Imaging of the Liver


Prognosis of the hepatic nodules that show hypovascular in arterial phase and hypointense in hepatobiliary phase on gadoxetic acid enhanced MRI

Yoshihiko Ooko, Fumihiko Kanai, Sadahisa Ogasawara, Tenyu Motoyama, Eiichiro Suzuki, Akinobu Tawada, Tetsuhiro Chiba, Osamu Yokosuka

Department of Gastroenterology ond Nephrology, Graduate School of Medicine, Chiba university, Chiba, Japan

Purpose: to investigate the prognosis of the hepatic nodules that show hypovascular in arterial phase and hypointense in hepatobiliary phase on gadoxetic acid enhanced MRI. Material and Method: From February 2008 to March 2012, 1614 patients were performed total 2656 gadoxetic acid enhanced MRIs in our institution.43 patients with 53 hepatic nodules less than 15mm that show hypovascular in arterial phase and hypointense in hepatobiliary phase and without hepatocellular carcinoma (HCC) were retrospectively identified from medical records. The lack of arterial flow was confirmed at least 2 modalities (gadoxetic acid enhanced MRI, dynamic CT, contrast enhanced ultrasounds and/or CT during hepatic angiography).1-year cumulative incidence rate of appearance of arterial flow in those nodules, 1-year cumulative incidence rate of diagnosis of HCC in the nodules and in the patients was analyzed. Result: 43 patients with 58 nodules were analyzed.22 male and 21 females were included. Median size of the nodules was 9mm (range 4-15mm). Median observation period was 522 days (range 126-1580days).1-year cumulative incidence rate of appearance of arterial flow in those nodules was 8.6%.1-year cumulative incidence rate of diagnosis of HCC in the nodules was 17.2%.1-year cumulative incidence rate of diagnosis of HCC in the patients was 18.6%. Conclusion: 1year cumulative incidence rate of diagnosis of HCC in the nodules less than 15mm that show hypovascular in arterial phase and hypointense in hepatobiliary phase was low. The follow-up may be better for those nodules.


The following people have nothing to disclose: Yoshihiko Ooka, Fumihiko Kanai, Sadahisa Ogasawara, Tenyu Motoyama, Eiichiro Suzuki, Akinobu Tawada, Tetsuhiro Chiba, Osamu Yokosuka


Identification of hepatocellular carcinoma patients who will benefit from 18F-fludeoxyglucose-positron emission tomography imaging for disease staging and determination of optimal treatment strategies

Etsushi Kawamura1, Susumu Shiom2' Kohei Kotani2, Atsushi Hogihara1, Hideki Fujii1, Sawako K. Uchida1, Shuji Iwai1, Hiroyasu Morikawa1, Joji Kawabe2, Mosoru Enomoto1, Yoshiki Murokomi1, Akihiro Tamori1, Norifumi Kawada11

1Department of Hepatology Graduate School of Medicine, Osaka City University, Osaka, Japan; 2Department of Nuclear Medicine, Graduate School of Medicine, Osaka City University, Osaka, Japan

Background and Aim: Accurate tumor staging for hepatocellular carcinoma (HCC) has become increasingly important with treatment advances. In extrahepatic metastasis (EHM) cases, radical treatment was not indicated previously, but sorafenib has been recently recommended. In advanced stages, although 18F-fiudeoxygIucose-positron emission tomography (FDG PET) can detect intra- and extrahepatic lesions with a single noninvasive scan, when this scan should be performed to determine treatment strategies is not clear. To set the indices for PET timing, we analyzed the clinical characteristics of HCC patients with FDG-avid primary lesions (PL) or EHM. Methods: Data from 64 consecutive HCC patients who underwent PET between April 2005 and November 2012 were retrospectively analyzed. The patient cohort had a median age of 74 [49-87] years; 41 men and 23 women; 7 initial occurrences and 57 recurrences; 14 chronic hepatitis and 50 cirrhosis cases; 46 HCV, 4 HBV, 4 alcoholic, 3 NASH, 1 AIH, and 6 unknown cases; and 15 Barcelona Clinic Liver Cancer (BCLC) stage 0, 26 stage A, 10 stage B, 12 stage C, and 1 stage D cases. Patients were not treated for HCC for a month before PET. Univariate and multivariate analyses were used to determine the factors associated with FDG-avid PL or EHM detection. Results: The sensitivity, specificity, and accuracy of PET were 36%, 100%, and 39% for detecting PLs (p = 0.276) and 88%, 75%, and 80% for detecting EHMs (p < 0.0001), respectively. PET detected FDGavid PLs in 34% of cases (22/64: 82% [18/22], hypervascular PLs; 50% [11/22], serum a-fetoprotein levels [AFP] \≤ 200 ng/mL, and 50% [11/22], beyond the Milan criteria). EHMs (7 in lymph nodes, 5 in the lung, 4 in the abdominal wall, 3 in bone, and 4 others [including overlapping]) were detected in 33% of cases (21/64: 38% [8/21], maximum standardized uptake value (SUV) of PL ≥4.0). Recommended treatments were changed (mostly from curative treatment to sorafenib) in 25% of cases (16/64) on the basis of a treatment algorithm according to BCLC staging using PET results. On logistic multivariate analyses, AFP level ≥200 ng/dL (odds ratio [OR] 11.2, p = 0.002) and being beyond the Milan criteria (OR 10.5, p = 0.008) were independent factors for FDG-avid PL detection. SUV of PL ≥4.0 was independent factor for FDG-avid EHM detection (OR 4.3, p = 0.045). Conclusions: In patients with a high AFP level or those beyond the Milan criteria, PET should be considered to evaluate HCC spread. PET detected EHMs at a high rate in patients with a high SUV of PL. Thus, in such patients, PET complements conventional imaging in BCLC staging and determining treatment strategies.


Akihimo Tamori - Grant/Research Support: MSD

The following people have nothing to disclose: Etsushi Kawamura, Susumu Shiomi, Kohei Kotani, Atsushi Hagihara, Hideki Fujii, Sawako K. Uchida, Shuji Iwai, Hiroyasu Morikawa, Joji Kawabe, Masamu Enomoto, Yoshiki Murakami, Norifumi Kawada


Virtual ultrasonography constructed by Gd-EOB-DTPAenhanced magnetic resonance imaging is useful to avoid bile duct injury during radiofrequency ablation

Yohei Koizumi1, Masashi Hirooka1,Hironori Ochi1,Yoshio Tokumoto1,Masanori Abe1, Fujimasa Tada1,Atsushi Hiraoka3, Hiroaki Tanaka2, Takaharu Tsuda2, Teruhito Mochizuki2, Yoichi Hiasa1

1Department of Gastroenterology and Metabology, Ehime University Gaduote School of Medicine, Toon, Japan; 2Department of Diagnostic and Therapeutic Radiology, Ehime University Graduate School of Medicine, Toon, Japan; 3GastroenteroIogy Center, Ehime Prefectural Central Hospital, Matsuyama, Japan

Background/Aims: Virtual ultrasonography from gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOBDTPA)-enhanced magnetic resonance imaging (MRI) was established to evaluate bile duct anatomy on ultrasonography. The aim of this study was to prospectively evaluate the contribution of this virtual technology to the safety and utility of radiofrequency ablation (RFA) Methods: This study was approved by our institutional review board and informed consent was obtained from all patients prior to any study-related procedures. Bile duct anatomy was assessed in 201 patients who underwent Gd-EOB-DTPA-enhanced MRI for the evaluation of hepatic tumors. Eighty-one of these patients subsequently underwent RFA assisted by ultrasound imaging. In 23 patients, the tumor was located within 5 mm of the central bile duct, as demonstrated by MRI. Results: Virtual ultrasonography constructed using Gd-EOB-enhanced MRI was able to visualize the common bile duct, left hepatic duct and right hepatic duct in 96.5%, 94.0%, and 89.6% of cases, respectively. The right anterior sectoral duct (74.1%) and the right posterior sectoral duct (80.1%) were also identified on virtual ultrasonography. The left lateral sectoral duct and left medial sectoral duct were able to be seen in 74.6% and 67.7% of cases, respectively. The target hepatic tumor nodule and biliary duct could be detected with virtual ultrasonography in all patients, resulting in no severe complications after RFA. Conclusions: The course of the bile ducts can be recognized on conventional ultrasound by referencing virtual ultrasonography constructed by Gd-EOB-DTPAenhanced MRI. This imaging technology is useful in avoiding bile duct injury during RFA.


The following people have nothinq to disclose: Yohei Koizumi, Masashi Himooka, Hironori Ochi, Yoshio Tokumoto, Masanomi Abe, Fujimasa Tada, Atsushi Himaoka, Himoaki Tanaka, Takahamu Tsuda, Temuhito Mochizuki, Yoichi Hiasa


Usefulness and challenges associated with the spleen stiffness via Virtual Touch Quantification for the prediction of liver fibrosis

Himonomi Tonoko1,2, Hiroko lijima1,2, Masahiro Yoshida2, Chikage Nakano2, Kenji Hashimoto2, Noriko Ishii1, Tomoko Aoki1, Yukihisa Yuri1, Kazunori Yoh1, Akio Ishii1,Tomoyuki Takashima1,Yoshiyuki Sokai1, Nobuniro Aizawa1, Kazunari iwata1, Naoto Ikeda1,Yoshinori Iwata1, Hirayuki Enomoto1, Masaki Saito1, Shuhei Nishiguchi1,Mai Nakano1

1Division of Hepatobiliary and Pancreatic, Medicine, Department of Internal Medicine, Hyogo College of Medicine, Nishinomiyacity Japan; 2 Ultrasound Imaging Center' Hyogo College of Medicine' Nishinomiya' Jopon

Background and Aim Virtual Touch Quantification (VTQ) can be used to easily measure spleen stiffness (SS) by referring to the corresponding B-mode image without restricting the measurement distance. However, the usefulness and challenges associated with the measurement of SS for the prediction of liver fibrosis stage are not well documented. In the present study, we aimed to evaluate SS by VTQ for the prediction of liver fibrosis. Patients and Methods From December 2010 to February 2013, 352 patients (162 men and 190 women) with chronic liver disease confirmed by liver biopsy were evaluated by VTQ for the measurement of liver stiffness (LS) and SS (average age 55.8 ± 13.5 years; 90 patients with hepatitis B, 179 with hepatitis C, and 1 with hepatitis B and C; 76 patients had non-B non-C hepatitis). The New Inuyama Classification was used to evaluate the degree of hepatitis. The distribution of liver fibrosis stages was as follows: stage F0 (n =15), F1(n =134), F2 (n = 66), F3 (n = 73), and F4 (n = 64). VTQ measurements were performed using the Siemens Acuson S2000 ultrasound system. SS values were compared with clinical parameters including measurements of LS; platelet count; levels of AST, ALT, bilirubin, hyaluronic acid, and albumin; prothrombin time; and APRI. Results The LS and SS values corresponding to each fibrosis stage were 1.16 and 2.40 for stage F0, 1.14 and 2.33 for stage F1, 1.34 and 2.44 for stage F2, 1.53 and 2.54 for stage F3, and 2.30 and 3.18 for stage F4, respectively. Significant differences between stages F3 and F4 were observed for both LS and SS values (P < 0.0001). SS values showed the highest correlation with LS values (r = 0.595, P < 0.0001). The area under the receiver operating characteristic curve for SS to distinguish between fibrosis was the highest among all the parameters (SS = 0.918; LS = 0.905; hyaluronic acid = 0.830; APRI = 0.772; platelet count = 0.738; prothrombin time = 0.738). However, for SS measurements, 20% (n = 3) of F0 and 16% (n = 22) of F1 patients fell above the F4 cutoff levels; these rates were higher than those for LS (0% of F0; 3% of F1). All cases with high SS values and F0 and F1 stages had a small spleen except for 1 severely obese F1 patient. Conclusion SS measurements obtained using VTQ could be a good predictor of liver fibrosis stage, although the occurrence of false positive results should be carefully considered in cases with small spleens.


