Autoimmune and Chronic Cholestatic Disorders of the Liver

Clinical Features and Epidemiology

Although there are significant regional differences, the worldwide incidence of AIH appears to be increasing. A nationwide registry-based cohort study performed in Denmark between 1994 and 2012 showed an AIH incidence of 1.68 per 100,000 population per year, though it doubled over the course of the study period. In Japan, the point prevalence of AIH increased from 8.7 per 100,000 in 2004 to 23.9 per 100,000 in 2016. A similar registry-based cohort study in Finland reported an incidence of 1.1 per 100,000 person-years and showed an overall point prevalence of 14.3 per 100,000 at the end of 2015. In this work, 76% of AIH cases occurred in women, which is similar to the proportion described in other studies. A population-based study from Korea showed a similar incidence rate of 1.07 per 100,000 persons, or a prevalence of 4.82 per 100,000 persons, while others performed in Alaskan Natives showed a point prevalence of 42.9 per 100,000, again highlighting marked regional variation in the incidence of AIH. ^,

Patients of any age may be affected by AIH. For AIH type 1, there appears to be an incidence peak at around 70 years of age, along with a smaller peak in young adults between 10 and 25 years of age. Overall, AIH is rare in children, with an incidence of only 0.23 per 100,000. Nevertheless, in the juvenile population AIH type 1 is the most common form of AIH and is diagnosed 5.5 times more frequently than AIH type 2. There are important differences between AIH type 1 and AIH type 2 with regard to patient demographics and clinical course. AIH type 2 patients with LKM1 antibodies are more likely to present at a younger age (median age 7.4 years) and frequently exhibit more aggressive, severely active disease than those with classical AIH type 1. It is also extremely rare in adults in the United States, but fairly common in southern Europe and the United Kingdom, likely reflecting differences in genetic susceptibility as a result of regional variations in major histocompatibility complex (MHC) hapoltypes. ^,

Though not performed routinely for diagnostic purposes, AIH is strongly associated with certain HLA types. In northern European and North American populations HLA DRB1∗0301 (DR3) appears to be the primary susceptibility haplotype, with DRB1∗0401 (DR4) also showing a strong association with AIH type 1. In a large Dutch cohort, 75% of AIH patients were HLA DR3 and DR4 positive. These patients had significantly higher serum IgG levels than those with other HLA haplotypes, and patients with HLA DR3 were more likely to require liver transplantation. In addition to DR3, DRB1∗1301 is frequently associated with AIH in some South American populations, while association with DRB1∗0404 and DRB1∗0405 has been reported in Mexican and Japanese populations. ^, A single-nucleotide polymorphism in the 3’ untranslated region of HLA-DPB1 (rs9277534), which has been associated with the clearance of hepatitis B viral infection as well as the development of acute graft-versus-host disease following hematopoietic stem cell transplantation, may also contribute to AIH susceptibility, at least in Japanese populations.

AIH type 2 is associated with somewhat different HLA haplotypes than AIH type 1. Like AIH type 1, HLA DR3 is significantly associated with susceptibility to AIH type 2 in one cohort (relative risk 4.25). However, the HLA DQB1∗0201 allele appears to be the primary genetic determinant of AIH type, conferring the highest relative risk (6.40) of the HLA DR and DQ phenotypes in this study.

Common polymorphisms in other non-MHC genes, including those coding for TNF-α and CTLA-4, have also been implicated as risk factors for the development of AIH. The CTLA-4 gene includes a single base-exchange polymorphism at position 49 in exon 1. The G allele at this position is more common in patients with AIH type 1 than in controls, and the GG genotype may be associated with higher mean serum transaminase levels in patients with AIH. Though not universal, this association appears to hold true across multiple ethnic backgrounds and may be related to deficient CTLA-4–mediated inhibition of T-lymphocyte proliferation in patients with the GG genotype. ^,

Autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) is a monogenic disorder caused by biallelic point mutations in the autoimmune regulator gene AIRE that is characterized by chronic mucocutaneous candidiasis, hypoparathyroidism, and adrenal insufficiency related to loss of self-tolerance to endocrine organ tissue-specific antigens. Historically, APECED-associated hepatitis (APAH) was thought to be a relatively rare manifestation of this disorder. However, a recent study identified APAH in 42% of the APECED patients in their cohort, with about one-third showing histological and biochemical evidence of liver injury before exhibiting other features of systemic immune dysfunction. In contrast with other forms of AIH, antinuclear and anti-smooth muscle antibodies are rare in the setting of APAH. Rather, liver microsomal antibodies (LMs) targeting CYP1A2 appear to be highly specific for APAH, but lack sensitivity, somewhat limiting their diagnostic utility. Despite difference with classic AIH, the majority of patients with APAH respond to typical immunosuppressive induction and maintenance regimens.

