Characteristics and outcome of autoimmune liver disease in Asian children

Transcription

1 DOI /s ORIGINAL ARTICLE Characteristics and outcome of autoimmune liver disease in Asian children Way S. Lee Su H. Lum Chooi B. Lim Sze Y. Chong Kim M. Khoh Ruey T. Ng Kai M. Teo Christopher C. M. Boey Jayalakshmi Pailoor Received: 30 March 2014 / Accepted: 23 June 2014 Ó Asian Pacific Association for the Study of the Liver 2014 Abstract Background Little is known about autoimmune liver disease (AILD) in Asian children. We studied the clinical features and predictors of outcome in childhood AILD in an Asian population. Methods Retrospective review of AILD [autoimmune hepatitis type 1 and 2 (AIH1, AIH2), primary sclerosing cholangitis (PSC) and autoimmune sclerosing cholangitis (ASC)] seen at two pediatric liver units in Malaysia. Results At presentation, 17 (56 %) of the 32 children [19 females, 59 %; median (range) age 7.7 ( ) years] with AILD (AIH1 = 18, AIH2 = 5, PSC = 0, ASC = 9) had liver cirrhosis. At final review [median (range) duration of follow-up 4.8 (0.4 12) years], 24 patients (75 %) survived with a native liver. Twenty-one (66 %) were in remission; 19 (AIH1 = 11; AIH2 = 4, ASC = 4) were on prednisolone and/or azathioprine, one on cyclosporine and another on mycophenolate mofetil. Three (AIH1 = 3) were in partial remission. Of the two who underwent liver W. S. Lee (&) S. H. Lum R. T. Ng K. M. Teo C. C. M. Boey Department of Paediatrics, University Malaya Medical Center, Kuala Lumpur, Malaysia leews@um.edu.my W. S. Lee University Malaya Paediatric and Child Health Research Group, University Malaya, Kuala Lumpur, Malaysia C. B. Lim S. Y. Chong K. M. Khoh Paediatric Unit, Selayang Hospital, Batu Caves, Selangor, Malaysia J. Pailoor Department of Pathology, University Malaya Medical Center, Kuala Lumpur, Malaysia transplantation (LT; 6.5 %; both ASC), one died of primary graft failure after LT. Six patients (19 %) died without LT (acute liver failure, n = 1; end-stage liver disease, n = 5). The overall survival rate (native liver and survival post-lt) was 78 %. A delay in seeking treatment adversely affected the final outcome [survival with native liver vs. LT or death (duration between onset of disease and treatment; median ± standard error) = 2.5 ± months vs ± 13.3 months; p = 0.012]. Conclusions Although remission was achieved in the majority of patients with prednisolone and/or azathioprine therapy, delay in seeking diagnosis and treatment adversely affects the outcome of childhood AILD in Malaysia. Keywords Outcome Introduction Autoimmune liver disease Asian children Autoimmune liver disease (AILD) is an umbrella term for a group of diseases characterized by unresolving inflammation of the liver of unknown cause and includes conditions such as autoimmune hepatitis type 1 and 2 (AIH1, AIH2), primary sclerosing cholangitis (PSC) and autoimmune sclerosing cholangitis (ASC) [1 3]. The hallmark of these diseases is the combination of inflammatory liver histology, circulating non organ-specific autoantibodies and raised immunoglobulin G (IgG) in the absence of a known cause [3], which is believed to be the outcome of a complex interplay between immune dysregulation and environmental triggers in genetically susceptible hosts [4]. AILD occurs in both adults and children. Based on limited epidemiological studies, the incidence of AIH1 among individuals of Caucasian origin in Europe and

2 North America ranges from /100,000/year [5, 6]. The incidence is considerably lower in Asia [7]. Before the widespread use of immunosuppressive therapy, as many as 40 % of patients with severe disease died within 6 months of diagnosis [8, 9]. Prednisolone and azathioprine have been shown to improve survival and are now established as standard immunosuppressive therapy [10]. The available evidence indicates that these agents are as efficacious in children as in adults, although there are as yet no randomized, controlled, treatment trials in children [3]. However, even with the use immunosuppressive therapy, between 3 18 % of children with AILD would eventually require liver transplantation (LT) [3, 11 13]. The clinical phenotype and outcome of AIH varies with ethnic background [14 16]. Adults of Hispanic origin have the highest prevalence of cirrhosis [15], while Asians have poorer survival outcomes that may require the use of alternative or more aggressive therapy [16]. Most of the current knowledge about the clinical features and outcome of childhood AILD is based on studies from Caucasian children [1, 2, 6, 11 13, 17, 18]. Little is known about this condition in non-caucasian [19] or Asian children [20]. This present observational study attempts to bridge this knowledge gap by describing the characteristics and outcome of children with AILD encountered consecutively at two pediatric hepatology centers in Malaysia. Patients and methods The University of Malaya Medical Center (UMMC), Kuala Lumpur, and Selayang Hospital (SH), Selangor, are referral centers for pediatric liver diseases in Malaysia. SH is the only liver transplant center in Malaysia. This is a retrospective review of all children who were consecutively diagnosed to have AILD at these two centers from January 1999 to December The present study was approved by the institutional ethics review committees of both centers (UMMC and NMRR ). Data collection The following information was collected from a review of patient medical records: demographic data, presenting clinical features, biochemical parameters, histological and radiological findings, treatment response and outcome. All liver biopsies were