Portal Hypertension: Experimental


Combination of Granulocyte colony stimulating factor (GCSF) and recombinant erythropoietin (rEPO) improves transplant free survival in patients of decompensated liver cirrhosis

Chandan K. Kedarisetty1, Ankit Bhardwaj2, Chhagan Bihari3, Archana Rastogi3, Nirupma Trehanpati2, Shiv K. Sarin1

1Hepatology, Institute of Liver and Biliary Sciences, New Delhi, India; 2Research, Institute of Liver and Biliary Sciences, New Delhi, India; 3PathoIogy, Institute of Liver and Biliary Sciences, New Delhi, India

Background: Median survival of decompensated cirrhosis is < 2 ears. The only effective treatment is liver transplantation. We had recently shown that use of GCSF enhanced survival in acute- on-chronic liver failure patients (Gastroenterology 2012). EPO has been shown to improve liver regeneration in posthepatectomy and fulminant liver failure rat models. Aim: The present study was to study whether the use of G-CSF and Darbopeitin alpha (DPO) could lead to improved survival and liver regeneration in decompensated cirrhotics. Methods: Consecutive patients with decompensated cirrhosis were enrolled after approval by IRB, to receive 5μg/kg GCSF on day 1,2,3,4,5 and then every third day (total 12) and a weekly dose 40 IU DPO for 1 month (Group A, n=29) or matched placebo (Group B, n= 26). Standard of care continued in both groups. Transjugular liver biopsy (TJLB) and hepatic venous pressure gradient (HVPG) were done at baseline and in a proportion of patients at 1 month. Follow up was done on days 2, 4, 7, 14, 28 and then at month 3, 6, 9, and 12. Results: The two groups were well matched in baseline characteristics. The most common etiology was alcohol, followed by cryptogenic and viral hepatitis. There was significant improvement in CTP scores in Group A from baseline (10.9 ±1.2) to at one (9.67± 1.8), three (9.1 ± 1.7) and six (8.8 ± 1.4) months of follow-up (p < 0.005). After combination therapy significant reduction was seen in reguirement of albumin infusion in Group A from baseline to one month (p- 0.045). There was reduction in need for large volume paracentesis in Group A. Baseline alpha fetoprotein levels were similar in both groups (7.3 ± 6.9 vs 6.6 ± 3.6) . At 1 month, it was 6.6 ± 3.6 ng/ml in Group A and 4.7 ± 2.7 in Group B (p- 0.018), suggestive of enhanced regeneration in Gr. A. No change in HVPG levels was seen from baseline to 1 month (p-0.26). Transplant free survival at 6 months was 82. % in Group A and 50% in Group B (p- 0.009) and at 12 months was 68.9% and 46.2% respectively (p- 0.049). The actuarial probability of survival at 6 and 12 months was significantly higher in Group A but not in Group B. Immunohistochemistry in repeat TJLB of a patient in Group A showed an increase in proportion of CD 34+ cells, CD 133+ cells, Ki-67 and proliferating cell nuclear antigen (PCNA) suggestive of liver regeneration. No adverse events were noted. Conclusions: This is the first RCT (NCT01384565) showing clinical benefit and an improved transplant free survival by using a combination of growth factors; G-CSF and rEPO, in patients with decompensated cirrhosis. The therapy merits further evaluation in patients waiting for liver transplantation.


The following people have nothing to disclose: Chandan K. Kedarisetty, Ankit Bhardwaj, Chhagan Bihari, Archana Rastogi, Nirupma Trehanpati, Shiv K. Sarin


Simvastatin prevents liver microvascular dysfunction and attenuates liver injury in rats with biliary cirrhosis submitted to hemorrage/resuscitation

Cintia Z. Meireles, Marcos Pasarin, Juan José Lozano, Hector Gar cia-Caldero, Jordi Gracia-Sancho, Juan Carlos Garcia-Pagan, Jaime Bosch' Juan G. Abraldes

Barcelona Hepatic Hemodynamic Laboratory. Liver Unit. Hospital Clinic. University of Barcelona. CIBERehd, Barcelona, Spain

Bleeding is associated with high mortality in cirrhosis. Baseline liver function is the main prognostic determinant, and further deterioration in liver function is a major cause of death in these patients. Microvascular dysfunction might contribute to liver injury in cirrhosis, and therapeutic strategies improving liver microcirculation might have a protective effect on liver function. Decompensations, such as infections or bleeding, might aggravate microvascular dysfunction. However, the effects of bleeding on the intrahepatic microvasculature in cirrhosis are largely unknown. The aims of this study were to evaluate the impact of hemorrhage/resuscitation (H/R) on the cirrhotic microcirculation, and whether a drug that improves liver microcirculation has hepatoprotective effects during H/R. Methods: The study was performed in 3 groups of rats: A: controls, B: rats with biliary cirrhosis (CBDL), C: CBDL rats pre-treated with 3 doses (5 mg*Kg-1*day-1) of simvastatin. Rats were submitted to H/R: 60' shock (target MAP of 55 mmHg), followed by 45' resuscitation (with blood/plasma expander 1: 1, aiming at 85% of baseline MAP) and a 120' observation period, or to a sham procedure of egual duration. Subseguently, livers were isolated and perfused, and liver endothelial function was assessed by the response to increasing doses of acetylcholine. Changes in liver transcriptome induced by H/R were assessed with microarrays. Results: Blood withdrawal induced hypovolemic shock of similar intensity in all 3 groups of rats. At the end of the procedure in vivo portal pressure was similar in H/R rats and in those submitted to the sham procedure. In contrast, H/R significantly impaired endotelial-dependent vasorelaxation in cirrhotic (p=0.035) but not in control livers. H/R induced a similar increase in ALT in control and cirrhotic rats (p=0.172), whereas the increase in AST was 10 times higher in cirrhotic than in control rats (p=0.007). Simvastatin prevented the decrease in endotelial-dependent vasorelaxation induced by H/R, and reduced by half the increase in ALT and AST (p<0.05). Transcriptomics showed a marked upregulation of genes related to inflammatory response after H/R in cirrhotic livers, but not in controls, and this was blunted by simvastatin. Conclusion: H/R aggravates liver microvascular dysfunction in cirrhosis, and upregulates liver inflammatory pathways. This does not occur in control livers. Simvastatin prevents H/R-induced liver endothelial dysfunction, and attenuates liver injury and the upregulation of liver inflammatory response, suggesting that it might have potential for protecting the cirrhotic liver during bleeding complications.


Juan Carlos Garcia-Pagan - Grant/Research Support: GORE

Jaime Bosch - Advisory Committees or Review Panels: Intercept pharma; Consulting: Chiasma, Gilead Science, Norgine, ONO-USA; Grant/Research Support: Gore

The following people have nothing to disclose: Cintia Z. Meireles, Marcos Pasarin, Juan José Lozano, Hector Garcia-Caldero, Jordi Gracia-Sancho, Juan G. Abraldes


The role of splenic reservoir monocytes in pulmonary vascular monocyte accumulation in experimental hepatopulmonary syndrome

Wei Wu, Junlan Zhang, Wenli Yang, Binggian Hu, Michael B. Fallon

The university of Texas Health Science Center, Houston, TX

Introduction: Pulmonary intravascular monocyte infiltration plays a significant role in the development of angiogenesis in experimental hepatopulmonary syndrome (HPS) after common bile duct ligation (CBDL) through production of vascular endothelial growth factor-A (VEGF-A). Circulating and hepatic levels of monocytes/macrophages are also increased after CBDL, but the origins of these cells are not defined. Recently, splenic reservoir monocytes have been identified as a major source of monocytes that accumulate in injured tissues and contribute to wound healing. Whether spleen monocytes contribute to monocyte alterations after CBDL has not been defined. Aim: To evaluate monocyte alterations in the circulation, lung and liver in experimental HPS and assess the effect of splenectomy on monocyte levels and pulmonary vascular and hepatic abnormalities. Methods: Sham and 3 week CBDL animals (with or without splenectomy at the time of CBDL) were studied. Lung and liver monocytes were assessed with a specific pan-monocyte marker CD68 and a specific resident macrophage marker CD163 (Western and IH). Peripheral blood monocytes (CD68) were measured by FACS after purification (Ficoll gradient). Lung changes were evaluated with VEGF-A levels (Western) and angiogenesis (factor VIII IH). Liver changes were evaluated with portal venous pressure (PVP), fibrosis (a-SMA, Western or IH) and bile duct proliferation (CK19, Western or IH). Results: Relative to sham, monocyfe levels increased in the lung (CD68) and liver (CD68 and CD163) after CBDL and were accompanied by elevated circulating monocyte numbers (CD68). Splenectomy significantly decreased monocyte levels in the lung (72% reduction, p<0.05 vs CBDL) and in the peripheral circulation (64% reduction, p<0.05 vs CBDL), but significantly increased liver monocyte levels (CD68 and CD163) (1.5±0.08 fold vs CBDL, p<0.05). Lung VEGF-A levels (25% reduction, p<0.05 vs CBDL) and angiogenesis (52% reduction, p<0.05 vs CBDL) significantly decreased after splenectomy, but hepatic άSMA and CK19 levels were increased and portal pressure was unchanged. Conclusion: Increased circulating levels and lung infiltration of monocytes occur after CBDL and are reduced after removal of splenic reservoir in association with improvements in lung vascular abnormalities. In contrast, liver monocytes with resident markers increase and hepatic alterations are exacerbated. Splenic reservoir monocytes are a major source for lung monocytes in experimental HPS, while liver monocytes have different origins. Tissue specific monocyte alterations may have a significant impact on the pathogenesis of pulmonary complications of liver disease.


