Diagnosis and management of autoimmune hepatitis

This guideline has been approved by the American Association for the Study of Liver Diseases (AASLD) and represents the position of the Association. Clinical practice guidelines are defined as “systematically developed statements to assist practitioner and patient decisions about appropriate heath care for specific clinical circumstances.”1 (All references are available in the Supporting Information.) These guidelines on autoimmune hepatitis provide a data-supported approach to the diagnosis and management of this disease. They are based on the following: (1) formal review and analysis of the recently-published world literature on the topic [Medline search]; (2) American College of Physicians Manual for Assessing Health Practices and Designing Practice Guidelines;2 (3) guideline policies, including the AASLD Policy on the Development and Use of Practice Guidelines and the American Gastroenterological Association Policy Statement on Guidelines;3 and (4) the experience of the authors in the specified topic. These recommendations, intended for use by physicians, suggest preferred approaches to the diagnostic, therapeutic and preventive aspects of care. They are intended to be flexible, in contrast to standards of care, which are inflexible policies to be followed in every case. Specific recommendations are based on relevant published information. To more fully characterize the quality of evidence supporting the recommendations, the Practice Guidelines Committee of the AASLD requires a class (reflecting benefit versus risk) and level (assessing strength or certainty) of evidence to be assigned and reported with each recommendation.4 The grading system applied to the recommendations has been adapted from the American College of Cardiology and the American Heart Association Practice Guidelines, and it is given below (Table 1). AASLD, American Association for the Study of Liver Diseases; AIH, autoimmune hepatitis; ALT, alanine aminotransferase; ANA, antinuclear antibody; AST, aspartate aminotransferase; CYP1A2, cytochrome P450 1A2; HCV, hepatitis C virus; IBD, inflammatory bowel disease; IgG, immunoglobulin G; LKM-1, liver/kidney microsome type 1; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; SMA, smooth muscle antibodies. Autoimmune hepatitis (AIH) is a generally unresolving inflammation of the liver of unknown cause. A working model for its pathogenesis postulates that environmental triggers, a failure of immune tolerance mechanisms, and a genetic predisposition collaborate to induce a T cell–mediated immune attack upon liver antigens, leading to a progressive necroinflammatory and fibrotic process in the liver.5,6 Onset is frequently insidious with nonspecific symptoms such as fatigue, jaundice, nausea, abdominal pain, and arthralgias at presentation,7 but the clinical spectrum is wide, ranging from an asymptomatic presentation8,9 to an acute severe disease.10,11 The diagnosis is based on histologic abnormalities, characteristic clinical and laboratory findings, abnormal levels of serum globulins, and the presence of one or more characteristic autoantibodies.12-16 Women are affected more frequently than men (sex ratio, 3.6:1).17-19 and the disease is seen in all ethnic groups20-34 and at all ages.21,35-44 There are no robust epidemiological data on AIH in the United States. In Norway and Sweden, the mean incidence is 1 to 2 per 100,000 persons per year, and its point prevalence is 11 to 17 per 100,000 persons per year.45,46 A similar incidence and prevalence can be assumed for the Caucasian population of North America. Data on the natural progression of untreated disease are derived principally from experiences published prior to the widespread use of immunosuppressive agents for AIH and before the detection of the hepatitis C virus (HCV).47-54 These studies showed that as many as 40% of patients with untreated severe disease died within 6 months of diagnosis,47,49 and that survivors frequently developed cirrhosis, esophageal varices and subsequent hemorrhage.47,49,50,55 An acute onset of illness is common (∼40%),56-63 and an acute severe presentation, characterized by hepatic encephalopathy within 8 weeks of clinical symptoms, is sometimes seen.10,11,58,64-68 Three randomized, controlled treatment trials established that prednisone alone or in combination with azathioprine improved symptoms, laboratory tests, histological findings, and immediate survival.48-50 These studies led to the acceptance of immunosuppressive regimens as the standard in treatment, and supported an autoimmune pathogenesis of the disease. However, these studies were completed decades ago before the discovery of HCV. Therefore, HCV infection could not be excluded in these studies and one can assume that several of these patients were indeed infected with HCV. Liver transplantation has also evolved as an effective treatment for the decompensated patient, and the 5-year patient and graft