DDW 2007:  Tuesday Abstracts:


Topic:  Hepatitis B

590. Age Greater that 40 is a Stronger Predictor of Advanced Chronic Hepatitis B (HBV) than Current North American HBV Treatment Algorithms That Demonstrate Low Predictive Value Regardless of HBeAg, Viral Load, or Alanine Aminotransferase (ALT).

V. Araya; R. Restrepo; D. Ng; R. Koka; K. D. Rothstein; M. Orrego; S. J. Munoz



Little data exists linking treatment algorithms with histological severity.

Aim: Determine whether the North American HBV treatment algorithms and/or patient demographics correlate or predict the severity of chronic HBV.


Method: Inclusion criteria:

1.     HBV-infected patients with at least two ALT’s, detectable HBV DNA quantitative levels, and eAg status obtained over a minimum of 6 months;

2.     No prior antiviral therapy;

3.     A METAVIR scored liver biopsy within 2 years of the serological data.


Peak ALT and HBV DNA levels were used to define ALT and VL categorization.


Exclusion criteria:

1.     HIV or HCV co-infection;

2.     Iron overload;

3.     Hepatocellular carcinoma.


Definitions: High viral load (HVL): ≥104 copies/ml (c/ml) if eAg -, or ≥ 105 c/ml if eAg+; low viral load (LVL): < 104 c/ml if eAg -, and < 105 c/ml if eAg+. For analysis, patients were distributed among 4 groups depending on their viral load and eAg status.


·        Groups 0: eAg (-) and LVL;

·        Group 1: eAg (-) and HVL;

·        Group 2: eAg (+) and LVL;

·        Group 3: eAg (+) and HVL.


Spearman correlation, Fisher’s exact test, and logistic regression analysis were employed for statistical modeling.



425 patients seen between 2001 and 2006 were evaluated. 96 patients met inclusion criteria. 61% were Asian, 59% males, mean age 47 (67% Age ≥ 40), 58% had persistently normal ALT, 65% were eAg negative.


No correlation existed between biopsy grade or stage, and eAg status, ALT status, gender, or ethnicity. Groups 0 and 3 correlated with biopsy grade (P=0.03 Fisher’s Exact Test, FET), but not stage (P=0.46 FET).


In Group 0, 37% had Stage ≥ 3 and 18% had Grade ≥ 3. The only variable demonstrating statistically significant correlation with stage and grade was Age ≥ 40 (P=0.004 FET).


Stage ≥ 3 was seen in 20% age <40 and 51% age ≥ 40 and Grade ≥ 3 was seen in 17% age < 40, and 36% age ≥ 40.


Logistic regression did not show that any Group, ALT status, Viral load level, or eAg status was able to predict stage or grade ≥ 3. Age ≥ 40 had the highest area under the ROC at 0.69 for advanced histological stages (3 or 4).


Combining the variables Age ≥ 40, Groups 1-3, and abnormal ALT yielded an area under the ROC of 0.71 for Stage ≥ 3 and 0.77 for Grade ≥ 3.



Current treatment algorithms for hepatitis B do not correlate nor predict the severity of chronic HBV. Since available HBV treatment can improve HBV infection severity, accurate histological diagnosis is important for risk assessment. Thus, patients age ≥ 40 may benefit most from a liver biopsy.


740. Contribution of common variants in the complement factor H gene to fibrogenesis in chronic hepatitis C virus (HCV) infection. 

F. Grünhage; H. Keppeler; H. Wasmuth; M. Odenthal; U. Drebber; H. P. Dienes; C. Hellerbrand; T. Sauerbruch; F. Lammert



Topic:  Liver Transplantation - General

T1028. The Effect of Liver Transplantation on Autonomic Dysfunction in Patients with End-Stage Liver Disease. 

E. J. Carey; D. D. Douglas; H. E. Vargas; T. J. Byrne; J. L. Rakela; V. Balan; D. C. Mulligan.



Autonomic dysfunction is a recognized complication of end-stage liver disease but there is little information on how liver transplantation (LT) affects this problem.



To prospectively evaluate autonomic function in patients with end stage liver disease before and after liver transplantation. 



