Jgh2615

Journal of Gastroenterology and Hepatology (2001) 16, 1297–1302
OUTCOME OF AUTOIMMUNE HEPATITIS IN CHILDREN Long-term outcome of autoimmune hepatitis in children
OMAR I SAADAH, ARNOLD L SMITH AND WINITA HARDIKAR Department of Gastroenterology and Clinical Nutrition, Royal Children’s Hospital, Melbourne, Abstract
Background and Aim: Autoimmune hepatitis (AIH) is a chronic disease of unknown etiology, which
usually progresses to cirrhosis if not diagnosed and treated promptly. Data on long-term follow up in
children with AIH are scant. The aim of this study is to assess the long-term outcome of autoimmune
hepatitis in children with respect to clinical and laboratory features at presentation.
Methods: Data were extracted from the medical records of patients presenting over a 28-year period
(1972–2000) to the Royal Children’s Hospital, Melbourne, Australia. Additional information was
obtained by interviewing patients, and their current physicians. Of the 30 patients (22 females, mean
age 9 years) identified, 18 had type I, three had type II, four had autoimmune–polyendocrinopathy
syndrome type 1, one had infantile giant-cell hepatitis associated with Coomb’s-positive hemolytic
anemia, and four were seronegative (antinuclear antibody (ANA), smooth muscle antibody (SMA) and
liver–kidney microsomal antibody (LKM)).
Results: Clinical features at presentation included hepatomegaly (86%), jaundice (66%) and
splenomegaly (50%). Initial investigations revealed a median serum bilirubin level of 55 mmol/L (range
6–425), median aspartate aminotransferase level of 678 IU (range 70–2548), and abnormal clotting in
33% of patients. Liver biopsies were performed on all patients at presentation and 11 showed cirrhosis
(36%). The mean follow-up period was 10.0 ± 7.8 years with 43% being followed for > 10 years. Only
two patients died and one required transplantation. Fourteen (50%) patients continue to be on low
dose prednisolone with azathioprine, two (7%) are on prednisolone alone, and six (21%) are on no
therapy. When the cirrhotic and non-cirrhotic patients were compared, the albumin level at presenta-
tion was significantly lower in the cirrhotic group (P = 0.01). Of the patients who were cirrhotic at pres-
entation, six (54%) remain compensated with a mean follow-up period of 8 years. All 24 patients
currently under follow up are engaged in age-appropriate activities including school, part- or full-time
work.
Conclusion: Autoimmune hepatitis has a favorable long-term outcome with a transplant-free survival
rate of 90% over a mean period of 10.0 ± 7.8 years (range: 0.5–23), and a normal or near-normal lifestyle
irrespective of presenting clinical, laboratory or histological features.
2001 Blackwell Science Asia Pty Ltd
Key words: autoimmune hepatitis, children, liver–kidney microsomal antibodies, polyendocrinopathy
syndrome, smooth muscle antibodies.
INTRODUCTION
presence of circulating non-organ and liver-specificautoantibodies.1 Autoimmune hepatitis is a disorder of unknown etiol- Steroids, with or without azathioprine, have convinc- ogy characterized histologically by dense mononuclear ingly altered the outcome in most patients.2–6 Without cell infiltrates in the portal tracts, progressive destruc- treatment, autoimmune hepatitis (AIH) often pro- tion of the hepatic parenchyma, and serologically by the gresses to cirrhosis, and in the more severe cases, carries Correspondence: Dr W Hardikar, Department of Gastroenterology and Clinical Nutrition, Royal Children’s Hospital, Flem- ington Road, Parkville, Victoria 3052, Australia. Email: [email protected] Accepted for publication 1 August 2001.
a high incidence of mortality and a low rate of sponta- The follow-up data collected included: presence of jaundice, splenomegaly, current immunosuppression, Reported series of autoimmune hepatitis in children complications of treatment, survival, liver transplanta- in the literature are scant.8–12 Two main forms have been tion, and age-appropriate activity, including work and described in children: type I associated with antinuclear and/or antismooth muscle (ANA/SMA) antibodies, and Statistical analysis was performed by using Stata type II is associated with antiliver-kidney microsomal Statistical software: Release 6.0. (Stata Corporation, antibodies type 1 (LKM-1). The proposed target for College Station, TX, USA). Fisher’s exact test was used antismooth muscle antibodies in AIH is the actin micro- for categorical variables, and Wilcoxon’s rank-sum filament, while antismooth muscle antibodies directed (Mann–Whitney) test was used for continuous vari- to non-actin microfilaments can be found in other ables. Results are expressed as either median or mean conditions, like viral hepatitis. The anti-liver–kidney value ± SD. A P value < 0.05 was considered to be microsomal antibody type 1 is usually directed against Although these two types of AIH have different clin- ical, biochemical and histological features, they have been reported to have similar severity and outcome.11 In the two largest series of children with AIH pub- Thirty patients with AIH were identified during the lished,11,12 one reported a median follow up of 5 years study period. Eighteen patients had type I with ANA while the other reported a mean follow up of 4 years and/or SMA antibodies, three patients had type II with and 10 months. Only one of the studies11 attempted to LKM-1 antibodies, and four patients had AIH, which identify factors predictive of long-term outcome.