survivals now exceed 80%.69-74 The diagnostic criteria for AIH and a diagnostic scoring system were codified by an international panel in 199375 and revised in 199913 (Table 2). The clinical criteria for the diagnosis are sufficient to make or exclude definite or probable AIH in the majority of patients. The revised original scoring system was developed as a research tool by which to ensure the comparability of study populations in clinical trials (Table 3),13 and can also be applied in diagnostically challenging cases not readily captured by the descriptive criteria.13 The treatment response is graded in the revised original scoring system, and a score can be rendered both before and after treatment (Table 3).13 A pretreatment score of 10 points or higher, or a posttreatment score of 12 points or higher, indicate “probable” AIH at presentation. A pretreatment score of 10 points has a sensitivity of 100%, a specificity of 73%, and diagnostic accuracy of 67%.76 A pretreatment score of 15 points, indicative of “definite AIH” has a sensitivity of 95%, a specificity of 97%, and a diagnostic accuracy of 94%.76 A retrospective study supports the usefulness of the revised original system in children with AIH.77 A simplified scoring system has been proposed recently to ease clinical application78 and is based on the presence and level of autoantibody expression by indirect immunofluorescence, serum immunoglobulin G (IgG) concentration, compatible or typical histological features, and the absence of viral markers (Table 3).78 In three recent retrospective studies, the simplified scoring system performed with high sensitivity and specificity in the diagnosis of AIH, but it has yet to be validated in prospective studies.76,79,80 The diagnosis of AIH requires the presence of characteristic clinical and laboratory features, and the exclusion of other conditions that cause chronic hepatitis and cirrhosis (Table 2).13 The clinical assessment should include an evaluation of alcohol consumption and the use of drugs known to be hepatotoxic. The laboratory assessment should include determinations of the levels of serum alanine (ALT) or aspartate (AST) aminotransferases, alkaline phosphatase (AP), albumin, total or γ-globulin, IgG, and bilirubin (conjugated and unconjugated). AIH can be asymptomatic in 34%-45% of patients.8,9,269 Typically, these patients are men and have significantly lower serum ALT levels at presentation than do symptomatic patients.8 Histological findings, including the frequency of cirrhosis, are similar between asymptomatic patients and symptomatic patients. Because as many as 70% of asymptomatic patients become symptomatic during the course of their disease,8,9 asymptomatic patients must be followed life-long, preferably by an expert, to monitor for changes in disease activity. In children, the gamma glutamyl transferase level may be a better discriminator of biliary disease, specifically primary sclerosing cholangitis (PSC), than the AP level, which can be elevated due to bone activity in the growing child.77 Neither the gamma glutamyl transferase nor AP levels, however, discriminate between the presence or absence of cholangiopathy in children with AIH.36 The conventional serologic markers of AIH should also be assessed, including antinuclear antibody (ANA), smooth muscle antibody (SMA), antibody to liver/kidney microsome type 1 (anti-LKM1) and anti-liver cytosol type 1 (anti-LC1) (Table 4).12-16 Diagnostic evaluations should be undertaken to exclude hereditary diseases (Wilson disease and alpha 1 antitrypsin deficiency), viral hepatitis, steatohepatitis and other autoimmune liver diseases that may resemble AIH specifically primary biliary cirrhosis (PBC) and PSC.12,13,36,81,82 Liver biopsy examination at presentation is recommended to establish the diagnosis and to guide the treatment decision.12,13,15,16 In acute presentation unavailability of liver biopsy should not prevent from start of therapy. Interface hepatitis is the histological hallmark (Fig. 1), and plasma cell infiltration is typical (Fig. 2).83-87 Neither histological finding is specific for AIH, and the absence of plasma cells in the infiltrate does not preclude the diagnosis.84 Eosinophils, lobular inflammation, bridging necrosis, and multiacinar necrosis may be present.55,86,87 Granulomas rarely occur. The portal lesions generally spare the bile ducts. In all but the mildest forms, fibrosis is present and, with advanced disease, bridging fibrosis or cirrhosis is seen.55,83-85 Occasionally, centrizonal (zone 3) lesions exist (Fig. 3),10,60-62,88-91 and sequential liver tissue examinations have demonstrated transition of this pattern to interface hepatitis in some patients.62 The histological findings differ depending on the kinetics of the disease. Compared to patients with an insidious onset, patients with acute severe hepatic failure exhibit more interface and lobular hepatitis, lobular disarray, hepatocyte necrosis, central necrosis