See “Management of Autoimmune Hepatitis in Children: How Many Steps Away From Common Agreement?” by Cicalese and Iorio on page 364.
Autoimmune hepatitis (AIH) in childhood is a progressive chronic inflammatory liver disease with an unknown aetiology. It accounts for approximately 2% to 5% of chronic liver diseases in children (1) and for 10% to 20% among adults (2), being most frequent in girls and women (3). The diagnosis is based on histology, clinical signs, and biochemistry, including increased levels of serum immunoglobulins and the presence of autoantibodies. On the basis of the autoantibody profile, AIH is divided into 2 types (4). In type 1 anti-nuclear (ANA) and/or anti-smooth muscle (SMA) antibodies are found. An association with the tissue type human leucocyte antigen (HLA)-DR3 has been noted for this type. Type 2 is rare and anti-liver kidney microsomal antibody is present. In this subtype an association with HLA-DRB1 and HLA-DQB1 tissue types has been described (5). Autoantibodies are not detectable in a minority of patients with AIH (1). As a part of an overlap syndrome AIH is frequently associated with other chronic inflammatory diseases, including inflammatory bowel disease or autoimmune sclerosing cholangitis (ASC) (6,7). This is seen in 6% to 8% of patients with AIH (7). At presentation the manifestations of AIH are heterogeneous. Nonspecific symptoms of varying severity such as fatigue, lethargy, malaise, anorexia, nausea, abdominal pain, and itching are common (5). In 25% to 50% a presentation similar to that of acute hepatitis may be seen, including fever, jaundice, abdominal pain, and hepatosplenomegaly (1). The histology is characterised by dense mononuclear and plasma cell infiltration of the portal tract (8).
The first-line treatment of AIH is prednisolone alone or in combination with azathioprine. In case of severe adverse effects or if remission cannot be achieved, second-line medication with 6-mercaptopurin, tacrolimus, cyclosporine, or mycophenolate mofetil (MMF) may be used (9,10). Most patients need lifelong treatment and few achieve permanent remission (5). Combination treatment with azathioprine and prednisolone has been shown to improve on histhopatholgical evaluation the fibrosis scores, with an arrest in its progression and no development into cirrhosis (11). On rare occasions liver transplantation may be necessary, but AIH has been reported to recur after transplantation in 10% to 35% of the patients (12,13).
This article describes the clinical and biochemical characteristics of a population-based paediatric cohort diagnosed with AIH and evaluates the efficacy of early treatment strategies.
In the period between January 1993 and September 2009, 33 children diagnosed as having AIH were treated at the paediatric department of the National University Hospital, Rigshospitalet, Copenhagen, Denmark. This is a referral centre for liver diseases in children from the eastern part of Denmark, with approximately 2.5 million citizens. To locate further cases in the area, all of the paediatric departments in eastern Denmark were asked about AIH cases. None of these departments treated patients who were not already known to the referral centre. Thus, the present patient material can be regarded to be population based. The patients were identified prospectively and the data were collected by retrospective examination of the files. Infections with hepatitis A, B, C, D, E, cytomegalovirus, and Epstein-Barr virus were excluded by serology. Only 3 patients were tested for hepatitis E and 5 for hepatitis D. They were found to be hepatitis E and hepatitis D negative. Alfa-1-antitrypsin deficiency was excluded by plasma levels of alpha-1-antitrypsin and, in 3 cases, also by gene analysis (PiZZ and PiSS). Wilson disease was excluded by measurements of ceruloplasmine, serum-copper, urine-copper, and/or absence of Kayser-Fleischer ring. In a single patient Wilson disease workup was not performed; however, he was found to be SMA positive, had a liver biopsy that indicated AIH with plasma cells in the portal tract and fibrosis grade 3, and responded well to immunosuppressive treatment. All measurements of autoantibodies were carried out at the Statens Serum Institute of Copenhagen. All of the liver biopsies were performed at Rigshospitalet and were examined by a liver pathologist. The ultrasound reports were conducted at the referring hospital or at Rigshospitalet, and all were rescored based on the liver parenchyma being normal, nodular, or coarse, according to the William ultrasound scoring scale (WUSS) (14).
