Autoimmune hepatitis following hepatitis A virus infection

Journal of Hepatology 1995; 23: 204-208 Printed in Denmark· All rights reserved

Copyright © Journal of Hepatology 1995

Journal of Hepatology

ISSN 0168-8278

Case Report

Autoimmune hepatitis following hepatitis A virus infection* Hans-Iko Huppertz, Ulrich Treichel2 , Andrea Maria Gassel 1, Reinhard Jeschke and Karl-Hermann Meyer zum Buschenfelde 2 Children's Hospital and 1Institute of Pathology, University of Wurzburg, and 21st Department of Internal Medicine, University of Mainz, Germany

A 7-year-old patient is reported who suffered from fatigue and jaundice due to chronic hepatitis. He had acquired hepatitis A virus infection in his community and communicated the disease to his German family 4 weeks later. While the other family members recovered from acute viral hepatitis A, the patient presented 10 weeks after the onset of hyperbilirubinemia (12 mg/dl) with the histology of chronic hepatitis, absence of markers for viral persistence, presence of autoantibodies against smooth muscle (1:320) and the asialoglycoprotein receptor (1:600), and marked hypergammaglobulinemia (3700 mg/dl), leading to the diagnosis of autoimmune hepatitis. The patient received im-

hepatitis (AIH) is characterized by and piecemeal necrosis, certain autoantibodies and hypergammaglobulinemia (1). In children as well as in adults, hepatitis is called "chronic" when it lasts at least 10 weeks and is capable of progression towards cirrhosis (1). Autoantibodies associated with autoimmune hepatitis are directed against a variety of non-organ-specific and organ-specific target antigens, i.e. undefined nuclear ("ANA"), smooth muscle ("SMA") and cytokeratin (soluble liver antigen "SLA") epitopes, against enzymes (P450 IID6, a target for antibodies against microsomal liver-kidney antigens "LKM"), crude liver preparations (liver-specific protein "LSP") and the hepatic asialoglycoprotein receptor (ASGPR) (2). Hypergammaglobulinemia is usually marked and associated with disease severity. Viral infections due to hepatitis B or C viruses, cytomegalovirus or EpsteinBarr-virus must be excluded. The disease may occur in AUTOIMMUNE

J-\.. periportal lymphocytic inflammation

* Dedicated to the memory of Dr. Attarin Huppertz.

munosuppressive therapy, symptoms of liver disease disappeared, and autoantibodies cleared from circulation. The case is discussed in the context of a putative virus-induced autoimmune hepatitis in childhood. Autoimmune hepatitis may be induced by an external trigger. Hepatitis A virus infection is one of probably several triggers that may induce autoimmune hepatitis in predisposed individuals.

Key words: Autoimmune hepatitis; Hepatitis A virus; Pathogenesis. © Journal of Hepatology.

predisposed patients encountering an unknown external trigger. A predisposing factor in Caucasians is the RLA haplotype AI, B8, DR3 (3). In addition, a defect in suppressor-inducer T-cells controlling immune responses against the asialoglycoprotein receptor (ASGPR) has been described (4). AIR may be initiated by drugs including tienilic acid that was found to induce the so-called anti-LKM-2 antibodies (5). Viruses causing acute hepatitis due to viral damage of hepatocytes or immune reactions against viral antigens have been suggested as the putative crucial event for the development of autoimmune hepatitis, for instance by molecular mimicry, as has been found between the major LKM target P450 IID6 and herpes simplex virus type I (6). Recently three cases have been reported that occurred in close temporal association with hepatitis A virus infection (7,8). Obviously, RAV infection is a very rare antecedent of AIR. In this report we describe a patient with AIR following a community-acquired hepatitis A virus infection.

Received 17 August; revised 24 October; accepted 23 November 1994

Case Report

Correspondence: Dr. Hans-Iko Huppertz, PD, Universitatskinderklinik, losef-Schneider-Str. 2, D-97080 Wiirzburg, Germany.