and submassive necrosis, but less fibrosis and cirrhosis.10,92,93 Interface hepatitis. The limiting plate of the portal tract is disrupted by a lymphoplasmacytic infiltrate. Hematoxylin and eosin stain; magnification, ×200. Plasma cell infiltration. Plasma cells, characterized by cytoplasmic halo about the nucleus, infiltrate the hepatic parenchyma. Hematoxylin and eosin stain; magnification, ×400. Median centrilobular zone 3 necrosis. Centrilobular zone 3 necrosis associated with a mononuclear inflammatory infiltrate. Hematoxylin and eosin stain; original magnification, ×200. Some patients exhibit features of both AIH and another disorder such as PSC, PBC, or autoimmune cholangitis, a variant syndrome.94-100 Certain histologic changes such as ductopenia or destructive cholangitis may indicate the presence of one of these variant types.101 In these cases, the revised original scoring system can be used to assist in diagnosis (Table 3).13,76 The findings of steatosis or iron overload may suggest alternative or additional diagnoses, such as nonalcoholic fatty liver disease, Wilson disease, chronic hepatitis C, drug toxicity, or hereditary hemochromatosis.84,85,101 Differences between a definite and probable diagnosis of AIH by the diagnostic scoring system relate mainly to the magnitude of serum IgG elevation, titers of autoantibodies, and extent of exposures to alcohol, medications, or infections that could cause liver injury.13,76,78 There is no time requirement to establish chronicity, and cholestatic clinical, laboratory, and histologic changes generally preclude the diagnosis. If the conventional autoantibodies are not detected, a probable diagnosis can be supported by the presence of other autoantibodies such as atypical perinuclear anti-neutrophil cytoplasmic antibody (atypical pANCA) or those directed against soluble liver antigen (anti-SLA).102,103 ANA, SMA, anti-LKM1, and anti-LC1 constitute the conventional serological repertoire for the diagnosis of AIH (Table 4).12-16,104-109 In North American adults, 96% of patients with AIH have ANA, SMA, or both,110 and 4% have anti-LKM1 and/or anti-LC1.111 Anti-LKM1 are deemed more frequent in European AIH patients and are typically unaccompanied by ANA or SMA.112 They are possibly underestimated in the United States.113 Anti-LKM1 are detected by indirect immunofluorescence, but because they may be confused with antimitochondrial antibody (AMA) using this technique, they can be assessed by measuring antibodies to cytochrome P4502D6, the major molecular target of anti-LKM1, using commercial enzyme-linked immunosorbent assays (ELISA). Autoantibodies are not specific to AIH104-109 and their expressions can vary during the course of the disease.110 Furthermore, low autoantibody titers do not exclude the diagnosis of AIH, nor do high titers (in the absence of other supportive findings) establish the diagnosis.110 Seronegative individuals may express conventional antibodies later in the disease114-118 or exhibit nonstandard autoantibodies.104-109,119 Autoantibody titers in adults only roughly correlate with disease severity, clinical course, and treatment response.110 In pediatric populations (patients aged ≤18 years), titers are useful biomarkers of disease activity and can be used to monitor treatment response.120 When tested on rodent tissues, an autoantibody titer of 1:40 is significant in adults, whereas in children titers of 1:20 for ANA and SMA, and 1:10 for anti-LKM1, are clinically relevant, because autoantibody reactivity is infrequent in healthy children.13 If present in high titer, anti-LKM1 strongly support the diagnosis of AIH, even if liver biopsy is precluded by other clinical considerations. The mainstay technique for autoantibody screening is indirect immunofluorescence on composite sections of freshly frozen rodent stomach, kidney and liver.14 This technique not only permits the detection of ANA, SMA, anti-LKM1, and AMA but also suggests the presence of other autoantibodies of an evolving clinical importance, such as antibody to liver cytosol type 1 (anti-LC1)111,121 and antibody to liver kidney microsome type 3 (anti LKM-3).122,123 Confirmation of the presence of the latter autoantibody is obtained with assays detecting antibodies to their molecular targets, formiminotransferase cyclo-deaminase (FTCD) and family 1 UDP-glucuronosyl-transferases (UGT1A), respectively (Table 4). Other autoantibodies that may be useful in classifying patients who lack the conventional serological findings are anti-SLA124-128 and atypical pANCA.119,129,130,131-139 Atypical pANCA, originally considered specific for PSC and inflammatory bowel disease (IBD),124,125 are frequently present in patients with AIH,126,127 and occasionally can be the only autoantibodies detected (Table 4).128 ANCA typically do not coexist with anti-LKM1.127 Recent evidence indicates that the target of atypical pANCA is located within the nuclear membrane. For this reason, a more suitable designation