The clinical presentation of AIH is remarkably protean and ranges from completely asymptomatic to fulminant hepatic failure. Approximately 25% of patients present with nonspecific signs and symptoms that resemble acute viral hepatitis and include fatigue, abdominal pain, nausea, fever, and jaundiced mucus membranes and skin. During periods of activity, serum transaminases are frequently markedly elevated and may exceed 1000 IU/mL. In contrast, serum alkaline phosphatase levels usually fall within the normal range or are only mildly elevated. Hypergammaglobulinemia and elevated serum IgG are frequently present and are considered to be defining features of AIH, though this is not universal, even in cases presenting with fulminant hepatitis.

Acute severe AIH resulting in massive hepatic necrosis and acute liver failure associated with coagulopathy and encephalopathy is a relatively rare manifestation of AIH. An immune-mediated process should, however, be considered in the differential diagnosis of all patients presenting with fulminant hepatic failure, as a dramatic response to immunosuppressive medical therapy is possible, and delay in the administration of corticosteroids has been associated with poor outcomes. Histological features alone are, unfortunately, not sufficiently specific to diagnose this entity, though the presence of prominent zone 3 necroinflammatory activity and a plasma cell–rich inflammatory infiltrate is said to be characteristic. ^, Correlation of these histological features with the presence of circulating autoantibodies along with careful exclusion of infectious, toxic, and metabolic etiologies may improve diagnostic accuracy.

At the other end of the spectrum, approximately one-third of patients with AIH are asymptomatic at the time of diagnosis. Patients with asymptomatic AIH are more likely to be men and have significantly lower bilirubin and transaminase levels compared with their symptomatic counterparts, though histological changes including degree of necroinflammatory lobular activity and fibrosis are similar. Approximately 70% of asymptomatic AIH patients eventually develop clinical symptoms. Response to immunosuppression is excellent, slightly better than that seen in symptomatic populations, with no treatment failures observed, although relapse following cessation of therapy occurs at a similar rate.

Approximately 34% of patients with AIH have established cirrhosis at the time of presentation, highlighting the chronic nature of this disease. The proportion of symptomatic and asymptomatic patients that have histologically confirmed cirrhosis during their initial workup is similar and is, perhaps not surprisingly, associated with poorer 10-year survival than those without advanced fibrosis. A similar proportion of children with AIH have established cirrhosis at the time of presentation. ^,

Pathogenesis

AIH is chronic inflammatory syndrome associated with a defect in suppressor T-cell function that leads to loss of self-tolerance and the development of a T-cell–mediated immune response that primarily targets hepatocytes. Although most cases occur de novo without a clear triggering event, some medications and infections appear to be associated with development of AIH. Hepatitis A virus (HAV) infection, in particular, is strongly suspected to be a trigger for AIH. In genetically susceptible individuals, abnormal T-helper activation and the production of antibodies targeting the asialoglycoprotein receptor (ASGPR) continue to occur following viral clearance. A similar association has been found with the Epstein-Barr virus (EBV). Other possible hepatotropic and non-hepatotropic viral triggers have been suggested but have a less clear association with the development of subsequent immune-mediated liver injury. ^,

An ever-expanding list of drugs have been cited as proposed triggers of autoimmune liver disease. True drug-induced AIH (DI-AIH) is a self-perpetuating process that develops in patients not previously diagnosed with AIH. In one large series, nitrofurantoin and minocycline were the most common medications associated with DI-AIH and produced a clinical syndrome virtually indistinguishable from de novo AIH. DILI may also occur in the setting of methyldopa or hydralazine use, although only one-half of these cases show an immune-mediated phenotype resembling AIH with associated hypergammaglobulinemia and circulating autoantibodies. More recently, DI-AIH has been linked with biological agents targeting TNF-α such as infliximab, adalimumab, and etanercept and was reported as a serious adverse event in a large international pharmacovigilance database. Although drug discontinuation and the administration of corticosteroids frequently result in clinical improvement, as with de novo AIH, long-term immunosuppression may be required, and DI-AIH–related deaths have been reported.

Numerous circulating autoantibodies have been identified in patients with AIH, but these lack specificity and play only a minor role in disease pathogenesis. ANAs targeting chromatin and ribonucleoprotein complexes are identified in nearly 80% of patients with AIH, but are also found in patients with other chronic liver diseases, including PBC, PSC, and NASH, dramatically limiting their diagnostic utility. ASMAs show somewhat higher specificity for AIH than ANAs and are moderately sensitive in the diagnosis of AIH. ASMAs target cytoskeletal components, including the microfilament F-actin, and are found in approximately 70% of AIH patients. Higher autoantibody titers (greater than 1:40) have better diagnostic specificity, but the presence of ASMAs in otherwise asymptomatic patients with normal liver function tests generally does not portend the development of AIH. As discussed earlier, the presence of ANAs and ASMAs defines AIH type 1.

