Autoimmune hepatitis, from mechanisms to therapy

Abstract

In 1950, Waldenström was the first to describe a chronic form of hepatitis in young women. Subsequently, the disease was found to be associated with other autoimmune syndromes and was later termed “lupoid hepatitis” because of the presence of antinuclear antibodies. In 1965, it became designated by Mackay et al. as “autoimmune hepatitis” at an international meeting, at which the general concept of autoimmunity was endorsed by the scientific community. In the early 1960s and 1970s, the value of immunosuppressive therapy with glucocorticoids and/or azathioprine was well documented in several studies. The original association of autoimmune hepatitis (AIH) and HLA alleles, which has remarkably stood the test of time, was published in 1972. In the 1970s and 1980s, several autoantibodies were identified in patients with autoimmune hepatitis directed against proteins of the endoplasmatic reticulum expressed in liver and kidney and against soluble liver antigens. Subsequently, the molecular targets of these antibodies were identified and more precisely characterized. In the last two decades many additional pieces of the AIH puzzle have been collected leading to the identification of additional antibodies and genes associated with AIH and to the emergence of new therapeutic agents. Meanwhile, the immunoserological and genetic heterogeneity of AIH is well established and it has become obvious that clinical manifestations, disease behavior, and treatment outcome may vary by racial groups, geographical regions and genetic predisposition. Currently, the International Autoimmune hepatitis group is endorsing multi-center collaborative studies to more precisely define the features at disease presentation and to define prognostic indices and appropriate treatment algorithms. Given the importance of serological testing, the IAHG is also working on guidelines and procedures for more reliable and standardized testing of autoantibodies. (Hepatology 2006;43;S132–S144.)

Autoimmune hepatitis (AIH) is a chronic inflammatory disease of the liver, characterized by a loss of tolerance against hepatocytes leading to the destruction of hepatic parenchyma.1 Based on limited epidemiological data, the prevalence of AIH is estimated to range between 50 and 200 cases per 1 million in Western Europe and North America among the Caucasian population. In 1992 an international panel met in Brighton, UK, to establish diagnostic criteria for AIH, because it was recognized that several features including histological changes and clinical presentation can often be found in other liver disorders such as biliary, viral, drug and alcohol related liver diseases. According to this approach the diagnosis depends on a combination of suggestive features together with the exclusion of other causes of chronic liver diseases. In 1999 the International AIH group (IAIHG) reviewed the descriptive criteria and the scoring system. According to the revised criteria of the IAHG the diagnosis of AIH relies on the following points (Table 1).2

Table 1. International Diagnostic Criteria for the Diagnosis of Autoimmune Hepatitis

Parameter	Score
Gender
Female	+2
Male	0
Serum biochemistry
Ratio of elevation of serum alkaline phosphatase vs. aminotransferase
>3.0	−2
1.5-3	0
<1.5	+2
Total serum globulin, γ-globulin or IgG
Times upper normal limit
>2.0	+3
1.5-2.0	+2
1.0-1.5	+1
<1.0	0
ANA, SMA or LKM-1 (titers by Immunofluorescence on rodent tissues)
>1:80	+3
1:80	+2
1:40	+1
<1:40	0
AMA
Positive	−4
Negative	0
Seropositivity for other defined autoantibodies
pANCA, anti-LC-1, anti-SLA/LP, anti-ASGPR, anti-sulfatide	+2
Hepatitis viral markers (HAV, HBV, HCV)
Negative	+3
Positive	−3
History of drug usage
Yes	−4
No	+1
Alcohol (average consumption)
<25 gm/day	+2
>60 gm/day	−2
Genetic factors: HLA DR3 or DR4
Yes	+1
Other autoimmune diseases
Yes	+2
Liver histology
Interface hepatitis	+3
Predominant lymphocytic infiltrate	+1
Rosetting of liver cells	+1
None of the above	−5
Biliary changes	−3
Other changes	−3
Response to therapy
Complete	+2
Relapse	+3