Michael B. Fallon - Advisory Committees or Review Panels: Bayer/Onyx; Grant/Research Support: Ikaria Therapeutics, Gilead, ANADYS, Mochida, Eaisi, Research Triangle Institute

The following people have nothing to disclose: Wei Wu, Junlan Zhang, Wenli Yang, Binggian Hu


Endothelin-1 (ET-1) stimulates pulmonary microvascular endothelial cell CX3CL1 expression and in vitro angiogenesis: a link between hepatic ET-1 and lung angiogenesis in experimental hepatopulmonary syndrome

Junlan Zhang, Wenli Yang, Binggian Hu, Wei Wu, Michael B. Fallon

The university of Texas Health Science Center, Houston, TX

Introduction: Hepatic production and release of endothelin-1 (ET-1) binding to overexpressed lung microvascular endothelin B (ETB) receptors, stimulates eNOS/NO activation via intracellular Ca2+ signaling pathways and leads to pulmonary vasodilation in experimental hepatopulmonary syndrome (HPS). ET-1/ETB receptor activation is also associated with lung microvascular accumulation of pro-angiogenic monocytes and angiogenesis although the mechanisms are unknown. We have recently found that lung monocyte adhesion and angiogenesis in HPS involves interaction of endothelial CX3CL1 with monocyte CX3CR1, although whether ET-1/ ETB receptor activation influences these events is unknown. Aim: To evaluate the effects of ET-1 on pulmonary vascular endothelial cell CX3CL1 expression and angiogenesis in an in vitro model of the HPS lung microvascular endothelium. Methods: Rat pulmonary microvascular endothelial cells (RPMVEC) were transfected with adenoviral-CMV-ETB receptor (1,000 VP/cell) for 24-48 hours to recapitulate in vivo alterations of HPS. ET-1 was added in the absence or presence of a selective ETB receptor antagonist (BQ788) and specific inhibitors for PLC/InsP3/Ca2+/calmodulin pathway (U73122, 2-APB, BAPTA-AM, W7), MERK/ERK (U0126) and Akt (wortmannin). Cell and supernate CX3CL1 expression was measured by RT-PCR and ELISA. Akt and ERK phosphorylation was assessed by Western blots. In vitro angiogenesis was measured by guantifying endothelial cell tubular structure formation. Results: Relative to untreated cells, ET-1 stimulation resulted in an ETB receptor dependent increase in CX3CL1 mRNA (ET-1: 8.12±0.01 vs BQ788+ET-1: 1.88±0.08, fold-control, p<0.05) and supernate CX3CL1 protein levels (ET1: 2080±70 vs BQ788+ET-1: 670±40, pg/ml, p<0.05). This effect was blocked by inhibitors of PLC/InsP3/Ca2+/calmodulin and MERK/ERK but not of Akt. ET-1 induced ERK activation was Ca2+ pathway independent. ET-1 administration also increased tubular structure formation, which was inhibited by BQ788 or blocking Ca2+ and MERK/ERK pathways. In addition, pre-treatment of RPMVECs with a CX3CR1 neutralizing antibody partially reduced ET-1 stimulated tube formation. Conclusion: ET-1/ETB receptor activation directly modulates pulmonary vascular endothelial cell CX3CL1 expression and in vitro angiogenesis via two independent intracellular signaling pathways, Ca2+ and MERK/ERK. These observations provide a direct link between hepatic ET-1 production and pulmonary vascular alterations in experimental HPS in vivo.


Michael B. Fallon - Advisory Committees or Review Panels: Bayer/Onyx; Grant/Research Support: Ikaria Therapeutics, Gilead, ANADYS, Mochida, Eaisi, Research Triangle Institute

The following people have nothing to disclose: Junlan Zhang, Wenli Yang, Binggian Hu, Wei Wu


Sarcopenia of cirrhosis is mediated by hyperammonemia mediated by a myostatin

Samjhana Thapaliya2, Ari Garber1, Stephen L. Welle3, Srinivasan Dasarathy1,2

1Gastroenterology and hepatology, Cleveland Clinic Foundation, Cleveland, OH; 2Pathobiology, Cleveland Clinic, Cleveland, OH; 3Physiology, University of Rochester, Rochester, NY

Background. We have previously shown that transcriptional upregulation of myostatin (mstd) in the skeletal muscle of cirrhotics and associated with sarcopenia or loss of muscle mass. We now demonstrate that myostatin is a critical intermediate in mediating hyperammonemia mediated sarcopenia. Methods. The post developmental myostatin knockout mice were generated by floxing the 3rd exon of myostatin gene (codes for entire active protein), a transgene encoding tamoxifen-activated Cre recombinase was introduced and mice with floxed (f/f) and normal myostatin genes (w/w) were derived from f/w founders. At 4 months, mice were fed tamoxifen for 2 weeks. At 3 m after tamoxifen, myostatin deletion (mstd-/-) was confirmed by genotyping. This model has been previously validated for increased myofibrillar protein synthesis and increased muscle mass and strength. This model has been previously validated for increased myofibrillar protein synthesis and increased muscle mass and strength. Hyperammonemia was induced by daily intraperitoneal injection of ammonium acetate (2.5 mmol/kg/d) for 28 days. Control animals were administered vehicle alone. Whole body weight, skeletal muscle weight and muscle strength were determined using standard methods used in our laboratory. Results. Myostatin expression, total protein and RNA content and translational efficiency in C2C12 myotubes were significantly lower during hyperammonemia. Myotube size was significantly lower during hyperammonemia (Figure 1). Hyperammonemia resulted in a reduction in whole body weight and gastrocnemius muscle mass (Figure 2) in the control mice while in the myostatin knockout mice (mstd-/-), body weight and gastrocnemius weight were significantly higher. Grip strength was also significantly higher in the mstd/- mice compared to the hyperammonemic mstd+/+ mice (Figure 3). Body weight, muscle mass and muscle strength in the mstd-/- mice were not significantly altered by hyperammonemia. Conclusions. Myostatin is a critical gene responsible for sarcopenia and muscle weakness during hyperammonemia in cirrhosis. Strategies targeting the hyperammonemia-myostatin link would form novel therapeutic options to reverse malnutrition and sarcopenia in liver disease.



The following people have nothing to disclose: Samjhana Thapaliya, Ari Garber, Stephen L. Welle, Srinivasan Dasarathy


Novel Vasopressin 1a Receptor (V1A-R) Partial Agonists Induce Vasoconstriction Without The Risk Of Ischemia: Comparison With V1a-R Full Agonists

Edward Cable, Geoffrey Harris, Kazimierz Wisniewski, Hiroe Tariga, Susan X. Song, Jeannine Toy, Halina Wisniewska, Mark Lu, Steven Hoffmaster

Ferring Research Institute, San Diego, CA

Purpose: Vasopressin (AVP) and terlipressin have been used to treat cardiovascular complications in portal hypertensive patients. Vasopressin analogs are thought to act by causing specific vasoconstriction of the splanchnic circulation (and skin) via V1a receptor activation. Reduction in splanchnic blood flow reduces portal pressure. Although terlipressin and AVP have been used clinically, both agents have a narrow therapeutic index that limits their use to hospitalized patients. The aim was to design V1a partial agonists with capped vasoconstrictive activity that provide clinically meaningful vasoconstriction over a broad dose range without the risk of causing tissue hypoxia. Methods: a) In vitro, recombinant cell lines expressing human and rat V1A-R and V2R; b) Ex vivo, human and rat mesenteric resistance arteries; and c) In vivo, skin blood flow (SBF) and lactate in normal rats and portal pressure measurements in portal hypertensive rats. Results: A series of V1a-R partial agonist compounds were identified. In in vitro assays at the V1a-R, compounds demonstrated a significantly reduced Emax over a broad concentration range compared to AVP and other full agonists. Partial agonists were tested in vivo in a rat skin blood flow model (SBF) to assess if the reduction in in vitro Emax translated into reduced maximal vasoconstriction. Compounds that maximally decreased SBF < 60% (compared to V1a-R full agonists) demonstrated no evidence of tissue hypoxia, while those >60% increased blood lactate. Partial agonists also decreased the diameter of human resistance arteries, with an Emax significantly lower than full agonists. When tested in portal hypertensive rats, a significant decrease in portal pressure was observed that hit a plateau that could be maintained over a broad dose range. Conclusions: A novel series of V1a-R partial agonists have been discovered that produce moderate vasoconstriction without ischemia over a broad dose range. In patients with portal hypertension, such agents could be used to decrease portal pressure and treat or prevent many cardiovascular complications, such as HRS-1 &2, ascites, refractory ascites and post paracentesis induced circulatory dysfunction and the intrinsic safety of such agents could expand the use of vasconstriction therapy in these patients potentially outside of the hospital.


