cholangiohepatitis

.]. Comp. Path. t998, Vol. 119. 135 147

I m m u n o h i s t o c h e m i c a l Characterization of the Lesions of Feline Progressive Lymphocytic Cholangitis/Cholangiohepatitis

J

M.J. Day 1)epartment of Patkolog~ and MicrobioIogy, [~iversity of Bristol, LangfordBS18 7DU, Uniled Kingdom Summary The histopathological features of liver biopsies from 20 cats with progressive lymphocytic cholangitis/cholangiohepatitis are reported. These biopsies were subject to immunohistochemical investigation for expression of CD3, CD79, Major Histocompatibility Complex (MHC) Class ii molecules, and tkline IgG, tgM and IgA. Livers from five normal cats, which were also examined, showed constitutive expression of MHC Class II by sinusoidal Kupffer cells and bile duct epithelium, in addition to a population of portal, and bile duct inter-epithelial, CD3 + T lymphocytes. In liver biopsies from cats with the active phase of lymphocytic cholangitis/cholangiohepatitis (n = 11), the portal lymphocytes were predominantly CD3 + T cells that infltrated bile duct epithelium and periportal hepatic parenchyma. CD79 + B lymphocytes formed distinct aggregates or tbllicles within the regions of T-cell infiltration. Low numbers of plasma cells were present, and these predominantly expressed IgA. MHC Class II was expressed by Kupffer cells, infiltrating T and B lymphocytes and macrophages. There was membrane and cytoplasmic Class II expression by bile duct epithelium, some vascular endothelium, and fibroblasts within areas of fibrosis. In liver biopsies from cats with chronic lymphocytic cholangitis/cholangiohepatitis (n=9), there was less inflammation, but the composition of the infiltrates was similar to that in the active phase of disease. The findings provide further evidence for an immunemediated pathogenesis in progressive lymphocytic cholangitis/cholangiohepatitis.

9 1998 W.B. Saunders Company Limited

Introduction Progressive lymphocytic cholangitis/cholangiohepatitis (PLCC), first described in a series of 21 cats in the United Kingdom (Lucke and Davies, 1984), remains an important cause of feline hepatic disease in this country. It occurs predominantly in younger cats (4 years and under), and the most common clinical features are ascites, jaundice and hypergammaglobulinaemia. This disease is diagnosed histopathologically and there is progression through two distinct stages of microscopical pathology. In the active stage, there is marked lymphocytic inflammation of portal tracts, particularly surrounding and infiltrating bile ducts, with occasional extension to periportal hepatic parenchyma ("piecemeal necrosis"). This is accompanied by bile duct proliferation and 0021 9975/98/(/60135 + [ 3 $I 2.0(//0

9 [998 W.B. Saunders C(nnpally Limilrd

136

M.J. Day

early portal-to-portal bridging fibrosis. The chronic stage is characterized by prominent monolobular fibrosis but also by a reduction in the intensity of lymphocytic infiltration and extent of bile duct proliferation. The aetiology of the disease is unknown. A genetic predisposition was suggested by the over-representation of Persian cats in the definitive study, and an immune-mediated pathogenesis has been proposed. Preliminary studies failed to reveal evidence of mitochrondrial, smooth muscle, thyroid or parietal cell autoantibodies, although antinuclear antibody was demonstrated in a single case (Lucke and Davies, 1984). Concurrent pancreatic inflammation is not a feature of feline PLCC. The clinical presentation and histopathological appearance of PLCC are distinct from those of other forms of feline inflammatory liver disease, which have recently been reviewed (in reports of studies by Gagne et al. [1996] and Weiss et al. [1996] ). In these studies, two histopathological variants of "portal hepatitis" were defined, and the major clinical signs in affected cats were anorexia, weight loss, lethargy and vomiting. Lymphocytic portal hepatitis, which occurred mainly in cats over 15 years of age (Weiss et al., 1996), was characterized by lymphoplasmacytic infiltration of portal areas, portal fibrosis and bile duct proliferation. There was no cholangitis, involvement ofperiportal hepatocytes or monolobular fibrosis, although the disease was reported to be progressive. By contrast, in cholangiohepatitis there was neutrophil-dominated inflammation of portal areas and bile ducts, which sometimes extended through the limiting plate to the hepatic parenchyma. This form of disease occurred most frequently in cats aged 11 15 years, and was associated with concurrent inflammatory bowel disease or pancreatitis (Weiss et al., 1996). In the proposed pathogenesis of cholangiohepatitis, an ascending bacterial infection of the biliary system is thought to lead to secondary immune-mediated disease after bacterial elimination (Zawie and Garvey, 1984). In the chronic stages of cholangiohepatitis (mixed cholangiohepatitis), there may be infiltration by lymphocytes and plasma cells, with fibrosis and bile duct proliferation, that progresses to monolobular fibrosis. There is a paucity of information on feline inflammatory liver disease and the aetiopathogenesis of these disorders is poorly understood. The aims of the present study were to define the nature and distribution of the lymphocytic infiltrate in that form of the disease recognized in the United Kingdom (PLCC), and to determine whether local expression of Major Histocompatibility Complex (MHC) Class II molecules has a role in the pathogenesis of this disease. Materials and Methods Case Material

