Abstract
Autoimmune hepatitis (AIH) is a progressive inflammatory disorder of unknown etiology, characterized histologically by interface hepatitis, serologically by the presence of non-organ specific autoantibodies, biochemically by elevated aminotransferases and serum IgG, and clinically by response to immunosuppressive treatment in the absence of other known causes of liver disease [1].
Introduction
Autoimmune hepatitis (AIH) is a progressive inflammatory disorder of unknown etiology, characterized histologically by interface hepatitis, serologically by the presence of non-organ specific autoantibodies, biochemically by elevated aminotransferases and serum IgG, and clinically by response to immunosuppressive treatment in the absence of other known causes of liver disease [1].
The spectrum of chronic inflammatory diseases of the liver extends from acute hepatitis to chronic hepatitis and finally to cirrhosis. In 1950, Dr. Waldenstrom described a form of chronic hepatitis occurring predominantly in young women with arthralgias, myalgia, hepatosplenomegaly, amenorrhea, skin rashes, fluctuating course, and invariably fatal outcome. The term “lupoid” hepatitis was coined after the detection of anti-nuclear antibodies (ANAs), the then positive test for lupus erythematosus in some of these individuals. The identification of anti-smooth muscle antibody (ASMA) in 1966 led to the nomenclature of “autoimmune chronic active hepatitis” for the first time, in order to distinguish it from systemic lupus erythematosus [2]. The discovery of hepatitis A and B viruses allowed hepatitis caused by these viruses to be excluded. Histologically, the term “chronic persistent hepatitis” was used when the mononuclear inflammation was limited to the portal tracts, while the term “chronic active hepatitis” was used to characterize infiltration of the adjacent hepatic parenchyma (interface hepatitis) [3]. Widespread acceptance of the autoimmune basis of this condition was accepted only after controlled trials demonstrated response to immunosuppression and a link with human leukocyte antigens (HLA) HLA-B8 and HLA-DR3 was established. The discovery of hepatitis C virus in 1989 led a panel of international experts, the International Autoimmune Hepatitis Group (IAIHG), to formulate several recommendations regarding the diagnosis and classification of AIH. The IAIHG developed a scoring system to weigh each clinical, laboratory, and histological finding at presentation as well as the response to corticosteroid therapy [4].
Epidemiology
Autoimmune hepatitis is regarded as an important paradigm for the study of autoimmunity and is an important etiology of chronic hepatitis affecting both children and adults. The incidence in children ranges from 0.23 to 0.4 per 100,000 and the prevalence ranges from 2.4 to 9.9 per 100,000 [5]. The incidence and prevalence are ~4–5 fold higher in adults. There is a female predominance, encompassing 60–75% of AIH cases in children and 70–95% in adults. The disease is seen in all races and ethnicities, with Alaskan natives presenting with a high frequency of icteric hepatitis and Hispanics with a high rate of cirrhosis. African Americans have a more progressive disease and African American children transplanted for AIH have a higher rate of recurrence compared to other races [6]. Autoimmune hepatitis is classified into subtypes based on the presence of circulating autoantibodies: type 1 is characterized by the presence of ANA and/or ASMA (or anti-actin antibody) and type 2 by the presence of anti-liver–kidney microsomal antibody (ALKM) or liver cytosol type 1 antibody (LC-1). The ratio of incidence of AIH type 1 and type 2 in North and South America and Japan is 6–7:1 while it is 1.5:1 in Europe and Canada.
Pathogenesis
Liver injury in AIH occurs through uncontrolled cellular (T cell) and humoral (B cell) autoimmune responses targeting hepatocytes. The trigger of this autoimmune phenomenon is not known and a leading hypothesis is that the hepatocyte injury is initiated by an infection or environmental toxin, followed by an abnormal autoimmune response in the genetically predisposed individual [7, 8]. The identification of molecular mimicry between virus and hepatocyte proteins adds credence to this theory [9].
