Book on hepatitis from page 466 to 475

Book on hepatitis from page 466 to 475

466  Hepatology 2012
antibodies (LKM), soluble liver antigen antibodies (SLA) or ANA. In cholestatic
cases cholangiography detects sclerosing cholangitis. In an overlap syndrome the
classical appearance of the individual disease component is mixed with features of
another autoimmune liver disease. Immunoglobulins are usually elevated in all
autoimmune liver diseases.
Regarding a therapeutic strategy the leading disease component is treated. In an
overlap syndrome presenting as hepatitis, immunosuppression with prednisone (or
combination therapy with azathioprine) is initiated. In cholestatic disease
ursodeoxycholic acid is administered. Both treatments can be combined when
biochemistry and histology suggest a relevant additional disease component
(Chazouilleres 1998). Validated therapeutic guidelines for overlap syndromes are
not available. It is important to realize that treatment failure in AIH may be related
to an incorrect diagnosis or an overlap syndrome of autoimmune liver diseases
(Potthoff 2007). Several studies show that treatment of the AIH component of
overlapping syndromes is important to avoid progression to cirrhosis (Chazouilleres
2006, Gossard 2007, Silveira 2007, Al-Chalabi 2008).
Liver transplantation
In approximately 10% of AIH patients liver transplantation remains the only life-saving option (Strassburg 2004). The indication for liver transplantation in AIH is
similar to that in other chronic liver diseases and includes clinical deterioration,
development of cirrhosis, bleeding esophageal varices and coagulation
abnormalities despite adequate immunosuppressive therapy (Neuberger 1984,
Sanchez-Urdazpal 1991, Ahmed 1997, Prados 1998, Tillmann 1999, Vogel 2004).
There is no single indicator or predictor for the necessity of liver transplantation.
Candidates for liver transplant are usually patients who do not reach remission
within 4 years of continuous therapy. Indicators of a high mortality associated with
liver failure are histological evidence of multilobular necrosis and progressive
hyperbilirubinemia. In Europe, 4% of liver transplants are for AIH (Strassburg
2009). The long-term results of liver transplantation for AIH are excellent. The 5-year survival is up to 92% (Sanchez-Urdazpal 1991, Prados 1998, Ratziu 1999) and
well within the range of other indications for liver transplantation. The European
liver transplant database indicates 76% survival in 5 years and 66% survival after 10
years (1647 liver transplantations between 1988 and 2007). When these numbers are
considered it is necessary to realize that patients undergoing liver transplantation
usually fail standard therapy and may therefore have a reduced life expectancy after
liver transplant compared to those who achieve stable complete remission on drug
therapy.
Recurrence and de novo AIH after liver transplantation
The potential of AIH to recur after liver transplantation is beyond serious debate
(Schreuder 2009). The first case of recurrent AIH after liver transplant was reported
in 1984 (Neuberger 1984), and was based upon serum biochemistry, biopsy findings
and steroid reduction. Studies published over the years indicate that the rate of
recurrence of AIH ranges between 10-35%, and that the risk of AIH recurrence is
perhaps as high as 68% after 5 years of follow-up (Wright 1992, Devlin 1995, Götz
1999, Milkiewicz 1999, Manns 2000, Vogel 2004). It is important to consider the
criteria upon which the diagnosis of recurrent AIH is based. When transaminitis is
Autoimmune Liver Diseases: AIH, PBC and PSC  467
chosen as a practical selection parameter many patients with mild histological
evidence of recurrent AIH may be missed. It is therefore suggested that all patients
with suspected recurrence of autoimmune hepatitis receive a liver biopsy,
biochemical analyses of aminotransferases as well as a determination of
immunoglobulins and autoantibody titers (Vogel 2004). Significant risk factors for
the recurrence of AIH have not yet been identified although it appears that the
presence of fulminant hepatic failure before transplantation protects against the
development of recurrent disease. Risk factors under discussion include steroid
withdrawal, tacrolimus versus cyclosporine, HLA mismatch, HLA type, and LKM-1
autoantibodies. An attractive risk factor for the development of recurrent AIH is the
presence of specific HLA antigens that may predispose toward a more severe
immunoreactivity. In two studies recurrence of AIH appeared to occur more
frequently in HLA-DR3-positive patients receiving HLA-DR3-negative grafts.