Edward Cable - Employment: Ferring Research Institute

Geoffrey Harris - Employment: Ferring Pharmaceuticals

Kazimierz Wisniewski - Employment: Ferring Research Institute

Jeannine Toy - Employment: Ferring Research Institute

Halina Wisniewska - Employment: Ferring Research Institute

The following people have nothing to disclose: Hiroe Tariga, Susan X. Song, Mark Lu, Steven Hoffmaster


Association of Gut Microbiota with Organ Failure, ACLF and Mortality in Cirrhotic Patients admitted with Infections

Jasmohan S. Bajaj1, Phillip Hylemon2, Douglas M. Heuman1, Arun J. Sanyal1, Jason M. Ridlon2, Patrick M. Gillevet3

1Gastroenterology, Hepatology and Nutrition, Virginia Commonwealth University and McGuire VA Medical Center, Richmond, VA; 2Mic; obioIogy, Virginia Commonwealth University and McGuire VA Medical Center, Richmond, VA; 3Microbiome Analysis Center, George Mason University, Manassas, VA

Background: Infections, especially those following bacterial translocation, are the leading cause of death and acute-onchronic liver failure (ACLF) in cirrhosis. Progressive changes in gut microbiome accompanying cirrhosis may predispose to infections, organ failure and death Aim: to evaluate differences in gut microbiota between cirrhotics with and without an infection matched on MELD scores and if changes in microbiota can predict 30 day outcomes of organ failure, ACLF and death. Methods: Cirrhotics hospitalized with infections were age/MELD-matched with cirrhotics hospitalized without infections in a case-control study. Stool was collected on admission and was analyzed using multi-tagged pyroseguencing. Patients were followed for 30 days for development of organ failure (shock, ventilation, dialysis, grade 3/4 HE), ACLF (≥2 organ failure) or mortality. Gut microbiota composition was compared between groups and those who did/did not develop organ failures, ACLF or death. Results: 44 infected cirrhotics (age 56yrs, 45%HCV, MELD 18, 66% on PPI, 45% on rifaximin and 8% on SBP prophylaxis) were matched with 44 non-infected patients (age 57 yrs, 50%HCV, MELD 18, 63% on PPI, 37% rifaximin, 15% SBP prophylaxis; p>0.05 on all comparisons). Infections were UTI (32%), SBP (32%), skin (16%), respiratory (13%) and spontaneous bacteremia (7%). There was 43% organ failure, 25% ACLF and 26% mortality in infected patients and none in 30 days in the uninfected group. Microbiota differences: In infected patients, median abundance of beneficial autochthonous bacteria was significantly lower (Lachnospiraceae 14 vs 3%, Clostridiales Incertae Sedis XIV 2 vs 0%, Ruminococcaceae 5 vs 1%, Veillonellaceae 2 vs 0%) and potentially pathogenic bacteria were higher (Enterobacteriaceae 0 vs 2%, Coriobacteriaceae 0.2 vs 0.5%) (all p<0.001). Beneficial bacteria (Lachnospiraceaee 5 vs 3%, Veillonellaceae 5 vs 2%, p<0.01) were significantly lower in those who developed organ failure. Similarly beneficial bacterial taxa were lower in those with ACLF (Clostridiales Incertae Sedis XIV 2 vs 0% and Leuconostocaceae 2 vs 0%, p<0.01). There was a higher abundance of Propionibacteriaceae (0 vs 2%) and Halomonadaceae (0 vs 2%) in those who died within 30 days compared to survivors. Conclusions: Despite controlling for MELD score, PPI, rifaximin and SBP prophylaxis, the gut microbiota in infected patients had significantly lower beneficial autochthonous bacterial taxa and higher potentially pathogenic ones compared to those without infections. Gut microbiota alterations at admission are associated with subseguent development of organ failures, ACLF and death within 30 days.


Jasmohan S. Bajaj - Advisory Committees or Review Panels: Salix, Merz, otsuka, ocera, grifols, american college of gastroenterology; Grant/Research Support: salix, otsuka, grifols

Douglas M. Heuman - Consulting: Bayer, Grifols, Genzyme; Grant/Research Support: Exilixis, Novartis, Bayer, Bristol Myers Sguibb, Scynexis, Ocera, Mannkind, Salix, Globeimmune, Roche, SciClone, Wyeth, Otsuka, Ikaria, UCB, Celgene, Centocor, Millenium, Osiris; Speaking and Teaching: Otsuka, Astellas

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

Patrick M. Gillevet - Management Position: BioSpherex LLC

The following people have nothing to disclose: Phillip Hylemon, Jason M. Ridlon


Role of angiogenic factors in serum of patients with hepatopulmonary syndrome

Sarah Raevens1, Stephanie Coulon1, Christophe Van Steenkiste2, Roos Colman3, Hans Van Vlierberghe1, Anja M. Geerts1, Thomas Perkmann4, Thomas Horvatits5, Valentin Fuhrmann5, Isabelle CoIIe6'1

1Deparment of Gastroenterology and Hepatology, Ghent University Hospital, Ghent, Belgium; 2Department of Gastroenterology and Hepatology, Maria Middelares Hospital, Ghent, Belgium; 3Department of Biostatistics, Ghent University, Ghent, Belgium; 4Department of Laboratory Medicine, Medical University Vienna, Vienna, Austria; 5Department of Gastroenterology and Hepatology, Medical University Vienna, Vienna, Austria; 6Department of Gastroenterology and Hepatology, Algemeen Stedelijk Ziekenhuis Aalst, Aalst, Belgium

Introduction-Hepatopulmonary syndrome (HPS) is a complication seen in 4 to 47% of patients with chronic liver disease and results in a significant increase in morbidity and mortality. HPS is caused by intrapulmonary vascular dilatations and direct arterio-venous connections (also known as shunts or collaterals) with a devastating influence on the physiology of the gas exchange. No effective medical therapies are available, only liver transplantation can cure HPS. The mechanisms responsible for these pathologic changes are still poorly understood. Many reports suggest an imbalance between vasodilators and vasoconstrictors. In recent literature evidence mounts for a process of (neo-) angiogenesis in the pathogenesis of HPS. Aim-The aim of this study was (1)to identify possible angiogenic factors in serum of patients with HPS and (2) to study the possibility to predict the presence or absence of HPS by these angiogenic factors. Methods-Multiplex assays (Meso Scale Discovery, Gaithersburg, USA) were used to measure the concentration of several angiogenic factors in the serum of patients with HPS (n=30) and without HPS (n=30). The following factors were measured: cKit, VEGFR2 (Growth Factor Panel II), bFGF, PIGF, sFIt1, Tie2, VEGF, VEGFC, VEGFD (Angiogenesis Panel I), eselectine, p-selectine, ICAM3, Throm, (Vascular Injury Panel I), CRP, SAA, ICAM1 and VCAM1 (Vascular Injury Panel II). The diagnosis of HPS was made on the basis of alveolo-arterial oxygen gradient above 20 and positive contrast echocardiography. Results-Patients with (n=30) and without HPS (n=30) had a similar MELD score (mean MELD score of 12.84 vs.12.52; p=0.852) and PaCO2 (34.33 vs.36.37; p=0.059). PaO2 values were significantly lower in the HPS group compared to the group without HPS (78.23 vs.85.67; p=0.009). Logistic regression was performed on all our results and we established a model with each predictor. Based on Area Under the Curve (AUC) data and p-values the best predictors were determined: VCAM1 (AUC=0.932; p<0.001)and ICAM3 (AUC=0.741; p=0.003). Combining these 2 factors results in an AUC of 0.993 and after cross-validation the AUC was still 0.988. Moreover, when adding age to the logistic regression model as an independent variable, VCAM1 and ICAM3 were still significant predictors for HPS. In the population with HPS, a moderate Spearman correlation between ICAM3 and ChildPugh Score was observed (r=0.48; p=0.007). Conclusion-Our results indicate that VCAM1 and lCAM3 might be promising biomarkers for predicting the presence of HPS. Combining these 2 factors result in an AUC of 0.993. However, the use of these biomarkers should be validated in a larger group of patients.


