- Email: [email protected]
Autoimmune hepatitis overlapping with primary sclerosing cholangitis in five cases
THE AMERICAN JOURNAL OF GASTROENTEROLOGY Copyright © 1998 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 93, No. 5, 1998 ISSN 0002-9270/98/$19.00 PII S0002-9270(98)00104-X
Autoimmune Hepatitis Overlapping With Primary Sclerosing Cholangitis in Five Cases Alistair N.B. McNair, Manus Moloney, Bernard C. Portmann, Roger Williams, and Ian G. McFarlane Institute of Liver Studies, King’s College Hospital and School of Medicine and Dentistry, London, United Kingdom
Objective: We report five cases (four male; median age 20 yr, range 14 –38 yr) of an autoimmune hepatitis/ primary sclerosing cholangitis overlap syndrome. The patients presented with jaundice, elevated serum aminotransferase and alkaline phosphatase activities, hyperglobulinemia with high immunoglobulin G (IgG) levels, circulating antinuclear and/or smooth muscle autoantibodies (> 1:40), and moderate to severe interface hepatitis on liver biopsy (with biliary features in four). Methods: All five fulfilled criteria for diagnosis of “definite” autoimmune hepatitis and showed marked responses to prednisolone and azathioprine therapy, with relapses occurring during reduction or withdrawal of treatment. Cholangiographic features of primary sclerosing cholangitis were found in three patients at presentation and after intervals of 7 and 14 yr in the other two. Only two had evidence of inflammatory bowel disease. Diagnostic criteria for identifying those patients who may benefit from immunosuppressive therapy were reviewed. Results: Review of the literature revealed only 11 similar cases that were sufficiently well described for comparison. However, in contrast to these and the present cases, preliminary data from other studies have suggested a marked association with ulcerative colitis and a poor response to immunosuppressive therapy. Conclusions: It is recommended that the possibility of an autoimmune hepatitis/primary sclerosing cholangitis overlap syndrome responsive to immunosuppressive therapy should be considered in any patient presenting with a hepatitic illness with hyperglobulinemia, antinuclear or smooth muscle autoantibodies, and biliary changes on liver biopsy. Cholangiography should be considered in such patients. (Am J Gastroenterol 1998; 93:777–784. © 1998 by Am. Coll. of Gastroenterology)
“hepatitic” pattern of serum biochemical liver tests, hyperglobulinemia with selective elevation of serum immunoglobulin (IgG), circulating autoantibodies, a histological picture of periportal necroinflammation (interface hepatitis) with no biliary changes, and a marked response to corticosteroid therapy (1). Diagnosis is based on these clinical and laboratory features and exclusion of other possible causes of liver disease (2, 3). Recent recommendations are that patients should be categorized as having definite or probable AIH according to several clinical, laboratory, and histological criteria (1). Primary sclerosing cholangitis predominantly affects men and is characterized by elevations of the cholestatic liver enzymes, abnormalities in serum and liver indices of copper metabolism, cholangiolitic features on liver biopsy, and characteristic cholangiographic features in the extrahepatic and/or intrahepatic biliary tracts. It may occur alone or in association with ulcerative colitis (4, 5). The spectrum of bile duct damage in PSC ranges from small duct PSC with histological features of fibrous-obliterative cholangitis and negative cholangiogram to intrahepatic and/or extrahepatic duct involvement with characteristic diffuse, multifocal strictures and segmental dilatation or beading. Purely intrahepatic liver disease seems to be rare. Intralobular hepatic inflammation is not usually considered a feature of PSC. Although there are anecdotal reports of some patients responding to corticosteroid therapy, in general the response is poor. However, some of the histological and immunological features of AIH are occasionally seen in PSC and there have been case reports of the simultaneous occurrence of AIH and PSC (6 –12). In an attempt to clarify the diagnostic criteria, it has been recommended that cases of PSC/AIH overlap should not be considered as part of the spectrum of AIH (1). Nonetheless, an overlap syndrome does exist and there is a need to define criteria for this condition. We report here five cases that illustrate the spectrum of features associated with this interesting syndrome, all of whom fulfilled criteria for definite AIH (1), but also had evidence of PSC.
INTRODUCTION Autoimmune hepatitis (AIH) and primary sclerosing cholangitis (PSC) are generally considered to be two distinct idiopathic liver disorders. Autoimmune hepatitis is a disease that predominantly affects women and is characterized by a
MATERIALS AND METHODS The five cases reported here (Table 1) represent all of those with the features described who had regularly attended our clinic during the past 5 yr and for whom sufficient data
Received July 18, 1997; accepted Jan. 19, 1998. 777
778
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998
TABLE 1 Demographic, Biochemical, Immunological, and Histological Data at Presentation in the Five Cases Case Number 1 Age (yr) Gender Albumin (# 35 g/L) Total globulins (# 35 g/L) IgG (# 18.6 g/L) IgM (# 2.0 g/L) AST (# 50 IU/L) Bilirubin (# 20 mmol/L) ALP (# 120 IU/L) ALP:AST ratio Gamma GT (# 55 IU/L) Autoantibody titers ANA SMA Liver histology Grade Stage
38 M 27 51 38.3 1.7 252 168 1245 2.08 638
2
3
20 F 33 52 29.3 1.7 320 46 1216 1.58 313
1:40 1:40
1:2560 1:80
4 3
2 3
26 M 34 47 NA 2.5 1750 72 156 0.04 NA 1:320 Neg 4 3
4 14 M 29 52 45.0 1.0 996 77 505 0.21 196 1:80 1:160 3 1
5 18 M 35 62 34.4 1.3 304 43 1115 1.52 639 Neg 1:320 0 1
Abbreviations: AST 5 aspartate aminotransferase, ALP 5 alkaline phosphatase, Gamma GT 5 g-glutamyl transferase (upper normal limits in parentheses); ANA 5 antinuclear antibodies (homogeneous pattern), SMA 5 smooth muscle antibodies, NA 5 data not available. Histological grading and staging is as described in Materials and Methods.
were available for this analysis. None had a history of excessive alcohol consumption or of exposure to other potentially hepatotoxic agents, and viral screens (HBsAg, antihepatitis C virus [HCV], IgM anti-HAV, CMV, EBV) were negative in all (performed retrospectively for anti-HCV and HCV-RNA on stored sera where necessary). All were also seronegative for antimitochondrial (AMA) and liver-kidney microsomal (LKM-1) autoantibodies on every occasion when they were tested. Metabolic liver diseases were excluded by appropriate investigations, including routine serum biochemical screening for ferritin, copper, ceruloplasmin, and a-1-antitrypsin phenotype. Other than ulcerative colitis (in two cases) none of the patients or their first-degree relatives had any other diseases thought to have an immunological basis. The histological findings described herein were as reported at the time of liver biopsy and were confirmed on review of all biopsies by one of us (B.C.P.) for the present study. Histological activity was graded for severity of necroinflammation and staged for degree of fibrosis according to Batts and Ludwig (13). According to this scheme, for lymphocytic piecemeal necrosis, grade 0 5 none; 1 5 minimal, patchy; 2 5 mild; 3 5 moderate; 4 5 severe; and for fibrosis stage, 0 5 none; 1 5 portal only; 2 5 periportal; 3 5 septal; 4 5 cirrhosis. Antinuclear (ANA), smooth muscle (SMA), AMA, LKM-1, and perinuclear-staining antineutrophil cytoplasmic (pANCA) autoantibodies were sought by standard indirect immunofluorescence by the hospital’s routine clinical immunology service. Sera were screened at 1:10 dilution and positive samples were titrated out to negativity by doubling dilution.
