Diagnosing Autoimmune Hepatitis in Children: Is the International Autoimmune Hepatitis Group Scoring System Useful?

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2004;2: Diagnosing Autoimmune Hepatitis in Children: Is the International Autoimmune Hepatitis Group Scoring System Useful? REGAN L. EBBESON* and RICHARD A. SCHREIBER Division of Gastroenterology, *Department of Pediatrics, The University of British Columbia and BC Children s Hospital, Vancouver, British Columbia, Canada Background & Aims: In 1999, the International Autoimmune Hepatitis Group (IAIHG) modified a scoring system to differentiate adult patients with definite or probable autoimmune hepatitis (AIH) from those with other forms of chronic liver disease. We assessed the use of the scoring system in children. Methods: Twenty-eight pediatric patients with AIH and/or sclerosing cholangitis were reviewed. Clinical, laboratory, and histologic data were collected to score patients both before and after standard treatment. Results: There were 8 boys and 20 girls. The median age at diagnosis was 11 years (range, 2 16 years). Twenty-one of 28 children were diagnosed with AIH, 4 as isolated primary sclerosing cholangitis (PSC), and 3 as overlap syndrome. At presentation, 18 of 21 (86%) with AIH scored as definite AIH and 3 of 21 (14%) scored as probable. No patient clinically diagnosed as AIH scored as other. Seven of 28 patients had proven PSC. All patients with isolated PSC scored as other. The 3 with overlap syndrome scored as definite AIH. When the -glutamyltranspeptidase (GGT) ratio was substituted for the alkaline phosphatase (ALP) ratio, 5 patients were reclassified from definite to probable AIH. Four of these 5 had an incomplete response to therapy, and 2 of 4 have confirmed overlap syndrome. Conclusions: The IAIHG scoring system has a use in children. Patients who fall into the other category should have cholangiographic imaging. Using the GGT ratio instead of the ALP ratio in the IAIHG score may improve the specificity for children, identifying those likely to have biliary disease. When GGT is used, patients classified as needing probable pretreatment should be considered for biliary imaging. In 1992, the International Autoimmune Hepatitis Group (IAIHG) developed a set of descriptive criteria and a scoring system that it recommended could be applied to differentiate adult patients with definite or probable autoimmune hepatitis (AIH) from those with other forms of chronic liver disease such as primary biliary cirrhosis or primary sclerosing cholangitis (PSC). 1 Several large studies found the scoring system to have an overall diagnostic sensitivity and specificity ranging from 97% to 100% and 44% to 87%, respectively. 2 The main deficiency of the system was its inadequate specificity with respect to excluding a diagnosis of AIH in those with biliary disease. The IAIHG published a modified scoring system in Both Kaya et al. 3 and Omagari et al. 4 concluded in separate reports that the modified scoring system might better discriminate between PSC and AIH and more precisely define the overlap between these 2 conditions. The use of the modified IAIHG scoring system in pediatric patients has not been well defined. Indeed, some of the criteria applied in the scoring scheme have little practical significance to pediatric patients. 2 For example, the level of alcohol intake rarely is relevant to the evaluation of children with chronic liver disease. Moreover, the IAIHG scoring system uses alkaline phosphatase (ALP) rather than -glutamyltranspeptidase (GGT) as the enzymatic indicator of a cholestatic process. A high serum ALP level may be a less specific marker for cholestatic liver disease in children because it also may reflect bone activity in the growing child. In the present study, we applied the IAIHG scoring system retrospectively to pediatric patients with chronic immune liver disease to determine its use in classifying children as having either definite or probable AIH or PSC. We gave consideration to the exclusion of average alcohol intake and analyzed the variations in classification when GGT was substituted for ALP in the scoring system. Abbreviations used in this paper: AIH, autoimmune hepatitis; ALP, alkaline phosphatase; ANA, antinuclear antibody; ASMA, anti-smooth muscle antibody; ERCP, endoscopic retrograde cholangiopancreatography; GGT, -glutamyltranspeptidase; IAIHG, International Autoimmune Hepatitis Group; PSC, primary sclerosing cholangitis; UNL, upper limit of normal by the American Gastroenterological Association /04/$30.00 PII: /S (04)