Autonomic Reflex Screen (ARS) was performed on 26 patients with end stage liver disease (ESLD) prior to transplantation. The ARS tests sudomotor, cardiovagal, and adrenergic nervous system function and includes sudomotor axon reflex test (QSART) at 4 sites, heart rate response to Valsalva maneuver and deep breathing, blood pressure and heart rate responses to head-up tilt. A 10-point composite autonomic score (CAS) was calculated from these data.



26 patients (21M, 5F) with ESLD had ARS prior to liver transplantation.   Average age was 55 years, indication for liver transplant was HCV (14), cryptogenic (4), alcohol (3), other (5), average MELD score at ARS was 17. Three patients had diabetes, two without any end-organ damage and one with mild CAD.


Prior to liver transplantation, 88.5% of patients had evidence of autonomic dysfunction. Mean CAS was 2.6 (range 0-10), median 2. Sudomotor function was disturbed in 64%, cardiovagal in 58%, and adrenergic in 38%. No correlation was found between MELD score or presence of diabetes and pre-LT CAS. There was a trend towards more autonomic dysfunction in patients whose etiology of liver disease included alcohol abuse, p=0.05.


18 patients had repeat ARS an average of 9 months after liver transplantation.  Pre-LT CAS in this group was 2.9 (range 1-10) and post-LT CAS was 1.9 (range 0-5). Two patients with normal pre-LT scores remained normal post-LT. Thirteen patients with pre-LT autonomic dysfunction had stable or improved scores after LT. Three patients had worsened autonomic dysfunction after LT.



Autonomic dysfunction is common in patients with ESLD, with over 88% affected. Sudomotor or cardiovagal dysfunction is present in over 60% of patients and adrenergic dysfunction occurs in almost 40%. Although most patients show improvement in autonomic function after LT, this effect is not universal. We await the post-LT ARS results of the remaining patients in this study.


Topic:  Cirrhosis - General

T1038. Prevalence of Complications Of Liver Cirrhosis In Patients With HCV Infection And Diabetes Mellitus. 

S. Memon; N. K. Khatri; U. Soomro; W. Jafri; R. Ghori.



There is evolving evidence that the presence of diabetes mellitus (DM) increases the prevalence and severity of various complications of liver cirrhosis due to HCV infection. Prospectively we evaluated this previously unreported novel observation in our setting.



To evaluate the prevalence of complications of liver cirrhosis in patients with HCV infection and diabetes mellitus.



Patients with liver cirrhosis were enrolled who visited our clinic from June 2006 to September 2006. A questionnaire was filled regarding the presence of diabetes mellitus and other complications of liver cirrhosis, whether occurred in past or presented in emergency room or clinic.



Out of total 304 patients, 53 (17.4%) had DM with 181 (59.5%) men and 123 (40.5%) women. There was no significant difference in mean duration of cirrhosis, mean age, mean body weight and mean body mass index (BMI) in diabetes and non-diabetes patients. The prevalence of DM was significantly higher in men compared to women (p value 0.04). No difference of Child’s class in both groups. History of variceal bleeding was significantly lower (p value 0.029) in DM patients (13.2%) compared to non diabetic(27.5%), whereas there was significant difference in the prevalence of ascites, hepatic encephalopathy and hepatocellular carcinoma due to presence or absence of DM.



Presence of DM does not increase complications of liver cirrhosis, rather it decreases the prevalence of variceal bleeding, hence our results do not support the above novel observation.


Topic:  Cirrhosis – General

T1042. Spontaneous Bacterial Peritonitis: Assessment of In-Hospital Mortality Rates and Predictive Factors of Adverse Clinical Outcomes with Contemporary Medical Management. 

K. Julka; S. L. Flamm.



Spontaneous bacterial peritonitis (SBP) is a serious complication of ESLD. In-hospital mortality rates of ~35% are often quoted but mainly arise from old reports. Because of increased awareness, earlier dx and therapy and improved abx and ICU care, we speculated that in-hospital mortality from SBP has improved with contemporary medical management.



We sought to assess the current in-hospital mortality from SBP and to determine which clinical factors currently predict mortality in this pt population.