The scoring system proposed by the International Autoimmune Hepatitis Group (IAIHG) was applied as recommended (1), whereby a diagnosis of definite AIH is made when the aggregate score is . 15 before treatment and . 17 when response to treatment is included in the assessment, and scores of 10 –15 before and 12–17 after treatment are interpreted as probable AIH. Histocompatibility antigen (HLA) phenotype data were included in the scoring assessment of the three patients for whom these data were available. Ratios of serum alkaline phosphatase to aspartate aminotransferase (ALP:AST ratios) were calculated from the degree of elevation of these enzyme activities above the upper normal limits. All patients had serum biochemical liver tests performed on more than one occasion before institution of therapy and results for the day on which the ALP was highest were used for this analysis. For comparison a retrospective analysis of ALP and AST values at presentation was similarly undertaken in 100 consecutive patients with definite AIH (with no evidence of biliary complications) attending our clinic during the same period. Case 1 This man was referred in 1986 with a 4-month history of jaundice and weight loss. Laboratory findings at presentation are shown in Table 1. Liver biopsy showed features of severe hepatitis with extensive bridging and multiacinar collapse (grade 4, stage 3), together with an acute cholangiolitic component with marked ductular proliferation. Two months after starting on prednisolone 50 mg/day serum globulins, AST, and bilirubin had returned to normal but ALP and gammaglutamyl transferase (GGT) remained elevated at 277 and 593 IU/L, respectively. Azathioprine (100
AJG – May 1998 mg/day) was introduced and the steroids were gradually reduced to a maintenance dose of 7.5 mg/day. Within the following year all of his liver tests returned to normal and a liver biopsy showed inactive disease with severe bridging fibrosis (grade 0, stage 3). He experienced relapses in 1989 and again in 1990 during attempts to reduce immunosuppressive therapy. On the second occasion he had abdominal pain and the cholestatic liver enzymes remained abnormal (GGT 862–1460 IU/L and ALP 250 –714 IU/L) after AST returned to normal. Endoscopic retrograde cholangiopancreatography (ERCP) revealed shortening and irregularity of the intrahepatic bile ducts compatible with primary sclerosing cholangitis (PSC). The gallbladder was noted to be grossly enlarged and an elective cholecystectomy was performed. Gallbladder histology revealed chronic cholecystitis. In 1992, he developed bloody diarrhea and colonoscopic biopsies showed features of ulcerative colitis. His condition deteriorated over the next few months, with increasing jaundice and encephalopathy. He died before liver transplantation could be performed. Postmortem liver histology showed a marked periportal and lobular hepatitis (grade 4, stage 3) with ductular proliferation and prominent bridging collapse. Case 2 This woman presented in 1994 with a 4-week history of painless jaundice and pruritus, with no other significant findings on examination. Ultrasonography revealed an apparently normal liver with no evidence of bile duct dilatation. Liver biopsy showed a chronic hepatitis (grade 2, stage 3) with biliary epithelial damage and attenuation, periductular fibrosis, severe ductopenia, and copper-associated protein deposits typical of stage III PSC. ERCP revealed abnormal narrowing and stricturing of the extra- and intrahepatic biliary tree (Fig. 1). There was no evidence of ulcerative colitis on rectal biopsy. Serum bilirubin, AST, and IgG normalized within 1 month of starting on prednisolone 20 mg/day, azathioprine 75 mg/day, and ursodeoxycholic acid 300 mg/day, but ALP and GGT remained elevated at 200 –300 IU/L. She had two biochemical relapses 15 and 27 months later, during attempts to reduce the steroids , 10 mg/day, but has remained well over the past 18 months taking this dose (with 75 mg/day azathioprine). Her AST, bilirubin, and IgG have normalized but ALP (196 IU/L) and GGT (270 IU/L) remain elevated and she is still ANA positive (1:640). Case 3 In 1976, this man was referred with a 4-month history of dyspepsia and fluctuating jaundice. Liver biopsy showed severe chronic hepatitis with piecemeal necrosis and bridging fibrosis (grade 4, stage 3) (Fig. 2A). Treatment with prednisolone 15 mg/day and azathioprine 75 mg/day led to normalization of all biochemical parameters and he became ANA negative. Liver biopsy 1 yr later showed only minimal residual activity with some scarring (grade 0, stage 2). He
AIH/PSC OVERLAP
779
FIG. 1. Endoscopic retrograde cholangiopancreatograph in Case 2, showing irregularities of intra- and extrahepatic bile ducts suggestive of primary sclerosing cholangitis, typical of the findings in the other four cases.
relapsed in 1980 during an attempt to withdraw all treatment. Liver biopsy at this time showed severe chronic hepatitis with prominent lobular activity (grade 4, stage 3) and mild cholangitic features. Within 4 months of increasing the steroids to 10 mg/day and reintroducing azathioprine (75 mg/day), all liver tests were again completely normal and follow-up liver biopsies (in 1982 and 1983) revealed only grade 1 chronic hepatitis with stage 3 fibrosis, with an occasional cholangiolitic focus. In 1984, he defaulted therapy and relapsed again. Remission was regained on reintroduction of therapy, but ALP and GGT remained slightly elevated (, twice upper normal limits). A liver biopsy in 1988 showed mild patchy piecemeal necrosis (grade 2, stage 3) and, for the first time, features of severe (but focal) destructive cholangitis and pericholangitis (Fig. 2B). A further biopsy in 1991 showed significant deterioration, with extensive postnecrotic bridging fibrosis and predominantly cholangitic features, including marked inflammation around septal-sized bile ducts (Fig. 2C). ERCP revealed a grossly abnormal intrahepatic biliary tree with reduced-caliber bile ducts and multiple areas of narrowing and irregularity consistent with PSC. He currently remains asymptomatic on 75 mg/day azathioprine alone. Case 4 This patient presented in 1981 (aged 14 yr) with a 6-wk history of right upper quadrant abdominal discomfort and fluctuating jaundice. Liver biopsy showed moderate to severe periportal necroinflammation with prominent lobular
780
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998
FIG. 2. Liver histology in Case 3. (A) Initial biopsy at presentation, showing extensive bridging parenchymal collapse and fibrosis with prominent interface hepatitis. (B) Twelve yr later. Note expanded and bridged portal areas showing severe cholangitic inflammation with bile duct damage and destruction (arrows). (C) Three yr after (B), showing marked chronic inflammation around larger, septal-sized bile duct branches.
AJG – May 1998
AIH/PSC OVERLAP
781
FIG. 2. Continued.