was negative for ANA/SMA and LKM-1 antibodies.
In this study, we describe 28 years of experience with When classified according to the International Auto- AIH, with special emphasis on long-term outcome immune Hepatitis Group scoring system,1 15 patients had definite and the other 10 had probable AIH.
In addition, four patients had autoimmune polyen-docrinopathy syndrome type 1 (APS-1), and onepatient had infantile giant cell hepatitis with associated autoimmune Coomb’s-positive hemolytic anemia.
Patients with AIH presenting to the Gastroenterologyand Hepatology service at the Royal Children’s Hospi- Clinical data
tal, Melbourne between 1972 and 2000 were includedin the present study. Data were collected by reviewing The clinical features at presentation are summarized in the medical records, phone follow up with the patients Table 1. Twenty-two patients were female (73%), with themselves and their families, as well as the medical care a mean age at presentation of 9.4 ± 4.2 years (range provider for those who graduated from pediatric care.
All patients were seronegative for hepatitis B surface A family history of extrahepatic autoimmune disease antigen. Hepatitis C infection was excluded in all in the first-degree relatives was positive in seven patients who presented after the availability of the patients; systemic lupus (two), rheumatoid arthritis hepatitis C virus ELISA assay (15 patients), and retro- (two), autoimmune thyroiditis (one) and in siblings spectively during follow up for the graduates, with the exception of three patients who were lost to follow up The onset of the disease followed three patterns: (i) and two patients who died before being tested. Other acute onset of hepatitis was observed in nine patients, possible causes of hepatitis were excluded by using progressing to fulminant hepatic failure in one. There was a history of previous episodes of hepatitis-like At diagnosis, patients were treated with prednisolone illness in three of the patients (2, 3 and 6 years prior to starting at 2 mg/kg per day (maximum of 60 mg/day), diagnosis); (ii) insidious onset of anorexia, fatigue and followed by a gradual tapering according to the clin- mild intermittent jaundice was observed in 13 patients, ical response and transaminase activity. Azathioprine with two of these found to have evidence of disturbed (1–2 mg/kg per day) was added to prednisolone, if an synthetic function of the liver and ascites at the time increase in the aspartate aminotransferase (AST) level of presentation; and (iii) an incidental finding of was noticed during the tapering of the prednisolone abnormal liver function tests or hepatosplenomegaly dose. Patients were followed up regularly with close during follow up for another medical problem was seen monitoring of the disease activity both clinically and by liver function tests, aiming to achieve control with Included in this series are four patients with APS-1, minimal immunosuppressive therapy. Treatment was with clinical characteristics being shown in Table 2. Also discontinued if the patient maintained remission for a included is one patient with autoimmune Coomb’s- minimum of 2 years, as defined by the absence of clini- positive hemolytic anemia, who presented at 9 months cal symptoms, normal transaminases, and absence of of age with anemia, and was treated with prednisolone.
necroinflammatory activity on a liver biopsy. Patients At 11 months of age, he developed jaundice, had ele- unresponsive to the immunosuppressive therapy, with vated liver enzymes and deteriorating liver function. A deterioration in liver function, were listed for liver trans- liver biopsy showed severe hepatitis with giant cell plantation when this became available in 1988.
transformation. He was treated unsuccessfully with Long-term outcome in autoimmune hepatitis Clinical features at presentation in 30 patients with prednisolone, azathioprine, cyclosporine and intra- venous immunoglobulin. Repeated liver biopsy prior toimmunoglobulin treatment showed progression to cir- rhosis. The patient started to improve very slowly over the subsequent 5 years. He is currently 7-years-old, on2 mg prednisolone and 25 mg azathioprine with normal growth, normal liver function tests and hemoglobin, but All three patients with insulin-dependent diabetes mellitus had this diagnosis made between 2 and 4 years prior to the diagnosis of autoimmune hepatitis. One of these was not treated with steroids as she had verypoorly controlled diabetes. The other two patients received full dose prednisolone without any significant Laboratory data
The median level for serum bilirubin at presentation mmol/L (range 6–425, normal value (nv): < 20).