However, this association was not confirmed in all studies. There have not been
conclusive data to support the hypothesis that a specific immunosuppressive
regimen represents a risk factor for the development of recurrent AIH (Gautam
2006). However, data indicate that patients transplanted for AIH require continued
steroids in 64% versus 17% of patients receiving liver transplants for other
conditions (Milkiewicz 1999).
Based on these results and other studies it would appear that maintenance of
steroid medication in AIH patients is indicated to prevent not only cellular rejection
but also graft-threatening recurrence of AIH (Vogel 2004). Steroid withdrawal
should therefore be performed only with great caution. The recurrence of AIH is an
important factor for the probability of graft loss. Apart from hepatitis C and primary
sclerosing cholangitis a recent report found AIH recurrence to represent the third
most common reason for graft loss (Rowe 2008). Transplanted patients therefore
require a close follow-up and possibly an immunosuppressive regimen including
steroids, although this is controversial and not backed by prospective studies
(Campsen 2008).
In addition to AIH recurrence the development of de novo autoimmune hepatitis
after liver transplantation has been reported (Kerkar 1998, Jones 1999a, Salcedo
2002). The pathophysiology of this is also elusive. From a treatment point of view
de novo  autoimmune hepatitis, which appears to occur mostly in patients
transplanted with PBC but may just be the serendipitous occurrence of AIH, is
responsive to steroid treatment (Salcedo 2002).
Primary biliary cirrhosis
Introduction
Primary biliary cirrhosis (PBC) is a chronic inflammatory, cholestatic disease of the
liver with an unknown cause. The clinical observation of a broad array of immune-mediated symptoms and phenomena suggests the disease to be of autoimmune
etiology, in the course of which progressive and irreversible destruction of small
interlobular and septal bile ducts progressively and irreversibly ensues (Table 5). As
in other autoimmune diseases PBC affects women in over 80% of cases and is
associated with varying extrahepatic autoimmune syndromes in up to 84%. These
extrahepatic manifestations of immune-mediated disease include the dry gland
syndrome (sicca syndrome with xerophthalmia and xerostomia) but also collagen
468  Hepatology 2012
diseases, autoimmune thyroid disease, glomerulonephritis and ulcerative colitis
(Table 6).
Table 5. Clinical profile of primary biliary cirrhosis (PBC).
Sex  90% female
Age   40-59 years
pruritus
jaundice
skin pigmentation
Elevation  alkaline phosphatase (AP), aspartate aminotransferase (AST), bilirubin, IgM
antimitochondrial antibodies (AMA)
associated immune-mediated syndromes
Liver biopsy
-  cellular bile duct infiltration
-  granulomas possible
-  copper deposits
Table 6. Extrahepatic immune-mediated syndromes in PBC and overlap with
rheumatic diseases.
dry gland “sicca” syndrome
Sjögren’s syndrome
rheumatoid arthritis
autoimmune thyroid disease
renal tubular acidosis
mixed connective tissue disease (MCTD)
polymyositis
polymyalgia rheumatica
pulmonary fibrosis
CREST syndrome
systemic lupus erythematosus (SLE)
pernicious anemia
ulcerative colitis
exogenous pancreatic insufficiency
myasthenia gravis
The striking female predominance (Donaldson 1996, Mackay 1997, Uibo 1999)
and familiar clustering of PBC (Kato 1981, Jones 1999b, Tsuji 1999) suggest that
inheritable genetic factors play a role in this disease. This has focussed attention on
the immunogentics of PBC in order to further define host risk factors (Manns 1994).