Descriptive Criteria

AIH predominantly affects women at any age.3 The clinical appearance ranges from absence of symptoms, in which cases the disease is discovered when abnormal liver enzymes are obtained during routine laboratory tests to a severe, sometimes, fulminant course of the disease.1 In most cases the onset is usually uncharacteristic with fatigue, fluctuating jaundice, right upper quadrant pain, and arthralgia. An association with extrahepatic autoimmune diseases such as rheumatoid arthritis, autoimmune thyroiditis, ulcerative colitis and diabetes mellitus is common.4 Biochemical evaluation reveals in most cases a more striking elevation of aminotransferase values than those of bilirubin and alkaline phosphatase. In order to discriminate against cholestatic diseases the scoring system includes negative weightings for ratios of the elevation of serum alkaline phosphatase to aspartate or alanine aminotransferase activities that exceeded 3.0. Another characteristic biochemical finding of AIH is a marked elevation of serum globulins, in particular γ-globulins, which is present in most cases. Serum concentrations of alpha-1-antitrypsin, copper and ceruloplasmin should be normal, however some patients might be included with abnormal serum copper and ceruloplasmin values provided that Wilson disease has been excluded. Autoantibodies are one of the distinguishing features of AIH. Seropositivity is considered at titers greater than 1:40 in adults and 1:20 for SMA and 1:10 for LKM1 in children. Antinuclear antibodies (ANA), anti smooth muscle antibodies (SMA) and anti liver-kidney microsomal autoantibodies (LKM-1) are regarded as definite antibodies. Other “defined” autoantibodies are those for which there are published data relating to methodology of detection and relevance to AIH. These include perinuclear antineutrophil cytoplasmic antibodies (pANCA), antibodies to liver-cytosol type 1 (anti-LC-1), autoantibodies to soluble liver antigen/liver pancreas antigen (anti-SLA/LP) and antibodies to the asialoglycoprotein receptor (anti-ASGPR). Serological evaluation should include testing for anti-mitochondrial antibodies (AMA) to exclude primary biliary cirrhosis (PBC). Although the histological appearance of AIH is characteristic, there is no specific histological feature capable of proving the diagnosis. Histological features usually include periportal hepatitis with lymphocytic infiltrates, plasma cells, and piecemeal necrosis. Histological changes of bile duct injury such as destructive cholangitis and ductopenia are not recognized as typical changes in AIH, and the scoring system includes a negative weighting. Mild biliary changes however might be seen in patients with AIH and should not necessarily alter diagnosis or treatment strategies in patients with otherwise satisfying criteria for AIH.5 Genetic factors are involved in the pathogenesis of AIH. In previous studies human leukocyte antigens (HLA) DRB1*0301and DRB*0401 have been identified as most relevant for susceptibility and/or severity to AIH in Caucasian patients.6 Viral hepatitis should be excluded by the use of reliable, commercially available tests. The exclusion of current hepatitis A, B and C infection is considered adequate in most cases. A history of moderate to heavy alcohol intake or recent use of known hepatotoxic drugs does not exclude AIH if liver damage continues after abstinence of alcohol or withdrawal of the drug. The scoring system includes a negative weighting for both histories to exclude probable AIH in those patients.

AIH, autoimmune hepatitis; IAIHG, International AIH group; ANA, antinuclear antibodies; SMA anti smooth muscle antiboides; LKM-1, anti liver-kidney microsomal autoantibodies; pANCA, perinuclear antineutrophil cytoplasmic antibodies; anti-LC-1; antibodies to liver-cytosol type 1; anti-SLA/LP; soluble liver antigen/liver pancreas antigen; anti-ASGPR, antibodies to the asialoglycoprotein receptor; AMA, anti-mitochondrial antibodies; PBC primary biliary cirrhosis; HLA, human leukocyte antigens; PSA, primary sclerosing cholangitis; KC-1, liver-cytosol type 1; UGT1A, uridine 5′-diphosphate glucuronosyltransferase; IH, immune hepatitis.

Validation of the Scoring System

The validity of the scoring system has been evaluated in several major studies with more than 1000 adult and pediatric patients.2 The overall diagnostic accuracy of the “first” scoring system for the diagnosis of AIH was 89%. The new scoring system seems to more precisely differentiate between biliary diseases and AIH and application of the revised system in PBC and PSC patients revealed significantly less patients scored for definite and probable AIH.7, 8

Natural Course

Overall there are only scarce data describing the natural history of untreated AIH, keeping in mind that the last placebo-controlled immunosuppressive treatment trial has been published in 1970s.9–12 These patients had been screened for epidemiological risk factors for viral hepatitis, but have not been screened for HCV infection and they had not been characterized by the now standardized diagnostic criteria. Nevertheless these studies revealed that untreated AIH had a very poor prognosis and 5- and 10-year survival rates of 50% and 10% have been reported. They furthermore demonstrated that immunosuppressive treatment significantly improved survival of patients with chronic active hepatitis. The 10-year overall survival of AIH patients is now estimated to range between 80% and 93%.3, 13

