiro Nouso, Koji Miyahara, Yuuki Morimoto, Yasuto Takeuchi, Hiroaki Hagihara, Kenji Kuwaki, Hideki Onishi, Fusao Ikeda, Yasuhiro Miyake, Shinichiro Nakamura, Hidenori Shiraha, Akinobu Takaki, Koichi Takaguchi, Takahisa Sato, Shinpei Sato, Shuntaro Obi, Kazuko Hirose, Maho Amano, Shin-Ichiro Nishimura

2138

  1. Top of page

Serum Insulin-Like Growth Factor-1 Level Is an Independent Predictor of Recurrence and Survival in Early Hepatocellular Carcinoma: A Prospective Cohort Study

Eun Ju Cho1, Jeong-Hoon Lee2, Yun Bin Lee2, Su Jong Yu2, Jeong Min Lee3, Kyung-Suk Suh4, Yoon Jun Kim2, Jung-Hwan Yoon2, Chung Yong Kim2, Hyo-Suk Lee2;
1Department of Internal Medicine, Kangwon National University Hospital, Chuncheon-si, Republic of Korea; 2Internal Medicine and Liver Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea; 3Radiology, Seoul National University College of Medicine, Seoul, Republic of Korea; 4Surgery, Seoul National University College of Medicine, Seoul, Republic of Korea

Abstract Purpose: Insulin like-growth factor-1 (IGF-1) reflects hepatic synthetic function and plays an important role in the development and progression of various cancers. In this study, we investigated whether pretreatment serum IGF-1 levels predict time-to-recurrence (TTR) and overall survival (OS) in patients with early-stage hepatocellular carcinoma (HCC) after curative treatment. Experimental Design: Consecutive HCC patients who had undergone surgical resection, radiofrequency ablation, or percutaneous ethanol injection as curative treatments of early HCC were included from two prospective cohorts and the training set (n=101) and the validation set (n=91) were established. Serum samples were collected before treatment and the levels of IGF-1 and IGF-binding protein-3 (IGFBP-3) were analyzed with regard to their associations with recurrence and survival. Results: In the training set, patients with low IGF-1 levels showed significantly shorter TTR (median, 14.6 months; 95% confidence interval [CI], 1.8–27.5) than patients with high IGF-1 levels (median, 50.8 months; 95% CI, 36.9–64.7; P<0.001) during a median follow-up period of 52.4 months. In the multivariate analysis, low levels of IGF-1 was an independent predictor of recurrence (hazard ratio [HR], 2.49; 95% CI, 1.52–4.08; P<0.001). Furthermore, together with high serum alpha-fetoprotein and multiple tumors, low levels of IGF-1 remained an independent predictor of poorer survival (HR, 8.00; 95% CI, 1.94–33.01; P=0.004). Applied to the independent validation set, low serum IGF-1 levels maintained its prognostic value for shorter TTR and OS. Conclusions: Low baseline IGF-1 levels independently correlated with shorter TTR and poorer survival in patients with early-stage HCC after curative treatment.

Disclosures:

Jeong Min Lee - Advisory Committees or Review Panels: Bayer healthcare

The following people have nothing to disclose: Eun Ju Cho, Jeong-Hoon Lee, Yun Bin Lee, Su Jong Yu, Kyung-Suk Suh, Yoon Jun Kim, Jung-Hwan Yoon, Chung Yong Kim, Hyo-Suk Lee

2139

  1. Top of page

The value of Golgi Protein 73 as a marker for differentiating between solid benign and malignant liver tumours

Mirelle Bröker, Jan IJzermans, Caroline Witjes, Anneke J. van Vuuren, Robert A. de Man;
Erasmus Medical Univercity Center, Rotterdam, Netherlands

Good diagnosis is essential in primary focal solid liver lesions, as the differential diagnosis consists of benign liver tumours such as Hepatocellular adenoma(HCA) and Focal Nodular Hyperplasia(FNH), but also of malignant tumours such as Hepa-tocellular Carcinoma(HCC). A promising new marker is Golgi Protein 73(GP73). While several studies have described GP73 as being specific for patients with HCC, they only included patients with liver cirrhosis and/or healthy people as controls. We determined the predictive value of GP73 in differentiating between solid benign and malignant liver tumours. The 264 patients in this study included 88 patients with HCC, 88 with HCA, and 88 with FNH. We collected a blood sample from each patient, used a quantitative ELISA assay to measure GP73 levels, and compared GP73 levels in the HCC patients with those in patients with benign liver tumours. The ROC-curve, sensitivity and specificity of GP73 were calculated and compared with those of AFP(Fig 1). The mean value of GP73(IU/ml) differed significantly between the three groups, 4.7 in patients with HCC, 2.1 in HCA and 1.8 in FNH (P<0.001). With regard to the ability to differentiate between HCC-patients from patients with an HCA and FNH, the area under ROC was 0.701 for GP73 and 0.912 for AFP. Sensitivity was 60% for GP73 and 77 % for AFP; specificity was 65% and 96% respectively. Although the literature suggests that GP73 is a valuable serum marker in patients with HCC the diagnosis of a benign liver tumour should not be excluded. The moderate sensitivity and specificity makes it unsuitable for a good discriminating between malignant and benign tumours.

Figure 1, ROC-curves comparing GP73 (dash line) and AFP (straight line).

Disclosures:

Robert A. de Man - Advisory Committees or Review Panels: crucell; Grant/Research Support: biotest, gilead

The following people have nothing to disclose: Mirelle Bröker, Jan IJzermans, Caroline Witjes, Anneke J. van Vuuren

2140

  1. Top of page

Increased expression of leptin-receptor in Inflammatory Type Hepatocellular Adenoma: the link with obesity

Jing Han1, Hironori Kusano1,3, Peter J. Zwiers1, Koert P. deJong2, Grietje Molema1, Annette S. Gouw1;
1 Pathology & Medical Biology, University Medical Center Groningen, Groningen, Netherlands; 2Surgery, division of HPB and Liver Transplantation, University Medical Center Groningen, Groningen, Netherlands; 3Pathology, Kurume University School of Medicine, Kurume, Japan

The inflammatory-type hepatocellular adenoma (IHCA), a common variant among the 4 subtypes of HCA, is a benign primary hepatocellular neoplasm which is associated with obesity. The pathogenetic mechanism behind this association is largely unknown. In this study we investigated the gene expression profiles of the adipokines leptin and adiponectin and their receptors in 25 samples of IHCA. All patients were females and their median body-mass-index was 30.1. Using real time RT-PCR and immunohistochemistry, leptin and adiponectin mRNA was not detectable in IHCA, in the non-lesional liver parts and normal control livers, whereas ample quantities were found in control adipose tissue. Increased gene expression of the leptin receptor (LR) was found in tumor compared to non-lesional liver, whereas AdiR2, the liver relevant adiponectin receptor-2 was decreased. On immunohistology expression of LR and AdiR2 was predominantly observed on sinusoidal and vascular endothelial cells (EC). Based on these findings we performed a complementary study of in vitro leptin stimulation of EC in which we found increased expression of LR and interleukin-6 by EC. In IHCA we also found immunohistologic nuclear expression of phosphorylated-STAT3, signifying downstream activation of lep-tin/LR signaling in both tumor and non-lesional liver parts. Conclusion: LR expression and activation of the leptin/LR pathway are increased in IHCA, probably due to obesity related hyper-leptinemia which can induce LR expression on tumor EC. These findings provide a first step in the study of the association between IHCA and obesity.

Disclosures:

The following people have nothing to disclose: Jing Han, Hironori Kusano, Peter J. Zwiers, Koert P. de Jong, Grietje Molema, Annette S. Gouw

2141

  1. Top of page

Categorization of non-B, non-C hepatocellular carcinoma patients using hierarchical clustering

Ryosuke Tateishi1, Takeshi Okanoue2, Naoto Fujiwara1, Kiwamu Okita3, Kendo Kiyosawa4, Masao Omata5, Hiromitsu Kumada6, Norio Hayashi7, Kazuhiko Koike1;
1 Department of Gastroenterol-ogy, The University of Tokyo Graduate School of Medicine, Tokyo, Japan; 2Saiseikai Suita Hospital, Suita, Japan; 3Shimonoseki Koh-sei Hospital, Shimonoseki, Japan; 4Nagano Red Cross Hospital, Nagano, Japan; 5Yamanashi Prefectural Hospital Organization, Kofu, Japan; 6Toranomon Hospital, Tokyo, Japan; 7Kansai Rosai Hospital, Amagasaki, Japan

Background and Aims: Non-B, non-C hepatocellular carcinoma (HCC), defined as both HBsAg and HCVAb negative, is a mix of patients with various backgrounds. Several risk factors such as alcohol consumption, obesity, fatty liver, and diabetes frequently overlap each other. We intended to classify non-B, non-C HCC patients using hierarchical clustering. Methods: We enrolled 2023 patients (mean age = 69.0+/-9.3 y, M/F=1614/409) from a nationwide survey on non-B, non-C HCC after excluding those with specific etiology including autoimmune, metabolic and vascular liver diseases and those with missing values. We performed an unsupervised hierarchical clustering analysis with following variables: age, gender, presence of hypertension, dyslipidemia, fatty liver, and diabetes mellitus, alcohol consumption, BMI, serum albumin, total bilirubin, AST, ALT, ALP, and GGT, platelet count, and pro-thrombin time. We also performed a survival analysis using Kaplan-Meier method based on generated categories. Results: Patients were divided into category 1 (male moderate drinker with normal liver enzymes and good liver function; N=497), 2 (male moderate drinker with fatty liver, obesity, and mildly impaired liver function; N=279), 3 (relatively young heavy drinker with elevated liver enzymes and poor liver function; N=602), 4 (female non-drinker with obesity and cirrhosis; N=305), and 5 (non-drinker with metabolic syndrome). Survival analysis showed significantly worse outcome in category 3 than other categories (5-year survival rate = 33.4% vs 52.3%, P<0.001 by log-rank test). Conclusions: Hierarchical clustering elucidated categories of typical alcoholic and nonalcoholic liver disease as well as those characterized by moderate alcohol intake with various degrees of obesity-related complications.

Thumbnail image of

Disclosures:

Ryosuke Tateishi - Grant/Research Support: Eisai Co. Ltd.

Kiwamu Okita - Consulting: KOWA

Kendo Kiyosawa - Consulting: MSD, Ootsuka, Tanabe Mitsubishi

Masao Omata - Consulting: Bristol-Myers Squibb, Boehringer Ingelheim, Otsuka Pharmaceutical, Bristol-Myers Squibb, Boehringer Ingelheim, Otsuka Pharmaceutical; Speaking and Teaching: Bristol-Myers Squibb, Pfizer, Roche, Bristol-Myers Squibb, Pfizer, Roche

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

Kazuhiko Koike - Speaking and Teaching: Bristol-Myers Squibb

The following people have nothing to disclose: Takeshi Okanoue, Naoto Fuji-wara, Norio Hayashi

2142

  1. Top of page

Alcohol-related hepatocellular carcinoma in France: analysis from the CHANGH study

Charlotte E. Costentin1, Thomas Decaens1, Pierre Lahmek2, Claire Becker3, Robert Bader4, Anne-Bérangère Marks5, Bertrand Con-dat6, Moana Gelu-Simeon7, Hélène Labadie8, Arnaud Pauwels9, Joanna Pofelski10, Jacques Arnaud Seyrig11, Jean Louis Coc-quart12, Jacques Denis13, Isabelle Rosa14; 1Hôpital Henri Mondor, Creteil, France; 2Hôpital Emile Roux, Limeil Brevannes, France; 3Centre Hospitalier de Lens, Creteil, France; 4CHG E.Muller, Mul-house, France; 5Hôpital de Valenciennes, Valenciennes, France; 6Hôpital Saint Camille, Bry-sur-Marne, France; 7CHU de Pointe-À-Pitre, Pointe-À-Pitre, France; 8Hôpital Delafontaine, Saint Denis, France; 9Centre hospitalier de Gonesse, Gonesse, France; 10Cen-tre Hospitalier de la Région d'Annecy, Pringy, France;
11Centre hospitalier du Centre Bretagne, Pontivy, France; 12Centre Hospitalier Oloron Sainte Marie, Oloron Sainte Marie, France; 13Centre Hospitalier Sud Francilien, Corbeil-Essonnes, France; 14Centre Hospitalier Intercommunal de Créteil, Créteil, France

The CHANGH study is a French observational cohort collecting data on clinical features and treatment allocation of hepatocellular carcinoma (HCC). The aim of this study was to compare the clinical features at diagnosis and the therapeutic allocation of alcohol-related (group A) and non alcohol-related (group B) HCCs. Patients and methods: 103 French hepatology units included 1207 patients between May 2008 and October 2009. Inclusion criteria was onset of a new HCC. Diagnosis of HCC relied on histology or radiological Barcelona criteria. Results: Patients with both etiologies (alcohol+another cause of liver disease) were exluded. 894 patients were analyzed: 582 (65%) in group A and 312 (35%) in group B (including 62 HBV (20%), 87 HCV (28%), 60 NAFLD (19%), 11 hemochromato-sis (3%), 14 with another etiology (5%) and 78 with mixed etiology (alcohol excluded, 25%)). At inclusion, mean age, α-fetoprotein levels > 500 ng/dl, presence of extra-hepatic metastasis were similar in both groups. Male gender and presence of esophageal varices were more frequent in group A compared to group B: 91% vs 67% and 68 % vs 43% respectively (p<0.0001). HCC was detected during a screening program in 119 patients in group A (20%) and in 80 (26%) in group B (p=0.12). Single HCC was detected in 195 (34%) patients in group A and in 132 (43%) patients in group B (p=0.008). Diffuse HCC was more frequent in group A (n=106, 18%vsn=34, 11%) (p=0.011). Child-Pugh score was higher in group A (class A: 37 %; B: 37 %; C: 20 %) than in group B (class A: 49 %; B: 18 %; C: 8 %) (p=<0.0001). BCLC stage was evaluable in 716 patients. Comparison of BCLC stage in group A versus group B was different but overall comparison didn't reach statistical significance (p=0.069). 172 (30 %) patients in group A and 82 (26%) in group B met the Milan criteria (p=0.313). Performance Status grade 0 or 1 was less frequently observed in group A (145; 55%) than in group B (11 8; 63%) (p=0.016). Potentially curative treatment was performed in 126 (22%) vs 92 (30%) patients in groups A and B, respectively (p=0.009).Among alcohol-related HCC, 305 occurred in abstinent patients (52%). More curative treatment were performed in abstinent patients (26%) vs non abstinent patients (18%) (p=0.026). There was no difference between group A abstinent patients and group B patients (26% vs 29% p=0.260). Conclusion: Our study show that alcohol-related HCC is more frequent in male gender, more frequently diffuse, detected in patients with more impaired performance status and less suitable for curative therapy. Finally, abstinence is a predictive factor of curative treatment.

Disclosures:

Jacques Denis - Consulting: ROCHE, France, ROCHE, France, ROCHE, France, ROCHE, France; Speaking and Teaching: BMS, BMS, BMS, BMS

The following people have nothing to disclose: Charlotte E. Costentin, Thomas Decaens, Pierre Lahmek, Claire Becker, Robert Bader, Anne-Bérangère Marks, Bertrand Condat, Moana Gelu-Simeon, Hélène Labadie, Arnaud Pauwels, Joanna Pofelski, Jacques Arnaud Seyrig, Jean Louis Cocquart, Isabelle Rosa

2143

  1. Top of page

Prognostic Factors of Radiofrequency Ablation for Hepatocellular Carcinoma: A Nationwide Cohort Study

Teng-Yu Lee1,2, Hsiu-Jon Ho2, Ken N. Kuo3, Jaw-Town Lin4, Hsien-Yuan Lane1, Ming-Shiang Wu5, Chun-Ying Wu2,1;
1Graduate Institute of Clinical Medical Science, China Medical University, Taichung, Taiwan; 2Internal Medicine, Taichung Veterans General Hospital, Taichung, Taiwan; 3Center for Health Policy Research and Development, National Health Research Institutes, Miaoli, Taiwan; 4School of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan; 5Division of Gastroenterology, National Taiwan University Hospital, Taipei, Taiwan

Prognostic factors such as large tumor size and viral factors have not been fully evaluated in hepatocellular carcinoma (HCC) following radiofrequency ablation (RFA). We aimed to investigate prognostic factors of RFA by a nationwide cohort study. Based on Taiwan National Health Insurance Research Database, 50,791 patients with newly diagnosed HCC were screened between July 2004 and December 2011. After excluding combination therapy, vascular invasion, extrahepatic metastasis or follow-up period less than 6 months, 2,261 patients who received RFA as curative therapy were identified. Cumulative incidences of and hazard ratios (HRs) for HCC recurrence and patient mortality were analyzed. With a median follow-up of 33.6 months, 5-year HCC recurrence and overall survival rates were 64.0% (95% confidence interval (CI): 61.5–6.5%) and 52.6% (95% CI: 49.7–55.5%), respectively. On multivariate analysis, tumor size >5 cm (HR 1.73; 95% CI, 1.35–2.23), hepatitis B virus infection (HR 1.34; 95% CI, 1.17–1.54), hepatitis C virus infection (HR 1.46; 95% CI, 1.26–1.68) and cirrhosis (HR 1.24; 95% CI, 1.09–1.42) were associated with an increased risk of HCC recurrence, and NSAIDs or aspirin use (HR 0.86; 95% CI, 0.76–0.97) were associated with a decreased risk. In addition, increased age (HR 1.02; 95% CI, 1.01–1.03), tumor size >5 cm (HR 1.61; 95% CI, 1.21–2.16), alcoholic liver disease (HR 1.38; 95% CI, 1.01–1.88), cirrhosis (HR 1.50; 95% CI, 1.26–1.79), liver decompensation (HR 1.84; CI, 1.49–2.27) and diabetes (HR 1.50; CI, 1.23–1.83) were associated with an increased risk of patient mortality. Conclusion: HCC patients with prognostic factors such as tumor size >5 cm should be carefully selected for RFA. Chronic viral infection was associated with tumor recurrence after RFA, and antiviral treatment could be further studied.

Prognostic factor analysis of HCC recurrence

harlotteHR95% CIP value
Age per year1.001.00–1.010.24
Male gender1.120.98–1.260.09
Tumor size   
&#8806;5 cm1  
>5 cm1.731.35–2.23<0.01
Chronic viral infection   
None1  
HBV1.341.17–1.54<0.01
HCV1.461.26–1.68<0.01
Alcoholic liver disease0.870.65–1.150.31
Cirrhosis1.241.09–1.42<0.01
Diabetes1.060.90–1.240.50
Metformin use0.970.82–1.140.70
Statin use0.860.67–1.090.21
NSAIDs or aspirin use0.860.76–0.97<0.05

Disclosures:

The following people have nothing to disclose: Teng-Yu Lee, Hsiu-Jon Ho, Ken N. Kuo, Jaw-Town Lin, Hsien-Yuan Lane, Ming-Shiang Wu, Chun-Ying Wu

2144

  1. Top of page

The expressions of cancer stem cell markers and EMT-associated genes in peripheral blood during the periop-erative period are related to postoperative recurrence of hepatocellular carcinoma

Gi Hong Choi1, Gwangil Kim2, Dai Hoon Han1, Young Nyun Park2, Jin Sub Choi1;
1Department of Surgery, Yonsei University College of Medicine, Seoul, Republic of Korea; 2Department of Pathology, Yonsei University College of Medicine, Seoul, Republic of Korea

Purpose: To investigate the correlations of messenger RNA (mRNA) expression levels of cancer stem cells (CSCs) markers and epithelial-mesenchymal transition (EMT)-associated genes in peripheral blood and their dynamic changes during the peri-operative period with postoperative recurrence of hepatocellular carcinoma (HCC). Method: Between April 2011 and February 2012, a total of 64 consecutive patients who underwent curative resection. The mRNA expression levels of three surface markers for CSCs (K19, EpCAM, and CD44) and two EMT-associated genes (Snail and Twist) in peripheral blood and HCC tissues were evaluated using a real-time reverse tran-scription-polymerase chain reaction (RT-PCR). Peripheral blood samples were collected before operation, after operation and at 7 days after operation. These samples were also obtained from 23 control subjects. To evaluate the dynamic changes of these markers, the median changed value of each marker in control patients was used as the cutoff value to divide study subjects into elevation and no elevation groups. The cutoff values of K19, EpCAM and CD44 were 1.99, 2.94 and 1.02, respectively. Results: Eighteen patients (34.9%) developed recurrence during the median follow-up, 16 months after resection. Although HCC tissues demonstrated higher levels of mRNA expression of K19 and CD44 in patients with recurrence than patients without recurrence, the preoperative peripheral blood mRNA levels of three CSC markers were not associated with postoperative recurrence. The elevation group of K19 in immediate postoperative peripheral blood tended to be associated with lower disease-free rate (p=0.082). The elevation of both K19 and CD44 was one of the independent poor prognostic factors for disease-free survival after curative resection (p=0.007). The increase in the expression of Twist or Snail mRNA in peripheral blood at postoperative seventh day was associated with lower disease-free survival rate (p=0.150). On analyzing the correlations between EMT and CSC markers, it was found that the patients with the elevation of K19 mRNA level and increased expression of EMT-associated genes (n=14) showed significantly lower disease-free survival rate after curative resection (p=0.003). Conclusion: The baseline mRNA expression levels of CSC markers in preoperative peripheral blood did not affect postoperative recurrence. If the elevation of K19 mRNA expression levels were combined with either the elevation of CD44 mRNA expression or increased expression of EMT-associated genes, it had significant correlations with postoperative recurrence and more likely represented true circulating CSCs in postoperative peripheral blood.

Disclosures:

The following people have nothing to disclose: Gi Hong Choi, Gwangil Kim, Dai Hoon Han, Young Nyun Park, Jin Sub Choi

2145

  1. Top of page

Could bile duct invasion be considered as the same as vascular invasion be in staging of hepatocellular carcinoma?

Yong Keun Park, Bong-Wan Kim, HeeJeong Wang;
Department of Surgery, Ajou University School of Medicine, Suwon, Republic of Korea

(Background/Aim) Hepatocellular carcinoma (HCC) presenting as bile duct invasion (BDI) is rare, and the independent prognostic value of this finding is unclear. According to Japanese TNM staging system for HCC, BDI is considered to carry the same prognostic significance as portal vein invasion (PVI) and hepatic vein invasion (HVI). However, AJCC TNM and the Barcelona clinic liver cancer staging system contain no mention of BDI. The current study mainly aimed to evaluate the predictive power of BDI in a cohort of patients with HCC who underwent surgical treatment. (Materials/Methods) 728 consecutive HCC patients underwent surgical treatment from 1994 to 2011 at our institution were enrolled in this study. Clinicopathological and oncologic outcome data were retrospectively reviewed. (Results) 28 patients with combined HCC and cholangiocarci-noma were excluded in this analysis. Overall and disease-free 5 year survival rates of the remaining 700 patients were 63.3% and 42.6%. After a median follow-up of 33 months, 379 patients (52%) developed cancer recurrence and 261 patients (35.8%) died. 42 cases (6% of 700) had HCC with BDI, and those 5 year survival rate were 48%. Independent risk factors for recurrence included tumor size > 5 cm, microvascular invasion, PVI, HVI, intrahepatic metastasis, microscopic positive resection margin and cirrhosis. Tumor size > 5 cm, microvascular invasion, intrahepatic metastasis, microscopic positive resection margin, tumor rupture and cirrhosis were significantly associated with decreased overall survival. However, BDI had no significant predictive value in multivariate analysis. Both AJCC TNM and Japanese TNM had good discriminating ability in the whole cohort, but the latter lost its prognostic power in a cohort of 42 patients who had HCC with BDI. (Conclusions) BDI was not revealed to be independent risk factor in patients with HCC who underwent surgical treatment. Japanese TNM staging system that considers BDI as the same as vascular invasion seems to be less appropriate for stratifying HCC patients, especially with BDI.

Disclosures:

The following people have nothing to disclose: Yong Keun Park, Bong-Wan Kim, Hee Jeong Wang

2146

  1. Top of page

The overall applicability of radical or effective hepatocellular carcinoma treatments was unaffected by old age

Koki Sato, Takamasa Ohki, Tomoharu Yamada, Mari Yamagami, Yasuhide Yamamoto, Kentaro Kojima, Arizumi Toshihiko, Michi-haru Seki, Nobuo Toda, Kazumi Tagawa;
gastroenterology, Mitsui memorial hospital, Tokyo, Japan

Purpose: The number of elderly hepatocellular carcinoma (HCC) patients is expected to increase. We compared the presenting features between elderly (≧80 years) HCC patients and the controls (<80 years). Methods: We enrolled 974 HCC patients, who were diagnosed and treated with appropriate therapeutic procedures between May 1984 and December 2012 (last follow-up date, 2013 May 31th). The patients were stratified into two groups on the basis of age (≧80 years, or <80 years). We compared baseline characteristics including age, sex, blood tests, and tumor related factors. We also elucidated cumulative reccurence and survival rates using Kaplan-Meire method. Cumulative reccurence rate was assessed in case of patients who were treated with RFA as initial therapy. Multivariate analysis was performed to clarify significant factors which affect on recurrence and survival. Results: The median observation period was 2.99 (range 1.47–5.12) years. There were 64 patients in the elderly group and 910 patients in the controls. A total of 597 patients received RFA as initial treatment, 36 (56%) in the elderly group and 561 (61%) in the controls (p=0.40). Elderly patients were more likely to be females (p=0.01), whereas, HBV infection were less common (p=0.03). Tumor size were significantly higher (29.5 [21–50] mm vs. 25 [18–35] mm, p=0.02) in the elderly group. Serum albumin level, total bilirubin level, ALT level were significantly higher in the controls (p=0.04, p <0.01, p=0.01, respectively), whereas, platelet count was significantly higher in the elderly group (p=0.01). There were no significant differences between the two groups in other parameters. There were no significant differences in survival and recurrence rates between the two groups. Adjusting for significant factors in univariate analysis, multivariate analysis indicated that tumor size, tumor numbers, serum albumin level, AST level, and Des-gamma-carboxy pro-thrombin level (HR 1.02; 95% confidence interval [CI], 0.92–1.12; p< 0.01, HR 1.03; 95% CI, 0.97–1.09; p=0.14, HR 0.47; 95% CI, 0.36–0.58; p< 0.01, HR 1.02; 95% CI, 1.01–1.03; p=0.04, HR 1.01; p=0.01, respectively) as independent factors which affect on the survival. Multivariate analysis is also indicated that RFA as initial treatment, tumor numbers and serum alpha-fetoprotein level (HR 1.08; 95% CI, 1.05–1.11; p=0.01, HR 1.01; p=0.03) as independent factors of HCC reccurence. Age over 80 years neither affected survival nor reccurence. Conclusion: Survival and reccurence rates of HCC were unaffected by age over 80 years. Even if patients are extremely old age, radical treatment for HCC might improve the survival rate in elderly patients.

Disclosures:

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

2147

  1. Top of page

LC-MS profiling of N-glycans derived from human serum samples for biomarker discovery in hepatocellu-lar carcinoma

Habtom W. Ressom1, Minkun Wang1, Yi Zhao1, Tsung-Heng Tsai1, Cristina Di Poto1, Yunil Hu2, Shiyue Zhou2, Mahlet Tadesse3, Dina H. Ziada4, Kirti Shetty5, Yehia Mechref2;
1Oncology, Georgetown University, Washington, DC; 2Chemistry and Biochemistry, Texas Tech University, Lubbock, TX; 3Mathematics & Statistics, Georgetown University, Washington, DC; 4Tropical Medicine and Infectious Diseases, Tanta University, Tanta, Egypt; 5Medicine, Georgetown University, Washington, DC

BACKGROUND: Glycoproteins such as alpha-fetoprotein are relevant to liver pathology because of the major influence of liver on the homeostasis of blood glycoproteins. The heterogeneity due to the dynamic nature of glycosylation, however, has raised significant challenges in the analysis of glycoproteins. An alternative strategy is to evaluate the levels of glycans detached from proteins. The objective of this study is to identify candidate N-glycan biomarkers in serum that distinguish hepa-tocellular carcinoma (HCC) cases from patients with liver cirrhosis. METHODS: Prospectively evaluated adult patients with liver cirrhosis were recruited from the hepatology practice at Georgetown University Hospital (US cohort, 94 patients) and Tanta University Hospital (Egyptian cohort, 89 patients). All patients were diagnosed to have cirrhosis on the basis of established clinical, laboratory and imaging criteria. About half of the patients in each cohort were diagnosed to have HCC (cases) based on either histology or characteristic imaging features. Controls were required to be HCC free for at least 6 months from the time of study entry. Sample preparation consists of release, purification, reduction and permethylation of N-glycans in serum. The samples were then analyzed by using liquid chromatography-mass spectrometry (LC-MS). Following data preprocessing, significant ions were selected. Putative gly-can structures of these ions were assigned based on the number of five monosaccharides: N-acetylglucosamine (GlcNAc), man-nose (Man), galactose (Gal), fucose (Fuc), and N-acetylneu-raminic acid (NeuNAc). RESULTS: We identified 14 N-glycans with statistically significant difference between HCC cases and cirrhotic controls. Half of these were previously reported as biomarkers for HCC (see table). These glycans represent possible modifications including hyperfucosylation, increased branching and bisecting N-acetylglucosamine. The majority of these glycans was down-regulated in HCC vs. cirrhosis (fold change > 2 and p-value < 0.05) in both the US and Egyptian cohorts. CONCLUSIONS: LC-MS profiling of N-glycans removed from serum proteins allows relative quantification of hundreds of oligosaccharides and identification of reliable candidate biomarkers for HCC.