Patients were classified as having acute hepatitis symptoms at onset if they presented with >3 symptoms characteristic of acute hepatitis, including fever, jaundice, abdominal pain, diarrhoea, nausea, hepatosplenomegaly, and vomiting.
After diagnosis all of the patients received first-line treatment consisting of either prednisolone (1–2 mg · kg−1 · day−1) alone or prednisolone in combination with azathioprine (1–2 mg · kg−1 · day−1). The patients were seen as outpatients weekly where blood samples were drawn for tests until the prednisolone dose had been tapered to a proximally 0.5 mg/kg. When alanine aminotransaminase (ALT) levels had stabilised, the patients were seen every 1 to 3 months at which time blood samples were drawn for tests.
Remission was defined as an ALT ≤ 50 μ/L on a minimum of 2 occasions, at least 1 month apart. Relapse was defined as an ALT > 100 μ/L after achieved remission, and the stable remission was defined as an ALT ≤ 50 μ/L in a patient not receiving any immunosuppressive treatment. An insufficient response to treatment was defined by the following criteria: need for second- or third-line treatment with 1 of the following 4 drugs: 6-mercaptopurine, tacrolimus, cyclosporine, or MMF; or first-line treatment for >18 months without remission; or liver transplantation.
Quantitative data were presented as number, median, range, or percentages. The Fisher exact test was used for the calculation of statistical significance.
The demographic data are described in Table 1, and the laboratory findings are described in Table 2. Of the 33 children with AIH, 16 were girls and 17 were boys. Twenty-nine patients (87.9%) were diagnosed as having type 1, 2 (6.1%) with type 2, and in 2 (6.1%) patients no autoantibody were detected. The median duration of the symptoms before diagnosis was 4 weeks. In 23 (69.7%) patients the symptoms at presentation were similar to those of acute viral hepatitis; however, in 16 (69.6%) of these patients the liver biopsies showed cirrhosis. On physical examination 12 (36.4%) patients had hepatomegaly and 2 (6.1%) had splenomegaly. In 1 patient (3%) elevated transaminases was found to be the only sign of liver disease.
All of the patients had increased transaminases. Twenty-six (78.8%) patients had hyperbilirubinaemia and 24 (77.4%) had increased international normalization ratio (INR). Twenty-nine (87.9%) patients had an elevated immunoglobulin G (IgG), 4 (12.1%) had a normal IgG, and 15 (45.5%) had a high immunoglobulin M (IgM). Immunoglobulin A (IgA) was low in 3 (9.1%) patients, whereas 1 of these had type 2 AIH. In 31 (93.9%) patients positive autoantibodies were detected. Two patients were found to have no detectable autoantibodies and in those, the diagnosis of AIH was solely based on histology and the response to immunosuppression.
Other autoimmune diseases were diagnosed in 12 (36.3%) patients. These included ASC (n = 6), type 1 diabetes mellitus (n = 1), psoriasis (n = 1), alopecia areata (n = 1), ulcerative colitis (n = 1), Addison disease (n = 1), Crohn disease (n = 1), and pernicious anaemia (n = 1). In 11 of the patients the other autoimmune disorders were diagnosed after the diagnosis of AIH. One patient was diagnosed as having both ASC and Crohn disease. One patient had insulin-dependent diabetes diagnosed before the diagnosis of AIH. In 11 (33.3%) patients, autoimmune diseases were present among parents or siblings, and among them 1 parent had AIH. Seven (21.2%) patients experienced episodes with intermittent arthralgia without evidence of synovitis.
