IgG4-Related Sclerosing Cholangitis

Study

Country

Cohort

N. of patients

IgG4-SC N (%)

AIP N (%)

Kanno et al. [43]

Japan

AIP

311 (33.8)

918 (100)

Ghazale et al. [26]

USA

IgG4-SC and AIP

53 (100)

49 (92)

Huggett et al. [32]

AIP and IgG4-SC

115

69 (60)

106 (92)

Lin et al. [53]

China

IgG4-RD

118

21 (17.9)

45 (38.1)

Inoue et al. [34]

Japan

IgG4-RD

235

(13)

142 (60)

Fernandez-Codina et al. [22]

Spain

IgG4-RD

30 (4)

142 (60)

Campochiaro et al. [10]

Italy

IgG4-RD

4 (10)

17 (41)

Key: N number, AIP autoimmune pancreatitis, IgG4–SC IgG4-sclerosing cholangitis

Disease Pathogenesis

The pathological mechanisms underlying IgG4-SC are not yet fully understood. The raised serum IgG4, lymphoplasmacytic infiltration seen in disease lesions, and the response to steroids and immunosuppressive agents indicate that aberration of the immune response is central. What triggers and sustains the inflammatory process is not clear, but several mechanisms have been proposed including autoimmunity against a self-antigen, molecular mimicry, or chronic antigen exposure triggering immune dysregulation. Advances in our understanding of the genetic background and the immunological environment of patients, are beginning to unravel disease pathogenesis.

Genetic Susceptibility

No studies to date have focused on the genetics of IgG4-SC patients specifically. Evidence is growing that AIP patients have a genetic background that makes them susceptible to disease development. Single nucleotide polymorphisms in genes encoding immune factors including cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) and Fc receptor-like 3 (FcR-3) have been reported to be associated with AIP development or recurrence [14, 88, 90]. Class II human leukocyte antigen (HLA) alleles HLA DRB1_0405 and DQB1_0401 were identified to be associated with AIP [44]. A Korean study found that substitution on position 57 on HLA DQB1 was associated with disease relapse in AIP [71]. It is likely that variation in class II alleles involved in antigen presentation can influence predisposition to disease and its course.

Autoantigens

A role for autoimmunity is supported by the presence of a T-Cell, B cell, and antibody-rich infiltrate in disease lesions. Multiple candidate autoantibodies and autoantigens have been investigated in AIP, although none have been found to be specific for the disease. Antibodies against carbonic anhydrase II and lactoferrin, which are expressed widely in exocrine organs, have been reported in 73 % and 54 % of AIP patients, respectively [4, 69]. Anti-carbonic anhydrase II antibodies were found to correlate with serum IgG4 levels [4]. Another purported mechanism of disease pathogenesis is molecular mimicry between sequences found in alpha-carbonic anhydrase of the bacterium Heliobacter pylori and carbonic anhydrase II [28]. Other candidate antibodies detected at lower levels in AIP include anti-carbonic anhydrase IV, pancreatic secretory trypsin inhibitor, amylase IV, heat-shock protein 10 and plasminogen binding protein [5, 19, 23, 52, 82].

The Role of B Cells and the IgG4 Molecule

The presence of IgG4-positive plasma cells in disease lesions and raised serum IgG4 levels seen in the majority of patients are indications that B cells and antibody production are important in IgG4-SC pathogenesis. The B lymphocyte-depleting agent rituximab has been used with success to treat IgG4-SC patients refractory to steroids and conventional immunosuppressants [11, 12, 45, 46, 55]. Recent work has identified circulating oligoclonal IgG4-positive plasmablasts in patients with active IgG4-RD, which remit after treatment with rituximab and re-expand during relapse [56, 57, 94, 95]. Relapse of IgG4-RD after B-cell depletion with rituximab infers that the reemergence of IgG4-positive plasmablasts are derived from either a subset of memory B cells that survive rituximab therapy or newly generated naïve B cells that interact with a yet unidentified antigen or pathogenic T-Cell repertoire, unaffected by rituximab.