Cellular Immunity
Autoimmune hepatitis occurs when there is loss of immune tolerance to hepatocyte proteins (or antigens). The inflammatory milieu within the diseased liver is composed of excessive T cells, scattered macrophages, and plasma cells. The subsets of CD4 helper and CD8 cytotoxic T cells are expanded and damage hepatocytes through production of proinflammatory cytokines (e.g., interferon-γ and tumor necrosis factor-α) or through granzyme-perforin mechanisms, respectively. Subsequent activation of effector cells such as macrophages and natural killer cells contributes to the ongoing hepatocyte injury. Autoreactive T cells specific to a limited number of “self” proteins have been identified and are thought to perpetuate the ongoing injury. Asialoglycoprotein receptor (ASGP-R) is expressed only in hepatocytes within the liver and is believed to be a major pathogenic antigen targeted by autoreactive T cells. Other hepatocyte antigens that have been associated with activation of autoreactive T cells include liver cytosolic antigen (formiminotransferase cyclodeaminase) and soluble liver antigen [7–9].
The ability of autoreactive T cells to proliferate is controlled by the regulatory T cell (Treg). Diminished quantity and function of Tregs in patients with AIH have recently been described and this allows autoreactive T cells to thrive and activate inflammatory responses. Quantitative deficiencies in Tregs inversely correlate with autoantibody titers in children with AIH. In addition, functional deficiencies of Tregs lead to the decreased ability to inhibit T cell proliferation and interferon-γ production, resulting in increased hepatocyte injury [10].
Humoral Immunity
Activation of B cells and the subsequent generation of autoantibodies from plasma cells aid in defining AIH. A complete list of autoantibodies and the target autoantigens that have been discovered in AIH are outlined in Table 20.1. Autoantibodies function mainly as biomarkers of disease, but in limited investigations have also been shown to contribute to antibody-dependent cell-mediated cytotoxicity of hepatocytes. Type 1 AIH is associated with positive ANA and/or ASMA titers. However, ANA is the most non-specific autoantibody biomarker of AIH and can be positive in up to 15% of healthy people without autoimmune disease. In order to increase specificity of this ANA for AIH, serum titers >1:80 are often used to define positivity. Because ASMA targets include many proteins within the cytoskeleton, like ANA, it is not specific to AIH and is found in other autoimmune diseases. In AIH type 1, ANA and ASMA are both positive in approximately 50% of patients; ANA only positivity occurs in 10–15% and ASMA only positivity in 30–35%. Type 2 AIH is associated with anti-LKM-1, which is specific to the cytochrome P450 IID6 (CYP2D6) protein within the cytoplasm of hepatocytes and in proximal renal tubules [11]. Approximately 50% of patients with AIH type 2 will also be positive for LC-1 and in 10% of patients with type 2 AIH, only LC-1 is detectable. The target of LC-1 is formiminotransferase cyclodeaminase, an enzyme involved in folate metabolism that is mainly expressed in the liver. Levels of LC-1 correlate with disease severity (degree of hepatocyte injury, rapid progression to cirrhosis) and have been associated with other concurrent autoimmune diseases within patients. An autoantibody that may be present in both types of AIH is anti-SLA/LP. The target of the anti-SLA/LP antibodies is a UGA serine transfer RNA-associated protein. Anti-SLA is present in ~ 20% of AIH type 1 and 50% of AIH type 2 cases. If anti-SLA is present, it predicts a more severe course of disease [11–13]. Both types of AIH commonly have ASGP-R (~75% of type 1 and ~40% of type 2) and this autoantibody may be the only detectable one found in patients who are negative for the conventional autoantibodies. Finally, perinuclear anti-neutrophil cytoplasmic antibodies (pANCA) have been identified in the majority of patients with AIH type 1 and are a heterogeneous group of antibodies that recognize a multitude of cytoplasmic proteins within neutrophils [13].