Studies have suggested an instability of lymphocytic DNA in PBC patients (Notghi
1990). Immunogentic analyses, however, have only come up with relatively weak
associations with specific human leukocyte antigen haplotypes. An additional
hypothesis is an alteration of bile acid composition and bile fluid composition,
which would indicate a role for transporter proteins in the development of PBC.
Bicarbonate rich bile is believed to be protective for biliary epithelium.
Definition and prevalence of PBC
Primary biliary cirrhosis is an inflammatory, primarily T cell-mediated chronic
destruction of intrahepatic microscopic bile ducts of unknown etiology (Strassburg
2000). It affects women in 80% of cases who exhibit elevated immunoglobulin M,
antimitochondrial antibodies directed against the E2 subunit of pyruvate
Autoimmune Liver Diseases: AIH, PBC and PSC  469
dehydrogenase (PDH-E2), a cholestatic liver enzyme profile with elevated alkaline
phosphatase, gamma glutamyltransferase as well as serum bilirubin levels, and a
variable course of disease leading to cirrhosis over the course of years or decades. A
prominent feature is the presence of extrahepatic immune-mediated disease
associations. In later stages pronounced fatigue, pruritus, marked
hyperbilirubinemia and the consequences of portal hypertension such as ascites,
bleeding esophageal varices, and encephalopathy develop (Strassburg 2004).
The prevalence is estimated at 65 per 100,000 in women and 12 per 100,000 in
men with an incidence of 5 per 100,000 in women and 1 per 100,000 in men. The
prevalence and incidence appear to vary regionally. An increase of PBC incidence
in recent years may be the result of more specific testing of antimitochondrial
antibody reactivity (Strassburg 2004).
Diagnostic principles of PBC
Suspicion of PBC arises when cholestasis and cirrhosis are present in middle-aged
women (Figure 2). Ultrasound is employed to rule out mechanical cholestasis. The
presence of antimitochondrial antibodies (AMA) against PDH-E2 is diagnostic of
PBC. AMA against E2 subunits of members of the inner mitochondrial membrane-expressed oxoacid dehydrogenase complex (PDH, branched chain ketoacid
dehydrogenase [BCKD], and ketoglutarate dehydrogenase [OADC]) are present in
95% of PBC patients. AMA-negative PBC can exhibit antinuclear autoantibodies
with specificity against nuclear dot antigen (SP100), a 210 kDa nuclear membrane
protein (gp210), or nucleoporin p62. In AMA-negative PBC a biopsy is indicated to
contribute to the establishment of the diagnosis; in the presence of AMA against
PDH-E2, histology is used primarily for the staging of cirrhosis and is not necessary
(Strassburg 2004).
Diagnostic role of AMA in PBC
The main aim of AMA determinations is the detection of PBC-specific AMA and
the exclusion of AMA of low diagnostic relevance for the disease. As a screening
test the determination of AMA using indirect immunofluorescence testing on rat
kidney cryostat sections or immobilized Hep-2 cells (Strassburg 1999). The indirect
immunofluorescence on rat kidney sections leads to the staining of the distal and
proximal tubuli (note: proximal staining only is indicative of liver/kidney
microsomal antibodies, LKM). When positive AMA immunofluorescence is
detected, further analysis should include subclassification using molecularly defined
antigen preparations. The detection of PDH-E2, BCKD-E2 can be achieved by
ELISA using recombinant antigen or reference sera. If both are negative, testing
should include OGD-E2. The final step is performed using Western Blot Analyses
to confirm the findings. By Western Blot the indicative 74 kDa (PDH-E2), 52 kDa
(BCKD-E2) and 48 kDa (OGD-E2) bands can be visualized. This multi-step
regimen secures a rational and reliable diagnosis of PBC-specific AMA excluding
those found in drug-induced and infectious diseases.
In the majority of cases the determination of anti-PDH-E2 is sufficient to secure
the diagnosis. Studies will have to evaluate whether the future application of a
single PDH-E2 ELISA as highly specific screening test in suspected PBC represents
an efficient and economic diagnostic approach.