Isabelle Colle - Advisory Committees or Review Panels: MSD, Janssen, MSD, Gilead; Patent Held/Filed: Trombogenics; Speaking and Teaching: BMS, Janssen

The following people have nothing to disclose: Sarah Raevens, Stephanie Coulon, Christophe Van Steenkiste, Roos Colman, Hans Van Vlierberghe, Anja M. Geerts, Thomas Perkmann, Thomas Horvatits, Valentin Fuhrmann


Inhibition of neuronal apoptosis and axonal regression ameliorates sympathetic atrophy and hemodynamic alterations in portal hypertensive rats

Nahia Ezkurdia1, Imma Raurell1, Sarai Rodriguez1, Antonio Gonzalez1, Rafael Esfebon1,2, Joan Genescá1,2, Marίa Martell1

1Liver Unit-Internal Medicine, Institut de Recerco Vall d'Hebron, Hospital Universitario Vall d'Hebron, Barcelona, Spain; 2CIBERehd, Instituto de Salud Carlos Ill, Madrid, Spain

Background: In portal hypertension, pressure changes detected in the portal area trigger an afferent vagal signal to brain cardiovascular nuclei. From there, an efferent response via cholinergic sympathetic pre-ganglionic neurons reaches the superior mesenteric ganglia (SMG) inducing sympathetic atrophy in post-ganglionic neurons and arterial splanchnic vasodilation. Aim: 1) To investigate alterations in neuromodulators and signailing pathways leading to axonal regression or apoptosis; 2) to test the effects of the stimulation of neuronal proliferation and survival pathways by using a tyrosine kinase receptor A (TrkA) agonist, gambogic amide, on experimental portal hypertension Methods: 1)SMG samples from portal vein ligated (PVL, n=7) and sham (n=7) rats, for the analysis of neuromodulators and their signalling pathways by Western blot and immunofluorescence; 2) Hemodynamic determinations of daily treated PVL with gambogic amide (PVLGA, n=8) or vehicle (PVLV, n=8), and sham (n=8) rats. Sympathetic atrophy in superior mesenteric arteries was analyzed by tyrosine hydroxylase immunohistochemistry and neuromodulator expression in SMG by Western blot. Results: An increased expression in nerve growth factor (NGF) (p=0.018), its precursor proNGF (p=0.024), semaphorin3A (Sema3A) (p<0.001) and its receptor neuropilin (p=0.003) was found in PVL respect to sham rats. The active form of TrkA was decreased in PVL (p=0.03). In contrast, the increased expression in receptor p75NTR(p=0.007) and in p75NTR/TrkA ratio in PVL (p=0.029) pointed to the activation of p75NTR/ProNGF/Sema3A pathway leading to apoptosis and axonal retraction. Moreover, a significant overexpression in cleaved caspase7 (p>0.001), inactive poly(ADP-ribose) polymerase (PARP) and Rho Kinase (ROCK) expression supported an apoptotic process in PVL rats. Gambogic amide administration in PVL resulted in amelioration in hemodynamic alteration (significant increase in mean arterial pressure and superior mesenteric artery resistance, and decrease in blood flow) and in sympathetic atrophy comparing to PVLV, through the activation of pTrkA (p=0.012) and its signalling MAPK/AKT pathways (p<0.05). Moreover, reduction in ROCK (p>0.001) and PARP (p=0.04) expression in PVLGA rats indicated inhibition of the apoptotic cascade and axonal retraction in this group. Conclusion: The adrenergic alteration and sympathetic atrophy in mesenteric vessels during portal hypertension is caused by neuromodulation alterations leading to post-ganglionic sympathetic regression and apoptosis, and contributing to splanchnic vasodilation. Inhibition of these pathways ameliorates sympathetic atrophy and hemodynamic alterations.


Rafael Esteban - Speaking and Teaching: MSD, BMS, Novartis, Gilead, Glaxo, MSD, BMS, Novartis, Gilead, Glaxo, Janssen

The following people have nothing to disclose: Nahia Ezkurdia, Imma Raurell, Sarai Rodriguez, Antonio Gonzalez, Joan Genesca, Maria Martell


Rapid absorption of dietary long-chain fatty acids with changes to intestinal fatty acid transporters in liver cirrhosis

Yasunori Yamamoto, Yoshio Ikeda, Eiji Tsubouchi, Hiroki Utsunomiya, Miyake Teruki, Hiroaki Nunoi, Morikazu Onji, Bunzo Matsuura, Yoichi Hiasa

Department of Gastroenterology and Metabology, Ehime University Graduate School of Medicine, Toon, Japan

Background & Aims: Dietary long-chain fatty acids (LCFA) represent an important risk factor for the development of hepatosteatosis, inflammation and hepatocarcinogenesis. Changes in dietary LCFA intestinal absorption have been suggested to occur in patients with liver cirrhosis (LC). We aimed to identify any such changes and to clarify the mechanisms involved. Patients and Methods: Participants comprised 20 healthy subjects (Controls) and 53 patients with LC (28 with portal hypertension (PH(+)LC), 25 without portal hypertension (PH(-)LC)).13C palmitate (a LCFA) and octanoate (a mediumchain fatty acid (MCFA)) were administered directly to the duodenum using gastrointestinal endoscopy. Breath 13CO2 levels were then measured to guantify metabolized fatty acids. We also evaluated the expression of LCFA transporters in intestinal specimens from these groups. Results: A rapid increase in metabolized 13C palmitate over the first 120 min after administration was observed in patients with LC, greater than that seen in Controls (P < 0.01). This pattern of early LCFA metabolism in LC patients resembled that seen with the metabolism of 13C octanoate, a MCFA absorbed via the portal vein. We performed an additional assay of 10 patients with PH(+)LC who were treated with balloon-occluded retrograde transvenous obliteration (BRTO) for portal-systemic shunt. In the early phase (0-120 min) after loading, 13CO2 excretion increased in all patients after BRTO treatment (P<0.001). Portal blood flow thus affected the dynamics of palmitate absorption in the early phase. In contrast, serum concentrations of apolipoprotein B-48 in these patients did not change between before and after BRTO treatment, indicating that the transport of LCFAs to lymph vessels via chylomicrons remained unaltered. Moreover, jejunal levels of fatty acid transporter protein 4 and microsomal triglyceride transfer protein, which are associated molecules in chylomicron formation, were decreased in PH(+)LC patients compared with the Controls. On the other hand, we also observed increased expression of CD36, a key molecule in the facilitative uptake of LCFAs, in jejunal blood vessels in the PH(+)LC group. These findings suggest that LCFA absorption in patients with PH(+)LC would be altered, resulting in LCFA processing via capillary blood vessels similar to MCFA. Conclusion: Rapid absorption of dietary LCFA occurs in LC patients. In patients with LC, a proportion of dietary LCFAs are absorbed via the portal vessels in a similar manner to MCFAs. Such altered LCFA processing is likely related to the promotion of intestinal absorption of LCFAs by CD36 that is upregulated in the blood vessels of the jejunum.


The following people have nothing to disclose: Yasunori Yamamoto, Yoshio Ikeda, Eiji Tsubouchi, Hiroki Utsunomiya, Miyake Teruki, Hiroaki Nunoi, Morikazu Onji, Bunzo Matsuura, Yoichi Hiasa


Mechanisms involved in the development of thrombocytopenia in rats with cirrhosis and ascites

Jose Ignacio Fortea, Cristina Ripoll, Carolina Fernandez Mena, Marta Puerto, Juan Antonio Rodriguez-Feo, Javier Vaguero, Rafael Bañares

Medicina-Ap. Digestivo, Hospital General Universitario Gregorio Marañon-IiSGM-CIBERehd, Madrid, Spain