Putative glycan structuresCohortsFold change (HCC vs. Cirrhosis)
GlcNAc 4 Man 3 Gal 2Egypt, US[DOWNWARDS ARROW]2.3,[DOWNWARDS ARROW]2.5
GlcNAc 4 Man 3Egypt, US[DOWNWARDS ARROW]2.7, [DOWNWARDS ARROW]3–8
GlcNAc 5 Man 3 FucEgypt, US[DOWNWARDS ARROW]2.7, [DOWNWARDS ARROW]2.2
GlcNAc 5 Man 3 GaiNeuNAcEgypt, US[DOWNWARDS ARROW]4- 2.4. [DOWNWARDS ARROW]4- 2.7
GlcNAc4Man3FucEgypt[DOWNWARDS ARROW]2.8
GlcNAc 5 Man 3 Gal 3 FucNeuNAc 3US[DOWNWARDS ARROW]2.1
GlcNAc4Man3GalFucUS[UPWARDS ARROW]1.9

2148

  1. Top of page

Prognostic significance of VEGF in the diagnosis of Hepatocellular carcinoma

Felix Garcia-Pajares, Carolina Almohalla, Alba L. Vargas, Irene Peñas, Fernando Santos Santamarta, Ramón Sánchez Ocaña, Maria Antonia Vallecillo, Gloria Sanchez-Antolin, Agustin Caro-Paton;
HEPATOLOGY. LIVER TRANSPLANT UNIT, HOSPITAL RIO HORTEGA, Valladolid, Spain

Objective: To investigate the expression level of serum vascular endothelial growth factor (VEGF) in patients with hepatocellular carcinoma (HCC) according to tumor size and comparing its expression in patients with liver cirrhosis (LC). Methods: Serum VEGF levels were measured using enzyme -linked immunosor-bent assay (ELISA) in 40 patients with LC and in 48 cirrhotic patients diagnosed of HCC. Clinical data were retrospectively collected from all the patients: Cirrhosis etiology, number and size of the lesions, serum alpha-fetoprotein and VEGF level. We compared the results obtained in patients with tumor size of less than 5 cm versus those equal to or greater than 5 cm. Results: The mean age was 59.5 (+ / -11.9) in LC patients and 66.6 (+ / -9.6) in CHC respectively. Alcohol was the most common cause of cirrhosis in the LC group (57.5%) and the Hepatitis C virus (47.8%) in the group of patients with HCC. In the HCC group 58.7% had one lesion, 21.7% had 2, 2.17% had 3 and 17.4% had 4 or more lesions. The average size was 4.27 + / -2.20 cm. The average serum level of VEGF in HCC patients was 212.4 pg / ml, significantly higher than the LC patients (P = 0.001) Table 1. VEGF serum levels were not correlated with alpha-fetoprotein levels. Furthermore VEGF was lower in CHC <5 cm compared to tumors greater than or equal to 5 cm (Table 2). The 5-year mortality was also significantly higher in the group of patients with VEGF> 220 pg / mL (p <0.001). Patients with serum VEGF> 220 pg / ml had worse survival compared to those with lower VEGF levels (P = 0.002). The serum VEGF level was an important prognostic factor in the multivariate analysis. .Conclusions: VEGF levels are significantly higher in cirrhotic patients with HCC. However, it is not correlated to alpha-fetoprotein levels. Serum tumor markers may be a new marker to diagnose the early HCC

VEGF AND alpha-fetoprotein LEVELS

 HCC PATIENTSCIRRHOTIC PATIENTS 
alpha-fetoprotein (ng/mL)32.252.8P<0.00I
VEGF (pg/mL)212.4157.6P<0.001
VEGF IN HCC<5cm160,88  
VEGF IN HCC>=5cm277,4789474  

Disclosures:

The following people have nothing to disclose: Felix Garcia-Pajares, Carolina Almohalla, Alba L. Vargas, Irene Peñas, Fernando Santos Santamarta, Ramón Sánchez Ocaña, Maria Antonia Vallecillo, Gloria Sanchez-Antolin, Agustin Caro-Paton

Disclosures:

Kirti Shetty - Grant/Research Support: Ikaria, Novartis, Onyx-Bayer, Hyperion; Speaking and Teaching: Merck-Schering Plough, Salix, Gilead, Onyx

The following people have nothing to disclose: Habtom W. Ressom, Minkun Wang, Yi Zhao, Tsung-Heng Tsai, Cristina Di Poto, Yunil Hu, Shiyue Zhou, Mahlet Tadesse, Dina H. Ziada, Yehia Mechref

2149

  1. Top of page

Risk Prediction of Hepatocellular Carcinoma in Patients with Cirrhosis: Validation of The ADRESS-HCC Risk Model

Ju Dong Yang3,1, Jennifer A. Flemming2, Joseph J. Larson4, Norah Terrault2, Lewis R. Roberts1, W. Ray Kim1;
1Mayo Clinic College of Medicine, Rochester, MN; 2Division of Gastroenterology and Hepatology, University of California San Francisco, San Francisco, CA; 3Department of Internal Medicine, University of Arkansas For Medical Science, Little Rock, AR; 4Division of Biomedical Statistics and Informatics, Mayo Clinic College of Medicine, Rochester, MN

Background: The ADRESS-HCC score was developed to predict risk of hepatocellular carcinoma (HCC) in patients awaiting liver transplantation. Consisting of 6 variables (age, diabetes, race, etiology, sex and CTP score), the score identifies patients with high enough risk to warrant surveillance. A score of 4.67 corresponds to an annual incidence of 1.5%, the level recommended for surveillance. Aim: We validate the ADRESS-HCC score in an independent cohort of patients with cirrhosis. Methods: All cirrhotic patients (n=428) enrolled in the Hepatitis C Antiviral Long-term Treatment against Cirrhosis (HALT-C) Trial were included in this analysis. ADRESS-HCC score was calculated as described in the figure. The Cox regression analysis was used to examine performance of the model predicting the incidence of HCC. Results: The mean age was 50.1 years. Men accounted for 72% and non-Hispanic whites 71%. Diabetes was present in 28%. The mean CTP score was 5.3. After a median follow-up of 5 years, 29 patients developed HCC. The mean ADRESS-HCC score was 4.96 (range 3.45- 6.30) with corresponding 1 year risk of 2.0% (range 0.44%-7.62%). In the figure, when patients were divided into three groups based on the risk score (quartile 1, low risk; quartiles 2/3, intermediate risk; and quartile 4, high risk), patients with intermediate risk (hazard ratio [HR], 5.8; 95% confidence interval [CI],1.1–106.1; P=0.03) or high risk (HR, 13.7; 95%CI, 2.8–246.6; P<0.01) has increased risk of HCC compared to low risk patients. At the threshold score of 4.67, the sensitivity to detect HCC within 5 years was 96.6%, indicating that virtually all patients that develop HCC could be identified by the score. Conclusion: The ADRESS-HCC risk model is useful in estimating risk of HCC development in an independent cohort of HCV patients with cirrhosis. This score may be used to improve effectiveness of surveillance for HCC.

Thumbnail image of

Disclosures:

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

Lewis R. Roberts - Advisory Committees or Review Panels: Inova; Grant/Research Support: Bristol Myers Squibb, Bayer, Nordion; Speaking and Teaching: Nordion

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

The following people have nothing to disclose: Ju Dong Yang, Jennifer A. Flemming, Joseph J. Larson

2150

  1. Top of page

MESIAH Score, BCLC, and other Staging Models as Predictors of Survival of Hepatocellular Carcinoma in HIV-Infected Patients

Mamta K. Jain1, Luciana O. Kikuchi2, Beatriz Minguez3, M. Ventura3, Eugenia Vispo4, Ting-Yi Chen6, Maaz B. Badshah 17, Emma E. Page7, Lynn E. Taylor8, Kristen M. Marks9, Marina Nunez10, Maria D. Hernandez11, ZibaJalali12, Erica Palys13, Nazia Qazi14, Marianne Harris5, Rena K. Fox15, Ayse Aytaman16, Judith Aberg6, Norbert Brau17, 18; 1 University of Texas Southwestern Medical Center, Dallas, TX; 2Universidade de São Paulo, São Paulo, Brazil; 3Hospital Universitario Val D'Hebron, Barcelona, Spain; 4Hospital Carlos III, Madrid, Spain; 5St. Paul's Hospital & University of British Columbia, Vancouver, BC, Canada; 6Bellevue Hospital Center & NYU School of Medicine, New York, NY; 7Chelsea and Westminster Hospital & Imperial College, London, United Kingdom; 8Miriam Hospital & Brown University, Providence, RI; 9Weill Cornell College of Medicine, New York, NY; 10Wake Forest University, Winston Salem, NC;
11University of Cincinnati, Cincinnati, OH; 12Cedar Sinai Medical Center & UCLA School of Medicine, Los Angeles, CA; 13VA Greater Los Angeles HCS & UCLA School of Medicine, Los Angeles, CA; 14Washington, DC VA Medical Center & George Washington University, Washington, DC; 15Universityof California San Francisco, San Francisco, CA; 16VA New York Harbor HCS & SUNY Downstate Health Science Center, Brooklyn, NY; 17Bronx VA Medical Center, Bronx, NY; 18Divisions of Infectious Diseases and Liver Diseases, Mount Sinai School of Medicine, New York, NY

Background: Recently, the MESIAH score (Model to Estimate Survival In Ambulatory HCC Patients) has been proposed as predictor of survival in patients with HCC. It has not been validated in HCC patients with HIV infection, which is the objective of this study. Methods: Overall, 216 HIV-infected patients with HCC were retrospectively identified from 1995–2013 in 38 centers in North and South America, Europe, and Australia. Predictability of survival using the MESIAH score was compared with 4 other published prediction models: BCLC (Barcelona-Clinic-Liver-Cancer), CLIP (Cancer-of-the-Liver-Italian-Program), J IS (Japan-Integrated-System) and Okuda. Univariate analysis was determined by Kaplan Meier analysis for univariate analysis of categorical variables and unadjusted Cox regression for continuous variables. Multivariate analysis was performed by Cox regression. Results: Among the 216 patients, there were 123 deaths with a median survival of 11.3 months (95% confidence interval, 8.3 - 14.3 months). The median MESIAH score was 4.034 (range, 2.23 - 7.39, interquartile range, 3.22 -4.94) with 80% of patients having a performance status (PS) of 0–1. In univariate analysis, all 5 models significantly (all p< 0.0001) predicted survival. In multi-variable step-wise Cox regression analysis using the 5 models and controlling for age, CD4+ cell count, log HIV viral load and MELD score, both MESIAH (H.R., 0.39; p<0.001) and BCLC (H.R., 0.69; p=0.001), but not CLIP, JIS, and Okuda, remained independent predictors of survival. Median MESIAH was significantly low in BCLC stage A vs. B (3.17 vs. 4.30, p<0.001), but there was no difference between stage B, C, and D. Within all BCLC stages, MESIAH further differentiated survival using the median as cut-off (BCLC A, median survival, 41.8 vs. 22.2 months, p=0.035, log rank; BCLC B, 16.9 vs. 7.6 months, p=0.001; BCLC C, 16.4 vs. 3.2 months, p<0.001; BCLC D, 7.4 vs. 1.2 months, p=0.001 ). Conclusion: In HIV-infected patients, MESIAH and BCLC complement each other in predicting survival of HCC in HIV-infected patients. In all BCLC stages, MESIAH can further differentiate survival.

Disclosures:

Mamta K. Jain - Advisory Committees or Review Panels: Merck, Vertex, Boehringer Ingelheim ; Grant/Research Support: Vertex, Bristol-Myers Squibb, Pfizer, Janssen, Gilead, Boehrnger Ingelheim , Viiv Healthcare, EMD Serono

Luciana O. Kikuchi - Speaking and Teaching: Bayer, Bayer, Bayer, Bayer

Kristen M. Marks - Grant/Research Support: Vertex, BMS, Boerhinger-Ingelheim, Janssen, Gilead; Speaking and Teaching: BMS

Marina Nunez- Consulting: Gilead

Norbert Brau - Advisory Committees or Review Panels: Janssen; Grant/Research Support: BMS, Gilead, Vertex; Speaking and Teaching: Vertex, Onyx

The following people have nothing to disclose: Beatriz Minguez, M. Ventura, Eugenia Vispo, Ting-Yi Chen, Maaz B. Badshah, Emma E. Page, Lynn E. Taylor, Maria D. Hernandez, Ziba Jalali, Erica Palys, Nazia Qazi, Marianne Harris, Rena K. Fox, Ayse Aytaman, Judith Aberg

2151

  1. Top of page

Trends of Hepatocellular Carcinoma (HCC) in HIV-Infected Patients over Time, 1995 - 2013

Kristen M. Marks1, Emma E. Page9, Nicolás Merchante3, Luciana O. Kikuchi5, Beatriz Minguez6, Eugenia Vispo7, Marianne Harris4, Maaz B. Badshah2, Mamta K. Jain8, Sonja Marcus2, David E. Kaplan10, Mark Nelson9, Pablo Barrero7, Juan A. Pineda3, Norbert Brau2,11; 1Weill Cornell College of Medicine, New York, NY; 2Bronx VA Medical Center, Bronx, NY; 3Hospital Universitario de Valme, Sevilla, Spain; 4St. Paul's Hospital & University of British Columbia, Vancouver, BC, Canada; 5Universidade de São Paulo, São Paulo, Brazil; 6Hospital Universitario Vall d'Hebron, Barcelona, Spain; 7Hospital Carlos III, Madrid, Spain; 8University of Texas Southwestern Medical Center, Dallas, TX; 9Chelsea and Westminster Hospital & Imperial College, London, United Kingdom; 10Philadelphia VA Medical Center & University of Pennsylvania, Philadelphia, PA;
1 1Mount Sinai School of Medicine, New York, NY

Background: The incidence of HCC overall has been rising worldwide. Cases of HCC in HIV-positive patients have only recently been reported, and their frequency over time as well as trends over time is unknown. Methods: HIV-infected patients with HCC (AASLD criteria) were retrospectively identified from 1992–2011 in 38 centers in North and South America, Europe, and Australia. Time of diagnosis was divided into earlier (1995–2005) and later years (2006–2011). Results: Among 221 HIV-infected patients with HCC, the number of cases rose steadily between 1995 and 2011. Compared to diagnosis pre-2006 (n=102), patients with a diagnosis 2006 or later tended to be diagnosed through screening more often (63% vs. 52%, p=0.097), were on antiretroviral therapy more often (90% vs. 79%, p=0.022), and had lower HIV viral load (mean, 1.98 vs. 2.80 log copies/mL, p<0.001), higher median CD4+ cells (356 vs. 272 per mm3, p=0.001), and lower median AFP levels (87 vs. 727 ng/mL, p=0.001). They also tended to have effective HCC therapy more often (64% vs. 51%, p=0.053), mostly because of more frequent use of surgical resection (16% vs. 5%) and of sorafenib, which became available in 2007 (12% vs. 0%). There was no difference in age (mean, 51 years), etiology of HCC (HCV, 77%; HBV, 22%, non-viral, 1%), and tumor staging (BCLC stages C&D, 56% vs. 54%). Survival was significantly longer 2006–2011 (median, 16.2 vs. 7.4 months, p=0.004). In multi-variable Cox regression analysis, only screening, log HIV viral load, AFP <200 ng/mL, BCLC staging, and effective HCC therapy were independently predictive of survival but not year of diagnosis before or after 2006. Conclusion: In HIV-infected patients with HCC, a diagnosis in 2006–11 was associated with better survival than one in 1995–2005. This was explained by better HIV viral control, more frequent screening and HCC therapy, as well as a lower AFP level.

Disclosures:

Kristen M. Marks - Grant/Research Support: Vertex, BMS, Boerhinger-Ingelheim, Janssen, Gilead; Speaking and Teaching: BMS

Luciana O. Kikuchi - Speaking and Teaching: Bayer, Bayer, Bayer, Bayer

Mamta K. Jain - Advisory Committees or Review Panels: Merck, Vertex, Boehringer Ingelheim ; Grant/Research Support: Vertex, Bristol-Myers Squibb, Pfizer, Janssen, Gilead, Boehrnger Ingelheim , Viiv Healthcare, EMD Serono

David E. Kaplan - Grant/Research Support: Merck, Bayer

Mark Nelson - Advisory Committees or Review Panels: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead; Consulting: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead; Grant/Research Support: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead, Roche; Speaking and Teaching: Boehringer Ingelheim, Janssen, MSD, BMS, Abbott, Viiv, Gilead

Norbert Brau - Advisory Committees or Review Panels: Janssen; Grant/Research Support: BMS, Gilead, Vertex; Speaking and Teaching: Vertex, Onyx

The following people have nothing to disclose: Emma E. Page, Nicolás Merchante, Beatriz Minguez, Eugenia Vispo, Marianne Harris, Maaz B. Badshah, Sonja Marcus, Pablo Barrero, Juan A. Pineda

2152

  1. Top of page

Inflammatory-based panels of markers do not predict post-translpant recurrence of hepatocellular carcinoma (HCC) in patients within Milan criteria

Emmanuel Tsochatzis1, Hasitha S. Wijewantha1, Manuel Rodríguez-Perálvarez1, Pinelopi Manousou1, Evangelia Fatourou1, Pieri Giulia1, Vasilios Papastergiou1, Eleni Theocharidou1, Douglas Thorburn1, David W. Patch1, James O'Beirne1, Tim Meyer2, Andrew K. Burroughs1;
1The Royal Free Sheila Sherlock Liver Centre and UCL Institute of Liver and Digestive Health, Royal Free Hospital, London, United Kingdom; 2Department of Oncology, Royal Free and UCL, London, United Kingdom

Background: Hepatocellular carcinoma (HCC) is the main indication for 15% of liver transplants. Currently the Milan criteria are widely used for listing patients, however HCC recurrence occurs in a subset of patients who fulfill these criteria. Chronic inflammation is associated with the pathogenesis of HCC and indeed published inflammation-based scores have been associated with survival in HCC cohorts. We evaluated the association of inflammation-based index (IBI), neutrophil-to-lymphocyte-ratio (NLR) and platelet-to-lymphocyte ratio (PLR) with post-transplant HCC recurrence in consecutive patients with HCC transplanted within the Milan criteria. Patients and methods: We analysed data from consecutive patients who were transplanted for HCC as main indication between 1995 and 2010 in our unit. Laboratory, epidemio- logical, radiological and histological data were analysed. Survival was evaluated using multiple regression analysis. All inflammatory scores were computed using laboratory values at the day of the transplant. IBI was scored from 0–2 (sum of albumin >35/<350/1 points and CRP <10/>100/1 points). Results: 150 patients with HCC were transplanted in the study period, of which HCC was an incidental finding on explant his tology in 19 (13%). The mean age was 54±7 years and 125 (83%) were males. Chronic hepatitis C was the primary cause of liver disease in 40% of patients, followed by chronic hepatitis B (23%) and alcohol (13%). Mean Child-Pugh score was 77±2.5. Mean NLR was 2.9±2.2, and 19 patients (12.7%) had NLR >5 which was associated with unfavorable outcome in other series. IBI scores of 0, 1 and 2 were present in 35%, 46% and 18% of patients respectively. In total, 21 patients (14%) had HCC recurrence at a mean of 32±23 (median 30) months post-transplant. None of the three panels (NLR, PLR and IBI) was associated with HCC recurrence in univariate or multivari-ate analysis or with post-transplant survival. HCC post-transplant recurrence was independently associated with no neo-adjuvant transarterial therapy (OR 4.334, 95%CI 1.257–14.939; P=0.020) and the total size of tumour nodules radio-logically (OR 1.035, 95%CI 1.005–1.066; P=0.024). Conclusions: Inflammatory-based panels do not predict post transplant HCC recurrence in patients transplanted within Milan criteria. Neo-adjuvant transarterial therapies and total tumour size are the only independenlty associated factors.

Disclosures:

The following people have nothing to disclose: Emmanuel Tsochatzis, Hasitha S. Wijewantha, Manuel Rodríguez-Perálvarez, Pinelopi Manousou, Evangelia Fatourou, Pieri Giulia, Vasilios Papastergiou, Eleni Theocharidou, Douglas Thor-burn, David W. Patch, James O'Beirne, Tim Meyer, Andrew K. Burroughs

2153

  1. Top of page

A product of alpha-fetoprotein and protein induced by vitamin K absence-II as a powerful predictor of the prognosis and recurrence of hepatocellular carcinoma after hepatectomy

Toshiya Kamiyama1, Hideki Yokoo1, Tatsuhiko Kakisaka1, Tatsuya Orimo1, Kenji Wakayama1, Hirofumi Kamachi1, Yosuke Tsuruga1, Kenichiro Yamashita2, Tsuyoshi Shimamura3, Satoru Todo2, Aki-nobu Taketomi1;
1 Gastroentelorogical Surgery I, Hokkaido University, Sapporo, Japan; 2Transplantation Surgery, Hokkaido University, Sapporo, Japan; 3Organ transplantation, Hokkaido University Hospital, Sapporo, Japan

Purposes: To evaluate an oncological implication of a product of alpha-fetoprotein (AFP) and protein induced by vitamin K absence or antagonism factor II (PIVKA-II) in patients with hepatocellular carcinoma (HCC). Methods: Data was prospectively collected from 516 consecutive patients who underwent curative primary hepatectomy for HCC between 1998 and 2010. AP-factor—a product of the serum levels of AFP and PIVKA-II— was evaluated for 2-year survival by ROC analysis to determine the cut-off values about AP-factor. Patient survival (PS), recurrence-free survival (RFS) and risk factors were analyzed according to the preoperative AP-factor. Results: The patients were classified into 3 groups by this cut-off values of AP-factor of 100000 (AUC=0.74607, Sensitivity=63.27%, Speci-ficity=77.41 % ); AP1 (n = 206; AFP < 200 ng/ml and PIVKA-II < 100 mAU/ml), AP2 (n = 152; AFP x PIVKA-II < 100000), and AP3 (n = 158; AFP x PIVKA-II > 100000). Among the factors evaluated during clinicopathological examination, AP-factor was significantly related to pathological factors such as differentiation, microvascular portal vein invasion, microvascu-lar hepatic vein invasion, and microscopic intrahepatic metastasis (p < 0.0001, p < 0.0001, p < 0.0001, and p < 0.0001, respectively). PS of AP1, AP2, and AP3 at 5 years were 82.7%, 78.8%, and 41.3%, respectively. PS of AP1 and AP-2 were significantly higher than those of AP3 (p < 0.0001 and p < 0.0001, respectively). RFS of AP1, AP2, and AP3 at 5 years were 34.0%, 40.7% and 17.1 %, respectively. RFS of AP3 were significantly lower than those of AP1 and AP2 (p < 0.0001 and p < 0.0001, respectively) Including the significant factors by univariate analysis to identify the risk factors for survival and recurrence, multivariate analysis was performed, which showed that albumin (p = 0.0056), AP-factor (p = 0.0062) and pathological factors: portal vein invasion (p = 0.0027), hepatic vein invasion (p = 0.0056), intrahepatic metastasis (p = 0.005) were independent risk factors for survival, and that tumor number (p < 0.0001), AP-factor (p = 0.0161), and non cancerous liver (p = 0.012) were for recurrence. Recurrence in AP1 patients tended to occur only in the liver, whereas that in AP3 patients tended to occur in extrahepatic sites, including or excluding the liver (p < 0.0001). Conclusion: AP-factor was closely related to differentiation, microscopic portal vein and hepatic vein invasion; AP-factor was selected as an independent factor for survival and recurrence by multivariate analysis. Patients hopeful of obtaining good outcomes after hepatectomy could be selected by measuring their AP-factor.

Disclosures:

The following people have nothing to disclose: Toshiya Kamiyama, Hideki Yokoo, Tatsuhiko Kakisaka, Tatsuya Orimo, Kenji Wakayama, Hirofumi Kamachi, Yosuke Tsuruga, Kenichiro Yamashita, Tsuyoshi Shimamura, Satoru Todo, Akinobu Taketomi

2154

  1. Top of page

Incidence of hepatocellular carcinoma in chronic hepatitis B according to risk factors: Meta-analysis with data from 55 trials and 19.550 patients

Maja Thiele1, Lise L. Gluud2, Annette D. Fialla1, Emilie K. Dahl2, Aleksander Krag1;
1Department of Gastroenterology and Hepatol-ogy, Odense University Hospital, Odense C, Denmark; 2Depart-ment of Medicine, Copenhagen University Hospital, Gentofte, Hellerup, Denmark

Introduction Incidence of hepatocellular carcinoma (HCC) in chronic hepatitis B differs according to various risk factors. In order to assess prognosis and indication for screening incidence rates adjusted for risk factors are needed. Methods We performed a systematic review on the incidence of HCC in untreated patients with chronic hepatitis B in randomised controlled trials and observational studies. Electronic and manual searches were combined. HCC person-time incidence rates with standard deviations were calculated from random effects meta-analyses. We performed subgroup analyses on cirrhosis status, gender, HBeAg status and hepatitis C coinfection. Sensitivity analyses included trial design, whether HCC screening was performed, age of included patients, region of trial and proportion of patients with inflammatory activity and viral replication. Heterogeneity was evaluated using Cochrane Q statistics. Meta-regression analyses were performed to test for study-level covariates. Calculations were performed in STATA version 12 (StataCorp, TX, USA). Results A total of 19.550 patients from 55 trials and studies were included (three randomised trials, 26 prospective cohorts, and 26 case control studies on antiviral treatment). A total of 885 HCCs occurred. The overall HCC incidence was 4.3% per person-year. Patients with cirrhosis had a 35-times higher risk of HCC than patients without cirrhosis (incidence 7.1% versus 0.2%). HBeAg-nega-tive patients and patients co-infected with hepatitis C had high risks of HCC (table). In sensitivity analyses study design played an important role in the incidence assessment (randomised trials 2.0%, prospective cohorts 2.8% and case control series 11.7%). No statistical heterogeneity was found. None of the covariates included in the meta-regression analyses had an effect on the incidence estimate. Conclusion Untreated chronic hepatitis B patients without cirrhosis carry a low risk of HCC and the value of surveillance can be questioned. The risk is 35 times higher in patients with cirrhosis.

 Number of trialsIncidence (person-years)±SD
  1. SD, standard deviation

Overall554.3%2.2%
Cirrhosis337.1%3.6'/;
Precirrhosis30.2%0.1%
Male101.1%0.6%
Female40.4%0.2%
HBeAg+42.0%1.0%
HBeAg-716.0%8.2%
Hepatitis C coinfection85.6%2.9%
Randomised trials32.0%1.0%
Prospective cohorts262.8%1.4%
Case control series2611.7%6.0%

Disclosures:

Lise L. Gluud - Independent Contractor: Merck

The following people have nothing to disclose: Maja Thiele, Annette D. Fialla, Emilie K. Dahl, Aleksander Krag

2155

  1. Top of page

Baseline alfa-fetoprotein level is a strong predictor of vascular invasion and survival in patients with hepato-cellular carcinoma

Omer Ekinci1,2, Bulent Baran2, Asli Cifcibasi Ormeci2, Ozlem Mut-luay Soyer2, Suut Gokturk2, Arzu Poyanli3, Mine Gulluoglu4, Cetin Karaca2, Filiz Akyuz2, Kadir Demir2, Fatih Besisik2, Sabahattin Kaymakoglu2;
1Department of Internal Medicine, Istanbul University, Istanbul Faculty of Medicine, Istanbul, Turkey; 2Department of Gastroenterohepatology, Istanbul University, Istanbul Faculty of Medicine, Istanbul, Turkey; 3Department of Radiology, Istanbul University, Istanbul Faculty of Medicine, Istanbul, Turkey; 4Department of Pathology, Istanbul University, Istanbul Faculty of Medicine, Istanbul, Turkey

Aim: To investigate clinical, etiological, and prognostic features of patients with hepatocellular carcinoma (HCC). Methods: Patients with HCC who were followed-up between 2001 and 2011 were included in the study. Diagnosis was established by histopathological and/or radiological criteria. We retrospectively reviewed clinical/laboratory data, etiology of primary liver disease, imaging characteristics and treatments of patients. Barcelona Clinic Liver Cancer (BCLC) stage was determined at initial diagnosis. Predictor variables of vascular invasion and prognostic factors that affect overall survival were investigated by multiple regression analyses. Kaplan-Meier survival analysis was done to find out treatment effect on survival. Results: There were 545 patients with HCC (449 male, mean age 59.5±10) included in the study. Patients with viral hepatitis (454 patients/83.3%) were prevalent and 68 patients (12.5%) were non-cirrhotic. 335 (61.5%) patients had a single nodule and the remaining had multinodular or diffuse HCC. Patients who had regular follow-ups and screening with AFP-ultrasonography were diagnosed at an earlier BCLC stage (p<0.001). Overall median survival was 16 (13–19) months. There were 146 (%26.8) patients with an AFP>400ng/ml. Extrahepatic metastasis and macroscopic vascular invasion were diagnosed in 26 (4.8%) and 37 (6.8%) patients, respectively. Independent predictor variables of vascular invasion were AFP>200ng/ml (OR: 2.97, 95%CI 1.42–6.21, p=0.004), total tumor diameter (TTD) >5cm, (OR: 4.30, 95%CI 1.45–12.75, p=0.008). Baseline AFP level>200 ng/ml (HR: 1.68, 95%CI 1.33–2.12, p<0.001), diffuse-type HCC (HR: 2.10, 95%CI 1.29–3.43, p=0.003) and BCLC stage (stage B, HR: 3.91, 95%CI 2.44–6.29; stage C, HR: 34, 95%CI 20–56; stage D, HR: 108, 95%CI 65–177; p<0.001) were detected as independent baseline variables associated with overall survival. Surgical treatments as resection and transplantation were found to be associated with a better survival according to other loco-regional treatments (log-rank, p<0.001). Conclusion: Baseline AFP level is an important predictor variable for both vascular invasion and overall survival. Surgical resection and transplantation are treatment of choice with best survival rates and a delayed diagnosis of HCC significantly diminishes the chance of survival by making curative treatments unavailable.