inflammation and ductular proliferation (grade 3, stage 1). Prednisolone therapy (80 mg/day, reducing to 10 mg/day over a 6-month period) led to good clinical and biochemical response. Liver biopsy in 1982 showed mild portal and lobular inflammation with minimal piecemeal necrosis (grade 2), focal cholangiolitis, mild ductular proliferation, and stage 2 fibrosis. His AST had risen to 150 IU/L, so azathioprine (50 mg/day) was added, leading to a reduction in the AST (, twice upper normal limit) but with continuing marked elevations of the cholestatic liver enzymes (ALP 210 – 665 IU/L, GGT 163–399 IU/L, and IgG 17.8 –35.3 g/L), and with ANA rising to 1:640. In 1988, he developed diarrhea. Colonoscopic biopsies were consistent with ulcerative colitis, and liver biopsy showed portoseptal fibrosis with pericholangitis, ductular proliferation, and copper-associated protein deposition indicative of a chronic pathological process involving the bile ducts. ERCP revealed changes compatible with intrahepatic PSC. Immunosuppressive treatment was gradually withdrawn. In 1992, he relapsed with elevated AST and biliary enzymes and high titers of ANA. Prednisolone and azathioprine were restarted with improvement in liver function and he is currently asymptomatic (for both his liver and bowel diseases), but ALP and GGT have remained elevated (at 227 and 600 IU/L, respectively). Case 5 This male patient presented in 1995 with a 3-month history of upper abdominal discomfort (associated with a
short-lived episode of diarrhea), weight loss and abnormal liver tests (Table 1). Liver biopsy showed a mixed portal inflammatory infiltrate with florid cholangitis and pericholangitis (grade 0, stage 1), and ERCP demonstrated changes compatible with PSC affecting predominantly the intrahepatic biliary tree but with some extrahepatic changes. Colonoscopy and colonic biopsies were normal. A trial of ursodeoxycholic acid (300 mg b.i.d.) led to no significant improvement in his liver biochemistry after 1 month and he was started on prednisolone at 30 mg/day. Within 1 month the ALP fell to 469 IU/L and AST to 55 IU/L. An attempt to reduce his steroids led to a relapse, so azathioprine was added (75 mg/day, increasing later to 150 mg/day) and prednisolone gradually reduced to 10 mg/day. After 18 months on this regimen, he is asymptomatic with normal IgG, ALP, bilirubin, and AST but his GGT is still elevated at 136 IU/L. A recent liver biopsy showed portal fibrosis with mild residual inflammatory infiltrate in the portal tracts (grade 0, stage 1) but no evidence of the previously noted cholangitic changes. RESULTS These five patients were mainly young (median age 20 yr, range 14 –38 yr), predominantly male (four cases), and presented with constitutional symptoms and jaundice. Serum AST activities were elevated . 5 times the upper normal limit in all and ALP . fourfold in four cases (Table 1). All had hyperglobulinemia with . 1.5-fold selective
782
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998 TABLE 2 Additional Data on the Five Cases at Presentation and During Follow-up Case Number
HLA phenotype Ulcerative colitis pANCA AIH score At presentation After treatment Follow-up (yr)
1
2
3
4
5
B8-DR3 Yes Positive
ND No Positive
A1-B8-DR3 No Positive
A1-B8-DR3 Yes Positive
ND No Positive
17 22 7
16 18 3
18 23 21
20 25 16
16 21 2
pANCA 5 perinuclear staining antineutrophil cytoplasmic antibodies, ND 5 not determined.
elevations in serum IgG concentrations and ANA (homogeneous pattern) and/or SMA titers $ 1:40. All were pANCA positive and were seronegative for AMA. The three who had been HLA phenotyped had the B8-DR3 haplotype, with two having the full extended haplotype A1-B8-DR3 (Table 2). Liver histology at presentation showed periportal lymphocytic piecemeal necrosis of varying severity in all but one, with marked cholangiolitic changes in three cases. Of the two patients without pronounced biliary features at presentation, one (Case 4) had two biopsies during the first 2 yr that showed mild biliary changes, but it was only when he developed ulcerative colitis 7 yr later that significant histological and cholangiographic features compatible with PSC were seen. The other (Case 3) was diagnosed as having definite AIH and has been managed as such throughout the 21 yr of his disease. During that time he has had seven liver biopsies. The first indication of biliary complications came with the finding of occasional focal cholangitic changes in his third biopsy (4 yr after the first). Three further biopsies (at 6, 7, and 12 yr after presentation) showed similar focal biliary changes and it was not until 15 yr after the initial diagnosis that histological and cholangiographic features suggestive of PSC were observed. All patients showed marked clinical and biochemical responses to treatment with prednisolone and azathioprine, and four had relapses during subsequent reductions or withdrawal of therapy. Histological improvement during treatment was also demonstrated in these four cases, although progression of the biliary lesions was evident. In the fifth case, follow-up biopsy after 1 yr on immunosuppressive therapy revealed no evidence of intrahepatic biliary disease. Application of the IAIHG scoring system for features at presentation revealed that all five cases achieved scores above the recommended minimum for a diagnosis of definite AIH, and this was reinforced when subsequent response to treatment was taken into account (Table 2). One feature of this system is the assignment of a negative score of 3 points for ALP:AST ratios . 3.0. All but one had serum ALP activities . 500 IU/L, but AST activities were also significantly elevated and none had a ratio . 3.0 (Table 1). Comparative analysis of the 100 consecutive cases of def-
inite AIH revealed that nine had ALP activities . 500 IU/L (median 625, range 500 –1280) but only one of these had a ratio . 1.5 (at 1.62). Of the remaining 91 (with ALP activities , 500 IU/L), all had ALP:AST ratios , 3.0 and only four had ratios . 1.5 (1.53, 1.60, 2.05, and 2.20, respectively). DISCUSSION The five cases described here do not conform to a clearly defined category of liver disease. All met criteria for definite autoimmune hepatitis (AIH) but had histological and radiological features of primary sclerosing cholangitis (PSC). In addition, they all showed a marked response to immunosuppressive therapy, which is usually of little benefit in PSC (5, 14). In these and other respects they appear to be distinct from the well-recognized overlap between PBC and AIH, of which there have been numerous reports (15, 16 –22). The overlap between PSC and AIH has been reported much less frequently (6 –12), although distinguishing between the two conditions in children can be particularly difficult (23–25), and there is no consensus on presenting features or whether corticosteroids are of benefit. Preliminary data on eight cases from the Mayo Clinic (10) indicated a significantly poorer response to corticosteroids than in patients with definite AIH, and an almost invariable association with ulcerative colitis. Other reports (6 –9, 11, 12) have documented a good response to immunosuppressive therapy and a less marked association with colitis, as in the present study. Because it is clear, however, that some patients with PSC/AIH do respond very well to immunosuppressive therapy, a method for identifying them is required. Titers of ANA or SMA, serum concentrations of IgG, or other immunological parameters do not seem to be particularly useful discriminators in this situation. In a recent analysis of 114 adults with PSC, Boberg et al. (26) reported that serum IgG was elevated in . 60% (although only 15% had levels . 1.5 times the upper normal limit) and approximately 20% had ANA or SMA at titers $ 1:40, but only two patients fulfilled criteria for definite and two for probable AIH (both of whom responded to immunosuppressive ther-
AJG – May 1998
AIH/PSC OVERLAP
apy). pANCA is associated with PSC (27) but these antibodies reportedly also occur in up to 90% of patients with AIH (28 –31). Similarly, the HLA B8-DR3 haplotype, found in the three present cases who were HLA-typed and in the three cases reported by Gohlke et al. (12), is associated with both AIH and PSC. The only presenting feature that distinguished the present cases from classical AIH was the finding of cholangitic changes on liver biopsy, and this was only in three cases. Of the remaining two patients, liver biopsy in one (Case 4) revealed ductular proliferation with no obvious bile duct damage and the other (Case 3) had no histological evidence of biliary disease. It was only some time after the initial diagnosis of AIH that liver biopsies in these two patients showed evidence of definite biliary involvement and that features of PSC were revealed on ERCP. It may be that their disease progressed with time from a pure AIH towards a mixed picture of AIH/PSC. Indeed, of the 11 cases of AIH/PSC overlap reported in the literature who were sufficiently well documented for comparison (6 –9, 11, 12), seven were initially diagnosed as AIH and subsequently found to have features of PSC after intervals of 2–9 yr, three cases presented with features of both conditions, and only one was initially considered to have PSC. A high serum ALP activity might be a useful parameter for raising suspicions of an overlap syndrome because four of the present patients had ALP activities . 500 IU/L at presentation, but also so did 9% of our AIH patients. However, none of the AIH patients with ALP activities . 500 IU/L or ALP:AST ratios . 1.5 subsequently developed any evidence of biliary disease over a median follow-up of 14 yr (range 3–20 yr) and, in contrast to the present AIH/PSC overlap cases, ALP activities rapidly normalized in all during the initial response to immunosuppressive therapy. Nonetheless, we would suggest that the likelihood of an overlap syndrome should be considered in any patient with features of definite or probable AIH who has a high serum ALP activity and/or an ALP:AST ratio . 1.5 and that ERCP should be performed in such patients, particularly if there are any biliary changes evident on liver biopsy. A trial of immunosuppressive therapy, with or without ursodeoxycholic acid, would seem to be justified in such cases. ACKNOWLEDGMENT We thank Dr. Peter Donaldson of our Institute for kindly providing the HLA phenotyping data on these patients. Reprint requests and correspondence: Dr. I. G. McFarlane, Institute of Liver Studies, King’s College Hospital, Denmark Hill, London SE5 9RS, UK.