In contrast, AIH type 2 is defined by the presence of anti-LKM1 antibodies, targeting CYP2D6; anti–LC-1 antibodies, targeting the folate-metabolizing enzyme formiminotransferase cyclodeaminase (FTCD); and LKM3, targeting uridine diphosphate glucuronosyltransferase (UGT), which is involved with drug and bilirubin metabolism. ^, Anti-LKM1 antibodies are generally found in children in the absence of other autoantibodies. As previously discussed, the HLA DQB1∗0201 allele is the strongest genetic modifier conferring susceptibility to AIH type 2, although circulating LKM1 may also be associated with the DRB1∗0701 allele. ^, Interestingly, DRB1∗0701 is also associated with persistent HCV infection, at least in northern European Caucasian populations, and low titers of anti-LKM1 may be seen in individuals with chronic HCV, again suggesting the possibility of an infectious trigger in genetically susceptible individuals. However, these antibodies appear to target a different epitope than those present in de novo AIH type 2. Currently, the role of the humoral immune response as a mediator of hepatocellular injury in AIH is thought to be limited, at best.

Hepatic injury in AIH is, instead, believed to be primarily a T-cell–mediated process. Impairment of regulatory T-cell (T-reg) function results in a loss of immune tolerance to hepatocellular self antigens, leading to the initiation and propagation of an effector T-cell response. Decreased activity and lower numbers of T-regs in patients with AIH result in an increase in the proliferation of effector T cells as well as their production of proinflammatory cytokines like interferon-gamma (IFN-γ) and IL-4. Although the targeted self antigen is not known for most cases of de novo AIH, CYP2D6, SEPSECS, and FTCD have been well-characterized as the antigenic targets of both B cells and effector T cells in AIH type 2. ^, These observations have led some to hypothesize that molecular mimicry (i.e., when an immune response directed against pathogenic organisms or xenobiotics inadvertently targets structurally similar self antigen) may contribute to the development of AIH. ^, For example, there is significant sequence homology between CYP2D6 residues 193 to 212 and the HCV and CMV epitopes displayed by antigen-presenting cells, suggesting that viral infection is capable of initiating an inflammatory cascade that leads to a sustained immune response to self antigen. Further work is needed to clarify the complex interplay among the genetic, environmental, and immunological factors that play a role in the pathogenesis of AIH.

Pathology

Gross Findings

Gross examination of the explanted liver for AIH is typically uninformative. Established cirrhosis is present in approximately one-third of AIH patients at the time of initial presentation and frequently shows a mixed micronodular/macronodular pattern that is indistinguishable from that seen in the setting of chronic viral hepatitis or burnt-out steatohepatitis. The surface of the liver appears shrunken and nodular. Sectioning reveals tan-brown to bile-stained regenerative nodules that are enveloped by variably prominent, firm, gray-white fibrous tissues ( Fig. 48.1A ). The gross appearance of the liver in the setting of a fulminant clinical presentation is similarly nonspecific and is indistinguishable from toxic/drug-mediated injury and viral hepatitis. The hepatic capsule may appear contracted or shriveled. The cut surface frequently shows large areas of hemorrhage and soft necrosis. Islands of residual/regenerating hepatocellular parenchyma are occasionally identifiable in an otherwise necrotic liver ( Fig. 48.1B ).

Microscopic Findings

The histological appearance of the liver in AIH shows a great deal of variability with respect to grade and stage of the disease, but it tends to show a chronic hepatitis pattern of injury with elements of interface and lobular activity. The portal inflammatory infiltrate is typically rich in both lymphocytes and plasma cells, although variable numbers of histiocytes, eosinophils, and neutrophils may also be present. Plasma cells, in particular, are conspicuous in most cases, frequently forming clusters of greater than 5 cells and outnumbering the lymphocytic component of the inflammatory infiltrate. These plasma cells tend to be immunoreactive for IgG. Although some authors have suggested that a preponderance of IgG-producing plasma cells supports a diagnosis of AIH over PBC, these observations are not specific to AIH, and interpretation of IgG:IgM ratios are challenging, limiting their practical application in routine clinical practice. ^, Importantly, the absence of a plasma cell–predominant infiltrate does not exclude a diagnosis of AIH and is occasionally encountered in otherwise typical cases of AIH.

Bile duct inflammation may be seen as a component of AIH, but it does not necessarily imply the presence of an overlap syndrome (see later discussion). In one series, approximately 24% of patients with otherwise classical AIH showed biliary changes, including destructive cholangitis and ductopenia, but responded well to immunosuppression and lacked other histological and serological features of PBC. ^, Typically, bile duct changes in AIH are characterized by the presence of occasional mononuclear inflammatory cells with focal, if any, appreciable epithelial injury and the absence of florid duct lesions ( Fig. 48.2 ). Cholestasis is generally not a feature of AIH , though it may be seen in severe cases with marked necroinflammatory activity.

In mildly active cases of AIH, there may be only a subtle, mild increase in portal lymphocytes and plasma cells associated with focal interface activity ( Fig. 48.3 ). With increasing disease activity, more portal inflammatory cells breech the limiting plate of hepatocytes to involve the periportal lobular parenchyma ( Fig. 48.4A ). In areas of interface hepatitis, evidence of hepatocellular injury, characterized by cellular swelling and ballooning degeneration, should be evident ( Fig. 48.4B ). Overt periportal hepatocellular necrosis and acidophil bodies ( Fig. 48.4C ) may also be identified and has historically been referred to as piecemeal necrosis. Collectively, more than 80% of AIH cases show at least focal interface hepatitis involving most portal areas (Ishak grade A2). However, this histological finding alone is insufficiently specific to discriminate AIH from non-immune causes of acute hepatitis, which exhibit interface activity with a similar frequency. The presence of abundant plasma cells and plasma cell clusters appears to be a better indicator of immune-mediated injury ( Fig. 48.5 ).