Autoantibody | Autoantigen(s) |
---|---|
Anti-nuclear antibody (ANA) | Single-stranded/double-stranded DNA, histones, centromere |
Anti-smooth muscle (or actin) antibody (ASMA) | Actin, vimentin, desmin |
Anti-liver–kidney microsomal antibody (LKM) | Cytochrome P450 IID6 (CYP2D6) |
Anti-soluble liver antigen/liver–pancreas antigen (anti-SLA) | UGA serine transfer RNA-associated protein |
Liver cytosol type 1 antibody (LC-1) | Formiminotransferase cyclodeaminase |
Perinuclear anti-neutrophil cytoplasmic antibody (pANCA) | Nuclear lamina proteins |
Asialoglycoprotein receptor antibody (ASGP-R) | Asialoglycoprotein receptor |
Perinuclear anti-neutrophil cytoplasmic antibodies have low specificity for AIH and are commonly found in primary sclerosing cholangitis and viral hepatitis. Therefore, pANCA is more helpful in the setting of autoimmune overlap with primary sclerosing cholangitis rather than in the first-line diagnostic work-up of AIH. The rate of autoantibody-negative AIH is ~10%, assuming that ANA, ASMA/anti-actin, ALKM, anti-SLA, and anti-LC-1 have been tested. Autoantibody levels vary during the course of disease and if the initial autoantibody screen is negative, repeat testing may allow for correct subtype classification.
Genetic Predisposition
The strongest association between genes and autoimmune diseases has been found within the HLA locus on the short arm of chromosome 6. Professional antigen-presenting cells (macrophages, dendritic cells, B cells) contain HLA class II proteins that form a complex with autoantigen(s) and present antigen to naive CD4 T cells, leading to activation and differentiation of autoreactive T cells. Certain HLA types are associated with AIH and increase the risk of developing disease. In the USA and Europe, the majority of adults with AIH type 1 express either HLA-DRB1*0301 (encodes HLA-DR3) or HLA-DRB1*0401 (HLA-DR4), or both. Susceptibility to AIH type 2 is associated with expression of HLA-DRB1*0701 (HLA-DR7) and HLA-DRB1*0301. Furthermore, HLA-DRB1*0701 expression in AIH type 2 has been associated with a more severe form of disease [14]. Ethnicity does appear to play a role in AIH and with increased globalization, there have been increasing numbers of reports of different genetic associations in non-Caucasoid populations which may account for variation in the clinical characteristics and outcomes [15].
Animal Models of Autoimmune Hepatitis
Animal models of human diseases can be powerful tools in aiding understanding of disease pathogenesis and response to novel therapies. Various models have been employed to study AIH but to date no one model recapitulates all aspects of the human disease. Mouse models that affect hepatic tolerance include targeted gene deficiency (knockout mice) or autoantigen immunization strategies. Mice that are deficient in programmed death-1, a key molecule for maintenance of tolerance, and that were neonatally thymectomized to deplete Tregs, develop a severe hepatitis that mimics human AIH [16]. The hepatitis is associated with induction of ANAs, hepatic T cell infiltrates, and lobular necrosis. This model demonstrated the significance of Tregs in disease, as replenishment of Tregs into these thymectomized mice rescued them from severe hepatitis. In another model, immunizing naive mice with supernatants from liver homogenate (autoantigen pool) resulted in experimental hepatitis that was strain dependent, suggesting genetic predisposition to disease. Adoptive transfer of activated T cells from the diseased mice into naive recipients resulted in initiation of experimental hepatitis, suggesting that effector T cells play a direct pathogenic role in liver injury [16].
Based on the fact that a target autoantigen has been identified for AIH type 2, many investigators have focused on this type of AIH in order to study disease pathogenesis. One model entails immunizing mice with plasmids containing human CYP2D6 (target of anti-LKM-1), human FTCD (target of anti-LC-1), mouse Ctla4 (encoding cytotoxic T-lymphocyte antigen 4, which facilitates antigen uptake by antigen-presenting cells), and Il12 (encoding interleukin-12, a Th1-skewing proinflammatory cytokine) [17]. When autoantigens and interleukin-12 were used to break tolerance, antigen-specific autoantibodies were detected, serum aminotransferases were mildly elevated, and an interface hepatitis composed of CD4+ and CD8+ T cells and scattered plasma cells was observed. Another model of AIH type 2 used transgenic mice that express human CYP2D6 on hepatocytes. Tolerance was broken with administration of an adenovirus-CYP2D6 vector and mice developed focal hepatocyte necrosis and a chronic hepatitis with fibrosis. The hepatic lesion was associated with a specific immune response to an immunodominant region of CYP2D6 and a cytotoxic T cell response to adenovirus-CYP2D6-infected target cells [18]. All of the models described above provide information on key players in the autoimmune response, but none demonstrated the chronic remitting–relapsing nature of human AIH.
