276A AASLD A B S T R A C T S HEPATOLOGY October 1995

677 INTRA-ORGAN VARIABILITY OF HEPATIC IRON CONCENTRATION (HIC) WITH HEPATITIS C. MP Bronner*. KV Kowdlev. TH Carlson. M Lin. RF Labb~. Departments of Pathology* & Medicine, Division of Hepatology; University of Washington, Seattle, WA. Purpose: Recent studies indicate that elevated hepatic Fe may interfere with interferon therapy of hepatitis C virus CHCV); Histologically, there is great variability of iron distribution in HCV hepatic disease. Variability of hepatic Fe distribution was therefore studied quantitatively to determine if small samples, such as those obtained by needle biopsy, can accurately reflect the true status of total hepatic iron load. Methods: Archived paraffin tissue blocks were studied from reseeted livers of the most recent l I consecutive patients undergoing transplantation at our institution (6 HCV and 5 other etiologies, including 1 non-alcoholic steatohepatitis, 2 alcehulic, and 2 cryptoganic eirrhoses). Different sampling techniques were used on caeh piece of tissue, to recapitulate small needle biopsies (ram 2 samples from one comer of each piece of liver) vs. larger 50/am thick microtome sections of the whole block surface (era 2 samples). Each block contained from 1 to 3 pieces of tissue, measuring up to 25x20x2mm, and 3 to 5 blocks were available from each patient. The samples were deparaffiuized, dried to a constant weight, digested with nitric acid, neutralized and assayed for Fe speetrophotometrically. Results: The variability of Fe distribution is shown by the coefficient of variation (%CV) for the two types of specimens assayed. As shown in the table, a greater degree of Fe variability (%CV) was noted in HCV cirrhosis. Greater variability, regardless of diagnosis, was noted in the small biopsy- sized samples (mean %CV's combining all diagnoses = 44% for small vs. 29% for large samples).

Hepatic Iron Concentration (HIC) (pg/g dry weight)

Diagnosis (n) I Bionsv-like Small Samnles Microtome Large Samuies Meat Mean± SO~a~ ~cV IMean± SO In~ ~CV ~CV

HCVCinhosis (6)11~7~79(42) I 46 1168:~:351 ¢20) I 37 42 OtharCirrhosis (51!1242:1:751(34)1 41 1617:L-4-29 o8) I 20 31

Conehialons: 1) In agreement with our earlier findings using fresh tissue, a large variation in Fe distribution exists in reseeted liver tissue using paraffin blocks. 2) The degree of HIC variability is independent of degree of Fe overload (variable whether HIC is low or high, data not shown above). 3) HCV patients have more variability in Fe distribution than those with cirrhosis due to other etiologies. 4) HIC determined by needle biopsy will not accurately assess the total hepatic Fe load, especially in HCV patients. These results suggest that quantitative HIC determination from needle biopsies will not accurately guide interferon therapy.

6 7 8 HEPATIC LIPID PEROXIDATION IN CHRONIC HEPATITIS C. M Kikuyama, Y Kobayashi, T Kawasaki, and T Yoshimi. Second Department of Internal Medicine, Hamamatsu University School of Medicine. Shizuoka, Japan.

Oxidative stress may contribute to hepatic damage in patients with chronic hepatitis C. Lipid peroxidation can occur during oxidative stress, with the resultant generation of a variety of reactive aldehydic products. Some of these aldehydic products can form covalent adducts with proteins, resulting in the formation of pr0tein-carbonyls. Ah'n: The aim of this study was to determine whether patients with chronic hepatitis C had evidence of increased hepatic lipid peroxidation, measured in liver biopsies using a histochemical procedure which detects protein-carbonyls. Patients and Methods: Twenty-two patients (18 male and 4 female, mean age 44 yr, range 23-67 yr) with chronic hepatitis C were included. As coatrols, 2 patients with fatty liver were studied. Cryostat sections of liver biopsies were stained histochemically for protein-carbonyls using the cold Schifl's method (Am J Pathol 129: 295, 1987). The area of cold SchifPs reaction products was calculated as the percent area of the total liver section by computerized image analysis, and its value was defined as the Schiff ratio. Hepatic macrophages were stained in liver sections using anti-CD68 antibody. Histological activity was quantitatively analysed using Knodell's histological activity index (HAl) scoring system. The HAI ranged from 2 to 10. Results: Liver biopsies from patients with chronic hepatitis C demonstrated histoehemical staining for protein-carbonyls in the portal area, and the Schiff ratio correlated positively with indices of hepatic necrosis and inflammation in the HAl (r=0.51: p<0.05).; Control biopsies had no Schiff- positive cells. Based on staining of sequential liver sections, most cells which were positive for protein-carbonyls also stained for CD68, a mucrophage marke~ Hepatocytes did not stain positively for protein-carbonyls. The Schiff ratio was correlated with serum y -glutamyltranspeptidase activity (r=0.71: p<O.01), but did not correlate with serum ALT activity or with the degree of hepatic fibrosis. Conclusions: This is the first demonstration of increased hepatic levels of protein-carbonyls in patients with hepatitis C. Protein-carbonyls were found predominantly in macrophages in the portal zone of the liver These results indicate that patients with chronic hepatitis C have evidence of increased hepatic lipid peroxidation, which might contribute to hepatic injury.