INTRODUCTION: Thrombocytopenia is a freguent finding in patients with cirrhosis. Although it traditionally attributed to an increased destruction of platelets by the spleen due to portal hypertension, the contribution of other factors such as an altered production of thrombopoietin (TPO) in the liver are being increasingly explored. Interestingly, we have recently observed that some patients with cirrhosis present higher levels of TPO in peripheral blood than in suprahepatic veins, suggesting an extrahepatic production site of TPO. Our aims were to investigate the mechanisms of thrombocytopenia and to identify potential sites of TPO production in an experimental model of carbon tetrachloride (CCl4)-induced cirrhosis in rats. METHODS: Male Sprague-Dawley rats with liver cirrhosis and ascites induced by the administration of CCl4 by oral gavage twice per week were compared with control rats receiving water (n= 5-6 per group). Two weeks after the last dose of CCl4, we performed a hemodynamic study to measure portal and mean arterial pressure. Arterial blood and tissues samples were taken from liver, spleen, kidney, lung, bone marrow and proximal intestine. Blood cell count was performed by an automated analyzer, and confirmed in peripheral blood smears. Megakaryocytopoiesis was assessed through bone marrow histological examination. TPO mRNA expression was investigated in tissue samples by real time-RT PCR with Tagman probes. RESULTS: Cirrhotic rats showed significantly lower platelets (840 +/- 66.82 vs 444 +/-180.9 10x3/uL, p=0.004), higher total WBC count (1.2 +/- 0.39 vs 10.8 +/- 7.75 10x3/uL, p<0.01), and an increased portal pressure (6.02 +/-1.41 vs 15.34 +/-1.46 mm Hg, p=0.004). Cirrhotic rats also showed an increased spleen-to-body weight (BW) ratio (0.17 +/- 0.01 vs 0.43 +/- 0.05) and a reduced liver-to-BW ratio (3.45 +/0.33 vs 2.27 +/- 0.40, p 0.008), suggesting the presence of hypersplenism and liver atrophy. Only the WBC count (R2=0.765, p=0.022) and the liver-to-BW ratio (R2=0.96, p=0.0005), but not the spleen-to-BW ratio, correlated with the platelet count among cirrhotic rats. The gene expression of TPO in the spleen, kidney, lung and intestine of cirrhotic rats was similar or lower to that found in control rats, and we did not observed any compensatory increase of megakaryocytes in bone marrow as assessed with H&E staining. CONCLUSIONS: Our results suggest that the presence of infection/inflammatory response and the decrease of functional liver mass contribute to the development of thrombocytopenia in cirrhosis. We were unable to identify an induction of the gene expression of TPO at any of the extrahepatic organs studied.


The following people have nothing to disclose: Jose Ignacio Fortea, Cristina Ripoll, Carolina Fernandez Mena, Marta Puerto, Juan Antonio Rodriguez-Feo, Javier Vaguero, Rafael Bañares.


Can Spleen Stiffness reflect Changes in Portal Pressure after Liver Transplantation?

Jun Liong Chin, Grace Chan, John D. Ryan, Aiden McCormick

Liver Unit, St. Vincent's University Hospital, University College Dublin, Dublin, Ireland

Introduction Spleen stiffness measured by transient elastography has been shown to correlate with hepatic venous pressure gradient and can predict esophageal varices. Elevated spleen stiffness in cirrhosis has been attributed to splenic tissue hyperplasia and fibrosis, portal hypertension and its conseguent hyperdynamic circulation. Aims The aim of this study was to evaluate changes in spleen stiffness after orthotopic liver transplantation (OLT) when portal hypertension resolves. Methods Twenty patients on the OLT waiting list were recruited prospectively. Spleen and liver stiffness were measured with Fibroscan before and at 2 to 4 weeks after OLT. Criteria applied for spleen stiffness measurement were similar to liver stiffness (≥10 measurements; ≥60% success rate; interguartile range, IQR<30% of median). For patients with ascites, spleen stiffness was measured in the right lateral decubitus position. Results 90% (18/20) of patients have esophageal varices on endoscopy or both esophageal and/or splenic varices on CT imaging. Spleen stiffness decreased significantly after OLT, from a median of 75.0(IQR 63.9-75.0) kPa (n=18) before transplant to 33.8(27.9-42.4) kPa (n=17) at weeks 2 to 4 after OLT (p<0.0001). As expected, liver stiffness measurements were higher before OLT at 44.3(27.4-75.0) kPa (n=7) and falls to 8.1(6.2-9.9) kPa (n=18) after transplant. Conclusion Spleen stiffness can measure changes in portal pressure after liver transplantation and decreases significantly when portal hypertension resolves.

Changes to spleen and liver stiffness measured by transient elastography before and after orthotopic liver transplantation (OLT)



The following people have nothing to disclose: Jun Liong Chin, Grace Chan, John D. Ryan, Aiden McCormick


Commensal Gut Flora Drives The Expansion Of Proinflammatory T Cells In The Small Intestinal Mucosa In Rats With CCl 4 Cirrhosis

Leticia Muñoz1 , 2, Marίa-José Borrero1, Maria Ubeda1 , 2, Margaret Lario1, David Daz1, Elia Aguado-Fraile3, Elisa Conde3, Laura Gorcίo-Bermejo3, Lourdes Lledó4, Melchor Alvorez-Mon5'1, Agustin Albillos6'1

1Medicine and Medical Specialties, University of Alcalá, Alcalá de Henares, Spain; 2CIBERehd, ISCIII, Madrid, Spain; 3Pathology, University Hospital Ramón y Cajal, IRYCIS, Madrid, Spain; 4Microbiology and Parasitology, University of Alcolá, Alcalá de Henares, Spain; 5lmmune System Diseases and Oncology, University Hospital Principe de Asturias, Alcalá de Henares, Spain; 6Gastoenterology, University Hospital, Ramón y Cajal, CIBERehd, IRYCIS, Madrid, Spain

Increases in enteric bacterial load and intestinal permeability are among the factors promoting gut bacterial translocation to mesenteric lymph nodes in cirrhosis. The intestinal immune system has a dual ability to avoid the invasion by resident bacteria or pathogens, and preserve the symbiotic relationship between host and microbiota. Deregulated, local immune responses driven by commensal bacteria are implicated in the increased intestinal permeability in chronic inflammatory bowel conditions. We aimed to examine in rats with non-ascitic (n=15) and decompensated (ascitic) CCl4-cirrhosis (n=19): i) the distribution, activation status and function of the T lymphocytes in the lamina propria of the small intestine (T-LPLs), and ii) the relationship of the abnormalities with the enteric bacterial load. Methods: T-LPLs from rats with CCl4-cirrhosis and controls (n=15) were analysed by four-colour flow cytometry. Terminal ileum samples were obtained for immunohistochemical staining. Total intestinal aerobic count in the ileal stool (log10 CFU/g of stool) was determined. Results: The small intestinal mucosa of ascitic cirrhotic rats showed intense atrophy, edema and wide villi, whereas that of non-ascites cirrhotic rats was close to normal. Compared to controls, T-LPLs lymphocytes (Th, Tc, gamma delta T and Treg) were similarly increased in ascitic and non-ascitic cirrhotic rats. Ascitic cirrhotic rats showed greater (p<0.05) freguencies of IFN gamma-secreting Th(31 ±6 vs 24±8%), Tc- (62±11 vs 46±11%), and IL17-secreting Th-LPLs (48±13 vs 29±18%) than non-ascitic cirrhotic rats. The freguencies of the latter parameters were similar in non-ascitic cirrhotic rats and controls (25±11,41 ±16, 39±14%, respectively). Compared to controls, the bacterial enteric load was greater (p<0.01) in ascitic than in non-ascitic cirrhotic rats and controls (6.7±0.8, 6.0±0.2, 5.5±1.1 CFU/g stool, respectively). Bowel decontamination with antibiotics reduced the enteric bacterial load (from 6.8 ± 0.2 to 3.6±1.3 CFU/g stool) and intestinal inflammation in ascitic cirrhotic rats. Conclusion: In cirrhosis, the mucosa of the small intestine is heavily infiltrated by T lymphocytes following a Th1 and Th17 pattern. Intestinal inflammation runs in parallel with the progression of cirrhosis, and is driven by commensal gut flora.


The following people have nothing to disclose: Leticia Munoz, Maria-José Borrero, Maria Ubeda, Margaret Lario, David Diaz, Elia Aguado-Fraile, Elisa Conde, Laura Garcia-Bermejo, Lourdes Lledó, Melchor Alvarez-Mon, Agustin Albillos


Combination of fibrosis markers (ELF) and macrophage activation markers (sCD163) accurately predicts portal hypertension in patients with liver cirrhosis

Thomas D. Sandahl1, Rie Mcgrail2, Holger J. Møller2, Søren Møller3, Flemming Bendtsen4, Hendrik V. Vilstrup1, Henning Grønbæk

1Medicine V (Hepatology and Gastroenterology), Aarhus University Hospital, Aarhus, Denmark; 2Department of Clinical Biochemistry Aarhus University Hospital, Aarhus, Denmark; 3Departments of Gastroenterology and Clinical Physiology & Nuclear Medicine,, Hvidovre University Hospital, Hvidovre, Denmark; 4Department of Gastroenterology, Hvidovre University Hospital, Hvidovre, Denmark

Abstract: Background: The increased intrahepatic resistance in cirrhotic portal hypertension supposedly is due to a dynamic inflammatory component and a structural fibrotic component. We have shown that blood soluble (s)CD163, a marker of inflammatory Kupffer cell activation, can predict portal hypertension. We explored whether the prediction of the portal hypertension can be improved by including fibrosis markers. Aim: To study blood sCD163 in combination with serologic markers of liver fibrosis as biomarkers for portal hypertension in patients with cirrhosis. Methods: We conducted liver vein catheterisations with measurement of the hepatic venous pressure gradient (HVPG) and at the same time measured blood concentrations of sCD163 and hyaluronic acid, tissue inhibitor of metalloproteinase 1, and procollagen III aminopeptide (the socalled Extended Liver Fibrosis package) in 79 cirrhosis patients (Child-Pugh CP-A: n=26, CP-B: n=28, CP-C: n=26) and 21 healthy persons. The liver status was described by the ChildPugh and MELD-scores. Results: sCD163 and all fibrosis markers increased in a stepwise manner with the CP-score (p<0.001, r>0.56, all; p<0.003) Receiver Operator Characteristics analysis showed that all of the individual analyses were able to predict portal hypertension (HVPG ≥ 10 mmHg) with AUROC's of > 0.80. Combining the inflammatory and fibrosis markers into one score, optimized by multiple linear regression analysis, improved the ROC analysis to an AUROC of 0.93 with a sensitivity of 93% and specificity 75%. Conclusion: Combining the two categories of markers accurately predicted significant portal hypertension in patients with liver cirrhosis. This may be clinically useful and supports the concept of inflammatory and fibrotic components of hepatic blood flow resistance.