Diagnosis and Scoring System
The diagnosis of AIH is based on the findings of elevated aminotransferases and serum IgG, a positive autoantibody, liver histology consistent with AIH (i.e., interface hepatitis), and a response to immunosuppressive treatment, in the absence of other known causes of liver disease. The diagnosis of AIH can be challenging, as there is no single pathognomonic test. Early diagnosis and appropriate management is important based on the findings of high mortality when AIH goes untreated. A significant proportion of “cryptogenic cirrhosis” and acute liver failure of indeterminate origin are thought to be secondary to AIH. A diagnostic scoring system was created by the IAIHG in 1993, revised in 1999 and a simplified scoring system was established in 2008 (Table 20.2) [19]. The diagnostic scoring systems are currently used as a research tool by which to ensure the comparability of study populations in clinical trials and can also be applied in diagnostically challenging cases not readily captured by the descriptive criteria. In children, a meta-analysis of four studies pertaining to the accuracy of the simplified criteria in the diagnosis of AIH revealed a sensitivity of 77% and a specificity of 95% [20]. There were 17% false negatives associated with seronegative AIH. Limitations to the scoring system include the inaccuracy of scoring in the setting of either concurrent primary sclerosing cholangitis (PSC), recurrent AIH post-transplantation, fulminant liver failure, and AIH with non-alcoholic fatty liver disease.
Parameter | Score |
---|---|
Autoantibodya | |
ANA or ASMA (anti-actin antibody) ≥1:40 | 1 |
ANA or ASMA (anti-actin antibody) ≥1:80 | 2 |
LKM antibody ≥1:40 | 2 |
Positive Anti-SLA | 2 |
Total serum IgG | |
>ULN | 1 |
≥ 1.1× ULN | 2 |
Liver histology | |
Compatible with autoimmune hepatitis: lymphocytic infiltrates, chronic hepatitis | 1 |
Typical of autoimmune hepatitis:b interface hepatitis (portal tract lymphocytes and plasma cells infiltrating into parenchyma); emperipolesis;c hepatic rosette formation | 2 |
Viral hepatitis absent | 2 |
Pretreatment overall score | |
Definite diagnosis | ≥7 |
Probable diagnosis | ≥6 |
ULN: upper limit of normal.
a Addition of points for all autoantibodies, maximum 2 points.
b Must have all three features to be considered “typical.”
c Emperipolesis is active penetration by one cell into and through a larger cell.
Clinical Features
Children with AIH often present before puberty and those with AIH type 2 usually present at a very young age. There is a variable mode of presentation, ranging from being asymptomatic and diagnosed after an incidental finding of elevated aminotransferases, to a presentation with fulminant liver failure. In more than one-third of children, presentation can be indistinguishable from that of an acute viral hepatitis, with non-specific symptoms of malaise, anorexia, nausea, vomiting, or abdominal pain, followed by the onset of jaundice. In another third, onset may be insidious, with progressive fatigue, weight loss, and intermittent jaundice lasting for several weeks or years before diagnosis. In ~10% of patients, complications of portal hypertension including splenomegaly, ascites, or variceal bleeding are the presenting symptoms. A small percentage may present with acute liver failure, with an international normalized ratio (INR) >2 despite vitamin K therapy, with or without encephalopathy. Children with AIH type 1 are more likely to present with cirrhosis while those with AIH type 2 present more frequently with fulminant liver failure. The course of the disease is relapsing with flares and spontaneous remissions, often resulting in a delayed or missed diagnosis.
When obtaining the history, it is important to inquire about medications the child has been taking in the past year, particularly minocycline, which is a known trigger of an “autoimmune-like hepatitis” [21]. Children with AIH may have other autoimmune disorders including celiac disease, thyroiditis, vitiligo, type 1 diabetes, juvenile inflammatory arthritis and inflammatory bowel disease. It is essential to inquire about symptoms and observe for signs related to these associated autoimmune diseases. In addition, there is a family history of autoimmune disease in up to 40% of patients with AIH. On physical examination, children may be jaundiced with an enlarged liver or spleen. In those with decompensated disease, there may be ascites and petechiae. Laboratory tests may show evidence of hypersplenism with a low white cell count and thrombocytopenia.