Up to 30% of adult patients have histological features of cirrhosis at diagnosis.3, 14 Recent data indicate that only a small number of patients develop cirrhosis during therapy and that fibrosis scores are stable or improve in up to 75% patients.15 Progression of fibrosis mainly occurs in patients with increased inflammatory activity under steroid therapy.15 The frequency of remission, and treatment failure are comparable in patients with and without cirrhosis at presentation.13 Noteworthy, however, the presence of cirrhosis at baseline significantly increases the risk of death or liver transplantation.3, 16

Almost half of children with AIH already have cirrhosis at the time of diagnosis.4, 17 Long-term follow-up revealed that only a few children can completely stop all treatment and about 70% of children receive long-term treatment until they reach adulthood.4, 18 About 15% of cases develop chronic liver failure and undergo transplantation before the age of 18 years.

In elderly patients a more severe initial histology grade has been reported, but the frequency of definite cirrhosis seems not to be different from younger patients.19, 20 Response to immunosuppression is similar in older and younger patients and up to 90% of the older patients reach remission. It is still a matter of debate whether older people might need higher19 or lower21 doses of corticosteroids. In a study from the United Kingdom, 42% of the elderly patients with AIH received no immunosuppressive therapy and the prognosis appeared no worse than in younger, usually treated patients indicating that steroid therapy has to be individualized.22

Finally, clinical presentation and outcome might vary between racial groups and geographical regions. Cirrhosis at presentation is more frequent in black North American AIH patients than in their white North American counterparts.23 They are also younger at presentation similar to patients from Brazil and Argentina.24 Africans, Asians and Arabs also have an earlier disease onset than patients from Northern Europe.25 These patients and Alaskan natives additionally appear to have a higher frequency of cholestatic laboratory findings and of acute icteric disease.26

The risk of hepatocellular carcinoma in autoimmune liver disease varies considerably between the different diseases PBC, primary sclerosing cholangitis (PSC) and AIH. Particularly, PSC can be complicated by cholangiocarcinoma, gall bladder carcinoma and hepatocellular carcinoma.27 In contrast occurrence of HCC in patients with AIH is a rare event and develops only in long-standing cirrhosis.28

Autoantibodies

Autoantibodies are assessed in liver diseases mainly to diagnose autoimmune liver diseases, but may additionally be useful tools to monitor disease activity.29 The lack of standardization of autoimmune liver serology has led to a low sensitivity and specificity of these relevant serological tests. Accordingly, the IAIHG has established an international committee to define guidelines and develop procedures and reference standards for more reliable testing.30 ANA, SMA and LKM are pivotal components for the diagnosis of AIH and should be first tested in suspicious patients.2 Antibodies against liver-cytosol type 1 (LC-1), pANCA, SLA/LP and the asialoglycoprotein receptor may be helpful to extend the diagnosis of AIH in patients who present without these antibodies. Additionally, antibodies against cardiolipin, chromatin and Saccharomyces cerevisiae have been described in AIH, which, however, do not define patients with a distinctive clinical phenotype (Table 2).31–33

Table 2. Molecular Definitions of Autoantigens Relevant to the Serological Diagnosis of Autoimmune Hepatitis

Antibody	Target antigen in AIH	Liver disease	Associated with in AIH
ANA	Centromere, Ribonucleoproteins (+ -complexes)	AIH, PBC, PSC, Drug-induced hepatitis, HCV, HBV, NASH
SMA	Actin, Tubulin, vimentin, desmin, skeleton	Same as ANA	HLA A1-B8-DR3, younger patients, poorer prognosis
pANCA	Unknown	AIH, PSC
LKM-1	Cytochrome P450 2D6	AIH-2, HCV
LKM-2	Cytochrome P450 2C9	Ticrynafen-induced hepatitis
LKM-3	UGT1A	HDV, AIH-2, APECED, HCV
LM	Cytochrome P450 1A2	Dihydralazine-induced hepatitis
		APECED, HCV
	Cytochrome P450 2A6	APECED
LC-1	Formiminotransferase cyclo-deaminase	AIH-2, HCV	Disease activity, younger patients
SLA/LP	UGA repressor tRNA-associated protein	AIH, HCV	More severe course, relapse after drug withdrawal
ASGPR	Asialoglycoprotein receptor	AIH, PBC, Drug-induced hepatitis, HCV, HBV, HDV	Disease activity
Chromatin	Chromatin		Disease activity, relapse after drug withdrawal
CLA	Cardiolipin	AIH, HCV, HBV	Disease activity
ASCA	Saccharomyces cerevisiae	AIH, PSC, PBC

Antibodies to Nuclear Antigens (ANA)