More recently, the presence of hyaline droplets in the cytoplasm of Kupffer cells has been described as a novel diagnostic clue that is highly specific for the diagnosis of AIH. In the initial series, well-circumscribed bodies with a glassy appearance were identified in the cytoplasm of Kupffer cells in all of the AIH patients in their cohort and showed outstanding specificity. These hyaline globules are periodic acid–Schiff (PAS) positive, diastase resistant ( Fig. 48.6A ), and immunoreactive for various immunoglobins, including IgG. Similar results have been reported by other groups, some of which advocate for the inclusion of Kupffer cell hyaline globules as a typical histological feature in the simplified diagnostic criteria for AIH. ^,

The presence of intact lymphocytes and plasma cells within the cytoplasm of viable hepatocytes (emperipolesis) is identified in the majority of AIH cases ( Fig. 48.6B ). Data from some groups indicate that emperipolesis is significantly more common in the setting of AIH than it is in chronic viral hepatitis, going so far as to suggest that it is superior to the presence of plasma cells and interface hepatitis as a histological indicator of AIH. Others contest that emperipolesis is difficult to reliably interpret in routine clinical practice and is, instead, correlated with inflammatory grade, regardless of etiology. ^, Nevertheless, emperipolesis is currently included with the presence of interface hepatitis and the formation of hepatocyte rosettes as “typical” histology for AIH in the simplified scoring system for AIH (discussed further later).

In addition to increased portal inflammation and interface hepatitis, AIH is typically associated with an element of lobular necroinflammatory activity as well. In cases with mild lobular activity, occasional aggregates of lymphocytes, plasma cells, and ceroid-laden Kupffer cells are identified in association with small foci of hepatocellular dropout ( Fig. 48.7A ). Acidophil bodies may also be encountered in the lobules, along with swollen hepatocytes with pale, rarefied cytoplasm and usually have no zonal predilection. In cases with severe lobular activity, more extensive lymphoplasmacytic inflammation is typically encountered and may be associated with large swaths of bridging or confluent hepatocellular necrosis ( Fig. 48.7B ). Regenerative hepatocellular rosettes may also be seen in cases with severe necroinflammatory lobular activity and are considered to be a typical histological feature of AIH that is purported to aid in its discrimination from chronic viral hepatitis ( Fig. 48.8 ). The significance of hepatocellular rosettes has, however, been called into question by one group who identified unequivocal rosettes in only 33% of AIH cases as well as in a similar proportion of cases without AIH. Further refinement of the histological criteria considered to be typical or compatible with AIH will undoubtably occur in the future.

AIH occasionally presents with a centrilobular-predominant pattern of hepatocellular injury that may or may not be accompanied by increased lymphoplasmacytic portal inflammation and interface hepatitis, but responds well to immunosuppressive therapy ( Fig. 48.9 ). Some have hypothesized that this centrizonal pattern of injury represents an early form of the disease with the potential to progress to more classic AIH, although the recurrence of a predominantly centrizonal pattern of injury during periods of relapse suggests that it may be a distinct variant of immune-mediated liver injury. Prominent centrilobular changes, namely the presence of a plasma cell–enriched inflammation and central perivenulitis, with massive hepatocellular necrosis may also be seen in the setting of AIH-associated acute liver failure. Recognition of this pattern of injury is essential because initiating immunosuppressive therapy early dramatically improves prognosis.

Although cases of acute-onset AIH undoubtably exist, AIH is generally considered to be a chronic disease, and fibrosis is often present at the time of initial clinical presentation. The natural history liver fibrogenesis in AIH follows a similar pattern to that seen in chronic viral hepatitis, beginning with mild expansion of the portal tracts ( Fig. 48.10A ) and progressing to periportal and bridging fibrosis, and ultimately established cirrhosis ( Fig. 48.10B ). From a histological perspective, it is often not possible to differentiate cirrhosis that has developed in the setting of AIH from other etiologies, although interface hepatitis and a plasma cell–rich inflammatory infiltrate may occasionally persist. Attempts should, nevertheless, be made to accurately characterize the immune-mediated nature of the disease, as fibrogenesis is a dynamic process, and there is mounting evidence that immunosuppressive medical therapy can stabilize or reduce fibrosis stage in a subset of patients. ^,

The surgical pathologist’s role in the diagnosis of AIH is relatively straightforward, as the histological pattern of injury is only one of the components used to predict a patient’s likelihood of having the disease. According to the simplified diagnostic criteria for AIH, typical histological features of AIH include the presence of interface hepatitis, emperipolesis, and hepatocellular rosette formation, while features compatible with AIH are characterized by a chronic hepatitis pattern of injury, but fall short of that which is considered to be typical (see Table 48.3 ). Correlating these histological findings with laboratory studies for circulating AIH-associated antibodies and serum IgG has excellent specificity for the diagnosis of AIH. Other groups have proposed modified scoring systems that consider prominent plasma cells and Kupffer cell hyaline globules as typical features of AIH or require exclusion of biliary features using copper special stains and immunohistochemical stains for CK7, which purportedly identify cases of AIH that would have been misdiagnosed by applying the simplified score. ^, Further validation is, however, required, particularly in the setting of overlap syndromes in which features of biliary injury and chronic cholestasis coexist with typical features of AIH.