Disclosures:

The following people have nothing to disclose: Omer Ekinci, Bulent Baran, Asli Cifcibasi Ormeci, Ozlem Mutluay Soyer, Suut Gokturk, Arzu Poyanli, Mine Gulluoglu, Cetin Karaca, Filiz Akyuz, Kadir Demir, Fatih Besisik, Sabahattin Kaymakoglu

2156

  1. Top of page

Comparison of Clinical Outcomes between the recurrent advanced Hepatocellular Carcinoma patients with Sorafenib treatment and the patients with Sorafenib +Interferon combination therapy

Hisakazu Doi;
Gastroenterology and Hepatology, Nagahama Red Cross Hospital, Nagahama city, Japan

Introduction: Therapeutic options are limited for patients with recurrent advanced HCC which could not meet the criteria of recommended locoregional and systemic therapies and the prognosis of such cases are poor. The aim of this study was to report initial experience with combination therapy of sorafenib and interferon for the patients with recurrent advanced HCC.Methods: All consecutive patients with recurrent advanced HCC unsuitable for locoregional and systemic therapies who treated with sorafenib mono-therapy(400mg, 800mg) or combination therapy of sorafenib(200mg, 400mg) and inter-feron(PEGIFN90μg, IFN alfa 300 MU) between 2009 -2012 were evaluated. We divided the patients into two groups: control group (n=21) received sorafenib only and treatment group (n = 11) received sorafenib and interferon. Survival rate, adverse events and factors related to treatment effect were analyzed retrospectively. Overall survival was calculated using the Kaplan Meier method. Toxicity was graded according to the NGICTC-AEv3.0. Antitumor effect was evaluated by RECISTv1 .1 criteria. Results: Treatment group was 11 patients (78% male, median age 69.8 yrs, Child A/B/C: 5/4/2/, HCV/HBV/others: 8/1/2, previous treatment TACE/RFA/OPE: 8/7/0), and control group was 21 patients (62% male, median age 74.9 yrs, Child A/B/C:13/8/0, HCV/HBV/others : 17/3/2, previous treatment TACE/RFA/OPE: 11/7/1). Median treatment duration was 14.6 months (range, 2.5 to 43) in treatment group and 10.0 months (range, 1 to 28.3) in control group. Tumor control rate (CR, PR or SD) in both groups were similar (treatment group: 8/11 72.3%, and control group 14/2166.7%, P=0.7258). The median overall survival was 32.6 months among the 11 patients in the treatment group and 12.7 months among the 21 patients in the control group (P<0.05). In addition, patients who achieved PR or SD in the treatment group revealed a significant longer survival as com pared with control group who also achieved PR or SD (40 months vs. 17.4 months, (P<0.05). During the treatment periods, adverse events were mild (grade 1/2 diarrhea, HFS, anemia: 60%, 20%, 80%, respectively) in the treatment group. Conclusion: As compared with sorafenib alone, Interferon plus Sorafenib combination therapy was associated with a significant longer survival without the additional toxicity for the patients with recurrent advanced HCC following previous treatment. Larger studies are needed to confirm the role of this treatment.

Disclosures:

The following people have nothing to disclose: Hisakazu Doi

2157

  1. Top of page

Serum amyloid A (SAA)-positive hepatocellular neoplasms and SAA-negative focal nodular hyperplasia-like nodules arising in alcoholic cirrhosis

Motoko Sasaki1, Norihide Yoneda2, Yasunori Sato1, Yasuni Nakanuma1;
1 Human Pathology, Kanazawa University Graduate School of Medicine, Kanazawa, Japan; 2Rasiology, Kanazawa University Graduate School of Medicine, Kanazawa, Japan

Background/Aims: It is well known that hypervascular hepatocellular nodular lesions resembling hepatocellular carcinoma (HCC) are sometimes detected by imaging modality in patients with alcoholic cirrhosis. Hepatocellular adenoma (HCA) usually arises in the absence of significant fibrosis. However, we have recently reported a unique type of hypervascular hepatocellular lesion, SAA-positive hepatocellular neoplasm (SAA-HN), in alcoholic cirrhosis that shares histological features and an immunoreactivity for SAA with inflammatory HCA. Herein, we characterized 50 hypervascular hepatocellular nodules in 20 patients (3 women and 17 men, age ranged 40–67 yrs, mean±SD, 51.5± 8.1 yrs) with alcoholic cirrhosis. Methods: Twenty patients with hypervascular hepatocellular nodules associated with alcoholic cirrhosis were retrieved from our pathological files (1997–2013). The hepatocellular nodules were multiple (>3) in 16 patients. The immunoreactivity for SAA, glu-tamine synthetase (GS) and glypican-3 was used for characterization of nodules. Furthermore, sixteen hypervascular hepatocellular nodules were examined on the magnetic resonance (MR) imaging with gadolinium ethoxybenzyl diethylene-triaminepentaacetic acid (Gd-EOB) enhancement. Results: Thirty-three nodules (4–35mm in diameter, mean±SD, 13.0± 7.4mm) in 15 patients (3 women and 12 men, age ranged 40–67 yrs, 49.6±8.0yrs) were diagnosed as SAA-HNs. The remaining seventeen nodules (5–25mm, mean±SD, 10.2± 4.2mm) in 8 patients (2 women and 6 men, age ranged 41–62 yrs, 53.1±8.4yrs) were focal nodular hyperplasia (FNH)-like nodules. The FNH-like nodules showed no immunoreactivity for SAA and showed a map-like staining pattern of GS, similarly to FNH. Three patients had both SAA-HNs and FNH-like nodules. SAA-HNs showed significantly more extensive sinusoidal dilatation, inflammatory reaction, abnormal thick arteries and cellular atypia than SAA-negative FNH-like nodules (p<0.05). SAA-HNs and FNH-like nodules did not show an overexpres-sion of GS and the immunoreactivity for glypican-3. On the MRI with Gd-EOB enhancement, eight of 12 SAA-HNs (67%) showed slightly hypointensity in the hepatobiliary phase similarly to HCC. In contrast, all 4 FNH-like nodules showed iso-intensity in the hepatobiliary phase and there was significant difference between these 2 lesions on the MRI with Gd-EOB enhancement (p<0.05). Conclusions. This study further clarified characteristics of SAA-HNs arising in alcoholic cirrhosis. SAA-HNs sometimes co-exist with FNH-like nodules and may show different findings on Gd-EOB enhanced MR imaging.

Disclosures:

The following people have nothing to disclose: Motoko Sasaki, Norihide Yoneda, Yasunori Sato, Yasuni Nakanuma

2158

  1. Top of page

An isoform of des-gamma-carboxy prothrombin (NX-PVKA) is a prognostic marker of hepatocellular carcinoma

Satoru Takeji, Masashi Hirooka, Yohei Koizumi, Hironori Ochi, Yoshio Tokumoto, Fujimasa Tada, Masanori Abe, Morikazu Onji, Yoichi Hiasa;
Department of Gastroenterology and Metabology, Ehime University Graduate School of Medicine, Toon, Japan

Background and aims: Serum des-gamma-carboxy prothrombin (DCP) is an established tumor marker of hepatocellular carcinoma (HCC) that can be identified using MU-3 antibody, which mainly reacts with the 9–10 glutamic acid (Glu) residues of DCP (conventional DCP). Other variants of DCP with fewer Glu residues, such as NX-PVKA, can be detected using P-11 and P-16 antibodies. We examined the clinical characteristics associated with NX-PVKA, and whether NX-PVKA could serve as a useful marker in patients with HCC and with liver cirrhosis but without HCC. Methods: This study investigated 197 and 102 patients with and without HCC, respectively, who were admitted to our hospital between 2001 and 2010. No patients had been administered with warfarin. Levels of NX-PVKA, conventional DCP, NX-PVKA-R (Conventional DCP/NX-PVKA), alpha-fetoprotein (AFP), L3 fraction of AFP (AFP-L3) and various clinical parameters were measured before starting treatment for HCC, as well as during the follow-up period for the patients without HCC. Moreover, significant factors associated with the survival of patients with HCC among these tumor markers, as well as other clinical and tumor factors, were investigated. Results: Analysis of receiver operating characteristic (ROC) curves showed that the NX-PVKA cut-off required to differentiate HCC is 24 mAU/mL, and that the sensitivity and specificity were 0.787 and 0.559, respectively. In patients without HCC, NX-PVKA was significantly associated with Child-Pugh classification (p = 0.0012). Significant associations were also observed with Conventional DCP (p = 0.0365) and NX-PVKA-R (p = 0.0032), but not with AFP and AFP-L3. NX-PVKA was the most significant predictor of HCC prognosis (hazard ratio = 81.32, p < 0.0001). The level of NX-PVKA was significantly associated with platelet count, prothrombin time, C-reactive protein, sex, maximum tumor size, number of nodules and portal venous invasion by HCC. A prognostic risk-of-mortality (ROM) score with which to estimate the overall survival of patients with HCC based on the NX-PVKA level and other clinical parameters was also established. The correlation coefficient between the ROM score and the actual overall survival of the patients with HCC was r = 0.686, which was better than that of the CLIP (r = 0.537) and Tokyo (r = 0.530) scores. Conclusion: NX-PVKA is associated with tumor factors and hepatic functional reserve and represents a useful tumor marker for evaluating the prognosis of patients with HCC. Therefore, NX-PVKA could serve as a clinical marker for evaluating tumor status and estimating the survival of patients with HCC.

Disclosures:

The following people have nothing to disclose: Satoru Takeji, Masashi Hirooka, Yohei Koizumi, Hironori Ochi, Yoshio Tokumoto, Fujimasa Tada, Masanori Abe, Morikazu Onji, Yoichi Hiasa

2159

  1. Top of page

The Effect of HBV DNA level on The Development of Hepatocelluar Carcinoma in Patients with Hepatitis B Virus-associated Liver Cirrhosis during Entecavir Treatment

Oh Sang Kwon, Hyeonsu Park, Jong Joon Lee, Young Kul Jung, Duck Joo Choi, Yun Soo Kim, Ju Hyun Kim;
Internal Medicine, Gachon University Gil Medical Center, Inchon, Republic of Korea

Background and aims: One of primary goals of oral antiviral agents is the suppression of hepatocellular carcinoma (HCC) development in patients with chronic hepatitis B. This study aimed to verify the effect of hepatitis B virus (HBV) DNA level on the development of HCC in patients with liver cirrhosis (LC) during entecavir (ETV) treatment. Methods: A total 268 naïve patients with HBV-associated LC was treated by ETV for at least 6 months, between March 2007 and December 2012. All patients had HBV DNA level over than 4 log 10 copies/mL and ALT level over than 40 IU/mL, because of regulation of Korea national health insurance. Exclusion criteria were 1) the ETV was skipped more than 3 months and 2) HCC was developed within 6 months after ETV treatment. Standard screening tests using alpha-fetoprotein and image tools were done 4–6 months intervals for the detection of HCC. HBV DNA levels (by real time polymerase chain reaction) were also checked 6 months interval. The cumulative incidence (by Kaplan-Meier method) and the risk factors of HCC (by Cox-regression analysis) were investigated. Results: The final 198 patients were enrolled. The mean age was 52±9 years old and 63.6% of patients was male. The mean follow up period was 31 ±16 (7–67) months. HCC was developed in 29 patients. The cumulative incidence of HCC was 5% in 1 year, 11% in 2 years, 13% in 3 years, 24% in 4 years, and 29% in 5 year. In the univariate analysis, the last peak HBV DNA that was defined as the highest HBV DNA level during 1 year just before HCC diagnosis or last visiting day in patients without HCC [hazard ratio (HR): 1.338, 95% CI: 1.165–1.536, p<0.001] was the only risk factor of HCC. HBV DNA levels at initial and last visit (in HCC cases, the level just before detection of HCC) during ETV treatment were not the risk factor. In addition, age, sex, ALT, HBeAg positivity, Child-Pugh class, the model of end stage liver disease score, smoking and alcohol drinking were not the risk factor. Although age, sex, and initial HBV DNA level were not the risk factor, those variables were included in the multivariate analysis. Finally, the age (HR: 1.042, 95% CI: 1.000–1.084, p=0.048) and the last peak HBV DNA (HR: 1.383, 95% CI: 1.188–1.611, p<0.001) were the only risk factors of HCC. The mean of the last peak HBV DNA level was 1.16 (0–8.4) in no HCC group and 2.02 (0–7.4) log 10 copies/mL in HCC group. Conclusions: The last peak HBV DNA is a significant risk factor of HCC during ETV treatment in HBV-associated LC. As fast as we can, HBV DNA should be suppressed for the prevention of HCC development.

Disclosures:

The following people have nothing to disclose: Oh Sang Kwon, Hyeonsu Park, Jong Joon Lee, Young Kul Jung, Duck Joo Choi, Yun Soo Kim, Ju Hyun Kim

2160

  1. Top of page

Hypovascular hepatocellular carcinoma is associated with more advanced disease than classically enhancing hepatocellular cancer, a case series

David N. Fitch1, Kumar Sandrasagaran2, Marwan Ghabril1, Paul Y. Kwo1;
1Division of Gastroenterology, Indiana University Medical Center, Indianapolis, IN; 2Indiana University School of Medicine, Indianapolis, IN

The diagnosis of HCC is typically based on radiographic findings found on CT and MRI that utilize the classic enhancement and washout properties. However, there are limited data on the role of dynamic CT and MRI in the diagnosis of hypovascular HCC. Our AIM was to determine what clinical differences exist at the time of presentation between populations with classically enhancing HCC and those without classic enhancement (hypovascular). METHODS: Single center retrospective analysis from 2002–2012 of all HCC patients. All patients diagnosed with HCC without classic enhancement were compared with two size-matched HCC patients with classic enhancement. Data abstracted included date of presentation, diagnosis, transplantation, number of imaging studies required for diagnosis, imaging modality used, size of lesion at presentation, use of biopsy, MELD score, Barcelona Clinic Liver Cancer staging system (BCLC), Child-Pugh score, AFP and etiology of liver disease. RESULTS: 24 patients with hypovascular HCC met our criteria and were compared to a size-match control group of 48 patients with HCC with classic enhancement . Results are show in the table below. Conclusions: In those who present with hypovascular HCC, we observed a greater requirement for liver biopsy, a longer duration to diagnose in those followed for greater than 1 year, as well as, higher AFP levels and more advanced disease. Better strategies are needed to identify patients with hypovascular HCC.

 Hypervascular (n=48)Hypovascular (n=24)P value
Diagnosed by biopsy95%95%0.001]
Presentation to diagnosis interval > 12 mentwere2%17%0.02
Child-Pugh Class ABC44% 48% 8%17% 37% 46%0.001
Barcelona Clinic Liver Cancer 0 A B C D8% 31% 23% 29% 8%0% 21% 4% 25% 50%0.001
AFP > 1000 at presentation6%26%0.03
Received liver transplant48%29%0.013

Disclosures:

Kumar Sandrasagaran -Consulting: Religen Corporation; Grant/Research Support: Siemens Medical Solutions, Religen Corporation

Marwan Ghabril - Grant/Research Support: Salix

Paul Y. Kwo - Advisory Committees or Review Panels: Abbott, Anadys, Novartis, Merck, Gilead, BMS, Janssen; Consulting: Vertex; Grant/Research Support: Roche, Vertex, GlaxoSmithKline, Merck, BMS, Abbott, Idenix, Vital Therapeutics, Gilead, Vertex, Merck, Idenix; Speaking and Teaching: Merck, Merck

The following people have nothing to disclose: David N. Fitch

2161

  1. Top of page

Validation of sub-staging classification of patients with intermediate Hepatocellular Carcinoma (BCLC-B) treated with conventional Transarterial Chemoembolization

Marco Biolato, Andrea Zanché, Vittoria Vero, Simona Racco, Brigida E. Annicchiarico, Massimo Siciliano, Maurizio Pompili, Gian Ludovico Rapaccini, Antonio Gasbarrini, Antonio Grieco;
Internal Medicine, Catholic University of Rome, Rome, Italy

Background and Aims: The intermediate stage of Barcellona Clinic Liver Cancer (BCLC) staging system for hepatocellular carcinoma (HCC) encloses a prognostically heterogeneous population. A sub-staging of BCLC-B has been recently proposed by Bolondi et al. (table), in order to refine prognosis and to tailor treatment allocation. Aim of this study is to validate prognostic capacity of sub-staging proposal on italian single-center database of HCC patients treated with conventional chemoembolization (TACE). Methods: 128 patients affected by HCC on BCLC-B stage and treated with conventional TACE in the period 1997–2008 at our institution were divided in four subgroups (B1-B4) according to the proposal sub-classification. Follow-up was updated to September 2012 (86% died). The survival of each group was assessed and compared using Kaplan-Meier method and log-rank test. Results: Overall survival of whole population was 31.5 months (95% C.I. 25.9–37.0). Number of patients in BCLC subgroups was B1 = 27, B2 = 69, B3 = 15, B4 = 17. Each stage appeared associated with different median overall survival (p < 0.05 between groups), namely B1 = 42.3 months (95% C.I. 30.6–53.8), B2 = 25.6 months (95% C.I. 21.6–29.6), B3 = 26.6 months (95% C.I. 8.7–44.5), B4 = 30.2 months (95% C.I. 19.8–40.7). The 3-years survival were: B1 = 44.2%; B2 = 22.9%; B3 = 15.2%; B4 = 35.3% (p<0.05). Conclusions: The new substaging proposal is able to refine prognosis of intermediate patients with HCC treated with conventional TACE. The prognosis of patients in B3 seems to depend mainly on the tumor while that of patients in B4 on the underlying cirrhosis, so further studies are needed to confirm the actual prognostic gradient of these two substages.

 BlB2B3B4
Child-Pugh score5–6-75–678–9*
Up-to-7 criterionINOUTOUTANY
Performance status0000–1
Portal vein thrombosisNONONONO
1st optionTACETACE or TARE Best supportive care
AlternativeOLT, TACE+ablationSorafenibResearch trials, TACE, sorafenib0LT**

*with severe/refractory ascites and/or jaundice; ** only if Up-to-7 IN and PS 0; OLT=ortotopic liver transplantation, TACE=transar-terial chemoembolization, TARE= transarterial radioembolization.

Bolondi L et al. Heterogeneity of Patients with Intermediate (BCLC B) Hepatocellular Carcinoma: Proposal for a Subclassification to Facilitate Treatment Decisions. Sem Liver Dis 2012;32:348–359.

Disclosures:

The following people have nothing to disclose: Marco Biolato, Andrea Zanché, Vittoria Vero, Simona Racco, Brigida E. Annicchiarico, Massimo Siciliano, Mau-rizio Pompili, Gian Ludovico Rapaccini, Antonio Gasbarrini, Antonio Grieco

2162

  1. Top of page

Utility of percutaneous radiofrequency thermal ablation therapy combined with transarterial chemoembolization for hepatocellular carcinoma

Kaori Ono1, Tsutomu Tamai1, Eriko Toyokura1, Takeshi Hori2, Kazuhiro Sakurai2, Kazuaki Tabu1, Akihiko Oshige1, Kohei Oda1, Dai Imanaka1, Kotaro Kumagai1, Seiichi Mawatari1, Akihiro Mori-uchi1, Hirofumi Uto1, Makoto Oketani1, Akio Ido1, Hirohito Tsub-ouchi3;
1Digestive Disease and Life-style Related Disease, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan; 2Hepatology, Kagoshima Teishin Hospital, Kagoshima, Japan; 3Department of HGF Tissue Repair and Regenerative Medicine, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan

[Objective] Percutaneous radiofrequency thermal ablation (RFA) therapy for hepatocellular carcinoma (HCC) has been established as a modality that establishes excellent local control is already standardized as a high cure for local control ability. Although RFA has been combined with transarterial chemoembolization (TACE) at our institutions for the purpose of raising the local control rate, the efficacy of this approach has not examined until now. We examined the utility of RFA therapy combined with TACE for HCC. [Methods] In this retrospective review, 148 HCC tumors in 103 patients treated with RFA from April 2007 to June 2012 at two medical institutions were examined. The study cohort had a median age of 70 years (range, 45–91), and consisted of 59.2% males and 40.8% females. There were 8, 82, and 13 patients with hepatitis B, hepatitis C, and non-HBV and non-HCV hepatitis, respectively. The Child-Pugh score (CP) was A in 86 patients and B or C in 17 patients. There were 49 hypovascular and 99 hypervascular HCCs. The median diameter was 15 mm (range, 6–44). There were 76 single tumors and 72 multiple tumors, and 64 first time tumors and 84 recurrences. RFA was combined with TACE to treat 83 HCCs. [Results] The overall survival rate was 98.9% at 1 year, 81.3% at 3 years, and 64.2% at 5 years. The overall local recurrence rate, include both RFA alone and RFA plus TACE, was 18.2% at 1 year, 42.3% at 3 years, and 47.5% at 5 years. Local recurrence of hypervascular HCCs (n=99) tended to be controlled by RFA plus TACE, and this was associated with tumor diameter <20 mm (P<0.01), des-γ-carboxy pro-thrombin (DCP) <35 mAU/ml (P=0.01), and ablated margin >5 mm (P=0.02). Multivariate analysis showed that treatment with RFA plus TACE (P=0.037; Hazard Ratio (HR), 2.426) and ablated margin >5 mm (P=0.025; HR, 2.332) significantly contributed to the control of local recurrences for hypervascular HCCs. On the other hand, RFA plus TACE controlled intrahep-atic distant recurrence significantly (P=0.03), and the cumulative survival rate was also good (P=0.04). RFA plus TACE (P<0.01) and CP score of A (P<0.01) were associated with a high cumulative survival rate for single, hypervascular HCCs (n=46). Multivariate analysis identified RFA plus TACE (HR, 17.245; P=0.02) and CP score of A (HR 19.256; P=0.01) were independent prognostic indicators. [Conclusion] RFA combined with TACE controls local recurrences and the intrahepatic distant recurrence. The data suggests it is possible to improve the cumulative survival rate by combining RFA and TACE.

Disclosures:

Makoto Oketani - Grant/Research Support: Bristol-Meyers Squibb

Hirohito Tsubouchi -Grant/Research Support: MSD, Chugai Pharmaceutical, Kan Research Institute, Daiichi-Sankyo, Eisai, Tanabe Mitsubishi; Speaking and Teaching: MSD, Tanabe Mitsubishi

The following people have nothing to disclose: Kaori Ono, Tsutomu Tamai, Eriko Toyokura, Takeshi Hori, Kazuhiro Sakurai, Kazuaki Tabu, Akihiko Oshige, Kohei Oda, Dai Imanaka, Kotaro Kumagai, Seiichi Mawatari, Akihiro Moriuchi, Hirofumi Uto, Akio Ido

2163

  1. Top of page

Accuracy of [18F]fluorocholine (18F-FCH) PET/CT for the characterization of de novo liver nodules detected in cir-rhotic patients: a prospective pilot study

Matteo A. Manini1, Virgilio Longari2, Eleonora Grassi1, Luigia Flo-rimonte2, Massimo Iavarone1, Angelo Sangiovanni1, Davide Viganò1, Paolo Gerundini2, Massimo Colombo1;
1 1st Division of Gastroenterology 1, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, University of Milan, Milano, Italy; 2UO Nuclear Medicine, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, University of Milan, Milano, Italy

Background and aim: The updated recommendations of AASLD for the diagnosis of the hepatocellular carcinoma (HCC) in patients with cirrhosis enable to increase the number of radio-logically diagnosed HCC. Nevertheless, the need of histological confirmation and whole-body staging remains. 18Fluoro-labelled choline-based tracers (18F-FCH) have been recently used as oncological PET probes to detect increased lipogenesis in cancer cells. Therefore, we prospectively assessed the accuracy of 18F-FCH PET/CT for the diagnosis and staging of > 1cm HCC. Methods: All Child-Pugh A or B cirrhotic patients with a de novo liver nodule detected during ultrasounds (US) surveillance were consecutively included. Computed Tomography scan (CT), Magnetic Resonance Imaging (MRI) and fine-needle biopsy were sequentially performed to characterize the nodule, using the diagnostic gold standard of 2010 AASLD criteria. PET/CT was performed at the end of radiological investigation. Whole-body PET/CT images were acquired 15, 30 and 60 minutes after a tailored 18F-FCH ev administration. A true positive result was an increased uptake of 18F-FCH at a qualitative evaluation by two expert radiologists. The liver positive findings on PET/CT were subsequently compared with the nodule location on conventional imaging. Results: Between October 2011 and April 2012, 17 patients, 11 (65%) males, 66 yr (43–83), 15 (88%) anti-HCV, with 20 liver nodules detected by US, mean diameter 1.7 (1.0–3.4) cm, 14 (70%) single, 16 (80%) 1–2 cm, 4 (20%) >2 cm, were enrolled. CT and/or MRI led to discover 4 additional nodules missed by US (mean 1.3 cm; range 1.0–2.4 cm) in 3 (18%) patients. The diagnosis of the 24 liver nodules was: 15 (63%) HCC, 13 (87%) by radiological criteria and 2 (13%) by histology; 2 (8%) low grade dysplastic nodules; 7 (29%) macroregenerative nodules. 18F-FCH PET/CT showed a sensitivity, specificity, PPV, NPV and accuracy for the diagnosis of HCC of 53%, 100%, 100%, 56% and 71%, respectively. All the positive PET/CTs showed an increased uptake of the nodule 15 minutes after injection, while 30 and 60 minute acquisitions had a sensitivity of 33% and 27%, only. One well-differentiated HCC undiag-nosed by radiology showed a positive 18F-FCH PET/CT. No meaningful localization of 18F-FCH was detected in other tissues. Conclusion: 18F-FCH PET/CT is a promising technique for the detection of HCC, but its accuracy still needs to be evaluated in larger series. 15 minutes imagine acquisitions allowed to identify correctly all the HCCs testing positive at 30 and 60 minutes. Therefore, PET/CT images could be acquired early after 18F-FCH administration, making unnecessary subsequent acquisitions

Disclosures:

Massimo Colombo - Advisory Committees or Review Panels: BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, Janssen Cilag, Achillion; Grant/Research Support: BRISTOL-MEYERS-SQUIBB, ROCHE, GILEAD, BRISTOL-MEYERS-SQUIBB, ROCHE, GILEAD; Speaking and Teaching: Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX, Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX

The following people have nothing to disclose: Matteo A. Manini, Virgilio Lon-gari, Eleonora Grassi, Luigia Florimonte, Massimo Iavarone, Angelo Sangio-vanni, Davide Viganò, Paolo Gerundini

2164

  1. Top of page

Serum ferritin level as a risk factor for hepatocellular carcinoma recurrence after radiofrequency ablation

Koji Uchino1, Hayato Nakagawa1, Ryosuke Tateishi1, Mayuko Kondo1, Naoto Fujiwara1, Tatsuya Minami1, Masaya Sato1, Kenichiro Enooku1, Yuji Kondo1, Yoshinari Asaoka1, Haruhiko Yoshida1, Shuichiro Shiina2, Kazuhiko Koike1;
1 Department of Gas-troenterology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; 2Department of Gastroenterology, Juntendo University, Tokyo, Japan

Backgrounds and Aims: Hepatic iron accumulation is considered to be a cofactor that influences liver injury and hepatocarcinogenesis. The aim of this study was to determine whether serum ferritin level is a risk factor for hepatocellular carcinoma (HCC) recurrence after radiofrequency ablation (RFA) with curative intent. Methods: We prospectively measured serum ferritin levels in 580 primary HCC patients (mean age=69.1 ±9.5 yrs, M/F=371/209) who underwent RFA between 2006 and 2010. They were divided into 4 groups according to serum ferritin levels: G1 (<50 ng/mL, N=137), G2 (51–100 ng/mL, N = 111),G3 (101–200 ng/mL, N=150) and G4 (≧201 ng/mL, N = 182). Correlation between serum ferritin and other clinical factors were analyzed. We assessed the impact of serum ferritin level on HCC recurrence after RFA among patients whose tumor was solitary and smaller than 2 cm. Results: Serum ferritin correlated with AST (rho=0.31, p<0.001), ALT (r=0.41, p<0.001), platelet count (rho=-0.11, p=0.006), total bilirubin (rho=0.17, p<0.001), hemoglobin (rho=0.35, p<0.001), tumor number (rho=0.10, p=0.01) and alpha-fetoprotein (rho=0.10, p=0.01). During the mean follow-up period of 1.9 years, recurrence developed in 416 patients. The primary tumor was solitary and smaller than 2 cm in 154 of 580 patients. Among these 154 patients, the median time to recurrence was significantly longer in G2 than in other groups (5.7 vs 2.1 years, p=0.01). In multivariable analysis, serum ferritin level group other than G2 was selected as a risk factor of recurrence (Hazard ratio 0.50; 95%CI 0.27–0.91; p=0.02), together with male sex, prothrombin time, AST, tumor size, and des-γ-carboxy-prothrombin. Conclusion: Serum ferritin level, possibly reflecting the degree of hepatic inflammation and fibrosis, is a negative risk factor for recurrence after RFA among patients with solitary small HCC, when in the range of 51 to 100 ng/mL.