REFERENCES 1. Johnson PJ, McFarlane IG. Meeting Report: International Autoimmune Hepatitis Group. Hepatology 1993;18:998 –1005. 2. Czaja AJ. Clinical aspects of autoimmune hepatitis in North America. In: Nishioka M, Zeniya M, eds. Autoimmune hepatitis. Amsterdam: Elsevier, 1994:27– 43.
783
3. Ohta Y, Onji M. Clinical aspects of autoimmune hepatitis in Japan. In: Nishioka M, Toda G, Zeniya M, eds. Autoimmune hepatitis. Amsterdam: Elsevier, 1994:45–53. 4. Ludwig J, La Russo NF, Wiesner RH. The syndrome of primary sclerosing cholangitis. In: Popper H, Schaffner F, eds. Progress in liver diseases. Philadelphia: Saunders, 1990:555– 66. 5. Ueno Y, La Russo NF. Primary sclerosing cholangitis. J Gastroenterol 1994;29:531– 43. 6. Minak G, Sutherland LR, Pappas C, et al. Autoimmune chronic active hepatitis (lupoid hepatitis) and primary sclerosing cholangitis in two young adult females. Can J Gastroenterol 1988;2:22–7. 7. Rabinowitz M, Demetris AJ, Bou-Abboud CF, et al. Simultaneous occurrence of primary sclerosing cholangitis and autoimmune chronic active hepatitis in a patient with ulcerative colitis. Dig Dis Sci 1992; 37:1606 –11. 8. Leggett BA, Hallam A, Powell EE, et al. Autoimmune chronic active hepatitis evolving to primary sclerosing cholangitis: evidence for a common pathogenesis. Hepatology 1992;16:191A (abstract). 9. Lawrence SP, Sherman KE, Lawson JM, et al. A 38 year old man with chronic hepatitis. Semin Liver Dis 1994;14:97–105. 10. Roberts S, Poterucha JJ, Gores GJ, et al. Overlap between type 1 autoimmune hepatitis and primary sclerosing cholangitis. Hepatology 1995;22:129A. 11. Wurbs D, Klein R, Terracciano LM, et al. A 28-year-old woman with a combined hepatitic cholestatic syndrome. Hepatology 1995;22: 1598 –1605. 12. Gohlke F, Lohse AW, Dienes HP, et al. Evidence for an overlap syndrome of autoimmune hepatitis and primary sclerosing cholangitis. J Hepatol 1996;24:699 –705. 13. Batts KP, Ludwig J. Chronic hepatitis. An update on terminology and reporting. Am J Surg Pathol 1995;19:1409 –17. 14. Lee YM, Kaplan MM. Primary sclerosing cholangitis. N Engl J Med 1995;332:924 –33. 15. Kloppel G, Seifert G, Lindner H, et al. Histopathological features in mixed types of chronic aggressive hepatitis and primary biliary cirrhosis. Virchows Arch Pathol Anat 1977;373:143– 60. 16. Taylor SL, Dean P, Riely C. Primary autoimmune cholangitis. An alternative to antimitochondrial antibody-negative primary biliary cirrhosis. Am J Surg Pathol 1994;18:91–9. 17. Ben-Ari Z, Dhillon AP, Sherlock S. Autoimmune cholangiopathy: Part of the spectrum of autoimmune chronic active hepatitis. Hepatology 1993;18:10 –5. 18. Michieletti P, Wanless IR, Katz A, et al. Antimitochondrial antibody negative primary biliary cirrhosis: A distinct syndrome of autoimmune cholangitis. Gut 1994;35:260 –5. 19. Colombato LA, Alvarez F, Cote J, et al. Autoimmune cholangiopathy: The result of consecutive primary biliary cirrhosis and autoimmune hepatitis? Gastroenterology 1994;107:1839 – 43. 20. Okuno T, Seto Y, Okanque T, et al. Chronic active hepatitis with histological features of primary biliary cirrhosis. Dig Dis Sci 1987;32: 775–9. 21. Goodman ZD, McNally PR, Davis DR, et al. Autoimmune cholangitis: A variant of primary biliary cirrhosis. Clinicopathologic and serologic correlations in 200 cases. Dig Dis Sci 1995;40:1232– 42. 22. Lacerda MA, Ludwig J, Dickson ER, et al. Antimitochondrial antibody-negative primary biliary cirrhosis. Am J Gastroenterol 1995;90: 247–9. 23. El-Shabrawi M, Wilkinson ML, Portmann B, et al. Primary sclerosing cholangitis in childhood. Gastroenterology 1987;92:1226 –35. 24. Mieli-Vergani G, Lobo-Yeo A, McFarlane BM, et al. Different immune mechanisms leading to autoimmunity in primary sclerosing cholangitis and autoimmune chronic active hepatitis of childhood. Hepatology 1989;9:198 –203. 25. Maggiore G, Veber F, Bernard O, et al. Autoimmune hepatitis associated with antiactin antibodies in children and adolescents. J Pediatr Gastroenterol Nutr 1993;17:376 – 81. 26. Boberg KM, Fausa O, Haaland T, et al. Features of autoimmune hepatitis in primary sclerosing cholangitis: An evaluation of 114 primary sclerosing cholangitis patients according to a scoring system for the diagnosis of autoimmune hepatitis. Hepatology 1996;23:1369 –76. 27. Snook JA, Chapman RW, Fleming K, et al. Anti-neutrophil nuclear antibody in ulcerative colitis, Crohn’s disease and primary sclerosing cholangitis. Clin Exp Immunol 1989;76:30 –3.
784
MCNAIR et al.
28. Hardarson S, LaBrecque DR, Mitros FA, et al. Antineutrophil cytoplasmic antibody in inflammatory bowel and hepatobiliary diseases: High prevalence in ulcerative colitis, primary sclerosing cholangitis, and autoimmune hepatitis. Clin Microbiol Immunol 1993;99:277– 81. 29. Mulder AHL, Horst G, Haagsma EB, et al. Prevalence and characterization of neutrophil cytoplasmic antibodies in autoimmune liver diseases. Hepatology 1993;17:411–7.
AJG – Vol. 93, No. 5, 1998 30. Vidrich A, Lee J, James E, et al. Segregation of pANCA antigenic recognition by DNase treatment of neutrophils: Ulcerative colitis, type 1 autoimmune hepatitis and primary sclerosing cholangitis. J Clin Immunol 1995;15:293–9. 31. Targan SR, Landers C, Vidrich A, et al. High-titer antineutrophil cytoplasmic antibodies in type-1 autoimmune hepatitis. Gastroenterology 1995;108:1159 – 66.