ANA were the first autoantibodies to be associated with AIH and led Mackay to create the term “lupoid” hepatitis as early as 1956. ANA are the most non-specific marker of AIH and can also be found in PBC, PSC, viral hepatitis, drug related hepatitis, and alcoholic and non-alcoholic fatty liver disease.34, 35 Additionally, serum ANAs have been reported to occur in up to 15% of the general healthy populations from different countries predominantly among older age groups. In AIH, they are routinely determined by indirect immunofluorescence on a rodent multi-organ substrate panel that should include kidney, liver and stomach. HEp2 cells should no longer be used at the screening stage due to the high positivity rate in healthy controls. ANA are readily detectable as a nuclear staining with a homogenous or sometimes speckled pattern. The target antigens are heterogeneous and are incompletely defined in AIH. ANA have been found to be reactive with centromers, ribonucleoproteins, cyclin A, histones and many other antigens.36, 37 Subtyping of the various ANA specificities offers so far no diagnostic or prognostic advantage. No liver specific or liver disease specific ANA has been discovered so far.

Antibodies to Smooth Muscle Antigen (SMA)

SMA are frequently found in AIH, and are directed against components of the cytoskeleton such as actin and non-actin components including tubulin, vimentin, desmin, and skeletin.34 They frequently occur in high titers in association with ANA and are, like ANA, present in a variety of liver and non-liver diseases such as rheumatic diseases. A particular specificity against F-actin has been described for SMA in AIH.38 SMA autoantibodies are detected by indirect immunofluorescence on cryostat sections similar to ANA. The examination of the kidneys is of importance since this allows the visualization of the vessels, glomeruli and tubule pattern, which is more specific for AIH. Actin positive patients with SMA are reported to be more frequently positive for the HLA A1-B8-DR3 haplotype, to be younger at disease onset and have a poorer prognosis than actin-negative patients with SMA.38

Autoantibodies to Liver-Kidney Microsomal Antigens (LKM)

Autoantibodies against microsomal proteins form a heterogeneous group and are associated with several immune mediated diseases including AIH, drug induced hepatitis, the autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED or APS-1) syndrome, and chronic hepatitis C and D infection.34 Antibodies to LKM are regarded as serological markers of AIH type 2 and were first discovered in 1973 by indirect immunofluorescence and are reactive with the proximal renal tubule and hepatocellular cytoplasm.18, 39–41 Subclassification is achieved by ELISA and Western blot, preferably with recombinant antigens. The 50 kD antigen of LKM-1 was identified as the cytochrome mono-oxygenase P450 2D6 (CYP2D6). In AIH, LKM-1 recognizes the linear CYP2D6₁₉₆-₂₁₈, CYP2D6₂₅₄-₂₇₁, and CYP2D6₃₂₁-₃₅₁ sequences as immunodominant epitopes (Fig. 1).42–46 Conformational epitope mapping on CYP2D6 revealed that the C-terminal portion of the molecule accounts for all LKM-1 reactivity. Interestingly, the sequence between 316 and 327, which is most likely exposed on the surface of the molecule, appears to be are region capable of differentiating LKM-1 activity in AIH and HCV and may represent a key target for the autoantibody.47 The mechanisms of development and the pathogenic role of anti-LKM-1 antibodies in autoimmune hepatitis are however not yet resolved.34 Several recent studies shed more light on the initial suggestion that molecular mimicry might be involved in generation of LKM-1 antibodies.42 Antibodies directed against the CYP2D6 present in HCV and LKM-1 pos. patients can cross-react with HCV proteins (NS3, NS5a). The conformational epitope targeted by the autoantibodies of HCV and LKM-1 pos patients (aa 254-288) is located in the same region as the major linear epitope recognized by AIH-2 patients.48 This cross-reactivity could be the result of molecular mimicry at the B cell level, which through an initial reactivity against viral proteins would lead to reactivity against a self-protein. Interestingly another sequence on HCV from the E1 protein has also been identified to cross-react with the major B-cell autoepitope aa257-271 of Cyp2D6, but only in LKM1 positive HCV patients who possessed the HLA allotype B51 indicating that virus/self cross-reactivity is dependent on a specific immunogenetic background.49

Figure 1. Three-dimensional structure of CYP2D6: major and minor epitopes found in autoimmune hepatitis and hepatitis C (Figure modified from Sugimura et al.46).