reassessed by a single pathologist. Diagnostic criteria Diagnosis of AIH was made according to the diagnostic criteria defined by International Autoimmune Hepatitis Group [3, 21]: (1) a raised IgG level and presence of organspecific and non-organ-specific autoantibodies; (2) presence of interface hepatitis and portal plasma cell infiltration on liver histology; (3) the absence of other liver diseases of known etiology [21]. AIH was classified into two subtypes according to seropositivity for autoantibodies: patients with the presence of antinuclear antibodies (ANA, titer C 1:20) and/or antismooth muscle antibodies (anti-sma, titer C 1:20) were classified as having AIH1 and those with the presence of liver-kidney-microsomal type 1 antibody (LKM1, titer C 1:10) as having AIH2 [3]. All the antibodies were detected using the immunofluorescence method (Innova Biosciences, Cambridge, UK). The diagnosis of PSC was made in patients who had a cholestatic biochemical profile with demonstration of characteristic multifocal stricturing and dilatation of intrahepatic and/or extrahepatic bile ducts on cholangiography in the absence of a secondary cause of sclerosing cholangitis [22]. Secondary causes of sclerosing cholangitis, which include cholangiocarcinoma, choledocholithiasis and intrahepatic metastases, were excluded by hepatic imaging studies such as CT scan of the abdomen. The diagnosis of Langerhans cell histiocytosis was established by the presence of pancytopenia, typical histology of the affected tissue/organ and imaging study of the hepatobiliary system. ASC was diagnosed in patients who had features of PSC as well as positive ANA and/or anti-sma, elevated IgG and liver biopsy abnormalities consistent with AIH [2]. Serological tests for viral hepatitis A, B, C and E, cytomegalovirus as well as Epstein-Barr virus excluded the presence of these infections. Wilson disease was excluded by ophthalmologic examination for Kayser-Fleischer rings together with measurements of ceruloplasmin, serum copper and urinary copper before and after penicillamine challenge. Alpha-1-antitrypsin deficiency was excluded by the absence of characteristic clinical features and determination of plasma levels of alpha-1-antitrypsin level and phenotype. Pattern of presentation Patients were deemed to have an acute hepatitis pattern if they had presenting features of fever, jaundice, vomiting, abdominal pain and diarrhea indistinguishable from clinical features of acute viral hepatitis [13]. Criteria for acute liver failure were the presence of evidence of acute liver injury with coagulopathy unresponsive to vitamin K prothrombin time [(PT) C 15 s or international normalized ratio (INR) C 1.5 with encephalopathy or PT C 20 s or INR C 2.0 without encephalopathy] in the absence of any known pre-existing chronic liver disease [23].

3 Histology Liver biopsy was performed at initial referral in most patients. Histopathology was re-evaluated by a single pathologist. The number of portal tracts available for review was noted. The presence of interface hepatitis, lymphoplasmacytic infiltrate and rosette formation was recorded [3, 13]. Biliary features, including ductular transformation, biliary epithelial damage and sclerosing lesions, were graded as absent, mild, moderate or severe [2]. Cholangiography Cholangiography was performed by means of either magnetic resonance cholangiopancreatography (MRCP) or endoscopic retrograde cholangiopancreatography (ERCP). Cholangiography was performed in patients who developed features of cholestasis such as raised conjugated bilirubin and/or gamma glutamyltranspeptidase during the course of monitoring their disease. Treatment In both centers, prednisolone with or without azathioprine was the first-line therapy for patients with AIH [3]. Azathioprine, started at an initial dose of 0.5 mg/kg body weight and gradually increased to 2 mg/kg body weight, was added when maintenance of remission could not be sustained without steroid toxicity [3]. Pre-treatment testing of thiopurine methyltransferase (TPMT) levels could not be performed as this test was not available. Patients with inadequate response to first-line therapy were given second-line therapy consisting of mycophenolate mofetil (MMF) or cyclosporine (CsA) [3]. CsA was started at a dose of 5 mg/kg/day, with a targeted therapeutic (trough) level of ng/ml [11]. The initial dose of MMF was 20 mg/kg/day, increasing to a maximum of 40 mg/kg/day [24]. Levels of MMF were not monitored. Prior to 2010, high-dose ursodeoxycholic acid (UDC, up to 30 mg/kg/ day) was also administered in addition to immunosuppression as the standard therapy for PSC and ASC [3]. Subsequent to emerging evidence about the possible deleterious effect of UDC in patients with PSC in 2009 [25], this practice was gradually abandoned in both centers. Response to treatment and remission were defined as normalization of liver enzymes. Partial remission referred to improvement in liver enzymes without normalization [3]. Relapse was defined as the presence of elevation of liver enzymes in isolation or in combination with histological evidence of disease activity when this was available [3]. Patients who remained in continuous remission for 2 years together with a documented normalization of IgG levels were offered discontinuation of therapy. Treatment was restarted in patients who relapsed after discontinuation of treatment [3]. Assessment of prognostic indicators For classification of the final outcome, patients were divided into two groups. Group 1 comprised patients who survived without LT (either in full or partial remission