Clinical Hepatitis: Etiology, Diagnosis, and Natural History

Posters – Sunday May 21, 2006 8:00 - 5 Hepatitis C

Clinical Hepatitis: Etiology, Diagnosis, and Natural History

Abstract S1911 – The gut microflora is involved in ethanol metabolism and ethanol-induced impairment of the epithelial barrier in rodents

L. Ferrier; F. Berard; L. Debrauwer; C. Chabo; P. Langella; L. Bueno; J. Fioramonti

Background

Hepatic toxicity of alcohol is associated with endotoxemia and increased intestinal permeability. Since the colon is the main reservoir of endotoxins, impairment of the epithelial barrier may lead to endotoxemia. Alcohol diffuses easily from blood to colonic lumen, and colonic bacteria are able to metabolize alcohol into acetaldehyde, known to be a potent mast cell activator and to impair the intestinal barrier. This study aimed at investigating whether alcohol metabolism by the colonic flora may be responsible for the alcohol-induced impairment of the colonic mucosal barrier.

Methods

Male Wistar rats were given an acute intragastric administration of ethanol (1.5, 3.0 and 4.5 g/kg), or isocaloric amounts of dextrose (controls). A group was treated with non absorbable antibiotics (ampicillin and neomycin ; ATB) in drinking water for ten days. The involvement of mast cells was assessed by treating another group with the mast cell stabilizer, doxantrazole. Gut paracellular permeability (GPP) was assessed by the passage of orally administered ⁵¹Cr-EDTA into urine, and ethanol and endotoxin blood levels were measured. Ethanol metabolism by microflora was evaluated by measuring acetaldehyde in colonic contents. Finally, colonic strips from rats pretreated or not with doxantrazole, were mounted in Ussing chambers to measure permeability to dextran.

Results

An acute administration of 3.0 and 4.5 g/kg ethanol increased GPP (3.2±0.2% and 3.4±0.4%, respectively, vs. 2.2±0.2% for dextrose, p<0.05), which was significantly reduced either by doxantrazole or ATB treatment. Two hours after intake, ethanolemia was 17.4±0.2 mmol/L, and reached 34.3±0.2 mmol/L (p<0.001) under ATB treatment. Portal endotoxins were increased 90 minutes after ethanol intake (789.3±73.7 IU/L vs. 382.2±34.2 IU/L, p<0.01). Under ATB, endotoxemia returned to basal values. Acetaldehyde lumenal concentration within the colon (22.2 ±1.5 μmol/L) reached 132.6 ±31.6 μmol/L (p<0.05) and only 86.2 μmol/L±10.9 under ATB treatment. In Ussing chambers, 17 mmol/L ethanol were ineffective. Acetaldehyde dose-dependently increased dextran flux from 120 μmol/L concentration, its effects being inhibited by doxantrazole.

Conclusion

The increase in alcoholemia observed under ATB treatment indicates an important role of colonic bacteria in alcohol metabolism in rats. Colonic epithelial barrier alterations by acute ethanol intake involve colonic bacterial metabolism of ethanol into acetaldehyde, which in turn induces mast cell degranulation. These alterations may be responsible for an excessive lumen-to-blood endotoxin crossing, frequently observed in patients suffering of alcoholic liver disease

Abstract S1912 – Accuracy of indices utilizing common lab values in predicting hepatic fibrosis and cirrhosis in hepatitis C patients in a community setting

S. Ekici; J. Kim; B. Wolfman; T. Sorra; A. Khdair; I. Grosman

Several indices utilizing commonly available lab tests have been proposed for predicting the presence of significant fibrosis or cirrhosis in patients with chronic hepatitis C (CHC). However, their applicability outside the setting of referral centers has not been well studied.

Aim:

To assess the accuracy of the aspartate aminotransferase (AST)-platelet (PLT) ratio index (APRI), AST/alanine aminotransferase (ALT) ratio (AAR), age-PLT index (API) (Poynard et al. J Viral Hepatol 1997;4:199-208) and the Hepatitis C Antiviral Long-Term Treatment Against Cirrhosis Trial model (HALT-CM) (Lok et al. Hepatology 2005;42:282-292)in predicting significant fibrosis or cirrhosis in our cohort.