Three (9.1%) patients developed oesophageal varices, confirmed by gastroscopy, ultrasound, or computerized tomography scan. Six (18.2%) patients were found to have possible ASC, of whom 5 were girls. Only 2 had a magnetic resonance cholangiopancreatography to confirm the diagnosis of ASC. All were commenced on ursodeoxycholic acid. In further 2, the diagnosis was suspected because of a continuous increased bilirubin and in the last 2 patients a continuous increased GGT.
Thirty-two patients had undergone a liver biopsy performed at onset. In 1 patient a biopsy was not taken, but because this patient had increased IgG and strongly positive SMA, and because other causes of liver diseases were excluded, this patient was included in the study. All of the biopsies confirmed the diagnosis based on the presence of interface hepatitis and the presence of plasma cells in the portal tract. In 24 (75%) patients, the biopsy showed cirrhosis and in 7 (21.8%) patients, it showed fibrosis. In the 24 patients with cirrhosis, clinical symptoms had lasted for a median of 42 days (range 1 day–5 years).
One hundred sixty ultrasound examinations were performed. Fibrosis and/or cirrhosis were seen in 17 (51.5%) patients. Four (12.1%) had signs of fibrosis at diagnosis but developed normal parenchyma during their treatment. Fifteen (45.5%) patients had splenomegaly at diagnosis, and in 2 the splenomegaly disappeared during treatment. Overall, 27 (81.8%) children had an ultrasound performed that, at some point, showed an abnormality, with splenomegaly to be the most frequent finding.
Treatment and Response to Treatment
Treatment, follow-up, and response are described in Table 3. Twenty (60.6%) patients were treated initially with prednisolone alone and 13 (39.4%) with a combination of prednisolone and azathioprine. Of the 20 patients, 18 (90%) were treated for 3.6 months (median; range 7 days–3.7 years) before azathioprine was added to their treatment. Six patients received treatment with prednisolone and MMF because of insufficient response to prednisolone and azathioprine, and 5 of them moved on to receive tacrolimus because of insufficient response to MMF. The change in treatment to tacrolimus was performed 7.6 years (median; range 1.5–8.5 years) after start of treatment. None of the patients developed hepatotoxicity either in the beginning of the treatment or later.
Twenty (60.6%) patients were treated with prednisolone and azathioprine at the time of remission, whereas 8 (24.2%) received only prednisolone at remission. The median time from start of treatment to first remission was 5.4 months (range 12 days–6.7 years). No relation was found between the type of treatment (prednisolone and azathioprine vs prednisolone alone) inducing remission and the risk of later relapse.
At the end of a median follow-up of 54.2 months (range 11–11.9 years), 5 (15.2%) patients were in stable remission without any immunosuppressive maintenance drug, 16 (48.5%) patients were in remission (12 in treatment with prednisolone and azathioprine and 4 with prednisolone alone), and 12 (36.3%) patients were not in remission. The median follow-up of remission was 19.4 months (range 2.7–33 months)
Furthermore at the end of follow-up, 96.9% of the patients had achieved remission at some point. A total of 10 (30.3%) patients had an insufficient treatment response. One patient with type 1 AIH had received a liver transplantation at the age 17 years, 6 years and 9 months after the diagnosis of AIH. Another patient, also with type 1 AIH, is presently listed for liver transplantation. None of the patients have died.
The last major survey describing 73 children with AIH followed at a tertiary centre was published in 1996 (8). The present study was designed to be a descriptive population-based study of all of the paediatric patients with AIH in the eastern part of Denmark. Because data collection was retrospective, limitations in the data set do occur; however, all of the patients were seen at the same centre, which may have minimised the weakness of the retrospective setup.
In the present study the median duration of symptoms before diagnosis was 4 weeks, but in two-thirds of patients the first liver biopsy showed cirrhotic changes. This seems to indicate a lag phase between the onset of disease and the diagnosis in which AIH is a relatively silent disease with limited clinical manifestations. This is further emphasised by the fact that approximately 50% of the patients had splenomegaly at diagnosis, indicating that the disease had been long-standing. This is in accordance with previous findings (15) and emphasises the urgency of treatment despite a short duration of symptoms.