with or without immunosuppressive therapy). Group 2 consisted of patients who died as a result of the disease without LT or those who survived with a LT. The following clinical and laboratory parameters were compared between the groups: gender and age at onset of disease, interval between the onset of first symptoms and diagnosis as well as biochemical and hematological parameters at diagnosis. Statistical analysis Data were managed with IBM SPSS statistical package version for Windows. Dichotomous measures were compared by means of 2 test. If the minimum expected frequency requirements for 2 test were not met, the Fisher exact test was used instead. The Mann- Whitney U test was used for continuous variables. Multivariate logistic regression was used to weigh the parameters found to be significant in the univariate analysis. Statistical significance was set at a p \ Results A total of 32 children who were referred to the two centers during the study period fulfilled the diagnostic criteria for AIH 1 and 2, PSC and ASC (UMMC, n = 25; SH, n = 7). Of these, 18 (56 %) children had AIH1, 5 (16 %) had AIH2, and 9 (28 %) had ASC. No cases of PSC were encountered. There were patients from all major ethnic groups in Malaysia (Malays, n = 14; Chinese, n = 10; Indians, n = 7; native Sabahan, n = 1; Table 1), with a female preponderance (n = 19, 59 %). The median age at presentation was 7.7 years (range years). Clinical features at onset of the disease The median duration between the onset of the first symptom and referral was 2.0 months (range 2 days 8 years). The two most common modes of presentation were acute hepatitis (n = 13, 41 %) and insidious onset of symptoms

4 Table 1 Demographic, clinical characteristics and laboratory data of 32 Malaysian children with autoimmune liver disease AIH1 AIH2 ASC All Number of patients (%) 18 (56) 5 (16) 9 (28) 32 (100) Female (%) 12 (67) 4 (80) 3 (33) 19 (59) Age at diagnosis [median, in years (range)] 7.7 ( ) 8.7 ( ) 5.9 ( ) 7.7 ( ) Duration of symptoms [median, in months (range)] 2.0 ( ) 0.9 ( ) 24 ( ) 2.0 ( ) Pattern of presentation Acute hepatitis (%) (41) Insidious onset (%) (34) Complications of chronic liver disease (%) (13) Incidental findings (%) (13) Acute liver failure (%) (6) Associated autoimmune disorders (%) 4 (22) 1 (20) 3 (33) 8 (25) Systemic lupus erythematosus Inflammatory bowel disease a Autoimmune thyroiditis Type 1 diabetes mellitus Family history of autoimmune disorders (%) 3 (16) 0 (0) 1 (14) 4 (14) Antinuclear antibody Positive Negative Antismooth muscle antibody Positive Negative Antiliver, kidney, microsomal antibody Positive Negative Evidence of cirrhosis at presentation 8 (44) 3 (60) 6 (67) 17 (53) AIH1 autoimmune hepatitis type 1, AIH2 autoimmune hepatitis type 2, ASC autoimmune sclerosing cholangitis a One patient has Crohn s disease and two have ulcerative colitis (n = 11, 34 %), which included jaundice (n = 10), abdominal pain (n = 3), anorexia (n = 2), pruritus (n = 2) and loss of weight (n = 1). Four children (13 %) were diagnosed during the course of evaluation and follow-up for other problems. Three (9 %) were under monitoring for other autoimmune diseases while one was being investigated for poor weight gain. Two patients with AIH1 presented with acute liver failure (6 %). Presenting symptoms at the onset of disease were fever (n = 8, 25 %), weight loss (n = 6, 19 %) and anorexia (n = 6, 19 %). Two thirds of the patients had jaundice and liver enlargement (n = 21, 66 % each). Common signs at the onset of disease were splenomegaly (n = 12, 44 %), stigmata of chronic liver disease (n = 7, 22 %), pruritus (n = 5, 16 %) and ascites (n = 4, 13 %). One child had encephalopathy at presentation. A total of 17 patients (53 %) had evidence of cirrhosis at diagnosis. A diagnosis of cirrhosis was based on clinical manifestations (n = 4, all had bleeding esophageal varices), imaging studies (n = 7), liver biopsy (n = 12) or any combination of the three. Associated autoimmune and other disorders One quarter of the children (n = 8, 25 %) had an associated autoimmune disorder (Table 1). Three patients each had SLE and inflammatory bowel disease [IBD; UC = 2, Crohn s disease (CD) = 1]. The onset of AILD preceded the onset of other autoimmune disorders in four children, while in the remaining four patients the autoimmune conditions preceded the onset of AILD. Two children had pre-existing medical conditions not related to AILD. One boy had a right pelvi-ureteric junction obstruction requiring surgical re-implantation of the ureter, while another girl had a germinoma of the brain, requiring surgical removal and chemotherapy. The latter patient was in clinical remission of the tumor at the time of the diagnosis of AIH1. Family history Three patients (9 %) had a first-degree relative and another a second-degree relative with other autoimmune