72 consecutive patients diagnosed with "suppurative peritonitis" from 1/00 - 12/04 in a tertiary care ctr were identified by ICD-9 codes. Patients were excluded from analysis if a source of abd infection was identified such as perforation or abscess. 52 patients had SBP as defined by diagnostic paracentesis during hospitalization with a cell ct of >250 PMNs in the setting of confirmed cirrhosis. Demographic information and numerous clinical factors including etiology of ESLD, Child's class, clinical symptoms including encephalopathy (HE), GI bleeding, renal failure, bacteremia and hospital day of SBP diagnosis were documented. The endpoint was in-hospital mortality. Statistical analysis was performed to assess predictive value of clinical factors related to mortality.



52 patients (31M:21F) with mean age 56.3 (range 32-81)were analyzed. EtOH (21), HCV (14),  and 3 each HBV,HCV, AIH, Cryptogenic, PSC, and other were etiology of ESLD in the majority of patients. 23 were dx'ed on day 1 (D1), 16 on D2-D5, and 13 after D5 of hospitalization. The overall mortality rate was 15/52 (29%). Renal failure (47% vs 5%) bacteremia (50% vs 20%), concomitant GI bleeding (56% vs 23%) and Child's class C vs B (33% vs 10%) were predictive of increased mortality. Presence of clinical symptoms including HE (31% vs 23%), ascites cx + (35% vs 25%)and day of diagnosis (30% D1 vs 25% D2-5 vs 31% >D5) were not predictive.



1.     Despite early diagnosis/therapy and improved abx and ICU care, the in-hosp mortality rate for SBP has not improved.

2.     Predictive factors of adverse outcomes including renal failure, concomitant GI bleeding, bacteremia and Child's class are unchanged from early reports.

3.      New strategies such as more aggressive primary prophylaxis must be sought to improve mortality from SBP.


Topic:  Cirrhosis – General

T1043. Endoscopic Band Ligation May Not Be Inferior to Endoscopic Injection Sclerotheraphy in Cardiac Variceal Bleeding, Not Fundal Variceal Bleeding. 

M. Jang; J. Jang; H. Kim; H. Kim; W. Shin; J. Lee; K. Kim; J. Park; J. Lee; H. Kim.



Endoscopic injection sclerotherapy (EIS) with cyanoacrylate has been used in conditions with both cardiac (GOV1) and fundal (GOV2/IGV1) variceal bleeding of gastric varices. However, cardiac varices are very different from fundal varices both hemodynamically and pathophysiologically. Therefore, therapeutic strategies should be individualized. This pilot study was aimed to evaluate the efficacy of endoscopic band ligation (EBL) in active cardiac variceal bleeding, not fundal variceal bleeding.



We consecutively enrolled a total of 23 patients with liver cirrhosis presenting with active cardiac variceal bleeding. They were treated by EBL (6/19, EBL group) or EIS (13/19, EIS group). Remained 4 patients could not undergo therapeutic procedures because of sudden cardiac collapse by massive bleeding. GOV1 was defined as the varices within 2-5cm of gastroesophageal junction.



Total patients were 95% of male and 52 (36-72) years old in median (range). The causes of liver cirrhosis were alcoholic (62%), HBV-associated (30%), HCV-associated (4%) and cryptogenic (4%), respectively. The patients had experienced early rebleeding (within 48 hours) in 17% (1/6) of the EBL group and 23% (3/13) of the EIS group (P=NS), and late rebleeding(within 6 months) in 17% (1/6) and 46% (6/13) in each group (P=NS). Rebleeding-free survival was not significantly different between EBL group and EIS group (165 days (1-450) vs. 90 days (1-870), P=0.39 on Kaplan-Meier curve). Mortality by cardiac variceal bleeding was not significantly different between 2 groups (17% vs. 8%, P=NS).



EBL may not be inferior to EIS in controlling an active cardiac variceal bleeding in this pilot study. Therefore, it will be warranted to perform the larger-scale study to reevaluate the role of EBL in active cardiac variceal bleeding, not fundal variceal bleeding.


Topic:  Current HCV Treatment – Side Effects

T1044. Eltrombopag Maintains Platelet Counts During Myelosuppressive Pegylated Interferon Alpha Treatment of Chronic Hepatitis C Virus Infection. 