A 7-year-old patient presented with malaise and anorexia for a few days and then jaundice appeared in November 1992 with dark urine and pale stools. Total bili-

204

Hepatitis A virus in autoimmune hepatitis

rubin in serum was 3.9 mg/dl and liver enzymes were markedly elevated (Fig. 1). IgM antibodies to hepatitis A virus (HAV) were found, establishing the diagnosis of acute hepatitis A. After a few days the boy felt better and the jaundice disappeared. The patient had probably contracted hepatitis A in the local soccer club where the public health physician had diagnosed several cases of hepatitis A in recent months. The boy in turn infected his younger non-identical twin siblings, who recovered after a few days of jaundice. Six weeks after the first onset of hyperbilirubinemia the patient became adynamic and jaundiced again, losing 3 kg of body weight. On examination he was icteric without scratch marks, showed some fatigue, but had undisturbed memory and intelligence. The liver was grossly enlarged and the abdomen was bulging with a tympanitic percussion sound. Serum total bilirubin level was 12.0 mg/dl. Results of further laboratory tests

ALTIGPT

IPc G I

[mg dl]~ 5000

[UIl)

500

0

I

8

I

6

~

4

4000

300

3000 ~

I I

I

100

2000

I 0

anti·HAV IgM

anti-8MA

NO

1

I

I I

1000 ~ I

2

I I

I

o~

'" I Immunosuppressive therapy 2

3

./-

NO

+

NO

4

5

6

7

8

0

I

I

9 10 11 [months] NO

anti ASGPR

NO

+

NO

NO

ALTfGPT «20 Uti)

48.

489

115

1.·

IgG

NO

3700

2350

1560

1430

1570

3 .•

12.0

1.7

01

0.2

0.2

«

10

~ I

400

200

Bilirubin [mg/dl]

are given in Fig. 1. Clotting factors were normal, except for a decrease in factor VIII level to 35% of normal. IgM antibodies to HAV were still present. Further hepatotropic infections, inducing hepatitis viruses (HV) B, C, D, E, F, Epstein-Barr virus, cytomegalovirus (CMV), leptospirosis, listeriosis and toxoplasmosis, were excluded by serology and polymerase chain reaction in serum for hepatitis C virus sequences. Metabolic diseases including Wilson's disease, alphalantitrypsin-deficiency, and cystic fibrosis were excluded. The boy showed hypergammaglobulinemia and was positive for rheumatoid factor. Autoantibodies against smooth muscle (1:320, tested by immunofluorescence (9)) and the asialoglycoprotein receptor (1:600, tested by Elisa and immunoblot analysis (10)) were positive (Fig. I). Antibodies to nuclear (ANA) or mitochondrial (AMA) antigens, to soluble liver antigen (SLA), and to liver-kidney-microsomes (LKM) were negative. HLA-typing was AI, A2, Bw57, Bw62, Cw3, Cw6, DR7, DRw8, DQw2. Liver ultrasound showed an enlarged liver and a dilated portal vein, but no intrahepatic cholestasis. Liver needle biopsy was performed 10 weeks after the onset of jaundice. Formaldehyde-fixed and routinely processed tissue sections showed dense portal tract infiltrates by mononuclear cells and few eosinophils. The infiltrates spread into the adjacent parenchyma showing piecemeal necrosis (Fig. 2). Fibrosis of portal tracts and the beginning of bridging fibrosis was found. The mononuclear cell infiltrates were characterized on snap frozen tissue sections analyzed by indirect immunoperoxidase staining with monoclonal antibodies against cell surface

1132mg/dl)

Bilirubin « 1 mg/dl)

Fig. 1. Laboratory findings in a 7-year-old patient with hepatitis A virus infection and ensuing autoimmune hepatitis. The bars represent bilirubin levels (mg/dl), the circles (-0-) give alanine aminotransferase (ALT/GPT in u/I) levels, the crosses (-x-) total serum immunoglobulin (G (IgG) levels. Gamma-glutamyltranspeptidase was elevated at the first examination during acute hepatitis A (time point 0) (267 u/l; normal values <14 U/l). Alkaline phosphatase was within normal range at all time points. Both anti-SMA and anti-ASGPR autoantibodies showed declining titers. Immunosuppressive therapy was established giving prednisone and azathioprine with stepwise dose-reduction. ND= not done, HAV=hepatitis A virus, SMA=smooth muscle antigen, ASGPR=asialoglycoprotein receptor.