The following people have nothing to disclose: Hironori Tanaka, Hiroko lijima, Masahiro Yoshida, Chikage Nakano, Kenji Hashimoto, Noriko Ishii, Tomoko Aoki, Yukihisa Yuri, Kazunori Yoh, Akio Ishii, Tomoyuki īakashima, Yoshiyuki Sakai, Nobuhiro Aizawa, Kazunari Iwata, Naoto Ikeda, Yoshinori Iwata, Hirayuki Enomoto, Masaki Saito, Shuhei Nishiguchi, Mai Nakano


Comparison of the diagnostic performance of Real time tissue elastography as strain elastography with that of FibroScan as shear wave elastography

Hiroyasu Morikawa, Sawako K. Uchida, Hideki Fujii, Atsushi Hagihara, Etsushi Kawamura, Shuji Iwai, Masaru Enomoto, Yoshiki Murokomi, Akihiro Tamori' Norifumi Kowodo; Deporfmenf of Hepatology, Osaka City University, Osaka, Japan

Background and Aim: Elastography, developed as a non-invasive tool to measure tissue elasticity. Recently it has become possible to measure liver fibrosis directly and non-invasively. Ultrasound (US) elastography is categorized into shear wave elastography and strain elastography. We have reported on the usefulness of Real time tissue elastography (RTE) as strain elastography in patients with chronic hepatitis C (CHC) [J Gastroenterol, 2011]. We show here a comparison of the diagnostic performance of RTE with that of FibroScan (FS) as shear wave elastography in patients with liver diseases. Patients and Methods: From October 2010 through May 2013, seven hundred and twenty seven liver disease patients received simultaneous RTE and FS routine examinations upon admission by a fixed sonographer. Etiologies of liver diseases were included hepatocellular carcinoma (n=255, 35%), CHC (n=245, 34%), chronic hepatitis B (n=55, 8%), non-alcoholic steatohepatitis (n=49, 7%), and others (n=123, 17%). RTE was performed using EUB-8500 and Ascendus, both with EUP-L52 Linear probe, 3-7 MHz (Hitachi Medical, Kashiwa). Details of the technical procedures have been described. For quantitative analysis, Mean and LF index as the image features of tissue elasticity were obtained from RTE images using a novel software Elasto_ver1.5.1. Results: Table 1 shows the following results, A) Rate of unreliable results of the procedures, B) Relationship between liver stiffness and laboratory data, and C) Diagnostic value of liver fibrosis from RTE and FS. A) In FS, unreliable results were obtained further than RTE. B) Simple regression analyses indicated that the correlations between elastography and indirect markers of fibrosis were not obtained higher than previous reports. C) The area under the receiver operating characteristic curve (AUC) for stage F0-2 was 0.80, 0.79 and 0.87 for LF index, Mean, and FS, respectively. The AUC for cirrhosis (F4) was 0.79, 0.78, and 0.84 for each of them. Conclusion: RTE and FS are useful for detecting the degree of fibrosis in patients with liver disease. Since these procedures were noninvasive, useful, and convenient, US elastography should become a standard clinical examination.

Table 1

AB, platelet countB'PTAPRIC, FO-2 vs F3-4C, F0-3 vs F4
  1. PT prothrombin time, APRI aspartate aminotransferase-to-platelet ratio index.

  2. *p<0.001, ns no significant difference.

number of patientsn=727n=493n=493n=493n=187n=187
LF index/r=0.23*r=0.36*nsAUC 0.80AUC 0.79
Mean/nsr^0.24*r^0.21*AUC 0.79AUC 0.78
FS11.6% (n=84)r=0.18*r=0.28*r=0.19*AUC 0.87AUC 0.84


Akihiro Tamori - Grant/Research Support: MSD

The following people have nothing to disclose: Hiroyasu Morikawa, Sawako K. Uchida, Hideki Fujii, Atsushi Hagihara, Etsushi Kawamura, Shuji Iwai, Masaru Enomoto, Yoshiki Murakami, Norifumi Kawada


Radiological And Pathological Correlation For Hcc Treated With Liver Transplantation

Mohomed R. AbdelfoffohHodeel Al-mono3, Mohomed Neimatallah4' Hussien Elsiesy1' Mohommed Al-seboyel1, Dieter C. Broering1

1liver tronsplont ond hepatobiliay surgery king foisol specialist hospital ond reoseorch center' Riyadh' Saudi Arabia; 2hepatobiliary and pancreatic surgery, University of Alexandria, Faculty of Medicine' Alexandria' Egypt; 3Deporfmenf of Pathology and Laboratory Medicine, king faisal specialist hospital and reoseorch cenfer, Riyadh' Soudi Arabia; ‘Department of Rodiology, king faisal specialist hospital and reasearch center, Riyadh, Saudi Arabia

Introduction: Hepatocellular carcinoma (HCC) is the fifth most common cancer in the world. The 5-year survival rate for untreated, symptomatic HCC is < 5%. In contrast, the 5- year survival rate in patients with cirrhosis following transplantation of small (2 cm) HCC is 80%. The detection of small HCC is clearly critical to patient outcome. Although many CT and magnetic resonance (MR) imaging studies have reported high diagnostic accuracy for HCC and Dysplastic Nodules (DN) in patients with cirrhosis, most of these have been limited by study design, incomplete pathologic correlation and suboptimal imaging techniques. Correlation between explant pathology and pretransplant radiology is of prime importance- among other factors- to choose the suitable line of treatment for liver nodules in cirrhotic patients including follow up, locoregional therapy, liver transplantation or paliative treatment. Methods: 100 patients who underwent liver transplantation at our institution between 2002 and 2013 for the presence of HCC were retrospectively reviewed. Liver transplantation was performed either as primary treatment or following bridging locoregional treatment. HCC was radiologically diagnosed. Radiological diagnosis was performed using one or two contrast enhanced dynamic imaging studies including Multidetector computed tomography (MDCT) and MRI. Pathological examination was made using whole liver explant examination by senior pathologists who have experience in liver pathology. Pathologists had knowledge about the pretransplant radiological findings. Radiological and pathological correlation was made between explant pathology and radiological findings. Correlation was made on per nodule level including size, location and nature. Results: A total 230 nodules were identified in explant pathology from 100 liver transplant patients. Overall; 208 nodules were radiologically identified before transplant (90.4%), while pretransplant imaging modalities failed to show 22 nodules (9.6%). Out of the missed nodules 10 HCC lesions and 12 dysplastic nodules (4.3% and 5.2 % out of all lesions respectively) were pathologically identified. Out of the identified nodules 59 were misinterpreted.114 nodules were found to be more than or equal to 2cm in maximum dimensions by pathology (group 1)compared to 67 between 1and 2 cm (group2) and 49 less than 1 cm (group 3). percentage of the missed or misinterpreted nodules was significantly less in first group as compared to the other two group, (p=0.000) Conclusion: pretransplant imaging modalities are very relaible in diagnosis of cirrhotic liver nodules specially in smalll lesions


Hussien Elsiesy - Speaking and Teaching: ROCHE, BMS, JSK

The following people have nothing to disclose: Mohamed R. Abdelfattah, Hadeel Al-mana, Mohamed Neimatallah, Mohammed Al-sebayel, Dieter C. Broering


Liver Stiffness Measurement Predicts Risk of Decompensation, Hepatocellular Cancer and Mortality in Patients with Chronic Liver Diseases: A Systematic Review and Meta-analysis

Siddharth Singh1, Larissa L. Fujii1, M. Hassan Murad3, Zhen Wang3, Sumeet Asrani4, Richard Ehman2, Patrick S. Kamath1, Jayant A. Talwalkar

1Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN; 2RadioIogy, Mayo Clinic, Rochester, MN; 3Knowledge and Evaluation Research Unit, Mayo Clinic, Rochester, MN; 4Baylor University Medical Center, Dallas, TX

Background: Liver stiffness measurement (LSM) as measured by elastography may independently predict the development of clinical outcomes in patients with chronic liver diseases (CLD). However, the results from individual studies are inconsistent. Aims: To perform a systematic review and meta-analysis of studies evaluating the association between LSM and subsequent risk of clinically relevant outcomes in patients with CLD. Methods: We performed a systematic literature search up to February 2013, for all cohort studies reporting the association between baseline LSM and subsequent development of decompensated cirrhosis, hepatocellular cancer (HCC) and/or mortality, in patients with CLD. When studies reported exposure grouped into categories to provide a dose-specific relative risk (RR) (using the lowest category as referent category), we imputed a risk estimate per unit of LSM, using linear trend metaanalytic statistical methodology. Summary adjusted RR estimates per unit of LSM and 95% confidence intervals (CI) were estimated using the random effects model. Results: Seventeen studies, reporting on 7058 patients with CLD, we included. In patients with compensated cirrhosis, baseline LSM was significantly associated with subsequent risk of hepatic decompensation, in a dose-dependent manner (6 studies; RR, 1.07; 95% Cl, 1.03-1.11). The results were stable across etiology and stage of CLD, geographic location and across different modalities of LSM. On meta-analysis of 9 studies in patients with CLD, high baseline LSM predicted future risk of development of HCC (RR, 1.11; 95% Cl, 1.05-1.18). The results were consistent across Asian and Western population, and studies which included patients with compensated cirrhosis only or all stages of CLD. Likewise, baseline LSM predicted the future risk of mortality (5 studies; RR, 1.22; 95% Cl, 1.05-1.43) as well as a composite of these outcomes (7 studies; RR, 1.32; 95% Cl, 1.16-1.51). Considerable heterogeneity was observed, primarily in the magnitude of effect and not in the direction of effect. Metaregression analysis was not able to explain heterogeneity based on stage of CLD, geographic location, level of adjustment for confounding variables in individual studies or method of data imputation. Conclusion: Based on meta-analysis, liver stiffness measurement is an independent predictor of the risk for decompensated cirrhosis, HCC and mortality in patients with CLD. These data suggest that LSM may be clinically useful in assessing prognosis among individuals with CLD.


Richard Ehman - Board Membership: Resoundant Inc; Management Position: Resoundant Inc; Patent Held/Filed: Mayo Clinic / GE, Mayo Clinic / GE; Stock Shareholder: Resoundant Inc.