G. M. Dusheiko; J. G. McHutchison; N. H. Afdhal; M. L. Shiffman; M. Rodriguez-Torres; S. Sigal; M. Bourliere; T. Berg; N. Blackman; F. M. Campbell; S. White.



Eltrombopag is an oral, non-peptide, small molecule thrombopoietin receptor (TPO-R) agonist. The safety, efficacy and pharmacokinetics of eltrombopag in HCV-infected subjects with thrombocytopenia precluding initiation of pegylated-interferon (PEG-IFN) and ribavirin have previously been reported. We have now examined the ability of eltrombopag to counteract the myelosuppressive effects of PEG-IFN on platelet counts in HCV-infected patients during treatment. A sub-analysis of data from study TPL102357 was performed to determine if eltrombopag is able to maintain platelet counts in thrombocytopenic subjects during PEG-IFN treatment thus avoiding deleterious dose reductions of PEG-IFN.



HCV positive subjects with compensated cirrhosis and platelet counts 20-70,000/uL were randomized (1:1:1:1) to receive 30mg (10 pts), 50mg (14 pts), 75mg (21 pts) eltrombopag or placebo once daily for 4 weeks (induction phase). Subjects achieving platelet counts of > 70,000/uL in the induction phase could initiate PEG-IFN/ribavirin therapy along with study drug for 12 weeks (maintenance phase). Platelet count assessments made at the baseline visit were compared to those made at the completion of study drug concomitant with PEG-IFN/ribavirin (Study Day 113).



A total of 74 subjects were enrolled and 49 of those successfully initiated PEG-IFN/ribavirin.

·        Eltrombopag effectively maintained platelet counts above 50,000/uL in up to 81% of patients during the first 12 weeks of antiviral treatment, counteracting the myelosuppressive effects of PEG-IFN.

·        Eltrombopag use avoided the need for PEG-IFN dose modification in approximately 90% of patients during the first 12 weeks of antiviral treatment.

·        Further research into the long-term use (48 weeks) of eltrombopag in patients with HCV –associated thrombocytopenia is warranted.


Topic:  Disease Progression – metabolic disorders

T1072. Outcome of Japanese patients with cirrhosis due to nonalcoholic steatohepatitis(NASH) and hepatitis C.  

S. Yatsuji; E. Hashimoto; A. Kabutake; M. Tobari; M. Taniai; K. Tokushige; K. Shiratori.



Background and aims:

Ethnic differences in the prevalence and features of NASH are well-documented. But, there is no information on the outcome of Japanese NASH patients. In this study we compared the outcome of Japanese with liver cirrhosis due to NASH(LC-NASH) with that of those with LC associated to hepatitis C virus infection (LC-HCV).


Patients and Methods:

We investigated the long-term morbidity and mortality of 48 patients with biopsy proven LC-NASH and 60 with biopsy proven LC-HCV who were not treated or did not respond to interferon. The end-points were survival, appearance of varices and HCC. Time to failure analysis (Kaplan-Meier), and log-rank analyses were used for across-group comparisons. The impact of baseline risk factors on survival and the development of specific complications were evaluated by logistic regression. Specific categorical features across different subsets were compared using χ2 test. Mann-Whitney test was used for across-group comparisons of numerical data. The patients were monitored every 4-6 months clinically, biochemically and ultrasonographically.



LC-NASH group: the median age was 64 y.o. (18-89 y.o.). There were 27 women (56%); 12 patients (25%) had a BMI>30, 32 (67%) a BMI>25, and 29 (60%) had diabetes. Thirty four (71%) patients were in Child-Pugh class A. During follow up (median 32.5 months; range 0.8-199), 11 patients developed some disease. Three died from HCC and 3 from a liver unrelated cause. Five patients developed HCC, 5 esophageal varices, 5 ascites, and 4 developed hepatic encephalopathy. The 5-year survival rate was 83%, and the cumulative probability of developing HCC at 5 years was 11%. LC-HCV group: the median age was 58 y.o. (33-73 y.o.). There were 31 women (52%); 2 patients (3%) had a BMI>30, 20 (33%) a BMI>25, and 16 (27%) had diabetes. Forty-three (72%) patients were in Child-Pugh class A. During follow up (median 50.7 months; range 0.4-189), 19 patients developed some disease. Six died from HCC and 2 from a liver unrelated cause. Fifteen patients developed HCC, 17 esophageal varices, 15 ascites, and 8 developed hepatic encephalopathy. The 5-year survival rate was 79%, and the cumulative probability of developing HCC at 5 years was 20%. There were no significant differences between the two groups regarding any parameter.