Disclosures:

Ryosuke Tateishi - Grant/Research Support: Eisai Co. Ltd.

Kazuhiko Koike - Speaking and Teaching: Bristol-Myers Squibb

The following people have nothing to disclose: Koji Uchino, Hayato Nakagawa, Mayuko Kondo, Naoto Fujiwara, Tatsuya Minami, Masaya Sato, Kenichiro Enooku, Yuji Kondo, Yoshinari Asaoka, Haruhiko Yoshida, Shuichiro Shiina

2165

  1. Top of page

Transarterial chemoembolization for hepatocellular carcinoma in liver transplant candidates with and without marginal hepatic reserve

David N. Tran1, Nicholas Fidelman1, Elisabeth Garwood1, Francis Y. Yao2, Robert Kerlan1;
1 Department of Radiology, University of California San Francisco, San Francisco, CA; 2Division of Gastroenterology and Hepatology, University of California San Francisco, San Francisco, CA

PURPOSE. To determine the prevalence of serious adverse events in liver transplant candidates with and without marginal hepatic reserve who underwent transarterial chemoembolization (TACE) for treatment of hepatocellular carcinoma (HCC). MATERIALS AND METHODS. 205 consecutive adult liver transplant candidates with HCC diagnosed according to AASLD guidelines underwent 351 segmental or subsegmental TACE procedures from 2005–2009 at our institution. Patients were subdivided into "low" and "high" risk groups for development of hepatotoxicity based on pre-TACE liver function. Patients in high risk group had at least one of the following criteria: pre-TACE bilirubin >2mg/dl, international normalized ratio (INR) >1.5, creatinine >1.2 mg/dL, platelet count <60,000/mL, Model for End-Stage Liver Disease (MELD) score >15, Child-Pugh class B or C, ascites, or portal venous thrombus. Hepatotoxicity was defined as new or worsening ascites, encephalopathy, or NCI Common Terminology Criteria for Adverse Events grade 3 or 4 toxicity of bilirubin, AST, ALT, creatinine or INR. Rate and risk factors for death or urgent liver transplantation within 6 weeks of TACE and irreversible hepa-totoxicity were determined using generalized estimating equation analysis. RESULTS. 133 patients who had at least one "high-risk" factor underwent 236 TACE procedures, while 72 patients who were considered "low-risk" underwent 115 TACEs. Irreversible hepatotoxicity developed as a result of 25 procedures (10.5%), and resulted in death or urgent liver transplant in 7 patients (5.2%). Conversely, 1 procedure (0.9%) in the "low-risk" group resulted in irreversible hepatotoxicity. Strongest predictors for irreversible hepatotoxicity were BCLC stage D HCC, INR > 1.6, presence of ascites on imaging, Child-Pugh class C cirrhosis, bilirubin >3mg/dL, albumin <2.6g/L, and AFP>1000ng/ml. Predictors for TACE-related death or urgent liver transplantation were BCLC stage D HCC and presence of ascites on imaging. CONCLUSION. TACE can be performed safely in liver transplant candidates with baseline hepatic dysfunction. However, poor hepatic reserve increases the risk of irreversible hepatotoxicity, which may lead to death or require urgent liver transplantation.

Disclosures:

The following people have nothing to disclose: David N. Tran, Nicholas Fidel-man, Elisabeth Garwood, Francis Y. Yao, Robert Kerlan

2166

  1. Top of page

Revisit of Alpha-fetoprotein as a Tumor Marker for Hepatocellular Carcinoma in Chronic Hepatitis B Patients Receiving Entecavir

Henry Lik-Yuen Chan, Grace LH Wong, Angeline OS Lo, Vincent W. Wong;
Department of Medicine & Therapeutics, Chinese University of Hong Kong, Hong Kong, Hong Kong

Background: The role of alpha-fetoprotein (AFP) as a biomarker for hepatocellular carcinoma (HCC) surveillance has been controversial. One reason for falsely elevated AFP is active hepatitis, which will reduce the specificity and impair the performance of AFP. As most high risk patients are treated by antiviral agents nowadays, one should expect the biochemical activities are well suppressed. In this study, we aimed to evaluate the role of AFP as a HCC surveillance tool among patients on entecavir. Patients and Methods: This was a retrospective-prospective cohort study of consecutive entecavir-treated patients in the outpatient clinic. All patients received entecavir 0.5 mg daily for at least 12 months. Regular HCC surveillance was performed with AFP and ultrasonography. All HCC cases diagnosed after at least 12 months of entecavir therapy were included. AFP at month -12, -9, -6, -3, and 0 (time of HCC diagnosis) from HCC cases and at corresponding time points from non-HCC cases were analyzed. Results: 1,531 patients were studied with a follow-up of 51 ±13 months. The mean age was 52±12 years; 1,099 (72%) patients were male and 332 (21.7%) patients had clinical evidence of cirrhosis. 57 (2.9%) patients developed HCC at 24 (range 12, 63) months of entecavir treatment. The median size of HCC was 3.3 (interquartile range 2.8, 4.2) cm. Alanine aminotransferase levels of patients were normalized after 6 months of entecavir. AFP started to increase 6 months before the diagnosis of HCC. The area under receiver operator characteristic curve of AFP was highest at month 0 (0.85, 95% confidence interval [CI]: 0.73–0.98), and remained satisfactory at month -3 (0.82, 95% CI: 0.79–0.91) and month -6 (0.79, 95% CI: 0.69–0.79). Using the conventional AFP cutoff of 20 mcg/l at month 0, the sensitivity and specificity to diagnose HCC were 38.6% and 98.9%, respectively. Adopting a lower AFP cutoff value of 6 mcg/l at month 0, the sensitivity was increased to 80.7%, while the specificity was decreased to 80.4%. At 6 months before the diagnosis of

HCC (month -6), the sensitivity and specificity of AFP for HCC were 75.4% and 70.7% at the cutoff of 6 mcg/l whereas the specificity remained very high (97.7%) at the AFP cutoff of 20 mcg/l. Conclusions: AFP is a specific tumor marker for HCC in chronic hepatitis B patients receiving entecavir therapy. Adopting a lower cutoff value of AFP level at 6 mcg/l would increase the sensitivity for HCC detection at a very early stage.

Disclosures:

Henry Lik-Yuen Chan - Advisory Committees or Review Panels: Gilead, Vertex, Bristol-Myers Squibb, Abbott, Novartis Pharmaceutical, Roche, MSD

Grace LH Wong - Advisory Committees or Review Panels: Otsuka, Gilead; Speaking and Teaching: Echosens, Furui

Vincent W. Wong - Advisory Committees or Review Panels: Otsuka, Roche Pharmaceuticals, Gilead, Abbott; Speaking and Teaching: Bristol-Myers Squibb, Novartis Pharmaceuticals, Echosens

The following people have nothing to disclose: Angeline OS Lo

2167

  1. Top of page

Efficacy and safety of sorafenib in patients aged 80 years and older with advanced hepatocellular carcinoma

Kohichiroh Yasui, Masayasu Jo, Takeshi Nishimura, Kanji Yam-aguchi, Yoshio Sumida, Hironori Mitsuyoshi, Masahito Minami, Yoshito Itoh;
Department of Molecular Gastroenterology and Hepa-tology, Kyoto Prefectural University of Medicine, Kyoto, Japan

Aim: Although sorafenib is the only approved systemic therapy for patients with advanced hepatocellular carcinoma (HCC), its efficacy and safety for elderly patients are not established. We aimed to assess the efficacy and safety of sorafenib therapy in patients aged 80 years and older with advanced HCC. Methods: This retrospective study included 185 patients with advanced HCC who received sorafenib therapy. The median age was 71 years (range, 22–91 years). Twenty-four patients (13%) were 80 years and older and 161 (87%) were less than 80 years old. There were 152 male patients (82%) and 33 female patients (18%). Chronic hepatitis C virus infection was the predominant cause of liver disease (60%). One hundred seventy-five patients (95%) had Child-Pugh class A and 10 (5%) had class B. The ECOG performance status was 0 for 128 (69%) patients and 1 or 2 for 67 (31%). Forty-seven patients (25%) had macroscopic vascular invasion and 73 (39%) had extrahepatic spread. Tumor response was assessed using the modified Response Evaluation Criteria in Solid Tumors (mRE-CIST). The primary outcome of this study was overall survival. Data were compared between patients aged 80 years and older and those aged less than 80 years. This study was approved by the ethics committees of each center, and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from each patient. Results: Median overall survival was 10.6 months in all patients. It was 11.7 months in patients aged 80 years and older and 10.5 months in those aged less than 80 years. Multivariate analysis revealed that overall survival was significantly associated with ECOG performance status, presence or absence of macroscopic vascular invasion and extrahepatic spread, baseline α-fetoprotein levels, and best tumor response. Three patients had a complete response, 22 had a partial response, 37 had a stable disease, and 73 had a progressive disease, and 50 were not evaluated for response. Major drug-related adverse events were hand-foot-skin reaction, fatigue, hypertension, diarrhea, and liver dysfunction. There were no significant differences in overall survival, tumor response, and frequency and severity of adverse events between patients aged 80 years and older and those aged less than 80 years. Although the mean dose of sorafenib was significantly lower in patients aged 80 years and older than those aged less than 80 years, the duration of the treatment and the discontinuation rate were not different between the two groups. Conclusions: The efficacy and safety of sorafenib were almost similar between patients aged 80 years and older and those aged less than 80 years.

Disclosures:

Kohichiroh Yasui - Grant/Research Support: AstraZeneca K.K., CHUGAI Pharmaceutical Co., Ltd., Dainippon Sumitomo Pharma Co., Ltd., Eisai Co., Ltd., FUJI-FILM Medical Co., Ltd., Merck Serono, MSD K.K., Otsuka Pharmaceutical Co., Ltd.

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

The following people have nothing to disclose: Masayasu Jo, Takeshi Nishimura, Kanji Yamaguchi, Yoshio Sumida, Hironori Mitsuyoshi, Masahito Minami

2168

  1. Top of page

Predisposing factors for recurrence of HBV-related small hepatocellular carcinoma after percutaneous radiofre-quency ablation

Won Sohn1, Yong-Han Paik1, Min Woo Lee2, Hyunchul Rhim2, Hyo Keun Lim2, Ju Yeon Cho1, Geum-Youn Gwak1, Moon Seok Choi1, Joon Hyeok Lee1, Kwang Cheol Koh1, Seung Woon Paik1, Byung Chul Yoo1;
1 Internal Medicine, Samsung Medical Center, Seoul, Republic of Korea; 2Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea

Background: Radiofrequency ablation as a curative therapy for hepatocellular carcinoma (HCC) is widely used. The aim of this study was to investigate predisposing factors for recurrence of HBV-related small HCC after radiofrequency ablation. Methods: A total of 173 patients underwent percutaneous radiofrequency ablation (RFA) for small of HBV-related HCC (<3cm in diameter) from January 2008 to December 2010 at Samsung Medical Center, Seoul, Korea. We analyzed the risk factors for recurrence of HCC after RFA. Results: The median follow-up duration was 27.0 months. A total of 92 patients (53%) were recurred after percutaneous RFA. Cumulative recurrence-free rates after RFA at 1 -, 3-, and 5-years were 81.1%, 46.4%, and 35.1%, respectively. A univariate analysis showed that predisposing factors for HCC recurrence were the multi-nodularity (hazard ratio (HR) 2.23, p=0.003), pre-procedural HBV DNA levels >2,000 IU/mL (HR 1.60, p=0.025), the presence of hepatitis B envelope antigen (HBeAg) (HR 1.56, p=0.037), Barcelona Clinic Liver Cancer (BCLC) stage A (HR 1.56, p=0.038). The independent risk factors for recurrence by mul-tivariate analysis were the multi-nodularity (HR 1.95, p=0.022), pre-procedural HBV DNA levels ≧2000 IU/mL (HR 1.55, p=0.041). Conclusions: Multi-nodularity and HBV DNA levels were related with the recurrence of HBV-related HCC after radiofrequency ablation.

Disclosures:

The following people have nothing to disclose: Won Sohn, Yong-Han Paik, Min Woo Lee, Hyunchul Rhim, Hyo Keun Lim, Ju Yeon Cho, Geum-Youn Gwak, Moon Seok Choi, Joon Hyeok Lee, Kwang Cheol Koh, Seung Woon Paik, Byung Chul Yoo

2169

  1. Top of page

The impact of sarcopenia on the recurrence of hepatocellular carcinoma in HCV-related liver cirrhosis

Saori Kamachi1, Yasushi Ide1, Chika Tsuji1, Shunya Nakashita1, Taiga Otsuka1, Yuichiro Eguchi1, Iwata Ozaki1, Atsushi Miyoshi2, Kenji Kitahara2, Toshihiko Mizuta1;
1 Internal medicine, Saga University, Saga, Japan; 2Surgery, Saga University, Saga, Japan

Background:Sarcopenia often occurs in patients with liver cirrhosis and is associated with poor prognosis. However, little is known about the association of sarcopenia with hepatocar-cinogenesis. We therefore assessed the relationship between sarcopenia and the recurrence of hepatocellular carcinoma (HCC) in patients with HCV-related cirrhotic.Methods:Ninety-eight (64 males; median age 71 years) patients free of HCC after local ablation or surgical resection between April 2004 and December 2012 were included. Muscle mass was analyzed using the L3 skeletal muscle index (L3 SMI), calculated by assessing muscle tissue areas on CT images at the third lumber vertebra and normalizing for stature. Cutoff values of the L3 SMI for sarcopenia were defined as <52.4 cm2/m2 for males and <38.5 cm2/m2 for females. The difference in HCC recurrence rate was analyzed using a Cox proportional hazards model. Results:Median L3 SMI was 45.5 cm2/m2 for males and 40.6 cm2/m2 for females, with 83% of males and 35% of females being sarcopenic. The baseline characteristics were comparable in the groups with and without sarcopenia, exception for gender, median age (73 vs. 70 years, p=0.03) and median BMI (21.6 vs. 24.5 kg/m2, p<0.01). The cumulative 1-and 3-year recurrence rates were 38% and 85%, respectively, in sarcopenic and 27% and 60%, respectively, in non-sar-copenic patients (p<0.05), with median times to recurrence of 15.7 and 31.7 months, respectively (p=0.03). Univariate analysis showed that sarcopenia (p=0.03) and age (p<0.01) were significantly associated with recurrence, whereas multi-variate analysis showed that only sarcopenia (p=0.03) was independently associated with recurrence of HCC.Conclu-sion:Sarcopenia may contribute to the recurrence of HCC. Although the detailed mechanisms are still unclear, nutritional or exercise intervention to improve the amount of skeletal muscle may prevent the recurrence of HCC.

Thumbnail image of

Disclosures:

The following people have nothing to disclose: Saori Kamachi, Yasushi Ide, Chika Tsuji, Shunya Nakashita, Taiga Otsuka, Yuichiro Eguchi, Iwata Ozaki, Atsushi Miyoshi, Kenji Kitahara, Toshihiko Mizuta

2170

  1. Top of page

Clinical prediction of a poor prognosis progenitor-like hepatocellular carcinoma subtype

Zhiying Lim1, Krishnakumar Madhavan2, Maureen Da Costa2, Yock Young Dan3;
1Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore; 2Division of Hepato-biliary & Pancreatic Surgery, National University Health System, Singapore, Singapore, Singapore; 3Department of Hepatology and Gastroenterology, National University Health System, Singapore, Singapore, Singapore

Background Prediction of prognosis for hepatocellular carcinoma (HCC) remains a challenge for management. Current prediction models of post-resection recurrence do not take into account tumor biology, limiting the full potential of personalized medicine that specifically targets the biology of HCC. It is postulated that post-resection recurrence is due to micrometastases occurring before the tumor is removed and this has a higher likelihood in a stem cell type of tumor. We hypothesize that clinical indices will allow identification of a primitive progenitor HCC subtype that will have a worse prognosis compared to other subtypes. This study aims to determine if alpha-fetoprotein (AFP) levels used in conjunction with tumor differentiation can better predict post-resection recurrence in HCC. Methods A retrospective review of 11 1 patients who underwent curative hepatectomy for HCC from June 2004 to February 2012 was conducted. An AFP cut-off value of 200ng/ml was used to classify patients into AFP-negative and AFP-positive groups. The AFP-negative group was further subdivided according to tumor differentiation to well-differentiated and poorly-differentiated groups. These patients were studied for prognostic significance of AFP levels and tumor differentiation in predicting early recurrence (defined as <3 years) using univariate analysis. Results Patient demographics and co-morbidities were comparable between both groups. There were 82 (73.9%) patients who were AFP-negative. Within the AFP-negative group, 50 patients had poorly-differentiated tumors and 21 patients had well-differentiated tumors. There was a significantly higher early recurrence rate (<3 years) after surgical resection in AFP-negative HCC (P=0.011). When the 1-year post-resection recurrence rate was studied in AFP-negative tumors, all of these patients that recurred belonged to the poorly differentiated group (P=0.037). Associated factors related to early recurrences were analyzed which revealed tumor size to be significant (P=0.023). Multivariate analysis showed that both AFP levels (P=0.037) and tumor size (P=0.036) were independent factors associated with early recurrence. All these factors do not predict survival. Conclusion While AFP has been reported as a predictor of recurrence, results have been controversial. This is likely due to differences in tumor differentiation as shown in this study. AFP-negative tumors with poorly differentiated histology predict an aggressive group with high potential of recurrence, which is likely due to progenitor-cell based HCC. Identification of this group will allow us to give tailored adjuvant chemotherapy that will hopefully improve outcomes.

Disclosures:

Yock Young Dan - Consulting: Merck, Sharp and Dohme

The following people have nothing to disclose: Zhiying Lim, Krishnakumar Madhavan, Maureen Da Costa

2171

  1. Top of page

The mitochondrial gene abnormalities and sustained pathological alterations in the liver after hepatitis C virus eradication

Nobuhiro Aizawa1, Hirayuki Enomoto1, Kazunori Yoh1, Akio Ishii1, Tomoyuki Takashima1, Yoshiyuki Sakai1, Kazunari Iwata1, Naoto Ikeda1, Hironori Tanaka1, Yoshinori Iwata1, Masaki Saito1, Hiroko Iijima1, Yuji Iimuro2, Jiro Fujimoto2, Shuhei Nishiguchi1;
1Division of Hepatobiliary and Pancreatic Diseases, Department of Internal Medicine, Hyogo collage of medicine, Nishinomiya, Japan; 2Department of Surgery, Hyogo collage of medicine, Nishinomiya, Japan

Background: Despite the improved histological findings in the liver after hepatitis C virus (HCV) eradication, hepatocellular carcinoma (HCC) occurs in some patients with a sustained viro-logical response (SVR) to interferon therapy. However, the mechanism of carcinogenesis in virus-eliminated liver tissues remains unclear. The aim of this study is to clarify the characteristics of SVR-related hepatocarcinogenesis. Subjects and Methods: A total of 31 patients with SVR were enrolled. Seven patients had HCC (Group A), and the remaining 24 patients did not (Group B). We pathologically examined the non-cancerous hepatic tissues of Group A patients, and compared those tissues with the liver tissues of Group B patients by electron microscopy (EM). We evaluated the degree of the dilatation of vesicular endoplasmic reticulum on a scale of 0 to 3. We also evaluated the mitochondrial alterations on a scale of 0 to 10 according to the following points: (1) Dense granules, (2) Paracrystalline inclusions, (3) Vacuole formation, (4) Irregularity of mitochondrial shapes, and (5) Lack of mitochondrial membrane. Furthermore, the numbers of mutations of mt-DNA D-loop areas were analyzed by a PCR-based direct sequence method. We examined the relationship between mt-DNA mutations in liver tissues and pathological alterations. This study was conducted in accordance with the Helsinki declaration and written informed consents were obtained from all patients before the study. Results: Regarding dilatation of vesicular endoplasmic reticulum, elevated scores (> 2 points) were found in 17% (4/24) of the Group B patients and in 100% (7/7) of the Group A patients. Regarding the morphological mitochondrial alterations, elevated scores (> 5 points) were found in 4.2% (1/24) of the Group B patients and in 71.4% (5/7) of the Group A patients. Scores of the alterations regarding the mitchondria and the vesicular endoplasmic reticulum were higher in the HCC Group than in the non-HCC Group. Mutations at 5 or more sites in mt-DNA D-loop areas were found in specimens from 3 patients in the Group A, and they had dilatation of vesicular endoplasmic reticulum and the morphological mitochondrial alterations. Conclusion: EM revealed obvious histological alterations in non-cancerous hepatic tissues of HCC patients. Their histological alterations after SVR should be correlated with hepatocarcinogenesis.

Disclosures:

The following people have nothing to disclose: Nobuhiro Aizawa, Hirayuki Enomoto, Kazunori Yoh, Akio Ishii, Tomoyuki Takashima, Yoshiyuki Sakai, Kazunari Iwata, Naoto Ikeda, Hironori Tanaka, Yoshinori Iwata, Masaki Saito, Hiroko Iijima, Yuji Iimuro, Jiro Fujimoto, Shuhei Nishiguchi

2172

  1. Top of page

Drug Eluting Bead Transcatheter Arterial Chemo Embolization (DEB TACE) in Hepatocellular Carcinoma (HCC) : Pathological and Radiological Correlation in a Liver Transplant(LT) Population

Ekta Gupta1, Tejas S. Kirtane1, Chiranjeev Dash2, Natalie Szpakowski1, Maunik Patel3, Won Kyoo Cho1, Reena Jha3, Justin Lee3, Kirti Shetty2;
1 Division of Gastroenterology, Washington Hospital Center/ Georgetown University, Washington DC, DC; 2Divi-sion of Gastroenterology, Georgetown University, Washington DC, DC; 3Georgetown University Hospital, Washington DC, DC

Background: While conventional TACE (cTACE) has been demonstrated to produce significant tumor necrosis, the pathological effects of DEB-TACE on HCC have not been well characterized. We undertook this analysis to (a) examine HCC response to DEB TACE utilizing explanted specimens following LT and (b) determine correlation between pathological and radiological assessment of tumor response. Methods: Patients undergoing LT at our center following DEB-TACE were identified through our prospectively maintained database (REDCap). All included patients for radiological-pathological correlation analysis had an index MRI (pre-DEB TACE), and at least one MRI between DEB-TACE and LT. Tumor response was evaluated using both radiological [EASL criteria as published- complete response (CR); partial response (PR); stable disease (SD); progressive disease (PD)] and pathological criteria [complete response (CR) - 100% necrosis; partial response (PR)-50–99% necrosis; stable disease (SD)- 1–49% necrosis; no response (NR)- no necrosis]. Objective response (OR) was defined as CR + PR; disease control (DC) as CR + PR + SD. The agreement between pathology and MRI in assessing CR, OR, and DC were measured using Cohen's Kappa (κ). Results: Between November 2008 and November 2012, we identified 37 patients who underwent DEB TACE for HCC followed by LT. Majority were males (73%) with mean age of 58.9 years ± 5.6 and mean MELD score of 12.2 ±4.5. Etiology of cirrhosis was hepatitis C virus in 78%. The overall rates of CR, OR, DC were similar by both pathological and radiological criteria (Table 1). However, in the 28 patients in whom both pathological and radiological criteria were available, the agreement between the two methods of measurement was poor for OR = 0.28) and DC (κ = -0.11); moderate for CR = 0.64). The detection of CR by MRI had sensitivity of 75% and specificity of 89%. Conclusions: 1. DEB TACE is an effective loco-regional therapy for HCC in a LT population based on pathological criteria (OR rate of 65%). 2.MRI has acceptable sensitivity and specificity when correlated with pathological examination to assess DEB TACE CR but not OR or DC

Table 1. Outcomes of DEB TACE for HCC

 Pathological (Criteria (n=34)Radiological Criteria (n=28)
 n%n%
CR1029829
PR12351036
SD721829
NR/PD51527
OR22651864
DC29852693

2173

  1. Top of page

Incidence and risk assessment for the development of hepatocellular carcinoma during long-term entecavir and lamivudine therapy in hepatitis B virus-related liver disease

Yun Nah Lee1, Young Seok Kim1, Sang Gyune Kim1, Sae Hwan Lee1, Young Don Kim2, Gab Jin Cheon2, Soung Won Jeong1, Jae Young Jang1, Hong Soo Kim1, Boo Sung Kim1;
1Digestive Disease Center and Research Institute, Department of Internal Medicine, SoonChunHyang University College of Medicine, Bucheon, Republic of Korea;2Departments of Internal Medicine, University of Ulsan College of Medicine, Gangneung Asan Hospital, Gangneung, Republic of Korea

Background and Aims: Chronic hepatitis B (CHB) patients are at increased risk for hepatocellular carcinoma (HCC). Although nucleos(t)ide analogues play a great role to reduce the incidence of HCC, it is not clear what factors contribute the occurrence of HCC even during continuing treatment. The aim of this study was to assess and compare the incidence and risk factor for the development of HCC during long-term entecavir (ETV) and lamivudine (LAM) therapy in patients with hepatitis B virus-related liver disease. Methods: Total 458 CHB patients who have been treated with nucleos(t)ide analogue(s) starting with entecavir (293 patients) and lamivudine (165 patients) for more than 1 year were included in this study. The incidence and the risk factors of HCC were analyzed. Results: HCC developed in 19 of 293 (6.5%) patients and 12 of 165 (7.3%) after median 29.7±12.4 and 28.5 ±25.7 months from starting ETV and LAM, respectively. The rate of LAM having undetectable HBV DNA at 1 year after treatment was lower when compared with the ETV-treated group (55.2% vs 87.7 %, p<0.001). The LAM also had a higher drug-resistant mutation rate (60.6% vs 1.4%, p<0.001). The cumulative incidence rates of HCC was not significantly difference between ETV and LAM (p=0.131). In logistic regression analysis, the significant risk factors related to HCC development were old age (over 50 years), presence of diabetes mellitus, cirrhosis and failure to normalize ALT (over 40 IU/L) in ETV group and old age (over 50 years), albumin lower than 3.8 mg/dL, and undetectable HBV DNA at 1 year after treatment in LAM group. In multivariated analysis, old age (OR 3.81; 95% CI 1.17 to 17.2; p=0.025), failure to normalize ALT (OR 4.33; 95% CI 1.59 to 12.0; p=0.005), liver cirrhosis (OR 8.15; 95% CI 2.20 to 52.7; p<0.001) were independently associated with HCC development in ETV group and old age (OR 5.56; 95% CI 1.18 to 26.22; p=0.03), albumin lower than 3.8 mg/dL (OR 11.38; 95% CI 2.98 to 43.45; p<0.001) and undetectable HBV DNA at 1 year after treatment (OR 5.57; 95% CI 1.60 to 19.43; p<0.007). Conclusion: In this study, ETV therapy showed more excellent efficacy in nucle-oside-naïve chronic hepatitis B patients than LAM therapy, but there was no difference between ETV and LAM therapy in cumulative incidence rates of HCC. Therefore, the virologic response does not seem to significantly reduce the overall incidence of HCC in patients with ETV therapy.