Methods:

Records for 149 consecutive patients with CHC who underwent biopsy at our institution from 2000 to 2005 were reviewed. Only treatment naïve patients, 18 or older, with no other concurrent liver disease or HIV were included. Lab values, including AST, ALT and international normalized ratio (INR) were obtained. All lab values were within 6 months of the biopsy date. APRI was calculated using the formula: ((AST/upper limit of normal)/PLT)x100. AAR equaled AST/ALT. API and HALT-CM were calculated using published formulas. Fibrosis (F) was staged using the modified Ishak scale (F0 to F4), with F≥2 representing significant fibrosis and F4 denoting cirrhosis. Published cut-off values for cirrhosis and significant fibrosis were used to determine sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for each test.

Results: 107 patients who met the criteria were identified; 54 were male. The mean age was 50±7.5; the mean fibrosis score was 1.8±0.9; 63% had significant fibrosis and 5% had cirrhosis. Sensitivity, specificity, PPV and NPV for each index are presented on the table.

Summary:

Both API and APRI>1.5 have low sensitivity but high specificity and PPV for significant fibrosis. APRI>2.0 was both sensitive and specific for cirrhosis. The APRI and HALT-CM could be used to reliably exclude cirrhosis. None of the indices studied had a high PPV for cirrhosis, most likely due to its low prevalence in our cohort. The AAR was not a useful predictor of cirrhosis. Other investigators have also questioned the validity of this parameter.

Conclusion

The results in our study done in a community teaching hospital validates the observations published from large referral centers.

		Sensitivity(%)	Specificity(%)	PPV(%)	NPV(%)
Prediction of Significant Fibrosis	APRI >0.5	78	68	80	64
	APRI >1.5	25	95	89	43
	API≥6	36	93	89	46
Prediction of Cirrhosis	APRI >1.0	100	75	16	100
	APRI >2.0	100	92	38	100
	HALT-CM≥20%	100	37	8	100
	HALT-CM≥50%	60	83	16	97
	AAR≥1.0	40	71	6	96

Abstract S1913 – Serum Hyaluronic Acid (HA), YKL-40, FibroSpect-II (FS-II) Test and Digital Quantification of Fibrosis (DQF) Correlate With Ishak Fibrosis Scores in Patients With Chronic Hepatitis C (HCV)

P. Mehta; R. Ploutz-Snyder; J. Nandi; R. A. Levine

Liver biopsy has limitations of sampling variation and interobserver bias. Direct serum markers of fibrosis such as HA, YKL-40 and FS-II (HA, TIMP-1 and α-2 macroglobulin) have been used as surrogate markers to assess liver fibrosis. We reported that DQF discriminates between varying stages of liver fibrosis in HCV and has excellent interobserver reliability (Liv Intl 2005; 25: 1142).

AIMS:

To correlate the accuracy of HA, YKL-40, FS-II and DQF parameters with Ishak fibrosis scores.

METHODS:

Ishak scores, DQF, HA, YKL-40 and FS-II were evaluated retrospectively in 96 patients with HCV. Mean liver biopsy length was ≧15 mm having ≧5 portal tracts. Statistical analysis was performed using SPSS (v.13.0.1)software. The Spearman's Rho statistic assessed the correlational relationships among Ishak score and the above parameters. Receiver operator characteristic (ROC) curves evaluated the accuracy of these parameters when compared to Ishak stages of fibrosis, as well as between incremental steps of the Ishak scale. Area Under the Curve (AUC) and false positive rates were determined for each predictor of fibrosis. Variable sample sizes for some of the analyses are due to pending data.

RESULTS:

All four diagnostic parameters were highly correlated with one-another (p<.01) and with Ishak scores (Table 1). ROC curves discriminated patients into various fibrosis categories (Table 2). The areas under the ROC curves were significant, indicating excellent diagnostic abilities for all four measures to detect the broad categories of the Ishak (4-6) vs (0-3), and Ishak (4-6) vs (2-3). All three parameters, except YKL-40, were able to differentiate between Ishak (2-3) vs (0-1). DQF showed the highest AUC of all measures tested for discriminating patients into these broad Ishak categories.