5 Table 2 Treatment and outcome of 32 Malaysian children with autoimmune liver disease conditions. A girl with AIH1 and SLE had a sister with SLE. The father of a girl with AIH1 also had AIH1 himself. A girl with AIH1 and autoimmune thyroiditis had several female family members with autoimmune thyroiditis. The grandfather of a boy with ASC had autoimmune thyroiditis. Laboratory features at presentation Nine children (28 %) in the present study had normal serum bilirubin levels (B 17 lmol/l) at initial presentation. Of these, five patients with no pre-existing disease were found to have raised liver enzyme levels during routine screening for health insurance, while four were found to have raised liver enzyme levels during monitoring of the pre-existing conditions (other autoimmune conditions, n = 3; poor weight gain, n = 1). The median alanine transferase level was 331 IU/l (range IU/l). Fourteen children (44 %) had an INR level C 1.5, with two (6 %) having an INR C 2.0. Antibodies AIH1 n (%) AIH2 n (%) ASC n (%) All n (%) Number of patients 18 (56) 5 (16) 9 (28) 32 (100) Alive, in remission (66) Prednisolone monotherapy Prednisolone? azathioprine MMF MMF then CsA Alive, in partial remission (9) MMF then CsA Non-adherence to therapy Alive after liver transplant (3) Died after liver transplant (3) Died without liver transplant (19) Fulminant failure Disease complication Progressive disease Refused continual therapy Refused initial therapy AIH1 autoimmune hepatitis type 1, AIH2 autoimmune hepatitis type 2, ASC autoimmune sclerosing cholangitis, CsA cyclosporin, MMF mycophenolate mofetil Of the 18 patients with AIH1, 15 were positive for ANA (titer ranged from 1:40 to 1:5,120; Table 1). The three patients who were negative for ANA were positive for anti- SMA. All five patients with AIH2 were positive for LKM1 antibody. Three of the five patients with AIH2 were also positive for ANA antibody. LKM1 antibody was negative in all 18 patients with AIH1. All nine patients with ASC were positive for ANA and negative for LKM1 antibody. Histology Liver biopsies of 29 patients were available for review. Overall, the liver architecture was normal in 12 (41 %) cases. In 11 (38 %) cases, a few portal tracts showed fibrosis. In the remaining six (21 %) cases, there were features of cirrhosis, four of which showed on-going inflammation at the portal tracts with interface hepatitis. Portal inflammation was noted in 25 (86 %) cases and was mild in 9, moderate in 9 cases and severe in the remaining 7 cases. Interface hepatitis was noted in 17 (59 %) cases and involved all the portal tracts in 10 cases. The inflammatory cell infiltrate included lymphocytes, plasma cells, neutrophils and eosinophils. Severe fatty change was noted in one case. Cholangitis was noted in one case. Mild bile duct proliferation was present in eight cases. Cholangiography Fifteen patients (47 %) underwent either MRCP (n = 14) or ERCP (n = 1). Of the seven children with ASC who had a cholangiographic study, six had features suggestive sclerosing cholangitis. Diagnosis of ASC was confirmed histologically in the remaining three children who either had a normal MRCP study (n = 2) or did not undergo cholangiography (n = 1). All three had histological features of small duct disease. Treatment and outcome The final status of all patients was ascertained (median duration of follow-up: 4.8 years; range years; Table 3). Of these, 24 patients (75 %) were still alive with their native liver (Table 3). All but one patient were started on therapy with steroids after diagnosis. Remission (n = 20; 66 %) Twenty-one patients (66 %; Table 2) were in full remission (AIH1, n = 13; AIH2, n = 4; ASC, n = 3): (1) two patients with AIH1 were on prednisolone monotherapy; (2) 11 patients with AIH1 and three with AIH2 were on prednisolone followed by the addition of azathioprine; (3) one patient with AIH2 was on CsA after failure to respond to azathioprine and MMF; (4) four patients with ASC were given UDC in addition to prednisolone and azathioprine. None of the patients were able to discontinue drug therapy. Nine (43 %) of the 21 patients had at least one episode of

6 relapse (median number of episodes = 2; range 1 10 episodes). Partial remission (n = 3; 9 %) One patient who had liver cirrhosis and portal hypertension was initially started on prednisolone and azathioprine. Therapy was subsequently changed to MMF. A repeat biopsy 3 years later showed persistent inflammatory activity with worsening of liver cirrhosis. Partial remission was achieved when CsA was started. The remaining two patients who were in partial remission were non-compliant to treatment. The treatment of an 11-year-old who had 3 years of therapy was discontinued by his parents. He had liver cirrhosis and was in partial remission. Another boy with AIH1 responded to initial immunosuppressive therapy (prednisolone and azathioprine) although he did not achieve full remission with normalization of liver enzymes. He was treated for a total of 3 years but refused further treatment. Liver transplant (n = 2; 6 %) Two patients with ASC underwent LT. The indication for LT in one child was unacceptable quality of life with intense pruritus. This 7-year-old boy died because of primary non-function of the liver graft. The second child, an 8-year-old girl, underwent LT for advanced liver disease and was alive 3 years post-lt. Deaths (n = 7; 22 %) A total of seven patients died, six without LT and one after LT. Of the six patients (19 %) who died without LT, three (AIH1, n = 2; ASC, n = 1) died after adequate immunosuppressive therapy but without LT (Table 2). The first was a 6-year-old girl with AIH1 who presented with acute liver failure after a 2 week history of jaundice. The INR was 2.7 on admission. She did not respond to methylprednisolone therapy and succumbed without LT. The second was a girl who with both AIH1 and SLE, who was in remission with prednisolone monotherapy but subsequently died of a ruptured appendicitis at another hospital. The third was a child with ASC who was first diagnosed at 8 years of age. She was treated with prednisolone, azathioprine and UDC. Following numerous relapses, she succumbed to progressive disease at 15 years of age without LT. Another three patients (ASC, n = 2; AIH1, n = 1) died without therapy or as a result of insufficient therapy. The first was a 5-year-old boy