Fig. 2. Liver tissue of a 7-year-old patient with autoimmune hepatitis 10 weeks after the beginning of jaundice due to hepatitis A virus infection: dense mononuclear infiltrate in portal tracts and piecemeal necrosis in adjacent parenchyma (arrows). Hematoxylin/eosine staining, original magnification 250X.

205

H. I. Huppertz et al. TABLE 1 Anti-asialoglycoprotein-receptor (ASGPR) reactivity in a 7-year-old boy with autoimmune hepatitis following hepatitis A virus infection and in his family Anti-ASGPR autoantibodies

Father (37 years) Mother (33 years) 1st twin sibling (5 years) 2nd twin sibling (5 years) Patient (7 years)

negative negative negative negative positive (1 :200)

Proliferative response to human ASGPR * (cpm) PHA

Medium

ASGPR

132073±6444 68073±3078 11567±847 19181±1566 27333± 1287

1327±235 871±57 116±45 248±98 48±12

950±232 699±50 160±23 14l±2l 34±13

* Proliferative response was assessed in peripheral blood lymphocytes as described earlier (17). Two micrograms/well of phytohemagglutinin (PHA) and medium alone served as controls. ASGPR was used in a concentration of 2 Jig/well. All determinations have been done in quadruplicate.

markers. The majority of mononuclear cells were Tcells (CD3+) of the helper/inducer type (CD4+). Some suppressor/cytotoxic T-cells (CD8+) were found in areas of hepatocyte necrosis. Macrophages (CD68 + ) were found in moderate numbers in portal tracts and in areas of parenchymal necrosis.' Only a very few Bcells (CD21 +) were detected. Viral antigens (CMV, HBV-surface and -core antigens) could not be demonstrated and polymerase chain reactions for HAV-RNA, HCV-RNA, and CMV-DNA were negative. Ten weeks after the onset of hyperbilirubinemia the diagnosis of autoimmune hepatitis (AIH) was established and the patient was treated with prednisone (60 mg/m 2) and azathioprine (I mglkg). During the following months liver size and levels of liver enzymes and bilirubin returned to normal. Three months later antibodies to smooth muscle and immunoglobulin levels declined. Seven months after the onset of hyperbilirubinemia, autoantibody titers against the asialoglycoprotein-receptor (anti-ASGPR) were still elevated while they were negative in his parents and siblings. At this time point, the proliferative response of peripheral blood lymphocytes against purified human ASGPR was negative in both the patient and the family members (Table 1). Finally, anti-ASGPR antibodies disappeared from circulation a further 6 months later when the child was completely healthy (Fig. 1). Immunosuppressive therapy is now scheduled to be terminated.

Discussion We have described a patient with the typical features of autoimmune hepatitis, including autoantibodies against smooth muscle and the asialoglycoproteinreceptor; hypergammaglobulinemia, absence of a persistent infection by a known virus, and histologic features of chronic hepatitis, including parenchymal necrosis and fibrosis. This is the first report of childhood onset AIH in association with preceding hepatitis A virus

206

(HAV) infection in a patient without a family history of AIH. Since there are no laboratory data available prior to the patient's contracting HAV infection, we cannot entirely exclude the possibility that he might have had underlying asymptomatic AIH. As induction of autoimmunity is not well characterized in man, HAV infection in our patient might have been one of several events leading to AIH. It is postulated that the pathogenesis of AIH requires a so far unknown trigger to set off a sequence of events leading to AIH in a predisposed individual. A prospective study following relatives of patients with AIH described two cases of AIH associated with HAV infection in patients who had first-degree relatives with AIH (7). In this study two of three patients with subclinical acute hepatitis A from a group of 58 healthy relatives of patients with AIH were found to develop AIH within 5 months. Interestingly, in these patients rising anti-ASGPR titers and a so-called suppressor T-cell defect specific for ASGPR-reactive Tcells were determined during the follow up. Recently, an adult patient with no family history of AIH was reported to have developed AIH after HAV infection (8).

