JOURNAL OF PATHOLOGY, VOL. 175 77-83 (1995)

INTRAHEPATIC EXPRESSION OF HEPATITIS C VIRUS ANTIGENS IN CHRONIC LIVER DISEASE

KAYHAN T. NOURI-ARIA, RICHARD SALLIE*, MASASHI MIZOKAMI~, BERNARD c . PORTMANN AND ROGER WILLIAMS

Institute of Liver Studies, King’s College School of Medicine and Dentistry, Bessemer Road, London SE5 9PJ, U. K ; ?Second Department of Medicine, Nagoya City University, Nagoya, Japan

Received 20 June I994 Accepted 23 September 1994

SUMMARY Localization of hepatitis C virus (HCV) antigens was studied in fresh frozen and formalin-fixed, paraffin-

embedded liver tissue by immunoperoxidase using monoclonal antibodies to nucleocapsid protein and polyclonal human immunoglobulin G purified from plasma containing antibodies to structural and non-structural antigens of hepatitis C virus. The results observed using monoclonal antibody to HCV core were similar to those of polyclonal IgG against HCV antigens in the majority of cases and both correlated well with HCV status as defined by ‘nested’ polymerase chain reaction. HCV antigens were detected in both hepatocytes and mononuclear cells. Using polyclonal human IgG, a small proportion of biliary epithelial cells were also positive in 6/29 patients. In most of the specimens examined, relatively few cells (1-5 per cent) were found to be positive for HCV antigens. The cryostat sections, using polyclonal IgG against HCV antigens, exhibited greater immunohistochemical staining, suggesting that the fixation and processing of the tissue may be a major factor in the conservation and the outcome of HCV antigen@) findings. However, the results using monoclonal antibodies may reflect the specificity of antigen expression.

KEY worms-Immunohistochemistry, polymerase chain reaction, hepatitis c virus, h e r tissue.

INTRODUCTION

Hepatitis C virus (HCV), the causative agent of the majority of cases of post-transfusion hepatitis, has recently been identified, cloned, and sequenced.’,’ The agent is a single-stranded RNA virus of about 10 kilobases in length, distantly related to the flaviviruses and pestiviruses.

While serological tests based upon non- structural and structural antigens of HCV have greatly advanced knowledge of the epidemiology of non-A, non-B hepatitis (NANBH),3-5 data relating to the distribution and expression of viral antigens in liver tissue are mounting but with some degree of A wide range of

“Present address: The Liver Unit, National Institutes of Health, Bethesda, Maryland, U.S.A.

Addressee for correspondence: Professor Roger Williams, CBE, MD, FRCP, FRCS, FRCPE, FRACP, FACP (Hon), Institute of Liver Studies, King’s College School of Medicine and Dentistry, Bessemer Road, London SE5 9PJ, U.K.

0 1995 by John Wiley & Sons, Ltd. CCC 0022-341 7/95/010077-07

monoclonal antibodies (MAbs) to different regions of HCV6,8910 or polyclonal antibodies to peptides (41AA of NS4 or 191AA of NS5),93’1 or serum IgG fraction against HCV antigens were used.’ In one study carried out in patients with HCV infection, MAb against recombinant core and NS3-encoded proteins failed to demonstrate immunostaining of liver tissue, whereas MAb to cl00 antigen showed clear immunostaining of hepatocytes.” In another study, MAb to core, envelope, and NS3 of HCV antigens gave a similar pattern of hepatocyte staining.6 In all of these

however, HCV antigen was found only in the cytoplasm of hepatocytes, despite the finding of both positive and intermediate replicative strands of HCV RNA within mononuclear cells using the polymerase chain reaction (PCR)”p’4 or in situ hybridization. l 5 Using the latter technique, we have reported HCV genome expression of hepatocytes, mononuclear cells, bile duct epithe- lium, and occasionally sinusoidal cells, suggesting

78 K. T. NOURI-ARIA ET AL.

a less restricted cellular tropism for this virus.16 Recently, Nuovo et al., using in situ PCR, have demonstrated HCV RNA in hepatocyte cytoplasm and nuclei, bile duct epithelial cells, and sinusoidal Kupffer cells.17

In this paper, we report the distribution of HCV antigens within the liver tissue as detected by immunoperoxidase using a cocktail of MAbs to HCV core protein and human IgG fraction against structural and non-structural products of the HCV genome in patients whose HCV RNA status was identified by nested PCR using liver tissue. We have also compared the effects of fixation (cryostat vs. formalin) on the outcome of the immunohisto- chemical staining.