6 7 9 SAFETY AND UTILITY OF PERCUTANEOUS LIVER BIOPSY (PLB) IN HEMOPHILIAC CHILDREN WITH HEPATITIS C(HcV). KB Sehwarz L Hesterber~,. J Boitnott. J Casella. and G Dover. Depts. of Pediatrics and Pathology, Johns Hopkins University School of Medicine, Baltimore, Md.

There is considerable controversy surrounding the indications for, safety of, and optimal method for performing LB in hemophiliac children with viral hepatitis. This issue is of great current concern given the large number of these children with chronic HCV. Furthermore, there is general consensus that alpha-interferon (IFN), should not be initiated without a baseline LB.

Ten hemophiliac male children (ages 7 - 14 years) with 2 + tests for HCV Ab (EIA 2.0 Abbott) recently underwent PLB as part of a study to investigate the efficacy of HCV polymerase chain reaction assay (PCR) to assess response to IFN. The study was approved by the FDA and by the Johns Hopkins Committee on Clinical Investigation. Factor VHI 50 + 5 IU/kg i.v. was given pre-biopsy by bolus infusion to achieve a level of >80%; patients were randomized to receive either bolus or continuous Factor VIII for the next 72 hours. Results were as follows: Factor VIII was 15 + 9% at baseline and 90 4- 6% 1 hour post-bolus infusion, p<0.00001. Hemoglobin was 12.0 4- 0.2g/dl at baseline and 11.9 4- 0.3§/dl at 24 hours. PLB was accomplished safely in all children without need lor transfusion or operative intervention. The first 8 biopsies were done utilizing fentanyl and midazolam for conscious sedation and the automatic Microvasive needle; 3/8 children experienced pain and anxiety, associated with the presence of perihepatic fluid collections noted by ultrasound in 2. Therefore, the biopsy protocol was modified to minimize movement of and trauma to the liver at the time of the biopsy by utilizing propofol, ultrasound guidance, and the Jamshidi needle, with excellent patient tolerance. Histology was assessed in blinded fashion by a pathologist unaware of the PCR results. 3 children had histologically normal livers and serum PCR was negative, suggesting remote infection. 7 had mild - moderate hepatitis although serum aminotransferases were abnormal in only 3. PCR was positive in all 7. Conclusion: PLB is safe and useful in hemophiliac children with HCV. Certain modifications can maximize patient tolerance.

6 8 0 LIVER INJURY AND FIBROSIS IN INTERFERON ~/ GENE K N O C K - * OUT MICE. Z Shi TA Stewart and DC Rockey. Liver Center Laboratory and the Department ot Mechcme, Umverslty ot'California, San Francisco, *Genentech Inc., South San Francisco, CA.

Liver fibrosis is characterized by increased deposition of matrix proteins through mechanisms which likely involve interaction between extracenular matrix signals and cytokines. Cytokines in particular appear to be important in cellular activation during liver injury and fibrosis. For example, we have previously shown that interferon ~/ (IFN)') is antiproliferative and antifibro- genie for rat hepatic lipocytes. The aim of the Current study was to examine the fibrogenic response oflFN7 gene deficient (IFNy-/-) mine to liver injury. Methods: Balb-c and C57BL/6 wild type (IFN'~+/+) and IFN~'-/- male mice weregavaged with carbon tetrachloride (CC14) or intraperitoneally injected with dimethylnitrosamine (DMN) for 4 wks. During administration of CC14, some Balb-c IFN't-/- mice were treated with recombinant murine IFN~/ (25,000 units/day) for 4 wks. Liver histology was assessed with Masson's Trichrome. Immunohistochemistry was performed with anti-desmin anti- body. Whole liver collagen type I mRNA was detected by RNase protection assay. Results: C57BL/6 IFNT-/- mice displayed greater acute liver injury and higher mortality (l.7-fold that of C57BL/6 IFN'/+/+ mice P<0.01) than their wild type counterparts. No significant differences in acute liver injury or mortality between Balb-c wild type and IFN't-/- mice were detected. After 4 doses of CC14, granulomas were more prominent in size and number in the livers of Balb-c IFNy-/- mice than Balb-c wild type mice. Extracellular matrix was more prominent in the livers of both Balb-c and C57BL/6 IFN't-/- mice than wild type controls. Whole liver collagen type I mRNA in Balb-c and C57BL/6 IFNy-/- mice was 2.07 and 1.g6-fold that of Balb-c and C57BL/6 wild type mine respectively (n=4, P<0.01 for Balb-c strain; n=3, P<0.01 for C57BL/6 strain). After administration of DMN for 4 wks, whole liver collagen type I mRNA in Balb-c and C57BL/6 IFNy-/- mice was 3.4 and 2.9-fold mat of Balb-c and C57BL/6 wild type mice respectively (n=4, P<0.05 for Balb-c strain). After repletion of IFN~', collagen type I mRNA expression after CCI4-induced liver injury in Balb-c IFN'/-/- mice was reduced to the level of that in the livers of Balb-c wild type mice (n=3). Conelnsion: C57BL/6 IFNy-/- mice were highly susceptible to CCL-induced acute liver injury. IFNv-/- mice (Balb-c and C57BL/6 strains) expressed significantly higher levels of collagen type I mRNA and developed more prominent fibrosis after CC14 or DMN treatment than their wild type counterparts. Collagen type I mRNA expression Of CCl4-induced liver injury was downregulateffwith IFNy treatment in Balb-c IFNy-/- mice. The data indicate that IFNy is a central cytokine involved in the pathogenesis of hepatic fibrosis and suggest that IFN 7 may modulate acute liver injury under certain conditions.