Jayant A. Talwalkar - Consulting: Lumena; Grant/Research Support: Intercept, Salix, Gilead

The following people have nothing to disclose: Siddharth Singh, Larissa L. Fujii, M. Hassan Murad, Zhen Wang, Sumeet Asmani, Patrick S. Kamath


Growing burden of HCV-related advanced liver disease related to enhanced Fibroscan-based screening

Mark Danta1,3, Dianne How-Chow3, Elizabeth Mclnnes4, Gail Matthews2,3, Gregory J. Dore2,3

1St Vincent's Clinical School, Sydney, NSW, Australia; 2Kirby Institute, Kirby Instiute, Sydney, NSW, Australia; 3St Vincent's Hospial' Sydney NSW Ausmoia; 4Australian Catholic University, Sydney, NSW, Australia

Background: The estimated burden of HCV-related advanced liver disease in Australia is escalating, related to high HCV prevalence and an “ageing cohort” effect. The introduction of Fibroscan-based liver disease staging has the potential to both increase and accelerate HCV disease assessment. The aim of this study was to evaluate the burden of cirrhosis through Fibroscan-based assessment. Methods: All initial Fibroscan assessments for HCV-infected patients were included, since incorporation into clinical assessment at St Vincent's Hospital, Sydney from late 2008-2012. The proportion of patients with Fibroscan-based cirrhosis (≥13.0 kPa) was determined for the total study period, and by year. Fibroscan score was then correlated with demographic, clinical and treatment data for the cohort. Results: Over the period 2009-2012, 884 HCV-infected patients (17% with HIV or HBV co-infection) underwent Fibroscan-based disease staging, with 1 33 (15%) identified with cirrhosis on their initial assessment. The cirrhotic cohort was older (52 v 49 years) and more likely male (77 vs 65%) compared with the non-cirrhotic cohort (≥13 kPa). Interestingly, there was no difference in HIV rate between cohorts. Among those with cirrhosis, Fibroscan score was 13-29 kPa (74%), 3049 kPa (21%), and 50+ kPa (5%). There was no correlation between Fibroscan score and ALT (Spearman's r=-0.26). The proportion of patients with cirrhosis on their initial assessment has been relatively stable (2009, 39/227 (17%); 2010, 44/284 (15%); 2011, 42/277 (15%)), however, the total number of patients with identified cirrhosis requiring clinical management is growing rapidly. Of the 63 (47%) cirrhotics treated, there was no difference between median Fibroscan score in those with an SVR (24 kPa) Vs no SVR (21 kPa) following treatment (Wilcoxian rank=0.62). Longitudinal followup revealed significant regression of fibrosis in 6 of 7 individuals following an SVR. Over the entire study period, 36 (27%) of the cirrhotic cohort required a hospital admission. Conclusion: Fibroscan-based staging has enhanced overall disease assessment and enabled identification of large numbers of patients with HCV-related cirrhosis requiring follow-up. Consequently, there is a growing need for clinical management programs directed towards HCV-related advanced liver disease which will require considerable further investment in HCVrelated clinical care.


Gail Matthews - Consulting: Viiv; Grant/Research Support: Gilead Sciences; Speaking and Teaching: BMS, MSD

Gregory J. Dore - Board Membership: Roche, Merck, Janssen, Gilead, BristolMyers Squibb, Abbvie; Grant/Research Support: Roche, Merck, Janssen, Gilead, Bristol-Myers Squibb, Abbvie, Vertex; Speaking and Teaching: Roche, Merck, Janssen, Gilead

The following people have nothing to disclose: Mark Danta, Dianne How-Chow, Elizabeth Mclnnes


Portal hypertension and collateral circulation can influence on the estimation of liver stiffness measurement by transient elastography

Moon Young Kim1,2, Soon Koo Baik1,2, Mee-Yon Cho3, Youn Zoo Cho1, Won Ki Hong1, Hye Won Hwang1, Jin Hyung Lee1, Myeong Hun Chae1,Seung Yong Shin1, Jung Min Kim1,Sang Ok Kwon1, Dong Joon Kim4, Ki Tae Suk4, Gab Jin Cheon5, Young Don Kim6, Doe Hee Choi6

1lnternal medicine, Yonsei University Wonju college of medicine, Wonju, Republic of Korea; 2Cell Theropy ond Tissue Engineering, Wonju Christian Hospifol' Yonsei University Wonju College of Medicine, Wonju, Republic of Korea; 3Pathology, Wonju Christian Hospital, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea; 4lnfernol Medicine' Chuncheon Sacred Heart Hospital, Hallym University College of Medicine, Chuncheon, Republic of Korea; 5lnfernol Medicine' Kangneung Asan Hospital, Ulsan University College of Medicine, Kangneung' Republic of Korea; 6Internal Medicine' Kongwon University School of Medicine, Chuncheon, Republic of Korea

Backgrounds and aim: Liver stiffness(LS) measurement using transient elastography has been proposed as a noninvasive method for the prediction of the severity of hepatic fibrosis. However, LSM is influenced by meal, hepatitis or cholestasis. Recently, some studies showed the relationship between LS and hepatic venous pressure gradient(HVPG). However the influence of portal hypertension on LS has not been known. So we evaluated the change of strength of relationship between LS and histologic grades after 3months propranolol treatment. Methods: LSM and HVPG were performed at baseline and after 3month propranolol treatment in 61 consecutive cirrhotic patients(male, 52 (85.2%)) who had biopsy proven cirrhosis with HVPG≥12mmHg were included. Linear regression analysis was performed for evaluation of relationship between ΔLS [%, (baseline LS - follow-up LSE) / baseline LS x 100] and ΔHVPG [%, (baseline HVPG - follow-up HVPG) / baseline HVPG x 100]. Results: The etiologies of cirrhosis were composed of alcohol and HBV(40 and 21 patients, respectively). The baseline mean HVPG and LS were 17.3±4.1mmHg(12-27) and 4o.3±18.3kPa, respectively and these showed significant correlation(r2 = 0.24, p < 0.0001). Follow up HVPG and LS(13.0±4.8mmHg(4-21) and 33.3±17.4kPa, respectively) after propranolol treatment also showed significant correlation(r2 = 0.46, p < 0.0001). A strong positive relationship between ΔLS(%) and ΔHVPG(%) was also observed in the overall population (r2=0.34, p < 0.0001). Thirty three patients(37/61, 60.5%) were propranolol responders. In responder group, baseline LS correlated with the baseline ΔHVPG(r2=0.29, p<0.001) and it more closely correlated with the HVPG after propranolol treatment(r2=0.58, p<0.001) but there was no correlation in nonresponders. Baseline LS correlated with the Laennec histologic grades(r2=0.27, p<0.001) and it showed more close correlation with histologic grades(r2=0.37, p<0.001) after propranolol treatment. In responder group, LS showed more significant improved correlation with the histologic grades(r2=0.27 vs. r2=0.40, p<0.001) after propranolol treatment, however there was no significant change in nonresponder(r2=0.23 vs. r2=0.27, p>0.05). Conclusion: The interval change of LS showed significant correlation with the change of HVPG after propranolol treatment. Improved correlation of adjusted LSM by propranolol treatment with histologic grades suggested that LS is also influenced by portal hypertension in patients with clinically significant portal hypertension. Key Words: Portal hypertension, Liver stiffness measurement, Propranolol, Cirrhosis.


The following people have nothing to disclose: Moon Young Kim, Soon Koo Baik, Mee-Yon Cho, Youn Zoo Cho, Won Ki Hong, Hye Won Hwang, Jin Hyung Lee, Myeong Hun Chae, Seung Yong Shin, Jung Min Kim, Sang Ok Kwon, Dong Joon Kim, Ki Tae Suk, Gab Jin Cheon, Young Don Kim, Dae Hee Choi


Cholangioscopy Performed by SpyGlass Direct Visualization System and Our 3 Years Experiences: What Is The Role In The Diagnosis of Indeterminate Biliary Lesions

Sodeffin HOlagu, Omer Senfurk, Gokfug Sirin, Alfoy Celebi

Gostroenterology, Kocaeli University Medical Faculty Hospital, Kocaeli, Turkey

Introduction: Difficulties in obtaining histological diagnosis in biliary strictures during endoscopic retrograde cholangiopancreatography (ERCP) necessitates the use of further diagnostic techniques. Aim: We aimed to assess the feasibility, clinical utility and the safety of cholangioscopy procedure. Method: Medical records of peroral cholangioscopies performed in Kocaeli University Gastroenterology department between November 2009 and December 2012 by using SpyGlass Direct Visualization System (SDVS), designed for peroral cholangioscopy (Boston Scientific, Massachusetts, USA), were analyzed prospectively. Cholangioscopy was performed in 51 patients (21 male and 30 female patients). Mean age was 57, 1(28-81) and 59, 6(41-84) for male and female patients respectively. Thirty-six patients with indeterminate biliary strictures and filling defects who had inconclusive results on previous biliary ductal tissue sampling. Results: Cholangioscopy was performed successfully in all patients. Cholangiocarcinoma was diagnosed in 25 patients. While brush cytology yielded the diagnosis of carcinoma in only four patient, biopsies taken by skybite yielded the diagnosis of carcinoma in 25 patients (good in 16, moderate in 7, poor in 2). In other 6 patients appearance was consistent with Primary Sclerosing Cholangitis (PSC). Benign strictures were detected in 9 patients. Three of these patients with benign strictures and eight other patients had choledocolithiasis. In patients with choledocolithiasis, the size and the number of stones were bigger than the ones reported on conventional imaging studies. In two patient appearance was consistent with Caroli Disease. Biliary tract was normal in four patients suspected of having Klatskin tumor (2), benign strictures and PSC. The overall accuracy of SpyGlass visual impression for differentiating malignant from benign ductal lesions was 80% (20/25). The accuracy of SpyBite biopsies for differentiating malignant from benign ductal lesions that were inconclusive on ERCP-guided brushing or biopsy was 75.6% (25/33) in an intent-to-treat analysis. Diagnostic SDVS procedures altered clinical management in 64% of patients. Spontaneous gallbladder perforation was observed three days after the procedure in the patient managed with choledochal balloon dilatation(Case two) and early cholangitis in two patients. Conclusion: SpyGlass cholangioscopy with SpyBite biopsies has a high accuracy with regard to confirming or excluding malignancy in patients with indeterminate biliary lesions


The followinq people have nothing to disclose: Sadettin HÜlaqu, Omar, Sentumk, Goktug Sirin, Altay Celebi


Liver stiffness by shear-wave elastography in Alcoholic Liver Disease

Luisa Vonghia1,2, Wim Verlinden1, Laura Cauwenberghs1, Anton Steinhauser 1,Jurgen L. Van Dongen1,Veerle Van Marck3, Paul A. Pelckmons1, Peter P Michielsen1, Sven M. Francque1

1Gastroenterology and hepatology, University Hospital Antwerp, Antwerp, Belgium; 2Department of Basic Medical Sciences, University of Bari, Bari, Italy; 3Department of Patology, University Hospital Antwerp, Antwerp, Belgium

Background: Liver stiffness (LS) measurement is a validated tool in the non invasive assessment of liver fibrosis in several chronic liver diseases including alcoholic liver disease (ALD). Cut-off values may differ according to the aetiology of the disease and the technique used. Real-time shear wave elastography (SWE) is an emerging ultrasound guided technique that allows a real time visualization of liver elastography that needs validation. Aim: To study the correlation bewteen liver stiffness measured by SWE and liver histology in patients affected by ALD. Methods: Patients affected by ALD who were scheduled for a liver biopsy after clinical evaluation were consecutively enrolled. Twentytwo patients met the inclusion criteria and underwent abdominal ulatrasound, real-time SWE using Airexplorer (Supersonic Shear Imagine S. A., Aix-en-Provence, France) and liver biopsy at one single time point. Liver histology was assessed using the NASH CRN scoring system. Results: The patients enrolled (mean age 52±11 years, mean BMI 24.7±5 Kg/m2) presented a diagnosis of cirrhosis in 45.5% of the cases. The mean alcohol intake in the study population was 73 ±88 g/day; 27% of the patients were abstinent at the moment of the liver assessment. LS measures increased concomitantly with the severity of fibrosis (Friedman test p<0.0001). LS was significantly higher in the F4 group versus F0-F3 (Mann-Withney test p=0.018). There was a positive correlation between LS and Child-Pugh and MELD scores (respectively: r=0.63, p=0.001 and r=0.54, p=0.09, Spearmann test). The analyses of receiver operating characteristics (ROC) curves for F0-F3 versus F4 showed an area under the ROC of 0.8 (95% Cl: 0.6-0.9). The cut off value of 13.1 kPa showed a sensitivity of 0.9 and a specificity of 0.6, a NPV of 0.8 and a PPV of 0.7. The intraobserver agreement was 0.9(95%CI: 0.86-0.96). Conclusion: SWE is a useful tool in the assessment of liver fibrosis in patients with ALD. A cut-off of 13.1 kPa has a high NPV and hence can reliably be used to exclude cirrhosis.