470  Hepatology 2012
Figure 2. Diagnostic algorithm of PBC including clinical presentation, ultrasound and
serology.
Therapeutic principles in PBC
Treatment leading to a cure of PBC is not available (Strassburg 2004).
Ursodeoxycholic acid (UDCA) (15 mg/kg body weight per day) has been shown to
improve serum biochemistry, histology and survival but has no effect on fatigue and
osteoporosis. It has immunomodulatory properties, alters cell signal transduction
and modifies hydrophilicity of the bile. UDCA should not be given in severe
cholestasis and during the first trimester of pregnancy. Immunosuppression in PBC
has shown disappointing results. Symptomatic therapy of the complications of PBC
includes management of pruritus (cholestyramine, induction with rifampicin, opioid
antagonists, serotonin antagonists), ascites (diuretics, beta blockers to control portal
hypertension), osteoporosis (vitamin D and calcium supplementation,
bisphosphonates in some), as well as endoscopic intervention for bleeding
esophageal varices. Fat-soluble vitamin replacement is suggested. When liver
cirrhosis-induced liver failure is progressive liver transplantation remains a
definitive therapeutic option. Ten-year survival rates are 75-80% and recurrence of
PBC after transplant occurs in 10-40%. Recurrence can be expected in 25-30%
(Rowe 2008, Strassburg 2009)
Immunosuppression in PBC
Corticosteroids:  Treatment with prednisolone can improve serum
aminotransferase activities, alkaline phosphatase and elevated immunoglobulins. It
does not lead to significant improvement of bilirubin, pruritus, or histology. In a
Autoimmune Liver Diseases: AIH, PBC and PSC  471
placebo-controlled study with 36 asymptomatic patients for over 1 year osteopenia
and cushingoid side effects were noted (Mitchison 1992).
Azathioprin:  The classical immunosuppressant azathioprin, which has a
pronounced effect in AIH, did not show significant effects in two different studies
and is not used in PBC (Christensen 1985).
Cyclosporin A: In a large study of 346 patients with a median observation time
of 2.5 years this classical transplant immunosuppressant did not show significant
effects on histological progression (Lombard 1993). Contrasting these findings, in a
small study with 20 patients who were treated for 2 years, histology improved,
which should however be viewed with caution (Wiesner 1990). Because of the
possibility of severe side effects cyclosporin A is not a recommended therapeutic
option.
D-penicillamine: Because PBC is characterized by copper accumulation in the
bile ducts the chelator d-penicillamine was studied. D-penicillamine also has
immunosuppressive and antifibrotic properties. It was tested on a total of 748
patients in 6 studies, without leading to a positive therapeutic effect. However, 30%
of patients had severe side effects (Bodenheimer 1985). D-penicillamine in PBC is
not recommended.
Colchicine:  Because of its antifibrotic and anti-inflammatory properties
colchicine was studied in 3 studies in the 1980s. Despite improvement of albumin,
bilirubin, aminotransferases and alkaline phosphatase, an improvement of clinical
symptoms and histology was not observed (Kaplan 1986, Warnes 1987,
Bodenheimer 1988). Severe side effects were not reported but an effect on long-term prognosis was not seen.
Methotrexate: Despite its known hepatotoxicity, methotrexate was used as an
immunosuppressant in PBC. In a placebo-controlled study with 60 patients, low-dose methotrexate (7.5 mg/week) led to an improvement of biochemical parameters
except for bilirubin but no effects were reported regarding necessity of liver
transplantation or survival (Hendrickse 1999). Hepatotoxicity was not observed.
Interstitial pneumonitis, which affects 3-5% of rheumatoid arthritis patients, was
observed in 14% of PBC patients. Methotrexate cannot be recommend outside of
scientific evaluations or studies.
In principle other immunosuppressants (Table 7) such as mycophenolic acid
(mycophenolate mofetil), tacrolimus (FK506) or even monoclonal antibodies
against the interleukin 2 receptor may represent interesting candidate strategies.