Vol. 93, No. 5, 1998 ISSN 0002-9270/98/$19.00 PII S0002-9270(98)00104-X
Autoimmune Hepatitis Overlapping With Primary Sclerosing Cholangitis in Five Cases Alistair N.B. McNair, Manus Moloney, Bernard C. Portmann, Roger Williams, and Ian G. McFarlane Institute of Liver Studies, King’s College Hospital and School of Medicine and Dentistry, London, United Kingdom
Objective: We report five cases (four male; median age 20 yr, range 14 –38 yr) of an autoimmune hepatitis/ primary sclerosing cholangitis overlap syndrome. The patients presented with jaundice, elevated serum aminotransferase and alkaline phosphatase activities, hyperglobulinemia with high immunoglobulin G (IgG) levels, circulating antinuclear and/or smooth muscle autoantibodies (> 1:40), and moderate to severe interface hepatitis on liver biopsy (with biliary features in four). Methods: All five fulfilled criteria for diagnosis of “definite” autoimmune hepatitis and showed marked responses to prednisolone and azathioprine therapy, with relapses occurring during reduction or withdrawal of treatment. Cholangiographic features of primary sclerosing cholangitis were found in three patients at presentation and after intervals of 7 and 14 yr in the other two. Only two had evidence of inflammatory bowel disease. Diagnostic criteria for identifying those patients who may benefit from immunosuppressive therapy were reviewed. Results: Review of the literature revealed only 11 similar cases that were sufficiently well described for comparison. However, in contrast to these and the present cases, preliminary data from other studies have suggested a marked association with ulcerative colitis and a poor response to immunosuppressive therapy. Conclusions: It is recommended that the possibility of an autoimmune hepatitis/primary sclerosing cholangitis overlap syndrome responsive to immunosuppressive therapy should be considered in any patient presenting with a hepatitic illness with hyperglobulinemia, antinuclear or smooth muscle autoantibodies, and biliary changes on liver biopsy. Cholangiography should be considered in such patients. (Am J Gastroenterol 1998; 93:777–784. © 1998 by Am. Coll. of Gastroenterology)
“hepatitic” pattern of serum biochemical liver tests, hyperglobulinemia with selective elevation of serum immunoglobulin (IgG), circulating autoantibodies, a histological picture of periportal necroinflammation (interface hepatitis) with no biliary changes, and a marked response to corticosteroid therapy (1). Diagnosis is based on these clinical and laboratory features and exclusion of other possible causes of liver disease (2, 3). Recent recommendations are that patients should be categorized as having definite or probable AIH according to several clinical, laboratory, and histological criteria (1). Primary sclerosing cholangitis predominantly affects men and is characterized by elevations of the cholestatic liver enzymes, abnormalities in serum and liver indices of copper metabolism, cholangiolitic features on liver biopsy, and characteristic cholangiographic features in the extrahepatic and/or intrahepatic biliary tracts. It may occur alone or in association with ulcerative colitis (4, 5). The spectrum of bile duct damage in PSC ranges from small duct PSC with histological features of fibrous-obliterative cholangitis and negative cholangiogram to intrahepatic and/or extrahepatic duct involvement with characteristic diffuse, multifocal strictures and segmental dilatation or beading. Purely intrahepatic liver disease seems to be rare. Intralobular hepatic inflammation is not usually considered a feature of PSC. Although there are anecdotal reports of some patients responding to corticosteroid therapy, in general the response is poor. However, some of the histological and immunological features of AIH are occasionally seen in PSC and there have been case reports of the simultaneous occurrence of AIH and PSC (6 –12). In an attempt to clarify the diagnostic criteria, it has been recommended that cases of PSC/AIH overlap should not be considered as part of the spectrum of AIH (1). Nonetheless, an overlap syndrome does exist and there is a need to define criteria for this condition. We report here five cases that illustrate the spectrum of features associated with this interesting syndrome, all of whom fulfilled criteria for definite AIH (1), but also had evidence of PSC.
INTRODUCTION Autoimmune hepatitis (AIH) and primary sclerosing cholangitis (PSC) are generally considered to be two distinct idiopathic liver disorders. Autoimmune hepatitis is a disease that predominantly affects women and is characterized by a
MATERIALS AND METHODS The five cases reported here (Table 1) represent all of those with the features described who had regularly attended our clinic during the past 5 yr and for whom sufficient data
Received July 18, 1997; accepted Jan. 19, 1998. 777
778
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998
TABLE 1 Demographic, Biochemical, Immunological, and Histological Data at Presentation in the Five Cases Case Number 1 Age (yr) Gender Albumin (# 35 g/L) Total globulins (# 35 g/L) IgG (# 18.6 g/L) IgM (# 2.0 g/L) AST (# 50 IU/L) Bilirubin (# 20 mmol/L) ALP (# 120 IU/L) ALP:AST ratio Gamma GT (# 55 IU/L) Autoantibody titers ANA SMA Liver histology Grade Stage
38 M 27 51 38.3 1.7 252 168 1245 2.08 638
2
3
20 F 33 52 29.3 1.7 320 46 1216 1.58 313
1:40 1:40
1:2560 1:80
4 3
2 3
26 M 34 47 NA 2.5 1750 72 156 0.04 NA 1:320 Neg 4 3
4 14 M 29 52 45.0 1.0 996 77 505 0.21 196 1:80 1:160 3 1
5 18 M 35 62 34.4 1.3 304 43 1115 1.52 639 Neg 1:320 0 1
Abbreviations: AST 5 aspartate aminotransferase, ALP 5 alkaline phosphatase, Gamma GT 5 g-glutamyl transferase (upper normal limits in parentheses); ANA 5 antinuclear antibodies (homogeneous pattern), SMA 5 smooth muscle antibodies, NA 5 data not available. Histological grading and staging is as described in Materials and Methods.
were available for this analysis. None had a history of excessive alcohol consumption or of exposure to other potentially hepatotoxic agents, and viral screens (HBsAg, antihepatitis C virus [HCV], IgM anti-HAV, CMV, EBV) were negative in all (performed retrospectively for anti-HCV and HCV-RNA on stored sera where necessary). All were also seronegative for antimitochondrial (AMA) and liver-kidney microsomal (LKM-1) autoantibodies on every occasion when they were tested. Metabolic liver diseases were excluded by appropriate investigations, including routine serum biochemical screening for ferritin, copper, ceruloplasmin, and a-1-antitrypsin phenotype. Other than ulcerative colitis (in two cases) none of the patients or their first-degree relatives had any other diseases thought to have an immunological basis. The histological findings described herein were as reported at the time of liver biopsy and were confirmed on review of all biopsies by one of us (B.C.P.) for the present study. Histological activity was graded for severity of necroinflammation and staged for degree of fibrosis according to Batts and Ludwig (13). According to this scheme, for lymphocytic piecemeal necrosis, grade 0 5 none; 1 5 minimal, patchy; 2 5 mild; 3 5 moderate; 4 5 severe; and for fibrosis stage, 0 5 none; 1 5 portal only; 2 5 periportal; 3 5 septal; 4 5 cirrhosis. Antinuclear (ANA), smooth muscle (SMA), AMA, LKM-1, and perinuclear-staining antineutrophil cytoplasmic (pANCA) autoantibodies were sought by standard indirect immunofluorescence by the hospital’s routine clinical immunology service. Sera were screened at 1:10 dilution and positive samples were titrated out to negativity by doubling dilution.