with ASC who responded initially to prednisolone and azathioprine. Due to noncompliance with therapy, the child succumbed 5 years after the initial diagnosis. The second was a 3-year-old boy with ASC who had experienced intermittent jaundice for 2 years. The family of this child who had signs of advanced cirrhosis at diagnosis refused treatment and the child succumbed without therapy. The third was a girl with AIH2 who was first diagnosed at 3 years of age. She did not continue follow-up and returned at 10 years of age with advanced liver cirrhosis. She was given intravenous methylprednisolone but succumbed while undergoing assessment for LT. Final outcome At final review, 75 % (n = 24) of patients were alive with their native liver, 20 % (n = 6) died without LT, 3 % (n = 1) were alive after LT while 3 % died after LT. The overall survival (both native liver and after LT) was 78 %. Prognostic indicators Univariate analysis was conducted to determine predictors of favorable (group 1; survival with native liver in either full or partial clinical remission; n = 24) or unfavorable outcome (group 2; LT or death as a result of end stage liver disease; n = 8) (Table 3). Patients in group 1 had a significantly shorter duration between the onset of first symptoms and seeking treatment compared to those in group 2 [median duration (month) ± standard error = 2.5 ± 2.9 vs ± 13.3; p = 0.012; Table 3]. Comparison with studies from Caucasian populations A search of the literature was made to ascertain the subtypes, characteristics and outcomes of childhood AILD (Table 4). Of the nine studies identified, one study that was based on a therapeutic trial on CsA on AIH did not provide sufficient clinical details and was excluded from analysis [11]. Seven of the eight studies selected were from Caucasian children, while the remaining one was from Iranian children. Overall, the presenting features and outcome of childhood AILD in the present study are fairly similar to those described in the Caucasian children. Generally, there is a female preponderance (49 96 %). The median age at diagnosis ranged from 7.1 years (present study) to 12.9 years (Cleveland, OH, USA) [17]. Acute liver failure was an an uncommon mode of presentation ( %). The survival rate with native liver ranged from 75 % (present study) to 100 % (Denmark) [18], while the overall survival (native liver and LT) ranged from 78 % (present study) to 100 % (Denmark) [18].

7 Table 3 Assessment of prognostic factors in childhood autoimmune liver disease Parameters All n = 32 Group 1 (alive with native liver) n = 24 Group 2 (died or with LT) n = 8 p All numerical values are expressed in median ± standard error a The duration of the onset of the first symptom and diagnosis Sex Males Females Age at diagnosis (years) 7.7 ± ± ± Presence of comorbidity Yes No Duration to diagnosis (months) a 3.4 ± ± ± Jaundice at diagnosis Yes No Signs of cirrhosis at diagnosis Yes No Albumin (g/l) 33 ± 1 33 ± 1 33 ± Total bilirubin (lmol/l) 74 ± ± ± Alanine transferase (IU/l) 369 ± ± ± Aspartate transferase (IU/l) 434 ± ± ± Alkaline phosphatase (IU/l) 435 ± ± ± Gamma glutamyltranspeptidase (IU/l) 176 ± ± ± International normalized ratio (INR) 1.4 ± ± ± Discussion The present study is the first of its kind on childhood AILD in Asian children. The current knowledge on childhood AILD is based largely on studies from Caucasian populations [1, 2, 6, 11 13, 17, 18]. There are only a few reports of AILD from non-caucasian children. Rafeey et al. [19] described 60 Iranian children with AIH. However, no detailed description on outcome was available, and no cases of ASC or PSC were included [19]. A study by Yeh et al. [20] on nine Taiwanese children with AIH did not distinguish between AIH 1 and AIH 2, nor did it include cases of ASC or PSC. The characteristics of childhood AILD from this Asian population are similar to what has been reported in Caucasian children [1, 2, 6, 11 13, 17, 18, 26, 27]. There is a female preponderance (59 % female). However, the median age at first diagnosis (7.1 years) was somewhat younger than that in similar studies (Table 4). Acute liver failure was uncommon [1, 12, 13, 17]. The proportion of patients who had liver cirrhosis (56 % in the present study) at initial diagnosis was also similar to those observed by others [12, 17]. AILD may be associated with other autoimmune conditions [3, 26]. Studies based predominantly on Caucasian populations have noted a strong association of AILD and IBD. In particular, PSC is strongly associated with UC [6]. In contrast, diseases such as SLE are rarely encountered in children with AILD in these populations. In adult and pediatric patients with SLE, AIH is an uncommon association [28]. Of the 361 reported cases of childhood AILD in Caucasian children in the literature [2, 3, 6, 12, 13, 17, 18], only three cases (0.09 %) of SLE were noted. However, in one study, AIH was seen in up to 10 % of Caucasian children with SLE [28]. There appears to be a different pattern in studies of children from Asian populations where SLE is more commonly encountered among children with AILD. In a study from Taiwan [20], 6 (67 %) of the 9 children with AIH had SLE, whereas 3 (10 %) of the 32 cases of childhood AILD had SLE. Several genes confer susceptibility to AIH [26]. In European and North American populations, the strongest associations are found within the HLA-DRB1 locus, including the alleles encoding the HLA-DR3 (DRB1*0301) and DR4 (DRB1*0401) [26]. The HLA-DR2 (DQB1*0501 and DQB1*0601) locus has been reported to be consistently associated with SLE in both Caucasian and Asian populations [29, 30]. Further genetic studies in patients with AILD and SLE are necessary to identify genes that