The followinq people have nothinq to disclose: Luisa Vonqhia, Wim Vemlinden, Laura Cauwenbemqhs, Anton Steinhause;, Jurgen L. Van Donqen, Veerle Van Mamck, Paul A. Pelckmans, Pete; P. Michielsen, Sven M. Fmancque


Body Mass Index, but not Waist Circumference, is an Independent Predictor for Invalid Transient Elastography Studies

Juanda Leo Hartono, Liang Tze Lim, Yock Young Dan, Guan Huei Lee, Seng Gee Lim, Kieron B. Lim

Gastroenterology and Hepatology, National University Hospital Singapore, Singapore, Singapore

Background/Aims Transient elastography (TE) is a well validated tool which is increasingly utilized in clinical practice to assess and quantify liver fibrosis. We aim to identify factors associated with invalid TE results in a tertiary referral center in a large prospective cohort study. Methods Consecutive patients who were referred for TE between September 2011 to March 2013 were included. Age, gender, body mass index (BMI) and waist circumference were recorded. An invalid result was defined as failure to capture 10 readings or interquartile range (IQR) of more than 30%. Patients were assessed with Fibroscan using a medium-sized (M) probe. Results Among the 1919 cases referred, valid results were acquired in 1851 (96.5%). Univariate analysis showed that high waist circumference (p=0.003) and high BMI (p=0.001) were associated with invalid results. Advanced age and female gender were not statistically significant. In multivariate analysis which included age, gender, BMI, and waist circumference, BMI was shown to be the only independent predictor for invalid results (Table 1). The number of invalid TE studies increased with increasing BMI (5.2% in BMI>25 vs 11.5% in BMI>30 vs 26.1% in BMI>35). Conclusion Body mass index is independently associated with invalid results for transient elastography. Patients with BMI >35 should consider other modalities to assess liver fibrosis.

Factors analysed for invalid results during transient elastography

FactorsInvalid ResultsvalueOR (95% CI)Multivariate (P value)
GenderMale39 (3.3%)1138 (96.7%)---
Female29 (3.9%)713 (96.1%)0.5271.11 (0.83-1.47)0.335
Mean Age (years)54.2±12.551.4+13.40.081-0.131
Mean Waist Circumference (inches)34.9 ± 5.2232.9 ±4.30.003-0.681
Mean BMI26.5±5.3624.2±3.90.001-0.020
BMI>25Yes (n=762)40 (5.2%)722 (94.8%)0.0011.51 (1.23-1.86)-
No (n=1157)28 (2.4%)1129 (97.6%)---
BMI>30Yes (n=148)(11.5%)131 (88.5%)<0.0012.5 (2.26-5.50)-
No(n=1771)51(2.9%)1720 (97.1%)---
BMI>35Yes (n=23)6(26.1%)17 (73.9%)<0.0019.61 (3.91-23.59)-
No (n=1896)62 (3.3%)1834 (96.7%)---


Yock Young Dan - Consulting: Merck, Sharp and Dohme

Kieron B. Lim - Consulting: Gilead, GlaxoSmithKline, AstraZeneca; Grant/Research Support: Novartis, Janssen

The followinq people have nothinq to disclose: Juanda Leo Hamtono, Lianq Tze Lim, Guan Huei Lee, Seng Gee Lim


Analysis of risk factors for aiming at early detection of hepatocellular carcinoma

Tomoko Aoki1,2, Hiroko Iijima1,2, Masahiro Yoshida1, Tomoyuki Takashima2, Nobuhiro Aizawa2, Kazunori Yo2, Kenji Hashimoto1,2, Chikage Nakano1,2, Naoto Ikeda2, Hironori Tanaka1,2, Masaki Saito2, Hirayuki Enomoto2, Shuhei Nishiguchi2

1Ultrasound Imaging Center, Hyogo College of Medicine, Nishi nomiya Japan; 2Division of Hepatobiliary ond Pancreatic Disease, Department of Internal Medicine, Hyogo College of Medicine, Nishinomiya, Japan

[Aim] Contracting viral hepatitis and progression to cirrhosis are the known risk factors for liver cancer in patients with chronic liver disease (CLD), though many cases develop into cancer with little clinical manifestation of cirrhosis. We studied the feasibility of using only non-invasive examination to identify high risk cases for cancer. [Methods] We studied 1203 CLD patients, 593 male, mean age 59.3, who underwent Virtual Touch Quantification (VTQ) in our hospital (January ‘07 to January ‘13) and performed multivariate analysis to identify factors associated with shear wave velocity (Vs). Total 849 cases had Shear Wave Elastography (SWE) and total 1717 cases had Fibroscan on the same day of liver biopsy. The correlation between these modalities and VTQ was determined. Among 1203 cases, cases without a history of cancer at the initial visit at our hospital were followed; during the follow-up, 45 cases developed cancer (mid-carc group), 860 cases did not develop cancer (never-carc group).298 cases had a history of cancer at the initial visit (past-carc group). We conducted multivariate analysis to compare the difference between three groups. [Results] In multivariate analysis, fibrosis proven by biopsy showed the strongest correlation with Vs (standardized partial regression coefficient p0.336, correlation coefficient r 0.558, p<0.001). Hyaluronic acid (β0.210), ΔGTP (β0.138), Plt (β0.189), pT-INR (β0.109) also showed significant correlations with Vs (p<0.05). The correlation between VTQ and SWE is expressed as SWE=0.59+0.85xVTQ, and a strong correlation was found (r 0.755, p<0.001). The correlation with Fibroscan is expressed as Fib=-6.05+10.97xVTQ, and was well correlated (r 0.690, p<0.001). In univariate analysis, significantly more women were in never-carc group, and more men in the other two groups (p<0.001). We found significantly higher complication rate of steatosis in never-carc group and significantly less in past-carc group. VTQ, fibrosis marker, tumor marker, glucose level were significantly lower in younger age in never-carc group compared to those in past/mid-carc groups (p<0.001). Liver biopsy showed significantly more number of F1 /A1 cases in never-carc group, and F4/A2 cases in mid/past-carc groups (p<0.05). Multivariate logistic regression analysis identified that stiff liver cases (Odds ratio [OR] 2.38/1.89) in elderly (OR 1.93/2.32) male (OR 2.77/4.24) cases were especially associated with higher risk of cancer development when compared never-carc group to mid/pastcarc groups (p<0.001). [Conclusion] Non-invasive fibrosis diagnosis correlated well with liver fibrosis and was suggested to be useful in cancer screening.


The following people have nothing to disclose: Tomoko Aoki, Hiroko lijima, Masahiro Yoshida, Tomoyuki Takashima, Nobuhiro Aizawa, Kazunori Yo, Kenji Hashimoto, Chikage Nakano, Naoto Ikeda, Hironori Tanaka, Masaki Saito, Hirayuki Enomoto, Shuhei Nishiguchi


Point Quantification Elastography technique for quantification of live

Erica Fiorini 1,3, Fabio Conti1, Elena Mazzotto1,3, Chiara De Molo2, Silvia Righi1, Gabriella Verucchi1, Antonietta D'Errico2, Marco Lenzi1, Claudia Sama1, Pietro Andreone1, Carla Serro1,3

1Medical and Surgical Sciences, University of Bologna, Bologna, Italy; 2Experimental, Diagnostic and Specialty Medicine, University of Bologna, Bologna, Italy; 3lnternal Medicine and Digestive Diseases, St. Orsola-Malpighi Hospital, Bologna, Italy

BACKGROUND AND AIM: Point quantification elastography (PQE) is a new shear wave-based elastography technique to assess liver fibrosis (LF) by measuring liver stiffness (LS) noninvasively. LS is expressed in Young's modulus. The aim of this single-center study was to assess the diagnostic accuracy of PQE in patients with chronic liver disease (CLD) using liver biopsy (LB) as the reference standard. METHODS: Between September 2012 and May 2013, we enrolled 123 consecutive patients (64 males, 59 females; mean age 50±13) scheduled for LB by referring physicians. On the same day, PQE using the ultrasound (US) system iU22 (Philips, Bothell, WA, USA) and USassisted LB were performed.10 PQE measurements were recorded, average LS (PQE-LS) was calculated. LF was staged according to the METAVIR system. In 69 patients, transient elastography (TE) data were also available. Aetiologies of CLD were HCV (57) or HBV infection (21), alcohol (2), non-alcoholic steatohepatitis (10), autoimmune hepatitis (3), primary biliary cirrhosis (2), primary sclerosing cholangitis (1), undefined (14) or a combination of the above aetiologies (13). RESULTS: PQELS was significantly correlated with LF stage (r = 0.647, p<0.001). Optimal cut-off values, sensibility (se) and specificity (sp) for the different levels of LF were determinated by analysis of receiver operating characteristic (ROC) curve: 4.7 kPa for mild LF (F1) (se 63.7%, sp 77.8%), 6.5 kPa for moderate LF (F2) (se 75.0%, sp 86.4%), 7.3 kPa for severe LF (F3) (se 88.6%, sp 86.2%) and 1O.2 kPa for cirrhosis (sensibility 89.5%, specificity 83.5%). There was a statistically significant correlation also between PQE-LS and TE-LS (r=0.796, p<0.001). In patients with PQE and TE data available, the diagnostic performance of the two techniques was assessed by the area under the ROC curve (AUC) analysis for F0 versus F1F4, F0-F1 versus F2- F4, F0-F2 versus F3-F4 and F0-F3 versus F4. AUCs were: 0.70 (95% confidence interval [Cl]: 0.510.89) for PQE and 0.73 (95% CI: 0.61-0.86) for TE, 0.89 (95% Cl: 0.81-0.97) for PQE and 0.91 (95% Cl: 0.85-0.98) for TE, 0.91 (95% Cl: 0.83-1.00) for PQE and 0.91 (95% Cl: 0.84-0.99) for TE and 0.88 (95% CI: 0.80-0.96) for PQE and 0.85 (95% Cl: 0.73-0.97) for TE when comparing F0 versus F1-F4, F0-F1 versus F2- F4, F0-F2 versus F3-F4 and F0-F3 versus F4, respectively. Differences among PQE and TE AUCs were not statistically significant. CONCLUSIONS: This study shows that PQE is a reliable noninvasive method to assess LF, with a diagnostic performance not significantly different from TE. Compared to TE, PQE has the advantage of compute measurements visualizing in real-time the explored area.