Differential Diagnosis

The pattern of injury seen in AIH is insufficiently specific to render a diagnosis based on histological findings alone. As such, the clinical exclusion of infectious, toxic, and metabolic etiologies is an essential component to the diagnostic workup of a patient with suspected AIH. Acute-onset AIH frequently shows extensive lobular and perivenular activity as well as perivenular necrosis. The differential diagnosis for an acute hepatitis pattern of injury includes infectious etiologies, such as acute viral hepatitis A, B, and E, as well as infection with the Epstein-Barr virus and other nonhepatotropic viruses. Acute viral hepatitis A, in particular, may show a pronounced plasma cell–rich portal inflammatory infiltrate, along with interface activity and periportal necrosis ( Fig. 48.11 ). Acute viral hepatitis E is also characterized by an expansion of the portal tracts by a dense inflammatory infiltrate, although it is typically largely lymphocytic and associated with bile duct injury and cholestasis. The presence of centrizonal necrosis may also cause diagnostic confusion with acute-onset AIH. Complicating matters further, patients previously diagnosed with AIH more frequently have antibodies to the hepatitis E virus than patients with other immune disorders and other forms of chronic viral hepatitis, and infection may contribute to severity of hepatocellular injury. Although application of the modified and simplified International Hepatitis Group scoring systems can effectively discriminate AIH from acute viral hepatitis, this is rarely a diagnostic dilemma because serological studies for circulating antibodies and quantitative polymerase chain reaction (qPCR) testing for viral genetic material can reliably identify an active infection. ^{, ,}

Distinguishing AIH from DILI is complicated because some commonly prescribed medications have the potential to induce immune-mediated liver injury. DI-AIH can occur unpredictably, occasionally many months after initial exposure, and ongoing injury can be sustained following drug withdrawal. Although some authors have proposed models that incorporate portal inflammation, portal plasma cells, intraacinar eosinophils, hepatocellular rosette formation, and cholestasis as a means to discriminate AIH from DILI, this may be challenging in routine clinical practice ( Fig. 48.12 ). ^, The clinical exclusion of recent exposure to known or suspected hepatotoxic drugs and the application of the well-validated modified scoring system almost certainly have better performance characteristics than histological assessment alone.

The clinical differential for a patient with metabolic risk factors and increased transaminases will frequently include both nonalcoholic fatty liver disease (NAFLD) and AIH. This is, at least in part, caused by the fact that ANAs and ASMAs are significantly more common in patients with NAFLD than they are in the general population. Autoantibody titers are, however, generally lower in patients with NAFLD than that seen in the setting of AIH, although there is significant overlap. In situations in which there is diagnostic uncertainty, liver biopsy may be performed and shows good diagnostic accuracy. ^, The distinction between AIH and uncomplicated steatosis is typically straightforward because the presence of steatotic hepatocytes associated with minimal portal and lobular inflammation is not a typical feature of AIH. Discriminating AIH from nonalcoholic steatohepatitis (NASH) may be more challenging in some instances because portal inflammation is increased in progressive NAFLD and NASH, which may also show concurrent lobular inflammation and associated hepatocellular injury. Interface hepatitis and plasma cell clusters would, however, be unusual in NASH. Conversely, the presence of hepatocellular ballooning and zone 3–predominant pericellular (“chicken wire”) fibrosis is said to favor a diagnosis of NASH, as these findings are not typical for AIH. Of course, given the prevalence of the metabolic syndrome in Western populations, AIH may coexist with NAFLD/NASH, which poses additional diagnostic challenges that are not adequately addressed by current scoring systems. ^,

Cases of AIH with prominent portal inflammation and less lobular activity may mimic chronic viral hepatitis. As discussed previously in the context of acute viral hepatitis, the exclusion of infection is easily accomplished with serological and PCR-based studies, so this is rarely a clinical question. Nevertheless, there are several morphological clues that may be helpful in suggesting the proper diagnosis if these studies have not been performed or are not yet available. The portal inflammatory infiltrate in chronic viral hepatitis C is typically rich in lymphocytes. Lymphoid aggregates are often particularly prominent and may have well-developed germinal centers. Interface hepatitis is often present, but it is usually only mild and focal, and lobular injury is nearly always minimal ( Fig. 48.13A ). In contrast, lobular and interface activity is usually prominent in AIH ( Fig. 48.13B ). The presence of numerous or clustered plasma cells is also said to be relatively specific for AIH, although they may also be seen in the setting of chronic viral hepatitis C. ^,