CONCLUSION:

Our data supports the overall ability of serum markers to discriminate between clinically relevant stages of fibrosis/cirrhosis.

(n=number)

Ishak

FS-II

DQF

YKL-40

Ishak

.762**

.878**

.537**

.779

FS-II

.762**

.674**

.630**

.853**

DQF

.878**

.674**

.371*

.686**

YKL-40

.537**

.630**

.371*

.643**

.779**

.853**

.862**

.643**

Ischak Scores

Discriminating Categories

AUC

P value

False Positive Rate/90% Sensitivity

False Positive Rate/75% Sensitiviy

4-6 vs 0-3

n=58

FS-II

DQF

YKL-40

.886

.919

.814

.872

.000

.355

.226

.355

.613

.161

.194

.290

.258

0-1 vs 2-3

n=31

FS-II

DQF

YKL-40

.746

.873

.579

.789

.023

.001

.465

.007

.833

.500

.833

.667

.417

.333

.667

.417

2-3 vs 4-6

n=46

FS-II

DQF

YKL-40

.845

.876

.770

.828

.000

.002

.000

.474

.368

.421

.737

.263

.316

.368

.421

Abstract S1914 – Co-stimulatory expansion of Hepatitis C virus-specific T cells: exploring the potential for immunotherapy in chronic HCV infection

J. Coleclough; Y. Li; M. Bonyhadi; B. Levine; B. Vance; R. H. Vonderheide; K. Chang

Background:

Hepatitis C virus (HCV) persists with impaired capacity for virus-specific T cells to expand and produce antiviral effector cytokines. Since IFN-alpha therapy does not generally increase HCV-specific T cell response (Kaplan et al, Hepatology 2005), approaches to enhance the number and function of HCV-specific T cells may complement the effectiveness of antiviral therapy.

Aim:

Based on the efficient T cell expansion demonstrated in-vitro with CD3/CD28 costimulation, we asked if functional HCV-specific T cells may be efficiently expanded in-vitro from peripheral blood mononuclear cells (PBMC) of chronically HCV-infected patients with CD3/CD28 costimulation and T cell growth factors (IL2, IL15).

Method:

PBMC from chronically HCV-infected (HCV Ab+/RNA+) and spontaneously HCV-recovered subjects (HCV Ab+/RNA-) were stimulated as follows: 1) rIL2 and HCV-derived peptides; 2) rIL2 and HCV peptides with anti-CD3/anti-CD28-conjugated costimulatory beads; 3) rIL2, HCV peptides, anti-CD3/anti-CD28 beads and rIL15. HCV-specific T cell expansion, phenotype and cytokine profile were monitored by FACS with MHC-peptide tetramers specific for 3 HCV and 1 influenza-derived HLA-A2 restricted CD8 epitopes, intracellular cytokine staining and CFSE.

Result:

Compared to rIL2 and peptide alone, overall expansion of cells in culture was enhanced by about 5 fold with CD3/CD28 costimulatory beads and 10 fold with additional rIL15 (median fold expansion in 2 weeks: 0.5 vs 2.5 vs 5.8). While highest HCV-specific CD8 T cell enrichment was observed with rIL2 and peptide alone, total number of expanded HCV-specific CD8 T cells tended to be greater with CD3/CD28 costimulation, especially with rIL15, due to increased total T cell expansion in the latter 2 conditions. Yet, compared to patients with spontaneous HCV clearance, expanded HCV-specific CD8 T cells demonstrated poor production of antiviral cytokines (IFNg, TNFa).

Conclusion:

· We conclude that HCV-specific CD8 T cells can be expanded from chronic HCV patients using CD3/CD28 costimulation and that this expansion may be further enhanced with rIL15.

· However, despite their increased expansion, HCV-specific CD8 T cells did not show efficient antiviral cytokine production.

· Further studies are ongoing to explore strategies to enhance expansion and function of HCV-specific T cells and mechanisms of T cell dysfunction in HCV persistence.

*This study is supported by RO-1 AI47519; XCyte Therapies Inc; Philadelphia VAMC; NIH/NIDDK Center of Molecular Studies in Digestive and Liver Diseases P30DK50306 and its Molecular Biology and Cell Culture Core Facilities. J.A.C is supported by the AGA Student Research Fellowship.