MATERIALS AND METHODS

Liver tissue was obtained from 29 patients (mean age 46.3 years; 18 male, 11 female) who underwent orthotopic liver transplantation for subacute fulminant hepatitis or chronic end-stage liver disease from a variety of causes. The hepatitis ‘C’ status of each resected liver was characterized by reverse transcription nested PCR performed on RNA extracted from fresh liver tissue. Patients were divided into two groups according to their HCV RNA status in the liver. The first group consisted of 15 HCV-RNA-negative patients [median serum aspartate aminotransferase (AST) level 380 IU/l]. Of these, five had acute liver failure [hepatitis A (n= l), hepatitis E (n= l), and para- cetamol overdose (n = 3)], three cryptogenic cir- rhosis, two Wilson’s disease, two primary biliary cirrhosis, two chronic hepatitis B with cirrhosis, and one the Budd-Chiari syndrome. Serum anti- HCV was tested in ten patients in this group and found to be positive in one. Fourteen HCV-RNA- positive patients were studied in the second group (median serum AST level 164 IUIl), ten of whom had liver disease attributed to chronic NANBH. The remaining four consisted of autoimmune chronic active hepatitis (n= I), primary biliary cir- rhosis (n=2), and alcoholic liver disease (n= l), all of whom had had multiple transfusions in the past. Thirteen of the 14 patients were tested for serum anti-HCV (United Biomedical Inc., Lake Success, N Y , U.S.A.) and nine were positive. Cirrhosis was present in 22/29 of these cases. Normal control tissue negative for HCV RNA was obtained from the unused portions of seven donor livers used for reduced liver transplantation. The specimens

from each of the three groups were then coded, randomized, and, examined blindly by immuno- histochemical staining.

IgG purijication and labelling Human plasma [the positive control plasma,

supplied with the antibody detection kit, United Biomedical Incorporation (UBI), New York, U.S.A., proven to have antibody to structural and non-structural products of the HCV genome and to be non-reactive with hepatitis B virus, human immunodeficiency virus, and human T-cell lym- phoma virus-1 and -21 and the negative control plasma provided in the kit were used in the present study. Following ion-exchange chromatography (Sephadex G-25, Pharmacia, Uppsala, Sweden) to dialyse and equilibrate plasma against 0.1 M phos- phate buffer (pH 8.0), immunoglobulins were frac- tionated over a diethylaminoethyl cellulose column (DE 52, Whatmann, Maidstone, Kent, U.K.). IgG fractions collected in the above buffer were pooled and the purity was confirmed by immunodiffusion. After quantitation [using a bicinhoninic acid (BCA-1) kit; Sigma Chemical Co., Dorset, UK], IgG was equilibrated in 0.1 M sodium bicarbonate buffer (pH 8 .5 ) prior to labelling with biotin [(long arm) NHS water-soluble (BNHSws); Vector Laboratories, Peterborough, UK], which was performed according to the manufacturer’s instructions. Briefly, 4Opl of a 25 mg/ml BNHSws solution was added to 10mg of protein and the mixture was incubated for 2 h at room tempera- ture. The reaction was stopped by addition of 10 mg of glycine and the sample was dialysed three times against phosphate-buffered saline (PBS), pH 7.2, for 24 h before use.

Nine MAbs (all IgGl except for one IgG2) to HCV core antigen were generated using a recombinant protein (cl 1, 22 kD)” and applied in immunohistochemistry. The specificity of each MAb was tested against short synthetic peptides covering the 1-120 amino acid region of HCV core protein using ELISA (Mizokami et al., personal communication).

Immunohistochemistry Portions of liver resected at transplantation were

snap-frozen in liquid nitrogen and stored at -70°C or fixed in formalin and embedded in paraffin. Cryostat sections (W3,um) were fixed in acetone and allowed to dry at room temperature.