In contrast with chronic viral hepatitis C, interface and lobular activity is often quite robust in chronic viral hepatitis B(HBV), particularly during the phase of immune clearance, or following the emergence of viral mutants or superinfection with the hepatitis D virus. ^, In these scenarios, periportal inflammation with interface hepatitis is often marked, and lobular necrosis can be extensive. Plasma cells are, again, thought to be uncommon in chronic viral hepatitis B compared with AIH, and viral surface and core antigen may be identifiable on H&E-stained tissue sections as cytoplasmic ground-glass inclusions and sanded nuclei, respectively. ^, These viral antigens can be identified with immunohistochemical stains performed on slides cut from formalin-fixed, paraffin-embedded tissue, but quantifying HBV DNA in a blood sample by qPCR is the most reliable means of detecting infection. Again, both the modified and simplified scoring systems are able to differentiate chronic viral hepatitis from AIH.

Given significant differences in treatment guidelines, AIH must be differentiated from PBC and PSC. From a biochemical standpoint, this is often not a challenge as AIH presents with a hepatitic pattern of liver injury, whereas PBC and PSC are characterized by cholestatic liver chemistries. The ratio of serum alkaline phosphatase (ALP) relative to transaminase levels is, in fact, a component of the modified scoring system, with lower AP levels favoring a diagnosis of AIH. The absence of significant periportal copper deposition and large numbers of CK7-positive intermediate hepatobiliary cells also argue against a chronic cholestatic process in newer AIH scoring schemes. Serum antimitochondrial antibodies (AMAs), particularly those targeting PDH-E2, are extremely sensitive and specific for PBC, although other AMAs may occasionally be detected in patients with AIH. From a histological perspective, although focal bile duct inflammation is not infrequently encountered in AIH, destructive cholangitis with bile duct loss and florid duct lesions is not a feature of this disease and readily distinguishes it from PBC in most cases. Similarly, the presence of concentric fibrosis with obliterative bile duct scarring and loss is seen in PSC but not in AIH. Abnormal cholangiographic studies demonstrating bile duct strictures also strongly favor a diagnosis of PSC.

Variant forms of AIH with superimposed chronic cholestatic features (overlap syndromes) are discussed later in this chapter.

Grading and Staging

No grading and staging scheme has been developed for AIH. The system developed by Batts and Ludwig in 1995 is still routinely used in clinical practice to grade disease activity in chronic hepatitis, including AIH. The degree of interface hepatitis and lobular inflammation and necrosis are separately graded based on the intensity and distribution of necroinflammatory activity, with the more severe lesion determining the overall grade ( Table 48.4 ). The degree of fibrosis is considered separately, with mild fibrous expansion of the portal tracts being characterized as stage 1, periportal fibrosis with fine strands of fibrosis in zone 1 as stage 2, septal fibrosis with connective tissue bridges linking anatomically distinct structures as stage 3, and established cirrhosis with nodular remodeling as stage 4.

Other scoring systems such as the Scheuer system, French METAVIR system, Knodell system, and Ishak system may also be used with the choice being dictated by the specific needs of the treating hepatologist and the pathologist’s experience with each system.

Natural History and Treatment

The majority of cases of AIH respond to immunosuppressive medical therapy, which should be considered in all patients with active hepatic inflammation and elevated transaminases. Induction immunosuppression therapy is typically corticosteroid-based with either prednisone or prednisolone. More recently, budesonide, a synthetic corticosteroid with high first-pass metabolism, has been shown to be highly efficacious at treating AIH while showing fewer steroid-specific side effects than prednisone. The utility of budesonide is severely limited, however, in patients with significant portosystemic shunting, including those with established cirrhosis.

Azathioprine may also be started at the initiation of induction therapy, although most centers delay administration for 2 weeks until steroid responsiveness is confirmed and testing for thiopurine methyltransferase activity excludes patients at high-risk for therapy-induced myelosuppression. Corticosteroids are gradually weened following normalization of serum aminotransferases and IgG. After 6 months, most patients can be maintained on azathioprine alone or another steroid-sparing agent like mycophenolate mofetil. ^,

Historically, AIH relapse following the withdrawal of immunosuppression was common. Patients with 2 years of sustained biochemical remission on monotherapy appear to have a greater likelihood of successfully discontinuing immunosuppression, although remission still occurs in approximately one-half of these patients. Higher serum transaminase levels during maintenance appears to be associated with a greater risk for relapse. Most cases of relapse occur during the first 3 months following cessation of immunosuppression, although late relapses may also occur. Biochemical response to further immunosuppressive therapy is typical following relapse.

The role of liver biopsy before the withdrawal of immunosuppressive therapy is somewhat controversial. Although the 2019 American Association for the Study of Liver Diseases (AASLD) practice guidelines acknowledge the utility of liver biopsy at excluding the presence of clinically unsuspected inflammation, histological examination is not mandatory in adults. This is, at least in part, due to the fact that in one study, nearly one-half of patients with a sustained biochemical response on maintenance therapy and an absence of active inflammation at the time of biopsy experience disease relapse. In contrast, prewithdrawal biopsy is still recommended in pediatric populations because the absence of inflammation appears to be associated with a greater likelihood of sustained remission.