An important question in understanding IgG4-RD pathogenesis is why IgG4 immunoglobulin and IgG4-positive plasma cells are expanded in a great majority of patients. Although it has been postulated that autoantibodies might induce an inappropriate immune response, candidates thus far are of the IgG1 rather than IgG4 subclass. Oligoclonal IgG4-positive clones have been identified in sequencing of whole blood in IgG4-SC patients, suggesting that only specific B cells are expanded [54]. However a generalized polyclonal IgG4 response to multiple common antigens has been demonstrated in IgG4-RD patients. This supports the alternative theory that increased IgG4 is an epiphenomenon, occurring as a result of the expansion of preexisting IgG4-switched B cells rather than being driven by a specific autoantigen [13].

It is unknown as to whether the IgG4 immunoglobulin is directly involved in driving the inflammation seen in disease lesions. IgG4 has anti-inflammatory properties due to its unique structure that allows exchange of its Fab arm, producing functional monomers that are unable to form large immune complexes [91]. Unlike the other gamma immunoglobulin subclasses, IgG4 is unable to activate complement [92]. Under physiological conditions, specific IgG4 responses occur to generate humoral tolerance after repetitive antigen stimulation, for example, in beekeepers that are repeatedly exposed to bee venom [1]. These tolerogenic properties argue that IgG4 molecules themselves are unlikely to be intrinsically harmful.

However, in other immune conditions including pemphigus vulgaris and myasthenia gravis, IgG4 antibodies are thought to be directly pathogenic [24, 33]. In a small study, IgG4 in sera from AIP patients bound with normal pancreatic and biliary epithelial tissue, indicating an interaction between IgG4 antibodies with a yet unidentified antigen [3].

T-Cell Immunological Response

CD4-positive T-Cells are necessary to support and coordinate IgG4-switched B-cell responses, but their role in IgG4-SC pathogenesis has not been fully elucidated. T-Cells are a component of the lymphoplasmacytic infiltrate in disease lesions and are likely to interact with the B cells when in close proximity.

T-helper type 2 (Th2) cells have been implicated in IgG4-RD pathogenesis. The Th2 cytokines IL-4, IL-5, and IL-13 have been detected at the messenger RNA level in IgG4-RD disease lesions, blood CD4-positive T-Cells in IgG4-RD patients, and in the bile of IgG4-SC patients [41, 60, 83, 100, 101]. A skew of circulating CD4-positive T-Cells towards a Th2 phenotype has also been reported [73]. It has been suggested that Th2 cells in IgG4-RD promote peripheral eosinophilia, raised serum immunoglobulin E (IgE), and IgG4 predominance, as Th2-associated cytokines IL-4 and IL-13 have been shown to promote immunoglobulin class switch toward the IgG4 subtype [72, 87]. However a recent report that blood Th2 cell expansion is restricted to IgG4-RD patients with atopy challenges the hypothesis of a Th2-driven response in IgG4-RD [56, 57]. Mast cells have been suggested as an alternative source of Th2 cytokines, based on their colocalization with IL-4 and IL-13 in IgG4-RD lesions from salivary glands [79, 80].

T follicular helper cells, which support B-cell differentiation into antigen-secreting cells in germinal centers, have also been implicated in IgG4-RD pathogenesis. Next-generation sequencing of the B-cell receptor immunoglobulin heavy chain repertoire of circulating plasmablasts in IgG4-RD patients has shown they have undergone extensive somatic hypermutation, a process for which T follicular helper cells are integral [56, 57]. A recent study has shown that circulating type 2 T follicular helper (Tfh2) cells are expanded in patients with IgG4-RD [2]. Tfh2 cells preferentially secrete Th2 cytokines [59] and could be the driver of the B-cell differentiation to IgG4-positive plasmablasts and plasma cells.

The T regulatory (Treg) cell-associated cytokine IL-10 and tumor growth factor beta (TGF-β) have been found in IgG4-RD lesions [87, 100, 101]. There is also evidence that Tregs are expanded in the circulation and tissue lesions in IgG4-SC and AIP [49, 51, 61]. IL-10 has been shown to preferentially switch immunoglobulin toward IgG4 rather than IgE, and TGF-β has been purported to contribute to the fibrosis seen in late stage disease [36, 78].