8 Table 4 Demography, clinical features and outcome of childhood autoimmune liver disease in selected series Study London, UK London, Melbourne, Australia Birmingham, UK a UK Cleveland, USA Denmark b Utah, USA Iran Kuala Lumpur, Malaysia (present study) Year (reference) 1997 (1) 2001 (2) 2001 (12) 2010 (13) 2010 (17) 2012 (18) 2013 (6) 2007 (19) 2014 Center Referral center Referral center Referral center Referral center Referral center Populationbased Populationbased Referral center Referral center Number of patients Nature of disease AIH only ASC only AIH only AIH and ASC AIH AIH and ASC AIH, ASC, AIH only AIH, ASC PSC Ethnicity European- Caucasoid European- Caucasoid European-Caucasoid European- Caucasoid European- Caucasoid and Black European- Caucasoid European- Caucasoid Iranian Asian Female (%) 39 (75) 15 (55) 22 (73) 60 (60) 25 (76) 16 (49) 43 (51) 52 (96) 19 (59) Age at diagnosis (median, years) 10.5 for AIH1 7.4 for AIH (mean) Not stated AIH 1 (%) 32 (62) 18 (60) 67 (67) 30 (91) 29 (88) 44 (52) c 40 (67) 18 (56) AIH 2 (%) 20 (38) 3 (10) 18 (18) 3 (9) 2 (6) 14 (23) 5 (16) ASC/AISC (%) 27 (100) 16 (16) 6 (18) b 12 (14) 9 (28) Others (%) 4 (13) were negative for ANA/SMA/LKM ALF at presentation (%) Follow-up (median, range in years) Outcome Alive with native liver (%) 2 (6) were negative for autoantibody PSC = 29 (34) 6 (11.5) 0 (0) 1 (3) 7 (7) 2 (6) 0 (0) 4 (7) 2 (6) 5 (0.3 19) 6 (2 16) Mean: 10.0 ± (2.6 20) 6.1 ( ) 4.5 ( ) Mean: 5.9 ( ) 40 (77 %) 23 (85 %) 27 (89 %) 83 (82 %) 22 5y:66% 10 y: 58 % 33 (100 %) 5-year survival rate: PSC: 78 % ASC: 90 % AIH: 87 % Range: 1 20 months 4.8 (0.2 12) 54 (96 %) 24 (75 %) Died, no LT (%) 3 (6) 0 (0) 2 (7) 0 (0) 2 (6) 0 (0) 2 (4) 6 (19) Alive post-lt (%) 6 (12) 4 (15) 1 (3) 15 (15) 5 (15) 0 (0) 0 (0) 1 (3) Died, post-lt (%) 3 (6) 0 (0) 0 (0) 3 (3) 4 (12) 0 (0) 0 (0) 1 (3)

9 Table 4 continued Cleveland, USA Denmark b Utah, USA Iran Kuala Lumpur, Malaysia (present study) London, Melbourne, Australia Birmingham, UK a UK Study London, UK 33 (100) 52 (96) 25 (78) 46 (88) 27 (100) 28 (92) 98 (97) 5 y: 88 % 10 y: 80 % Overall survival (%; native liver and LT) ASC autoimmune sclerosing cholangitis, PSC primary sclerosing cholangitis, LT liver transplant, ANA antinuclear antibody, SMA antismooth muscle antibody, LKM antiliver, kidney microsomal antibody Only 23 cases with ASC of the total 51 children were included for comparison. The remaining 28 children in the original cases had AIH 1 and 2 In this study, AISCs were described as associated autoimmune disorders AIH 1 and AIH 2 combined. AIH 1: autoimmune hepatitis type 1, AIH 2: autoimmune hepatitis type 2, ALF acute liver failure a b c confer susceptibility to both diseases of patients, particularly in Asian populations. The incidence of AIH1 among adult Caucasian populations in Europe and North America ranges from / 100,000/year [4, 5]. Estimates from Asia are much lower, with AIH having an estimated incidence of between 0.08 and 0.15 cases/100,000/year in Japan [31] and 0.52/ 100,000/year in Taiwan [32]. The incidence of AILD among Asian children is unknown. In the present study, 32 cases of AILD were reported from two referral centers in Malaysia over a 14-year period (2.3 cases/year). The population of Malaysia aged 14 years and below in 1999 and 2010 were 7,885,600 and 7,827,900, respectively (nearest available population census figures corresponding to the period of the present study), giving an average of 7,856,750 between 1999 and 2010 [33]. Thus, the estimated incidence of childhood AILD in Malaysia is 0.03 case/100,000 children/ year, while the estimated incidence of childhood AIH (AIH1 and 2) is 0.02 cases/100,000 children/year. However, it should be pointed out that the present study was hospital based and may be limited by referral bias, as only the more severe cases may have been referred. This present figure is most likely an underestimation of the true incidence of childhood AILD in Malaysia. Nevertheless, similar to studies in adult populations, AILD appears to be much more uncommon in Asian children than in the Caucasian children. Evidence from adult populations suggests that ethnicity may affect the outcome of AIH [14 16]. It has been shown that adult Hispanics with AIH had the highest prevalence of cirrhosis [15] and Asians had poorer survival outcomes [16]. There are no studies on the effect of ethnicity on the outcome of childhood AIH or AILD, as these conditions are much rarer in children than in the adults. Furthermore, the ability to make comparisons on the outcome of childhood AILD between various populations is hindered by wide variation among available studies in the literature with regard to inclusion criteria, the nature of cases reported, the referral patterns of the respective centers and the study population, i.e., whether it was community based or based on a clinical population at a referral center. The outcome of the present study is comparable to those reported by other referral centers (Table 4). The transplantfree survival rate was 75 %. This was adversely affected by several factors. An advanced stage of disease at initial presentation may have adversely affected the outcome of three patients, whereas adverse social and family situations adversely affected the adherence of therapy in another two patients. In addition, the lack of access to timely LT in those with end-stage liver failure also adversely affected the outcome [34]. Only two patients underwent LT, while another six patients died without LT.