The scoring system proposed by the International Autoimmune Hepatitis Group (IAIHG) was applied as recommended (1), whereby a diagnosis of definite AIH is made when the aggregate score is . 15 before treatment and . 17 when response to treatment is included in the assessment, and scores of 10 –15 before and 12–17 after treatment are interpreted as probable AIH. Histocompatibility antigen (HLA) phenotype data were included in the scoring assessment of the three patients for whom these data were available. Ratios of serum alkaline phosphatase to aspartate aminotransferase (ALP:AST ratios) were calculated from the degree of elevation of these enzyme activities above the upper normal limits. All patients had serum biochemical liver tests performed on more than one occasion before institution of therapy and results for the day on which the ALP was highest were used for this analysis. For comparison a retrospective analysis of ALP and AST values at presentation was similarly undertaken in 100 consecutive patients with definite AIH (with no evidence of biliary complications) attending our clinic during the same period. Case 1 This man was referred in 1986 with a 4-month history of jaundice and weight loss. Laboratory findings at presentation are shown in Table 1. Liver biopsy showed features of severe hepatitis with extensive bridging and multiacinar collapse (grade 4, stage 3), together with an acute cholangiolitic component with marked ductular proliferation. Two months after starting on prednisolone 50 mg/day serum globulins, AST, and bilirubin had returned to normal but ALP and gammaglutamyl transferase (GGT) remained elevated at 277 and 593 IU/L, respectively. Azathioprine (100
AJG – May 1998 mg/day) was introduced and the steroids were gradually reduced to a maintenance dose of 7.5 mg/day. Within the following year all of his liver tests returned to normal and a liver biopsy showed inactive disease with severe bridging fibrosis (grade 0, stage 3). He experienced relapses in 1989 and again in 1990 during attempts to reduce immunosuppressive therapy. On the second occasion he had abdominal pain and the cholestatic liver enzymes remained abnormal (GGT 862–1460 IU/L and ALP 250 –714 IU/L) after AST returned to normal. Endoscopic retrograde cholangiopancreatography (ERCP) revealed shortening and irregularity of the intrahepatic bile ducts compatible with primary sclerosing cholangitis (PSC). The gallbladder was noted to be grossly enlarged and an elective cholecystectomy was performed. Gallbladder histology revealed chronic cholecystitis. In 1992, he developed bloody diarrhea and colonoscopic biopsies showed features of ulcerative colitis. His condition deteriorated over the next few months, with increasing jaundice and encephalopathy. He died before liver transplantation could be performed. Postmortem liver histology showed a marked periportal and lobular hepatitis (grade 4, stage 3) with ductular proliferation and prominent bridging collapse. Case 2 This woman presented in 1994 with a 4-week history of painless jaundice and pruritus, with no other significant findings on examination. Ultrasonography revealed an apparently normal liver with no evidence of bile duct dilatation. Liver biopsy showed a chronic hepatitis (grade 2, stage 3) with biliary epithelial damage and attenuation, periductular fibrosis, severe ductopenia, and copper-associated protein deposits typical of stage III PSC. ERCP revealed abnormal narrowing and stricturing of the extra- and intrahepatic biliary tree (Fig. 1). There was no evidence of ulcerative colitis on rectal biopsy. Serum bilirubin, AST, and IgG normalized within 1 month of starting on prednisolone 20 mg/day, azathioprine 75 mg/day, and ursodeoxycholic acid 300 mg/day, but ALP and GGT remained elevated at 200 –300 IU/L. She had two biochemical relapses 15 and 27 months later, during attempts to reduce the steroids , 10 mg/day, but has remained well over the past 18 months taking this dose (with 75 mg/day azathioprine). Her AST, bilirubin, and IgG have normalized but ALP (196 IU/L) and GGT (270 IU/L) remain elevated and she is still ANA positive (1:640). Case 3 In 1976, this man was referred with a 4-month history of dyspepsia and fluctuating jaundice. Liver biopsy showed severe chronic hepatitis with piecemeal necrosis and bridging fibrosis (grade 4, stage 3) (Fig. 2A). Treatment with prednisolone 15 mg/day and azathioprine 75 mg/day led to normalization of all biochemical parameters and he became ANA negative. Liver biopsy 1 yr later showed only minimal residual activity with some scarring (grade 0, stage 2). He
AIH/PSC OVERLAP
779
FIG. 1. Endoscopic retrograde cholangiopancreatograph in Case 2, showing irregularities of intra- and extrahepatic bile ducts suggestive of primary sclerosing cholangitis, typical of the findings in the other four cases.
relapsed in 1980 during an attempt to withdraw all treatment. Liver biopsy at this time showed severe chronic hepatitis with prominent lobular activity (grade 4, stage 3) and mild cholangitic features. Within 4 months of increasing the steroids to 10 mg/day and reintroducing azathioprine (75 mg/day), all liver tests were again completely normal and follow-up liver biopsies (in 1982 and 1983) revealed only grade 1 chronic hepatitis with stage 3 fibrosis, with an occasional cholangiolitic focus. In 1984, he defaulted therapy and relapsed again. Remission was regained on reintroduction of therapy, but ALP and GGT remained slightly elevated (, twice upper normal limits). A liver biopsy in 1988 showed mild patchy piecemeal necrosis (grade 2, stage 3) and, for the first time, features of severe (but focal) destructive cholangitis and pericholangitis (Fig. 2B). A further biopsy in 1991 showed significant deterioration, with extensive postnecrotic bridging fibrosis and predominantly cholangitic features, including marked inflammation around septal-sized bile ducts (Fig. 2C). ERCP revealed a grossly abnormal intrahepatic biliary tree with reduced-caliber bile ducts and multiple areas of narrowing and irregularity consistent with PSC. He currently remains asymptomatic on 75 mg/day azathioprine alone. Case 4 This patient presented in 1981 (aged 14 yr) with a 6-wk history of right upper quadrant abdominal discomfort and fluctuating jaundice. Liver biopsy showed moderate to severe periportal necroinflammation with prominent lobular
780
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998
FIG. 2. Liver histology in Case 3. (A) Initial biopsy at presentation, showing extensive bridging parenchymal collapse and fibrosis with prominent interface hepatitis. (B) Twelve yr later. Note expanded and bridged portal areas showing severe cholangitic inflammation with bile duct damage and destruction (arrows). (C) Three yr after (B), showing marked chronic inflammation around larger, septal-sized bile duct branches.
AJG – May 1998
AIH/PSC OVERLAP
781
FIG. 2. Continued.