79 HCV ANTIGEN EXPRESSION OF LIVER TISSUE

Paraffin-embedded sections were cut (4,um), dewaxed, and rehydrated before immunostaining. Endogenous peroxidase was blocked by immersion of the sections in 1 per cent H,O, in methanol for 30min, followed by treatment in 0.1 per cent Triton-lOO/PBS for 5 min to unmask the antigens. Non-specific binding of IgG to Fc receptors was blocked by preincubation of liver sections in 10 per cent pooled normal human sera prior to incu- bation with polyclonal IgG or 10 per cent bovine serum albumin in PBS (pH 7.2) prior to staining with the MAb. The sections were then incubated with appropriate dilutions of MAb to HCV core or biotinylated IgG to structural and non-structural products of HCV for 1 h at room temperature. The sections were washed three times in PBS and those stained with MAb were further incubated with biotinylated anti-mouse Ig for another 30 min. The sections were then incubated with horseradish peroxidase-conjugated avidin-biotin complex (Vectastain, ABC ‘Elite’, Vector Laboratories) for 30 min at room temperature. After three washes in PBS, the reaction was developed using 5mg of diaminobenzidine tetrahydrochloride dehydride (DAB) (Aldrich Chemical Co. Ltd., Poole, U.K.) in 10 ml of PBS containing 0.3 per cent H,O, as a substrate. Finally, the sections were counterstained with haematoxylin and analysed microscopically. They were assessed blindly by two observers (KTNA and BP), who reached agreement in every case.

Expression of HCVprotein in insect cells The NS5 and nucleocapsid regions of the

HCV genome were inserted into a baculovirus vector (AcNPV), and insect cells (Spodoptera frugipedra) were infected with the recombinant virus (BCH10)I9 and used as positive controls. Uninfected cells and cells infected with wild-type baculovirus were used as negative controls. The cells were cultured in Tc 100 Fluid (Flow, Scotland) for 72 h at room temperature and then fixed in formalin, processed, and stained in a similar fashion to liver specimens.

Reverse transcription-polymerase chain reaction Liver tissue HCV-RNA was detected by reverse

transcription PCR using primers from 5’ untrans- lating region of the HCV genome as previ- ously described.2” Outer primers were JR12 (nt 1-20, 5’-GGCGACACTCCACCATAGAT-3’)

and JR19 (nt 197-216, 5’-CGCCCAAATCTC CAGGCATT-3‘); inner primers were JR13 (nt 35-53, 5’-GAACTACTGTCTTCACGCA-3‘) and JR14 (nt 140-16 1, 5’-GGCAATTCCGGTGTA CTCACC-3).21*22 To avoid false-positive results, PCR was performed with rigid adherence to the recommendations of Kwok and H i g ~ c h i ? ~ and with appropriate positive and negative controls.

RESULTS

The specificity of MAb to HCV core protein and biotinylated IgG against structural and non-structural products of the HCV genome was examined by the immunostaining of insect cells (Spodoptera frugipedra) infected with baculovirus vector expressing a fusion peptide consisting of NSS+core regions (BCHlO) of HCV. At high magnification, the cells showed strong granular staining forming a ring-like pattern around the nucleus using the above IgG (Fig. la). Similar results were obtained using MAb to the core protein of HCV. No staining was seen when insect cells infected with the wild-type baculovirus were incubated with the biotinylated IgG fraction (Fig. lb).

Using polyclonal human IgG HCV antigens, HCV antigens were detected in the hepatocyte cytoplasm of 13 out of 14 cryostat sections and in only three of the paraffin-embedded sections from patients with chronic liver disease whose liver tissue had been found to be HCV-RNA positive using ‘nested’ PCR (Table I). In most cases, diffuse cytoplasmic staining was observed, affecting 1-5 per cent of the hepatocytes (Fig. 2a). In a few instances, the staining pattern was predominantly granular; this was more common in the group with serum anti-HCV than in the anti-HCV negative group (possibly as a formation of immune com- plexes), although no statistical significant differ- ence could be demonstrated between the two groups. Biliary epithelial and mononuclear cells positive for HCV antigens were also observed in a small proportion of patients (Fig. 2b). Liver sections from two HCV-RNA-negative by PCR revealed that a significant proportion of liver cells expressed HCV antigen staining. One had chronic HBV infection and was serum anti-HCV positive and the second patient had Wilson’s disease.

At the beginning of this study, we had access to a limited number of cryostat sections which were used for immunostaining with polyclonal IgG

80 K. T. NOURI-ARIA ET AL.

HCV ANTIGEN EXPRESSION OF LIVER TISSUE 81

antibodies against HCV antigens. As a result, MAb to HCV core protein was only applied to paraffin-embedded sections. A significant propor- tion of livers positive for HCV RNA using PCR showed hepatocyte cytoplasmic staining for HCV core antigen in the HCV-positive group (9/13) (Fig. 3a). The proportion of hepatocytes express- ing HCV core ranged between 1 and 2 per cent. Mononuclear cell staining was also observed in four livers positive for HCV PCR (Fig. 3b), whereas no bile duct staining was seen using MAb to HCV core protein. Once again, two livers which were not identified to be HCV-RNA positive using PCR expressed HCV core protein in hepatocytes or mononuclear cells.