may be peripheral anti-neutrophil nuclear antibody (pANNA) (Table 4).102,103 Anti-SLA129 and anti–liver-pancreas (anti-LP),130 originally described as separate autoantibodies in AIH, were later found to target the same antigen and to represent a single serological entity. These antibodies are now referred to as anti-SLA or anti-SLA/LP. Their molecular target is a transfer ribonucleoprotein (Table 4).119,131,132 SLA has recently been renamed SEPSECS (Sep [O-phosphoserine] tRNA synthase) Selenocysteine Synthase. Anti-SLA are occasionally found in patients with AIH who are negative for ANA, SMA, and anti-LKM1,133 but are more commonly found in association with the conventional autoantibodies, especially if sensitive immunoassays are used.133-136 Anti-SLA are highly specific for the diagnosis of autoimmune liver disease,133 and their detection may identify patients with more severe disease and worse outcome.137-140 Commercial ELISAs are available for their detection. The conventional and nonstandard autoantibodies described in AIH are shown in Table 4. Figure 4 provides an algorithm for the use of autoantibodies in the diagnosis of AIH. The use of serological tests assisting in the diagnosis of AIH. Serological tests in the evaluation of acute or chronic hepatitis of undetermined cause. The initial serological battery includes assessments for antinuclear antibodies (ANA), smooth muscle antibodies (SMA), antibodies to liver/kidney microsome type 1 (LKM-1), and antimitochondrial antibodies (AMA). The results of these conventional tests then direct the diagnostic effort. If one or more tests are positive, the diagnosis of autoimmune hepatitis (AIH) or primary biliary cirrhosis (PBC) should be pursued. If these tests are negative, other serological assessments are appropriate, including tests for antibodies to actin (F-actin), soluble liver antigen/liver pancreas (SLA/LP), liver cytosol type 1 (LC-1), UDP-glucuronosyltransferases (LKM-3), the E2 subunits of the pyruvate dehydrogenase complex (PDH-E2), perinuclear anti-neutrophil cytoplasmic antibodies (pANCA). The results of these supplemental tests may suggest other diagnoses, including primary sclerosing cholangitis (PSC), or cryptogenic chronic hepatitis. Multiple genetic associations with AIH have been described in different ethnic groups.29,141-154 The primary genetic association is with the major histocompatibility complex locus, and associations of HLA alleles with disease predisposition, clinical phenotype, response to therapy, and outcome have been studied.18,155-168 AIH is a complex polygenic disorder169 unlikely to be transmitted to subsequent generations; thus, routine screening of patients or family members for genetic markers is not recommended. AIH may be present in patients with multiple endocrine organ failure, mucocutaneous candidiasis, and ectodermal dystrophy. Such patients have the rare genetic disorder autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED), caused by a single-gene mutation located on chromosome 21q22.3 that affects the generation of the autoimmune regulator (AIRE) protein.170 AIRE is a transcription factor expressed in epithelial and dendritic cells within the thymus that regulates clonal deletion of autoreactive T cells (i.e., negative selection). APECED has an autosomal recessive pattern of inheritance and lacks HLA DR associations and female predilection. The liver autoantigens associated with APECED are cytochrome P450 1A2 (CYP1A2), CYP2A6 in addition to CYP2D6.171-174 Antibodies to cytochrome P450 1A2 were previously called anti liver microsomal (anti-LM) antibodies (Table 4). This is the only syndrome involving AIH that exhibits a Mendelian pattern of inheritance, and genetic counseling for the patient and family members are warranted. Recommendations: 1. The diagnosis of AIH should be made when compatible clinical signs and symptoms, laboratory abnormalities (serum AST or ALT, and increased serum total IgG or γ-globulin), serological (ANA, SMA, anti-LKM 1, or anti-LC1), and histological (interface hepatitis) findings are present; and other conditions that can cause chronic hepatitis, including viral, hereditary, metabolic, cholestatic, and drug-induced diseases, have been excluded (Table 2). (Class I, Level B) 2. Diagnostically challenging cases that have few or atypical clinical, laboratory, serological or histological findings should be assessed by the diagnostic scoring systems (Table 3). (Class IIa, Level B) 3. In patients negative for conventional autoantibodies in whom AIH is suspected, other serological markers, including at least anti-SLA and atypical pANCA, should be tested. (Table 4; Fig. 4). (Class I, Level B) 4. In patients with AIH and multiple endocrine disorders, the APECED syndrome must be excluded by testing for the typical mutations in the AIRE gene. (Class I, Level C) Two types of AIH (type 1 and type 2) have been recognized based on serological markers112,129,130,175 but have not been established