Treatment protocols for AIH continue to evolve but are effective at inducing remission in most instances. Nevertheless, approximately one-third of AIH patients have established cirrhosis at the time of initial presentation, and AIH is considered to be an appropriate indication for liver transplantation. Graft and patient survival in this population is generally very good. However, AIH is reported to recur in 17% to 34% of allografts, but it can usually be managed with increased immunosuppression. ^,

Clinical Features and Epidemiology

The incidence of the disease ranges from 0.9 to 5.8 per 100,000 people. The key features of PBC are listed in Table 48.5 and compared with other cholestatic liver diseases. Approximately 90% of patients with PBC are female between 40 and 60 years of age (range, 20 to 80 years) ( Box 48.2 ). PBC occurs worldwide, and all races are susceptible. ^, There is some familial clustering of PBC, although less than 1% of first-degree relatives of patients with PBC develop the disorder.

Many patients with PBC are asymptomatic at presentation, detected primarily by the report of abnormal liver function tests. The early phase of disease is insidious, and symptomatic patients primarily report itching and fatigue. In later stages, the signs and symptoms of disease are either related to progressive cholestasis or cirrhosis. Among the former are pruritus, xanthomas, jaundice, and osteoporosis. The signs and symptoms of cirrhosis in PBC are the same as for other causes of cirrhosis: esophagogastric varices, ascites, spider angiomas, and splenomegaly.

PBC coexists with other autoimmune diseases; the strongest recognized associations are with connective tissue disorders such as rheumatoid arthritis, CREST syndrome ( c alcinosis, R aynaud disease, e sophageal dysmotility, s clerodactyly, and t elangiectasia), systemic lupus erythematosus, Sjögren syndrome, dermatomyositis, interstitial lung disease, and autoimmune thyroid disease. ^,

Pathogenesis

The etiology of PBC is favored to be related to an interaction between genetic predisposition and environmental triggers.

Pathology

Gross Pathology

In early-stage PBC, the liver is often slightly enlarged and variably bile stained. In later stages, macronodular cirrhosis develops and is often associated with an intense green hue that reflects progressive cholestasis ( Fig. 48.14 ). In contrast with PSC, cholangiectases and cholangitic abscesses are not seen.

Microscopic Pathology

Early and Intermediate Phase of Disease

The hallmark lesion of PBC is destructive cholangitis that affects interlobular and septal bile ducts (<70 to 80μm in diameter) and ultimately results in duct loss. Unlike PSC, the large and extrahepatic ducts are spared. The term nonsuppurative destructive cholangitis accurately describes these lesions, but the more succinct term florid duct lesion is more commonly used ( Figs. 48.15 to 48.18 ). Because these lesions are usually focal, they may be missed on liver biopsy. The florid duct lesion is characterized by an ill-defined cluster of epithelioid histiocytes (granuloma) accompanied by lymphocytes that are centered on septal and interlobular bile ducts, with histological evidence of bile duct damage and destruction. Well-formed sarcoidal-type granulomas are uncommon.

The bile duct damage in florid duct lesions is usually segmental in the longitudinal axis of the duct and in cross-section, with only a portion of the duct affected at one time. Granulomas are often poorly defined and tend to be associated with the bile duct in an eccentric fashion, although concentric involvement may also be seen. Granulomas may also be found in portal structures without an obvious connection to a bile duct. Less commonly, they may be found in hepatic lobules. Additional characteristic features, but with a less dramatic form of bile duct injury, include disruption of the bile duct basement membrane (highlighted on a PAS/D stain), intraepithelial lymphocytes and plasma cells, biliary epithelial cytoplasmic vacuolization, and occasional mitotic figures.

The result of the destructive cholangitis is disappearance of the duct, frequently leaving a lymphoid aggregate and sometimes periodic acid Schiff–positive amorphous basement membrane material as remnants of the lost duct. Because the bile ducts typically run parallel to hepatic artery branches, the finding of an unaccompanied artery in >50% of portal tracts is presumptive evidence of a vanishing bile duct disease. All of these changes are typically accompanied by bile ductular proliferation.

Bile duct changes are also typically accompanied by a dense portal lymphoplasmacytic infiltrate (see Fig. 48.17 ; Fig. 48.19 ), which is typically patchy in early stages and may contain lymphoid aggregates and follicles, conspicuous plasma cells, and a few eosinophils. Plasma cells may be particularly conspicuous, mimicking AIH. Biliary damage is likely primarily caused by the T-cell infiltrates, but B cells appear to participate, primarily in the early stages. When comparing AIH and PBC in terms of plasma cell subsets, IgG-positive cells strongly predominate over IgM-positive cells in AIH, but in PBC, IgM-positive plasma cells are much more conspicuous and may predominate. However, subtyping of immunoglobulin subclasses is seldom performed in clinical practice. The portal lymphocytic infiltrate may spill over into the lobules, mimicking interface hepatitis, the so-called biliary type piecemeal necrosis (see Fig. 48.19 ). Unlike AIH, hepatocyte necrosis at the interface is exceptionally uncommon.