Eleven patients had a raised globulin level of more than 1.5-fold the upper limit of normal (ULN), mean 46.0 ± 17.6, range 24–84, nv: 24–33 g/L. Immunoglob- ulin levels were performed in 16 patients. Immunoglob- ulin (Ig)G and IgM isotypes were high in nine and eleven patients respectively, and two had low IgA with some patients having more than one abnormality. The mean value for serum albumin was 37.0 ± 11.6 g/L (range 17–74, nv: 35–50). In 23 patients (76.6%), AST exceeded fivefold the ULN, with a median of 678 IU/L (range 70–2548, nv: < 50). Abnormal clotting at pre- sentation was observed in 10 patients (33%). Antinu-clear antibodies (ANA) were found to be positive in APS-1, Autoimmune polyendocrinopathy syndrome type 1.
Clinical, biochemical, and histological characteristics of patients with autoimmune polyendocrinopathy syndrome type-1 MucocutaneouscandidiasisAddisonAlopeciaSteatorrheaGrowth failure Pernicious anemiaGrowth failurePrimary ovarian failure AST, aspartate aminotransferase; Anti-LKM, anti-liver–kidney microsomal antibodies type 1; ANA, antinuclear antibodies; 16 patients. The pattern of immunofluorescence was AST to less than twice the ULN within the first reported in 12 of these patients; homogeneous in four, 3 months was observed in 25 patients (83%). Nineteen speckled in four, and mixed in four.
patients were Cushingoid in the early stages of pred-nisolone treatment, but improved with reduction of thedose. Other side-effects were: osteoporosis (six), serious Histological data
infection (four), hypertension (four), acne (three), epi-gastric pain (three), behavioral disturbance (three), All patients underwent a liver biopsy, obtained by the growth disturbance (two), proximal muscle weakness percutaneous route prior to starting treatment, except (one), and cataract (one). Azathioprine was required in for one patient in whom the procedure was delayed 14 patients after a median time of 3 months (range until after the commencement of steroid therapy 0.23–24) following steroid treatment. None of the because of severe blood clotting disturbance. The main patients had developed signs of significant bone marrow histological features at presentation were: portal tract suppression because of treatment with azathioprine.
inflammation (29), portal tract fibrosis (20), piecemealnecrosis (19), bridging necrosis (11), lobular inflam-mation (eight), and cirrhosis (11). Twenty follow-up liver biopsies were done in 15 patients. The first follow-up biopsy was done at a mean of 2.0 ± 1.5 years from Thirty patients with AIH were followed for a mean of the initial biopsy. Twelve patients demonstrated an 10.0 ± 7.8 years (range, 0.5–23). Twenty patients (66%) improvement of necroinflammatory changes, one had were followed up for ≥ 5 years, 13 (43%) for ≥ 10 years, no improvement, and two progressed to cirrhosis in Two patients died because of liver failure before liver Characteristics of patients presenting with cirrhosis transplantation became available. One of these patients are shown in Table 3. None of the four patients with had type I AIH while the other was seronegative. One APS-1 or the patient with autoimmune (AI) hemolytic patient (type I AIH) was transplanted 8 years after diag- anemia had cirrhosis at the time of presentation.
nosis because of portal hypertension, recurrent peri-tonitis and hepatorenal syndrome. He is alive and well5.5 years after the transplant without evidence of recur- Treatment
Fourteen relapses (all while weaning prednisolone) Twenty-nine patients were initially treated with pred- were recorded in 10 patients under our care, with nisolone with a starting dose of 2 mg/kg followed by median follow up of 8 years. The mean time to the first gradual weaning once liver function tests showed relapse was 3.9 ± 2.8 years after the initiation of steroid sustained improvement. The lowest dose of steroid treatment. Three patients had a subsequent relapse at a found to maintain remission was 5.2 ± 1.6 mg, which mean of 2.70 ± 0.23 years from the first relapse. All was achieved in a median time of 10 months (range cases responded to bolus treatment with prednisolone.
1–153 months) from the start of steroid treatment. The Among the survivors (Fig. 1), six patients ceased treat- initial response to prednisolone as defined by a drop of ment without relapse, and 18 are still on treatment Characteristics of patients presenting with and without cirrhosis AST, aspartate aminotransferase. *Values are expressed as mean ± SD. †Significant.