By contrast, the seven liver sections obtained from normal donors showed no immunostaining of any cell population. Immune-absorption experi- ments with preincubation of liver sections positive for HCV antigens with unlabelled human plasma (the original plasma from which the purified IgG was isolated) abolished the expression of HCV antigen staining. The absorption of the MAb with HCV core protein also reduced the immune stain- ing of liver tissues by 80 per cent. No specific pattern of staining was observed using the IgG fraction isolated from the negative control plasma (Fig. 4).

Histological analysis revealed that lymphoid aggregateslfollicles were to be found more fre- quently in the HCV-RNA-positive group [7/14 (50 per cent)] than in the HCV-RNA-negative group [2/15 (13 per cent)] (Fisher’s exact test, P=0.035). Bile duct damage was present in 5/14 (35 per cent) of those HCV-positive vs. 1/15 (6 per cent) of the HCV-negative group (P=0*06). There was also no significant difference for portal or lobular inflam- mation between the HCV-RNA-positive group [7/14 (50 per cent)] and the HCV-negative group [4/15 (26 per cent)] (P=0*135), nor was there a

significant difference between the serum AST levels in these two groups. Furthermore, no associ- ation was found between the hepatic HCV antigen expression (using either polyclonal IgG to struc- tural and non-structural antigens of hepatitis C or MAb to HCV core protein) and the histological analysis, as judged by lymphoid aggregates/ follicles, bile duct damage/loss, and lobular or portal inflammation.

DISCUSSION

Using MAb to HCV core protein and purified human IgG to structural and non-structural products of the hepatitis C virus genome and immunohistochemistry, we have demonstrated that HCV antigens are predominantly located within the cytoplasm of hepatocytes in liver tissues from patients with chronic HCV infection. A small proportion of mononuclear cells and bile duct epithelial cells were also found to express HCV antigens in some patients, supporting our previous findings and those of Nuovo et al., where the HCV genome was identified within similar cell popu- lations using in situ hybridization and in situ PCR, respectively. 16317

The extrahepatic replication of HCV has been reported by several groups, who, using either PCR on isolated mononuclear cells from peripheral

or in situ hybridization on liver tis- sues, 5,1 demonstrated the presence of positive and intermediate replicative strands of HCV RNA in these cells, indicating a wider cell tropism for HCV. The lack of detection of HCV antigens in cells other than hepatocytes (such as resident/ infiltrated mononuclear cells) in the above studies”” may be explained by the method of fixation or patient selection. All the patients in our study had end-stage liver disease and may have

Fig. I-Immunoperoxidase staining of insect cells infected with baculovirus expressing a fusion peptide consisting of NS5 + core regions of HCV (BCHIO) (a) or insect cells infected with wild-type baculovirus (b) using the IgG fraction of human plasma positive for anti-HCV. [Original magnification: (a) x 400, (b) x 2501

Fig. 2-Immunohistochemica1 staining of liver tissue from a patient with chronic HCV infection demonstrating diffused cytoplasmic hepatocytes (a) and bile duct epithelum (b) using the IgG fraction of human plasma positive for anti-HCV. (Original magnification x 320)

Fig. 3-Immunohistochemica1 staining of liver tissue from a patient with chronic HCV infection demonstrating hepatocyte cytoplasm (a) and mononuclear cells including plasma cells (b) using the MAb to HCV core protein. (Original magnification x 250)

Fig. 4-Immunohistochemical staining of liver tissue from a patient with chronic HCV infection using the IgG fraction of human plasma negative for anti-HCV. (Original magnification x 250)

82 K. T. NOURI-ARIA ET AL.

Table I-Imunohistochemical staining of cryostat and paraffin-embedded liver tissue using monoclonal antibodies to HCV core protein and polyclonal human IgG against HCV antigens

Paraffin Cryostat ~

Polyclonal IgG Polyclonal IgG Liver Serum MAb

No. Hep. MNC BD Hep. MNC BD Hep. MNC BD PCR HCV Anti-HCV Histology

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 21 28 29

ND - ve - ve

ND ND

+ ve - ve - ve ND - ve ND - ve ND + ve + ve + ve + ve + ve + ve - ve + ve - ve - ve + ve - ve + ve NA