Pietro Andreone - Advisory Committees or Review Panels: Roche, Janssen-Cilag, Gilead, MSD/Schering-Plough; Grant/Research Support: Roche, Gilead; Speaking and Teaching: Roche, MSD/Schering-Plough

The following people have nothing to disclose: Erica Fiorini, Fabio Conti, Elena Mazzotta, Chiara De Molo, Silvia Righi, Gabriella Verucchi, Antonietta D'Errico, Marco Lenzi, Claudia Sama, Carla Serra


Spleen stiffness measured by acoustic radiation force impulse (ARFI) imaging predicts survival in patients with liver cirrhosis

Johannes Vermehren, Annika Vermehren, Olga Zimmermann, Andreas Polta, Wolf P. Hofmann, Christoph Sarrazin, Stefan Zeuzem, Mireen Friedrich-Rust

Medizinische Klinik 1, Klinikum der J. W. Goethe-Universität, Frankfurt am Main, Germany

Background & Aim: Acoustic radiation force impulse (ARFI) imaging is a novel non-invasive ultrasound-based method for the evaluation of liver fibrosis and cirrhosis. The relevance of ARFI imaging for outcome of patients with liver cirrhosis has not been established so far. Methods: In this prospective study, consecutive patients with established liver cirrhosis from different etiologies underwent ARFI imaging of the liver and spleen and transient elastography (TE) of the liver. Liver and spleen stiffness measurements were compared with overall survival time or time to transplantation using Cox regression analyses. Results: 158 patients (mean age: 54± 11; male gender: 66.5%; Child-Pugh score B/C: 46%) were included in the study. The median liver and spleen stiffness measured by ARFI was 3.0 m/s and 3.7 m/s, respectively. The median liver stiffness measured by TE was 36.3 kPa. The mean duration of follow-up was 629±390 days. During the follow-up period 44/158 patients died and 25/158 underwent liver transplantation. There was a significant association between stiffness values obtained with all three methods and survival, i. e. patients with higher stiffness values showed shorter survival times. However, in a multivariate Cox regression analysis that included all three modalities, only spleen stiffness measured by ARFI was independently associated with survival (p=0.008; HR: 2.1, 95% Cl: 1.2-3.6). The AUROC of spleen stiffness measurements by ARFI for the prediction of survival was 0.64 (p=0.003). Conclusions: Spleen stiffness measured by ARFI predicts survival in patients with liver cirrhosis.


Christoph Sarrazin - Advisory Committees or Review Panels: Boehringer Ingelheim, Vertex, Janssen, Merck/MSD, Gilead, Roche, Boehringer Ingelheim, Achillion, JaPharmaceuticals, Merck & Co., Idenix, Janssen, Roche Pharma AG, Vertex Pharmaceuticals, Presidio, Santaris, Inc

Mireen Friedrich-Rust - Speaking and Teaching: Siemens Medical, Echosens

The following people have nothing to disclose: Johannes Vermehren, Annika Vermehren, Olga Zimmermann, Andreas Polta, Wolf P. Hofmann


Prospective and comparative study of diagnosis accuracy of liver biopsies and noninvasive markers (Aixplorer and Fibroscan) (ANRS HC EP 29)

Thomas Deffieux1, Jean-Luc Gennisson1, Laurence Bousguet2, Dalila Amroun2, Marion Corouge2, Simona Tripon2, Vincent Mallet2, Philippe Sogni2, Mathias Fink1, Mickael Tanter1, Stanislas Pol2

1ESPCI ParisTech CNRS UMR7587 INSERM U979, Langevin Institute, Paris, France; 2INSERM U1016, Université René Descartes, Service d'Hépatologie, Hôpital Cochin, Paris, France

Background and aims: Supersonic Shear Imaging (SSI) is a new guantitative elastography technigue allowing real-time bidimensional elasticity mapping of liver tissue (Aixplorer, Supersonic Imagine, Aix en Provence, France). In this study, we evaluated its performance for liver fibrosis staging in patients with chronic liver diseases who underwent a liver biopsy and compared the results with those of blood tests (Apri, Fib4, Forns index) and one-dimensional transient elastography (Fibroscan, Echosens, Paris, France). We also investigated a new ultrasonic imaging mode of viscosity measurements and its correlation with fibrosis, activity and steatosis levels. Patients and Methods: 120 patients with chronic liver disease (68 HCV or HBV, 14 with alcoholic liver disease, 9 with NASH, 7 with autoimmune hepatitis, 22 with other causes) were prospectively enrolled. The Metavir fibrosis score were : F0-1: n=63, F2: n=18, F3: n=21, F4: n=18. Among them, 117 patients had a SSI evaluation (probe SC6-1), 110 a Fibroscan (FS) and 94 had biochemical noninvasive markers (Apri, Fib4, and Forns index). The accuracy of SSI, FS and blood tests by comparison with the Metavir fibrosis score were assessed using receiver operator characteristic (ROC) curve analysis. We also estimated the liver viscosity using shear wave spectroscopy technigue and compared the results not only to the fibrosis levels but also to necroinflammatory activity and steatosis levels. Results: The table summarizes the areas under the ROC curves (AUROC) for the different tests in two populations: patients with viral hepatitis and all patients. Viscosity was found to be an average predictor of fibrosis (AUROC = 0.71 F≥ 2, 0.73 for F ≥ 3, and 0.8 for F = 4) but a poor predictor for both activity (AUROC = 0.43 A ≥1, 0.71 for A ≥ 2, and 0.68 for A = 3) and steatosis (AUROC = 0.38 for S ≥ 20%, 0.46 for S ≥ 30%, and 0.39 for S ≥ 40%). Conclusions: The SSI performance is eguivalent to Fibroscan for noninvasive evaluation of fibrosis, and superior to all noninvasive blood tests. They allow a fair delineation of patients (HCV or HBV) who need to be treated. Viscosity could participate in staging liver fibrosis but not steatosis or activity.


METAVIRF>2F>3F = 4F>2F>3F = 4
Viral hepatitisAll patients
AUROC SSI0.860.810.900.820.810.86
AUROC FS0.890.820.850.840.800.85
AUROC APRI0.740.670.650.740.700.70
AUROC Fib 40.720.690.700.760.710.77
AUROC Forns0.790.760.740.790.740.83
AUROC SSI + blood tests0.920.840.920.880.850.91
AUROC FS + blood tests0.90.840.870.870.820.9


Jean-Luc Gennisson - Consulting: Supersonic Imagine

Vincent Mallet - Board Membership: MSD, Janssen; Speaking and Teaching: Roche, Gilead, BMS

Philippe Sogni - Independent Contractor: Roche, MSD, BMS, GILEAD, JANSSEN, Mayoli-Spindler

Mathias Fink - Stock Shareholder: Supersonic Imagine Mickael Tanter - Stock Shareholder: Supersonic Imagine

Stanislas Pol - Board Membership: Sanofi, Bristol-Myers-Sguibb, Boehringer Ingelheim, Tibotec Janssen Cilag, Gilead, Glaxo Smith Kline, Roche, MSD, Novartis; Grant/Research Support: Glaxo Smith Kline, Gilead, Roche, MSD; Speaking and Teaching: Sanofi, Bristol-Myers-Sguibb, Boehringer Ingelheim, Tibotec Janssen Cilag, Gilead, Glaxo Smith Kline, Roche, MSD, Novartis

The following people have nothing to disclose: Thomas Deffieux, Laurence Bousguet, Dalila Amroun, Marion Corouge, Simona Tripon


MRI-estimated pancreatic fat is associated with insulin resistance and is greater in patients with non-alcoholic fatty liver disease than healthy controls

Niraj Patel1, Michael R. Peterson4, Grace Y. Lin4, Richele Bettencourt3, Claude B. Sirlin2, Rohit Loomba5,3

1Department of Internal Medicine, University of California at San Diego, San Diego, CA; 2Liver Imaging Group, Department of Radiology, University of California at San Diego, San Diego, CA; 3Division of Epidemiology, Department of Family and Preventive Medicine, University of California at San Diego, San Diego, CA; 4Department of Pathology, University of California at San Diego, San Diego, CA; 5Division of Gastroenterology, Department of Internal Medicine, University of California at San Diego, San Diego, CA

Background: Metabolic syndrome is associated with fat deposition in the liver and pancreas. Using a novel magnetic resonance imaging (MRI) technigue, we previously showed that pancreatic steatosis may be related to hepatic steatosis in patients with biopsy-proven non-alcoholic fatty liver disease (NAFLD) (Patel et al. APT 2013). This study was limited by the lack of a control group. In addition, the association between insulin resistance and pancreatic fat remains to be evaluated. Aim: To compare pancreatic fat in patients with NAFLD and healthy controls using a novel MRI technigue and to determine whether pancreatic fat is associated with hepatic steatosis and insulin resistance. Methods: A nested case-control study was derived from two cross-sectional studies of 43 adults with biopsy-proven NAFLD and 49 healthy controls who underwent clinical evaluation, biochemical testing and MRI. Pancreas and liver fat were guantified using a validated MRI measurement, the proton density fat fraction (PDFF). Results: Compared to controls, patients with NAFLD had a higher BMI (31.5 vs.25.5, P< 0.001) and higher proportion of males (55.8% vs.22.5%, P < 


0.001) (Table). MRI-determined pancreatic PDFF was significantly higher in the NAFLD group than healthy controls (8.5% vs.3.6%, P < 0.001). In multivariable-adjusted (age-sex-BMIdiabetes adjusted) linear regression models, pancreatic PDFF remained significantly higher in NAFLD patients than healthy controls (P = 0.003). Pancreatic PDFF showed significant correlation with liver PDFF (Spearman correlation = 0.57, P < 0.001). Among all patients, a homeostatic model assessment of insulin resistance (HOMA-IR) score above 2.5 (the median value) was associated with a higher pancreatic PDFF (7.3% vs.4.5%, P = 0.015) and liver PDFF (13.5% vs.4.0%, P < 0.001) than HOMA-IR < 2.5. Conclusion: Patients with NAFLD have higher pancreatic fat content than healthy controls, and there is a significant correlation between pancreatic fat and liver fat guantified by PDFF. In addition, insulin resistance is associated with increased pancreatic and liver fat.

  1. Standard deviation shown in parentheses. * P-value for MRI PDFF Pancreas and MRI PDFF Liver adjusted for differences in age, sex, BMI and diabetes.