Biochemically, aminotransferases are elevated with or without elevated serum gamma-glutamyl transferase (GGTP) and total protein is high with a low/normal serum albumin. Serum IgG is typically elevated and there are increased titers of ANA, ASMA, or anti-LKM. It is essential to obtain a liver biopsy to establish a diagnosis of AIH. The typical histological picture is interface hepatitis: a dense infiltration of the portal tracts consisting mainly of lymphocytes and plasma cells that extends into the liver lobules with destruction of the hepatocytes at the periphery of the lobule and erosion of the limiting plate. There may be bridging fibrosis or cirrhosis evident on trichrome stain (Figure 20.1).
Figure 20.1 Characteristic histological features of autoimmune hepatitis. (A) Interface hepatitis with portal tracts containing lymphoplasmocytic infiltrates that extend into the lobule. (B) Interface hepatitis at higher power, showing the penetration of inflammatory cells from the portal area into the periportal parenchyma (arrows denote plasma cells). The hepatocytes show cellular swelling and degenerative changes. (C) Bridging necrosis in higher grade autoimmune hepatitis with inflammatory cells infiltrating a band of confluent necrosis. (D) Fibrosis in autoimmune hepatitis with interportal bridging fibrosis (portal tracts indicated by arrows). (a–c, hematoxylin & eosin stain; d, Trichrome stain.)
In children with persistently elevated serum GGTP, biliary damage or sclerosis on histology, or non-responsiveness to immunosuppression, it is advisable to perform a magnetic resonance cholangiopancreatography (MRCP) to look for changes consistent with PSC. There may be overlap between AIH and PSC, which has been designated as autoimmune sclerosing cholangitis or AIH-PSC overlap syndrome [22]. In this setting there are characteristic cholangiographic features of PSC in combination with elevated IgG, positive autoantibodies, and interface hepatitis on liver biopsy. Although the relationship between AIH and PSC is not established, one may speculate that there is a spectrum ranging from “cholestatic” disease affecting the biliary system alone (PSC) to “inflammatory” disease affecting the hepatocytes (AIH). Differences between AIH types 1 and 2 and AIH/SC overlap syndrome are illustrated in Table 20.3.
Feature | Autoimmune Hepatitis | Overlap Syndromea | |
---|---|---|---|
Type 1 | Type 2 | ||
Median age (years) | 11 | 7 | 12 |
Females (%) | 75 | 75 | 55 |
Mode of presentation (%) | |||
Acute hepatitis | 47 | 40 | 37 |
Acute liver failure | 3 | 25 | 0 |
Insidious onset | 38 | 25 | 37 |
Complication of chronic liver disease | 12 | 10 | 26 |
Associated immune disease (%) | |||
Present | 22 | 20 | 48 |
Inflammatory bowel disease | 20 | 12 | 44 |
Family history of autoimmune disease | 43 | 40 | 37 |
Autoantibodies (%) | |||
Anti-nuclear antibody or anti-smooth muscle antibody | 100 | 25 | 96 |
Anti-liver–kidney microsomal antibody-1 | 0 | 100 | 4 |
Perinuclear anti-neutrophil cytoplasmic antibody | 45 | 11 | 74 |
Anti-soluble liver antigenb | 58 | 58 | 41 |
Blood proteins (%) | |||
Increased IgG | 84 | 75 | 89 |
Partial IgA deficiency | 9 | 45 | 5 |
Low C4 level | 89 | 83 | 70 |
HLA alleles, increased frequency (%) | |||
HLA-DR*0301 | Yes | Noc | No |
HLA-DR*0701 | No | Yes | No |
HLA-DR*1301 | No | No | Yes |
Liver/biliary features (%) | |||
Abnormal cholangiogram | 0 | 0 | 100 |
Interface hepatitis | 66 | 72 | 35 |
Intrahepatic biliary features | 28 | 6 | 31 |
Cirrhosis | 69 | 38 | 15 |
Remission after immunosuppression (%) | 97 | 87 | 89 |
a AIH/PSC: also known as autoimmune sclerosing cholangitis.
b Measured by radioligand assay.
c But increased in HLA-DR*0701-negative patients.