However, study data is currently lacking.
Table 7. Effects of immunosuppressants in PBC.
Biochemical
improvement
Histological
improvement
Survival  Side effects/toxicity
Corticosteroids  ++  ++  -  ++
Azathioprine  -  -  +  +
Cyclosporin A  ++  -  ++  ++
D-penicillamine  -  -  -  ++
Colchicine  ++  -  +  -Methotrexate  ++  +  -  +
472  Hepatology 2012
Ursodeoxycholic acid in PBC (UDCA)
In 1981, a positive effect of UDCA was first observed on elevated liver parameters,
the exact mechanism of which was unclear (Leuschner 1996). On one hand UDCA
leads to a modification of the bile acid pool to a more hydrophilic environment with
lower detergent-like properties, and it leads to increased bile flow. On the other
hand an immunomodulatory activity is suggested regarding HLA antigens expressed
on biliary epithelial cells and altered signal transduction (Paumgartner 2002). The
optimal dose in PBC patients appears to be 13-15 mg/kg. In a meta-analysis of 3
studies that looked at 548 patients with this dose, biochemical improvement and a
slower histological progression to fibrosis was observed (Poupon 1997). These
effects were only evident when follow up extended to 4 years. These data rely
heavily on the positive effects of a single study and it is not surprising that a
subsequent meta-analysis of 8 studies with 1114 patients failed to find positive
associations with UDCA therapy (Goulis 1999).
There are a number of problems with this. Doses varied and protocols included
patients with insufficient dosing, and follow up was under 2 years in some cases. In
a recently published analysis of 367 patients from 4 clinical cohorts an initiation of
UDCA therapy in early stages of PBC (stage I-II) and a treatment duration of 2
years led to a retardation of histological progression, which argues for an early
initiation of UDCA therapy after diagnosis even in the absence of fibrosis or
cirrhosis. UDCA was also shown to improve biochemistry, delay portal
hypertension and varices, and currently has no therapeutic alternative (Poupon
2003). No convincing effect was demonstrable on osteopenia and extrahepatic
manisfestations of PBC. An interesting side effect appears to be the significant
reduction of colonic epithelial proliferation. UDCA therapy is not associated with a
higher prevalence of colonic polyps and appears to delay their reappearance after
polypectomy (Serfaty 2003).
Therapy in non-responders and combination strategies
Non-response is usually defined as a failure to lower cholestatic enzyme activities
or to reach normalisation of these parameters. In patients in whom alkaline
phosphatase and gamma glutamyltransferase activities are not lowered by UDCA
therapy, increased morbidity and progression is likely. Alternative therapeutic
strategies can be considered.
Steroids and UDCA: The combination of immunosuppressants and UDCA was
looked at in smaller studies and included the use of prednisolone (Leuschner 1996),
azathioprine (Wolfhagen 1998) and budesonide (Leuschner 1999, Angulo 2000)
(Table 7). In a randomised, controlled study with 30 patients, who received 10 mg
prednisolone/day an improvement of inflammatory activity was reported (Leuschner
1996). A study with 9 mg budesonide/day showed in 39 patients not only
biochemical but also histological improvement (Leuschner 1999). In an open study
with 22 patients a deterioration of osteopenia was noted (Angulo 2000). The
combination of budesonide and UDCA may have additional beneficial effects
related to the activation of the anion exchanger AE2, which may serve to alter
biliary composition and produce a more protective bicarbonate rich bile.
Sulindac and UDCA: In an open study with 23 patients and incomplete response
to UDCA over 12 months treated with UDCA or UDCA and sulindac a trend
Autoimmune Liver Diseases: AIH, PBC and PSC  473
towards histological improvement and biochemical improvement were reported in
the combination group (Leuschner 2002).
Colchicine and UDCA: In 3 studies the combination of colchicine and UDCA
were studied for 24 months in a total of 118 patients (Raedsch 1992, Ikeda 1996,
Poupon 1996). Although mild biochemical improvement was noted, the effect of
longer treatment remains unclear. Because of the biliary elimination of colchicine
combinations with bile acids, there may be potentially toxic effect.