Formalin-fixed, paraffin wax-embedded tissues from 20 cases of feline PLCC were extracted from the archive of the Comparative Pathology Laboratory, Department of Pathology and Microbiology, University of Bristol. Tissues, which were selected on the basis of suitability for serial sectioning for immunohistochemistry, included six necropsy and 14 wedge biopsy samples. Needle-core biopsies could not be serially

Feline Lymphocytic Cholangiohepatitis

137

sectioned and wcre therefore excluded. Samples were collected between August 1984 and June 1997, Five samples were submitted from referral clinics at the University of Bristol, and 15 samples were submitted from veterinary practices in the Bristol area. A haematoxylin and cosin (HE)-stained section was prepared from each block tbr histopathological examination, and unstained sections (4pm) were mounted on poly-l-lysine-coated slides for imnmnohistochemical examination. Control sections of normal feline liver were obtained from tive cats submitted for routine necropsy. These animals had no evidence of hepatic disease and sections of liver had normal histological structure. Sections of normal feline lymph node and spleen were similarly obtained and included as positive controls lbr each antibody used in the immunohistochemical studies.

ImmunohistochemisOy Sections from each control and test tissue were stained with a panel of cross-reactive antisera specific for the pan-T lymphocyte marker CD3, the pan-B lymphocyte and plasma cell marker CD79, and M H C Class I1. Antisera specific for feline IgG (Fc specific), t~line lgM (Fc specific) and t~line IgA (Fc specific) were also used. Negative control sections were incubated with normal mouse, rabbit or goat serum in place of specific antibodies. The cross-reactive antisera specific [br conserved peptide sequences of CD3 and CD79 (.]ones et al., 1993) have been validated previously for use in the cat (Day, 1995; Callanan et al., 1996) and were confirmed to identify appropriate areas of feline lymph node and spleen in this study (data not shown). The crossreactive antiserum specific tbr M H C Class II, which has been validated for use in the dog (Day, 1996), reacted similarly with feline macrophages, dendritic cells, and T and B lymphocytes in the tissues examined here (data not shown). Sections tbr immuimhistochemistry were pre-treated by passage through graded alcohol and exposure to hydrogen peroxide (0'5%) in methanol to block endogenous peroxidasc activity. Antigen unmasking to facilitate staining tbr CD3 or feline immunoglobulins was accomplished by incubation with 0'1% trypsin tbr 45 rain at 37~ Antigen unmasking for sections to be stained for CD79 and M H C Class II was accomplished by exposure to microwaves (7 min, medium power, 650W microwave oven) after immersion in citrate bufl~w (10raM, pH 6"0). Immunohistochemical staining for CD3 was carried out by serial incubation at room temperature with normal goat serum (diluted 1 in 5 in phosphate-buffered saline [PBS] 0"01M, pH 7-4; 30 rain), rabbit anti-human CD3 peptide (Dako, Glostrup, Denmark; 1 in 200 in PBS; 4h), biotin-labelled goat anti-rabbit IgG (Sigma, Poolc, Dorset; 1 in 20 in PBS; 30 rain) and avidin-peroxidase complex (Sigma; 1 in 20 in PBS; 30 min). In all immunohistochemical procedures, sections were washed twice (15 min) in PBS between each incubation, and antibody labelling was "visualized" by means of diamiuobenzidine (l)ako) with peroxide, bclbre a light counterstain with Mayer's hacmatoxylin. Immunohistochemical staining fur CD79 or M H C Class II necessitated serial incubation with normal goat serum (as above), murine monoclonal anti-human CD79 peptidc (1)ako; 1 in 100 in PBS; 4h) or murine monoclonal anti-human M H C Class II (Dako; 1 in 20 in PBS; 4h), followed in each case by biotin-labelled goat antimouse IgG (Sigma; 1 in 20 in PBS; 30 rain) and avidin-peroxidase complex as above. Feline IgG staining necessitated serial incubation with normal goat serum (as above), rabbit anti-cat IgG (Nordic Laboratories, Tilburg, The Netherlands; 1 in 50 in PBS; 30 min) and peroxidase-conjugated goat anti-rabbit IgG (Sigma; 1 in 100 in PBS; 30 rain). Staining for IgM or lgA was carried out by incubation with normal rabbit serum (1 in 3 in PBS, 30 rain), goat anti-cat lgM or goat anti-cat IgA (Nordic Laboratories; both 1 in 50 in PBS; 30 min) and peroxidase-conjugated rabbit antigoat IgG (Sigma; 1 in 200 in PBS; 30 rain).