2 936 EBBESON AND SCHREIBER CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 10 Patients and Methods All patients having a diagnosis of AIH or sclerosing cholangitis between 1984 and 2003 were identified by the health records department at BC Children s Hospital. A total of 36 cases were reviewed. All of these patients had been referred to the Division of Pediatric Gastroenterology. Diagnoses of AIH were based on the patients clinical and biochemical presentation, liver histology, and their response to therapy. Diagnoses of sclerosing cholangitis were based on either endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopancreatography findings. A diagnosis of overlap syndrome was made when the patient had clinical and histologic findings in keeping with AIH, but also was found to have sclerosing cholangitis confirmed by cholangiographic study. There were 25 cases of AIH, 6 cases of isolated PSC, 3 cases of overlap syndrome, 1 case of secondary sclerosing cholangitis, and 1 case of cholelithiasis identified. Four of the cases of AIH were excluded from this study because the patients had an uncorrected coagulopathy at the time of diagnosis and no liver biopsy examination was performed before treatment initiation. Two of the cases of PSC were excluded because a liver biopsy examination, required for the IAIHG scoring system, was not performed. The one case of cholelithiasis confirmed by ERCP was excluded. The one case of presumed secondary sclerosing cholangitis was in a 14-month-old child with severe combined immunodeficiency syndrome who had not yet had an ERCP and was not included in this study. Thus, 28 cases were included in this study: 21 AIH, 4 isolated PSC, and 3 overlap syndromes. Data collected from the time of diagnosis included sex, age at onset, serum ALP level, aspartate transaminase level, alanine transaminase level, GGT level, serum bilirubin level, gamma immunoglobulin (IgG) level, antinuclear antibody (ANA) level, anti-smooth muscle antibody (ASMA) level, antimitochondrial antibody level, hepatitis viral markers, drug history, average alcohol intake, liver histology at diagnosis, other autoimmune diseases, family history of autoimmune diseases, other defined autoantibodies, and HLA DR typing if tested. Response to therapy was determined according to definitions provided by the IAIHG. Patients were scored and classified as definite AIH, probable AIH, or other, both before and after treatment according to the revised scoring system published by the IAIHG (Table 1). 2 For calculation of the ALP and GGT ratios in the IAIHG scoring system, the values for the upper limit of normal of these enzymes were both age and sex adjusted in accordance with the reference ranges provided by the BC Children s Hospital laboratory handbook. The study was approved by the University of BC ethics review committee. Results A total of 28 cases were included. There were 8 boys and 20 girls. The median age at diagnosis was 11 years (range, 2 16 years). The pretreatment score for each Table 1. IAIHG Revised Scoring System for the Diagnosis of AIH Parameters Score Female sex ALP:AST (or ALT) ratio a Serum globulins or IgG levels greater than normal ANA, ASMA, or LKM-1 1:80 1:80 1:40 1:40 0 AMA positive 4 Hepatitis viral markers Positive Negative Drug history Positive 4 Negative Average alcohol intake 25g/day 60 g/day Liver histology Interface hepatitis Predominantly lymphoplasmacytic infiltrate Rosetting of liver cells None of the above 5 Biliary changes Other changes Other autoimmune disease(s) Optional additional parameters Seropositivity for other defined autoantibodies HLA DR3 or DR4 Response to therapy Complete b Relapse Interpretation of aggregate scores Pretreatment Definite AIH 15 Probable AIH Posttreatment Definite AIH 17 Probable AIH AST, aspartate transaminase; ALT, alanine transaminase. Data from the International Autoimmune Hepatitis Group Report. 2 a The ALP:AST (or ALT) ratio relates to the degree of increase above the upper limit of normal (unl) of these enzymes (i.e., (ALP/ ALPunl)/(AST/ASTunl). b A complete response is defined as marked improvement of symptoms and return of AST or ALT, bilirubin, and immunoglobulin values completely to normal within 1 year and sustained at least a further 6 months on maintenance therapy or marked improvement of symptoms together with at least 50% improvement of all liver test results during the first month of treatment, with AST or ALT levels continuing to decrease to less than twice the upper limit within 6 months during any reductions toward maintenance therapy.