Holger J. Moller - Grant/Research Support: Danish Council for Strategic Research; Patent Held/Filed: Aarhus University

Henning Grønbæk - Advisory Committees or Review Panels: Novartis; Grant/Research Support: NOVO Nordisk; Speaking and Teaching: Eli Lilly, Ipsen

The following people have nothing to disclose: Thomas D. Sandahl, Rie Mcgrail, Søren Møller, Flemming Bendtsen, Hendrik V. Vilstrup


Virtual Portal Vein Pressure from Anatomic CT Angiography Results from the Prospective Multicenter VIRGIN (VIRtual Portal Vein Pressure AGainst INvasive Evaluation) Study

Xiaolong Qi1, Zhiwei Li2, Fangyu Zhou1, Chaohui Xiao2, Lingxiang Yu2, Tongling Ding2, Jiale Huang1, Yanjie Zhu1, Guoxin Fan1, Surong Xing3, Weiguo Xu3, Wenbin Ji3, Huisong Chen1, Shenglan Wang1, Gongwen Li4, Changging Yang1

1Division of Gastroen terology and Hepatology, Tongji Hospital, Tongji University School of Medicine, Shanghai, China; 2Division of Hepatobiliary Surgery, 302 Hospital of Chinese People's Liberation Army, Beijing, China; 3Division of Radiology, Tongji Hospital, Tongji University School of Medicine, Shanghai, China; 4Tongji University School of Civil Engineering, Shanghai, China

Objective: The aim of this study is to test the feasibility and diagnostic accuracy of virtual portal vein pressure (vPVP) for predicting invasive portal vein pressure (PVP). Methods: The VIRGIN (VlRtual Portal Vein Pressure AGainst INvasive Evaluation) study was a two site, prospective, analytical study. The protocol was approved by the local ethics committee. All participants gave written informed consent. The diagnostic performance of vPVP involved in 21 participants who underwent Doppler ultrasound, CT angiography and invasive PVP measurement, all of which were interpreted in blinded fashion. The diagnostic accuracy of vPVP, portal vein velocity and diameter for predicting PVP were analyzed. Results: A representative example of vPVP interpretation is presented in Figure 1A. On average, the vPVP values deviated from PVP by ± 2.02 mmHg (mean difference= +0.617, SD= 2.410) shown in Figure 1B. The vPVP and PVP were closely correlated (Pearson's correlation coefficient= 0.818, p< 0.001; Spearman's rank correlation= 0.802, p< 0.001) in Figure 1C. In addition, the diagnostic value of portal vein velocity was modest (Spearman's rank correlation= 0.576, p< 0.01; Pearson's correlation coefficient= 0.460, p< 0.05) in Figure 1D. The diameter was poor correlated with PVP with no statistical significance. Conclusions: We have developed a model predicting PVP based on anatomic CT angiography. Although the accuracy of vPVP is limited with the small sample size of patients, the novel assessment offers a potential prediction of PVP without the need for invasive measurements.

Figure 1 Diagnostic accuracy of vPVP and Vpv

A) Three dimensional models of hepatic vein (HV) and portal vein (PV) (left); Mathematical models meshed with internal tetrahedral (middle); PVP and vPVP values (right). B) Bland-Altman Plot: demonstrate the difference between vPVP and PVP plotted against the mean value (black line). Two red lines indicate 2 SD above and below mean delta; C) A good correlation of vPVP and PVP; D) A modest diagnostic value of Vpv. Vpv= portal vein velocity.



The following people have nothing to disclose: Xiaolong Qi, Zhiwei Li, Fangyu Zhou, Chaohui Xiao, Lingxiang Yu, Tongling Ding, Jiale Huang, Yanjie Zhu, Guoxin Fan, Surong Xing, Weiguo Xu, Wenbin Ji, Huisong Chen, Shenglan Wang, Gongwen Li, Changging Yang


Regulatory T cells (Tregs) expansion and reversal of CD4 and CD8 positive lymphocytes distribution pattern within the portal vein - A novel inflammatory aspect of portal hypertension

Tomer Adar1, Yair Edden2, Shimon Shteingart1,Mahmud Mahamid1, Ellen Broide3, Ariella B. Shitrit1, Eron Goldin1

1Department of Gastroenterology, Digestive disease institute, Shaare Zedek Medical Center, Jerusalem, Israel; 2Department of General Surgery, Digestive Diseases Institute, Shaare Zedek Medical Center, Jerusalem, Israel; 3Flow cytometry unit, Department of immunology, Shaare Zedek Medical Center, Jerusalem, Israel

INTRODUCTION The portal vein drains both the spleen and intestines. Both organs are rich in inflammatory cells. Portal hypertension (p-HTN) may induce stress within these organs, and result in pro-inflammatory changes within the portal vein. This pro-inflammatory change within the portal vein may in turn aggravate hepatic inflammation and fibrosis, thus creating a vicious cycle. Furthermore, these inflammatory changes may be more pronounced within the portal vein. AIM To demonstrate the inflammatory aspects of portal hypertension within the portal and systemic circulation. . METHODS Adult Sprague -Dawley rats were divided into two groups. Portal hypertension (p-HTN) was induced in group A by partial portal vein ligation. Control rats in group B underwent a sham operation (laparotomy and portal vein exposure without ligation). Rats were sacrificed a week later with documentation of body and spleen weight and evidence of p-HTN. Spleens were biopsied and underwent H&E staining and evaluation by a pathologist (blinded to group allocation) for vascular congestion. During sacrifice, blood samples were drawn from both the portal vein (PV) and the inferior vena cava (IVC) to characterize lymphocytes subpopulations by flow cytometry. RESULTS Nine and 6 rats were included in groups A and B respectively, with a total of 14 and 13 samples from PV and IVC respectively. Portal HTN developed in group A as documented by development of macroscopic vascular collaterals, increased spleen/body weight compared to group B (p<0.05) and vascular congestion and stasis in histopathological examination of the spleen. Rats with p-HTN had significantly increased percentage of CD4+CD25+FOXP3+ regulatory T cells (Tregs) compared to control (7.86 vs.5.4, p<0.05). Additionally, a distribution pattern reversal was noted for both CD4 and CD8 positive lymphocytes. The PV / IVC ratio for CD4 was 1.03 and 0.88 in p-HTN and control rats (p<0.05) and for CD8 the PV/IVC ratio was 1.04 and 0.88 p-HTN and control rats (both p<0.05). DISCUSSION Portal hypertension is traditionally perceived as a hemodynamic phenomenon, whose pathogenesis is based on pressure gradients and collateral formation. This study demonstrates portal hypertension is also an inflammatory phenomenon, with a reversal of CD4 and CD8 distribution patterns (with a ration turning greater than 1), and an increased fraction of Tregs (despite the CD4 distribution change). This novel approach to p-HTN may improve our understanding of its pathogenesis, and may direct to future therapeutic targets.


The following people have nothing to disclose: Tomer Adar, Yair Edden, Shimon Shteingart, Mahmud Mahamid, Ellen Broide, Ariella B. Shitrit, Eran Goldin


Which noninvasive hepatic fibrosis test most effectively predicts portal pressure and survival in alcoholic liver disease patients?