138

M.J. Day

Table 1 Clinical features of cats with progressive lymphocytic cholangitis/cholangiohepatitis (PLCC)

Case

Age

Breed

Sex

Clinical features

Duration of clinical

~ears)

s~gn~

1 2 3

3 9 12

DSH DSH DSH

FN FN F

4 5 6 7 8 9 10

4 6 2 l '25 2 3 4

DSH DSH DSH DSH DLH DSH DSH

M FN M F NR NR MN

11 12 13

8 5 3

DSH DSH Persian

MN M M

14

3

NR

MN

15

6

DSH

MN

16 17 18 19 20

1'25 4 6 0'5 NR

NR Persian DSH Siamese DSH

F MN MN M MN

Icterus, ascites Raised ALP and ALT, hepatomegaly Inappetance, polydipsia, weight loss, icterus, raised ALP and A L T Weight loss Intermittent icterus Grand mal seizure, unresponsive to medication NR Anorexia, raised ALP and A L T Ventral oedema, ascites, hypoalbuminaemia Hepatomegaly, icterus, weight loss, dullness, raised ALP and ALT Icterus, hepatomegaly Recurrent episodes of ascites Icterus, acute collapse, nervous symptoms, death Lethargy, inappetence, weight loss, icterus, raised ALP and ALT Polydipsia, hepatomegaly, raised ALP and ALT Lethargy, weight loss, hepatmnegaly Icterus Icterus, hepatomegaly Stunted since birth, acute collapse Ascites

4 months 4 weeks 5 months 2 years

4 months 2 weeks 3 months

2 weeks

NR, not recorded; DSH, domestic short haired; DLH, domestic long haired; M, male; MN, neutered male; F, female; FN, neutered female; ALP, serum alkaline phosphatase; ALT, serum alanine aminotransferase. Specific values for serum ALP and ALT were not recorded on the submission forms, but in each case elevations were presumed to be significantly outwith the normal reference range for the cat, as determined by the testing laboratory.

Results Clinical Features of 20 Cats with PLCC

The cats (Table 1) ranged in age from 6 months to 12 years (mean 4"4___2"8 years; n = 19) and the population included 14 domestic short-haired cats, one domestic long-haired cat, two Persian cats and one Siamese (breed not recorded in two cases). There were six female (three neutered) and 12 male (seven neutered) cats (sex not recorded in two cases). The major clinical signs (Table 1) were recorded in 19 cases and included: icterus (nine cases), ascites (four cases), hepatomegaly (six cases), raised liver enzymes (six cases), and weight loss (five cases). Three cats died suddenly, and two of these had neurological signs. The duration of clinical signs, recorded in eight cases, ranged from 2 weeks to 5 months. Immunohistochemical Features of Normal Feline Livers

In normal feline liver, small to moderate numbers of CD3 + T lymphocytes were found in portal areas and throughout sinusoids. The portal T cells

Feline Lymphocytic Cholangiohepatitis

Fig. 1, Fig. 2. lqg. 3. Fig. 4.