10 In the present study, a delay in diagnosis after the onset of disease was found to be the only significant factor that adversely affected the final outcome. Unlike the study by Radhakrishnan et al. [17], who noted that the presence of liver cirrhosis at initial diagnosis adversely affected the outcome of children with AIH, the presence of liver cirrhosis at initial presentation did not adversely affect the outcome in the present study. UDC was previously used routinely in patients with PSC [25]. However, after the discovery that the use of high-dose UDC may have possible deleterious effects on the outcome of PSC, its use in patients with PSC or in patients with AILD who also have features of cholangitis has largely been abandoned [25]. We acknowledge the following limitations in the present study. First, the estimated disease prevalence based on cases referred to pediatric liver centers in Malaysia is almost certainly an underestimate. Second, patients with AIH negative for ANA, SMA or anti-lkm 1 may have been missed and labeled as having cryptogenic chronic hepatitis because of limitations in diagnostic testing [3]. It is well known that some patients with AIH are negative for the conventional autoantibodies, such as ANA, SMA and anti-lkm 1 [3]. Tests for other uncommon autoantibodies that are used in the diagnosis of AIH, such as LKM 3, antisoluble liver antigen/liver-pancreas (anti-sla/lp), antiliver cytosol 1 (anti-lc1), antipyruvate dehydrogenase subunit E2 (anti-pdh-e2) or perinuclear antineutrophil cytoplasmic antibodies (panca), were not available for patients in the present study [3]. In conclusion, childhood AILD is an uncommon condition among Asian children and is likely to be less common than in Caucasian populations. SLE is an important associated autoimmune condition and should be routinely excluded in Asian children with AILD. A delay in referral for diagnosis and treatment has been identified as a factor adversely affecting the outcome of childhood AILD in Malaysia. Despite some adverse social factors and poor adherence to therapy, the transplant-free survival of children with AILD in Malaysia is comparable to that reported by referral centers with predominantly Caucasian patient populations. These findings suggest that more aggressive immunosuppressive therapy is not indicated in Asian children with AILD. Acknowledgements Compliance with ethical requirements and Conflict of interest The authors declare that no human subjects or animals were involved in the conduct of the present research and that it was not a clinical trial. The present study was approved by the institutional ethics review committees of both centers (UMMC and NMRR ). Way Seah Lee, Su Han Lum, Chooi Bee Lim, Sze Yee Chong, Kim Mun Khoh, Ruey Terng Ng, Kai Ming Teo, Chiong Meng Boey and Jayalakshmi Pailoor declare that they have no conflict of interest. None of the authors received any external funding in relation to the preparation of this work. References 1. Gregorio GV, Portmann B, Reid F, Donaldson PT, Doherty DG, McCartney M, et al. Autoimmune hepatitis in childhood: a 20-year experience. Hepatology 1997;25: Gregorio GV, Portmann B, Karani J, Harrison P, Donaldson PT, Vergani D, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology 2001;33: Manns MP, Czaja AJ, Gorham JD, Krawit EL, Mieli-Vergani G, Vergani D, et al. AASLD Practice Guidelines: diagnosis and management of autoimmune hepatitis. Hepatology 2010;51: Liberal R, Longhi MS, Mieli-Vergani G, Vergani D. Pathogenesis of autoimmune hepatitis. Best Prac Res Clin Gastroenterol 2011;25: Boberg KM. Prevalence and epidemiology of autoimmune hepatitis. Clin Liver Dis 2002;6: Deneau M, Jensen MK, Holmen J, Williams MS, Book LS, Guthery SL. Primary sclerosing cholangitis, autoimmune hepatitis, and overlap in Utah children: epidemiology and natural history. Hepatology 2013;58: Onji M, Nonaka T, Horiike N, Moriwaki H, Muto Y, Ohta Y. Present status of autoimmune hepatitis in Japan. Gastroenterol Jpn 1993;28(Suppl 4): Mistilis SP, Skyring AP, Blackburn CR. Natural history of active chronic hepatitis: I. Clinical features, course, diagnostic criteria, morbidity, mortality and survival. Australas Ann Med 1968;17: Soloway RD, Summerskill WH, Baggenstoss AH, Geall MG, Gitnick GL, Elveback IR, et