Regional Factors Promoting Lymphocyte Recruitment

It has been suggested that factors local to the pancreatobiliary system may be at play in IgG4-SC, as it often occurs alongside AIP. Pathological specimens of IgG4-SC show severe inflammation in the peribiliary glands, which contain pancreatic acini [27]. In tissue specimens from AIP and IgG4-SC, the chemokine CCL1 was expressed highly at the messenger RNA level and was localized to the peribiliary glands and pancreatic duct epithelium. The expression of CCR8, the receptor for CCL1 found on Th2 and Treg lymphocytes, was also upregulated in IgG4-SC disease lesions [102]. Another study found that CXCR5, expressed on Tfh cells, and its ligand CXCL13 were upregulated in AIP tissue [20]. A variety of other chemokines have been found to be overexpressed in AIP and IgG4-SC tissue including CCL1, CXCL13, CCL17, CCL19, and CCL21, but their role in the disease is not yet clear [74].

Clinical Features and Natural History

Clinical Presentation

Patients with IgG4-SC are predominantly males in their seventh decade and most commonly present with obstructive jaundice, weight loss, and abdominal pain. Patients with concomitant pancreatic involvement can present with steatorrhea, indicative of exocrine insufficiency and/or diabetes [26, 32]. In others, biliary involvement might be found incidentally on cross-sectional imaging performed for another reason.

Patients should be asked about previous occupational exposure, especially “blue-collar work” and history of allergy and/or atopy. Both have been observed at increased rates in IgG4-RD, although their significance in disease pathogenesis remains unclear [16, 17, 38, 39].

Laboratory Findings

There is no single laboratory test that can accurately diagnose IgG4-SC. Liver function tests are often deranged. An obstructive pattern of raised alkaline phosphatase, gamma-glutamyltransferase, and bilirubin is most commonly observed. In addition, patients can also have a polyclonal hypergammaglobulinemia and raised serum IgG.

Serum IgG4 is raised in 70–74 % of patients at time of diagnosis [26, 32]. However, an elevated serum IgG4 is not specific to IgG4-RD and can also be raised in PSC and pancreatobiliary malignancy, which mimic IgG4-SC both clinically and radiologically [11, 12, 58, 70, 94, 95]. Several studies have investigated whether using a higher cutoff value for serum IgG4 increases its ability to distinguish IgG4-SC from PSC or CCA. Using a higher IgG4 value over four times the upper limit of normal increases the specificity or positive predictive value (PPV) to almost 100 % for IgG4-SC. Alternatively when serum IgG4 is raised between one and two times the upper limit of normal, using an IgG4 to IgG1 ratio rather than IgG4 in isolation has been shown to increase PPV and sensitivity for IgG4-SC in Dutch and UK cohorts [8, 70]. However these methods do not detect the group of IgG4-SC patients with a normal serum IgG4.

Serum IgE levels are raised in between 35 and 95 % of IgG4-RD patients. Furthermore, 25–30 % of patients have a peripheral blood eosinophilia. There is conflicting evidence as to whether patients with a history of allergy are more likely to have a raised IgE and/or eosinophilia, compared to nonallergic patients [17, 38, 39, 98].

No autoantibody has been found to be specific to IgG4-SC [76]. The tumor marker CA19-9 can be raised in both pancreatobiliary malignancy and IgG4-SC, making it a poor differentiator between the conditions [26]. Although bile IgG4 levels can be elevated in patients with IgG4-SC compared to other biliary disorders including PSC and CCA, it is not specific [93].

Imaging Features

Imaging alone is unable to make a firm diagnosis of IgG4-SC as features can mimic PSC, CCA, and pancreatic carcinoma. Imaging of the biliary tree via magnetic resonance cholangiopancreatography (MRCP) or endoscopic retrograde cholangiopancreatography (ERCP) can reveal IgG4-SC biliary strictures.