Methotrexate and UDCA: Several pilot studies and 3 randomized studies have
looked at methotrexate in combination with UDCA. In a recent randomized
placebo-controlled protocol with 60 patients a high rate of side effects without
therapeutic benefit was reported (Van Steenbergen 1996, Bach 2003).
Primary sclerosing cholangitis
Diagnosis of primary sclerosing cholangitis (PSC)
PSC is classically characterized by the progressive destruction of large intra- as well
as extrahepatic bile ducts and – contrasting with AIH and PBC – preferentially
affects male patients with a maximum age of around 25-45 (Strassburg 1996).
About 50-75% of the time, PSC is associated with ulcerative colitis. PSC is
clinically characterized by upper quadrant pain, pruritus, anorexia and fever, but up
to 50% of patients lack any symptoms (Weismüller 2008). The diagnosis is
established by a typical biochemical profile of cholestasis with elevations of
bilirubin, alkaline phosphatase and gamma glutamyl transferase, the characteristic
findings upon cholangiography and a typical biopsy showing ring fibrosis around
the bile ducts, which is not present in all patients. Serology regularly identifies
atypical anti-neutrophil cytoplasmic autoantibodies (xANCA) in up to 80% of
patients (Terjung 2000), although these are not disease specific and can also occur
in patients with ulcerative colitis without PSC. There is a significant association of
PSC with cholangiocarcinoma (10-20%) and colorectal cancer (9% in 10 years). In
a subgroup of patients, small bile duct PSC may be present (Broome 2002), which
lacks typical strictures and pruning of the biliary tree upon cholangiography. In
these cases the diagnosis can be established in the presence of the typical
association with ulcerative colitis in male patients by performing a liver biopsy
(Figure 3).
Differential diagnosis: sclerosing cholangitis
The finding of macroductal sclerosing cholangitis can be brought about by a number
of conditions, which include ischemia, liver transplantation complications, and
drugs. Of note are two additional differential diagnoses that require attention
(Figure 4): secondary sclerosing cholangitis (Gelbmann 2007, Esposito 2008, von
Figura 2009, Al-Benna 2011) and IgG4-associated cholangitis (Webster 2009,
Clendenon 2011, Takuma 2011, Zhang 2011).
Secondary sclerosing cholangitis is an entity with severe infection of the biliary
tree that develops in some patients following systemic infections and sepsis who are
treated with aggressive intensive care unit management. IgG4-associated cholangitis
is an immune-mediated entity often with high plasma levels of IgG4 and IgG4
expression in biliary cells obtained upon brush biopsy. The latter can be treated with
474  Hepatology 2012
immunosuppression and should be diagnosed because of an available medical
therapy.
Figure 3. Diagnostic algorithm of PSC including clinical presentation.
Figure 4a. Examples of different entities of sclerosing cholangitis. A. PSC showing
multiple strictures with narrowing (black arrows) and prestenotic dilatation (white arrows) and
an endoscopic aspect of purulent biliary infection at the biliary papilla.
Autoimmune Liver Diseases: AIH, PBC and PSC  475
Figures 4b, 4c. Examples of different entities of sclerosing cholangitis. B. Secondary
sclerosing cholangitis (SSC) with a similar intrahepatic picture but also biliary casts (dotted
arrows) that can be extracted endoscopically (right panel). C. Cholangiogram of autoimmune
(AIC) IgG4-associated cholangitis mimicking PSC. Black arrows show narrowing, white arrows
show dilatations.
Association of PSC with inflammatory bowel disease
A clinical hallmark of PSC is the high number of patients suffering from
inflammatory bowel disease (IBD). In several studies with 605 PSC patients in the
US (Mayo Clinic), UK (King’s College) and in Sweden. IBD was found in 71%,
73% and 81% of PSC cases (Boberg 1998, Bergquist 2002). In our own experience