as valid clinical or pathological entities.13 A proposed third type (type 3) has been abandoned, as its serologic marker (anti-SLA) is also found in type 1 AIH and in type 2 AIH.176-179 Type 1 AIH is characterized by the presence of ANA, SMA or both, and constitutes 80% of AIH cases.175 Seventy percent of patients are female, with a peak incidence between ages 16 and 30 years.180,181 Fifty percent of patients are older than 30 years, and 23% are at least 60 years old.38,43,44,181 Associations with other autoimmune diseases are common (15%-34%); these include autoimmune thyroid disease, synovitis, celiac disease, and ulcerative colitis.43,44,182 At the time of diagnosis, cirrhosis is present in ∼25% of patients.183,184 Antibodies to SLA have emerged as possible prognostic markers that may identify patients with severe AIH who are prone to relapse after corticosteroid withdrawal.134,137-140,179,185 Type 2 AIH is characterized by the presence of anti-LKM1112 and/or anti-LC1 and/or anti-LKM-3. Most patients with type 2 AIH are children, and serum immunoglobulin levels are usually elevated except for the concentration of IgA, which may be reduced.112 Concurrent immune diseases are common,112 progression to cirrhosis occurs,112 and an acute severe presentation is possible.58,64 Recommendations: 5. Classification of autoimmune hepatitis into two types based on the presence of ANA and SMA (type 1 AIH) or anti-LKM1 and anti-LC1 (type 2 AIH) can be used to characterize the clinical syndrome or to indicate serological homogeneity in clinical investigations. Anti-LKM1 antibodies should be routinely investigated to avoid overlooking type 2 AIH. (Class IIa, Level C) PSC and PBC can have clinical, laboratory, histological, and genetic findings that resemble those of AIH,95,206-212 and AIH can have features that resemble each of these cholestatic syndromes.36,81,82,213-217 These nonspecific shared features can confound the codified diagnostic scoring system.13,76,78 The prevalence of AIH among patients with PSC was determined to be 21%-54% using the original scoring system,218,219 but this prevalence decreased to 8% in PSC when the revised original scoring system was applied.206,220,221 Application of the original scoring system in a retrospective review of 141 patients with PBC showed that 19% and 0% scored as probable and definite AIH, respectively.222 Clinical judgment is required to determine the predominant phenotype of the disease and to manage the process appropriately.95,223 AIH patients may demonstrate serological features that suggest another diagnosis. AMA occur in about 5% of AIH patients in the absence of other biliary features (“serological overlap”),178,224-228 and their presence may confound the clinical diagnosis. AMA may disappear226 or persist as long as 27 years without an evolution into PBC.227 The revised original scoring system can render a diagnosis of “probable AIH” in these patients, if other features of AIH are sufficiently strong.229,230 Other acute and chronic liver diseases of diverse etiologies that can have serological features of AIH include alcoholic231 and nonalcoholic fatty liver disease,232,233 acute234 and chronic54,235-241 viral hepatitis, and drug-induced hepatitis.242,243 Drugs such as minocycline,244-246 diclofenac,247,248 infliximab,249 propylthiouracil,250 atorvastatin,251 nitrofurantoin,252 methyl dopa,253 and isoniazid254 can cause a syndrome that resembles AIH replete with autoantibodies that generally disappear after discontinuation of the drug. Similarly, an AIH-like clinical syndrome has been associated with various herbal medications255-258 and with vaccination.259-261 Manifestations of AIH vary among ethnic groups. African-American patients have a greater frequency of cirrhosis at presentation than do white Americans.26,31,32 Alaskan natives exhibit a higher frequency of acute icteric disease than non-native counterparts,27 whereas Middle Eastern patients commonly have cholestatic features.28 Asian patients typically present with late onset, mild disease,20,262 whereas South American patients are commonly children with severe liver inflammation.21,22 Aboriginal North Americans have a disproportionately high frequency of immune-mediated disorders, cholestatic features, and advanced disease at presentation,33,34 and Somali patients are frequently men with rapidly progressive disease.30 Socioeconomic status, healthcare access, and quality of care are additional factors that must be considered when assessing nonclassical disease manifestations within racial groups.31,32,263,264 AIH can have an acute severe presentation that can be mistaken for a viral or toxic hepatitis.10,11,58,64,65,67,68,265 Sometimes autoimmune hepatitis may present as acute liver failure. Corticosteroid therapy can be effective in suppressing the inflammatory activity in 36%-100% of patients,11 whereas delay in treatment can have a strong negative impact on outcome.265-267 In addition, unrecognized chronic