In some patients, relapsing hepatitis A has been described showing clinical and histologic features reminiscent of chronic hepatitis in the continued presence of infectious virus and specific IgM antibodies (11, 12). However, in our patient IgM antibodies to HAV were already negative at the time of biopsy 10 weeks after the onset of jaundice. The liver histology of our patient showing parenchymal necrosis and early bridging fibrosis was distinct from the histology of acute viral infection. Immunohistochemical features of the portal tract mononuclear infiltrate were consistent with findings described in pediatric AIH (13). Moreover, liver tissue of our patient did not harbour genomic se-

Hepatitis A virus in autoimmune hepatitis

quences of HAY. Thus, the disease was not due to persistent HAV infection. The patient's serum contained circulating smooth muscle antibodies and anti-ASGPR antibodies as markers of AIH. The presence and titers of antibodies to the ASGPR reflect the clinical course of AIH (10). However, anti-ASGPR antibodies have also been found in patients with acute viral hepatitis (10). Interestingly, although the twin siblings of the patient also developed acute viral hepatitis A, only the patient with AIH displayed circulating anti-ASGPR antibodies. This is consistent with Vento's finding that antiASGPR antibodies declined within a few weeks in patients with HAV infection who recovered spontaneously, while levels of this autoantibody remained elevated in those patients developing AIH (7, 14). Therefore, in our case the formation of these organspecific autoantibodies was clearly associated with the entity "AIH", suggesting a similar prognostic value of these autoantibodies, as reported for anti-LSP (liverspecific protein) (15). . While antibody titers are elevated during active inflammation, 50% of patients with AIH in remission clear anti-ASGPR autoantibodies from circulation (16). In the present case circulating anti-ASGPR were found at the time of peak activity of AIH. However, while low titers of anti-ASGPR antibodies were found after 6 months of immunosuppressive therapy and subsequent induction of clinical remission? peripheral Tcell response to the ASGPR was absent (Table 1). This could be explained by a low sensitivity of this assay, which uses T-cell proliferation in vitro. Alternatively, a T-cell independent formation of anti-ASGPR antibodies after initiation of immunosuppressive therapy might exist. A third explanation could be that induction of anti-ASGPR antibodies by ASGPR-specific Tcell clones, as observed in vitro (17), could take place locally in the liver, and reactive T-cells might not appear in peripheral blood. The viral induction of autoimmunity might be due to non-specific stimulation of the immunological network, since Ruiz-Moreno and coworkers have shown that interferon treatment of NANB-hepatitis might initiate AIH in children (18). In summary, HAV infection is a rare event at the onset of AIH. However, this report supports the putative role of this hepatotropic virus infection as a trigger of autoimmune liver disease which can occur in childhood. ASGPR is a liver-specific autoantigen which may represent an important candidate for initiating or perpetuating the autoimmune liver disease. The proposed pathogenetic sequence of events leading to AIH might be a model for other causes of AIH without a known trigger. In addition, anti-ASGPR auto anti-

bodies might be useful markers for monitoring the development of AIH and might help in guiding therapy in patients with chronic liver disease after acute viral infection, but without evidence for viral persistence.

Acknowledgements We are grateful to Dr. Haaf and Dr. Ebert (Tauberbischhofsheim) for contributing clinical data of the patient, to Prof. Manns (Hannover) for measuring autoantibodies and to Prof. Seelig (Karlsruhe) for performing polymerase chain reaction for hepatitis A virus sequences in liver biopsy specimen. Finally, we thank Dr. Michael Schilsky (New York) for reading the manuscript.