al. Clinical, biochemical, and histological remission of severe chronic active liver disease: a control study of treatment and early prognosis. Gastroenterology 1972;63: Murray-Lyon IM, Stern RB, Williams R. Controlled trial of prednisolone and azathioprine in active chronic hepatitis. Lancet 1973;I: Cuarterolo M, Ciocca M, Velasco CC, Ramonet M, Gonzalez T, Lopez S, et al. Follow-up of children with autoimmune hepatitis treated with cyclosporine. J Pediatr Gastroenterol Nutr 2006;43: Saadah OI, Smith AL, Hardikar W. Outcome of autoimmune hepatitis in children. J Gastroenterol Hepatol 2001;16: Chai PF, Lee WS, Brown RM, McPartland JL, Foster K, McKeirnan PJ, et al. Childhood autoimmune liver disease: indications and outcome of liver transplantation. J Pediatr Gastroenterol Nutr 2010;50: Czaja AJ. Special clinical challenges in autoimmune hepatitis: the elderly, males, pregnancy, mild disease, fulminant onset, and non-white patients. Sem Liver Dis 2009;29: Wong RJ, Gish R, Frederick T, Bzowej N, Frenette C. The impact of race/ethnicity on the clinical epidemiology of autoimmune hepatitis. J Clin Gastroenterol 2012;46: Zolfino T, Heneghan MA, Norris S, Harrison PM, Portmann BC, McFarlane IG. Characteristics of autoimmune hepatitis in patients who are not of European caucasoid ethnic origin. Gut 2002;50: Radhakrishnan KR, Alkhouri N, Wolrley S, Arrigain S, Hupertz V, Kay M, et al. Autoimmune hepatitis in children impact of cirrhosis at presentation on natural history and long-term outcome. Dig Liver Dis 2010;42:

11 18. Vitfell-Pedeparsen J, Jorgensen MH, Muller K, Heilmann C. Autoimmune hepatitis in children in Eastern Denmark. J Pediatr Gastroenterol Nutri 2012;55: Rafeey M, Kianrad M, Hasani A. Autoimmune hepatitis in Iranian children. Indian J Gastroenterol 2007;26: Yeh SH, Ni YH, Jeng YM, Chen HL, Wu JF, Chang MH. Emerging importance of autoimmune hepatitis in children in Taiwan, an endemic area for viral hepatitis. Pediatr Neonatol 2009;50: Alvarez F, Berg PA, Bianchi FB, Bianchi L, Burroughs AK, Cancado EL, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999;31: Chapman R, Fevery J, Kalloo A, Nagorney DM, Boberg KM, Shneider B, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology 2010;51: Lee WS, McKiernan PJ, Kelly DA. Etiology, outcome and prognostic indicators of childhood fulminant hepatic failure in the United Kingdom. J Pediatr Gastroenterol Nutri 2005;40: Aw MM, Dhawan A, Samyn M, Bargiota A, Mieli-Vergani G. Mycophenolate mofetil as rescue treatment for autoimmune liver disease in children: a 5-year follow-up. J Hepatol 2009;51(1): Lindor KD, Kowdley KV, Luketic VAC, Harrison ME, McCashland T, Harnois D, et al. High dose ursodeoxycholic acid for the treatment of primary sclerosing cholangitis. Hepatology 2009;50: Liberal R, Grant CR, Mieli-Vergani G, Vergani D. Autoimmune hepatitis: a comprehensive review. J Autoimmun 2013;41: Mieli-Vergani G, Heller S, Jara P, Vergani D, Chang MH, Fujisawa T, et al. Autoimmune hepatitis. J Pediatr Gastroenterol Nutr 2009;49: Irving KS, Sen D, Tahir H, Pilkington C, Isenberg DA. A comparison of autoimmune liver disease in juvenile and adult populations with systemic lupus erythematosus a retrospective review of cases. Rheumatology 2007;46: Chai HC, Phipps ME, Chau KH. Genetic risk factors of systemic lupus erythematosus in the Malaysian population: a mini review. Clin Develop Immunol 2012;. doi: /2012/ Azizah MR, Ainol SS, Kong NC, Normaznah Y, Rahim MN. HLA antigens in Malay patients with systemic lupus erythematosus: association with clinical and autoantibody expression. Korean J Int Med 2001;16: Toda G, Toda G, Zeniya M, Watanabe F, Imawari M, Kiyosawa K, et al. Present status of autoimmune hepatitis in Japan correlating the characteristics with international criteria in an area with a high rate of HCV infection. Japanese National Study Group of Autoimmune Hepatitis. J Hepatol 997;26: Koay LB, Lin CY, Tsai SL, et al. Type 1 autoimmune hepatitis in Taiwan: diagnosis using the revised criteria of the International Autoimmune Hepatitis group. Dig Dis Sci 2006;51: Department of Statistics, Malaysia gov.my. Accessed 26 Mar 2014) 34. Lee WS, Chai PF, Lim KS, Lim LH, Looi LM, Ramanujam TM. Outcome of biliary atresia in Malaysia a single centre study. J Paediatr Child Health 2009;45: