AA

SL

DA

bst

ract

s

Sa1686

Diastolic Dysfunction Predicts Mortality After Transjugular IntrahepaticPortosystemic Shunt CreationGabriel Lang, Stacy Banerjee, Thuong VanHa, Helen S. Te, Nancy Reau, Donald M.Jensen

Introduction: Cirrhotic patients with diastolic dysfunction appear to be extremely sensitiveto the increased preload that occurs after TIPS placement and liver transplantation. Unfortu-nately, many of the specific clinical variables that predict the high rate of mortality followingTIPS remain undocumented. Our goal was to examine the relationship between patientswith abnormal left ventricular (LV) diastolic function and mortality following TIPS. Methods:A retrospective chart review of patients who had undergone TIPS placement between 2005and 2009 was performed. Of the 80 patients identified, 43 patients (39%) had a twodimensional echocardiogram performed within the 6 months prior to TIPS. These patientswere subsequently divided into 2 groups based on the presence or absence of diastolicdysfunction based on the American Society of Echocardiography's guidelines. Patients withseptal e' <8 (measured by tissue Doppler) and LA volumes >34 ml/m2 were classified ashaving abnormal LV diastolic function. Subsequently, through the social security death indexdata base, we identified which patients survived to the 3 month time point after TIPScreation. Analysis of potential predictors of response including presence of diastolic dysfunc-tion, etiology of cirrhosis, age, indication for TIPS, MELD score, and Child Pugh score wereperformed with Chi-squared analyses to determine their association with mortality at threemonths. Results: Of the 43 patients with echocardiographic data, the mean age was 56 yearsold, with 56% being male. 20 patients (47%) were identified as having diastolic dysfunction.Overall, 14 (33%) patients diedwithin 3months of TIPS. There were no intergroup differenceswith regard to cause of cirrhosis or TIPS indication. Of the patients with diastolic dysfunction,11 (55%) died within 3 months compared with only 3 (13%) patients without diastolicdysfunction. Based on our analysis, diastolic dysfunction assessed 6 months prior to TIPSprocedure is a significant predictor of mortality (p=.003). Conclusion: Our results indicatethat diastolic dysfunction diagnosed by echocardiographic criteria within 6 months of TIPScreation predicts mortality at 3 months. The categorization of diastolic dysfunction by LAvolumes and septal e' provides clinicians with important information regarding patientoutcomes prior to TIPS placement.

Sa1687

Dissociation of Phosphoinositide Hydrolysis and Cellular Contractility inHepatic Stellate Cell ActivationQing Yu, David Yoo, Don C. Rockey

During liver injury, stellate cells transform from a quiescent phenotype into “activated”myofibroblast-like cells. This activation program is characterized the acquisition of smoothmuscle-specific and cytoskeletal proteins and concomitant increases in cell motility andcontractility - and this feature appears to be important in the pathogenesis of the woundingresponse. The family of vasoactive endothelin (ET) peptides are prominent mediators ofstellate cell contractility. Here, we have explored the mechanism of ET receptor signalingin stellate cells during activation. Methods. Retired male breeder Sprague-Dawley rats weresubjected to bile duct ligation (BDL) or carbon tetrachloride (CCl4) gavage to induce liverinjury and wound healing. Primary stellate cells were isolated and cultured using standardmethods. ET receptor signaling was assessed by measuring phsophatidylinositol-specificphospholipase C (PI-PLC) activation (PI-PLC hydrolyzes phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2] to inositol 1,4,5-trisphosphate and diacylglycerol) using Dowexresin columns and high performance liquid chromatography (HPLC). Results. ETA and ETBreceptor mRNAs and binding sites were upregulated in stellate cells following both formsof liver injury. Stimulation of stellate cells with ET-1 led to an increase in phosphoinositidelevels from 10 DPM/μg protein in normal stellate cells to 39 DPM/μg protein in BDL or 29DPM/μg protein in CCl4 stellate cells (p<0.05, BDL or CCl4 vs. normal). While inhibitionof ETB receptor signaling in activated hepatic stellate cells by BQ-788 significantly decreasedphosphoinositide hydrolysis, inhibition of ETA receptor by BQ-123 had no effect. Inhibitionof PI-4 kinase by wortmannin significantly blunted the production of PIP and PIP2 as wellas generation of IP2 and IP3 after activation with ET-1, but had no effect on stellate cellcontractility. Stimulation of either ET receptor in activated stellate cells resulted in myosinlight chain phosphorylation and stellate cell contraction, while inhibitors of intracellular orextracellular sources of calcium had no effects on ET-1 mediated stellate cell contraction.Also, inhibition of myosin light chain kinase (MLCK) with ML7 (1μM) had no effect onET-1 mediated contraction. Conclusions. Despite enhanced ET receptor expression followingliver injury and stellate cell activation, the corresponding increase in ET-mediated PI-PLCsignaling in activated stellate cells plays little or no role in cell contractility. Different ET

S-918AASLD Abstracts

receptors may play different role in PI-PLC signaling pathways, but the data point to Ca2+-independent signaling pathways in ET-induced myofibroblast contraction.