Differential Diagnosis
The differential diagnosis of chronic hepatitis in children is broad, including but not limited to AIH, AIH-PSC overlap syndrome, PSC, chronic infectious hepatitis (hepatitis B, hepatitis C), α1-antitrypsin deficiency, Wilson disease, drug-induced liver injury, celiac disease and non-alcoholic fatty liver disease. Diagnostic work-up in order to exclude these non-AIH diseases includes hepatitis B surface antigen, hepatitis C antibody, α1-antitrypsin level and phenotype, ceruloplasmin, and liver ultrasound. In addition, MRCP (or endoscopic retrograde cholangiopancreatography) may be warranted if there is concern for PSC with or without AIH.
As mentioned above, it is important to obtain a history of recent exposure to medications such as minocycline or nitrofurantoin, as these and other medications have been proposed to trigger AIH-like liver injury. Mechanisms to explain this association involve covalent binding of a drug metabolite to a hepatocyte protein with resultant formation of a “neoantigen” that is recognized as foreign, eliciting an autoimmune response [21]. Minocycline-induced AIH-like liver injury usually occurs acutely within two years after drug initiation, with a range of three days to six years. Symptoms include jaundice, lethargy, anorexia, and abdominal discomfort. In addition, signs of serum sickness (fever, rash, joint pains) may co-exist. Over 90% of patients will be ANA positive with hypergammaglobulinemia, while only 25% will be ASMA positive, and ALKM positivity has not been described. Histologically, minocycline-induced autoimmune-like hepatitis mimics classical AIH, with the exception that cirrhosis at presentation has not been reported. The mainstay of treatment includes stopping the offending medication and administering corticosteroids. The overwhelming majority of patients can be weaned off immunosuppression and relapses are rare. Inhibitors of tumor necrosis factor (infliximab, adalimumab) and immune checkpoint inhibitors are now included in the list of medications and biologic agents that can cause AIH-like liver injury [23, 24].
Autoimmune hepatitis may be associated with another underlying autoimmune disease or immunodeficiency state. Other autoimmune diseases that have been reported concurrently with AIH include celiac disease, thyroiditis, vitiligo, insulin-dependent type 1 diabetes, inflammatory bowel disease and vasculitis. The autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) syndrome is caused by mutations of AIRE, encoding the autoimmune regulator. Patients with APECED syndrome present in the first decade of life with at least two of the following conditions: recurrent candidiasis, polyendocrinopathy (hypoparathyroidism, adrenal insufficiency), AIH, or dystrophy of dental enamel and nails. Up to 20% of these patients have been shown to have AIH in association with autoantibodies against cytochromes CYP1A2 and CYP2A6 [25]. This syndrome is the only one involving AIH that has an autosomal recessive pattern of inheritance. Other immune-dysregulated states associated with AIH include immunodysregulation polyendocrinopathy enteropathy X-linked (IPEX) syndrome, common variable immunodeficiency, and hyper-IgM syndrome. Common variable immunodeficiency is associated with hypogammaglobulinemia and the liver manifestation of nodular regenerative hyperplasia. Hyper-IgM syndrome is caused by mutations in genes that affect immunoglobulin class switch recombination and is associated with elevated IgM and low IgG levels. The X-linked hyper-IgM syndrome, or CD40L deficiency, is the most common and most severe form of hyper-IgM syndrome. This ligand is expressed on activated T-lymphocytes and is necessary for T cells to induce immunoglobulin isotype switching in B cells. Typically, male infants present with recurrent bacterial and opportunistic infections. The most common liver disease found in hyper-IgM syndrome is secondary sclerosing cholangitis with chronic Cryptosporidia biliary infection; however, AIH and liver and biliary tract carcinomas have also been reported [26]. Finally, types 1 and 2 AIH may have a genetically determined isolated deficiency of complement component C4, and a partial IgA deficiency has been noted in patients with type 2 AIH. The initial evaluation of AIH should, therefore, also include a complete blood cell count with differential and levels of all serum immunoglobulins (IgG, IgM, IgA).