3 October 2004 USEFULNESS OF IAIHG SCORING SYSTEM 937 Figure 1. Pretreatment scores. Figure 2. Posttreatment scores. of the 21 AIH patients using the IAIHG scoring system ranged from (Figure 1). Of these, 18 of 21 patients (86%) scored 15 points, thus satisfying the score for a diagnosis of definite AIH. Three patients each scored 15 points, thus placing them in the probable AIH category (10 15 points). No patient who carried a clinical diagnosis of AIH scored as other ( 10 points). Of the 7 patients who had proven sclerosing cholangitis (6 by ERCP, 1 by magnetic resonance cholangiopancreatography), 4 had a diagnosis of isolated PSC and 3 held a diagnosis of overlap syndrome. All patients who had a diagnosis of isolated PSC scored as other (range, 1 9 points). The 3 patients with a diagnosis of overlap syndrome scored in the definite AIH category pretreatment (Table 2). All 3 patients went on to have an incomplete response to therapy. In the posttreatment scoring, patients 23 and 24 remained in the definite AIH category after treatment despite their incomplete response to therapy. Patient 22 was reclassified to the probable category after treatment. All of these patients underwent ERCP because of their incomplete response to therapy. Patients 22 and 23 had both intrahepatic and extrahepatic biliary changes consistent with PSC, and patient 24 had changes limited to the intrahepatic biliary system. Figure 2 shows the distribution of the IAIHG scores after treatment in the 28 patients. Of the 3 patients who were classified as probable AIH pretreatment, one was reclassified to the definite category posttreatment. The remaining 2 patients stayed in the probable AIH category posttreatment, each with a score of 17 points, respectively. One of these patients was a 14-year-old girl with a normal IgG level, high-titer ANA level, negative ASMA level, and a liver biopsy specimen showing interface hepatitis with bile duct proliferation. She had an incomplete response to therapy with prednisone. An ERCP was normal. She subsequently was diagnosed with systemic lupus erythematosus. The other patient was a 3-year-old boy with increased IgG levels, high-titer ANA and ASMA levels (1:1280 and 1:320, respectively), and a liver biopsy specimen showing both interface hepatitis and bile duct proliferation with onion skinning also seen around some bile ductules. He has not yet had an ERCP, both because of his young age and because of his complete response to prednisone therapy. Two patients who were classified as definite AIH pretreatment were Table 2. Scoring Profiles of the 3 Patients With Overlap Syndrome Parameter Patient 22 Patient 23 Patient 24 Female ALP:AST ratio 0 IgG level greater than normal ANA, ASMA, or LKM-1 Negative viral serology Negative drug history Average alcohol intake 25 g/day Total histologic score 0 Other autoimmune disease (UC) (CD) Total pretreatment aggregate score (Alopecia) 18 Response to treatment IR IR IR AST, aspartate transaminase; UC, ulcerative colitis; CD, Crohn s disease; IR, incomplete response.