Eun Ju Cho1, Jeong-Hoon Lee2, Moon Young Kim3, Jeong-Ju Yoo2, Won-Mook Choi2, Young Youn Cho2, Yuri Cho2, Dong Hyeon Lee2, Yun Bin Lee2, Su Jong Yu2, Yoon Jun Kim2, Hyo-Suk Lee2, JungHwan Yoon2, Soon Koo Baik3

1Department of Internal Medicine, Kangwon National University Hospital, Chuncheon, Republic of Korea; 2Department of Internal Medicine and Liver Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea; 3Department of Internal Medicine,Yonsei University Wonju College of Medicine, Wonju, Republic of Korea

Objectives: We investigated which noninvasive fibrosis test most effectively reflects hepatic venous pressure gradient (HVPG) and predicts prognosis in patients with alcoholic liver disease (ALD). Methods: A total of 195 consecutive patients with ALD were included. Biochemical indices and liver stiffness assessed by transient elastography (TE) were compared with HVPG. Results: The diagnostic values of liver stiffness in detecting clinically significant portal hypertension (CSPH; HVPG≥10 mmHg) was significantly higher (AUROC=0.87±0.03) than those of biochemical indices (i. e. APRI, FIB4, P2/MS and platelet count/spleen diameter ratio; all P<0.001). In multivariate analysis, liver stiffness was most significantly correlated to HVPG (P<0.001), whereas other biochemical indices were not. On the other hand, the prognostic values of liver stiffness for liver-related death (AUROC=0.73±0.07) did not differ from those of FIB4 (0.78±0.04), HVPG (0.70±0.07) and APRI (0.69±0.04). Significant risk factors for liver-related death were Child score (hazard ratio [HR]=2.35, P<0.001), varices >F2 (HR=5.85, P=0.002) and FIB4 (HR=1.11, P=0.03), but not liver stiffness and HVPG. For all-cause death, age and FIB4 were independent predictors in compensated patients (P=0.02 and <0.001, respectively), whereas Child score was in decompensated patients (P<0.001). Conclusions: In patients with ALD, liver stiffness most accurately predicts CSPH, but did not improve the prognostic values of traditional risk factors for mortality, whereas FIB4 was independent predictor for liver disease-related death and all-cause death.

Fig 1. ROC curves for the detection of (A) CSPH and (B) liverrelated death



The following people have nothing to disclose: Eun Ju Cho, Jeong-Hoon Lee, Moon Young Kim, Jeong-Ju Yoo, Won-Mook Choi, Young Youn Cho, Yuri Cho, Dong Hyeon Lee, Yun Bin Lee, Su Jong Yu, Yoon Jun Kim, Hyo-Suk Lee, Jung-Hwan Yoon, Soon Koo Baik


Hemodynamic effects of the Mas receptor agonist AVE0991

Sabine Klein, Robert Schierwagen, Tilman Sauerbruch, Jonel Trebicka; Internal Medicine I, University of Bonn, Bonn, Germany

Background: The Renin-Angiotensin-system (RAS) is involved in pathogenesis of liver cirrhosis and portal hypertension. Interestingly, an alternative RAS pathway (Angiotensin-(1-7)/Masreceptor) counterbalances the vasocontrictive effect of the classical RAS arm (Angiotensin-II/AT1-receptor) mediating vasodilation. In this context we showed that the mas-receptor agonist Angiotensin-(1-7) lowered portal pressure, while its antagonist (A779) increased the portal pressure of cirrhotic rats. Here we investigated the hemodynamic effects of AVE099, a non-peptidic mas-receptor agonist, in cirrhotic BDL rats with portal hypertension. Methods: In wild-type rats we induced secondary biliary cirrhosis using bile-duct ligation (BDL). After 4-6 weeks, when rats developed ascites as a sign of portal hypertension, portal and mean arterial pressure were measured invasively, and the regional blood flows were investigated with the coloured microspheres technigue. The coloured microspheres were injected before and 30 minutes after the i. v. injection of 0.1mg/kg AVE0991. Results: Before AVE0991 application, BDL rats showed the characteristic hemodynamic pattern of an increased portal pressure and hepatic resistance, whereas the mean arterial pressure was reduced. Injection of AVE0991 reduced significantly intrahepatic resistance. At the same time AVE0991 decreased cardiac output. This led to a significant reduction of portal pressure of about 20%. These effects were observed in cirrhotic BDL, but not in sham-operated rats. Discussion: Modulation of the alternative RAS pathway by stimulationg the Mas-receptor might be a new tool for treatment of portal hypertension in liver cirrhosis.


The following people have nothing to disclose: Sabine Klein, Robert Schierwagen, Tilman Sauerbruch, Jonel Trebicka


Altered Portal Venous flow response to terlipressin in BDL rats, measured non-invasively using phase-contrast MRI

Manil Chouhan1,2, Alan Bainbridge4, Simon Walker-Samuel3, Mark F. Lythgoe3, Shonit Punwani2, Stuart Taylor2, Rajiv Jalan1, Nathan Davies1, Mookerjee P Rajeshwar1

1Liver Failure Group, UCL Institute for Liver and Digestive Health, University College London, London, United Kingdom; 2UCL Centre for Medical Imaging, University College London, London, United Kingdom; 3UCL Centre for Advanced Biomedical Imaging, University College London, London, United Kingdom; 4Department of Medical Physics, University College London Hospital NHS Trust, London, United Kingdom

Background and aims: Derrangements in portal venous (PV) flow are known to occur in models of liver disease with portal hypertension, but studies have been limited by the invasive nature of measurement methods. Phase-contrast (PC) MRI can be used to measure bulk vessel flow but has never been applied to study changes in PV flow in small rodent models. Terlipressin is used clinically to reduce splanchnic and PV flow and whilst in healthy subjects, the hepatic arterial buffer response maintains perfusion, in subjects with liver disease this response is impaired. In this study, we have used PCMRI to measure changes in portal flow after terlipressin administration and studied the responses in a rat model of liver disease. Methods: Eight Sprague-Dawley rats were randomised to bile duct ligation (BDL) procedure (n=4) or sham surgery (n=4). Studies were conducted after 4 weeks, based on previous measurements of portal hypertension in the BDL model. PV flow was measured using respiratory-gated 2D PCMRI (2 mm slice thickness, 10° flip angle and 22 cm/s velocity encoding) with a 9.4T Agilent scanner. Terlipressin was administered intravenously as a bolus dose of 10mcg/100g and PC-MRI measurements repeated seguentially for 30-40 minutes post-administration. Bulk PVF was normalised to explant liver weight. Data was analysed using paired and unpaired t-tests. Results: After 4 weeks, sham vs BDL body weight (530.5±37.5g vs 427.0±41.5g) and liver weight (19.4±1.38g vs 32.4±2.18g) was significantly different (p<0.05). Baseline PV flow in sham (143.53±14.42ml/min/100g)vsBDL (79.51 ±44.74ml/min/100g) rats approached significance (p=0.059). The reduction in PV flow post-terlipressin, was significant in sham animals (mean reduction of 63.48±14.28ml/min/100g; p<0.05), and close to statistical significance in BDL rats (mean reduction of 55.44±35.71ml/min/100g; p=0.053). Significant differences in post-terlipressin nadir PV flow in sham (80.05±20.18ml/min/100g) vs BDL (24.07±14.66mI/m in/100g) rats were demonstrated (p<0.05). Conclusions: This is the first non-invasive PCMRI PV flow study in a model of liver disease, and our findings highlight: (i) reduced baseline PV flow in a fibrotic model affirming a significant contribution of intra-hepatic resistance to the development of portal hypertension, (ii) BDL PV flow demonstrates a different, more labile haemodynamic response to terlipressin compared to sham-operated rats and (iii) the utility of PV flow guantification in the development of novel non-invasive biomarkers for portal hypertension.


Stuart Taylor - Grant/Research Support: Wellcome Trust Rajiv Jalan - Consulting: Ocera Therapeutics, Conatus Nathan Davies - Patent Held/Filed: UCL

The following people have nothing to disclose: Manil Chouhan, Alan Bainbridge, Simon Walker-Samuel, Mark F. Lythgoe, Shonit Punwani, Mookerjee P. Rajeshwar


Comparison of vWF-Antigen and ELF-Score in prediction of portal hypertension

Andreas Maieron1, Stephanie Hametner1, Alexandra Etschmaier2, Monika Ferlitsch2, Rainer SchÖfl1, Alexander Ziachehabi1, Thomas Hufnagl3, Doris Trubert-Exinger4, Arnulf Ferlitsch2

1Gastroenterology and Hepatology, Elisabeth Hospital, Linz, Austria; 2Gastroenterology and Hepatology, Medical University Vienna, Vienna, Austria; 3Siemens Diagnsotics, Siemens Diagnostics, Vienna, Austria; 4Laboratory Medicine, St PÖlten Hospital, Si PÖlten, Austria