- ve

- ve

ALF ALF PBC cry. Cry. Wil. CHB CHB PBC ALF Bud. Wil. Cry. ALF ALF CLD CLD ACH CLD CLD CLD CLD CLD CLD PBC PBC ALD CLD CLD

Hep. =Hepatocytes; MNC=mononuclear cells; BD=bile duct epithelium; ALF=acute liver failure; PBC=primary biliary cirrhosis; Cry=cryptogenic cirrhosis; CHB=chronic hepatitis B; Bud=Budd Chiari syndrome; Wil=Wilson’s disease; CLD=chronic liver disease; ACH=autoimmune hepatitis; ALD=alcoholic hepatitis; NA=not available; ND= not determined.

had a higher rate of infectivity and different cell populations within the liver. We have also studied large sections from hepatectomy specimens, which considerably increased the chance of detecting antigens whose distribution is particularly patchy throughout the liver tissue.

Bile duct damage is a frequent finding in chronic hepatitis due to HCV infection;432s but is uncom- mon in other forms of chronic viral hepatitis. The detection of HCV antigens in those structures may shed some light on the pathogenesis of the bile duct damage which, unlike that observed in allograft rejection, graft versus host disease or

primary biliary cirrhosis, does not seem to be associated with aberrant expression of HLA-DR antigens.26 One possible explanation for the dis- crepancy between the results of the present study and those published in the literature with regard to bile duct staining for HCV antigens may be the non-specific binding of the IgG to bile duct epithe- lium in the present study. However, this seems unlikely, since only a small proportion of patients with HCV infection and none of the control liver sections showed bile duct staining. Another expla- nation may be cross-reactivity between the polyclonal IgG used in the present study and a

HCV ANTIGEN EXPRESSION OF LIVER TISSUE 83

virallhost ‘neo-antigen(s)’ expressed on the liver tissue during HCV infection. At least two such antigens (GOR2’ and interferon-induced host anti- gens2* following treatment of patients with chronic NANBH) have been reported in patients with HCV infection.

The discrepancy between the results of immuno- histochemistry and PCR found in two patients in this study may be explained by the extensive tissue processing required to extract RNA, which may have resulted in loss of viral genome and a false- negative result. An alternative explanation would be the presence of an inhibitory factor(s) reported to be present in the liver tissue (both frozen and formalin-fixed), causing false negativity of the target nucleic acids when PCR is applied.29

ACKNOWLEDGEMENTS

We thank Drs Michinori and Kyoko Kohara of the Institute of Medical Science, Tokyo, Japan for providing monoclonal antibodies to HCV core and Dr B. Rodgers of The Wellcome Foundation, Beckenham, Kent for his generous assistance in providing insect cells infected with baculovirus vectors.

1.

2.

3.

4.

5.

6.

7.

X.

REFERENCES Choo Q-L, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M. Isolation of a cDNA clone derived from a blood borne non-A, non-B viral hepatitis genome. Science 1989; 244: 359-362. Kato N, Hijikata M, Ootsuyama Y, et al. Molecular cloning of the human hepatitis C virus genome from Japanese patients with non-A, non-B hepatitis. Proc Nut1 Acad Sci USA 1990; 87: 9524-9528. Carson JA, Tedder RS, Briggs M, et al. Detection of hepatitis C viral sequences in blood donations by ‘nested’ polymerase chain reaction and prediction of infectivity. Lancet 1990 335 1419-1422. van der Poel CL, Reesink HW, Schassberg W, et al. Infectivity of blood seropositive for hepatitis C virus antibodies. Lancet 1990; 335 558-560. Simmonds P, Zhang LQ, Watson HG, et ul. Hepatitis C quantifi- cation and sequencing in blood products, haemophiliacs and drug users. Lancet 1990; 336 : 1469-1472. Hiramatsu N, Hayashi N, Naruna Y, et al. Immunohistochemical detection of hepatitis C virus-infected hepatocytes in chronic liver disease with monoclonal antibodies to core, envelope and NS3 regions of the hepatitis C virus genome. Hepatology 1992; 16: 306-3 I 1. Krawczynski K, Beach MJ, Bradley DW, el al. Hepatitis C virus antigen in hepatocytes: immunomorphologic detection and identifi- cation. Cmwmvzterology 1992; 103 622-629. Yamada G, Nishimoto H, Endou H, et al. Localization of hepatitis C viral RNA and capsid in human liver. Dig Dis Sci 1993; 3 8 882-887.