Disclosures:

The following people have nothing to disclose: Yun Nah Lee, Young Seok Kim, Sang Gyune Kim, Sae Hwan Lee, Young Don Kim, Gab Jin Cheon, Soung Won Jeong, Jae Young Jang, Hong Soo Kim, Boo Sung Kim

Disclosures:

Kirti Shetty - Grant/Research Support: Ikaria, Novartis, Onyx-Bayer, Hyperion; Speaking and Teaching: Merck-Schering Plough, Salix, Gilead, Onyx

The following people have nothing to disclose: Ekta Gupta, Tejas S. Kirtane, Chiranjeev Dash, Natalie Szpakowski, Maunik Patel, Won Kyoo Cho, Reena Jha, Justin Lee

2174

  1. Top of page

Monitoring of Plasma Vascular Endothelial Growth Factor after Sorafenib Administration as Predictor of Survival in Advanced Hepatocellular Carcinoma

Kaoru Tsuchiya1, Yutaka Yasui1, Shuya Matsuda1, Masaru Muraoka1, Nobuharu Tamaki1, Shoko Suzuki1, Takanori Hosokawa1, Ken Ueda1, Hiroyuki Nakanishi1, Jun Itakura1, Yuka Takahashi1, Masayuki Kurosaki1, Yasuhiro Asahina2, 3, Namiki Izumi1
1Department of Gastroenterology and Hepatology, Musashino Red Cross Hospital, Tokyo, Japan; 2Department of Gastroenterology, Tokyo Medical and Dental University, Tokyo, Japan; 3Department of Liver Disease Control, Tokyo Medical and Dental University, Tokyo, Japan

Background and Aims: Predictive biomarkers for determining prognosis of patients with hepatocellular carcinoma (HCC) receiving sorafenib treatment are required because reduction of tumor size by sorafenib is rare even in patients with a favorable prognosis. Vascular endothelial growth factor (VEGF) is a sorafenib target and has been shown to have an independent prognostic value in a large cohort of patients with advanced HCC. Few studies have investigated serial changes in plasma VEGF concentrations following sorafenib administration. The aim of the present study was to investigate changes in plasma VEGF concentrations as a potential prognostic biomarker in patients treated with sorafenib. Methods: Between December 2009 and January 2013, 75 patients with advanced, inoperable HCC were treated with sorafenib at our center; their plasma VEGF and serum alpha-fetoprotein (AFP) concentrations were evaluated at baseline (before sorafenib administration) and at 4-weekly intervals until discontinuation of sorafenib. MDCT was performed before sorafenib administration, 1 month after, and every 3 month from then on. Radiological findings were evaluated by modified RECIST criteria. Results: Compared with pretreatment levels, plasma VEGF concentrations increased at 4 weeks following sorafenib administration in 55 of 75 (73. 3%) patients, but decreased at 8 weeks in 19 of 73 (26. 0%) patients. Patients whose plasma VEGF concentrations were maintained at <1. 5 times the level of 8 week after sorafenib administration (n=41) had longer MST than those >1. 5 times (n=34) (8. 8 months vs. 31. 4months; P <. 0001). Best radiological findings were Complete Response(CR)(n=5), Partial Response(PR) (n=20), Stable Disease(SD) (n=43), Progressive Disease(PD) (n=7). According to therapeutic assessment after 12 weeks following sorafenib administration, maintaining plasma VEGF (HR4. 97, P<. 0001) and serum AFP (HR 5. 29, P=0. 0007) concentrations <1. 5 times the level of week 8 were significantly independent predictive factors of overall survival. Patients maintaining both plasma VEGF and serum AFP concentrations <1. 5 times (n=19) showed excellent survival (18 patients survive, median; 15. 8months), even though their radiological findings were not diagnosed as complete response or partial response. Conclusions: Plasma VEGF concentrations after sorafenib administration may predict the prognosis of patients with advanced HCC. Monitoring of plasma VEGF and serum AFP concentrations is clinically useful for patients treated with sorafenib.

Disclosures:

Namiki Izumi - Speaking and Teaching: MSD Co., Chugai Co., Daiichi Sankyo Co., Bayer Co., Bristol Meyers Co.

The following people have nothing to disclose: Kaoru Tsuchiya, Yutaka Yasui, Shuya Matsuda, Masaru Muraoka, Nobuharu Tamaki, Shoko Suzuki, Takanori Hosokawa, Ken Ueda, Hiroyuki Nakanishi, Jun Itakura, Yuka Takahashi, Masayuki Kurosaki, Yasuhiro Asahina

2175

  1. Top of page

Moderately causative role of occult hepatitis B virus infection (occult HBV) in the development of non- hepatitis B virus (HBV) infection, non- hepatitis C virus (HCV) infection (NBNC) hepatocellular carcinoma (HCC)

Hiroshi Imamura1, Norie Yamada2, Hiroshi Yotsuyanagi3, Seiji Kawasaki1, Tadatoshi Takayama4, Masakazu Yamamoto5, Norihiro Kokudo6, Takanobu Kato7, Shinji Tanaka8, Shigeki Arii8
1Department of Hepato-Biliary-Pancreatic Surgery Juntendo School of Medicine, Tokyo, Japan; 2Department of Internal Medicine, Center for Liver Diseases, Seizankai Kiyokawa Hospital, Tokyo, Japan; 3Department of Infectious Diseases, University of Tokyo, Tokyo, Japan; 4Department of Digestive Surgery, Nihon University School of, Tokyo, Japan; 5Department of Surgery, Institute of Gastroenterology, Tokyo Women's Medical University, Tokyo, Japan; 6Department of Hepato-Biliary-Pancreatic Surgery, University of Tokyo, Tokyo, Japan; 7Department of Virology II, National Institute of Infectious Diseases, Tokyo, Japan; 8Department of Hepato-Biliary-Pancreatic Surgery, Tokyo Medical and Dental University, Tokyo, Japan

Introduction: Occult HBV is defined by the presence of HBV DNA in the serum or liver in the individuals negative for HBV surface antigen (HBsAg). It is debatable whether occult HBV is a risk factor to HCC development in NBNC cohort. In Japan, 60%, 20 %, and 20 % of patients with HCC are HCV- related, HBV- related, and NBNC, respectively. Aim: We investigated whether occult HBV contributes to the development of NBNC HCC. Methods: The study group comprised of 58 consecutive patients who underwent hepatectomy for NBNC HCC; while the control group comprised of 61 consecutive patients undergoing liver resection for diseases not related to primary liver cancer including HCC, i. e., metastatic liver tumor, gall bladder cancer, etc. Patients under 40 years old were excluded from the study considering that the nationwide HBV vaccination program to newborn form HBV carrier mother started in 1986. Ten patients with HBsAg (+) HCC were also studied. HBV surface antibody (HBsAb) and core antibody (HBcAb) was measured. Serum sample and liver tissue sample from tumorous as well as non-tumorous parts of the liver were obtained from the study group; and serum and liver tissue from the non-tumorous part were taken from the control group. HBV serum DNA and HBV transcripts in the liver tissue were examined by RTD-PCR using specific primers of surface (S), core (C), and X regions. Occult HBV was defined by the detection of HBV DNA in at least one PCR assays in either serum or liver tissue samples. Results: Male/female ratios were 51/7 vs. 29/32 (p<0. 0001) and median (range) ages were 65(43-86) vs. 69(46-83) years (p=0. 020), the study vs. control groups, respectively. In the study group, 26 out of 55 (47 %) patients were HBcAb (+) and 12 of them (54 %) were HBsAb (+). Occult HBV was identified in 20 (34 %) and 11(18 %) patients in the study and control group, respectively (Odds ratio [95 % C. I.]): 2. 39 (1. 02-5. 58). Overall, 15 out of 37 (41 %) patients with HBcAb (+) and 16 out of 80 (20 %) patients with HBcAb (-) patients were identified as those with occult HBV (p=0. 025). The levels of intrahepatic HBV DNA in the study and the control groups were lower than those in HBV-related HCC group. Only 4 of the 58 (7%) of the study group and 3 of the 61 (5%) of the control group had intrahepatic levels comparable to HBV group, presumably defined over 1 HBV DNA copy per 1 microgram of total DNA. Conclusions: Occult HBV may be a moderate contributor to the development of NBNC HCC. HBcAb positivity cannot be used as a surrogate marker for occult HBV. Low intrahepatic HBV DNA levels in the majority of cases suggest that the role of occult HBV might be different from HBsAg (+) HCC.

Disclosures:

Norihiro Kokudo - Grant/Research Support: Dainihon Sumitomo Pharmacy, Bayer, Chugai Pharmaceutical, Translational Research Informatice Center

The following people have nothing to disclose: Hiroshi Imamura, Norie Yamada, Hiroshi Yotsuyanagi, Seiji Kawasaki, Tadatoshi Takayama, Masakazu Yamamoto, Takanobu Kato, Shinji Tanaka, Shigeki Arii

2176

  1. Top of page

Safety and outcomes by disease etiology in sorafenibtreated uHCC patients in clinical practice: final analysis of GIDEON (Global Investigation of therapeutic DEcisions in hepatocellular carcinoma and Of its treatment with sorafeNib)

Arun J. Sanyal1, Riccardo Lencioni2, Sheng-Long Ye3, Masatoshi Kudo4, Alan Venook5, Jean-Pierre Bronowicki6, Xiao-Ping Chen7, Lucy Dagher8, Junji Furuse9, Jean-Francois H. Geschwind10, Laura Ladron de Guevara11, Christos Papandreou12, Tadatoshi Takayama13, Seung Kew Yoon14, Keiko Nakajima15, Jorge A. Marrero16
1Internal Medicine/Division of Gastroenterology Hepatology and Nutrition, Virginia Commonwealth University, Richmond, VA; 2Division of Diagnostic Imaging and Intervention, Pisa University Hospital and School of Medicine, Pisa, Italy; 3Liver Cancer Institute and Zhongshan Hospital, Fudan University, Shanghai, China; 4Department of Gastroenterology and Hepatology, Kink University School of Medicine, Osaka, Japan; 5University of California, San Francisco, CA; 6Department of Gastroenterology and Hepatology, INSERM U954, University Hospital, University of Lorraine, Nancy, France; 7Hepatic Surgery Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China; 8Policlίnica Metropolitana, Caracas, Venezuela, Bolivarian Republic of; 9Kyorin University School of Medicine, Tokyo, Japan; 10Vascular and Interventional Radiology, Johns Hopkins University School of Medicine, Baltimore, MD; 11 Hospital Angeles Clinica Londres, Mexico City, Mexico; 12University Hospital of Larissa, Larissa, Greece; 13Department of Digestive Surgery, Nihon University School of Medicine, Tokyo, Japan; 14The Catholic University of Korea, Seoul, Republic of Korea; 15Global Medical Affairs, Bayer HealthCare Pharmaceuticals, Montville, NJ; 16Department of Internal Medicine, Division of Digestive and Liver Diseases, University of Texas, Southwestern Medical Center, Dallas, TX

Background: GIDEON is a global, prospective, non-interventional study. Completion of GIDEON creates a large database of >3000 sorafenib (Sor)-treated unresectable hepatocellular carcinoma patients (pts), allowing for evaluation across different subgroups, including disease etiology. Methods: Baseline characteristics were collected on pts for whom a decision to treat with Sor had been made in clinical practice. Adverse events (AEs), dosing, and outcomes data were collected during follow-up. Results: In the safety population (n=3202), the most common disease etiologies were hepatitis B virus (HBV, 37%), hepatitis C virus (HCV, 33%), and alcohol use (26%). Regionally, Asia-Pacific had the highest HBV incidence (82%); alcohol use was most common in the US (39%). More HBV than HCV pts had advanced Barcelona Clinic Liver Cancer (BCLC) stage and extrahepatic spread (Table). Non-alcoholic steatohepatitis was not commonly reported: 0. 3% (HBV pts), 0. 6% (HCV pts), 1. 6% (alcohol use pts). HCV pts had higher Child-Pugh score than HBV pts (Table). More HCV pts had Eastern Cooperative Oncology Group performance status ≥2 (45%) than HBV pts (10%). A similar number of pts received prior locoregional treatment or transarterial chemoembolization across etiological subgroups; more HBV pts received prior surgery (Table). Sor safety profile was generally comparable across etiologies. HCV pts had a slightly higher rate of drug-related AEs (74%) compared with HBV (56%) or alcohol use pts (68%), and a higher rate of drug-related grade 3 or 4 AEs (HCV 29%, HBV 18%, alcohol use 25%). In the intent-to-treat population (n=3213), median overall survival (OS; months) from Sor initiation and time from initial diagnosis to death (months) were longer in HCV (12. 1 and 36. 5, respectively) than HBV (9. 8 and 23. 6) and alcohol use pts (9. 7 and 22. 9). Further analyses evaluating the influence of prior therapy on outcome patterns across etiological subgroups are planned. Conclusions: As previously reported, disease etiology varies by region. Sor safety and prior treatment patterns are comparable across etiological subgroups. Shorter OS in HBV pts may reflect the natural disease history.

Overall a'b (n=3202)HBVa (n=1170)HCV a (n=1053)Alcohola (n=1418)
  1. a

    Missing pts not tabulated; bPts may have multiple responses; cAt start of Sor therapy; dNon-evaluable pts not tabulated

Disease characteristicsc (% of n)Child-Pugh scored A B61. 5 20. 8 2. 365. 2 18. 6 1. 659. 6 23. 8 3. 258. 8 24. 1 2. 5
BCLC stage ABCD7. 119. 8 52. 0 5. 413. 6 4. 99. 9 25. 3 44. 6 6. 15. 9 19. 7 53. 8 5. 0
Extrahepatic spread39. 749. 931. 040. 0
Previous therapy (%of n)Prior locoregional treatment58646355
Prior transarterial chemoembolization47554845
Prior surgery21271617

Disclosures:

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

Riccardo Lencioni - Consulting: Jennerex Inc.; Speaking and Teaching: Bayer Healthcare

Masatoshi Kudo - Advisory Committees or Review Panels: Bayer HealthCare; Grant/Research Support: Bayer HealthCare

Alan Venook - Grant/Research Support: Bayer Pharm, Onyx

Jean-Pierre Bronowicki - Consulting: Merck, Janssen, Boehringer Ingelheim, Gilead, BMS, Bayer, Novartis, GSK, Merck, Janssen, Boehringer Ingelheim, Gilead, BMS, Bayer, Novartis, GSK; Speaking and Teaching: Roche, Merck, Janssen, BMS, Bayer, Roche, Merck, Janssen, BMS, Bayer

Junji Furuse - Consulting: Chugai, Eisai, Chugai, Eisai; Speaking and Teaching: Bayer, Taiho, Eli Lylli, Bayer, Taiho, Eli Lylli

Jean-Francois H. Geschwind - Consulting: Biocompatibles/BTG, Bayer HealthCare, Guerbet, Nordion, Delcath, Prescience Labs, Jennerex; Grant/Research Support: Biocompatibles/BTG, Bayer HealthCare, Philips Medical, Nordion, Guerbet, DOD, NCI-ECOG, NIH-RO1; Stock Shareholder: PreScience Labs Founder/CEO

Laura Ladron de Guevara - Consulting: Bayer

Christos Papandreou - Advisory Committees or Review Panels: Bayer

Keiko Nakajima - Employment: Bayer HealthCare Pharmaceuticals; Stock Shareholder: Bayer HealthCare Pharmaceuticals

Jorge A. Marrero - Advisory Committees or Review Panels: Bayer, Onyx; Grant/Research Support: Bayer, BMS

The following people have nothing to disclose: Sheng-Long Ye, Xiao-Ping Chen, Lucy Dagher, Tadatoshi Takayama, Seung Kew Yoon

2177

  1. Top of page

Multimodality Treatment Of Hepatocellular Carcinoma In A Single Tertiary Referral Centre

Michela Triolo1, Angelo Sangiovanni1, Matteo A. Manini1, Massimo Iavarone1, Sara Vavassori1, Cristina Della Corte1, Raffaeila Romeo1, Laura Virginia Forzenigo2, Antonio F. Nicolini2, Giorgio Rossi3, Massimo Colombo1
11st Division of Gastroenterology Fondazione IRCCS Ca, Granda Ospedale Maggiore Policlinico, University of Milan, Milano, Italy; 2Division of Radiology Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, University of Milan, Milano, Italy; 3Division of Surgery and Liver Transplant, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, University of Milan, Milano, Italy

Introduction: Background & Aims: Multimodality treatment of hepatocellular carcinoma (HCC) is a common clinical practice. Comorbidities, liver and no-liver related conditions, led to an incomplete adherence to the AASLD guidelines for the treatment of HCC. The impact of this behavior is not fully investigated. The aim is to define the clinical impact of multimodality treatment of HCC in cirrhotic patients attending a single tertiary referral centre Methods: 383 consecutive cirrhotic patients with a de-novo diagnosis of HCC, 281(73%) males, mean age 66 (31-86), 298 (78%) Child-Pugh A, 201(52%) BCLC A0-3, 58 (15%) A4, 67 (18%) B and 57 (15%) C, were observed between 200/ and 2011. HCC treatments were decided by a multidisciplinary team of experts on intention-to-treat according to the AASLD guidelines. Restaging was performed every 1-3 months by CT/MRI after treatment. Patient and tumor characteristics and blood tests were considered as predictors of survival and tested by univariate and multivariate Cox proportional hazards model (STATA 10. 0 Statistical Package). Results: During a mean trial time of 36 months 107 (28%) patients died and 35 (9%) were lost. Overall 1, 3 and 5 yr survival was 92%, 68% and 57 % respectively (99%, 84% and 71% for BCLC A, 94%, 52%, 43% for BCLC B, 59%, 16% and 0% for BCLC C). First line treatment was: 37 (10%) liver transplantation (OLT), 56 (15%) resection, 142 (37%) local ablation, 89 (23%) chemoembolization, 34 (9%) sorafenib, 12 (3%) best supportive care (BSC), 13 (3%) not specified. Adherence to AASLD guidelines was in 267 (72%) patients. 62 (17%) received a treatment recommended by AASLD for a more advanced stage, 41(11%) for a less advanced stage. Complete response was achieved in 152 (41%) after first line treatment (including 37 OLT), in 44 (12%) after second line, in 11 (3%) after third line. HCC recurred after OLT in 3 (8%). Independent predictors of survival were: treatment choice (HR 1. 70 95% CI 1. 40-2. 04, p <0. 000), BCLC (HR 2. 10, 95% CI 1. 61-2.95 p<0. 000), AFP > 100ng/mL, (HR 1. 53, 95% CI 1. 19-1.97 p=0. 001). Conclusion: In the clinical practice of a tertiary referral centre, adherence to AASLD guidelines for HCC treatment was in two third of the patients and a complete response in more than half of the cases. The multidisciplinary decision of treatment choice is one of the strongest predictors of HCC survival.

Disclosures:

Massimo Colombo - Advisory Committees or Review Panels: BRISTOL-MEYERSSQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, Janssen Cilag, Achillion; Grant/Research Support: BRISTOL-MEYERS-SQUIBB, ROCHE, GILEAD, BRISTOLMEYERS-SQUIBB, ROCHE, GILEAD; Speaking and Teaching: Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX, Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX

The following people have nothing to disclose: Michela Triolo, Angelo Sangiovanni, Matteo A. Manini, Massimo Iavarone, Sara Vavassori, Cristina Della Corte, Raffaella Romeo, Laura Virginia Forzenigo, Antonio F. Nicolini, Giorgio Rossi

2178

  1. Top of page

Prognostic value of anthropometric features assessed by CT-scan in patients with advanced hepatocellular carcinoma treated by tyrosine kinase inhibitors

Jean-Charles Nault1, Frederic Pigneur2, Charlotte E. Costentin3, Lambros Tselikas2, Alain Luciani2, Alexis Laurent4, Ariane Mallat3, Christophe Duvoux3, Alain Rahmouni2, Thomas Decaens3
1Inserm U674, Paris, France; 2Radiology, Henri Mondor Hospital, APHP, Crete", France; 3Liver unit, Henri Mondor Hospital, APHP, Creteil, France; 4Hepatobiliary surgery, Henri Mondor Hospital, APHP, Creteil, France

Introduction: Anthropometric (sarcopenia, subcutaneous and visceral fat) assessment by CT-scan have been associated with survival in cancer treated by antiangiogenic therapies. We aim to explore the prognostic value of anthropometric features in patients with advanced hepatocellular carcinoma (HCC) treated by tyrosine kinase inhibitors. Material and methods: Clinical, biological data and treatment-related toxicity were recorded in a monocentric cohort of 52 patients treated for advanced HCC by tyrosine kinase inhibitors (sorafenib or brivanib). 48 patients had a CT-scan available before treatment for anthropometric measures: visceral fat area, subcutaneous fat area, L3 muscle area, psoas L3 area. Sarcopenia was defined according to an international consensus. Survival analysis was performed using cox univariate/multivariate analysis. Results: Patients were mainly men (86 %) of median 69 years old with underlying cirrhosis (71%). The etiology of the underlying liver disease was alcohol in 25 %, HBV in 29%, HCV in 10%, NASH in 8 %, mixed etiology in 17% and unknown etiology in 11% of the cases. 14 patients (27%) had grade 2-4 diarrhea and 10 patients (19%) had grade 2- 4 hand foot syndrome. 6% of the patients had a body mass index (BMI) <20, 46% a BMI between 20-25, 33% were overweight (BMI 25-30) and 8% obese (BMI>30). The median of visceral fat area was 50cm2/m2 for men and 26cm2/m2 for women and the median value of subcutaneous fat aera was 67cm2/m2 for men and 78cm2/m2 for women. 70% of patients (34/48) had sarcopenia and 35% had both overweight or obesity and sarcopenia. Sarcopenia was not correlated with treatment-related toxicity. Median overall survival was 10. 4 months and median time to progression was 5. 7 months. Patients with high visceral fat area had a shorter median overall survival compared to patients with low visceral fat area but the difference was not significant (9 months vs 15 months, P=0. 15 log rank). In univariate analysis, hand foot syndrome, level of alkaline phosphatase (AP) and alphafoetoprotein were significantly associated with overall survival. In multivariate analysis, hand foot syndrome (HR: 0, 33 [0, 14; 0, 8] P=0, 01)and a low level of AP (HR: 0, 42 [0, 23; 0, 80] P=0, 008) were independently associated with an increase overall survival. Conclusion: Sarcopenia assessed by CT-scan has a high prevalence (70%) in patients with advanced HCC treated by tyrosine kinase inhibitors, even in patients with overweight and obesity. However, anthropometric measures have no prognostic value in our cohort. Presence of hand-foot syndrome and a normal alkalin phosphatase level are associated with an increased overall survival.

Disclosures:

Christophe Duvoux - Advisory Committees or Review Panels: Novartis, Roche, Novartis, Roche, Novartis, Roche, Novartis, Roche; Speaking and Teaching: Astellas, Astellas, Astellas, Astellas

The following people have nothing to disclose: Jean-Charles Nault, Frederic Pigneur, Charlotte E. Costentin, Lambros Tselikas, Alain Luciani, Alexis Laurent, Ariane Mallat, Alain Rahmouni, Thomas Decaens

2179

  1. Top of page

Chemoembolization of hepatocellular carcinoma with M1 drug eluting beads: safety and effectiveness

Massimo Iavarone1, Antonio F. Nicolini2, Michela Triolo1, Silvia Crespi2, Angelo Sangiovanni1, Massimo Colombo1

11st Division of Gastroenterology, Fondazione IRCCS Ca' Granda, Milan, Italy; 2Division of Radiology, Fondazione IRCCS Ca' Granda, Milan, Italy

BACKGROUND/AIMS: Transarterial chemoembolization (TACE) is one of the most utilized treatment for hepatocellular carcinoma (HCC) on cirrhosis. This preliminary study assesses the safety and efficacy of transarterial chemoembolization using 70-150 μm drug eluting beads (DC BeadM1-TACE, Biocompatibles UK), an embolizing device with a narrower distribution avoiding premature proximal occlusion of tumor-feeding vessels and increasing drug delivery to the tumor. METHODS: All consecutive patients submitted to DC BeadM1-TACE for HCC in our Center were included. At least one sizable and untreated lesion should have been detectable at baseline CTscan. Clinical and analytical data were recorded at 24 and 48 h, 7, 14 and 30 days after first DC BeadM1-TACE. Response was assessed by CT-scan after one month according to RECIST criteria modified for HCC (mRECIST). Those patients who did not obtained a complete response, were subsequently submitted to a second DC BeadM1-TACE. RESULTS: Thirty-six compensated cirrhotics (90% male, mean age 72 years, 69% HCV, 70% Child-Pugh A) with unresectable HCC (28% BCLC A4 and 42% BCLC B) received chemoembolization with doxorubicin loaded DC BeadM1. No major complications occurred during the study treatment period, 47% patients developed abdominal pain after 20 minutes to 3 hours following the procedure, 6% patients developed a temporary arterial hypertensive crisis and 11% patients developed the classical post TACE syndrome. Among 27 patients with CT-scan available after 1 month, 11% had complete response, 67% partial response, 19% stable disease and progression was observed in 3%. CONCLUSIONS: Chemoembolization using DC BeadM1 is an effective procedure with a favourable safety profile and interesting results in terms of tumor necrosis.

Disclosures:

Massimo Colombo - Advisory Committees or Review Panels: BRISTOL-MEYERSSQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, GILEAD, Janssen Cilag, Achillion; Grant/Research Support: BRISTOL-MEYERS-SQUIBB, ROCHE, GILEAD, BRISTOLMEYERS-SQUIBB, ROCHE, GILEAD; Speaking and Teaching: Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX, Glaxo Smith-Kline, BRISTOL-MEYERS-SQUIBB, SCHERING-PLOUGH, ROCHE, NOVARTIS, GILEAD, VERTEX

The following people have nothing to disclose: Massimo Iavarone, Antonio F. Nicolini, Michela Triolo, Silvia Crespi, Angelo Sangiovanni

2180

  1. Top of page

Hepatic toxicity following chemoembolization with doxorubicin-eluting beads in patients with HCC and marginal hepatic reserve

R. P. Lokken1, Nicholas Fidelman1, Francis Y Yao2, Robert Kerlan1; 1Department of Radiology, University of California San Francisco, San Francisco, CA; 2Division of Gastroenterology and Hepatology, University of California San Francisco, San Francisco, CA

PURPOSE. To determine the prevalence and risk factors of irreversible hepatic toxicity after chemoembolization with doxorubicin-eluting beads (DEB-TACE) for hepatocellular carcinoma (HCC) in a patient population with high prevalence of hepatic dysfunction. MATERIALS AND METHODS. 48 consecutive adult liver transplant candidates with HCC diagnosed according to AASLD guidelines underwent 76 segmental or subsegmental DEB-TACE procedures from 2010 to 2012 at our institution. Median baseline MELD score was 13, and 55 procedures (72%) were done in the setting of Child-Pugh B or C cirrhosis (20 procedures in the setting of Child-Pugh C cirrhosis). 22 procedures (29%) were for BCLC stage C disease, and 20 procedures (26%) were for BCLC stage D disease. Hepatotoxicity was defined as new or worsening ascites, encephalopathy, or NCI Common Terminology Criteria for Adverse Events grade 3 or 4 toxicity of bilirubin, AST, ALT, creatinine or INR. Rate and risk factors for death or urgent liver transplantation within 6 weeks of DEB-TACE and irreversible hepatotoxicity were determined using generalized estimating equation analysis. RESULTS. 24 DEB-TACE procedures (32%) resulted in hepatic toxicity, which was irreversible after 15 (19. 7%) of the procedures. Of these 24 procedures, 8 resulted in irreversible laboratory abnormalities, 8 caused new moderate or severe ascites, whereas 1 led to new irreversible encephalopathy. Prevalence of irreversible hepatotoxicity was 35% for patients with ChildPugh class C cirrhosis. There were no deaths, but 1 patient with hepatotoxicity required transplantation within 6 weeks of DEBTACE. Strongest predictors for irreversible hepatotoxicity were Child-Pugh class C cirrhosis, platelet count <60, 000/ml, portal vein flow abnormality (thrombosis, TIPS, or hepatofugal flow), MELD>15, Eastern Cooperative Oncology Group performance status of 2, and presence of bilobar disease. CONCLUSION. DEB-TACE appears to be safe in a patient population with marginal hepatic reserve, however patients with Child-Pugh C cirrhosis should be approached with caution if transplant is not an option.