References l. Johnson PJ, McFarlane IG. Meeting Report: International

Autoimmune Hepatitis Group. Hepatology 1993; 18: 9981005. 2. Meyer zum Buschenfelde KH, Lohse AW, Manns M, Poralla T. Autoimmunity and liver disease. Hepatology 1990; 12: 354-63. 3. Krawitt EL, Albertini RJ. Immunogenetic studies of autoimmune liver diseases. In: Krawitt EL, Wiesner RH eds. Autoimmune Liver Diseases. New York: Raven Press, 1991: 6373. 4. Vento S, Eddleston A; Autoimmunity and liver diseases. In: Popper H, Schaffner F. Progress in Liver Diseases, vol. IX, Philadelphia: WB Saunders, 1990: 335-343. 5. Homberg JC, Adner C, Abuaf N. A new anti-liver-kidney microsome antibody (anti-LKM-2) in tienilic acid-induced hepatitis. Clin Exp Immunol 1983; 55: 232-8. 6. Manns M, Griffin KJ, Sullivan KD, Meyer zum Buschenfelde KH. LKM-l autoantibodies recognize a short linear sequence in P450 IID6. J Clin Invest 1991: 88: 1370-78. 7. Vento S, Garofano T, DiPerri G, Dolci L, Concia E, Bassetti D. Identification of hepatitis A virus as a trigger for autoimmune chronic hepatitis type 1 in susceptible individuals. Lancet 1991; 337: 1183-7. 8. Rahaman SM, Chira P, Koff RS. Idiopathic autoimmune chronic hepatitis triggered by hepatitis A. Am J Gastroenterol 1994; 89: 106-8. 9. Manns M, Gerken G, Kyriatsoulis A, Staritz M, Meyer zum Buschenfelde KH. Characterisation of a new subgroup of autoimmune chronic active hepatitis by autoantibodies against a soluble liver antigen. Lancet 1987; i: 292-4. 10. Treichel U, Poralla T, Hess G, Manns M, Meyer zum Buschenfelde KH. Autoantibodies to human asialoglycoprotein receptor in autoimmune-type chronic hepatitis. Hepatology 1990; 11: 606-12. 11. de la Rubia L, del Pozo F, Carrasco M, de Rueda 00, Cebrero M, de Frias EG. Persistence of anti-HAV IgM antibodies and relapse of hepatitis type A in childhood. J Hepato11993; 17: 421-2. 12. Schiff ER. Atypical clinical manifestations of hepatitis A. Vaccine 1992; 10: 18-28. 13. Senaldi G, Portmann B, Mowat Ap, Mieli-Vergani G, Vergani D. Immunohistochemical features of the portal tract mononuclear cell infiltrate in chronic aggressive hepatitis. Arch Dis Child 1990; 67: 336-343.

207

H. I. Huppertz et al.

14. Vento S, McFarlane BM, McSorley CG, Ranieri S, GuilianiPiccari G, Da1 Monte PR, Verucchi G, Williams R, Chiodi F, McFarlane IG. Liver autoreactivity in acute virus A, B and non-A, non-B hepatitis. J C1in Lab Immunol 1988; 25: 1-7. 15. McFarlane IG, Hagarty HE, McSorley CG, McFarlane B, Williams R. Antibodies to liver-specific protein predict outcome of treatment withdrawal in autoimmune chronic active hepatitis. Lancet 1984; ii: 954-6. 16. Treichel U, Gerken G, Rossol S, Rotthauwe HW; Meyer zum Buschenfe1de KH. Autoantibodies against the human asialoglycoprotein receptor: effects of therapy in autoimmune and

208

virus-induced chronic active hepatitis. J Hepatol 1993; 19: 55-63. 17. Lohr H, Treichel U, Poralla T, Manns M, Meyer zum Buschenfelde KH, Fleischer B. Liver-infiltrating T helper cells stimulate the production of autoantibodies against the human asialoglycoprotein receptor in autoimmune chronic active hepatitis in vitro. Clin Exp Immunol 1992; 88: 45-9. 18. Ruiz-Moreno M, Rua J, Carreno V, Quiroga JA, Manns M, Meyer zum Buschenfelde KH. Autoimmune chronic hepatitis type 2 manifested during interferon therapy in children [letter]. J Hepatol 1991; 12: 265-6.