139

Portal tract from normal feline liver showing CD3 + T lymphocytes between epithelial cells of bile ducts (inter-epithelial T lymphocytes). Avidin-biotin imnmnohistochemistry, anti-CD3 peptidc. x 100. Intense lymphocytic infiltration of portal area. Active stage of progressive lymphocytic cholangitis/ cholangiohepatitis (PLCC) (Gase 4). HE. x 60. Porlal fibrosis, bile duct proliferation and associated ncutrophil infiltration, extending through limiting plate. Chronic stage of PI,(:(; (Case 14). HE. x 75. CD3 + T lymphocytes surrounding and infiltrating bile duct epithelium. Active stage of PLCC (Case [ l). Avidin-biotin immunohistochcmistry, anti-CD3 pcptide, x 125.

were often closely associated with bile ducts, either immediately beneath the basement membrane or between bile duct epithelial cells (inter-epithelial lymphocytes) (Fig. 1). These T cells expressed surface membrane MHC Class

140

M.J. Day Table 2 Histopathologieal features of liver from cats with PLCC

(,?l,~e

&mple type Portal lymphoNeutrophiland Bile duct Portalfibrosis Portal-to- Typeo/'di~ease plasma~ytic inmacrophagein proli/~ration portal fibro.ff,~ flammation and bile .filtration u,ith bile duct i~filtration duct pwliferation

1

PM

+

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

B B B B PM PM B B

+ +++ +++ ++ + + + +

B

+++

-

B B PM B B B B PM PM B

+++ + + + + + + + + +

+

PM, post m o r t e m ; B, biopsy; - ,

++ ++ + .

.

+ ++ ++ ++ +

++ + ++ + ++ + + + . +++ +++ +++ +++ ++ + +++ ++ + +++ +++

++ + + ++ + + + -

++ + + + + ++ + +

++ ++ +++ § ++ + +++ + + +++ +++

++ ++ +++ +++ ++ + +++ +++ + +++ +++

.

Active Active Active Active Active Active Active Active Active Active Active Chronic Chronic Chronic Chronic Chronic Chronic Chronic Chronic Chronic

not present; + , mild; + + , m o d e r a t e ; + + + , m a r k e d .

II. In addition, there was strong constitutive expression of M H C Class II (cytoplasmic) by sinusoidal Kupffer cells, and in two normal livers some large bile ducts had granular, intracytoplasmic expression of M H C Class II. B lymphocytes (CD79 § were not found in normal liver, but in two sections there were low numbers of CD79 § IgA plasma cells in some portal areas. In sections stained for immunoglobulin, there was outlining of the space of Disse and positive staining of secretion within the lumina of larger bile ducts.

Histopathological Features of Livers with PLCC The major histopathological features of liver from 20 cats with PLCC are presented in Table 2. Sections of liver from 11 cats had microscopical features consistent With the active stage of disease. In these cases there was lymphocytic inflammation of portal areas with infiltration and destruction of bile ducts, in addition to varying degrees of extension of the inflammatory process through the limiting plate to periportal parenchyma (Fig. 2). The inflammatory process was accompanied by mild to moderate portal-to-portal fibrosis with bile duct proliferation in nine of 11 cases. Bile duct proliferation was accompanied by local infiltration of neutrophils and macrophages in four cases. Sections of liver from nine cats had microscopical features consistent with the chronic stage of lymphocytic cholangitis/cholangiohepatitis. Lymphocytic inflammation of portal areas was restricted to small aggregates of cells surrounded by a dense fibrous connective tissue matrix. These lymphocytic

Feline

Summary

Ca.~e

Lymphocytic

Table 3 examination

of immunohistoehemical

CD3 + CI)79- B-cell T (ell~ agt~re~ale~ pre.~enl

1

+ + +

2 3 4

+ + + + + + + +

5 6

+ + + +

7 8 9 10 11

+ + + + + + + + + + + +

12 13 lJ 15 16 17 I8 19 20

+ + -- q+ + + + + + + + + + + +

Yes, also SmlgM" No Yes Yes, also S m l g M + or SmlgG + No Yes, also SmIgG + No Yes Yes Y('s Yes, also S m l g G + or SmlgM + No No Yes Yes Yes Yes Yes Yes Yes