Disclosures:

The following people have nothing to disclose: R. P. Lokken, Nicholas Fidelman, Francis Y. Yao, Robert Kerlan

2181

  1. Top of page

Effects of common variation in the hepatocanalicular phosphatidylcholine transporter ABCB4 and the bile salt export pump ABCB11 on genetic cholangiocarcinoma risk

Vincent Zimmer1, Florentina Mihalache2, 1, Marcin Krawczyk1, Bianca Simon1, Frank Grunhage1, Monica Acalovschi2, Frank Lammert1
1Department of Medicine I, Saarland University Medical Center, Homburg, Germany; 2Department of Medicine III, University Iuliu Hatieganu, Cluj-Napoca, Romania

Background: The pathogenesis of cholangiocarcinoma (CCA) as the second most common primary liver cancer is poorly characterized. However, the combined effects of cholestasis and chronic inflammation are considered to be central to CCA formation. Whereas information on inherited risk factors underlying CCA susceptibility is limited, genetic variation in hepatobiliary transporters has been suggested to modulate genetic risk. 1 Therefore we now aimed to investigate selected hepatobiliary transporter variants in an independent large European cohort of CCA patients. Patients & Methods: Overall, 226 individuals with CCA (males n =132, age 66 ± 12 years) originating from Germany (n =168) and Romania (n = 58) were genotyped. The control group consisted of 359 CCA- and PSCfree individuals (males n =160, age 61 ± 11 years). The common single nucleotide polymorphisms (SNPs) in the hepatocanalicular phosphatidylcholine transporter ABCB4 (c. 504C>T, c. 711T>A, p. R652G, and rs31653 in intron 26) and the bile salt export pump ABCB11 (c. 3084A>G and p. A444V) were genotyped by PCR-based assays with 5'-nuclease and fluorescence detection (TaqMan). 2 Results: All genotype frequencies except for rs31653 (p=0. 02) are in HWE (p > 0. 05). In contrast to initial reports, the association tests do not provide evidence for genetic risk modulation by either ABCB4 (rs2109505: odds ratio (OR) = 0. 95; rs1202283: OR = 0. 82; rs8187799: OR = 0. 76; rs31653: OR = 0. 78) or ABCB11 (c. 3084A>G: OR =1. 01; p. A444V: OR = 0. 94) hepatobiliary transporter variants (all P > 0. 05). Conclusions: Despite the strong functional candidacy, our data do not support a prominent role of common ABCB4 and ABCB11 variation on genetic CCA risk. However, taking into account data from paediatric cohorts and recent reports on CCA occurrence in families with ABCB4 mutations, resequencing of candidate gene loci appears warranted to adequately fully assess the potential contribution of hepatobiliary transporter variation on CCA susceptibility. 3, 4 References: 1. Wadsworth et al. Dig Dis 2011 2. Wasmuth et al. Gut 2007 3. Strautnieks et al. Gastroenterology 2008 4. Tougeron et al. J Hepatol 2012

Disclosures:

The following people have nothing to disclose: Vincent Zimmer, Florentina Mihalache, Marcin Krawczyk, Bianca Simon, Frank Grunhage, Monica Acalovschi, Frank Lammert

2182

  1. Top of page

Progenitor cell markers in hepatocellular carcinoma: clinico-pathological correlations and prognostic value

Evangelia Fatourou1, Despoina Karandrea2, 4, Ioannis Koskinas5, Marina Paleologou2, Ourania-Thaleia Syminelaki2, Menelaos Karanikolas7, Evangelos Felekouras6, Efstathios Antoniou6, Emanuel K. Manesis5, Joanna Delladetsima3, Dina Tiniakos2
1Institute of Liver Studies, King's College Hospital, London, United Kingdom; 2Laboratory of Histology-Embryology, MedicalSchool, National & Kapodistrian University of Athens, Athens, Greece; 31stDepartment of Pathology, MedicalSchool, National & Kapodistrian University of Athens, Athens, Greece; 4Department of Pathology, Aretaieion Hospital, MedicalSchool, National & Kapodistrian University of Athens, Athens, Greece; 52nd Department of Internal Medicine, MedicalSchool, National & Kapodistrian University of Athens, Athens, Greece; 61st and 2nd Department of Surgery, MedicalSchool, National & Kapodistrian University of Athens, Athens, Greece; 7Anaesthesiology, Washington University School of Medicine, Saint Louis, Washinghton, WA

Background/aims: Hepatocellular carcinoma (HCC) is a heterogeneous disease of distinct clinical subgroups. Keratin 19 (K19) immunopositivity has been proposed to represent hepatic progenitor cell (HPC) origin of HCC and has been correlated with poorly differentiated histology and aggressive tumor behaviour. Similarly, Epithelial Cell Adhesion Molecule (EpCAM) immunophenotype has been used to classify HCC into different subtypes with prognostic implication and may represent a possible biomarker for HPC origin. Methods: We assessed the expression of K19 and EpCAM by immunohistochemistry in a series of seventy-four Greek patients with HCC (mean age 65. 52±10. 8 years, male 72. 9%) followed up for 39. 6±25. 3 months in correlation with clinico-pathological parameters and survival. HCCs were considered K19-positive or EpCAM-positive if >5% tumor cells were immunostained. Results: K19-specific and EpCAM-specific immunoexpression were detected in tumor cells of 8/74(10. 81%) and 21/74 (28. 37%) HCCs, respectively. K19 immunopositivity tended to correlate with micro-vascular invasion (p=0. 065) but there was no correlation with other clinico-pathological parameters. EpCAM immunopositivity was significantly correlated with advanced TNM stage (p=0. 007). In univariate analysis, decreased recurrence-free survival (RFS) was associated with K19 immunopositivity (p<0. 001), advanced TNM stage > III (p=0. 024) and microvascular invasion (p=0. 005). Similarly, decreased overall survival (OS) was significantly associated with K19 immunopositivity (p=0. 03) and microvascular invasion (p=0. 014). In multivariate analysis, K19 positivity was the only independent predictor of RFS (OR=7. 534, 95%CI=2. 47322. 935; p<0. 001) and OS (OR=3. 317, 95%CI = 1. 1959. 208; p=0. 021). EpCAM positivity did not show any correlation with recurrence-free or overall survival. Conclusions: K19 but not EpCAM immunoexpression is an independent predictor of patient survival in a cohort of Greek patients with HCC and could be used to subgroup HCCs according to tumor aggressiveness.

Disclosures:

The following people have nothing to disclose: Evangelia Fatourou, Despoina Karandrea, Ioannis Koskinas, Marina Paleologou, Ourania-Thaleia Syminelaki, Menelaos Karanikolas, Evangelos Felekouras, Efstathios Antoniou, Emanuel K. Manesis, Joanna Delladetsima, Dina Tiniakos

2183

  1. Top of page

Hexokinase II Immunoexpression in Hepatocellular Carcinoma: Evidence for an Increased Level in Disease Progression, and Biologically Aggressive Tumors

Grace Guzman1, Rohini Chennuri1, Alexander Chan1, Bryan Rea1, Hui Xie1, Nissim Hay1, Scott Cotler2
1Department of Pathology M/C 847, Univ Illinois, Chicago, L; 2Medicine, Loyola University Maywood, IL

Abstract Background: The mitochondrial high affinity hexokinase type II (HKII) is expressed in hepatocellular carcinoma (HCC). Normal hepatocytes exhibit low affinity hexokinase (glucokinase [HKIV]) but during oncogenesis, there is a switch from HKIV to HKII expression. The aims of this study were to compare the immunoexpression of HKII in non-dysplastic cirrhosis (NDC), liver cell dysplasia in cirrhosis (LCDC), HCC, and normal liver control tissues, and to correlate HKII expression with clinical and histopathological parameters. Design: Immunohistochemistry was performed on a liver cancer progression tissue array consisting of specimens from 161 liver explants with cirrhosis, including 45 with HCC, 108 without HCC, 143 with dysplasia, and 8 normal liver control tissues. HCC specimens were characterized with regard to tumor grade and morphology. HKII expression was quantified as positive pixel counts/square millimeter (ppc/mm2) by image analysis. Results: There was a stepwise increase of HKII level from normal liver tissue to NDC, to LCDC, and to HCC (p=0. 001). HKII levels were significantly higher in areas of LCDC versus NDC (mean difference 13. 44 ±23. 30, p=0<. 001); and in LCDC and HCC versus NDC (mean difference 15 ±23. 93, p=0. 007). HKII levels were similar in LCDC and HCC (mean difference 7. 60 ±42. 43, p=0. 124). HKII levels were higher in grade 2 to 4 (71. 40 ±32. 66) versus grade 1 HCCs (48. 96 ± 35. 84) (p=0. 044); and in pleomorphic (85. 22 ±35. 85) versus nonpleomorphic HCC variants (57. 03 ±25. 33) (p=0. 041). P-values for LCDC and HCC versus NDC and tumor grade remained significant in multivariate analysis adjusting for lymphovascular tumor invasion, age, gender, etiology of chronic liver disease, body mass index (BMI), and ethnicity. Conclusions: Higher levels of HKII immunoexpression in LCDC and HCC compared to NDC suggest that upregulation of HKII is correlated with hepatocarcinogenesis in humans. Higher levels of HKII are associated with more aggressive histological features in HCC.

Disclosures:

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

The following people have nothing to disclose: Grace Guzman, Rohini Chennuri, Alexander Chan, Bryan Rea, Hui Xie, Nissim Hay

2184

  1. Top of page

Comparative Study of Percutaneous Radiofrequency Ablation and Hepatic Resection for Small, Poorly Differentiated Hepatocellular Carcinomas

Hiroya Iida, Tukasa Aihara, Sinichi Ikuta, Naoki Yamanaka
Meiwa hospital, Nishinomiya, Japan

Aim: Histologically, poorly differentiated hepatocellular carcinomas (HCC) are considered highly malignant. Here, we aimed to evaluate the relative efficacy and safety of hepatic resection or radiofrequency ablation (RFA) for treating this malignancy. Methods: Between April 2004 and May 2011, we enrolled 48 patients who had poorly differentiated HCCs that had been diagnosed postoperatively by pathological assessment. All the tumors had a maximum diameter <3 cm and all patients had <3 tumors. Fifteen of these patients underwent hepatic resection (HR group) and 33 patients underwent RFA (RF group). The patient background, tumor characteristics, overall survival rate, and recurrence free survival rate were assessed in both groups. Results The mean maximum tumor diameter was 2. 5 cm and 2. 0 cm in the HR and RF groups, respectively. The prothrombin activity level was 94% and 86% in the HR and RF groups, respectively. The 1-, 3-, and 5-year cumulative survival rates were 89. 1%, 68. 7%, and 68. 7%, respectively, in the HR group, and 59. 2%, 40. 9%, and 32. 7%, respectively, in the RF group. The 1-, 3-, and 5-year recurrence free survival rates were 85. 1%, 64. 8%, and 48. 6%, respectively, in the HR group, and 29. 0%, 7. 2%, and 7. 2%, respectively, in the RF group. There was a significant difference between these groups (p < 0. 05). Conclusion As hepatic resection has greater efficacy than RFA in the treatment of poorly differentiated HCC, even in cases with a small tumor size, we recommend its use for this malignancy.

Hepatic resection group (n=15)RFA group (n=33)p-value
Gender (malerfemale)14: 121: 120. 03
Etiology (HBV: HCV: NBNC)2: 12: 15: 22: 60. 54
Preoperative TAE (%) 40. 081. 80. 006
Maximum tumor diameter (cm)2. 5±0. 42. 0±1. 10. 003
Number of tumors1. 2±0. 61. 5±0. 90. 39
Albumin level (g/dl)4. 0±0. 43. 8±0. 30. 11
AST (IU/L)47±2650±190. 34
ALT (IU/L)54±5543±220. 84
Total bilirubin level (mg/dl)0. 9±0. 40. 7±0. 30. 09
Platelet count (x10000/μl)14. 3±4. 412. 8±6. 60. 17
Prothrombin activity (%)94±886±90. 008
AFP (ng/ml)3501820172±2960. 56
DCP (mAU/ml)665±1750311±9450. 13
ICGR 15 (%)15. 5±10. 819. 9±11. 10. 14
Child-Pugh (A: B: C)14: 1: 028: 5: 00. 41
Age (years)67±869±100. 62

Disclosures:

The following people have nothing to disclose: Hiroya Iida, Tukasa Aihara, Sinichi Ikuta, Naoki Yamanaka

2185

  1. Top of page

Influence of repeated radiofrequency ablation therapy on long-term liver function in patients with HCV-related hepatocellular carcinoma

Koichi Honda1, Masataka Seike1, Junya Oribe1, Mizuki Endo1, Mie Yoshihara1, Hiroki Syo1, Masao Iwao1, Masanori Tokoro1, Junko Nishimura1, Tetsu Mori1, Tsutomu Yamashita2, Satoshi Fukuchi2, Toyokichi Muro2, Kazunari Murakami1
1Gastroenterology, Oita University Yufu City, Japan; 2Gastroenterology and Hepatology, Oita Medical Center, Oita City, Japan

(Aim) Radiofrequency ablation (RFA) is currently recognized as an effective local treatment of hepatocellular carcinoma (HCC). However, there are a few reports of altered long-term liver function after RFA, especially it is unclear whether long-term liver function after repeated RFA treatments is affected. Furthermore, there is no report concerning the long-term liver function after RFA in patients achieving sustained viral response (SVR) by interferon therapy. This study examined the factors that influence the long-term changes in liver function after RFA for HCVrelated HCC. Moreover, the alteration of long-term liver function of SVR patients after RFA and the influence of repeated RFA were examined. (Method) This retrospective study included 176 patients who underwent RFA for HCV-related HCC. The inclusion criteria for patient selection were as follows: HCC occurring in HCV-related chronic liver disease; the presence of up to four nodules per patient, with each nodule having a diameter less than 5cm, and presence of tumors only in the liver, with complete necrosis achieved by treatment with RFA. HCC patients with a coexisting condition, chronic disease, or evidence of tumor progression were excluded. The enrolled patients were divided into two groups: patients achieving a SVR with interferon therapy for HCV (SVR group: 15 patients) and patients with persistent chronic active hepatitis (CAH group: 161 patients). When the HCC became uncontrollable, the observations were ended. Factors that influenced the changes in liver function after RFA were examined. The cumulative rate of a worsening Child-Pugh score (CPs) (2-point increase) was calculated using the Kaplan-Meier method and multivariate analysis was conducted using Cox's proportional hazards model. (Results) CPs worsening was identified in 37 patients during the study period (39. 3±26 months). The cumulative CPs aggravation rate was 8. 1% at 1 year, 20. 4% at 3 years, and 30. 4% at 5 years. During follow-up, no one in the SVR group showed CPs worsening and there was a tendency to be lower rate of CPs worsening in the SVR group compared with the CAH group (log-rank test, p=0. 056). Multivariate analysis identified a Cp class B (p=0. 000), platelet count <100000/mm3 (p=0. 011) and AST <40 IU/L (p=0. 031) as significant independent factors contributing to CPs worsening. Repeated RFA was not related to the aggravation of long-term liver function. (Conclusion) In the long term, liver function deteriorates gradually in patients with a poor hepatic functional reserve, low platelet count or high ASL level. Liver function was maintained in the patients who achieve SVR, even when RFA was performed repeatedly.

Disclosures:

The following people have nothing to disclose: Koichi Honda, Masataka Seike, Junya Oribe, Mizuki Endo, Mie Yoshihara, Hiroki Syo, Masao Iwao, Masanori Tokoro, Junko Nishimura, Tetsu Mori, Tsutomu Yamashita, Satoshi Fukuchi, Toyokichi Muro, Kazunari Murakami

2186

  1. Top of page

Immunotherapy Generates CD4+ and CD8+ T Cell Immune Responses to ASPH in Hepatocellular Carcinoma

Yoshito Tomimaru1, Stephen H. Gregory2, Howard Safran2, Kevin P. Charpentier3, Jack R. Wands1
1Department of Medicine, Liver Research Center, RI Hospital and the Warren Alpert Medical School of Brown University, Providence, RI; 2Department of Medicine, RI Hospital and the Warren Alpert Medical School of Brown University, Providence, RI; 3Department of Surgery, RI Hospital and the Warren Alpert Medical School of Brown University, Providence, RI

Background: Hepatocellular carcinoma (HCC) has a poor prognosis due to a high recurrence rate and lack of effective systemic therapy. Aspartate-β-hydroxylase (ASPH) is a highly conserved transmembrane protein overexpressed in 95% of human HCC and promotes cell migration and invasion. We have previously shown that immunization with ASPH-loaded dendritic cells (DCs) has anti-tumor effects and reduced HCC growth in murine models through activation of CD4+ and CD8+ antigen-specific T cells. Methods: In the present study, we determined if ASPH protein-loaded CD14+ monocyte derived DCs could lead to significant activation of antigen-specific CD4+ helper and CD8+ cytotoxic T lymphocytes (CTLs) in peripheral blood mononuclear cells (PBMC) derived from healthy donors and patients with HCC. DCs isolated from healthy donors and HCC patients were loaded with ASPH protein. DCs were also loaded with alpha-fetoprotein (AFP) (antigen control); antigen unloaded DCs were used as an additional control. Helper T cells and CTLs were co-incubated with the antigen-loaded DCs, and stimulation was subsequently evaluated by flow cytometric analysis using activation markers (CD154 and IFNy expression for helper T cell; CD137 and IFNy expression for CTL activity). Results: The percent of CD4+ T cells expressing CD154 and INFy was significantly increased following stimulation with ASPH-pulsed DCs compared to those stimulated with AFP-loaded DCs or DCs without antigen-loading. Furthermore, PBMCs incubated ASPH-pulsed DCs resulted in significantly increased CD137+ and INFy+ CTLs compared to stimulation with AFP-pulsed DCs or DCs without antigen loading. Conclusion: These results demonstrate that ASPH protein is a highly immunogenic antigen that can lead to significant T cell activation against ASPH expressing HCC tumors.

Disclosures:

The following people have nothing to disclose: Yoshito Tomimaru, Stephen H. Gregory, Howard Safran, Kevin P. Charpentier, Jack R. Wands

2187

  1. Top of page

Clinicopathological features of liver neoplasms with neuroendocrine differentiation in surgically resected liver tumors

Yoriko Nomura1, 3, Osamu Nakashima4, Masayoshi Kage1, Hisafumi Kinoshita2, Hirohisa Yano3
1pathology Kurume univercity, Kurume, Japan; 2Surgey Kurume university, Kurume, Japan; 3Department of pathology, Kurume University Kurume, Japan; 4Department of clinical inspection, Kurume university, Kurume, Japan

Background and Aim: In general, neoplasms with neuroendocrine differentiation including neuroendocrine tumors (NET) and neuroendcrine carcinomas (NEC) are rare, but the prognosis of NEC is often poor. In the liver, NET or NEC is not listed in WHO classification and its clinicopathological features, including frequency and biologic features, are largely unknown. In the present study, we addressed this issue using a large number of surgically resected liver tumors in our institute. Material and Methods: We used 1071 tumors from 913 patients who undergone curative hepatectomy for liver neoplasms, at Kurume University Hospital during January 2000 and April 2011. Paraffin sections were made from the entire parts of the tumor at the largest diameter and stained with hematoxylin-eosin. Pathological diagnosis was performed according to the established diagnostic criteria of NET (e. g., NET G1 and NET G2) and NEC (e. g., small cell and large cell) in other digestive organs. When NET or NEC was suspected by the observation of HE-stained sections, we performed immunostaining for chromogranin A, synaptopysin, Keratin 19 (K19), and Ki-67. All slides were evaluated at least 2 pathologists (Y. N. and O. N.). Clinicopathological features of NEC were also compared with those of 18 cases of poorly differentiated hepatocellular carcinoma (pHCC) in this series. Results: There were 4 cases (0. 37%) that consisted solely of NET G1, but there were no cases that consisted solely of NET G2 or NEC. There are 5 cases (0. 46%) showing the co-existence of large cell NEC with HCC, which was designated as NHCC. Large cell NEC components were often intermingled with pHCC component of the tumor, suggesting the transdifferentiation of HCC into large cell NEC. NET G1 and NEC components were positive for chromogranin A and synaptophysin. K19 expression was significantly higher in HCC component (60%) and NEC component (40%) of NHCC than in pHCC (11. 1%). Ki-67 labeling indices of NET G1, pHCC, and HCC and NEC components of NHCC were 2. 4%, 18. 5%, 12. 6%, and 58. 9%, respectively, and NEC component of NHCC was significantly higher than the others. There are no significant difference in prognosis and the other clinical features between pHCC and NHCC. Conclusion: The incidence of primary NET and NEC are very low in surgically resected liver tumors. LC NEC occurs as a histological component of HCC and does not influence the patient prognosis.

Disclosures:

The following people have nothing to disclose: Yoriko Nomura, Osamu Nakashima, Masayoshi Kage, Hisafumi Kinoshita, Hirohisa Yano

2188

  1. Top of page

HCC Features in Patients with "Healthy” Liver: The Hepatocat Group Experience

Francesca Romana Ponziani1, Emanuele Rinninella1, Maria Assunta Zocco1, Matteo Garcovich1, Brigida E. Annicchiarico1, Massimo Siciliano1, Francesca D'Aversa1, Laura Riccardi1, Valentina Cesario1, Mariachiara Campanale1, Nicoletta de Matthaeis1, Marco Biolato1, Felice Giuliante3, Gennaro Nuzzo3, Maria Vellone3, Roberto Iezzi2, Anna Maria De Gaetano2, Alfonso W. Avolio3, Salvatore Agnes3, Fabio Maria Vecchio4, Antonio Grieco1, Maurizio Pompili1, Gian Ludovico Rapaccini1, Antonio Gasbarrini1
1Internal Medicine and Gastroenterology, A. Gemelli Hospital, Rome, Italy; 2Radiology, A. Gemelli Hospital, Rome, Italy; 3Surgery, A. Gemelli Hospital, Rome, Italy; 4Histopatholoqy/ A. Gemelli Hospital, Rome, Italy

BACKGROUND Hepatocellular Carcinoma (HCC) is one of the most common complications of cirrhosis or chronic hepatitis (chronic liver disease, CLD), but it can also be diagnosed in patients with healthy liver (HL). AIM To report the features of HCC in an Italian cohort of patients with CLD or HL. PATIENTS AND METHODS 420 Caucasian patients diagnosed with HCC and affected by CLD or with HL were selected from the database of the HEPATOCAT multidisciplinary group for the treatment of HCC. Age, sex, number and size of lesions, presence of diabetes or HIV infection or other neoplasms, alcohol or drugs use, were evaluated. Chi-square and Student's t-tests were used to compare data; a two-sided P value of 0. 05 was considered statistically significant. RESULTS Patients' characteristics are shown in Table 1. 405 patients were affected by CLD, 15 presented with HL. There was no difference in sex, prevalence of diabetes, other neoplasms and HIV infection, alcohol and drug use between the two groups. Patients with HL were younger (mean age 58 vs. 65 years, p=0. 014). HCC in patients with HL was as always a single lesion (p=0. 044) larger than 5 cm or infiltrating the whole liver (p<0. 0001). Mean tumor dimension were 88 mm in patients with HL vs. 39 mm in patients with CLD (p<0. 0001). Moreover, alpha-fetoprotein (AFP) values were higher (7218 vs. 1620 ng/ml p=0. 028) CONCLUSION The present study shows that HCC in absence of underlying liver disease is more frequent in young patients, and presents often with single or infiltrating lesions of big size and high AFP levels. This could be the expression of a different process of carcinogenesis.

Patients'characteristics

AllCLDHLP value
Gender (Male/Female)331/89321/8410/50. 241
Age (mean 土 SD)65 ±1165 ±1158± 170. 014
years <=409/420 (2%)7/405 (17%)2/15(1. 3%)0. 004
years 40-60105/420(25%)99/405 (24%)6/15(40%)
years >60306/420 (73%)299/405 (73%)7/15(46%)
AFP (mean 土 SD)1809187001620 ±77007218 ±236610. 028
Single lesion228/420 (54%)216/405 (53%)12/15 (80%)0. 044
Multinodular/ Infiltrating192/420 (46%)189/405 (47%)3/15(7%)
Tumor size (max mm; mean 土 SD)40 ±2939 ±2888 ±31pく0. 0001
Tumor <5 cm322/420 (77%)321/405 (79%)1/15 (7%)p<0. 0001
Tumor >=5 cm98/420 (23%)84/405 (21%)14/15 (93%)
Alcohol use103/420 (24. 5%)102/405 (25%)1/15 (7%)0. 102
FD/HIV14/420 (3%)14/405 (3. 5%)0/15 (0%)0. 464
Diabetes81/420(19%)77/405 (19%)4/15(27%)0. 461
Other cancer18/420(4%)18/405(4%)0/15 (0%)0. 404

Disclosures:

The following people have nothing to disclose: Francesca Romana Ponziani, Emanuele Rinninella, Maria Assunta Zocco, Matteo Garcovich, Brigida E. Annicchiarico, Massimo Siciliano, Francesca D'Aversa, Laura Riccardi, Valentina Cesario, Mariachiara Campanale, Nicoletta de Matthaeis, Marco Biolato, Felice Giuliante, Gennaro Nuzzo, Maria Vellone, Roberto Iezzi, Anna Maria De Gaetano, Alfonso W. Avolio, Salvatore Agnes, Fabio Maria Vecchio, Antonio Grieco, Maurizio Pompili, Gian Ludovico Rapaccini, Antonio Gasbarrini

2189

  1. Top of page

HCC in Italian HBV/HCV Infected Patients: The Hepatocat Group Experience

Francesca Romana Ponziani1, Emanuele Rinninella1, Maria Assunta Zocco1, Matteo Garcovich1, Brigida E. Annicchiarico1, Massimo Siciliano1, Laura Riccardi1, Valentina Cesario1, Mariachiara Campanale1, Nicoletta de Matthaeis1, Marco Biolato1, Anna Maria De Gaetano2, Felice Giuliante3, Gennaro Nuzzo3, Maria Vellone3, Roberto Iezzi2, Alfonso W. Avolio3, Salvatore Agnes3, Fabio Maria Vecchio4, Antonio Grieco1, Maurizio Pompili1, Gian Ludovico Rapaccini1, Antonio Gasbarrini1
1Internal Medicine and Gastroenterology, A. Gemelli Hospital, Rome, Italy; 2Radiology, A. Gemelli Hospital, Rome, Italy; 3Surgery, A. Gemelli Hospital, Rome, Italy; 4Histopathology/ A. Gemelli Hospital, Rome, Italy

BACKGROUND Hepatocellular Carcinoma (HCC) is one of the most common complications of HCV and HBV-related liver disease. Young age, large and multinodular HCCs are typical of HBV patients, which accounts for 50% of the cases, while the presence of cirrhosis is the most common characteristic in HCV ones. AIM to report the epidemiology and features of HCC in an Italian single center comparing HBV and HCV patients. PATIENTS AND METHODS Two-hundred and sixty-six Caucasian cirrhotic patients with HCV or HBV cirrhosis or chronic hepatitis were selected from the database of the HEPATOCAT multidisciplinary group for the treatment of HCC. Patients with non HCC tumors were excluded from the analysis. Age, sex, BCLC, number and size of lesions, Child-Pugh score when applicable, vascular invasion and metastasis were evaluated. Chi-square and Student's t-tests were used to compare data; a two-sided P value of 0. 05 was considered statistically significant. RESULTS Patients' characteristics are shown in Table 1. 192 (72. 2%) patients were HBV infected and 74 patients (27. 8%) HCV infected. HBV-related liver disease seems associated to a younger age (p<0. 001) and larger HCCs (p=0. 005) than HCV, while no difference was found in degree of liver dysfunction, number of lesions and vascular invasion. Furthermore, metastases were more common among patients with HBV infection (p=0. 007). CONCLUSION In our series of Italian patients, HBV-related HCC is more common in young people and more aggressive than HCV-related one, since it presents with lesions of larger size and more prone to develop metastases.

Patients' characteristics

AllHBVHCVP value
Gender (Male/Female)203/6358/16145/470. 623
Age (mean ± SD)66 ±1062 ±1168 ±10<0. 001
years <=404/257 (1. 5%)3/71 (4. 2%)1/186(0. 5%)0. 013
years 40-6059/257 (21. 8%)22/71 (31%)37/186 (20%)
years >6059/257 (21. 8%)22/71 (31%)37/186 (20%)
cirrhosis261/266 (98%)71/74(96%)190/192 (99%)0. 105
CHILD A/B/C181/57/16/745/20/4/2136/37/12/50. 506
AFP (mean 士 SD)1479 ±4502676 ±10200996 ± 46970. 091
<=20113/226 (50%)34/65 (52%)79/161 (49%)0. 462
21-20066/226 (30. 5%)15/65 (23%)51/161 (32%)
201-100021/226 (9%)6/65 (9%)15/161 (9%)
>100026/226(11%)10/65 (15%)16/161 (10%)
BCLC 0/A/B/C/D35/118/58/34/216/32/21/10/529/86/37/24/160. 372
Single lesion143/266 (54%)39/74 (53%)104/192 (54%)0. 830
Multinodular123/266 (46%)35/74 (47%)88/192 (46%)
Tumor <5 cm209/263 (79. 5%)50/74 (67. 5%)159/189 (84%)0. 003
Tumor >5 cm54/263 (20. 5%)24/74 (32. 5%)30/189(16%)
Macrovascular invasion35/266 (13%)13/74(17. 5%)22/192(11. 5%)0. 079
Metastases11/266 (4%)7/74 (9. 5%)4/192 (21%)0. 007

Disclosures:

The following people have nothing to disclose: Francesca Romana Ponziani, Emanuele Rinninella, Maria Assunta Zocco, Matteo Garcovich, Brigida E. Annicchiarico, Massimo Siciliano, Laura Riccardi, Valentina Cesario, Mariachiara Campanale, Nicoletta de Matthaeis, Marco Biolato, Anna Maria De Gaetano, Felice Giuliante, Gennaro Nuzzo, Maria Vellone, Roberto Iezzi, Alfonso W. Avolio, Salvatore Agnes, Fabio Maria Vecchio, Antonio Grieco, Maurizio Pompili, Gian Ludovico Rapaccini, Antonio Gasbarrini

2190

  1. Top of page

Outcomes in Veterans with Hepatocellular Carcinoma (HCC) Receiving Trans-Arterial Chemoembolization (TACE) - a Single Center Experience

Amar Mahgoub2, Thomas J. Maust2, Kelly McMaken1, Astrid Knott12, Eric Dieperink1, 2, Paul Thuras1, 2, Christine Pocha1'2
1Minneapolis VAHCS, Minneapolis, MN; 2University of Minnesota, Minneapolis, MN

Purpose: TACE improves survival in patients with HCC. We investigated which patient and treatment related factors influ- ence outcome of TACE in Veterans with HCC. Methods: Between 2000-2011, 69 of 343 patients with HCC (ICD-9: 155.0 received at least 1 TACE procedure at the Minneapolis VA Health Care System. Retrospectively, data including demographic and clinical characteristics, treatment modality and outcome were collected. Chi-Square and logistic regression were used to determine factors associated with mortality. Results: Of the 69 all male patients, mean age of 61, 71% were Caucasian, 65% had hepatitis C, and 77% had compensated liver disease with a mean MELD score of 9 at the time of HCC diagnosis. All-cause mortality was 68% with a mean follow up time of 25 mos. The number of lesions (1, 2, 3, multifocal), the number of TACE procedures received (range 1-13), albumin < 3mg/dl, decompensated status (15/69), Child Pugh Score (A: 56, B: 13) and etiology of liver disease did not influence treatment outcome. In addition to TACE 38 patients received a combination of treatment modalities (radiofrequency ablation: 20, body radiation: 2, percutaneous ethanol injection: 8, percutaneous acetic acid injection: 2, sorafenib: 17) without significant impact on mortality. Of the total of 174 TACE procedures performed 27(15. 5%) used a single agent, 23 (13. 2%) used 2 agents, 123 (70. 7%) used 3 agents in combination and 1 used agents. Of the 3 agent combinations; cisplantin, doxorubicin, and mitomycin were used most commonly (79; 45. 4% of all TACE) followed by cisplantin, doxorubicin, and adriamycin (42; 24. 1% of all TACE). Thus, these two 3 drug combinations account for approximately 70% of all treatments. AFP < 20ng/ml correlated with improved survival; AFP > 400ng/ml predicted significantly lower survival (log rank 3. 8 df 1 p < 0. 05). Choice and combination of chemotherapy agents (95. 7% cisplatin, 95. 7% mitomycin, 66. 7% doxorubicin, 30. 4% adriamycin, 1. 4% metrotrexate) had no influence on survival. Severe adverse events related to TACE were rare (1 abscess, 1 death). Most common co-morbidities were cardiac disease (22%), substance use disorders (32%), alcohol abuse (51%), depression (35%), and metabolic syndrome (42%) but none increased all-cause mortality. Conclusions: AFP is the primary factor that predicted mortality and those with levels above 400ng/ml may require more aggressive treatment. Given the large number of drug combinations and overall small patient numbers conclusions about superiority of any drug or combination is limited. Further research on the ideal drug combination and patient selection is warranted.