inflammation and ductular proliferation (grade 3, stage 1). Prednisolone therapy (80 mg/day, reducing to 10 mg/day over a 6-month period) led to good clinical and biochemical response. Liver biopsy in 1982 showed mild portal and lobular inflammation with minimal piecemeal necrosis (grade 2), focal cholangiolitis, mild ductular proliferation, and stage 2 fibrosis. His AST had risen to 150 IU/L, so azathioprine (50 mg/day) was added, leading to a reduction in the AST (, twice upper normal limit) but with continuing marked elevations of the cholestatic liver enzymes (ALP 210 – 665 IU/L, GGT 163–399 IU/L, and IgG 17.8 –35.3 g/L), and with ANA rising to 1:640. In 1988, he developed diarrhea. Colonoscopic biopsies were consistent with ulcerative colitis, and liver biopsy showed portoseptal fibrosis with pericholangitis, ductular proliferation, and copper-associated protein deposition indicative of a chronic pathological process involving the bile ducts. ERCP revealed changes compatible with intrahepatic PSC. Immunosuppressive treatment was gradually withdrawn. In 1992, he relapsed with elevated AST and biliary enzymes and high titers of ANA. Prednisolone and azathioprine were restarted with improvement in liver function and he is currently asymptomatic (for both his liver and bowel diseases), but ALP and GGT have remained elevated (at 227 and 600 IU/L, respectively). Case 5 This male patient presented in 1995 with a 3-month history of upper abdominal discomfort (associated with a
short-lived episode of diarrhea), weight loss and abnormal liver tests (Table 1). Liver biopsy showed a mixed portal inflammatory infiltrate with florid cholangitis and pericholangitis (grade 0, stage 1), and ERCP demonstrated changes compatible with PSC affecting predominantly the intrahepatic biliary tree but with some extrahepatic changes. Colonoscopy and colonic biopsies were normal. A trial of ursodeoxycholic acid (300 mg b.i.d.) led to no significant improvement in his liver biochemistry after 1 month and he was started on prednisolone at 30 mg/day. Within 1 month the ALP fell to 469 IU/L and AST to 55 IU/L. An attempt to reduce his steroids led to a relapse, so azathioprine was added (75 mg/day, increasing later to 150 mg/day) and prednisolone gradually reduced to 10 mg/day. After 18 months on this regimen, he is asymptomatic with normal IgG, ALP, bilirubin, and AST but his GGT is still elevated at 136 IU/L. A recent liver biopsy showed portal fibrosis with mild residual inflammatory infiltrate in the portal tracts (grade 0, stage 1) but no evidence of the previously noted cholangitic changes. RESULTS These five patients were mainly young (median age 20 yr, range 14 –38 yr), predominantly male (four cases), and presented with constitutional symptoms and jaundice. Serum AST activities were elevated . 5 times the upper normal limit in all and ALP . fourfold in four cases (Table 1). All had hyperglobulinemia with . 1.5-fold selective
782
MCNAIR et al.
AJG – Vol. 93, No. 5, 1998 TABLE 2 Additional Data on the Five Cases at Presentation and During Follow-up Case Number
HLA phenotype Ulcerative colitis pANCA AIH score At presentation After treatment Follow-up (yr)
1
2
3
4
5
B8-DR3 Yes Positive
ND No Positive
A1-B8-DR3 No Positive
A1-B8-DR3 Yes Positive
ND No Positive
17 22 7
16 18 3
18 23 21
20 25 16
16 21 2
pANCA 5 perinuclear staining antineutrophil cytoplasmic antibodies, ND 5 not determined.
elevations in serum IgG concentrations and ANA (homogeneous pattern) and/or SMA titers $ 1:40. All were pANCA positive and were seronegative for AMA. The three who had been HLA phenotyped had the B8-DR3 haplotype, with two having the full extended haplotype A1-B8-DR3 (Table 2). Liver histology at presentation showed periportal lymphocytic piecemeal necrosis of varying severity in all but one, with marked cholangiolitic changes in three cases. Of the two patients without pronounced biliary features at presentation, one (Case 4) had two biopsies during the first 2 yr that showed mild biliary changes, but it was only when he developed ulcerative colitis 7 yr later that significant histological and cholangiographic features compatible with PSC were seen. The other (Case 3) was diagnosed as having definite AIH and has been managed as such throughout the 21 yr of his disease. During that time he has had seven liver biopsies. The first indication of biliary complications came with the finding of occasional focal cholangitic changes in his third biopsy (4 yr after the first). Three further biopsies (at 6, 7, and 12 yr after presentation) showed similar focal biliary changes and it was not until 15 yr after the initial diagnosis that histological and cholangiographic features suggestive of PSC were observed. All patients showed marked clinical and biochemical responses to treatment with prednisolone and azathioprine, and four had relapses during subsequent reductions or withdrawal of therapy. Histological improvement during treatment was also demonstrated in these four cases, although progression of the biliary lesions was evident. In the fifth case, follow-up biopsy after 1 yr on immunosuppressive therapy revealed no evidence of intrahepatic biliary disease. Application of the IAIHG scoring system for features at presentation revealed that all five cases achieved scores above the recommended minimum for a diagnosis of definite AIH, and this was reinforced when subsequent response to treatment was taken into account (Table 2). One feature of this system is the assignment of a negative score of 3 points for ALP:AST ratios . 3.0. All but one had serum ALP activities . 500 IU/L, but AST activities were also significantly elevated and none had a ratio . 3.0 (Table 1). Comparative analysis of the 100 consecutive cases of def-
inite AIH revealed that nine had ALP activities . 500 IU/L (median 625, range 500 –1280) but only one of these had a ratio . 1.5 (at 1.62). Of the remaining 91 (with ALP activities , 500 IU/L), all had ALP:AST ratios , 3.0 and only four had ratios . 1.5 (1.53, 1.60, 2.05, and 2.20, respectively). DISCUSSION The five cases described here do not conform to a clearly defined category of liver disease. All met criteria for definite autoimmune hepatitis (AIH) but had histological and radiological features of primary sclerosing cholangitis (PSC). In addition, they all showed a marked response to immunosuppressive therapy, which is usually of little benefit in PSC (5, 14). In these and other respects they appear to be distinct from the well-recognized overlap between PBC and AIH, of which there have been numerous reports (15, 16 –22). The overlap between PSC and AIH has been reported much less frequently (6 –12), although distinguishing between the two conditions in children can be particularly difficult (23–25), and there is no consensus on presenting features or whether corticosteroids are of benefit. Preliminary data on eight cases from the Mayo Clinic (10) indicated a significantly poorer response to corticosteroids than in patients with definite AIH, and an almost invariable association with ulcerative colitis. Other reports (6 –9, 11, 12) have documented a good response to immunosuppressive therapy and a less marked association with colitis, as in the present study. Because it is clear, however, that some patients with PSC/AIH do respond very well to immunosuppressive therapy, a method for identifying them is required. Titers of ANA or SMA, serum concentrations of IgG, or other immunological parameters do not seem to be particularly useful discriminators in this situation. In a recent analysis of 114 adults with PSC, Boberg et al. (26) reported that serum IgG was elevated in . 60% (although only 15% had levels . 1.5 times the upper normal limit) and approximately 20% had ANA or SMA at titers $ 1:40, but only two patients fulfilled criteria for definite and two for probable AIH (both of whom responded to immunosuppressive ther-
AJG – May 1998
AIH/PSC OVERLAP
apy). pANCA is associated with PSC (27) but these antibodies reportedly also occur in up to 90% of patients with AIH (28 –31). Similarly, the HLA B8-DR3 haplotype, found in the three present cases who were HLA-typed and in the three cases reported by Gohlke et al. (12), is associated with both AIH and PSC. The only presenting feature that distinguished the present cases from classical AIH was the finding of cholangitic changes on liver biopsy, and this was only in three cases. Of the remaining two patients, liver biopsy in one (Case 4) revealed ductular proliferation with no obvious bile duct damage and the other (Case 3) had no histological evidence of biliary disease. It was only some time after the initial diagnosis of AIH that liver biopsies in these two patients showed evidence of definite biliary involvement and that features of PSC were revealed on ERCP. It may be that their disease progressed with time from a pure AIH towards a mixed picture of AIH/PSC. Indeed, of the 11 cases of AIH/PSC overlap reported in the literature who were sufficiently well documented for comparison (6 –9, 11, 12), seven were initially diagnosed as AIH and subsequently found to have features of PSC after intervals of 2–9 yr, three cases presented with features of both conditions, and only one was initially considered to have PSC. A high serum ALP activity might be a useful parameter for raising suspicions of an overlap syndrome because four of the present patients had ALP activities . 500 IU/L at presentation, but also so did 9% of our AIH patients. However, none of the AIH patients with ALP activities . 500 IU/L or ALP:AST ratios . 1.5 subsequently developed any evidence of biliary disease over a median follow-up of 14 yr (range 3–20 yr) and, in contrast to the present AIH/PSC overlap cases, ALP activities rapidly normalized in all during the initial response to immunosuppressive therapy. Nonetheless, we would suggest that the likelihood of an overlap syndrome should be considered in any patient with features of definite or probable AIH who has a high serum ALP activity and/or an ALP:AST ratio . 1.5 and that ERCP should be performed in such patients, particularly if there are any biliary changes evident on liver biopsy. A trial of immunosuppressive therapy, with or without ursodeoxycholic acid, would seem to be justified in such cases. ACKNOWLEDGMENT We thank Dr. Peter Donaldson of our Institute for kindly providing the HLA phenotyping data on these patients. Reprint requests and correspondence: Dr. I. G. McFarlane, Institute of Liver Studies, King’s College Hospital, Denmark Hill, London SE5 9RS, UK.