IgG p[arma celh

141

Cholangiohepatitis

o f l i v e r t i s s u e f r o m 20 c a t s w i t h P L C C

Ig3d pla,~ma IgA plasma celb cels

Bile duct

membrane

Bile duct cytoplasmic

A.IHC Clms H ;~IHC Clas~ H

present

presenl

+

+

+ +

Yes

Yes

+ + +

+ + +

+ + + +

No No No

Yes Yes Yes

+ -

+

+ +

No Yes

No Yes

+ + + -

Yes No Yes No No

Yes No Yes Yes No No No Yes No

-

+

+ +

No No Yes No No

+ + + + + + + + + + +

+ + + + + + + + + + +

No No Yes No No No No Yes No

+

+ +

+ + +

, No cellular staining; + , scaltered individual cells; + + , low to m o d e r a l e n u m b c r o f cells; + + + , large n u m b e r o f cells.

aggregates were associated with degenerate bile ducts, with less clear evidence of bile duct infiltration than in the active phase of disease. The lymphoid cells were generally contained by the connective tissue matrix and did not infiltrate the hepatic parenchyma. The major feature of these sections was marked monolobular fibrosis and bile duct proliferation. The latter was accompanied by local infiltration of neutrophils and macrophages in five cases, and in two of these cases the inflammatory cells extended into hepatic parenchyma away from the area of bile duct proliferation (Fig. 3). In four of six cats subjected to necropsy, some details of other viscera were available. In all fbur animals the pancreas and intestinal tract were recorded as being apparently normal, and this was confirmed by microscopical examination in two cases. Immunohistochemical t~?atures of Livers with PLCC The major features arc recorded in Table 3. The inflammatory infiltrates of both active and chronic forms of the disease comprised both T and B lymphocytes in addition to plasma cells. In active disease, there were large

142

M.J. Day

numbers of CD3 + T cells within the portal infiltrates. These cells infiltrated bile ducts (Fig. 4) and actively migrated through the limiting plate into adjacent hepatic parenchyma, where they were observed to cluster around individual hepatocytes, with close membrane apposition (Fig. 5). By contrast, CD79 § B lymphocytes did not infihrate bile ducts or parenchyma, and within portal areas were largely present as aggregates surrounded by T cells (Fig. 6). These B-cell aggregates occasionally had a "follicular" appearance, with a distinct lymphoblastic centre, and in some instances these blastic B cells expressed surface membrane IgM or IgG in addition to CD79. Scattered T cells were found within these follicular areas. In all cases, both T and B lymphocytes expressed surface membrane M H C Class II molecules. T and B cells were occasionally observed within the lumina of sinusoids or dilated portal lymphatic vessels. CD79 + plasma cells with cytoplasmic IgG, IgM or IgA were also found within the lesions of active PLCC. These cells were generally scattered in low numbers throughout the ir/flammatory aggregates, and were particularly associated with the margins of B-cell follicular aggregates. IgA plasma cells were more frequently present than IgG- or IgM-bearing cells. There was no immunoglobulin expression by bile duct epithelium, but in some sections immunoglobulin staining outlined the space of Disse. There was strong, cytoplasmic M H C Class II expression by sinusoidal Kupfl~r cells and macrophages present within inflammatory lesions. Vascular endothelium and fibroblasts within areas of portal fibrosis also expressed this molecule. In addition, epithelium of larger bile ducts occasionally expressed membrane or cytoplasmic MHC Class II, or both (Figs 7 and 8); smaller, actively proliferating ducts were not stained. The chronic lesions of PLCC were characterized by a much reduced inflammatory infiltration, but a similar distribution of cell types was present. CD3 + T cells were largely restricted to aggregates within portal connective tissue, and there was little evidence of parenchymal infiltration. CD79 § B cells formed aggregates or follicles as above, but there was little evidence of concurrent expression of surface membrane immunoglobulin by these cells. Low numbers of scattered plasma cells were more consistently observed within the connective tissue of chronic PLCC, and these were predominantly IgMand IgA-bearing cells. In a single liver (case 19), numerous plasma cells of each class were identified. In areas of bile duct proliferation associated with infiltration of neutrophils and macrophages, there were scattered mixtures of T cells, B cells and plasma cells. M H C Class II expression was observed on sinusoidal Kupffer cells, lymphocytes and macrophages, and this molecule was also expressed by the epithelium of larger (non-proliferating) bile ducts in some cases.

Discussion The clinical signs and liver pathology of the 20 cats studied resembled those of the series of cats with PLCC reported by Lucke and Davies (1984), and further suggest that this disease is distinct from feline lymphocytic portal

Feline Lymphocytic Cholangiohepatitis

Fig. 5.