Disclosures:

The following people have nothing to disclose: Amar Mahgoub, Thomas J. Maust, Kelly McMaken, Astrid Knott, Eric Dieperink, Paul Thuras, Christine Pocha

2191

  1. Top of page

Predictors of Hepatocellular carcinoma recurrence post thermal ablation

Mohamed A. Chinnaratha1, Dharshan Sathananthan2, Puraskar Pateria3, Edmund Tse2, Gerry C. MacQuillan3, Jonathan Tibballs4, Alan J. Wigg1
1Hepatology/Liver Transplant Medicine unit, Flinders Medical Centre, Bedford Park, SA, Australia; 2Gastroenterology/Hepatoogy, Royal Adelaide Hospital, Adelaide, SA, Australia; 3Gastroenterology/Hepatology, Sir Charles Gairdner Hospital, Nedlands, WA, Australia; 4Radiology, Sir Charles Gairdner Hospital, Nedlands, WA, Australia

Percutaneous thermal ablation (PTA) is widely used as a curative option in subjects with early stage (BCLC-A) hepatocellular carcinoma (HCC). Despite successful ablation with complete radiological response, the risk of HCC recurrence remains high. Our primary aim therefore was to evaluate factors predicting the recurrence of HCC post PTA. METHODS: Multi-centre retrospective study of patients treated with PTA (Radiofrequency Ablation [RFA] and Microwave Ablation [MWA]) between Jan 2006 - Dec 2012. Subjects included were consecutive patients who had PTA with curative intent. Subjects who had other loco-regional therapies prior to PTA or with evidence of macrovascular invasion were excluded. Primary end point was the identification of factors predicting overall intrahepatic recurrence (IHR) using uni and multivariate analysis. IHR included both recurrence due to local tumour progression [LTP] and intrahepatic distant recurrence [IDR]. Secondary endpoints were rate of IHR (both LTP and IDR), recurrence free survival and the adverse event rate (<30 days from the procedure requiring hospitalization). RESULTS: Ninetythree subjects [mean age (±SD): 62. 7 (±10. 1) years, 77. 4% males] were included in the study. 91. 2% had cirrhosis and HCV (29%), HBV (18. 3%) accounted for majority of the liver disease. 11. 8% had more than one nodule and the overall mean (±SD) tumour diameter was 26. 1 (13. 3) mm. 73. 1% had RFA and the mean (±SD) follow-up duration was 421. 3 (±396. 9) days. Overall IHR rate was 55. 9% during the followup period with LTP in 33. 3%, IDR in 29% and 6. 5% had both. Overall median (±SE) recurrence free survival was 422 (±48) days. Poorly differentiated HCC was the only independent predictor of overall IHR [HR (95% CI): 6. 1(1. 9-19. 2), p=0. 002], LTP [9. 8 (2. 3-41. 3, p=0. 002] and IDR [5. 3 (1. 2-22. 9), p=0. 03]. There was a trend towards early IHR in patients having MWA compared to RFA [median (±SE) days: 399 (±32) v 554 (±111) days, p=0. 06). This was more evident in single tumours less than 30mm where the recurrence was significantly earlier in those having MWA [median (±SE) days: 399 (±37) v 568 (±120) days, p=0. 02]. Overall, 11. 8% had an adverse event and this was higher in the MWA group compared to RFA but, not significant (25% v 9. 7%, p=0. 14). There were no procedure related deaths in this cohort. CONCLUSION: Poorly differentiated HCC is an important, independent predictor of overall IHR, LTP and IDR post PTA. Trends towards earlier recurrence in patients having MWA, together with a higher adverse event rates in the MWA group raise concerns about the efficacy and safety of this technique relative to RFA in real world settings and require further study.

Disclosures:

The following people have nothing to disclose: Mohamed A. Chinnaratha, Dharshan Sathananthan, Puraskar Pateria, Edmund Tse, Gerry C. MacQuillan, Jonathan Tibballs, Alan J. Wigg

2192

  1. Top of page

Evaluation of hepatocellular carcinoma risk in patients with chronic hepatitis C by EOB-MR imaging

Shunsuke Nojiri, Mio Endo, Noboru Shinkai, Kei Fujiwara, Takashi Joh
Department of Gastroenterology and Metabolism, Nagoya City University Graduate School of Medical Sciences, nagoya, Japan

Background/aims: It is known that advanced fibrosis increases the risk of carcinogenesis, and it is important to diagnose liver fibrosis in order to predict the risk of liver cancer. Liver biopsy is the gold standard test to determine the grade of fibrosis, but there are some associated problems, such as sample error and severe complications. We recently reported the accuracy of staging fibrosis of chronic hepatitis of HCV infection using gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA) MR imaging, revealing that the EOB-MRI index decreases as fibrosis progresses, but there have been no studies predicting the risk of HCC using these methods. Thus, the aim of this study was to evaluate the efficacy of EOB-MR imaging as a predictor of HCC development. Materials and Methods: Between August 2008 and 2009, we studied 142 HCV-infected patients (male: 71, female: 71), excluding those with HCC or a past history of it, who underwent EOB-MRI in our hospital. The age, gender, alfa-fetoprotein (AFP), serum albumin, total bilirubin, prothrombin time, total cholesterol, platelet count, serum AST, ALT, and EOB-MRI index were investigated. The EOB-MRI index was calculated as we described previously: the ROIs (regions of interest) of the liver and the intervertebral disc were measured for each MR image (1. 5Tesla MR system) in plane (pre) and liver-biliary (post) phases. EOB-MRI index= (post-liver intensity / post-disc intensity) / (preliver intensity / pre-disc intensity). Results: The mean age was 66. 1 ±12 years. The median follow-up period was 3. 1 years. 109 patients were observed until the end of the study period (31 December, 2012). In the follow-up period, HCC occurred in 21 patients. The cumulative occurrence rates were 2. 1, 9. 1, and 14. 1% at 1, 2, and 3 years, respectively. The age, AST, and AFP were higher and albumin, total cholesterol, prothrombin time, platelet count and EOB-MRI index were lower in the HCC occurrence than in no-occurrence group at the baseline. A lower EOB-MRI index led to a significantly higher risk based on the hazard ratio (95%CI) compared to an EOB-MRI index≥1. 601; index≤ 1. 31: 17. 34 (2. 16-138. 7, P=0. 007), 1. 311 to 1. 38: 11. 5 (1. 2-103, p=0. 02), 1. 381 to 1. 50: 7. 24 (0. 89-58. 9, P=0. 06), 1. 501 to 1. 60: 2. 68 (0. 17-42. 9, P=0. 48). With an optimal cutoff value of the EOB-MRI index of 1. 46, on multivariate analysis, only an EOB-MRI index lower than 1. 46 was identified as an independent risk factor: HR 6. 80 (1. 51-30. 6, p=0. 012). Conclusion: The EOB-MRI index is associated with the risk of HCC occurrence in hepatitis C patients. The EOB-MRI index may be a substitute for liver biopsy when evaluating this risk.

Disclosures:

The following people have nothing to disclose: Shunsuke Nojiri, Mio Endo, Noboru Shinkai, Kei Fujiwara, Takashi Joh

2193

  1. Top of page

Prognostic Scoring Systems for Predicting PostChemoembolization Outcomes in Patients with NonMetastatic Hepatocellular Carcinoma

Yeonjung Ha1, Ju Hyun Shim2, Han Chu Lee2, Seungbong Han3, Kang Mo Kim2, Young-Suk Lim2, Dong Jin Suh2, 4
1Internal Medicine, Asan Medical Center, Seoul, Republic of Korea; 2Gastroenterology, Asan Liver Center, Asan Medical Center, Seoul, Republic of Korea; 3Biostatistics, Asan Medical Center, Seoul, Republic of Korea; 4Internal Medicine, Vievis Namuh Hospital, Seoul, Republic of Korea

Background & Aims: Transarterial chemoembolization (TACE) is an effective treatment option for patients with unresectable hepatocellular carcinoma (HCC). However, the treatment response and resulting survival is quite heterogeneous according to the tumor status and liver function. The purpose of this study is to construct prognostic nomograms capable of estimating individual probabilities of tumor progression and overall survival at specific times during serial TACE procedures. Methods: A total of 1, 181 consecutive patients with HCC who underwent repeated TACE treatments at Asan Medical Center were included in this study. No patient had extrahepatic metastasis or Child-Pugh class C disease. We assigned patients to two cohorts according to the first TACE date, i. e. a derivation set of 854 patients (January 2004 to December 2006) and a validation set of 327 patients (January 2007 to December 2007). The multivariate Cox proportional hazards models were developed and separately validated based on the pre-TACE clinical and laboratory covariates assessed for their association with one-year progression-free survival and three-year overall survival. The accuracy of the models was internally and externally evaluated using the time-dependent C-index of discrimination and a Hosmer-Lemeshow type test for calibration. Results: The median progression-free survival time of the derivation set was 13. 6 months and with a one-year progressionfree survival rate of 54. 2%. The three-year overall survival rate was 42. 8% with a median survival of 30. 2 months. Nomograms for progression-free survival and overall survival were built into the derivation set incorporating the following six independent prognostic factors: log10 (total tumor volume) calculated from 4/3 x 3. 14 (maximum radius of the nodule in cm)^3; tumor number; tumor type (nodular or infiltrative); portal vein invasion; Child-Pugh class (A or B); and (MELD score/10)^(−2). The models had good calibration and discrimination abilities with C-indices of 0. 76 (one-year progression) and 0. 80 (three-year survival), respectively. The results from the external validation using the test cohort also showed that these models performed well in terms of goodness-of-fit and discrimination (C-indices 0. 72 for one-year progression and 0. 80 for three-year survival). Conclusions: Our prognostic calculators which were developed to quantify the expected benefit of repeated TACE for non-metastatic HCC in terms of its progression and the survival outcomes can serve as useful clinical aids for counseling TACE-treated patients and for planning the further treatment schedule per patient.

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: Yeonjung Ha, Ju Hyun Shim, Seungbong Han, Kang Mo Kim, Young-Suk Lim, Dong Jin Suh

2194

  1. Top of page

Correlation between the Stromal Activated Hepatic Stellate Cells and Angiogenesis in Human Hepatocellular Carcinoma

Nan Lin1, Xu Linan2
1Department of Hepatobiliary Surgery, the Third Affiliated Hospital, Sun Yat-sen University Guangzhou, China; 2the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, China

Background and Aims: Angiogenesis is very important during the progression of hepatocellular carcinoma (HCC). In murine xenograft models, activated hepatic stellate cells (HSCs) are reported to be capable of promoting tumor angiogenesis. The effect of activated HSCs on angiogenesis in human HCC is not well understood. In the present study, we elucidated the correlation between activated stromal HSCs and angiogenesis in human HCC. Methods: We analyzed the activated HSCs in human HCC samples and the correlation with patients' clinicopathologic features and tumor angiogenesis using immunohistochemical methods. Results: We found that activated HSCs are mainly located in the blood sinus of HCC. The density of activated HSCs and microvessels was higher when the HCC was poorly differentiated or developed from liver cirrhosis. There was a positive correlation between the density of activated HSCs and microvessel density in HCC. Conclusions: Activated HSCs existed in human HCC stroma, the density of which was positively correlated with tumor microvessel density, indicating that activated HSCs participate in promoting angiogenesis in HCC.

Disclosures:

The following people have nothing to disclose: Nan Lin, Xu Linan

2195

  1. Top of page

Description of HNFlalpha-mutated hepatocellular adenomas with unusual morphological features

Christine Sempoux1, Nathalie Sturm2, Claire Castain, Valerie Paradis4, Nora Frulio5, Brigitte Le Bail3, 7, Jean Saric6, Charles Balabaud7, Paulette Bioulac-Sage3, 7
1Pathology CHU St Luc, Brussels, Belgium; 2pathology, CHU, Grenoble, France; 3Pathology, CHU Pellegrin, Bordeaux, France; 4Pathology, CHU, Clichy, France; 5Radiology, CHU St Andre, Bordeaux, France; 6Surgery, CHU St Andre, Bordeaux, France; 7Inserm U1053, Universite Bordeaux2, Bordeaux, France

Background. In general the correlation between HNF1 a mutation and the absence of liver fatty acid binding protein (LFABP) staining is excellent in hepatocellular adenomas (HCA). Moreover, LFABP negative HCA occurring mainly in women display very characteristic morphological features such as i) homogeneous steatosis often associated with clear cells, ii) absence of cellular atypia, and iii) no abnormal glutamine synthase (GS) staining. We recently observed two LFABP negative HCAs whose diagnosis could not be made on H&E due to the absence of classical histological criteria. Aim. To review LFABP negative HCAs lacking the classical histological criteria particularly the presence of steatosis. Patients. Cases of HNF1 a inactivated HCA (H-HCA) confirmed by LFABP negative staining from two series (Bordeaux, Brussels) and consults cases were reviewed (n= 80). This allowed the identification of 15 cases (3 men) in which steatosis was totally or largely absent. The diagnosis was either impossible to make on H&E or was barely suggested by the presence of a few fat droplets and/or clear cells. In 4 cases with multiple HCA/adenomatosis, the diagnosis was however likely because of the presence of other small nodules (at least one) typical of H-HCA. Five cases (2 men) were related to vascular diseases (Budd Chiari syndrome; portal vein thrombosis (2), one with umbilical vein catheterization; portal vein agenesis; Fontan procedure); among the 4 cases with borderline lesions or HCC, 3 belonged to this group. HCA were solitary (3. 5cm-11 cm) in 6 patients and multiple (up to adenomatosis) in 9. Results. Apart from the complete or almost complete absence of steatosis, two types of patterns were observed and were often mixed: 1- formation of rosettes (up to pseudo glands) particularly in zones without steatosis (80% of the cases) with rare cases expressing CK7 and normal or faint GS expression except in one case (patchy). Notably, only one case (with HCC expressing strongly GS) showed nuclear pcatenin staining. 2- multiple veins, sinusoidal dilatation, congestion, peliosis, fibrinoid deposits, mostly in foci, either combined or isolated, ranging from mild to severe, with normal or occasionally enlarged CD34 staining. Conclusions. a- HHCA with unusual features are rare (<20% of all H-HCA) and occur in some specific circumstances; b- H&E and MRI can miss the diagnosis of LFABP negative nodules mainly due to the lack of steatosis; c - the presence of rosettes is possibly more related to cholestasis than representing a criterion suggestive of malignancy; d- malignant transformation can, however, occur particularly in cases associated with vascular diseases.

Disclosures:

The following people have nothing to disclose: Christine Sempoux, Nathalie Sturm, Claire Castain, Valerie Paradis, Nora Frulio, Brigitte Le Bail, Jean Saric, Charles Balabaud, Paulette Bioulac-Sage

2196

  1. Top of page

The disappearance or decrease of tumor stain on contrast enhanced CT images after 2 weeks of sorafenib treatment predicts sorafenib efficacy in patients with advanced hepatocellular carcinoma

Teiji Kuzuya, Masatoshi Ishigami, Kazuhiko Hayashi, Takashi Honda, Hidemi Goto
Gastroenterology and Hepatology, Nagoya University, Nagoya, Japan

Objectives: The aim of this study was to investigate the relationship between clinical efficacy and the disappearance or decrease of tumor stain on contrast enhanced CT images after 2 weeks of sorafenib treatment for patients with advanced hepatocellular carcinoma (HCC). Methods: Subjects were 78 patients who began sorafenib therapy. Mean age was 67. 6 years, and HCC stage was III (n=32), IVa (n=23) and IVb (n=23). Contrast enhanced CT images were taken at baseline, after 2 weeks and after 6 weeks of sorafenib treatment. The antitumor responses were evaluated according to the modified RECIST criteria at 6 weeks after starting sorafenib therapy. Results: Responses by modified RECIST criteria were as follows; thirteen patients (1 8. 8%) achieved PR, 36 (52. 2%) had SD, and 20 (29. 0%) had PD. After 2 weeks of sorafenib treatment, the disappearance or decrease of tumor stain on contrast enhanced CT images was found in 45 (63. 4%) out of 71 patients who had hypervascular tumors. In patients with the disappearance or decrease of tumor stain on contrast enhanced CT images, the median time to progression (TTP) was significantly longer than in patients without the disappearance or decrease of tumor stain (2. 5 vs. 1. 5 months; hazard ratio, 2. 30; 95% CI, 1. 22 to 4. 33; p=0. 010). Conclusions: Evaluation of the disappearance or decrease of tumor stain on contrast enhanced CT images after 2 weeks of sorafenib treatment may be useful for predicting sorafenib efficacy in patients with advanced hepatocellular carcinoma.

Disclosures:

Kazuhiko Hayashi - Grant/Research Support: MSD, Astra Zenaca

Hidemi Goto - Grant/Research Support: MSD, Roche, Bayer, Bristol-Myers

The following people have nothing to disclose: Teiji Kuzuya, Masatoshi Ishigami, Takashi Honda

2197

  1. Top of page

Y90-Radioembolization plus sorafenib versus Y-90 Radioembolization alone for the treatment of unresectable hepatocellular carcinoma: a matched case-control study

Antonio Facciorusso1, Carlo Sposito1, Sherrie Bhoori1, Raffaele Romito1, Carlo Chiesa2, Marco Maccauro2, Tiziana Camerini3, Rosalba Miceli3, Carlo Morosi4, Carlo Spreafico4, Vincenzo Mazzaferro1
1Surgery and Hepatology, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy; 2Nuclear Medicine, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy; 3Trial Office and Biomedical Statistics, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy; 4Radiology, Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Milan, Italy

Trans-arterial radioembolization (TARE) with yttrium-90 induces tumor necrosis and delays progression of unresectable hepatocellular carcinoma (HCC). TARE+sorafenib combinations may further enhance objective tumor response. However, efficacy and safety of combo therapies have been scarcely studied. This study aimed at investigating in two case/time-matched cohorts the added value of sorafenib to TARE in unresectable HCC. Material and methods Out of 118 HCCs treated with TARE (Therasphere ®) between Dec 2010 and Dec 2012, a group of 15 patients who underwent the procedure while being on sorafenib (previously started in other Institutions) was identified (Group 1: study group). A second group of 30 patients who underwent TARE alone for similar tumor stages and well-balanced baseline liver function, was randomly extracted from a prospectively collected database (Group 2: controls). Treatment efficacy was investigated as progression-free survival (PFS) and overall survival (OS) by means of Kaplan Meier method and compared with the Log-rank test. Along with patients/tumor parameters, response rate (according to RECIST 1. 1 and EASL criteria), liver decompensation (LD) and time to LD were analyzed. Results Baseline characteristics of the two groups did not statistically differ in terms of liver function, tumor burden and performance status. 80% of patients in both groups were BCLC C stage because of portal vein thrombosis (PVT) and all patients were Child-Pugh A. Patients in Group 1 had started sorafenib at a median time of 2 months prior to TARE; median time on sorafenib was 9 months and median dose was 600 mg/day. Five patients (33%) withdrew sorafenib after TARE (four of them because of LD, one due to tumor progression) while the remaining 10 were on therapy at the time of death or last follow-up. No significant differences in patients' outcome among the two groups were identified (median PFS 6 months in TARE+sorafenib vs. 7 months in TARE alone, p=0. 935; median OS 10 vs. 10 months, p= 0. 711). Objective response rate (ORR: complete + partial response) according to RECIST was 45. 5% in Group 1 vs. 42. 3% in Group 2 (p=1. 0); ORR according to EASL was 10% in Group 1 vs. 40% in Group 2 (p=0. 12). LD occurred within 6 months from treatment in 53. 3% of Group 1 vs. 43. 3% in Group 2 (p=0. 526) and median time to decompensation was 5 vs. 8 months respectively (p=0. 518). No influence on results of PVT, dosing and tumor burden was found. Conclusion The association of sorafenib does not seem to delay progression nor prolong survival in HCC patients treated with TARE. Liver toxicity and decompensation does not differ among the two therapeutic regimens.

Disclosures:

Sherrie Bhoori - Speaking and Teaching: Bayer, Nordion Raffaele Romito - Speaking and Teaching: Nordion, Bayer Carlo Chiesa - Speaking and Teaching: NORDION M. D. S.

Carlo Morosi - Speaking and Teaching: Nordion

Vincenzo Mazzaferro - Advisory Committees or Review Panels: Bayer; Grant/Research Support: Nordion; Speaking and Teaching: Merck Serono S. p. A.

The following people have nothing to disclose: Antonio Facciorusso, Carlo Sposito, Marco Maccauro, Tiziana Camerini, Rosalba Miceli, Carlo Spreafico

2198

  1. Top of page

Predictors of Mortality in Patients with Hepatocellular Carcinoma undergoing Transarterial Chemoembolization

Pranab Barman1, Pratima Sharma1, Venkat Krishnamurthy3, 2, Jonathon Willatt3, 2, Heather McCurdy3, Richard H. Moseley3, 1, Grace L. Su3, 1
1Internal Medicine, University of Michigan, Ann Arbor, MI; 2Radiology, University of Michigan, Ann Arbor, MI; 3Internal Medicine, VA Ann Arbor Healthcare System, Ann Arbor, MI

Background: Transarterial chemoembolization (TACE) has been recommended as the treatment of choice for patients with Barcelona stage B hepatocellular carcinoma; however, community practice often varies from these AASLD guidelines. In this study, we sought to examine whether strict adherence to guidelines was important and to assess which factors determined outcome after TACE. Methods/Results: From the period of 2006 to 2012, 308 patients with newly diagnosed hepatocellular carcinoma were treated at the VA Ann Arbor Healthcare System. Of these, 109 patients underwent TACE procedure. The decision to undergo TACE was made with the recommendations of either the multidisciplinary liver tumor board (> 90%) and/or gastroenterology consultation (> 98%). Of these 109 patients, the median age was 60 years (48-90), 97% were males and 82% had chronic hepatitis C infection. The median size of the largest lesion was 4 cm, 51% were multifocal, and portal vein thrombosis was present in 7%. The median MELD score was 9 (6-1 8), and 61%, 34% and 5% were Child A, B and C cirrhotics, respectively. The Barcelona stage was A in 21%, B in 22%, C in 50% and D in 7% of the patients. The median ECOG score was 1. The median time from diagnosis to TACE was 81 days. Sixty two patients died after a median 333 days from the index TACE treatment. The un-adjusted 1-, 2-, and 3-year patient survival was 64%, 35%, and 24%, respectively. Hepatitis C (HR=2. 66; p=0. 016), CTp score (B vs. A: HR= 2. 51, p=0. 002; C vs. A: HR=7. 96, p<0. 0001) and the TNM stage (4 vs. 1: HR=11. 82, p=0. 005) were the independent predictors of mortality. Interestingly, Barcelona stage (p=0. 88) and performance status as measured by ECOG (0. 98) were not associated with mortality after TACE. Conclusions: In this community based, single VA center study, we found that a significant number of patients who were not Barcelona Stage B were treated with TACE. In patients undergoing TACE, we found that advanced TNM stage, poor liver synthetic function and chronic hepatitis C infection were better predictors of mortality then guideline directed decisions based on Barcelona stage. These factors may be useful as future guides to patient selection for TACE.

Disclosures:

Heather McCurdy - Speaking and Teaching: Onyx Pharmaceuticals, Bayer Health Care

The following people have nothing to disclose: Pranab Barman, Pratima Sharma, Venkat Krishnamurthy, Jonathon Willatt, Richard H. Moseley, Grace L. Su

2199

  1. Top of page

Hurricane Katrina's Impact on HCC in New Orleans

Reinaldo Quevedo1, Erica Kane2, Kristine Oines4, Joseph F. Buell3, Nathan J. Shores3
1Department of Medicine, Tulane University School of Medicine, New Orleans, LA; 2Department of Surgery Tulane University School of Medicine, New Orleans, LA; 3Department of Transplant Hepatology, Tulane University School of Medicine, New Orleans, LA; 4Tulane University School of Public Health, New Orleans, LA

This study aims to determine if the disruption of the New Orleans (NO) healthcare system by the devastation of Hurricane Katrina (2005) resulted in regional or racial disparity of HCC diagnosis and outcomes. All patients with HCC (n=8, 315) from NO [2000-2004 (Pre-Katrina) and 2006-2009 (Post-Katrina)] were identified from the Surveillance Epidemiology and End Results (SEER) national cancer database. Data from Atlanta (Atl), Detroit (Det.), and San Francisco (SF) were included for comparison. Data were analyzed with descriptive analysis, as well as Kaplan-Meier analysis to estimate 1-and 3yr HCC survival. Results: The cumulative observed cases (NO, Atl, Det., SF) of HCC were 3998 before Katrina and 4317 after with a median age of 60. 2 and 58. 7 years old, respectively. Of all patients included, 30. 3%(n=1212) were female and 22. 6%(n=904) were African American (AA) before Katrina versus 28. 7%(n=1240) and 26. 1%(n=1126) after (p=NS). Patients across all cities were diagnosed at an earlier stage post-Katrina than beforehand (p<0. 0001). Pre- and Post- Katrina HCC incidence was similar in NO (n=419 vs 352). However, Post-Katrina NO experienced a 7% increase in localized disease and an approximate 8. 5% and 4. 5% reduced incidence of regional and distant disease, respectively. AA's (n=2030) in all cities presented at an earlier age, 58. 4 vs 62. 9 (p<. 0001), and with more advanced disease than whites (n=4600) regardless of year (p=0. 014). One- and three-year survival improved significantly from 2000-2004 (1 yr: 32. 3%; 3yr: 17. 6%) to 2006-2009 (1yr: 51. 1%; 3 yr: 43. 8%) across all cites and races (p<. 0001). NO saw 15% and 24% increases in 1- and 3- yr survival respectively. Whites had a 1. 34-fold greater chance of living to one year compared to AA's (p< 0. 0001) post-Katrina. This trend was not influenced by city. Conclusion: 1-and 3-yr HCC survival and tumor stage at presentation improved in post-Katrina NO despite the disruption to the local healthcare system and was similar to other cities evaluated (Atl, Det., SF). AA's develop HCC at a younger age and suffer disparate survival compared to white subjects. However, this racial disparity was not limited to New Orleans and not exacerbated by Katrina.

1-and 3- Year Survival

Pre-Katrina All CitiesPost-Katrina All CitiesPre-KatrinaNewOrleansPost-KatrinaNewOrleansPre-KatrinaNewOrleans(White)Post-KatrinaNewOrleans(White)Pre-KatrinaNewOrleansPost-KatrinaNewOrleans
1-Year Survival32. 3%51. 1%31. 5%46. 4%34%50. 6%24. 6%39. 6%
3-Year Survival17. 6%43. 8%15%38. 8%16. 5%27. 4%11. 5%25%

Disclosures:

Nathan J. Shores - Advisory Committees or Review Panels: Gilead; Speaking and Teaching: Vertex, Merck, Salix

The following people have nothing to disclose: Reinaldo Quevedo, Erica Kane, Kristine Oines, Joseph F. Buell

2200

  1. Top of page

Expression of Ribosomal Protein S23 in Hepatocellular Carcinoma Tissue as Good Prognostic Marker

Chan Ran You1, 2, Myeong Jun Song1, 2, Sung Woo Hong1, 2, Wonhee Hur1, Jeong Won Jang1,2, Si Hyun Bae1, Jong Young Choi1,2, Sang Wook Choi1,2, Seung Kew Yoon1,2

1The Catholic Liver Research Center, Collegy of Medicine, The Catholic University of Korea, Seoul, Republic of Korea; 2Department of Internal medicine, Collegy of Medicine, The Catholic University of Korea, Seoul, Republic of Korea

Background: Ribosomal proteins (RP) involve in the regulation of apoptosis and carcinogenesis. In recent studies of RP in hepatocellular carcinoma (HCC), the overexpression of ribosomal protein L12, L27, L30 and l36 was found. We investigated the expression of ribosomal protein S23 (RPS23) and its effect on prognosis in postoperative HCC patients. Methods: We assessed the RPS23 expression in human hepatoma cell lines (HepG2, Hep3B, Huh7) using Western blots. Liver tissues were obtained from patients who had undergone curative resection for HCC. We included the 63 patients with early tumor stage (< UICC stage III) and good liver function (< Child Pugh score 6). The expressions of RPS23 in tumor tissue and non-tumor tissue were examined by immunohistochemical staining. According to tumor RPS23 expression, the patients were classified into two groups, tumor RPS23 (t-RPS23) expression group and nonexpression group. Results: RPS23 expression was higher in the human hepatoma cell lines than normal liver cells. The strongest expression of RPS23 was observed in HepG2 cells compared to Hep3B and Huh7 cells. RPS23 was expressed in 34 patients (54. 0%) on tumor tissues and 8 patients (13. 1%) in Non-tumor tissues (P=0. 016). Basal characteristics were not different between t-RPS23 expression group and non-expression group. The three years and five years survival rate was higher in t-RPS23 expression group (90. 2% vs. 65. 8% of 3 year survival rate, p=0. 007, 86. 1% vs. 55. 7% of 5 year survival rate, p=0. 01 0). On multivariate analysis, t-RPS23 expression (HR 10.95 P=0. 004) and tumor number (HR 9. 71, P=0. 001) were independent prognostic factors for 3 years and 5 years survival. Tumor recurrence during one year after resection was relatively low in t-RPS23 expression group, but not statistically significant (17. 8% vs. 36. 0%, p=0. 084). However, on multivariate analysis, t-RPS23 expression (HR 4. 18, P=0. 021) and serosa invasion (HR 4. 15, P=0. 019) were independent factors for 1 year tumor recurrence. Conclusions: The RPS23 expression of HCC patients was higher in tumor tissue than in nontumor tissues. The RPS23 expression in tumor tissue was good prognostic marker for resected HCC. RPS23 may be associated with good oncogene to have paradoxical function in HCC.