Additionally, a murine model of AIH-2 was generated by immunization of mice against the human autoantigens CYP2D6 and against formiminotransferase cyclodeaminase.50 Molecular mimicry between foreign and self-antigens resulted in a liver disease closely resembling AIH-2 in terms of histological, biochemical and serological changes. In another animal model, Christen et al. used mice transgenic for human CYP2D6.51 These mice expressing mouse and human CYP2D6 developed a hepatitis similar to AIH following infection with an adenovirus expressing human CYP2D6. During the hepatitis, these mice additionally developed anti-CYP2D6 autoantibodies recognizing the major CYP2D6 epitope associated with autoimmune hepatitis.42

Anti-LKM-2 antibodies are directed against CYP2C9 and are induced in ticrynafen-associated hepatitis.34 The antihypertensive drug Ticrynafen however has been withdrawn from the market due to severe side-effects and LKM2 antibodies are therefore only of historical interest.52 Anti-LKM-3 antibodies are detected in about 10% of AIH-2 patients alone or in combination with LKM-1 antibodies. They are directed against family 1 uridine 5′-diphosphate glucuronosyltransferase (UGT1A), which belong to the drug metabolizing enzymes located in the endoplasmatic reticulum. These antibodies were first detected in patients with hepatitis D infection and later in patients with AIH and HCV infection (Table 2).53–56

Autoantibodies to Soluble Liver Antigen/Liver Pancreas Antigen (Anti-SLA/LP)

Anti-SLA/LP antibodies are detectable by radioimmunoassay and enzyme linked immunosorbent assays (ELISA) but cannot be detected by immunofluorenscence.34 Anti-SLA/LP were initially considered to be highly specific for AIH, where they are detectable in about 10% to 30% of all patients. Though anti-SLA/LP antibodies are occasionally found in patients with AIH who are negative for ANA, SMA or anti-LKM, they are also frequently present in typical cases of AIH-1 and -2.57–60 Seropositivity for anti SLA/LP is associated with DR3 and DRB1*0301 and seronegativity with DRB1*0401 in AIH patients.61 Recent studies with more sensitive serological tests revealed that they are also detectable in children with autoimmune cholangitis and in patients with HCV infection.62, 63 Interestingly, the prevalence of anti-SLA/LP antibodies in HCV patients increases when anti-LKM-1 antibodies are present. Screening of cDNA expression libraries identified a UGA tRNA suppressor as anti-SLA target autoantigen.60, 64 Epitope mapping revealed a dominant immune reactivity directed to peptide p395-414 of SLA/LP and a less prominent immune response to 2 other epitopes adjacent to the dominant epitope.65 The exact function and its role in autoimmunity are unclear. So far no evidence could be found for molecular mimicry being the causative mechanism for the development of SLA/LP autoantibodies.

Antibodies to Liver-Cytosol Type 1 (LC1)

Anti-LC1 antibodies are found in up to 50% of patients with AIH type 2.34 Less frequently, anti-LC1 may be associated with SMA and ANA in sera from patients with AIH type 1 and chronic hepatitis C infection.66–68 In addition, anti-LC1 proved to be the only serological marker in 10% of patients with AIH.68 Anti-LC1 are visualized by indirect immunofluorescence, however their characteristic staining may be masked be the more diffuse pattern of LKM-1 antibodies. Therefore other techniques such as ouchterlony double diffusion, immunoblot and counter-immunoelectrophoresis are also used for their detection. The antigen recognized by anti-LC1 was identified as formiminotransferase cyclodeaminase (FTCD).69 FTCD is a bi-functional enzyme involved in the folate metabolism, and is most highly expressed in the liver.70 Anti-LC1 sera recognize conformational epitopes throughout the protein but linear epitopes were found exclusively within the C-terminal 146 amino acids.71, 72 Contrary to most other autoantibodies in AIH, anti-LC1 seems to correlate with disease activity and may be useful as a marker of residual hepatocellular inflammation in AIH.73

Antibodies to the Asialoglycoprotein Receptor (ASGPR)

Antibodies to ASGPR are observed in up to 90% of all patients with AIH and can coexist with ANA, SMA and anti-LKM-1.34 However, they are not disease specific and can also be found in viral hepatitis, drug-induced hepatitis and PBC.74 The asialoglycoprotein receptor is a liver specific glycoprotein of the cell membrane. Its main function is the internalization of asialoglycoproteins by binding a terminal galactose residue to coated pits. Levels of anti-asialoglycoprotein antibodies correlate with inflammatory disease activity and might be used as additional marker to monitor treatment efficacy.75–77

Antibodies to Neutrophil Cytoplasmic Antigens (pANCA)

pANCA are detected in 65% to 95% of sera from patients with AIH type 1 and additionally in sera from patients with PSC and with viral hepatitis.34 pANCA are detected by immunofluorescence. The target antigen in AIH is unknown and its role in AIH is unclear, but determination might be useful to identify patients formerly classified as having cryptogenic hepatitis.