In this prospective study, we found no significant differences between LC-NASH and LC-HCV concerning morbidity, including HCC and mortality.


Topic:  Epidemiology – Psychosocial

T1075. Severity of Depression Symptoms and Transaminase Levels are Related in Non-alcoholic Steatohepatitis (NASH) and Chronic Hepatitis C. 

G. S. Sayuk; S. El-Dirani; J. E. Elwing; P. Lustman; M. Lisker-Melman; J. S. Crippin; R. E. Clouse.



Depression is prevalent in patients with NASH and is associated with greater degrees of steatosis and inflammation on biopsy (Psychosom Med 2006; 68:563-9). Depression in patients without liver disease is associated with insulin resistance and elevated levels of inflammatory markers (e.g., IL-6, TNFa), a plausible explanation for its relationship with NASH. We measured current depression symptoms and transaminases to assess whether these indirect measures of hepatic inflammatory activity vary with degree of depression symptoms in NASH and in chronic HCV.



58 patients with NASH and 115 patients with HCV completed the 21-item Beck Depression Inventory (BDI) at an outpatient visit. Clinical data were extracted from chart review. AST and ALT were determined contemporaneously with BDI or extracted from recent laboratory data. Linear regression models were constructed with transaminases as dependent variables and demographic (age, sex, BMI) and liver disease factors (MELD score, encephalopathy, interferon) as independent variables that might confound a relationship with BDI. Both total BDI (tBDI) and the 13-symptom cognitive scale (cBDI) were examined to avoid confounding by overlap of liver disease with depression symptoms on the total measure.



AST and ALT correlated in both NASH (r=0.61, p=0.01) and HCV patients (r=0.68, p=0.01). Mean tBDI scores were similar in NASH and HCV patients (14.1 ±1.6 vs 15.0 ±1.1, p=0.6).


In NASH, 23 patients (39.7%) had elevated tBDI (≥16) and 16 (27.6%) had scores ≥10 on the cBDI; in HCV, 49 patients (42.6%) had elevated tBDI and 32 (27.8%) had high cBDI scores. AST was greater in NASH patients with elevated tBDI vs those with lower tBDI scores (AST: 59 ±6 vs 43 ±3, p<0.01) whereas ALT was not different (72 ±13 vs 61 ±6, p=0.4).


Similar findings were observed in HCV patients (AST: 102 ±13 vs 71 ±7, p=0.02; ALT: 115 ±18 vs 83 ±8; p=0.09). Regression analyses revealed that, in NASH patients, both tBDI and cBDI predicted AST (p=0.03 for each), while no other variable was significant. ALT was inversely predicted by age, the only variable retained in its models (p<0.01 in each). In HCV patients, tBDI (p=0.04) was the sole predictor of AST, and age again inversely predicted ALT (p=0.03).



·        Current depression symptom activity is associated with AST levels in both NASH and HCV patients, even after controlling for potentially confounding factors (including liver disease severity).

·        Correlation of liver histology and longitudinal measurements of transaminases with depression activity in HCV and NASH patients will be required to clarify the significance of the relationship and its causal directionality.

·        Clinical Implications:  These data serve as preliminary evidence of a potential relationship between depression and a biomarker of hepatic inflammation in two distinct chronic liver disease populations.


Topic:  Epidemiology

T1112. Barriers to Providing Health Services for HIV/AIDS, Hepatitis C Virus Infection, and Sexually Transmitted Infections in Substance Abuse Treatment Programs in the United States. 