9. Tsutsumi M, Urashima S, Takada A, Date T, Tandka Y. Detection of antigens related to hepatitis C virus RNA encoding the NS5 region in the livers of patients with chronic type C hepatitis. Hepatology 1994; 1 9 265-212.

10. Sansonno D, Dammacco F. Hepatitis C v i m 1 0 0 antigen in liver tissue from patients with acute and chronic infection. Hepatology 1993; 1 8 240-245.

11. Blight K, Rowland R, Hall PM, et ul. Immunohistochemical detection of the NS4 antigen of HCV and its relation to histo- pathology. Am J Pathol 1993; 143: 1568-1573.

12. Hsieh T-T, Yao D-S, Sheen I-S, Liaw Y-F, Pao CC. Hepatitis C virus in peripheral blood mononuclear cells. Am J Clin Parhol1992; 98 392-396.

13. Wang J-W, Sheu J-S, Lind J-T, Chen T-H. Detection of replicative form of hepatitis C virus RNA in peripheral blood mononuclear cells. J Infect Dis 1992; 166: 1167-1169.

14. Zignego AS, Macchia D, Monti M, et ul. Infection of peripheral blood mononuclear cells by hepatitis C virus. J Hepatol 1992; 1 5 382-386.

15. Lamas E, Baccarini P, Housset C, Kremsdorf D, Brechot C. Detection of hepatitis C virus RNA sequences in the liver using ‘in situ’ hybridisation. J Hepatol 1992; 1 6 219-223.

16. Nouri-Aria KT, Sallie R, Sangar D, et al. Detection of genomic and intermediate replicative strands of hepatitis C virus in liver tissue by ‘in situ’ hybridisation. J Clin Invest 1993; 91: 2226-2234.

17. Nuovo GJ, Lidonnici K, MacConnell P, Lane B. Intracellular localization of polymerase chain reaction (FCR)-amplified hepatitis C cDNA. Am J Surg Path01 1993; 17: 683-690.

18. Saito M, Hasegawa A, Kashiwakuma T, et al. Performance of an enzyme-linked immunosorbent assay system for antibodies to hepatitis C virus with two new antigens (clllc7). Clin Chem 1992; 38: 2434-2439.

19. Chiba J, Ohha H, Matsuura Y, et al. Serodiagnosis of hepatitis C virus (HCV) infection with an HCV core protein molecularly expressed by a recombinant bacnlovirus. Proc Nut1 Acad Sci USA 1991; 8 8 46414645.

20. Sallie R, Rayner A, Eddleston ALWF, Portmann B, Williams R. Detection of HCV RNA in formalin fixed, paraffin embedded liver tissue by nested polymerase chain reaction. J Med Virol 1992; 37: 31g314.

21. Erlich HA, Gelfand D, Sninsky JJ. Recent advances in the polymerase chain reaction. Science 1991; 252: 1643-1651.

22. Ulrich PP, Romeo JM, Lane PK, Kelly I, Daniel LJ, Vyas GN. Detection, semiquantitation, and genetic variation in hepatitis C virus sequence amplitude from the plasma of blood donors with elevated alanine aminotransferase. J Clin Invest 1990; 86 1609-1 6 16.

23. Kowk S , Higuchi R. Avoiding false positives with PCR. Nature 1989; 339 237-238.

24. Bach N, Thung SN, Schaffner F. The histological features of chronic hepatitis C and autoimmune chronic hepatitis: a compara- tive analysis. Hepatohgy 1992; 1 5 572-577.

25. Schener PJ, Ashrafzadeh P, Sherlock S , Brown D, Duesheiko GM. The pathology of hepatitis C. Hepatology 1992; 15 567-571.

26. Danque POV, Bach N, Schaffner F, Gerber MA, Thung SN. HLA-DR expression in bile duct damage in hepatitis C. Mod Pathol 1993; 6: 327-332.

27. Mishiro S, Hoshi Y, Takeda K, et al. Non-A, non-B hepatitis specific antibodies directed at a host derived epitope: implication for an autoimmune process. Luncet 1990; 336. 140&1403.

28. Ruiz-Moreno M, Rua MJ, Carreno V, Quiroga JA, Manns M, Meyer Zum Buschenfelde KH. Autoimmune chronic hepatitis type 2 manifested during interferon therapy in children. J Hepatol 1991 12: 265-266.

29. An SF, Fleming KA. Removal of inhibitor(s) of‘ the polymerase chain reaction from formalin fixed, paraffin wax embedded tissues. J Clin Pathol 1991; 4 4 924-927.