Abstract S1915 – Metabolic syndrome in veterans with chronic HCV infection correlates with NAFLD activity score.

P. K. Pandya; P. Shah; P. Callahan; D. Reker; S. C. Mathur

Background:

Several retrospective studies have estimated that histopathologic features of NAFLD are present in 4-7% of patients chronically infected with HCV and are associated with parameters of the metabolic syndrome (MS). Given the increased prevalence of insulin resistance and type 2 diabetes, we systematically evaluated the the clinical and laboratory parameters of MS and correlated these with concomitant histologic findings of NAFLD.

Methods:

Demographic, anthropometric, serologic, and metabolic parameters were studied in a cross-sectional study of treatment naïve veterans with chronic hepatitis C. Insulin resistance was defined as a HOMA > 2.5. MS was defined by the presence of 3 or more ATP III criteria. Serum levels of TNF-α, IL-6, adiponectin and leptin were measured. Liver biopsies were evaluated for degree and distribution(zone III) of steatosis, ballooning degeneration, mixed inflammation, and perisinusoidal fibrosis to calculate a NAFLD activity score(NAS) as recently proposed to assess the likelihood of NAFLD. Chronic hepatitis C inflammatory activity and fibrosis were determined according to the modified hepatic activity index (HAI). Histologic findings and NAS were correlated with clinical and laboratory markers of MS using Spearman correlation coefficient.

Results:

· Liver biopsies from 47 veterans (74.5% Caucasian and 14.9% African American; 96% male) were evaluated.

· The average age and BMI were 52.4 ± 6.8 and 28.4 ±6.0 respectively. 26 of the 47 veterans (53.3%) met criteria for MS. No histological features suggestive of NAFLD were present in 16 of 21 (76.2%) veterans without MS, compared to 11 out of 26 (42.3%) with MS. Findings highly suggestive of NAFLD (NAS≥5) were seen only in the group with MS (11.5%).

· In univariate analysis, the presence of MS showed statistically significant association with steatosis (p=0.0013) and NAS (p=0.010) but not with ballooning degeneration (p=0.24 or perisinusoidal fibrosis (p=0.44). The degree of ballooning degeneration showed an increasing trend with increasing HAI score, but this was not statistically significant.

· In addition, the NAS was associated with BMI (P=0.0006), TNF-alpa (p=0.012) and HOMA (p=0.013) but not with leptin (p=0.02, adiponectin (p=0.007), and IL-6 (p=0.002), but not with age (p=0.006).

· The NAS was associated with the presence and extent of perisinusoidal fibrosis (p=0.0006).

Conclusion:

· In chronic hepatitis C the presence of MS is significantly associated with steatosis and NAS.

· Findings highly suggestive of NAFLD were observed at in 12% of veterans, which is 2 times higher than reported previously.

· Although the NAS was associated with the presence and extend of perisinusoidal fibrosis, the presence of MS did not correlate with either perisnusoidal fibrosis or ballooning degeneration.

Abstract S1916 – Chronic Hepatitis C May Be The Most Common Association For Hepatic Granulomas

J. G. Martinez; S. Xiao; N. Snyder

Background:

Hepatic granulomas have been reported in up to 10 % of liver biopsie, most of which today are performed for staging of chronic hepatitis C (HCV).Increased frequency of hepatic granulomas may be associated with both mild and severe forms of HCV. It has been postulated that HCV may have a role in granuloma formation. (Yamatamoto. Hepato-Gastroenterol 42: 291; Emile,Human Pathology 24:1095 ; Goldi. Histopathology 28: 265).

Purpose:

We reviewed the cases of hepatic granulomas at our institution the last decade to determine 1) the causes of hepatic granulomas and 2) whether hepatic granulomas are associated with HCV.

Methods:

Following institutional review board approval, we used the data from our laboratory information system to identify all patients with the word “granuloma” or “granulomatous” in their final diagnosis or the pathologist’s description. When appropriate, reports, patient’s charts, and slides were reviewed, and information on special stains and cultures was obtained. Lipogranulomas were excluded from our study. The diagnosis of HCV was based the presence of serum HCV RNA.