As described in the literature, AIH presented in our population mostly to be acute viral hepatitis, and 12.1% of the patients did not show an increase of IgG (1,8). IgA deficiency is reported to be common in type 2 AIH, which was also observed within our study (9).
Type 1 AIH has been described to account for two-thirds of all of the cases and generally presents during adolescence, whereas type 2 AIH presents in younger children and during infancy (9). In our study only 6.1% were diagnosed as having type 2 AIH. Hence, the distribution between type 1 and type 2 AIH differed significantly from the literature. The different distribution in type 1 and 2 AIH is not readily explained but one possibility could be that 18.2% of our patients had ASC, all of them with type 1 AIH. It has been shown that ASC could be related to the same disease process as type 1 AIH (6). Only a few of our patients (6.1%) had no detectable autoantibodies at presentation, which is in accordance with the literature (1).
Approximately 40% of our patients had signs of other autoimmunity-related chronic inflammatory diseases. This is slightly higher than that reported by Gregorio et al (8), in which such problems were described in up to 21% of the patients. The reason for this discrepancy could be that we included ASC to be an associated autoimmune disease.
The initial treatment strategy varies between centres. Some provide monotherapy with prednisolone and add azathioprine only in the presence of serious adverse effects or in case of insufficient response to treatment. Others recommend that treatment from the first days with both prednisolone and azathioprine can be considered, but only in selected patients because of the risk of azathioprine-induced hepatotoxicity, particularly in severely jaundiced patients (9). Some other authors favour prednisolone in combination with azathioprine to be the initial treatment (16,17). To maintain remission some recommend azathioprine as monotherapy after the patient has achieved remission on prednisolone and azathioprine (18). Several studies have shown positive response rates with MMF in children who are resistant to or intolerant of standard immunosuppression (19).
In our study the median time from the start of treatment to the first remission was approximately 6 months. This is less compared with the literature, in which complete biochemical remission may require 8 to 12 months or longer, but it is similar to the study by Gregorio et al, wherein the median time was 7 months (1,8).
Twenty patients were treated initially with prednisolone alone, and at the time of remission, 60.1% of the patients in our study were taking both prednisolone and azathioprine, suggesting that in the majority of the patients it was necessary to combine prednisolone with azathioprine to achieve remission. This result can be biased by the fact that the patients were not randomised to the single agent versus the combination at the start of treatment. Mieli-Vergani et al described that 85% of the patients will eventually require the addition of azathioprine to achieve remission (9).
The present study suggests that first-line treatment with prednisolone combined with azathioprine versus prednisolone alone should be investigated in future randomised studies; however, we did find that the risk of relapse in patients treated with prednisolone alone was not significantly different from the risk of relapse seen in patients treated with prednisolone in combination with azathioprine at presentation.
A female predominance with a female:male ratio of 8:1 has been described at a tertiary centre, and according to the literature, AIH is most common in girls and women (9,20). Furthermore in the study by Gregorio et al (8), 75% of their patients with AIH were girls. This was, however, not seen in this Danish cohort and is not readily explained. The relatively high percentage of boys in this study was not due to the presence of an increased number of patients with ASC because 83.3% of the patients with possible ASC were girls. A likely explanation could be the small number of patients included in this study, which meant that our cohort of patients was not large enough to reveal a female predominance. Another possible explanation could be that our cohort of patients were population based and not from a tertiary centre.
At the time of this study, 15.2% of our patients had achieved stable remission, 1 had received a liver transplantation, and none of the patients had died. In the review by Gregorio et al (8), 1.5% of the patients had stable remission, 15% had received a liver transplant, and 11.5% had died. This study was performed in 1993, and it is possible that the difference between the results of the 2 studies is caused by a higher awareness of disease-related complications and a more aggressive treatment strategy with a faster change in treatment to tacrolimus and MMF in the nonresponding patients. Furthermore, the fact that we are comparing a population-based cohort of patients with a group of patients seen at a tertiary referral centre could also contribute to the differences observed.