Sa1820

Outcomes in Patients Receiving High Dose Ursodeoxycholic Acid Are NotDependent on Disease StatusMohamad Imam, Andrea A. Gossard, Kris V. Kowdley, Velimir A. Luketic, M. EdwynHarrison, Timothy M. McCashland, Alex Befeler, Denise M. Harnois, Roberta A.Jorgensen, Janice L. Petz, Jill C. Keach, Alisha C. DeCook, Felicity Enders, Keith D.Lindor

Background: A study evaluating the effectiveness of high dose ursodeoxycholic acid (UDCA)in the treatment of primary sclerosing cholangitis (PSC) surprisingly showed that UDCAincreases the likelihood of clinical deterioration. The question was raised whether this effectwas confined to patients with advanced disease. Our aim was to compare the probabilityof developing adverse clinical endpoints in patients with varying disease status and todetermine if disease status influenced the clinical response of those patients to UDCA.Methods: We conducted a retrospective review of one hundred fifty patients diagnosed withPSC. Patients were grouped in accordance to treatment (UDCA vs. placebo) and diseasestatus (cirrhotic-stage 4, fibrotic-stage 3/4, elevated bilirubin). Patients were then analyzedfor development of pre-established clinical endpoints that include death, transplantation,development of esophageal varices and cholangiocarcinoma. Based on treatment assignmentand baseline disease status Chi square tests were then performed to assess the presence ofa difference in occurrence of endpoints between both groups. Results: One hundred fiftypatients were included of which 49 patients developed endpoints during follow-up. Of thepatients developing endpoints 14 patients were stage 3 and 17 patients were stage 4. Infibrotic and cirrhotic patients no difference in the occurrence of endpoints existed betweenpatients receivingUDCA vs. placebo (p > 0.05). Other patients developing endpoints included8 patients with stage 2 disease and 10 patients with stage 1 disease. Interestingly, in patientswith less advanced disease (stage 1 or 2) there was an increased development of endpointsamongst patients using UDCA as compared to placebo (p < 0.01). Moreover, in the absenceof an elevated bilirubin level (total bilirubin ≤ 1.0 mg/dl) there was an increase in the rateof development of adverse events as compared to placebo; on the other hand patients withelevated bilirubin levels using UDCA were as likely to develop adverse events as patientson placebo. Conclusions: As previously shown high dose UDCA therapy is associated withhigher rates of serious adverse events in patients with PSC. As expected these outcomesoccur more commonly in patients with advanced disease whether on UDCA treatment orplacebo. Interestingly, the increased occurrence of adverse events with UDCA treatment ascompared to placebo is only significant in patients with early stage disease or normaltotal bilirubin.

Sa1821

Biochemical Response to Ursodeoxycholic Acid Predicts Long-Term Outcomein Japanese Patients With Primary Biliary Cirrhosis: A Validation StudyNobuaki Azemoto, Teru Kumagi, Masanori Abe, Ichiro Konishi, Bunzo Matsuura, YoichiHiasa, Morikazu Onji

Background: There is an ongoing need for predictors of long-term outcomes for patientswith primary biliary cirrhosis (PBC). Biochemical response to ursodeoxycholic acid (UDCA)has been introduced to predict long-term outcomes in Western countries. We have alsoreported that early biochemical response to UDCA predicted development of symptoms inasymptomatic patients with PBC [Ehime criteria (J Gastroenterol 2009; 44: 630), ≧70%reduction in GGT level above the upper limit of normal (ULN) or normal GGT level after6 months of treatment]. Aim: The aim of this study was to validate the criteria of biochemicalresponse to UDCA in predicting long-term outcomes among Japanese patients with PBC.Methods: A retrospective chart review was conducted for 227 Japanese patients with PBC.Patients taking UDCA with an observation period of more than 6 months were included inthe study. Data collection included demographics, biochemical and serological markers, andhistological stage. Four different criteria regarding biochemical response to UDCA werecompared and validated: 1) Ehime criteria; 2) Paris criteria (Hepatology 2008; 48: 871); 3)Barcelona criteria (Gastroenterology 2006; 130: 715); and 4) Rotterdam criteria (Gastroenter-ology 2009; 136: 1281). Results: In total, 138 patients (62% from the previous study) metthe inclusion criteria and underwent analysis. Overall, patients were given UDCA for a meanperiod of 55.5 months (range, 9-291 months) and were characterizable as middle-agedwomen (85.1% female; mean age at diagnosis, 58 years) in the early stage of the disease.Good biochemical response to UDCA was recorded in more than three-quarters of patientsaccording to the Ehime criteria (79.1%), Paris criteria (88.4%) and Rotterdam criteria(77.3%), but not according to Barcelona criteria (49.2%). Using the Ehime criteria, highlevels of bilirubin, AST, ALT and ALP at baseline, advanced histological stage and lack ofbiochemical response after 6 months of treatment were all associated with death or livertransplantation (P<0.05 in all). Multivariate analysis identified lack of biochemical responseaccording to Ehime criteria alone as an independent factor predicting treatment failure(odds ratio, 3.891; 95% confidence interval, 1.223 - 12.381; P=0.021). The Paris criteria[combination of 3 variables (ALP ≤3 x ULN and AST ≤2 x ULN and bilirubin ≤1 mg/dL)after 1 year of treatment] also predicted long-term outcomes in our population (P=0.003),whereas the Barcelona and Rotterdam criteria showed no such association (P=0.282 and P=0.553, respectively). Conclusions: Good biochemical response to UDCA according the Ehimeand Paris criteria is associated with long-term outcome in Japanese patients with PBC andallows identification of non-responders who may benefit from further trials. Finally, Ehimecriteria should be validated in a different patient cohort.