Four patterns of strictures have been recognized [66]. Type 1 describes a single distal common bile duct (CBD) stricture which can mimic pancreatic carcinoma or CCA. This appearance commonly occurs in IgG4-SC, particularly in association with AIP where the stricture may be caused by inflammation of both the pancreas and biliary wall [31]. Type 2 lesions can be divided into type 2a intrahepatic strictures with prestenotic dilatation and type 2b intrahepatic strictures without prestenotic dilatation and reduced bile duct branching. Both type 2 patterns can exhibit additional extrahepatic strictures, and appearances can be similar to PSC (Fig. 5.1). Unlike PSC, IgG4-SC strictures often show biliary dilation of over 10 mm proximal to a confluent narrowing in the distal CBD. Characteristic PSC features such as a beaded and pruned-tree appearance of the bile ducts are often absent in IgG4-SC [65]. A recent study of biliary appearance using MRI found continuous rather than skip lesions, and a single wall CBD thickness of over 2.5 mm favored IgG4-SC over PSC [85].

Fig. 5.1

MRCP of a patient with type 2 IgG4-SC with intra- and extrahepatic biliary dilatation

Type 3 IgG4-SC describes a distal CBD stricture and hilar hepatic stricture. Type 4 strictures involve the hilum only (Fig. 5.2). In a Japanese survey of IgG4-SC patients without pancreatic lesions, this was the commonest subtype [84]. Both type 3 and type 4 can mimic hilar CCA.

Fig. 5.2

MRCP of a patient with type 4 IgG4-SC with a hilar stricture, which is difficult to differentiate from hilar CCA

Other characteristic features of IgG4-SC lesions include symmetrical biliary wall thickening, smooth inner and outer margins, and a homogenous echo appearance of the internal bile duct wall. These can be characterized using conventional abdominal ultrasound, computed tomography (CT), endoscopic ultrasound (EUS), and intraductal ultrasonography (IDUS). Lesions can occur in regions where there is no identifiable biliary stricture on cholangiography [35, 50, 62]. Cross-sectional imaging can identify mass lesions in other organs caused by systemic IgG4-RD. CT pancreas can show a characteristic sausage-shaped appearance or mass lesions within the pancreatic parenchyma representative of AIP [35]. In one series, pancreatic abnormalities were the strongest predictor of correctly distinguishing IgG4-SC from PSC and malignancy [25].

Histopathological Features

Inflammatory lesions in IgG4-SC are usually distributed in the extrahepatic, hilar, and perihilar bile ducts but can also affect the small intrahepatic ducts and gallbladder.

Macroscopically the affected areas of the bile duct are diffusely thickened, with stenotic lumens, and in some cases appear as tumorous lesions [64, 100, 101]. In contrast to PSC, the biliary epithelium is relatively well preserved but inflammation can extend into local veins, glands, and nerves [99].

Microscopically, classical IgG4-SC lesions share the lymphoplasmacytic infiltrate, obliterative phlebitis, and storiform pattern of fibrosis seen in other IgG4-RD conditions [26, 99]. The lymphoplasmacytic infiltrate is T-Cell predominant with scattered B-cell aggregates (Fig. 5.3 left). Germinal centers are sometimes seen and many specimens have an eosinophilia. The presence of IgG4-positive plasma cells, however, is not sufficient for diagnosis, as they can be seen in other conditions. A biopsy specimen with a mean of >10 IgG4 plasma cells per high-power field (HPF) (Fig. 5.3 right) or an IgG4/IgG plasma cell ration of >40 % is suggestive and incorporated into diagnostic guidelines for IgG4-RD and IgG4-SC [15, 68, 75]. It should also be noted that some classical histopathological features might not be present on biopsy if insufficient amounts of tissue are obtained. In one series of transpapillary biopsy specimens collected from IgG4-SC strictures using IDUS, obliterative phlebitis was absent and >10 IgG4-positive cells per HPF was only observed in a minority [62].

Only gold members can continue reading. Log In or Register to continue