Subtypes of Autoimmune Hepatitis

Currently, it is a matter of debate whether AIH represents a heterogenous disease, which can be classified into distinct nosological entities; a question however, which is impossible to answer as long as the aetiology of AIH is not resolved. There have been several proposals to classify AIH according to different antibody profiles. The clinical utility of this classification is still uncertain and the IAHG regards such distinctions as too premature for routine use in everyday practice. For research purposes, a subdivision in three groups has been used for several years. According to this approach, AIH type 1 is characterized by the presence of ANA and/or anti-SMA antibodies. AIH type 2 is characterized by anti-LKM-1 and with lower frequency against LKM-3 antibodies (with or without ANA or SMA antibodies). AIH type 3 is characterized by autoantibodies against SLA/LP (with or without ANA or SMA antibodies). AIH type 1 represents the most common form of AIH, whereas AIH type 3 and particularly AIH type 2 are rare entities in adult patients.78, 79 AIH type 2 is more frequently seen in childhood, where it has been reported to represents up to 30% of all AIH patients. Clinically, the subgroups do not differ fundamentally, and AIH type 3 resembles AIH type 1 with respect to clinical characteristics and immunogenetic markers.59, 60, 80 However, several published studies confirmed the initial observation that patients with anti-SLA/LP antibodies display a more severe course of AIH.57, 58, 61, 63, 81, 82 Nevertheless, AIH subgroup “type 3” is most controversial and not recognized by the IAHG. The differences between AIH-1 and -2 are more prominent. AIH-2 displays a regionally variable prevalence with only 4% in the United States and up to 20% in Western Europe.18, 78 Patients with AIH-2 are younger at presentation, show a more severe course at onset and are more likely to progress to cirrhosis.4, 18, 78 Both entities of AIH are characterized by a high incidence of other organ specific immune mediated diseases, e.g., autoimmune thyroid disease and diabetes mellitus, which are not only detected in patients with AIH, but also in their first degree relatives.

Pathogenesis

The pathogenic mechanisms leading to liver cell destruction in AIH and to autoimmunity in general are as yet unknown despite extensive investigation into both humoral and cell-mediated aspects. Most of the common autoimmune diseases have an inherited element, and some autoimmune diseases thus cluster in families. Additionally, other environmental triggers and host factors must be relevant for the development of autoimmune diseases.83, 84 Historically, infectious agents have often been cited as triggers of autoimmune diseases and several viruses have been proposed to trigger AIH. However, in most cases there is still no firm evidence that a particular infectious agent is associated with the onset of disease. The difficult distinction is whether pathological immunity is truly directed against self and not against infectious non-self.

Genetic Susceptibility

The genetic background of AIH does not follow a Mendelian pattern and a conclusive role of a single genetic locus capable of explaining the etiology of AIH has not been identified. AIH is therefore considered as a complex trait like most other human diseases, which means that there are one or more genes acting alone or in concert to reduce or increase the risk of that trait. The heritable component of AIH is currently regarded as small. However, the absence of evidence does not mean evidence of absence and these data have yet to be established. Current problems and future strategies to tackle complex disease genetics, particularly in autoimmune liver diseases, have been recently summarized in several excellent reviews by P. Donaldson.85, 86 A key concept will be the use of high throughput genotyping assays with sufficient, well-characterized patients and adequately matched controls.

The most conclusive association with autoimmune hepatitis is related to the major histocompatibility complex alleles. The significance of HLA polymorphisms in terms of disease susceptibility and severity lies in the observation that most autoimmune diseases are T cell dependent and all T cell mediated responses are MHC restricted.87, 88 The MHC region is located on chromosome 6p21.3 and can be divided in 3 highly polymorphic regions. The MHC class I region encodes the HLA A, B, and Cw family and class II region encodes the HLA DR, DQ and DP family. The MHC class III region encodes many proteins of immunological interest such as tumor necrosis factor alpha and beta, the complement proteins C2, C4A, CAB and Bf, heat shock protein 70 and the MHC class I chain related proteins, Mic-alpha and Mic beta.