NAFLD Patients (n=43)Healthy Controls (n=49) P-value
Age (years) 48.4(11.9) 43.2 (20.0) 0.14
Sex (% male)55.8%22.5% <0.001
BMI (kg/m2) 31.5(4.6) 25.5 (7.2) <0.001
Diabetic (%)34.9%32.7%0.74
MRI PDFF Pancreas 8.5% (6.6)3.6% (2.3) 0.03*
MRI PDFF Liver 15.9% (6.7)2.5% (0.9)<0.001*


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

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

The following people have nothing to disclose: Niraj Patel, Michael R. Peterson, Grace Y. Lin, Richele Bettencourt


Quantification of hepatic blood flow using a high-resolution phase-contrast MR imaging seguence

Ashley Knight-Greenfield, Hadrien Dyvorne, Cecilia Besa, Nancy Cooper, Thomas D. Schiano, Bachir Taouli

Icahn School of Medicine at Mount Sinai, New York, NY

Background: Phase contrast Magnetic Resonance Imaging (PCMRI) is a non-invasive technigue used to measure blood flow in the liver; however long acguisition times can limit its use in the clinic. Purpose: To measure blood flow in the portal vein (PV) and hepatic artery (HA) using a high resolution highly accelerated compressed sensing (PC-SPARSE) technigue and to correlate hepatic flow parameters with the presence of PH (portal hypertension). Methods: This was a retrospective, IRB approved study in 76 patients (M/F 48/28, mean age 54 y) who underwent MRI including PC-MRI. Flow, mean velocity, and vessel area were measured in the PV and HA. Arterial fraction (ART = HA flow/[HA flow + PV flow]*100) was calculated. PH score was calculated based on MRI findings. Mann Whitney test and Spearman rank correlation coefficient were used to test sensitivity to PH and Child-Pugh class. ROC analysis was performed for detection of PH and Child-Pugh class B and C. Results: 40/76 patients were cirrhotic, and 37 patients with cirrhosis had PH. PV velocity and flow were significantly lower in PH (velocity: 8.9 ± 3.1 vs.12.3 ± 2.8, p < 0.001; flow: 16.3 ± 7.7 vs.19.7 ± 7.0, p = 0.03). PV velocity was also lower in patients with Child-Pugh class B and C vs. class A (7.7 ± 3.9 vs.10.1 ± 2.1, p = 0.047). PV velocity correlated negatively with PH score (r = -0.516, p < 0.001), and ART correlated positively with Child-Pugh class (r = 0.491, p = 0.045). PV velocity had an AUC of 0.795 for detection of PH. Conclusions: A highly accelerated compressed sensing phase-contrast MRI technigue produces high-resolution images for hepatic flow measurement. PV velocity is promising for detection of PH and monitoring of PH treatment. Prospective studies with HVPG correlation and 4D flow to further assess the HA and its utility in PH diagnosis should be performed.

65 year old patient with cirrhosis and portal hypertension. PV (portal vein): velocity 6.9 cm/s, flow 7.4 ml/s, both low. HA (hepatic artery): velocity 24.8 cm/s, flow 5.2 ml/s; ART (arterial fraction) is high at 41.1%


The following people have nothing to disclose: Ashley Knight-Greenfield, Hadrien

Dyvorne, Cecilia Besa, Nancy Cooper, Thomas D. Schiano, Bachir Taouli


Defining the risk of cirrhosis in patients with transfused haemoglobinopathies

Edward Shelton1, Lani Shochet2, Chia Pei Chong2, Jamie Cheong2, Sim Yee Ong1, Don Bowden2, Alexander Hodge1,Virginia H. Knight1,Sant-Rayn Pasricha2, Anouk Dev1

1Gastroenterology and Hepotology, Monash Health, Melbourne, VIC, Australia; 2Medical Therapy Unit (Thalassaemia Service), Monash Health, Melbourne, VIC, Australia

BACKGROUND: Transfused haemoglobinopathy (TH) patients are at significant risk of liver cirrhosis and its seguelae due to hepatic iron loading and transfusion related hepatitis C (HCV). Screening for liver fibrosis in this population is inadeguate using current methods - pathology, liver ultrasound and T2*MRI. Transient elastography (TE) non-invasively assesses liver stiffness and hence, risk of cirrhosis and has been validated in many clinical scenarios including viral hepatitis. It has been studied in small cohorts of patients with beta thalassemia. The present study aimed to evaluate the prevalence of cirrhosis in a cohort of adult TH patients using TE. METHODS: 128 TH patients were identified by enrolment at the State Thalassaemia reference centre (August-November 2012).63 patients (males 46%, B thalassemia major 95%, HCV Ab positive 54%) prospectively underwent TE. Liver ultrasound, T2*MRI and present and historical ferritin, data were collected. Associations between risk factors and logeTE were compared by linear regression, and associations between TE thresholds (>7.9kPa for F≥2, >10.3 for F≥3, >11.9 for F=4) versus normal, by logistic regression. RESULTS: 18/63 (29%) had evidence of fibrosis, including 7/63 (11%) with cirrhosis by TE (of whom the diagnosis was not been previously known in six). By multiple logistic regression present and 15 year-old ferritin levels, presence of HCV Ab and age independently predicted TE. Current GGT and bilirubin are also associated with high TE scores and may be useful biomarkers for cirrhosis in this population (Table 1). There was no association between liver or cardiac iron loading assessed by T2*MRI and TE Scores. CONCLUSIONS: Fibrosis and cirrhosis are common in patients with TH due to acguisition of blood borne viruses and iron overload, and screening with TE could be used to determine advanced fibrosis. Present and historic ferritin levels are associated with higher TE scores indicating the importance of past liver iron loading despite current improved iron chelation. Liver iron guantification with T2*MRI does not predict liver fibrosis. HCV and iron loading may have an additive effect in fibrosis progression. This population is at risk from chronic liver disease and should undergo appropriate assessment for advanced fibrosis.

Associations with logeTE

Multiple logistic regressionCoefficient [95% CI] B coefficient P value
Ferritin (current) 0.20 [0.07,0.32] 0.33 0.003
Ferritin (1998) 0.18 [0.05, 0.30] 0.28 0.006
Hepatitis C Ab +ve 0.18 [0.00, 0.36] 0.19 0.047
Age 0.02 [0.00, 0.03] 0.25 0.008
Bilirubin 0.02 [0.00, 0.03] 0.25 0.006
GGT 0.27 [0.13, 0.41] 0.38 <0.001


The following people have nothing to disclose: Edward Shelton, Lani Shochet, Chia Pei Chong, Jamie Cheong, Sim Yee Ong, Don Bowden, Alexander Hodge, Virginia H. Knight, Sant-Rayn Pasricha, Anouk Dev


The usefulness of contrast-enhanced ultrasonography in early detection of hepatocellular carcinoma viability after transarterial chemoembolization - pilot study

Moon Young Kim1,2, Soon Koo Baik1,2, Mee-Yon Cho3, Youn Zoo Cho1, Won Ki Hong1, Hye Won Hwang1, Jin Hyung Lee1, Myeong Hun Chae1,Seung Yong Shin1, Jung Min Kim1,Sang Ok Kwon1, Dong Joon Kim4, Ki Tae Suk4, Gab Jin Cheon5, Dae Hee Choi6;

1Internal medicine, Yonsei University Wonju college of medicine, Wonju, Republic of Korea; 2Cell Therapy and Tissue Engineering, Wonju Christian Hospital, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea; 3PathoIogy, Wonju Christian Hospital, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea; 4lnternal Medicine, Chuncheon Sacred Heart Hospital, Hallym University College of Medicine, Chuncheon, Republic of Korea; 5Internal Medicine, Kangneung Asan Hospital, Ulsan University College of Medicine, Kangneung, Republic of Korea; 6Internal Medicine, Kangwon University School of Medicine, Chuncheon, Republic of Korea

Background: The therapeutic effect after transarterial chemoembolization(TACE) is usually assessed by dynamic liver computer tomography(CT) scan at several weeks later from TACE. In general, compact dense deposition of lipiodol is accepted as success sign of TACE. However, dense deposition of lipiodol also could mask the viable HCC tissue enhancement in the CT scan. The size of 2nd generation microbubble ultrasonography contrast agent (UCA) is smaller than red blood cell as about 2.4μm so, the contrast-enhanced ultrasonography (CEUS) using 2nd generation microbubble UCA could be effective in detection of small part of viability and patency of vessel in HCC after TACE without interference in assessment by lipiodol. So, in this preliminary study, we investigated whether the arterial enhancement in CEUS at 4week after TACE can predict or early detect HCC viability compared to CT scan. Methods: Totally, 12 patients were enrolled in this study. They all received CEUS, CT scan and MRI at baseline and 4week, 12week after TACE. The primary end-point was HCC viable tissue detection in Gd-EOM-DTPA-enhanced magnetic resonance imaging (MRI) after 4weeks or 12weeks later. Arterial phase enhancement was defined as positive finding for remained viable HCC in CEUS or CT scan. The independent variable was the positivity of 4week CEUS. Results: Among 12 patients, 8 patients showed positive finding for primary end-point (MRI positive at 4week or 12week). At 4week, CEUS, CT and MRI showed positive findings in 8 (66.7%), 3 (25%) and 4 (33.3%) patients respectively. All Patients who had 4 week CEUS positive finding (n = 8) showed MRI positive and remained viable HCC at 4week or 12week (p=0.002). Among these 8 patients, 5 patients presented 4week CT scan negative. These 5 patients showed all remained HCC viable tissue, 1 patient showed 4week MRI positive finding and the other 4 patients showed 12week MRI positive. On the contrary, among patient who showed 4week CT scan positive without 4week CEUS positive, no one finally diagnosed as having viable HCC positive. Especially, among 8 patients of 4week CEUS positive, 4 patients (50%) did not presented 4week MRI positive and they all finally confirmed to have remained HCC tissue just by 12week MRI test. Conclusions: In assessment of therapeutic response of TACE, early 4week CEUS showed excellent result in diagnosis and prediction of remained viable HCC. However, this result was derived from just small samples as preliminary study and has to be followed by more advanced well designed large population study. Keywords: Hepatocellular carcinoma, transarterial chemoembolization, contrast-enhanced ultrasonography


The following people have nothing to disclose: Moon Young Kim, Soon Koo Baik, Mee-Yon Cho, Youn Zoo Cho, Won Ki Hong, Hye Won Hwang, Jin Hyung Lee, Myeong Hun Chae, Seung Yong Shin, Jung Min Kim, Sang Ok Kwon, Dong Joon Kim, Ki Tae Suk, Gab Jin Cheon, Dae Hee Choi


Noninvasive Detection of liver fibrosis with multiparametric Magnetic Resonance Imaging compared to Transient Elastography

Hadrien Dyvorne1, Guido H. Jajamovich1,M. Isabel Fiel1, Scott L. Friedman1, Douglas T. Dieterich1, Claudia Donnerhack1, Richard Ehman2, Bachir Taouli1

1Icahn School of Medicine at Mount Sinai, New York, NY; 2Mayo Clinic, Rochester, MN