143

C D - T lymphocytes migrating through the portal limiting plate, in close apposition to periportal

hepatocytcs. Active stage of PLCC (Case 9). Avidin-biotin immunohistochemistry, ami-CD3 peptide. • 500. Fig. 6. Aggregates of CD79" B lymphocytes within an area of T-cell infiltration. Active stage of PLCC (Case 10). Avidin-biotin immmmhistodlemistry, anti-CD79 pcptide. • 25. Fig. 7. MHC Class II expression by lymphocytcs and bile duct epithelium within an area of portal fibrosis in the chronic stag(: of PLCC (Case 19). Avidin-biotin immunohistochemistry, anti-MHC Class 11. x 60. Fig. ~-~. MHC Class II expression by bile duct epithelinm and lymphocytes. Active stage of PLCC (Case 1). Avidin-biotin immunohistochemistry, anti-MHC Class II. • 75.

144

M.J. Day

hepatitis or cholangiohepatitis (Gagne et aL, 1996; Weiss et al., 1996). Unlike the latter, PLCC is characterized by ascites and icterus, and is not associated with concurrent intestinal or pancreatic disease, which has been clearly demonstrated in cholangiohepatitis (Weiss et al., 1996). PLCC is a disease of younger cats, whereas both cholangiohepatitis and lymphocytic portal hepatitis occur in older animals (Weiss et al., 1996). A genetic predisposition of the Persian breed to PLCC is suggested from examination of the combined data of the United Kingdom studies (eight of 41 cats; 19.5%). The histopathological features of hepatic biopsies recorded here are consistent with the original descriptions of active and chronic phases of PLCC reported by Lucke and Davies (1984) and differ from descriptions of the lesions of lymphocytic portal hepatitis and cholangiohepatitis (Gagne et al., 1996). With these original criteria, the cats in the present series were readily classified as having active or chronic disease. However, bile duct proliferation appeared more prominent in chronic disease, which was not the case in the original series (Lucke and Davies, 1984). The small foci of neutrophil and macrophage infiltration observed in association with proliferating bile ducts are a recognized feature of PLCC (Lucke and Davies, 1984) and are distinct from the chronic stage of cholangiohepatitis (mixed cholangitis), in which there is neutrophil infiltration of bile ducts (Gagne et aL, 1996). Feline PLCC has a poorly defined aetiopathogenesis, but exposure to an initiating agent may lead to subsequent immune-mediated damage, which may be perpetuated in the absence of the initiating agent (Lucke and Davies, 1984). The nature of this initiating agent remains obscure. There is little evidence for an ascending bacterial infection, as concurrent intestinal or pancreatic disease is not recognized. Affected animals also give negative results in tests for feline leukaemia virus (Lucke and Davies, 1984). To date, evidence for the role of immune-mediated damage has been limited. Cats with PLCC have raised concentrations of serum IgG, but serum autoantibody specific for mitochondria, smooth muscle, thyroid or parietal cells has not been identified (Lucke and Davies, 1984). The major factor suggesting immune-mediated disease has been the nature of the inflammatory infiltrates, dominated by lymphocytes; these appear to target and destroy bile duct epithelium, and occasionally periportal hepatic parenchyma, after migration through the limiting plate. In other species, immune-mediated hepatic diseases, characterized by lymphocytic infiltration of bile ducts or piecemeal necrosis of periportal parenchyma, are well documented. Human immune-mediated cholangiopathies, such as primary biliary cirrhosis or primary sclerosing cholangitis, bear superficial resemblance to feline PLCC (Lucke and Davies, 1984), and some cases of chronic active hepatitis in man have a well-defined autoimmune basis (Vento et al., 1986). Canine chronic liver diseases, including chronic active hepatitis and chronic cholangiohepatitis, may also have an immunemediated pathogenesis (Fuentealba et al., 1997; Poitout et al., 1997). Although titres of serum autoantibodies (antinuclear antibody, anti-liver membrane antibody) are non-diagnostic for such disorders (Andersson and Sevelius, 1992), a recent study has demonstrated a peripheral blood mononuclear cell