Disclosures:

The following people have nothing to disclose: Chan Ran You, Myeong Jun Song, Sung Woo Hong, Wonhee Hur, Jeong Won Jang, Si Hyun Bae, Jong Young Choi, Sang Wook Choi, Seung Kew Yoon

2201

  1. Top of page

Noninvasive prediction of hepatocellular carcinoma development using the FIB-4 index in chronic hepatitis C patients

Nobuharu Tamaki, Masayuki Kurosaki, Yu taka Yasui, Shoko Suzuki, Takanori Hosokawa, Kaoru Tsuchiya, Hiroyuki Nakanishi, Jun Itakura, Namiki Izumi
Department of Gastroenterology and hepatology, Musashin Red Cross Hospital, Tokyo, Japan

Background: The FIB-4 index is a simple formula to predict liver fibrosis based on standard biochemical values, calculated by the following formula: (age [yr] x aspartate aminotransferase [U/L]) / ((platelet [109/L] x (alanine aminotransferase [u/L])1/2). This study aimed to evaluate the utility of the FIB-4 index and associated time-course changes as a predictor of hepatocellular carcinoma (HCC) development. Methods: A total of 1046 chronic hepatitis C patients that received a liver biopsy in Musashino Red Cross Hospital were investigated. All patients had received interferon therapy after the biopsy and not achieved sustain virological response. We examined a relation between the time-course changes of the FIB-4 index and HCC development. Change in the FIB-4 index (AFIB-4 index/year) was calculated by the following formula: AFIB-4 index/year = (the FIB-4 index 1 year after interferon therapy 一 the FIB-4 index at the biopsy)/interval between 1year after interferon therapy and the biopsy. Results: Mean follow up periods were 6. 0 years and HCC developed in 119 patients during follow up periods. Univariate analysis demonstrated that the FIB-4 index >3. 25 and AFIB-4/year >0. 3 were predictive factors for HCC development. Multivariate analysis revealed that these two factors were independent. Hazard ratio (HR) for the FIB-4 index >3. 25 was 1. 9 (p=0. 03) and AFIB-4/year >0. 3 was 2. 0 (p=0. 001). Patients with a FIB-4 index >3. 25 and a AFIB-4/year >0. 3 were defined as high risk, those with a FIB4 index <3. 25 and a AFIB-4/year <0. 3 were defined as low risk, and other patients were defined as intermediate risk. The 3-, 5, and 7-year cumulative incidence of HCC in patients with high risk group was 7. 6%, 21. 0%, and 30. 0%, respectively, whereas they were 3. 1%, 9. 7%, and 18. 9% in patients with intermediate risk group, and 0. 6%, 2. 9, and 4. 8% in patients with low risk group (p<0. 001) The HR of HCC development in patients at high risk was 7. 3 (95%CI: 4. 3-12. 5, p<0. 001). Conclusion: It was possible to define a group with high risk of HCC development by calculating the FIB-4 index and considering time-course changes in the FIB-4 index. Because measurement of the FIB-4 index is simple and easy to repeat, it is useful for non-invasive, real-time monitoring of HCC development.

Disclosures:

Namiki Izumi - Speaking and Teaching: MSD Co., Chugai Co., Daiichi Sankyo Co., Bayer Co., Bristol Meyers Co.

The following people have nothing to disclose: Nobuharu Tamaki, Masayuki Kurosaki, Yutaka Yasui, Shoko Suzuki, Takanori Hosokawa, Kaoru Tsuchiya, Hiroyuki Nakanishi, Jun Itakura

2202

  1. Top of page

Identification of autoantibody based serological marker for early diagnosis of hepatocellular carcinoma by serological proteome analysis combined with protein microarray

Yu Hong1, Bei Zhang1, Su-hong Chen2, Jiang Long3, Ji-dong Jia1, Jian Huang1
1Liver Research Center, Beijing Friendship Hospital. Capital Medical University, Beijing, China; 2Biotechnology Center, Beijing Institute of Radiation Medicine, Beijing, China; 3Department of Oncology Minimally Invasive Interventional Radiology, Beijing Youan Hospital, Beijing, China

Background and aims: Novel biomarkers for the early detection of human hepatocellular carcinoma (HCC) are urgently needed. The identification of tumor-associated antigens (TAAs) and autoantibodies for cancer immunodiagnosis has been of interest recently as an approach to cancer detection. The aim of the present study is to identify TAAs and autoantibodies as biomarkers for early detection of HCC by serological proteome analysis combined with protein microarray. Methods: Total protein was extracted from tissue samples of HCC and cell cultures of HepG2 and HepG2. 2. 15 cell lines. For screening of HCCassociated antigen, the extracted protein was separated by two-dimensional gel electrophoresis and then underwent immunological reaction with serum from HCC at early stage and with normal serum as controls. Recombinant proteins of screened potential HCC-associated antigens were prepared and robotically attached in ordered arrays on aldehyde-activated glass slides to make antigen microarray. A total of 138 serum samples were used for detection of autoantibody by the antigen microarray, including 78 cases of HCC at early stage, 30 cases of cirrhosis and 30 cases of healthy control. Results: 30 potential HCC-associated antigen proteins were identified by serological proteome analysis, in which 11 proteins(HSP60, CK1, retinal dehydrogenase 1, alpha 1 antitrypsin, disulfideisomerase, centromere protein F, tubulin beta-4B chain, alphaenolase, ATP synthase subunit beta, apoptosis-inducing factor, heterogeneous nuclear ribonucleoprotein)were used to make antigen microarray for further clinical evaluation. The results of microarray detection with clinical samples showed four proteins with obviously different prevalence of autoantibody against TAAs among HCC, cirrhosis and healthy control: alpha 1 antitrypsin(13%, 6%, 0%), disulfide-iso-merase(83%, 66%, 67%), centromere protein F(54%, 42%, 7%), and alpha-enolase (18%, 11%, 3%) respectively. Further analysis with more candidate antigens and largescale clinical samples are on going. Conclusion: Serological proteome analysis combined with protein microarray is a powerful method for screening of autoantibody based serological marker for early diagnosis of cancer. Our preliminary results implied potential value of some autoantibody based markers on early detection of HCC.

Disclosures:

The following people have nothing to disclose: Yu Hong, Bei Zhang, Su-hong Chen, Jiang Long, Ji-dong Jia, Jian Huang

2203

  1. Top of page

Outcome of patients with end stage liver disease and hepatocellular carcinoma in a rural tertiary care center in New England

Gladys P. Ayala1, Rolland C. Dickson1, Adrienne R. Deters1, Lisa M. Glass1, Arifa Toor1, David Axelrod2
1Gastroenterology and Hepatology, Dartmouth Hitchcock Medical Center, Lebanon, NH; 2Transplant Surgery, Dartmouth Hitchcock Medical Center, Lebanon, NH

Background: Rural residents with end stage liver disease (ESLD) and HCC may be disadvantaged for receiving Liver Transplantation (LT). AIM: To assess outcome of pts with ESLD and HCC referred to a rural tertiary care center without onsite LT. Methods: Between 2/2007 and 1/2012, all pts with ESLD and HCC referred to our Liver tumor Clinic (LTC) were considered for this retrospective review. LT candidacy was assessed. Suitable candidates had LT evaluation and were referred to an LT center. Data was extracted from our LTC Database and EMR with follow up to 6/3/201 3. Results: At time of referral 99/129 (77%) were not LT candidates: HCC out of Milan criteria (59%), medical (14%), sobriety <6 months (11%), older age (9%), lack of support (3%), other (2%). There was no difference in mean age (62 vs 61), male sex (83 vs 80%) and MELD (11 vs 10) between non-LT and LT candidates. LT candidates had earlier BCLCS, were more likely than the non-LT group to have NASH and to have received surveillance for HCC: 83% vs 44%, (p=0. 0002), (Tab1). There was no difference in insurance type between groups. Non-LT candidates, had higher 1yr mortality independent of BCLS. Of potential LT candidates 60% (18/30) were listed for LT, 40% (12/30) were not: psychosocial issues (8), denial by LT center (3), rapid tumor progression (1). Thus, 15% (1 9/129) of our referral cohort were listed, 10% (13/129) underwent LT. One pt in the non-LT group eventually received LT. Summary/Conclusions: HCC out of Milan criteria and psychosocial issues are critical barriers to LT for rural pts. Increased surveillance, early referral and psychosocial support is needed.

Management of Cirrhosis in LT and Non-LT candidates

TreatmentNon-LT Candidates (N=99)*Potential LT CandidatesP-value
  1. a

    No treatment due to: 4 lost to f/u, 6 decompensated, 1 infection, and 1 non-GI illness.

Transplantation1%60%0. 0001
Radiofrequency Ablation28%77%0. 0001
Chemoembolization12%20%0. 37
Y-9012%00. 06
Resection11%3%0. 29
Bland Embolization 8% 00. 20
Sorafenib 8% 00. 20
Hospice 7% 00. 20
Radiation 1% 01. 00
No Treatment 12% 00. 06
Tumor Within Milan Criteria40(40%)30(100%)0. 0001
Transplanted1(1%)12(40%)0. 0001
1 yr mortality42(42%)3(10%)0. 0009

Disclosures:

Rolland C. Dickson - Advisory Committees or Review Panels: Biotest David Axelrod - Management Position: XynManagement

The following people have nothing to disclose: Gladys P. Ayala, Adrienne R. Deters, Lisa M. Glass, Arifa Toor

2204

  1. Top of page

Morbidity and mortality following radiofrequency- and microwave ablation in patients with hepatocellular carcinoma

Shawna S. Boyle1, Bashar Aqel2, Elizabeth J. Carey2, David D. Douglas2, Jorge Rakela2, Hugo E. Vargas2, Thomas J. Byrne2

1Mayo Clinic in Arizona, Phoenix, AZ; 2Division of Hepatology, Mayo Clinic in Arizona, Phoenix, AZ

Background: Radiofrequency ablation (RFA) and microwave ablation (MWA) are commonly used treatments for hepatocellular carcinoma (HCC) in patients awaiting liver transplantation (LT) and in non-transplant candidates. A paucity of data exists regarding complications of ablation 一 particularly for MWA. Aim: The purpose of this study was to evaluate the morbidity and mortality of RFA and MWA in patients receiving those treatments for HCC. Methods: We reviewed the medical records of 55 patients with HCC who were treated with either RFA or MWA between 2008 and 2012. 3 patients who received concomitant intra-arterial therapy within 30 days of ablation were excluded from analysis. Post-procedure outcomes were analyzed for complications. Minor complications included: a) prolonged (> 1 day) hospital stay after treatment, b) unplanned emergency room or outpatient clinic visit for procedure-related symptoms, and c) prolonged post-procedure pain requiring medication beyond 5 days. Major complications included: d) re-admission within 30 days unless deemed unrelated to procedure, e) new portal vein thrombosis, f) irreversible hepatotoxicity (defined as liver decompensation not returning to baseline within 30 days), and g) death within 30 days. Results: A total of 56 ablations (on 52 patients) were performed: 25 RFA, 31 MWA. 26 patients (50%) were listed for LT at the time of treatment. 20 patients (38. 5%) underwent LT within 1 year of treatment, none of whom were transplanted urgently because of treatment-related decompensation. Specific complications are shown in table 1. Reasons for readmission within 30 days of treatment included: exacerbation of hepatic encephalopathy (2), hypoxia with respiratory difficulty (2), pneumonia (2), and abdominal pain (3). 30-day mortality rate was 5. 4% (3 patients: 1 RFA, 2 MWA). Conclusions: RFA and MWA have similar rates of post-treatment morbidity and mortality, with readmission within 30 days being the most common recorded complication for both modalities. Procedure related mortality is low overall but should be understood as a potential risk of therapy.

Description of Complications for RFA and MWA

RFA (N=25)MWA(N=31)P-value
MINOR COMPLICATIONS
Prolonged hospital stay2 (8%)1(3. 2%)N. S.
Unplanned ER / clinic visit3 (12%)2 (6. 5%)N. S.
Prolonged pain3(12%)2 (6. 5%)N. S.
MAJOR COMPLICATIONS
Re-admit 30 days6 (24%)3 (9. 7%)N. S.
Portal vein thrombosis01(3. 2%)N. S.
Irreversible hepatotoxicity1(4%)2 (6. 5%)N. S.
30-day Mortality1(4%)2 (6. 5%)N. S.

Disclosures:

Hugo E. Vargas - Advisory Committees or Review Panels: Eisai; Grant/Research Support: Merck, Gilead, Idenix, Novartis, Vertex, Janssen, Bristol Myers, Ikaria, Abbott

The following people have nothing to disclose: Shawna S. Boyle, Bashar Aqel, Elizabeth J. Carey, David D. Douglas, Jorge Rakela, Thomas J. Byrne

2205

  1. Top of page

A Serum-based Multiplex Assay (HCCplex) Improves Prediction of Early Hepatocellular Carcinoma Compared to AFP

Monica Schmidt2, 1, Michael W. Fried2
1UNC Liver Center & Giilings School of Global Public Health, University of North Carolina Chapel Hill, Graham, NC; 2UNC Liver Center, University of North Carolina, Chapel Hill, NC

Background: The incidence of hepatocellular carcinoma (HCC) is rising worldwide. A combination of biomarkers may identify HCC more accurately than a single biomarker. A multivariate index assay combining wheat germ agglutinin-reactive E-Cadherin, vascular endothelial growth factor and lymphatic vessel endothelial hyaluronan receptor 1 may be clinically useful in enhancing the sensitivity and specificity of alpha fetoprotein (AFP) for detecting HCC. Methods: We selected a training set of cirrhotic cases with HCC (n=20) or cirrhotic controls without HCC (n=37). Cirrhosis was defined by biopsy and/or clinical/imaging criteria. HCC was diagnosed in accordance with AASLD guidelines including characteristic imaging and/or histologic confirmation. The three-plex assay (HCCplex) uses an antibody-lectin sandwich array (ALSA) format that detects specific glycan post-translational modifications on the 3 proteins. The results of the three markers were combined into a logit model with total AFP using a cut-point of 0. 5 to define positives. The algorithm from the training set was applied to a replication set (24 HCC cases and 24 cirrhotic controls), operator blinded to status, using serum samples within 60 days from diagnosis. To test robustnesss of the model, the two sample sets were also combined and the algorithm applied. One patient was inadvertently misclassified and corrected based on clinical information after unmasking. Results: HCCplex predicted HCC more accurately than AFP in all studies (Table 1). Sensitivity of HCCplex was better than AFP alone in all analyses. We evaluated HCC with low AFP<20 & <40 ng/mL. Conclusions: HCCplex may improve identification of HCC compared to AFP alone. The specificity of HCCplex in AFP negative (<40ng/mL) HCC may be useful in determining if lesions are HCC versus dysplastic nodules. A larger sample size and appropriately designed validation study is warranted based on these preliminary results.

Performance of HCCplex

TestCohortnSensitivitySpecificityAUROC
HCCplexTraining57951000. 9865
AFP aloneTraining57901000. 950
HCCplex AFP<20ng/mLTraining57501000. 9054
HCCplex AFP<40ng/mLTraining40661000. 9369
HCCplexReplication4876700. 8229
AFP aloneReplication4856780. 7535
HCCplex AFP<20 ng/mLReplication3150840. 8596
HCCplex AFP<40 ng/mLReplication3557760. 6853
HCCplexCombined110881000. 901
AFP aloneCombined11067950. 8802
HCCplex AFP<20Combined7343950. 8267
HCCplex AFP<40Combined7841970. 8172

Disclosures:

Monica Schmidt - Grant/Research Support: Merck & Co.; Patent Held/Filed: HCCplex; Stock Shareholder: PleX Diagnostics

Michael W. Fried - Consulting: Genentech, Merck, Abbvie, Vertex, Janssen, Bristol Myers Squibb, Gilead; Grant/Research Support: Genentech, Merck, AbbVie, Vertex, Janssen, Bristol Myers Squibb, Gilead; Patent Held/Filed: HCCPlex

2206

  1. Top of page

Comparative Analysis of Treatment Response for Unresectable Hepatocellular Carcinoma: Radioembolization Versus 3D-CRT Combined with TACE

Jin Mo Yang, Jin Mo Yang, Jeong Won Jang, Si Hyun Bae, Jong Young Choi, Seung Kew Yoon

The Catholic University Liver Research Center, Department of Internal Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea

Background/Aims Radioembolization is a relatively new modality used for the treatment of intermediate to advanced Hepatocellular carcinoma(HCC) especially in cases with large sized tumors or portal vein tumor thrombosis(PVTT). We analyzed and compared the treatment response after radioembolization therapy with transarterial chemoembolization(TACE) combined with 3-dimensional conformal radiotherapy(3D-CRT). Methods A total of 38 intermediate to advanced stage HCC patients who underwent radioembolization or TACE with 3DCRT (40-45 Gy for 4-5 weeks) that was started one week after the 1st TACE at Seoul St. Mary's hospital between February, 2008 and December, 2012 were analyzed (Radioembolization 19, TACE+3D-CRT 19). Treatment response was evaluated according to the modified Response Evaluation Criteria in Solid Tumors (mRECIST). We assessed the treatment response at 3 months and 6 months after the initial treatment, and the overall survival. Results The median follow up period was 13. 8 months. In the radioembolization group, the mean age was 65 years (42-78), median tumor size was 6. 5cm (1. 5-12. 7), and the pVTT was presenting in 8 of 19 patients (42. 1%). Objective response was obtained in 12 of 19 patients (63. 2%) at 3 months and in 10 of 19 patients (52. 6%) at 6 months after the treatment and the median overall survival was 13. 9 months (4. 8-44. 6). In the TACE+3D-CRT group, the mean age was 56. 6 years (42-73), median tumor size was 10cm (5. 2-15. 6), and the PVTT was presenting in 14 of 19 patients (73. 7%). Objective response was obtained in 9 of 19 patients (47. 4%) at 3 months and in 7 of 19 patients (36. 9%) at 6 months after the treatment and the median overall survival was 11. 7 months (3. 7-60. 5). Multivariate analysis showed that there was no statistically significant difference in the treatment response and overall survival between the radioembolization and the TACE+3D-CRT groups. Conclusions In patients with old age and advanced stage HCC (large sized tumor and PVTT) radioembolization therapy showed relatively favorable treatment response. Further investigation in a larger number of patients would be necessary to evaluate efficacy of radioembolization therapy and for better patient selection.

Disclosures:

The following people have nothing to disclose: Jin Mo Yang, Jin Mo Yang, Jeong Won Jang, Si Hyun Bae, Jong Young Choi, Seung Kew Yoon

2207

  1. Top of page

Experience with Sorafenib treatment of Hepatocellular carcinoma in (Hcc) in real life patients in Three Brazilian centeres

Rogerio C. Alves1, Daniele E. Alves1, 2, Mônica V. Viana1, Eduardo A, Fonseca1, Carla A. Matos1, Mario Kondo1, Artur Malzyner2, Otavio Gampel1, Paula B. Poletti1
1Hepatology, Iamspe/A. C. Camargo, Sao Paulo, Brazil; 2Hospital Heliopolis, Sao Paulo, Brazil

Introduction: Hepatocellular carcinoma is a challenging tumor to treat, because tumor arises from cirrhosis and the patient has 1 diseases. HCC is the most important cause of death in patients with liver cirrhosis. Sorafenib is established as treatment for patients with compensated liver disease (Child Pugh A), but there are few studies for patients with more advanced cirrhosis. Methods: From May 2008, 152 patients initiated sorafenibe at 800 mg daily doses, 138 were treated or had been treated for more than 1 month and were recruited to data analysis on April 2012, regarding sex, age, association with diabetes, etiology of cirrhosis, treatment intervals, dose reductions, adverse events, response rates, Child-Pugh's classification, global survival and survival according Child-Pugh and A. C. L. C status. Results: The median age of study population of 152 patients (104 men and 48 women) was 63, 78 years. The cirrhosis etiology was 46% hepatitis C, 13% N. As. H, 10, 71% cryptogenic cirrhosis and 6. 15% with hepatitis B. 45. 7% of the patients were Child-Pugh A, 42. 4 Child-Pugh B and 8. 7% Child-Pugh C by Child-Pugh classification. 34% of the patients were diabetic, the median treatment interval of sorafenib was days 335 days(7-1080 days); Overall response rate was 66%: 37.5% had stable disease and 25% had partial response with tumor shrinkage by Recist criteria. Adverse events were related in 58. 47 % of the patients and 42. 7 % of them needed to have dose modifications during the treatment, The common adverse events were diarrhea: 48. 5%, nausea and vomiting: 31. 4 %, abdominal pain: 45. 7%, hand-foot syndrome 25. 7%, arterial hypertension 8. 5% and 8. 5 % bleeding. The incidence of adverse events didn't differ considering each Child-Pugh's subgroup. The median survival time was 13. 5 months. ' The one and two year survival preliminary rate were 75 % and 68% in Child-Pugh A patients and 40 and 10 % in Child-Pugh B patients (p=0. 002) respectively and for B. C. L. C. B was 13. 5 and C 14 months respectively Conclusion: Our results contributed to emphasize treatment safety of sorafenibe within Hcc patients with more advanced liver disease. In this study population, 50% of the patients had liver disease Child B/C and sorafenib was well tolerated (58 % had adverse events) in all subgroups with acceptable response and treatment intervals similar than Sharp study, despite of advanced liver disease, but the survival rates trends to be statistically higher in patients with Child A than Child B and in B. C. L. C B than C.

Disclosures:

The following people have nothing to disclose: Rogerio C. Alves, Daniele E. Alves, Monica V. Viana, Eduardo A. Fonseca, Carla A. Matos, Mario Kondo, Artur Malzyner, Otavio Gampel, Paula B. Poletti

2208

  1. Top of page

Morphological characteristics of Non-Alcoholic Steatohepatitis-induced Hepatocellular Carcinoma

Manuela G. Neuman1, 2, Marius Braun3, Laurie M. Blendis4, Stephen Malnick5, Lawrence Cohen6
1In Vitro Drug Safety and Biotechnology, Toronto, ON, Canada; 2Pharmacology and Toxicology, University of Toronto, Toronto, ON, Canada; 3Liver Institute, Rabin Medical Centre, Petah Tikva, Israel; 4Gastroenterology Souraski Medical Centre, Tel Aviv, Israel; 5Gastroenterology and Medicine, Kaplan Hospital, Rehovot, Israel; 6Gastroenterology, Sunnybrook HSC, Toronto, ON, Canada

INTRODUCTION: Non-alcoholic steatohepatitis (NASH) has been associated with fibrosis that may progress to cirrhosis and hepatocellular carcinoma. The purpose of this study was to examine hepatocytes and Kupffer Cells (persinusoidal macrophages) in liver biopsies of 3 families (5 males and 4 females) with non-cirrhotic and cirrhotic NASH to determine unique histological changes during a period of 2-7 years from diagnosis. AIMS and METHODS: In this study, hepatocytes, stellate cells and Kupffer Cells were analyzed using light and electron microscopy, and immunohistochemistry with specific anti-macrophage antibody staining of liver biopsies. Evaluation for HBV, HCV, autoimmune hepatitis, PBC, PSC, alpha-1 anti-trypsin disease, Wilson's disease, and hemochromatosis were negative. Abdo ultrasound demonstrated diffuse, extensive steatosis in all the patients and nodules consistent with hepatocellular carcinoma (HCC) in two patients. RESULTS: BMI for all patients was over 30, and alcohol consumption was not significant. In all liver biopsies diffuse, macrovesicularsteatosis involved 40 to 70% of samples. The lobular hepatocytes showed prominent ballooning degeneration. Mallory Denk-hyaline bodies (MDBs) in four patients along with foci of lobular inflammation. The MDBs were observed by electron microscopy and stained by cytokeratin 18. The trichrome stain revealed bridging fibrosis. In family #1, there was a three-fold increase in Kupffer Cells in the eldest sibling with NASH compared to livers of the younger siblings. In addition, lipid droplets were more numerous and larger in the hepatocytes and the collagen deposits were 5 times larger compared to the siblings. In family #2, the older brother had cirrhosis, minimal inflammation and mild fatty infiltration while the younger had fatty infiltration with no inflammation or fibrosis. In family #3, large lipid droplets and collagen deposition were observed in the siblings, one of whom had HCC. Over 2-4 year follow up, all patients had progressive severity of NASH one of whom developed HCC. CONCLUSION: The special finding in livers of patients with NASH was accumulation of groups of Kupffer Cells, which was not associated with focal necrosis. It is postulated that collections of these perisinusoidal macrophages in NASH are a unique response to chronic portal endotoxemia or bacteremia. The persistent activation of these macrophages could lead to the chronic release of cytokines and contribute to chronic inflammation, fibrosis and cirrhosis leading to HCC.

Disclosures:

The following people have nothing to disclose: Manuela G. Neuman, Marius Braun, Laurie M. Blendis, Stephen Malnick, Lawrence Cohen

2209

  1. Top of page

Performance Status has Minimal Impact on Patient Survival in BCLC Early Stage

Xinyi Zhou, Kathleen C. Shaw, Luisa Agie, Jonathan Fawcett, Katherine A. Stuart
Gastroenterology and Hepatology, Princess Alexandra Hospital, Brisbane, QLD, Australia

Background: The Barcelona Clinic Liver Cancer (BCLC) staging system is the most widely used treatment algorithm for Hepatocellular Carcinoma (HCC). Patient performance status (PS) is a key component of the revised BCLC criteria. Patients with an ECOG status of >1 are categorised as BCLC Stage C irrespective of tumour size and number. Other HCC staging systems place more emphasis on the characteristics of the tumour. Aims and methods: To assess the influence of PS on treatment selection and overall survival in patients with BCLC Early Stage HCC. A retrospective review was conducted of all patients with BCLC Early HCC who were treated with curative intent using radiofrequency ablation (RFA), surgical resection (SR) or liver transplantation (LT) at a single Australian liver transplant centre between January 2005 and June 2012. Early HCC was defined as a single tumour <6. 5 cm or <3 tumours, none greater than 4. 5cm in maximum diameter with a total tumour diameter of <8cm. Patients were divided into 2 groups based on their PS (ECOG 0-1 good PS group, ECOG >2 poor PS group). Results: 78 patients with BCLC Early HCC were identified. There was no difference in the number of tumours or total dimensions between the two groups. On multivariate analysis the only significant difference between the two groups was INR (p=0. 005). There was no difference in outcomes in terms of tumour recurrence (30% vs 13. 3% p=0. 45) and overall survival (67. 2% vs 64. 7%, p=0. 85). Kaplan Meier survival curve and log rank test showed no significant difference (p=0. 83). Conclusion: These results suggest that patients with BCLC Early stage HCC and poor PS are more likely to have cirrhosis complicated by portal hypertension and more severe liver disease. A greater proportion has LT as their initial curative treatment. Overall survival is comparable between patients with poor and good PS. This data supports the current practice of treating all BCLC Early HCC patients with curative intent irrespective of their PS.

Demographic and Clinical Characteristics

Good PS (n=61)Poor PS (n=17)p value
Age (mean±SD, yr)57. 9±6. 757. 9±90. 96
Male sex (%) 90%88%0. 99
Ethnicity (n, Caucasian/Asian/Other)46/10/516/1/00. 29
Treatment (n, RFA/SR/OLT)19/15/275/2/100. 52
Cirrhosis (%)92%100%0. 58
Portal hypertension (%) 77% 94%0. 17
MELD (mean士SD)11. 0±4. 514. 2±8. 60. 13
CP score (mean士SD)6. 2±1. 37. 8±2. 30. 001
ASA score (1&2 vs 3&4, n)41/205/120. 005
INR (mean± SD)1. 27±0. 181. 48±0. 340. 005
Non-liver co-morbidities (%) 44. 3%41. 2%0. 41

Disclosures:

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

The following people have nothing to disclose: Xinyi Zhou, Kathleen C. Shaw, Luisa Algie, Jonathan Fawcett