Among Caucasoid northern Europeans and North Americans two studies have identified a strong genetic association with HLA DRB1*0301 and the DRB*0401.89, 90 Table 3 summarizes confirmed associations of HLA DRB1 alleles in AIH patients from different populations. Based on these results, different models have been created whereby genetic susceptibility and resistance to AIH is best related to specific amino acid sequence motifs within DRB1 polypeptides. A lysine/arginine dimorphism at position 71 has been proposed for Northern European Caucasians and a valine/glycine dimorphism at position 86 for AIH patients from Argentina and Brazil. Japanese AIH patients were found to have a histidin on position 13 of the DRB1 alleles. These different models suggest the existence of different environmental factors triggering AIH in different populations. HLA alleles associated with AIH confer not only susceptibility towards AIH but also appear to influence the course of the disease. Most strikingly, patients with the DRB1*0301 were found to be younger at disease onset and have a higher frequency of treatment failure. Patients with HLA B8 were found to have a more severe disease and require a liver transplantation more frequently and patients with DRB1*0401-DRB4*0103 are at greater risk to develop additional autoimmune disorders.90–92 The presence of HLA DR3 is associated with a lower probability of reaching remission, more frequent relapses and need for transplantations.

Table 3. Association of HLA DRB1 Alleles With Autoimmune Hepatitis

HLA allele	Population	Adults/Children	# PAT.	# Contr.	Risk ratio	Ref.
DRB1*0301	Argentina	Children	122	208	3.0	24
DRB1*0301	North America, UK		297	236	3.39	6
DRB1*0301	India	Adults	20	113	3.79	159
DRB1*0404	Mexico	Mixed	30	175	7.71	160
DRB1*0405	Argentina	Adults	84	208	10.4	24
DRB1*0405	Japan	Adults	77	248	4.97	161
DRB1*1301	Argentina	Children	122	208	16.3	24
DRB1*1301	India	Adults	20	113	6.47	159
DRB1*1302	Argentina	Children	122	208	0.1	24
DRB1*14	India	Adults	20	113	3.25	159
DRB1*1501	North America, UK		297	236	0.52	6
DRB1*0301/DRB1*0301	North America, UK		297	236	7.61	6
DRB1*0301/DRB1*04	North America, UK		297	236	5.09	6

Genetic associations with MHC genes are not confined to the class I and II region, but include also genes of the MHC class III region. Of particular interest in AIH are SNPs of the TNF and CA genes.93–95 Both genes however exhibit strong linkage disequilibrium with the HLA 8.1 haplotype indicating that they might not be the primary susceptibility genes. This is further confirmed by findings in different population where the disease is not primarily associated with HLA-DRB1*03 and no association with TNFA*2 has been found.96

Besides the MHC susceptibility genes it is widely recognized that genes outside the MHC also contribute to the risk of autoimmunity. Of particular interest are genes, which are involved in the regulation of immune responses. In general, case-control studies have produced mixed results, with little consensus in most cases on whether any polymorphisms are actually associated with AIH. The most promising candidate in this regard was the cytotoxic T-lymphocyte antigen 4 (CTLA 4), which encodes an important negative regulatory molecule of the immune system and which has been linked to several autoimmune diseases, including AIH.97, 98 However, CTLA 4 gene polymorphisms do not confer susceptibility to all AIH patients in different geographic regions.99, 100 Finally, a number of other SNPs of various genes including cytokines, vitamin D receptor, CD45 and Fas receptor have been associated with susceptibility to AIH.101–106 Additional studies will be necessary to confirm these associations in different populations and to exactly link them to the pathogenesis of AIH.

Hepatitis in the Autoimmune Polyendocrinopathy-Candidiasis-Ectodermal Dystrophy (APECED)

Hepatitis has also been described in 10% to 20% of patients with APECED. APECED is a rare autosomal recessive disorder, which is characterized by an immune-mediated destruction of endocrine tissues, chronic candidiasis and additional ectodermal disorders.107 In contrast to many other autoimmune diseases, APECED is associated with mutations of a single gene, designated autoimmune regulator (AIRE) (Fig. 2).108 AIRE upregulates the transcription of certain organ-specific self-antigens in medullary thymic epithelial cells, and has a role in the negative selection of organ-specific thymocytes. The molecular mechanisms by which AIRE functions in these processes are still not well understood. So far, more than 50 different mutations of the AIRE gene have been identified and are distributed throughout the entire non-coding and coding region.109, 110 The variety of autoimmune diseases seen in APECED patients suggested that AIRE may contribute to the etiology of other autoimmune disorders. Recent studies however indicate that common mutations in the AIRE gene do not play a major role in autoimmune liver diseases, and are therefore a unique feature of APECED.111, 112 Similar to AIH-2, autoantibodies to CYP1A2 and CYP2A6 have been described as markers of an autoimmune liver disease in patients with APECED, but antibodies to tryptophan hydroxylase are best predictors for hepatitis in APECED.113–116