Introduction To assess the diagnostic value of multiparametric MRI including diffusion-weighted imaging (DWI), dynamic contrast-enhanced MRI (DCE MRI), MR elastography (MRE), compared to transient elastography (TE) for detection of liver fibrosis. Methods This study was approved by the local IRB and all subjects gave informed consent upon enrollment.48 subjects underwent MRI and TE exams.48 subjects were enrolled (9 volunteers and 39 with chronic liver disease). DWI was performed using 16 b values and diffusion decay was fitted to the intravoxel incoherent motion model to yield D (true diffusion), PF (perfusion fraction), D* (pseudo diffusion) and ADC (apparent diffusion). DCE MRI was acguired after gadolinium contrast injection and a dual input single compartment model to yield arterial, portal and hepatic flow (Fa, Fp, Ft), arterial fraction (ART), distribution volume (DV), mean transit time (MTT) and time to peak (TTP). Liver stiffness was measured with MRE (LSMRE) and TE (LS-TE). Comparisons between noninvasive modal-ities and fibrosis METAVIR stages findings on histopathology were performed using Spearman correlation. ROC analysis was performed to assess the performance of each technigue for the detection of moderate (F2-F4) or advanced (F3-F4) fibrosis. Results Correlations with fibrosis stage were significant for D (r =-0.58, p<0.001), ADC (r = -0.50, p=0.001), MTT (r = 0.44, p=0.011), TTP (r = 0.47, p=0.005), LS-MRE (r = 0.77, p<0.001) and LS-TE (r = 0.66, p<0.001). For detection of F2-F4, AUROC were 0.803, 0.756, 0.640, 0.869, 0.836, and 0.809 for D, ADC, MTT, TTP, LS-MRE and LS-TE respectively. For detection of F3-F4, AUROC were 0.815, 0.792, 0.719, 0.696, 0.970 and 0.809 for D, ADC, MTT, TTP, LS-MRE and LS-TE respectively (Fig.1). Conclusion MRI had excellent diagnostic performance for non invasive detection of liver fibrosis, egual or better than that of TE.

ROC curves for the defection of METAVIR F2-F4 and F3-F4.



Scott L. Friedman - Advisory Committees or Review Panels: Pfizer Pharmaceutical, Sanofi-Aventis; Consulting: Abbott Laboratories, Conatus Pharm, Exalenz, Genenetch, Glaxo Smith Kline, Hoffman-La Roche, Intercept Pharma, Isis Pharmaceuticals, Melior Discovery, Nitto Denko Corp., Debio Pharm, Synageva, Gilead Pharm., Ironwood Pharma, Alnylam Pharm, Tokai Pharmaceuticals, Bristol Myers Sguibb, Takeda Pharmaceuticals, Nimbus Discovery, Isis Pharmaceuticals; Grant/Research Support: Galectin Therapeutics, Tobira Pharm, Vaccinex Therapeutics; Stock Shareholder: Angion Biomedica

Douglas T. Dieterich - Advisory Committees or Review Panels: Gilead, Genentech, Janssen, achillion, idenix, Merck, Tobira, Boehringer Ingelheim, Tibotec, Inhibitex, Roche, Vertex

Richard Ehman - Board Membership: Resoundant Inc; Management Position: Resoundant Inc; Patent Held/Filed: Mayo Clinic / GE, Mayo Clinic / GE; Stock Shareholder: Resoundant Inc.

The following people have nothing to disclose: Hadrien Dyvorne, Guido H. Jajamovich, M. Isabel Fiel, Claudia Donnerhack, Bachir Taouli


Noninvasive Rapid Acguisition Technigue for Hepatic Fibrosis using Magnetic Resonance Elastography

Veeral Oza1, Arunark Kolipaka2, Suresh Chamarthi2, Robert B. Kirkpatrick1, Adam J. Hanje1, Douglas M. Levin1,Sylvester Black3, Anthony Michaels1

1Division of Gastroenterology, Ohio State University, Columbus, OH; 2Deparfment of Radiology, Ohio State University, Columbus, OH; 3Department of Transplant Surgery, Ohio State University, Columbus, OH

Purpose Magnetic Resonance Elastography (MRE) is a noninvasive modality for the detection of hepatic fibrosis. Currently, MRE reguires the patient to hold their breath for up to twenty two seconds in order to obtain robust stiffness maps (3 cycle). We are currently studying this modality as well as a rapid acguisition technigue that reduces the breath hold time to eleven seconds (1.5 cycle) to determine any significant difference in stiffness values between these two seguences. Materials and Methods Liver MRE was prospectively performed on sixteen non cirrhotic patients using a 1.5T/3T MRI scanner (Avanto/Tim-trio, Siemens Healthcare, Germany). Eight patients were healthy volunteers with no self-reported history of liver disease (control group). Eight patients had known underlying liver disease and underwent MRE as well as an indication liver biopsy (study group). MRE wave images were processed using online reconstruction to report a mean stiffness value (kPa). A percutaneous liver biopsy was performed within 30 days of the MRE. The Metavir scoring system was used by our Hepatobiliary pathologists who were blinded to the MRE results. Comparisons were made using Pearson's correlation for the fibrosis score and MRE stiffness value. Student's t-tests were performed to determine MRE stiffness values between the control and study groups, and the 1.5 and 3 cycle seguences. Results In the study group, 63% had hepatitis C, 25% had autoimmune liver disease, and 12% had nonalcoholic fatty liver disease. The correlation coefficient was 0.609 [95% CI −0.17, 0.92] for the 1.5 cycle MRE stiffness value and fibrosis stage. The correlation coefficient was 0.604 [95% CI −0.18, 0.92] for the 3 cycle MRE stiffness value and fibrosis stage. A significant difference based on stiffness values (p value <0.036) was determined between the control and study groups using 1.5 and 3 cycle seguences. Finally, no significant difference (p value = 0.43) in stiffness values was found between 1.5 and 3 cycle seguences. Conclusion Our experience thus far has shown that the fibrosis stage and MRE stiffness value are only moderately correlated. However, patients with underlying liver disease have a statistically significant higher MRE stiffness score than people without known liver disease. MRE is a safe and effective method for the assessment of liver fibrosis, and the rapid 1.5 cycle technigue appears to be as effective as the 3 cycle technigue.


Arunark Kolipaka - Grant/Research Support: Siemens Healthcare Inc; Speaking and Teaching: Shenzhen Institute of Advance Technology, Shenzhen, China, Society of Cardiovascular Medicine

Adam J. Hanje - Speaking and Teaching: Salix Pharmaceuticals Anthony Michaels - Speaking and Teaching: Merck

The following people have nothing to disclose: Veeral Oza, Suresh Chamarthi, Robert B. Kirkpatrick, Douglas M. Levin, Sylvester Black


Percutaneous ethanol sclerotherapy for huge symptomatic hepatic cysts: long-term follow-up results in a single center

Se Young Jang, Soo Young Park, Won Young Tak, Young Oh Kweon, Jung Gil Park, Sun Young Ahn, Yu Rim Lee, Eun Jeong Kang

Gastroenterolog/Hepatology, Kyungpook National University Hospital, Daegu, Republic of Korea

Background/aims: Although surgical resection has been a gold standard therapy for huge symptomatic hepatic cysts, it is an invasive procedure for a benign disease with moderate morbidity and mortality. Aspiration of cyst fluid and ethanol instillation has been tried as a minimally invasive management modality, but there are no long-term reports on efficacy and safety of this treatment. The purpose of this study is to evaluate the long-term treatment outcome of percutaneous ethanol sclerotherapy in patients with huge symptomatic hepatic cysts. Patients and methods: We followed-up 42 patients who had visited Kyungpook National University Hospital and underwent percutaneous ethanol sclerotherapy for symptomatic, enlarging hepatic cysts. We evaluated the success rate of ethanol sclerotherapy, serial changes in cyst volume, and adverse events related to the procedure. There are 10 male (23.8%) and 32 female (76.2%) patients. The median volume of the cysts were 1047.4 ml (median diameter 12.3 cm, ranging from 6 to 30 cm). Thirty-six patients had abdominal pain due to enlarging cysts and 2 patients had infection in hepatic cysts. After aspiration of hepatic cyst fluid, 99% ethanol was replaced into the cyst for 20 min in supine, bilateral decubitus and prone posi-tion. Patients were closely monitored for any adverse events for 12 hours. After discharge from hospital, abdomen computed tomogram scans or ultrasonography were followed-up every 6 months. The median follow-up period was 49.5 months. Results: Hepatic cysts were successfully managed in 38 patients (90.5%) by ethanol sclerotherapy. The symptom did not resolve without reduction of cystic volume after ethanol sclerotherapy in 3 patients and there was cyst infection after sclerotherapy in one patient. Three patients underwent surgical resection of hepatic cysts. Among 38 successfully treated patients, 5 patients had hepatic cyst completely disappeared and 33 patients showed reduction of hepatic cysts from 1047.0 ml (12.6 cm in diameter) to 17.2 ml (3.2cm in diameter). There were no immediate complications related to procedure except pain which was manageable with analgesics. Conclusion: Treatment of hepatic cyst is indicated when the cysts are enlarging, cause symptoms or associated with complications. We conclude that huge symptomatic hepatic cyst can be safely treated with percutaneous ethanol injection in selected patients.


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: Se Young Jang, Soo Young Park, Young Oh Kweon, Jung Gil Park, Sun Young Ahn, Yu Rim Lee, Eun Jeong Kang


Non-invasive assessment of liver fibrosis: Acoustic Radiatiation Force Impulse of the left lobe correlates best with Ishak histology score

Sebastiana Atzori, Nimzing Ladep, Yasmin Pasha, Heather Marcinkowski, Vanessa Tooley, Simon D. Taylor-Robinson

Imperial College, London, United Kingdom

Background: Recently, non-invasive assessment of liver disease has become important, since liver biopsy has cost implications and associated complications. This study aimed to compare two technigues for the non-invasive evaluation of liver fibrosis: Acoustic Radiation Force Impulse (ARFI) and transient elastography (TE) using liver biopsy as the gold standard comparator. Methods: 75 patients underwent same-day liver biopsy, and measurement of liver fibrosis with TE and ARFI. Liver biopsy was un-interpretable in 3 patients. Statistical analysis was conducted in 72 patients. Liver fibrosis and necroinflammatory activity were evaluated semiguantitatively according to the METAVIR scoring system and fibrosis was also staged according to the Ishak scoring system. ARFI technology is a software integrated in a conventional ultrasound machine. The measurement was taken at three different sites: 10 measurements in the left part of liver (ARFI L): 10 in the right part of the liver (ARFI R) and 10 in the spleen (ARFI S) for each subject. TE was performed with Fibroscan. Ten valid measurements were taken. Statistical analysis was conducted using graphPad PRISM and Medcalc. The histological staging was correlated with median ARFI and TE and values of the biopsy, both with Metavir and Ishak score. Pearson's correlation coefficient calculated. P values of less than 0.05 were considered statistically significant. Results: 75 consecutive patients with chronic liver disease were enrolled. Most of the patients had chronic hepatitis C (HCV, n=27); chronic hepatitis B (HBV, n=11); or non alcoholic fatty liver disease (n=17); the remainder had other aetiologies. Three patients had liver biopsy because of abnormal liver function tests. Correlations with Ishak scores were higher with ARFI values (Pearson r: ARFI L 0.63; ARFI R 0.68; TE 0.62 and with METAVIR scores (ARFI L r=0.63; ARFI R 0.33; TE 0.57). Correlation with serum aminotransferase/platelets showed a statistical significance for ARFI (ARFI L r=0.67; TE 0.4). In contrast, there were weak correlations with necroinflammatory score (ARFI r=0.23; TE 0.26) and steatosis score (ARFI r=0.24; TE 0.32). Diagnostically, there was a significantly better accuracy of ARFI of the left lobe (ARFI L), compared to ARFI measured in the right lobe (ARFI R) and TE (AUC ARFIL 0.90; ARFI R 0.74; TE 0.73). Conclusion: ARFI of the left lobe performs diagnostically better than TE and correlates well with histological scores of liver fibrosis. As with previous reports, we showed a interlobe variations of liver stiffness. Further validation of our findings is warranted.