Background: Clinically significant portal hypertension (CSPH), defined by hepatic venous pressure gradient (HVPG) ≥ 10 mm/Hg causes major complications. To improve survival of cirrhotic patients, early diagnosis is crucial. The diagnosis of CSPH by HVPG measurement is not comprehensively available. vWF-Ag has shown significant ability on the diagnosis of CSPH and is a predictor for mortality. ELF-Score, consisting of a panel of direct biomarkers such as hyaluronic acid (HA), procollagen Ill N-terminal propeptid (PIIINP) and den inhibitor of metalloproteinase 1(TIMP-1), can detect liver cirrhosis adeguately in most of the cases. Therefore we compared the diagnostic ability of vWF and ELF Score. Methods: Data of 189 patients underwent HVPG measurement mostly due to liver cirrhosis. The diagnosis of cirrhosis was established by either liver biopsy, transient elastography (TE) or typical radiological signs. All patients were categorized into 3 different HVPG groups (<10, ≥10-<20, ≥20 mmHg), vWF-Ag and ELF-Score were analysed. Blood samples were obtained during HVPG measurement. Patients' data were collected to evaluate Child Pugh Score (CPS), MELD score and D'Amico staging system. For comparison of diagnostic ability a ROC analysis was performed. Results: 189 patients (male: 140, median age 57.9 IQR: 49.9-66.3) underwent HVPG measurement. The aetiology of liver cirrhosis: ALD 39%, NASH 11.1%, HepC 28.6% and others 21.3%. CPS A: 61.4%, CPS B: 23.8%, CPS C: 5.8%, unknown: 9%. Median MELD in CPS A was 6.1, IQR: 3.6-8.4; in CPS B 10.9, IQR: 8.4-12.5 and CPS C 18, IQR: 15.4-28. According D'Amico classification 56.1% were compensated.30 (15.9 %) patients showed no CSPH, 159 (84.1%) showed CSPH. vWF and ELF-Score increase significantly throughout different HVPG categories (p<0.000 and p<0.03). ROC-analysis shows an AUROC of 0.82 (0.74-0.9) for vWF≥10mmHg with a sensitivity of 80.9% and a specifity of 76.7% at a vWF cut off of 215, and 0.677 (0.55-0.81) for ELF-Score with a sensitivity of 80.2 and a specifity of 44.4% at a cut off of 9.7. AUROC for HVPG≥20mmHg for vWF is 0.69 (0.61-0.78) with a sensitivity of 88.6% and a specifity of 31.1% at a cut off of 215. ELFScore shows an AUROC of 0.6 (0.48-0.75) with a sensitivity of 82.8% and a specifity of 25.6% at a cut off of 9.7. vWF and ELF-Score correlate significantly with CPS and D'Amico (p<0.000). Conclusion: vWF and ELF-Score both show significant correlation with different stages of liver disease and HVPG categories. Although ELF-Score consists of a panel of direct markers of fibrosis, vWF outplays ELF-Score in the diagnostic ability of detection of CSPH.


Andreas Maieron - Advisory Committees or Review Panels: MSD, Jannsen, BMS, BVdhringer Ingelheim, Gilead; Grant/Research Support: Roche; Speaking and Teaching: Roche, MSD, Jannsen, Gilead

Stephanie Hametner - Speaking and Teaching: MSD

Alexander Ziachehabi - Advisory Committees or Review Panels: MSD; Grant/Research Support: GILEAD; Speaking and Teaching: MSD

Thomas Hufnagl - Employment: Siemens Healthcare Diagnostics GmbH

The following people have nothing to disclose: Alexandra Etschmaier, Monika Ferlitsch, Rainer SchÖfl, Doris Trubert-Exinger, Arnulf Ferlitsch


Hepatic arterial vascular resistance (HAR) is decreased in CCl4-cirrhosis without and with portal hypertension a mechanism independent of changes in portal and sinusoidal resistances

Miriam Moller, Antje Thonig, Cristina Ripoll, Alexander Zipprich

Internal Medicine I, Martin-Luther-Universitiy Halle, Halle/Saale, Germany

The increase in hepatic arterial flow when there is a decrease in portal venous flow is known as the hepatic arterial buffer response. In cirrhosis with portal hypertension and elevated portal pressure there is a decrease in HAR, which is associated to high concentrations of nitric oxide and adenosine. Whether this lower HAR is also present in cirrhosis without portal hypertension is unknown. The aim of the study was to investigate the HAR and response to vasoconstriction in cirrhosis with and without portal hypertension. Methods: Cirrhosis was induced by CCl4. After 8 weeks rats develop cirrhosis without portal hypertension (8W) and after 12-14 weeks with portal hypertension (12W). Bivascular liver perfusion was performed after 8W and 12W and dose response curves of methoxamine were done in the presence or absence of LNMMA (nitric oxide blocker). HAR, portal vein resistance (PVR) and sinusoidial resistance (SVR) were measured. Western Blot and Immunostainings were done to measure HIF 1a expression, a measurement of hypoxia. Results: HAR in both groups of cirrhotic animals (8W and 12W) were lower compared to controls. In comparison to controls, PVR and SVR were higher in 12W, while no difference was seen in 8W. Dose response curves to methoxamine revealed lower HAR in cirrhosis (8W and 12W) compared to controls. LNMMA corrected the dose response curves in cirrhosis (8W and 12W) (figure 1). Western Blot shows increased protein expression of HlF1a in 8W and 12W and immunostaining revealed higher expression of HIF1a in cirrhosis. Conclusion: Hepatic arterial resistance is decreased in cirrhosis without and with portal hypertension independent of portal and sinusoidal resistances and therefore independent of the hepatic arterial buffer response. Vasodilatation of the hepatic artery in cirrhosis seems more related to hypoxic conditions (HIF 1 a) rather to portal or sinusoidal pressure.



The following people have nothing to disclose: Miriam Moller, Antje Thonig, Cristina Ripoll, Alexander Zipprich


Eight weeks of endurance exercise training improves aerobic capacity, muscle mass, and symptoms of fatigue in patients with Child-Pugh class A and B cirrhosis

Laura Zenith1, Mark Haykowsky2, Ailar Ramadi2, Milad Yavari2, Neha Meena2, Andrea Harvey1,Michelle Carbonneau1, Mang M. Ma1, Puneeta Tandon1

1Medicine, University of Alberta, Edmonton, AB, Canada; 2Rehabilitation Medicine, University of Alberta, Edmonton, AB, Canada

Background: Cirrhosis is associated with reduced muscle mass, guality of life and exercise tolerance, the latter measured objectively as decreased peak oxygen uptake (V〇2peak). V〇2peak and muscle mass depletion are independent and potentially modifiable predictors of mortality in these patients. The impact of endurance training (ET) has not been well studied in cirrhosis. Accordingly, we aimed to evaluate the safety and efficacy of 8-weeks of ET on V〇2peak, muscle thickness, fatigue and liver function in patients with cirrhosis. Methods: 15 clinically stable outpatients with Child Pugh class A or B cirrhosis and guideline based primary prophylaxis for high risk esophageal varices were randomly assigned to 8-weeks of ET (n=9) or usual care (UC, n=6). Cirrhosis was diagnosed by compatible imaging and clinical data. Exclusion criteria were hepatocellular carcinoma, non-hepatocellular carcinoma malignancy, significant cardiopulmonary disease, hemoglobin <110 g/L or physical impairment preventing cycle exercise training. Supervised ET (cycle exercise) was performed 3 days/week at 60 to 80% of heart rate reserve for 50 minutes/session. The UC was not given any exercise guidelines and continued with their normal activities of daily living. Cardiopulmonary exercise (V〇2peak), compression guadriceps muscle thickness (ultrasound held with firm pressure over muscle), chronic liver disease guestionnaire and MELD score were measured at baseline and after 8 weeks. Data are reported as change from baseline score for each group. Statistical analysis was performed using a one-way analysis of variance. Results: At baseline, the mean age was 56 ± 7 years with a mean Child Pugh score of 6.4 ± 1.5 (67% Child Pugh class A) and a mean MELD score of 10.0 ± 2.4. There was no significant difference in baseline characteristics (age, weight, protein intake, liver disease severity, V〇2peak or secondary outcome measures) between the ET or UC groups. No adverse events occurred during cardiopulmonary exercise testing or ET. The changes in V〇2peak (0.4 ±0.2 vs.0.0 ± 0.1, L/min, p<0.001), thigh circumference (1.2 ± 0.6 vs.0.0 ± 0.5 cm, p=0.001), compression guadriceps muscle thickness (0.15 ± 0.13 vs.0.013 ± 0.09 cm, p=0.047) was significantly greater after ET compared to UC. The change in fatigue score was significantly higher (i. e. less fatigue) after ET compared to UC (0.98 ± 1.1 vs. −0.3 ± 0.7, p=0.032). No significant between group differences were found in the change in MELD or change in ALT. Conclusions: Eight weeks of ET is a safe and effective therapy to improve V〇2peak, muscle mass and fatigue in patients with clinically stable Child Pugh class A or B cirrhosis.


The following people have nothing to disclose: Laura Zenith, Mark Haykowsky, Ailar Ramadi, Milad Yavari, Neha Meena, Andrea Harvey, Michelle Carbonneau, Mang M. Ma, Puneeta Tandon