Long-term outcome in autoimmune hepatitis families.16 Cytochrome P450 1A2 has been regarded asthe hepatic autoantigen in APS-1,17 and antibodiesagainst this antigen have important diagnostic implica-tions. Of particular interest in our series is that all fourpatients were easily controlled with prednisolone alone,and indeed this was ceased in one patient. Autoimmunehepatitis appeared to play a minimal role in the mor-bidity associated with this condition.
One patient had severe giant cell hepatitis with Coomb’s-positive autoimmune hemolytic anemia, andthis entity has been regarded as a form of AIH thatcarries high fatality if not recognized and treatedpromptly.18–20 This patient, currently 7 years old, hassurvived the initial fulminant course of his illness,adding to the three published cases who have survivedto date.
The optimum duration of AIH treatment with Follow-up (FU) data on 30 patients with autoim- immunosuppression is not very clear, and the recur- rence rate after discontinuation of immunosuppressionis high in both adults21,22 and children.9,10 We have beenunable to wean the majority of our patients off steroids;however, we are able to maintain them on low-dose (prednisolone alone in four and a combination of low- steroids in addition to azathioprine, which is compati- dose prednisolone and azathioprine in 14). Three were ble with a fairly normal life. It is important to note that lost to follow up after a period of 1.3, 4 and 15 years attempts to wean resulted in relapses (14 in 10 care, respectively. Among the 24 patients followed up, patients), suggesting that long-term steroid usage is five patients had splenomegaly within a mean follow up of 19.0 ± 4.5 years. One of these patients is jaundiced All patients in this series had a liver biopsy at pre- because of a recent flare up of his disease and is on sentation, and hence the presence of cirrhosis could be both prednisolone and azathioprine. Only one of these accurately determined. Patients presenting with cirrho- patients had GI bleeding and required variceal banding.
sis had a significantly lower albumin at presentation Three female patients have had children. All surviv- (P = 0.01) and tended to have a poorer outcome in ing patients are engaged in age-appropriate activity terms of mortality and the need for transplantation including school, part- or full-time work. Thirteen patients are still attending our clinic regularly.
The overall outcome of patients in this series, however, was very favorable with only two deaths andone patient requiring a transplantation; however, none DISCUSSION
of these events occurred in the group with APS-1, or inthe child with hemolytic anemia.When considering only This is one of only three large series of autoimmune the 25 patients with definite or probable AIH (types I, hepatitis in children, and has the longest follow-up II and seronegative), the mortality in this series of two period with a mean of 10 years.Type II AIH constituted (8%) is similar to the six (11%) reported by Gregorio only 13% of the study population in contrast to the et al.,11 and two (6%) reported by Maggiore et al.12 series by Gregorio et al.11 in which 38% were type II This occurred despite the relatively high percentage AIH. This may be because of the different genetic back- of patients with cirrhosis at presentation (44% in this grounds in Europe from where the other series origi- series, 59% of the cases by Gregorio et al. and 80% of nate. Although immunofluorescence for LKM is a the cases by Maggiore et al.), and suggests that these difficult test to perform, we do not feel that failure to patients may remain stable for many years. The major- detect it is the cause for our low number of type II. High ity of our patients are engaged in age-appropriate titers of SMA identified in most type I patients suggest activities including school, part- or full-time work, that these are likely to reflect true type I disease. In suggesting that morbidity from this disease does not addition, there were only four patients in whom typical autoantibodies could not be identified. These so-called In conclusion, our experience with childhood AIH seronegative patients have been described,14 and may would suggest that the majority of patients require long- be positive for one of the other markers including term therapy, however, survival with a reasonable hepatocyte-specific asialoglycoprotein receptor (ASGP- R), soluble liver antigen (SLA) and liver-cytosolicantigen, which were not tested for.
We also included in this series four patients with APS- REFERENCES
1, which is known to be associated with autoimmunehepatitis, and considered as an important determinant 1 Johnson PJ, McFarlane IG. Meeting report: International of fatality.15 The autosomal locus has been assigned to Autoimmune Hepatitis Group. Hepatology 1993; 18:
chromosome 21q22.3 by linkage analysis in 14 Finnish 2 Murray-Lyon IM, Stern RB, Williams R. Controlled trial 13 Manns MP. Recent developments in autoimmune liver of prednisolone and azathioprine in active chronic hepati- diseases. J. Gastroenterol. Hepatol. 1997; 12 (Suppl.):
tis. Lancet 1973; 1: 735–7.