Approximately 30% of our patients had an insufficient treatment response and 15.2% achieved stable remission, which is in accordance with the literature in which up to 20% of patients with type 1 AIH achieve stable remission (1). Although the percentage of stable remission in our study is similar to that of the literature, the percentage may still be considered to be low, and the adverse effects to long-term use of prednisolone are a concern. Therefore, the development of improved treatment strategies based on future randomised trials will be important for paediatric patients with AIH.
1. Squires R. Autoimmune hepatitis in children. Curr Gastroenterol Rep 2004; 6:225–230.
2. Samad TM, Zadeh HM, Mashrabi O. Survey of autoimmune hepatitis in children. Res J Biol Sci 2008; 3:238–248.
3. Vergani GM, Vergani D. Autoimunne hepatitis in children: what is different from adult AIH. Semin Liver Dis 2009; 29:297–306.
4. Dimitrios PB, Vergani GM, Vergani D. Autoantibodies and their antigens in autoimmune hepatitis. Semin Liver Dis 2009; 29:241–253.
5. Krawitt EL. Autoimmune hepatitis. N Engl J Med 2006; 354:54–66.
6. Gregorio G, Portmann B, Karani J, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology 2001; 33:544–553.
7. Beuers U, Rust C. Overlap syndromes. Semin Liver Dis 2005; 25:311–320.
8. Gregorio G, Portmann B, Reid F, et al. Autoimmune hepatitis in childhood: a 20-year experience. Hepatology 1997;25:541–47.
9. Mieli-Vergani G, Heller S, Jara P, et al. Autoimmune hepatitis. J Pediatr Gastroenterol Nutr 2009; 49:158–164.
10. Larsen FS, Vainer B, Eefsen M, et al. Low-dose tacrolimus ameliorates liver inflammation and fibrosis in steroid refractory autoimmune hepatitis. World J Gastroenterol 2007; 13:3232–3236.
11. Ferreira AR, Roquete MLV, Toppa NV, et al. Effect of treatment of hepatic histopathology in children and adolescents with autoimmune hepatitis. J Pediatr Gastroenterol Nutr 2008; 46:65–70.
12. Czaja AJ. Current and future treatments of autoimmune hepatitis. Expert Rev Gastroenterol Hepatol 2009; 3:269–291.
13. Chai PC, Way LS, Brown RM, et al. Childhood autoimmune liver disease: indication and outcome of liver transplantation. J Pediatr Gastroenterol Nutr 2010; 50:295–302.
14. Williams SGJ, Evanson JE, Barrett N, et al. An ultrasound scoring system for the diagnosis of liver disease in cystic fibrosis. J Hepatol 1995; 22:513–521.
15. Maggiore G, Bernard O, Hadchouel M, et al. Treatment of autoimmune chronic active hepatitis in childhood. J Pediatr 1984; 104:839–844.
16. Ishibashi H, Komori MD, Shimoda S, et al. Guidelines for therapy of autoimunne liver disease. Semin Liver Dis 2007;27:214–26.
17. Czaja AZ, Freese DK. Diagnosis and treatment of autoimmune hepatitis. Hepatology 2002; 36:479–497.
18. Banerjee S, Rahhal R, Bishop WP. Azathioprine monotherapy for maintenance of remission in pediatric patients with autoimmune hepatitis. J Pediatr Gastroenterol Nutr 2006; 43:353–356.
19. Aw MM, Dhawan A, Samyn M, et al. Mycophenolate mofetil as rescue treatment for autoimmune liver disease in children: a 5-year follow-up. J Hepatol 2009;51:156–60.
20. Al-Mahtab M, Rahman S, Shrestha A, et al. Case report of autoimmune hepatitis from a tertiary centre in Bangladesh. Bangladesh Liver J 2009; 1:53–54.