4 938 EBBESON AND SCHREIBER CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 10 Table 3. Profiles of the 5 Patients Who Were Reclassified Based on a Score Assigned Using the GGT/AST Ratio Rather Than the ALP/AST Ratio Parameter Patient 10 Patient 24 Patient 15 Patient 21 Patient 23 Female 0 0 ALP:AST ratio 0 GGT:AST ratio IgG level greater than normal ANA, ASMA, or LKM-1 Negative viral serology Negative drug history Average alcohol intake 25 g/day Total histologic score 4 0 (fibrosis) 4 (fibrosis) Other autoimmune disease 0 (Alopecia) (CD) (CD) (CD) Total pretreatment aggregate 16 Definite 18 Definite 16 Definite 19 Definite 17 Definite score using ALP:AST ratio Total pretreatment aggregate 12 Probable 14 Probable 12 Probable 15 Probable 15 Probable score using GGT:AST ratio Response to treatment CR IR IR IR IR ERCP Not performed Positive Inconclusive Not performed Positive NOTE. GGT/AST ratio or ALP/AST ratio relates to the degree of increase greater than the upper limit of normal (unl) of these enzymes (i.e., (GGT/GGTunl)/(AST/ASTunl) or (ALP/ALPunl)/(AST/ASTunl), respectively. CD, Crohn s disease; AST, aspartate transaminase; CR, complete response; IR, incomplete response. downgraded to probable posttreatment, each with a score of 16, respectively. In one (patient 15), a follow-up liver biopsy examination revealed no evidence of interface hepatitis and a new onset of biliary changes. A subsequent ERCP study was inconclusive because there was incomplete filling of the intrahepatic biliary tree. The other patient presented with cirrhosis and eventually required a liver transplant. There was no common bile duct disease noted at laparotomy and no histologic evidence of bile duct disease in the explanted liver. Because the level of alcohol intake rarely is relevant to children, we questioned whether the 2 points given for 25 g alcohol intake would favor the score toward a diagnosis of AIH, thereby negatively impacting on the use of the modified IAIHG scoring system in the pediatric population. When the 2 points for 25 g of alcohol intake were removed from the scoring system and the aggregate scores for definite or probable were recalculated at 13 and 8 13 points, respectively, none of the patients either before or after treatment had shifted their classification category. When we substituted the GGT ratio ([GGT/GGT unl]/[aspartate transaminase/aspartate transaminase unl], in which unl represents upper limit of normal) for the ALP ratio in the modified scoring system for the 25 of 28 patients for which a GGT ratio was available, 5 patients were reclassified from definite to probable AIH pretreatment (Table 3, Figure 3). Four of the 5 patients (80%) went on to have an incomplete response to therapy. Patients 23 and 24 had ERCP-confirmed sclerosing cholangitis and had a diagnosis of overlap syndrome. Importantly, both of these patients had been classified as definite AIH both pre- and posttreatment using the standard ALP ratio in the scoring system. Patient 15, as discussed earlier, was identified as probable posttreatment using the standard ALP ratio. Patient 21, who had an abdominal ultrasound with evidence of extrahepatic biliary tree thickening, is awaiting ERCP. Hence, using the GGT ratio rather than the ALP ratio in the IAIHG scoring system identified 4 patients who went on to have incomplete responses to therapy and showed histologic and radiologic signs of biliary disease. Discussion In the clinical assessment of children presenting with immune hepatobiliary disease, it is often challenging to exclude a diagnosis of PSC without invasive biliary tract imaging. In routine pediatric practice, an ERCP or other biliary tract imaging techniques are difficult to perform and may be associated with complications or Figure 3. Use of the GGT ratio vs. the ALP ratio.