3 Stellon AJ, Hegarty JE, Portmann B, Williams R. Ran- 14 Johanson PJ, McFarlane IG, McFarlane BM, Williams R.
domised controlled trial of azathioprine withdrawal in Autoimmune features in patients with idiopathic chronic autoimmune chronic active hepatitis. Lancet 1985; 23:
active hepatitis who are seronegative for conventional auto-antibodies. J. Gastroenterol. Hepatol. 1990; 5: 244–51.
4 Cook GC, Mulligan R, Sherlock S. Controlled prospec- 15 Ahonen P, Mellärniemi S, Sipilä I, Perheentupa J. Clini- tive trial of corticosteroid therapy in active chronic hepati- cal variation of autoimmune polyendocrinopathy-candidi- tis. Q JM 1971; 40: 159–85.
asis-ectodermal dystrophy (APECED) in a series of 68 5 Johnson PJ, McFarlane IG, Williams R. Azathioprine for patients. N. Engl. J. Med. 1990; 322: 1829–36.
long term remission in autoimmune hepatitis. N. Engl. J. 16 Aaltonen J, Bjorses P, Sandkuijl L, Perheentupa J, Med. 1995; 333: 958–63.
Peltonen L. An autosomal locus causing autoimmune 6 Maggiore G, Bernard O, Hadchouel M, Alagille D. Life disease: autoimmune polyglandular disease type I assigned saving immunosuppressive treatment in severe autoim- to chromosome 21. Nature Genet. 1994; 8: 83–7.
mune chronic active hepatitis. J. Pediatr. Gastroenterol. 17 Clemente MG, Obermayer-Straub P, Meloni A et al. Nutr. 1985; 4: 655–8.
Cytochrome P450 1A2 is a hepatic autoantigen in autoim- 7 Soloway RD, Summerskill WHJ, Baggenstoss AH et al. mune polyglandular syndrome type 1. J. Clin. Endocrinol. Clinical biochemical and histologic remission of severe Metab. 1997; 82: 1353–61.
chronic liver disease: a controlled study of treatment and 18 Bernard O, Hadchouel M, Scotto J, Odièvre M, Alagille early prognosis. Gastroenterology 1972; 63: 820–33.
D. Severe giant cell hepatitis with autoimmune hemolytic 8 Lidman K, Biberfeld G, Sterner G, Norberg R. Chronic anaemia in early childhood. J. Pediatr. 1981; 99: 704–11.
active hepatitis in children clinical and immunological 19 Brichard B, Sokal E, Gosseye S, Buts JP, Gadisseux JF, long-term study. Acta Paediatr. Scand. 1977; 66: 73–
Cornu G. Coombs-positive giant cell hepatitis of infancy: effect of steroids and azathioprine therapy. Eur J. Pediatr. 9 Maggiore G, Bernard O, Hadchouel M, Hadchouel P, 1991; 150: 314–17.
Odievre M, Alagille D. Treatment of autoimmune chronic 20 Weinstein T, Valderrama E, Pettei M, Levine J. Early active hepatitis in childhood. J. Pediatr. 1984; 104: 839–44.
steroid therapy for the treatment of giant cell hepatitis 10 Maggiore G, Bernard O, Homberg JC et al. Liver disease with autoimmune hemolytic anemia. J. Pediatr. Gastroen- associated with anti-liver-kidney microsome antibody in terol. Nutr. 1993; 17: 313–16.
children. J. Pediatr. 1986; 108: 399–404.
21 Hegarty JE, Nouria Aria KT, Portmann B, Eddleston 11 Gregorio GV, Portmann B, Reid F et al. Autoimmune ALWF, Williams R. Relapse following treatment with- hepatitis in childhood 20-years experience. Hepatology drawal in patients with autoimmune chronic active hepati- 1997; 25: 541–7.
tis. Hepatology 1983; 3: 685–9.
12 Maggiore G, Veber F, Bernard O et al. Autoimmune 22 Czaja AJ, Beaver SJ, Sheils MT. Sustained remission after hepatitis associated with anti-actin antibodies in children corticosteroid therapy of severe hepatitis B surface antigen and adolescents. J. Pediatr. Gastroenterol. Nutr. 1993; 17:
negative chronic active hepatitis. Gastroenterology 1987; 92:

Source: http://pressfile.kau.edu.sa/Files/0030234/Researches/36973_AIH.pdf