Yasmin Pasha - Grant/Research Support: Merz Pharmaceuticals GmbH, Frankfurt, Germany

The following people have nothing to disclose: Sebastiana Atzori, Nimzing Ladep, Heather Marcinkowski, Vanessa Tooley, Simon D. Taylom-Robinson


Is the 18F-FDG PET-CT useful in staging and treatment of HCC?

Suk Bae Kim, Il Han Song, Sun Young Cho, Young Kwang Choo, Sung Soo La, Hyoung Joon Kim

Division of Hepatology and Gastroenterology, Department of Internal Medicine, Dankook University College of Medicine, Cheonan, Republic of Korea

Background: 18F-FDG PET-CT(18F-fiuorodeoxygIucose positron emission tomography-computed tomography) has been widely used in many kinds of malignant tumors. However, the efficacy of 18F-FDG PET-CT in hepatocellular carcinoma(HCC) is still controversy. We aimed to evaluate the usefulness of 18F-FDG PET-CT in staging and treatment of HCC. Methods: We analyzed the HCC patients retrospectively who took 18F-FDG PETCT examination from January 2008 to December 2012. We compared the stage and treatment between before and after 18F-FDG PET-CT to know the efficacy on HCC. We reviewed the medical record, biopsy result, follow-up CT and follow-up data to know the confirmation of the extrahepatic metastasis which was suspected in 18F-FDG PET-CT. Results: Total 160 HCC patients were analyzed.27 patients (16.9%) of them were suspected as extrahepatic metastasis on 18F-FDG PET-CT. High FDG uptake on lung was observed on 18 patients.13 patients of them were already suspected as hematogenous lung metastasis in liver CT.3 patients of them were diagnosed as benign lesion on chest CT and biopsy. Lung masses of 2 patients were detected on only 18F-FDG PET-CT, but there was no changes of staging. High FDG uptake of extraabdominal L/N was found on 2 patients (inguinal and supraclavicular L/N). They were diagnosed as benign L/N enlargement. High FDG uptake on bone was found at 4 patients.3 patients of them showed bone metastasis in liver CT and 1 patient who had metastasis to mandible diagnosed as both adrenal metastasis without staging change.1 patient showed high uptake on prostate and confirmed as benign nodule on biopsy.1 patient with abdominal muscle metastasis was detected on both liver CT and 18F-FDG PET-CT. Skin metastasis was suspected on 1 patient and was confirmed as false positive high uptake of FDG. Conclusions 27 patients of 160 patients were suspected as extrahepatic metastasis on 18F-FDG PET-CT. However there was no change on staging and treatment after 18F-FDG PET-CT because most of them were already suspected on liver CT or confirmed as false positive on biopsy and on other confirmative examinations.


The following people have nothing to disclose: Suk Bae Kim, Il Han Song, Sun Young Cho, Young Kwang Choo, Sung Soo La, Hyoung Joon Kim


Heterogeneous intensity at Gadoxetic Acid-enhancedMR imaging (EOB-MRI) is a feature of early hepatocellular carcinoma (HCC) and associated with heterogenous OATP1B3 expression

Masayuki Nakano1, Tomoaki Ichikawa2, Hiroyuki Morisaka2, Utaroh Motosugi2

1Pathology, Ofunachuo Hospital, Kamakura, Japan; 2Radiology, Yamanashi University, Yamanashi, Japan

Background and Purpose: At EOB-MRI the early HCCs may show a heterogenous singal such as areas of different intensity (from high to low) within in the same lesion. To clarify this issue, we examined the immunocytochemical expression of OATP1B3, which is known to be associated with EOB-MRI intensity. Materials and Methods: Forty-one surgically resected HCCs, detected in as patients, were retrospectively studied. The series included 22 early HCC and 19 advanced small HCCs. All the patients had a EOB-MRI performed before the surgical resection. EOB-MRI signal intensity on the hepatobiliary phase was classified into uneven and even. Immunohistochemical staining was performed and evaluated as follows: 0: no intralesional staining; 1: weaker intralesional staining as compared to surroundings; 2: intralesional staining of the same intensity as surroundings; 3: intralesional staining stronger than surroundings. Results: Age and nodule size of early and advanced HCC were 69.9 and 68.8 yrs and 10.3 and 22.6mm, respectively. EOB-MRI intensity was uneven in 37% early HCC (8/22) and in 27% advanced HCC (5/19). In 7/8 early HCC showing an uneven EOB-MRI signal, OATP1B3 was expressed with a mixed pattern of staining (from 0 to 3 in the same case)(87.5%). In the remaining 14 cases early HCC showing a even signal only 5/14 (36%) cases showed a mixed pattern of OATP1B3 staining. In 5/19 (27%) advanced HCCs showing an uneven EOB-MRI intensity, OATP1 B3 was expressed with a mixed pattern of staining (from 0 to 3 in the same case) 2/5 (40%). Conclusion: Uneven intensity appearance at EOB-MRI in early HCC is not rare and might be related to the heterogeneous expression of OATP1B3. The current classification of EOB-MRI findings in early HCC into 3 groups (low, iso, hyper) does not take into account the possible combination of signals of different intensity in the same tumor, which seems to be feature of earlier than advanced HCC.


The following people have nothing to disclose: Masayuki Nakano, Tomoaki Ichikawa, Hiroyuki Morisaka, Utaroh Motosugi


The yield of Magnetic Resonance Cholangiopancreatography (MRCP) for the investigation of dilated bile ducts in patients with normal liver function tests (LFTs)

Shlomit Tamir1,2, Ofer Benjaminov1,2, Assaf Issachar1,2, Marius Broun1,2

1Rabin Medical Center, Petah Tikvah, Israel; 2Tel Aviv University, Tel Aviv, Israel

Purpose To evaluate the yield of MRCP for the investigation of biliary duct dilatation in patients with normal as compared to those with elevated LFTs. Method and materials This was a retrospective study conducted on MRCP scans of 68 consecutive patients (pts) referred to our tertiary medical center for the evaluation of biliary duct dilatation seen on previous imaging (CT, US). Biochemical data were collected from the medical records: ALT, AST, ALP, GGT, Total Bilirubin, and Direct Bilirubin. The cutoff for elevated LFTs was set at 1.5 times the upper normal limit of either biochemical parameter. Pathological findings were compared between two groups of patients with bile duct dilation: normal versus elevated LFTs. Results Normal LFTs were found in 47 pts and 21 pts had elevated LFTs. Of the 68 pts referred to MRCP for evaluation of biliary dilatation, 53 pts had biliary dilatation confirmed on MRCP. Of the 15 pts without bile duct dilatation, 8 pts were diagnosed with abnormalities in the biliary tree and 7 pts had a completely normal MRI study. MRCP demonstrated the cause of bile duct dilatation in 41 pts (60.3%), more commonly in pts with elevated (n=14, 66.7%) than normal (n= 27, 57.4%) LFTs. Benign pathologies which did not reguire further evaluation or treatment (periampulary diverticula, benign asymptomatic stricture) were demonstrated more commonly in pts with normal LFTs (14/47, 29.8%) than in pts with elevated LFTs (1/21,4.8%). Pathologic findings which reguired further evaluation or treatment (space occupying lesion, choledocholithiasis, severe stricture) were more commonly seen in the elevated LFTs group (13/21, 61.9%) than in the normal LFTs group (13/47, 27.7%), p=0.007. Malignancy was diagnosed in 3 pts. All of them had elevated LFT's (p=0.027). Conclusion MRCP is a valuable tool in the workup of biliary duct dilatation even in the setting of normal LFTs, as the probability of an obstructing pathological finding is as high as 27.7% in those patients. However, it is less likely to find a clinically significant finding, in patients with normal as compared to elevated LFTs. Appropriate criteria should be set for MRCP in patients with incidental biliary dilatation and normal LFTs, weighing the low but significant prevalence of obstructing pathology in these patients.


The following people have nothing to disclose: Shlomit Tamir, Ofer Benjaminov, Assaf Issachar, Marius Braun


The clinical application of a novel 3D virtual liver surgery simulation system

Yukio Oshiro1, Ryoichi Miyamoto1, Ken Nakayama1, Jun Mitani2, Nobuhiro Ohkohchi1

1Department of Surgery,Division of Gastroenterological and hepatobiliary Surgery, and Organ Transplantation, University of Tsukuba, Tsukuba, Japan; 2Department of Computer Science, Graduate School of System Information Engineering, University of Tsukuba, Tsukuba, Japan

Background and aims: The shape of the liver is changed during liver resection due to the patient's respiratory motion and surgical procedure. However, in a conventional method, the 3D liver model is fixed and rigid. Therefore, we aimed to develop a novel 3D virtual simulation system which represents the realtime deformation of the liver, and investigate whether the novel system is useful for hepatic surgery. Methods: i) We developed the novel simulation system “Liversim”,which is programmed to represent the real-time deformation of the liver. This system enables to operate a simulation of hepatectomy while observing the real-time deformation of the liver. In addition, we developed algorithms for cutting the liver and blood vessels. ii) The usefulness of the “Liversim” in 3 patients was assessed by guestionnaires. This investigation targeted medical students and young surgeons. We applied self-assessment scores on a scale of 1 to 10, where 10 indicates “extremely valuable or 100% consistency”. Questions were carried out before and after surgery. Results: i) Between April 2012 and May 2013, 13 patients received preoperative simulation on the “Liversim”. This system enables us to demonstrate the real-time deformation and resection of the liver. The portal and hepatic veins are easily detectable on the resection line. There was no obvious discrepancy during the resection process; i. e., emergence of branches of the portal vein or hepatic vein, depth and direction of the resection line; between our simulation and the real surgery. ii) 1. Scores about usefulness of the “Liversim” were 9.67±0.71 for the student group (n = 9) and 9.56±0.73 for the surgeon group (n = 9). All respondents thought that this system was valuable, with no significant differences between the two groups.2. The scores about the portal vein were 7.88±2.7 for the students (n = 8) and 8.67±0.58 for the surgeons (n = 3). The scores about the hepatic vein were 8.0±2.62 for the students (n = 8), and 8.33±0.58 for the surgeons (n = 3). The scores for the surgeons about the portal and hepatic vein were significantly higher than those for the students. It was revealed that students felt difficulty in recognizing the anatomy of the portal and hepatic vein. Conclusion: This novel system may be useful for realizing the hepatic anatomy and operative procedure to young surgeons and students.


The following people have nothing to disclose: Yukio Oshiro, Tyoichi Miyamoto, Ken Nakayama, Jun Mitani, Nobuhiro Ohkohchi