Feline Lymphocytic Cholangiohepatitis

145

proliferative response to liver m e m b r a n e antigens in dogs with chronic hepatitis (Poitout et al., 1997). Further evidence of an immune-mediated basis for some forms of" cholangitis and hepatitis has been provided by development of rodent models ibr these diseases. Rats immunized with bile duct epithelia in adjuvant develop lymphocytic cholangitis; this is mediated by cytotoxic T Iymphocytes and may be adoptively transferred to naive recipients by spleen cells from affected rats (Ueno et al., 1996). Similarly, in experimental autoimmune hepatitis, mice immunized with hepatocyte antigens in adjuvant develop T cell-mediated hepatocyte destruction (Lohse et al., 1992). The results of the present study provide further evidence for immunemediated processes in feline PLCC. The disease is characterized by marked portal infiltration of T lymphocytes bearing CD3 and M H C Class II. These T cells infiltrate the bile ducts and penetrate the limiting plate to become closely associated with individual hepatocytes. This "recognition event" may lead to cytotoxic destruction of target cholangiocytes or hepatocytes, or both, by mechanisms that may include induction of target-cell apoptosis. Of" note was the finding of a background population of CD3 + T cells in the liver of normal cats. These cells were located within sinusoids, and within portal areas adjacent to, or within, bile duct epithelium. Such T cells may be analogous to enteric inter-epithelial lymphocytes and be responsible fbr hepatic i m m u n e surveillance, thus having a role in the primary response to the initiating agents of PLCC. It is clear, however, that feline PLCC is not purely a T cell-mediated disease. There are significant populations of B lymphocytes within the hepatic lesions, and in individual cases there are large numbers of plasma cells, expressing a range of immunoglobulin classes. There is distinct compartmentalization ot" the B cells within the portal infiltrates; these cells form aggregates of small lymphocytes or follicles with tymphoblastic central regions. Such inflammatory follicles are documented in many inflammatory disorders, and lymphoid fbllicles were recorded in the lesions of feline lymphocytic cholangitis by Prasse et al. (1982). Although the B cells and plasma cells do not directly contact bile duct epithelium or migrate through the limiting plate to contact hepatocytcs, there may be an indirect role for local antibody production via the mechanism ot" antibody-dependent cell-mediated cytotoxicity (Cochranc el aL, 1976). Cats with PLCC have reactive lymphadenomegaly and serum hypergammaglobulinaemia, further suggesting activation of the B-cell compartment in this disease. The dominance of IgAbearing over IgG- and IgM-bearing plasma cells in these lesions is of note, and may reflect the role of the liver in the metabolism and circulation of IgA. Feline bile is rich in IgA (Barlough et al., 1981), but there is little documentation of pathways of IgA circulation in the liver of the cat. The present study suggests a role for regional expression of M H C Class II in the pathogenesis of this disease. There is strong constitutive expression of M H C Class II by feline KupiTbr cells and occasional expression of this molecule within the cytoplasm of normal feline bile-duct epithelia. This suggests a role for the bile duct epithelium in local antigen presentation and i m m u n e surveillance of the liver. In PLCC there is "up-regulation" of M H C Class II

146

M.J. Day

within the liver, analogous to that in the human cholangiopathies (LaRousso et al., 1984). Because feline T and B lymphocytes constitutively express M H C Class II (Rideout et al., 1990), the expression of this molecule is not an indication of lymphocyte activation in the cat. Bile duct epithelium in a number of cases of lymphocytic cholangitis showed both membrane and cytoplasmic expression of M H C Class II, and there was also Class II expression by vascular endothelium, portal fibroblasts and infiltrating macrophages. Contrary to expectation, this study did not identify a difference in the immunological events occurring within the liver of cats with active versus chronic forms of PLCC. Although the overall number of inflammatory cells was reduced in chronic disease, there were still T-cell aggregates, B-cell follicular aggregates and scattered plasma cells of each immunoglobulin class. Moreover, bile duct epithelium continued to express both membrane and cytoplasmic M H C Class II in chronic disease, as in the active form, suggesting that antigen presentation by these cells is maintained in the later stages of disease. In summary, the results add support to the suggestion that feline PLCC has, in part, an immune-mediated pathogenesis. Future studies should further investigate the phenotype of the infiltrating T lymphocytes (CD4 or CD8 expression), the cytokine profile elaborated by these cells, and the antigenic specificity and potential for cytotoxic function of this population.

Acknowledgments The author gratefully acknowledges the contribution of the numerous veterinary surgeons who submitted samples from affected cats, and thanks Dr Vanda Lucke for critical appraisal of the manuscript.