Figure 2. Structure of the autoimmune regulator (AIRE) which shows several characteristic motives of a transcription factor. (Figure from Kumar et al.¹⁶²)

Therapy

AIH is generally responsive to immunosuppressive treatment and it was the first chronic liver disease in which medical therapy improved survival. Corticosteroids are in most cases successfully used for treatment of AIH. Treatment should be instituted in all patients with a severe onset of diseases characterized by aminotransferase levels exceeding 5 to 10 times the upper normal level and or histological evidence of bridging or multilobular necrosis. Patients with liver failure or fulminant presentation who fail to improve under immunosuppressive therapy should be considered as candidates for liver transplantation.117 In patients without symptoms and only mild inflammation, the decision to treat must be weighted against the risk of medication and it is still a matter of debate whether all AIH patients require immunosuppressive therapy.118–120 A recent study revealed that the overall survival and survival to liver related endpoints of asymptomatic patients receiving no therapy was not different to that of the total cohort.3 Asymptomatic patients not receiving therapy require a close follow-up since 25% of this group developed subsequently symptoms.3 Treatment should generally follow defined therapeutic schedules, but in individual patients therapy is best tailored to the patient's presentation.

The goal of treatment is a complete biochemical and histological resolution of inflammation as well as the clinical remission of symptoms. This is achieved in approximately 70% to 80% of patients within the first 3 years.121, 122 Clinical and biochemical normalization usually occurs within 3-6 months, if the diagnosis is correct, but histological improvement may lag behind. About 9% percent of patients deteriorate despite compliance to treatment, 13% improve, but not to a degree to satisfy remission and 13% are intolerant to standard therapy.123, 124 In patients with histological proven normalization relapse after withdrawal of therapy occurs in about 80%. A sustained response after withdrawal of treatment is achieved in only 20% of patients. Antibodies to SLA/LP, asialoglycoprotein receptor and chromatin might be useful markers for relapse but this has to be confirmed. Treatment with prednisone monotherapy or in combination with azathioprine remains the standard (Tables 4 and 5).125 Both are equally effective and the decision for either strategy involves the consideration of patient profiles.126 If standard treatment fails or drug intolerance occurs, alternative therapies can be considered (Table 5). Several agents have emerged, especially from the transplantation setting, which might offer greater immunosuppression and which are better tolerated than prednisone and azathioprine. These include cyclosporine A, tacrolimus, cyclophosphamide, mercaptopurine, mycophenolate mofetil or deflazacort, which have been more or less successfully tested in several small studies with AIH patients (Table 4).127–133 Cyclosporine A and mycophenolate mofetil appear to be the most promising candidates out of these. Ursodeoxycholic acid (UDCA) is a hydrophilic bile acid with putative immunomodulatory capabilities. It is presumed to alter HLA class I antigen expression on cellular surfaces and to suppress immunoglobulin production.134 UDCA is a well-tolerated drug; however, its role in AIH therapy or in combination with immunosuppressive therapy is still unclear. Uncontrolled trials have shown a reduction in histological abnormalities, clinical and biochemical improvement in patients with mild disease but its use as adjuvant therapy in patients with severe disease has been disappointing.135–137

Table 4. Standard and Novel Therapies for Autoimmune Hepatitis

Standard Immunosuppressive Therapy	Alternative Immunosuppressive Therapies
Prednis(ol)one/or Azathioprine	Second Generation Corticosteroids
	Budenoside
	Deflazacort
	Calcineurin Inhibitors
	Cyclosporin A
	Tacrolimus
Non-immunosuppressive Therapy	mTOR Inhibitor
Ursodeoxycholic Acid	Rapamycin
	Antimetabolites
	Mycophenolate Mofetil
	Cyclophosphamide
	Methothrexate

Table 5. Approaches to Treatment of Autoimmune Hepatitis and Need for Alternative Drugs

Acknowledgements

The authors acknowledge helpful discussions with Christian Strassburg, Elmar Jaeckel, Tim Lankisch and Heike Bantel from our group at the Medical School of Hannover as well as Eric F. Johnson, Scripps Clinic and Research Foundation, La Jolla, CA (USA), and the members of the International Autoimmune Hepatitis Group (IAHG) for fruitful collaborations in recent years. Our own work was and is supported by the Deutsche Forschungsgemeinschaft, programmed projects SFB 311, SFB 244, SFB 280 and SFB 621 as well as individual grant to A.V., C.S., and H.B.