5 October 2004 USEFULNESS OF IAIHG SCORING SYSTEM 939 inconclusive results. 5,6 A scoring system with high specificity for AIH that accurately can identify patients with PSC would be of great clinical use, particularly in young children, in whom it would be most prudent to limit cholangiographic study to those patients in whom the yield would be highest. In the present study, we applied the IAIHG scoring system and found that no pediatric patient with a clinical diagnosis of AIH scored as other, and that none who had a clinical diagnosis of isolated PSC scored as probable or definite AIH. Thus, the IAIHG scoring system was highly effective at separating children with isolated PSC from those with AIH. However, 3 patients with both AIH and PSC scored as definite AIH pretreatment, and 2 of these patients remained in the definite category posttreatment despite an incomplete response to therapy. The IAIHG modified scoring system failed to indicate that these AIH patients also might have biliary tract disease. Few studies have reported on the use of the IAIHG scoring system in pediatric patients with chronic immune hepatobiliary disease. Gregorio et al. 7 applied the scoring system prospectively to 55 consecutive children presenting with clinical and/or biochemical evidence of liver disease and a serologic profile compatible with AIH. All 55 patients underwent liver biopsy examination and cholangiography. AIH was diagnosed in 28 patients. Autoimmune sclerosing cholangitis, defined as cholangiographic-proven sclerosing cholangitis with an increased IgG, ANA, or ASMA level, was diagnosed in 27 patients. In this study, 52% of children with autoimmune sclerosing cholangitis scored as definite AIH and 48% scored as probable AIH, indicating that the scoring system did not discriminate between AIH and autoimmune sclerosing cholangitis. Importantly, however, none of the 28 children who had only a diagnosis AIH scored as other. Moreover, 8 of 9 (88%) children who had a diagnosis of PSC, defined in their study as sclerosing cholangitis without autoantibodies, scored as other. Thus, as in our study, the IAIHG scoring system quite effectively separated those with AIH from those with isolated PSC, but failed to identify those AIH patients who also had biliary tract disease. In the IAIHG scoring system there is no distinction made in point allocation between adult patients with ANA SMA type 1 AIH and those with anti liver kidney microsomal antibody type 2 AIH. Indeed, the clinical use of classifying adult patients into these subtypes remains uncertain. 2 In children, type 2 AIH most often presents at a younger age and may portend a worse prognosis. None of the pediatric patients in our series had type II AIH. That both AIH and sclerosing cholangitis occur frequently in children has been noted by others. El- Shabrawi et al. 8 reported on 13 children with sclerosing cholangitis in whom 40% had a presentation that was indistinguishable from AIH. In these cases of sclerosing cholangitis, there was an increase of IgG level, positive ANA or ASMA level, and an inflammatory portal tract infiltrate or interface hepatitis. In a recently published long-term follow-up study of 52 children with PSC from the Mayo clinic, 14 of 40 (35%) children with PSC and available liver biopsy specimens met the IAIHG criteria for definite (2 of 14) or probable (12 of 14) AIH. 9 The investigators suggested that the expression of PSC in children may be skewed toward a hepatitic presentation instead of the predominantly cholestatic profile that occurs in adults. The increased prevalence of concurrent AIH and PSC in children highlights the need for developing a noninvasive diagnostic tool that would help to identify those AIH children with bile duct disease. Because the long-term survival in children with AIH overlapping with PSC differs from those with AIH alone, the need to differentiate these conditions also has prognostic importance. 7,10 In reviewing the IAIHG criteria, we questioned the applicability of some of the adult diagnostic criteria for the pediatric population. For example, although primary biliary cirrhosis is principally an adult disease, we chose not to exclude the 2 points deducted for a positive antimitochondrial antibody because, importantly, albeit rarely, cases of primary biliary cirrhosis have been described in children. 11 Yet, because for most pediatric patients the amount of alcohol intake is a completely irrelevant issue, we considered whether the IAIHG scoring system credit of 2 points for 25 g of alcohol intake might account for false-positive definite AIH aggregate scores in children evaluated for immune chronic hepatobiliary disease. When the points for alcohol were excluded and the aggregate scores for each category classification were readjusted, no patient was reclassified. Thus, our study found that the IAIHG score in children was not affected by the inclusion or exclusion of alcohol as a criterion. However, our study did show that substitution of the GGT ratio for the ALP ratio improved the ability to identify those children with AIH and concurrent biliary disease. There is a wide range of normal values for ALP in children owing to the contribution of bone growth, and high ALP levels in children frequently reflects bone activity rather than hepatobiliary disease. 9,10 It is well known that GGT values are a more sensitive marker for bile duct injury in pediatric patients. 10 A high serum