E. J. Bini; S. Kritz; L. S. Brown; J. Robinson; D. Alderson; J. Rotrosen.



Although substance abuse treatment programs are an important point of contact to provide health services to diagnose, treat, and prevent transmission of HIV/AIDS, hepatitis C virus (HCV) infection, and sexually transmitted infections (STI), many drug treatment programs do not offer these services. The aims of this study were to determine the proportion of substance abuse treatment programs that did not offer health services for HIV/AIDS, HCV, and STI, and to identify barriers to offering these services.



We conducted surveys of drug treatment program administrators and clinicians within the National Drug Abuse Treatment Clinical Trials Network across the United States to evaluate the availability of 4 medical services (medical history/physical examination, biological testing, medical treatment, medical monitoring), 4 non-medical services (provider education, patient education, patient risk assessment, patient counseling), funding, and other key elements involved in testing, evaluating, and caring for patients with HIV, HCV, and STI.



Completed surveys were received from 269 of the 319 treatment program administrators (84.3%) and 1,723 of the 2,210 randomly selected clinicians (78.0%). The majority of substance abuse programs were private not-for-profit organizations (78.8%) and were free-standing facilities (60.7%). One or more medical staff was present in 78.9% of programs, and 41.2% of programs reported a current patient census of more than 500 clients. Of the 8 medical and non-medical services, the median number of health services offered was 6.0 (interquartile range [IQR], 3.0 – 7.0) for HIV/AIDS, 4.0 (IQR, 2.0 – 7.0) for HCV, and 4.5 (2.0 – 7.0) for STI (p <0.01).


A high proportion of substance abuse treatment programs did not offer any of the 4 medical services either on-site or by referral for HIV/AIDS (29.8%), HCV (41.0%), or STI (39.8%) (p = 0.02). In contrast, a lower proportion of drug treatment programs did not offer any of the 4 non-medical services on-site or by referral for HIV/AIDS (4.7%), HCV (13.7%), or STI (13.6%) (p <0.01).


The most common barriers identified by program administrators and clinicians included funding, patient health insurance benefits, patient acceptance, and staff training, with funding identified as the single biggest obstacle to providing health services for these infections.



Health service delivery for HIV/AIDS, HCV, and STI is less than optimal in drug treatment programs, and there are numerous barriers to providing these services. Public health interventions to overcome these barriers to care afford an opportunity to enhance treatment and prevention.


Topic: Cirrhosis – General

T1333. Predictive Factors Of Hepatocellular Carcinoma In Cirrhosis. 

C. Fierbinteanu Braticevici; L. Ionita; L. Tribus; R. Usvat; A. Petrisor; A. Bengus.



The incidence of hepatocellular carcinoma (HCC) associated with cirrhosis have increased over the last decade. The recognition of the risk factors of hepatic carcinogenesis could improve the early diagnosis and prognosis of HCC.



To investigate the predictive factors for HCC occurrence in cirrhotic patients



Between 1998 and 2003, 365 cirrhotic patients, aged 35-65, admitted to the University Hospital Bucharest, were prospectively followed up during a 3-year period for early diagnosis of HCC. HCC was diagnosed in 31 of them. The predictive value of different risk factors was evaluated using the Kaplan-Meier method and Cox regression model. The clinical and biochemical parameters consisted of: age, sex, etiology of cirrhosis, body mass index (BMI), serum aminotransferase, serum glucose, serum lipid profile, INR, serum glutathione and platelet count.



The incidence of HCC was 3.6 % in the first year, 1.9% in the second and 4.7% in the third. Multivariate analysis demonstrated that: age (>55years), male sex, HCV (hepatitic C virus) etiology of liver disease, LDL- cholesterol < 100mg/dl, INR>1. 8 and platelet count <80 × 10 3 /mm 3were important predictors of HCC.


Comorbidities: obesity and diabetes mellitus also increased the HCC risk. According to the contribution of each of these factors, a clinico-biological score ranging between 0 and 5.23 was calculated to allow the detection of high risk patients.



·        Not all patients with cirrhosis had an equal risk for developing HCC.

·        Parameters like: age, sex, etiology, lipid profile, comorbidity, platelet count and INR interact to increase the risk of HCC and may be taken into account when selecting the cirrhotic patients to be screened for this type of tumor.