Results:

32 patients with hepatic epitheloid and/or giant cell granulomas were identified. Special stains for at least AFB and fungi were performed on 24, and tissue cultures on 5. Thirteen (40.6%) of the patients had HCV. Among the patients with HCV, four had other associations (HBV, sarcoidosis, or HIV). Overall, 9 patients had HCV only, 3 sarcoidosis, 3 histoplasmosis, 2 HIV only, 2 mycobacterium avium complex, 2 primary biliary cirrhosis, 1 Q fever, 1 mucormycosis, 1 tuberculosis, 1 coccidiomycosis, 1 Hodgkin’s Disease, 1 cryptococcus, 1 chronic hepatitis B (HBV), 2 HCV/HBV, and 2 with no associations. Polarizing light revealed crystalloid foreign bodies in only one of the 12 patients with HCV.

Conclusions:

HCV is the most common association with hepatic granulomas found on liver biopsy at our institution. This may partly reflect that a majority of our liver biopsies were done for chronic HCV. Most of our HCV patients had a history of intravenous drug use, and the granulomas could also result from foreign material. However, we were only able to identify a crystalline substance in one patient. While granulomas are uncommon in HCV and may sometimes be associated with other problems, the finding that 28% of our cases of hepatic granulomas had HCV as their only association poses the question whether granulomas are part of the histological spectrum of chronic HCV. Granulomas may be another manifestation of the immunological abnormalities seen in HCV.

Abstract S1917 – Helicobacter Pylori Infection does not have a Significant Impact on the Natural History or Response to Treatment of Chronic HCV

C. J. Christensen; G. K. Sood; D. Little; M. Azzouz

Introduction:

Helicobacter Pylori infection(HP) is seen with increased frequency in patients with chronic HCV infection. The impact of chronic HP on immune and cytokine response,and its relationship to several chronic diseases such as coronary atherosclerosis, lymphoma, cholelithiasis, chronic fatigue, cerebral ischemia and migraine are frequently recognized in recent medical literature. We studied the impact of HP on the natural history of HCV infection including degree of histopathological changes, response to treatment and the incidence of cholelithiasis.

Methods:

Patients treated for HCV through our hepatitis clinic(1999-2005) were retrospectively evaluated. 158 patients were included, 78 patients(53%) had evidence of HP infection(group 1) and 70 patients(47%) were negative(group 2). Data collected and compared on the two groups included race, sex, age, BMI, serum ferritin, genotype, AST, ALT, HCV RNA level, ETOH history, and Diabetes Mellitus(DM). Liver biopsies were compared regarding histological activity and fibrosis stage. Gallstones were documented by imaging and treatment response was compared between the two groups.

Results:

The two groups had no statistically significant differences with regard to ETOH history, prevalence of DM, HCV genotype or RNA level. Mean age for group 1 was lower (49.81 +/- 5.86) vs (51.86 +/- 6.54) p<0.046. HP prevalence was higher in the African American AA patients, 52/78 (70%) of group 1 compared to 22/70 (30%) of group 2 (chi square 18.32, p<0.001). Laboratory parameters were similar in the two groups including ALT (92.69 vs 97.23) AST (72.08 vs 70.64), and ferritin (289 vs 289). No significant histological differences were noted between the two groups with portal inflammation score (1.83 vs 1.74, p=0.49), interface hepatitis score (1.80 vs1.76, p=0.78), lobular hepatitis score (1.46 vs 1.46, p=0.96) and (fibrosis score 1.68 vs 1.8, p=0.59). Although gallstones were noted less frequently in HP infection 21% vs 29%, this was not statistically significant(p=0.23). Our sample data analysis revealed a correlation between AST and fibrosis (r=0.40).

Conclusion:

· Our sample did not demonstrate any significant impact of HP infection on the natural history of chronic HCV or response to treatment. HP infection was more prevalent in younger and African American patients.

· A larger size sample would be needed to make a conclusion regarding the need for HP testing and treatment prior to the initiation of anti-viral therapy in patients with chronic HCV and its impact on the formation of gallstones.

Abstract S1918 – </