References Andersson, M. and Sevelius, E. (1992). Circulating autoantibodies in dogs with chronic liver disease. Journal of Small Animal Practice, 33, 389 394. Barlough, J. E., Jacobson, R. H. and Scott, F. W. (1981). The immunoglobulins of the cat. Cornell Veterinarian, 71,397 407. Callanan, J. J., Jones, B. A., Irvine, J., Willett, B.J., McCandlish, I. A. P. and Jarrett, O. (1996). Histologic classification and immunophenotype of lymphosarcomas in cats with naturally and experimentally acquired feline immunodeficiency virus infections. VeterinaryPathology, 33, 264-272. Cochrane, A. M. G., Thomson, A. D., Moussouros, A. and Eddleston, A. L. W. (1976). Antibody-dependent cell-mediated (K cell) cytotoxicity against isolated hepatocytes in chronic active hepatitis. Lancet, i, 441 444. Day, M.J. (1995). Immunophenotypic characterization of cutaneous lymphoid neoplasia in the dog and cat. Journal of Comparative Pathology, 112, 79 96. Day, M.J. (1996). Expression of major histocompatibility complex Class II molecules by dermal inflammatory cells, epidermal Langerhans cells and keratinocytes in canine dermatological disease. Journal of Comparative Pathology, 115, 317-326. Fuentealba, C., Guest, S., Haywood, S. and Horney, B. (1997). Chronic hepatitis: a retrospective study in 34 dogs. Canadian VeterinaryJournal, 38, 365 373. Gagne, J. M., Weiss, D.J. and Armstrong, P.J. (1996). Histopathologic evaluation of feline inflammatory liver disease. Veterina~ Pathology, 33, 521 526. Jones, M., Cordell, J. L., Beyers, A. D., Tse, A. G. D. and Mason, D. Y. (1993).

Feline Lymphocytic Cholangiohepatitis

147

Detection of T and B cells in many animal species using cross-reactive antipeptide antibodies, journal oflmmunology, 150, 5429 5435. LaRusso, N. F., Weisner, R. H., Ludwig, J. and MacCarty, R. L. (1984). Current concepts primary sclerosing cholangitis. New Englandjournal of Aledicine, 310, 899 903. Lohse, A. W., Brunner, S., Kyriatsoulis, A., Manns, M. and Zumbuschent~lde, K. H. M. (1992). Autoantibodies in experimental autoimmune hepatitis. Journal of Hepatology, 14, 48 53. Lucke, V. M. and Davies, J. D. (1984). Progressive lymphocytic cholangitis in the cat. journal of Comparative Pathology, 25, 249 260. Poitout, F., Weiss, D.J. and Armstrong, P.J. (1997). Cell-mediated immune responses to liver membrane protein in canine chronic hepatitis. VeterinaryImmunology and Immunopathology, 57, 169 178. Prasse, K. W., Mahafley, E. A., DeNovo, R. and Cornelius, L. (1982). Chronic lymphocytic cholangitis in three cats. Velerina~Pathology, 19, 99 108. Rideout, B. A., Moore, P. F. and Pedersen, N. C. (1990). Distribution of MHC class iI antigens in feline tissues and peripheral blood. 77ssueAntigens, 36, 221 227. Ueno, Y., Phillips, J. O., Ludwig, J., Lichtman, S. N. and LaRusso, N. F. (1996). Development and characterization of a rodent model of immune-mediated cholangitis. Proceedings~the National Academy of Sciences ~the USA, 93, 216 220. Vento, S., O'Brien, C.J., McFarlane, B. M., McFarlane, I. G., Eddleston, A. L. W. and Williams, R. (1986). Lymphocyte T sensitization to hepatocyte antigens in autoimmune chronic active hepatitis and primary biliary cirrhosis. Evidence for different underlying mechanisms and different antigenic determinants as targets. Gastroenterology, 91, 810 817. Weiss, D. J., Gagne, J. M. and Armstrong, P. J. (1996). Relationship between inflammatory hepatic disease and inflammatory bowel disease, pancreatitis, and nephritis in cats. Journal of theAmerican Veterinar~MedicalAssocialion, 209, 1114 1116. Zawie, D. A. and Garvey, M. S. (1984). Feline hepatic disease. Veterinary Clinics of North American Small Animal Practice, 2, 1201 1230.

Received, February9th, 1998] Accepted,April 16th, 1998 J