The hepatic lobules usually do not show significant histological changes. In some cases, clusters of lymphocytes and plasma cells may aggregate in the lobules, although acidophil bodies are absent or inconspicuous ( Fig. 48.20A ). Scattered, usually small, noncaseating, lobular granulomas may be seen as well on occasion ( Fig. 48.20B ).

A subtle form of nodular regenerative hyperplasia has been observed in almost 50% of patients with early-stage disease, and it may be prudent to perform a reticulin stain in select cases. These changes may reflect portal venule damage by granulomas and contribute to the early development of portal hypertension. Portal hypertension may develop in noncirrhotic PBC patients, and some are associated with nodular regenerative hyperplasia.

Late Phase of the Disease

Regenerative nodules in PBC and PSC are usually not as regular and round as in other forms of cirrhosis. The nodules often have a garland-shaped, irregular outline, not unlike pieces of a jigsaw puzzle. A progressive loss of ducts may occur early in the course of disease, with ductopenia usually being prominent in the later phase of the disease ( Figs. 48.21 and 48.22 ). Florid duct lesions become less common with advancing stage, presumably reflecting loss of the target biliary epithelium.

Periportal hepatocyte feathery degeneration, which is thought to reflect bile acid (cholate) stasis, helps distinguish PBC and PSC from chronic hepatitis resulting from autoimmune, viral, or drug causes. Additional changes associated with cholestatic liver include Mallory bodies and deposition of copper in periportal/periseptal hepatocytes (see Fig. 48.22C ,D). The combined low-magnification effect of these features is a characteristic halo-like pattern surrounding regenerative nodules (see Fig. 48.22E ).

Staging Primary Biliary Cholangitis

Although liver biopsy remains the gold standard for staging fibrosis, noninvasive methods such as transient elastography have emerged as viable alternatives because of the issues related to sampling errors and the risk of serious adverse effects, albeit rare, associated with liver biopsies. Clinically, patients are often characterized as early-, intermediate- (bridging fibrosis), and late-stage disease. There are several more complex staging systems that are less often used in the clinic and more often applied in the context of clinical trials. The three histological grading systems are illustrated in Table 48.6 . The heterogeneity with regard to inflammation and fibrosis that is so characteristic of PBC compounded by sampling errors significantly limits the accuracy of these staging systems.

Differential Diagnosis

PBC typically causes a chronic cholestatic disease with insidious onset, a long asymptomatic phase, and slow progression over the course of years. A patient presenting with acute onset of jaundice is unlikely to have PBC. One of the challenges associated with the histological diagnosis of PBC lies in the patchy nature of the disease; thus sampling error is always a consideration when PBC is considered clinically. A positive copper stain is often helpful in establishing the cholestatic nature of the disease, although it does not distinguish PBC from other cholestatic diseases such as PSC.

Natural History

Asymptomatic patients with early-stage disease have a longer time to transplantation or disease-related death. Before widespread screening, the median survival of patients was 6 to 10 years. With most patients diagnosed early in the course of the disease and the introduction of ursodeoxycholic acid (UDCA), the overall median survival has significantly increased.

Hepatocellular carcinoma develops in 6% of patients with PBC per year, although the risk is probably not as high as that seen in cirrhosis from other causes, such as viral hepatitis, hemochromatosis, or alcohol. Surveillance is recommended every 6 to 12 months. Older patients, male gender, advance histological stage, and inadequate response to UDCA are associated with an elevated risk for developing hepatocellular carcinoma.

Treatment

The initial therapy of choice for PBC is long-term treatment with UDCA. Improvement in alkaline phosphatase is typically observed within 3 months of starting UDCA. ^, Approximately one-third of patients have an inadequate biochemical response. ^, UDCA is associated with a long-term survival benefit. In a systematic review, of 4845 patients, UDCA was associated with higher rates of transplant-free survival at 5, 10, and 15 years compared with no treatment.

The mechanisms underlying this favorable response with UDCA include (1) increased hydrophilic index of the circulating bile acid pool, (2) stimulation of hepatocellular and ductular secretions and thus elimination of bile acids, (3) protection against bile acid injury, and (4) immunomodulation and antiinflammatory effects of UDCA. Obeticholic acid is used for patients with an inadequate biochemical response to UDCA. Obeticholic acid can be used in combination with UDCA.

Of note, prior treatments included immunosuppressive drugs (e.g., azathioprine, d -penicillamine, corticosteroids, cyclosporine, methotrexate) and the antifibrotic agent colchicine, and they met with little success.

For patients with decompensated cirrhosis, orthotopic liver transplantation (OLT) has emerged as the most effective form of therapy. OLT is associated with excellent short-and long-term survival rates with 1-year survival rates of 90% to 95% (see Chapter 52 ). PBC may recur in the allograft. In one study of 571 patients who had liver transplantation for PBC, the rates of recurrence at 5 and 10 years were 18% and 31%, respectively, and recurrence was associated with higher risk of graft loss.