disease can exhibit a spontaneous exacerbation and appear acute.92 If extrahepatic endocrine autoimmune features are present in children with severe acute presentation the APECED syndrome must be excluded.268 Concurrent immune disorders may mask the underlying liver disease.16,17,38,43,44,182 Autoimmune thyroiditis, Graves' disease, synovitis and ulcerative colitis are the most common immune-mediated disorders associated with AIH in North American adults,43,44,180,270 whereas type I diabetes mellitus, vitiligo, and autoimmune thyroiditis are the most common concurrent disorders in European anti-LKM1+ AIH patients.112 In children with AIH, autoimmune sclerosing cholangitis can be present, with or without IBD.36 In adults with both AIH and IBD, contrast cholangiography showing biliary changes suggestive of PSC are present in 44% of patients.81 In adults with AIH but not IBD, magnetic resonance imaging indicating biliary changes are observed in 8% of patients.82 Unless bile duct changes are present, concurrent immune diseases typically do not affect the prognosis of AIH.81 Cholangiographic studies should be performed in patients with both AIH and IBD, as well as in children and adults refractory to 3 months of conventional corticosteroid treatment. In a prospective pediatric study, 50% of patients with clinical, serological and histological characteristics of AIH type 1 had bile duct abnormalities compatible with early sclerosing cholangitis on cholangiogram.36 Recommendations: 6. The diagnosis of AIH should be considered in all patients with acute or chronic hepatitis of undetermined cause, including patients with acute severe hepatitis. (Class I, Level C) 7. Cholangiographic studies should be considered to exclude PSC in adults if there has been no response to corticosteroid therapy after 3 months. (Class IIb, Level C) 8. All children with AIH and all adults with both AIH and IBD should undergo cholangiographic studies to exclude PSC. (Class I, Level C) Three randomized, controlled trials have demonstrated that patients with serum AST levels of at least 10-fold the upper limit of the normal range (ULN) or more than five-fold ULN in conjunction with a serum γ-globulin level more than two-fold ULN have a high mortality (60% at 6 month) if untreated. Furthermore, histological findings of bridging necrosis or multilobular necrosis at presentation progress to cirrhosis in 82% of untreated patients and are associated with a 5-year mortality of 45%.55,86,87 These laboratory and histological findings of disease severity at presentation are absolute indications for corticosteroid treatment (Tables 4 and 5).274,275 Incapacitating symptoms associated with hepatic inflammation, such as fatigue and arthralgia, are also absolute indications for treatment regardless of other indices of disease severity (Table 5). The natural history of autoimmune hepatitis is uncertain in patients who have no or only mild symptoms and in those who have mild laboratory and histological findings. Prospective, randomized, controlled treatment trials have not been performed in these patients, and their indications for treatment remain uncertain and highly individualized (Table 5).269,276 Asymptomatic individuals with inactive cirrhosis may have an excellent immediate survival without corticosteroid treatment.8,9 Other asymptomatic patients who do not have cirrhosis may have inactive disease, and their natural 10-year survival may exceed 80%.9 There are no guidelines that reliably identify this “safe” population who require no therapy. Spontaneous resolution is possible in some asymptomatic patients with mild disease, but these patients improve less commonly (12% versus 63%, P < 0.006) and more slowly than treated patients.269 Furthermore, untreated asymptomatic patients with mild disease have a lower 10-year survival than treated counterparts (67% versus 98%, P < 0.01).269 The frequency of spontaneous improvement must be counterbalanced against the frequency of serious drug-related complications when making the treatment decision (12% versus l4%).269 Since the mild autoimmune hepatitis can progress and a rapid and complete response to a normal end point can be anticipated, corticosteroid therapy is favored in asymptomatic mild disease, especially in young individuals who are likely to tolerate the medication satisfactorily.269 Patients likely to have a poor outcome are those at increased risk for drug intolerance, and they include individuals with advanced inactive cirrhosis, postmenopausal osteopenia or vertebral compression, emotional instability or psychosis, poorly controlled hypertension, low thiopurine methyltransferase activity, and brittle diabetes (Table 5).277 Corticosteroid therapy is effective only in patients who have clinical, laboratory or histological features of active liver inflammation. Patients with inactive or “burned out cirrhosis” cannot benefit from therapy,9 and they have an increased risk of drug-induced side effects bec