6 940 EBBESON AND SCHREIBER CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 10 GGT level usually reflects bile duct disease, although the enzyme level also may increase with certain medications such as anticonvulsants. Of considerable interest, the pediatric study from the Mayo clinic found the ALP value was within the normal range in 25% of their children with PSC whereas the GGT value was increased in 94%, including all those who had normal ALP activity. 10 Thus, we questioned whether in children the GGT ratio rather than the ALP ratio should be applied in the IAIHG classification system. When the scores were recalculated using the GGT ratio, 5 patients pretreatment moved from a score of definite to probable AIH and 4 of them went on to have an incomplete response to therapy. Of these 4 patients, 2 patients have confirmed PSC and the other 2 patients have evidence suggesting bile duct disease. The use of the ALP level in accordance with the IAIHG scoring system failed to identify these 4 patients. The use of the GGT ratio appeared to pinpoint those children with AIH who ultimately also were found to have PSC. In summary, despite several limitations of this study including its retrospective nature, its small sample size, and the fact that not every patient underwent bile duct imaging studies, our findings do support the use of the modified IAIHG scoring system for the diagnosis of AIH in children. Our results suggest that those patients who fall into the other category before treatment should have a biliary imaging study. The use of the GGT ratio instead of the ALP ratio may improve the specificity of the scoring system for pediatric patients, thus identifying those unlikely to respond to treatment and likely to have both AIH and PSC. When the GGT ratio was used, patients classified as probable pretreatment also should be considered for cholangiographic investigation. Further prospective multicentered studies with large sample size would allow for determination of the sensitivity and specificity of the GGT ratio in the IAIHG classification system for children and may provide more optimal guidelines for the practicing gastroenterologist in the assessment and management of these young patients. References 1. Johnson PJ, McFarlane IG. Meeting report: International Autoimmune Hepatitis Group. Hepatology 1993;18: International Autoimmune Hepatitis Group Report. Review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999;31: Kaya M, Angulo P, Lindor KD. Overlap of autoimmune hepatitis and primary sclerosing cholangitis: an evaluation of a modified scoring system. J Hepatol 2000;33: Omagari K, Masuda J, Kato Y, Nakata K, Kanematsu T, Kusumoto Y, Mori I, Furukawa R, Tanioka H, Tajima H, Koga M, Yano M, Kohno S. Re-analysis of clinical features of 89 patients with autoimmune hepatitis using the revised scoring system proposed by the International Autoimmune Hepatitis Group. Int Med 2000; 39: Fox VL, Werlin SL, Heyman, MB. Endoscopic retrograde cholangiopancreatography in children. Subcommittee on Endoscopy and Procedures of the Patient Care Committee of the North American Society for Pediatric Gastroenterology and Nutrition. J Pediatr Gastroenterol Nutr 2000;30: Ferrara C, Valeri G, Salvolini L, Giovagnoni A. Magnetic resonance cholangiopancreatography in primary sclerosing cholangitis in children. Pediatr Radiol 2002;32: Gregorio GV, Portman B, Karani J, Harrison P, Donaldson PT, Vergani D, Mieli-Vergani G. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology 2001;33: El-Shabrawi M, Wilkinson ML, Portmann B, Mieli-Vergani G, Chong SKF, Williams R, Mowat AP. Primary sclerosing cholangitis in childhood. Gastroenterology 1987;92: Feldstein AE, Perrault J, El-Youssif M, Lindor KD, Freese DK, Angulo P. Primary sclerosing cholangitis in children: a long-term follow-up study. Hepatology 2003;38: Batres LA, Maller ES. Laboratory assessment of liver function and injury in children. In: Suchy FJ, Sokol RJ, Balistreri WF, eds. Liver disease in children. Philadelphia, PA: Lippincott Williams & Wilkins, 2001: Dahlan Y, Smith L, Simmonds D, Jewell LD, Wanless I, Heathcote EJ, Bain VG. Pediatric onset primary biliary cirrhosis. Gastroenterology 2003;125: Address requests for reprints to: Richard A. Schreiber. Division of Gastroenterology, Rm K4-200, BC Children s Hospital, 4480 Oak Street, Vancouver BC Canada V6H 3V4. rschreiber@cw.bc.ca; fax: (604)