REFERENCES 1. Johnson PJ, McFarlane IG. Meeting Report: International Autoimmune Hepatitis Group. Hepatology 1993;18:998 –1005. 2. Czaja AJ. Clinical aspects of autoimmune hepatitis in North America. In: Nishioka M, Zeniya M, eds. Autoimmune hepatitis. Amsterdam: Elsevier, 1994:27– 43.
783
3. Ohta Y, Onji M. Clinical aspects of autoimmune hepatitis in Japan. In: Nishioka M, Toda G, Zeniya M, eds. Autoimmune hepatitis. Amsterdam: Elsevier, 1994:45–53. 4. Ludwig J, La Russo NF, Wiesner RH. The syndrome of primary sclerosing cholangitis. In: Popper H, Schaffner F, eds. Progress in liver diseases. Philadelphia: Saunders, 1990:555– 66. 5. Ueno Y, La Russo NF. Primary sclerosing cholangitis. J Gastroenterol 1994;29:531– 43. 6. Minak G, Sutherland LR, Pappas C, et al. Autoimmune chronic active hepatitis (lupoid hepatitis) and primary sclerosing cholangitis in two young adult females. Can J Gastroenterol 1988;2:22–7. 7. Rabinowitz M, Demetris AJ, Bou-Abboud CF, et al. Simultaneous occurrence of primary sclerosing cholangitis and autoimmune chronic active hepatitis in a patient with ulcerative colitis. Dig Dis Sci 1992; 37:1606 –11. 8. Leggett BA, Hallam A, Powell EE, et al. Autoimmune chronic active hepatitis evolving to primary sclerosing cholangitis: evidence for a common pathogenesis. Hepatology 1992;16:191A (abstract). 9. Lawrence SP, Sherman KE, Lawson JM, et al. A 38 year old man with chronic hepatitis. Semin Liver Dis 1994;14:97–105. 10. Roberts S, Poterucha JJ, Gores GJ, et al. Overlap between type 1 autoimmune hepatitis and primary sclerosing cholangitis. Hepatology 1995;22:129A. 11. Wurbs D, Klein R, Terracciano LM, et al. A 28-year-old woman with a combined hepatitic cholestatic syndrome. Hepatology 1995;22: 1598 –1605. 12. Gohlke F, Lohse AW, Dienes HP, et al. Evidence for an overlap syndrome of autoimmune hepatitis and primary sclerosing cholangitis. J Hepatol 1996;24:699 –705. 13. Batts KP, Ludwig J. Chronic hepatitis. An update on terminology and reporting. Am J Surg Pathol 1995;19:1409 –17. 14. Lee YM, Kaplan MM. Primary sclerosing cholangitis. N Engl J Med 1995;332:924 –33. 15. Kloppel G, Seifert G, Lindner H, et al. Histopathological features in mixed types of chronic aggressive hepatitis and primary biliary cirrhosis. Virchows Arch Pathol Anat 1977;373:143– 60. 16. Taylor SL, Dean P, Riely C. Primary autoimmune cholangitis. An alternative to antimitochondrial antibody-negative primary biliary cirrhosis. Am J Surg Pathol 1994;18:91–9. 17. Ben-Ari Z, Dhillon AP, Sherlock S. Autoimmune cholangiopathy: Part of the spectrum of autoimmune chronic active hepatitis. Hepatology 1993;18:10 –5. 18. Michieletti P, Wanless IR, Katz A, et al. Antimitochondrial antibody negative primary biliary cirrhosis: A distinct syndrome of autoimmune cholangitis. Gut 1994;35:260 –5. 19. Colombato LA, Alvarez F, Cote J, et al. Autoimmune cholangiopathy: The result of consecutive primary biliary cirrhosis and autoimmune hepatitis? Gastroenterology 1994;107:1839 – 43. 20. Okuno T, Seto Y, Okanque T, et al. Chronic active hepatitis with histological features of primary biliary cirrhosis. Dig Dis Sci 1987;32: 775–9. 21. Goodman ZD, McNally PR, Davis DR, et al. Autoimmune cholangitis: A variant of primary biliary cirrhosis. Clinicopathologic and serologic correlations in 200 cases. Dig Dis Sci 1995;40:1232– 42. 22. Lacerda MA, Ludwig J, Dickson ER, et al. Antimitochondrial antibody-negative primary biliary cirrhosis. Am J Gastroenterol 1995;90: 247–9. 23. El-Shabrawi M, Wilkinson ML, Portmann B, et al. Primary sclerosing cholangitis in childhood. Gastroenterology 1987;92:1226 –35. 24. Mieli-Vergani G, Lobo-Yeo A, McFarlane BM, et al. Different immune mechanisms leading to autoimmunity in primary sclerosing cholangitis and autoimmune chronic active hepatitis of childhood. Hepatology 1989;9:198 –203. 25. Maggiore G, Veber F, Bernard O, et al. Autoimmune hepatitis associated with antiactin antibodies in children and adolescents. J Pediatr Gastroenterol Nutr 1993;17:376 – 81. 26. Boberg KM, Fausa O, Haaland T, et al. Features of autoimmune hepatitis in primary sclerosing cholangitis: An evaluation of 114 primary sclerosing cholangitis patients according to a scoring system for the diagnosis of autoimmune hepatitis. Hepatology 1996;23:1369 –76. 27. Snook JA, Chapman RW, Fleming K, et al. Anti-neutrophil nuclear antibody in ulcerative colitis, Crohn’s disease and primary sclerosing cholangitis. Clin Exp Immunol 1989;76:30 –3.
784
MCNAIR et al.
28. Hardarson S, LaBrecque DR, Mitros FA, et al. Antineutrophil cytoplasmic antibody in inflammatory bowel and hepatobiliary diseases: High prevalence in ulcerative colitis, primary sclerosing cholangitis, and autoimmune hepatitis. Clin Microbiol Immunol 1993;99:277– 81. 29. Mulder AHL, Horst G, Haagsma EB, et al. Prevalence and characterization of neutrophil cytoplasmic antibodies in autoimmune liver diseases. Hepatology 1993;17:411–7.
AJG – Vol. 93, No. 5, 1998 30. Vidrich A, Lee J, James E, et al. Segregation of pANCA antigenic recognition by DNase treatment of neutrophils: Ulcerative colitis, type 1 autoimmune hepatitis and primary sclerosing cholangitis. J Clin Immunol 1995;15:293–9. 31. Targan SR, Landers C, Vidrich A, et al. High-titer antineutrophil cytoplasmic antibodies in type-1 autoimmune hepatitis. Gastroenterology 1995;108:1159 – 66.