Gut 2001;49:431–435 431 Outcome of liver disease in children with Alagille syndrome: a study of 163 patients Gut: first published as 10.1136/gut.49.3.431 on 1 September 2001. Downloaded from P Lykavieris, M Hadchouel, C Chardot, O Bernard Abstract to have a relatively good long term prognosis in Background and aims—Various opinions terms of liver disease45; however, it is now well have been expressed as to the long term recognised that some patients with AGS can prognosis of liver disease associated with present with severe complications of liver Alagille syndrome (AGS). disease.6–12 We therefore reviewed the charts of Patients and methods—We reviewed the 174 patients with AGS presenting in childhood outcome of 163 children with AGS and to evaluate the role of the liver condition in liver involvement, investigated from 1960 mortality, morbidity, and long term outcome. to 2000, the end point of the study (median age 10 years (range 2 months to 44 years)) being death, liver transplantation, or the Patients and methods last visit. One hundred and seventy four children with Results—At the study end point, of the 132 AGS (106 boys) were investigated at Bicêtre patients who presented with neonatal Hospital between 1960 and 2000. Twenty four cholestatic jaundice, 102 remained jaun- had a sibling aVected by AGS; seven of these diced, 112 had poorly controlled pruritus, siblings are included in this series as well as two and 40 had xanthomas; cirrhosis was oVspring of aVected mothers. All patients had found in 35/76 livers, varices in 25/71 at least three of the five major clinical features. patients, and liver transplantation had Thirteen children with severe neonatal jaun- been carried out in 44 patients (33%). dice underwent Kasai operation for suspected Forty eight patients died, 17 related to biliary atresia before the diagnosis of AGS was complications of liver disease. Of 31 established. Cholecystostomies with external patients who did not present with neonatal bile drainage and cholecystojejunal anastomo- cholestatic jaundice, five were jaundiced ses were performed in 24 and six patients, at the study end point, 17 had well control- respectively. End stage liver disease was led pruritus, and none had xanthomas; defined by the combination of the following cirrhosis was found in 6/18 patients, laboratory tests: serum bilirubin concentration varices in 4/11, and none underwent liver >300 µmol/l, serum albumin concentration http://gut.bmj.com/ transplantation. Nine patients died, two of <35 g/l, and prolonged prothrombin time in liver disease. In the whole series, actuarial spite of treatment with parenteral vitamin K. survival rates with native liver were 51% Investigations of portal hypertension included and 38% at 10 and 20 years, respectively, upper digestive endoscopy, ultrasonography, and angiography. Therapy included various and overall survival rates were 68% and antipruritus drugs over time and supplementa- 62%, respectively. Neonatal cholestatic tion as necessary with parenteral or oral fat jaundice was associated with poorer sur- on September 27, 2021 by guest. Protected copyright. Service d’Hépatologie soluble vitamins. Hypercaloric continuous vival with native liver (p=0.0004). Pédiatrique, Hôpital nocturnal enteral feeding was performed in 39 Conclusions—The prognosis of liver dis- de Bicêtre, 94275 Le children with severe growth and height retarda- ease in AGS is worse in children who Kremlin Bicêtre tion. Cedex, France present with neonatal cholestatic jaun- P Lykavieris dice. However, severe liver complications M Hadchouel are possible even after late onset of liver STATISTICAL ANALYSIS O Bernard 2 disease, demanding follow up throughout Categorial data were compared using the ÷ test with Yates’ correction if indicated. Con- Unité INSERM 347, life. 94276 Le Kremlin (Gut 2001;49:431–435) tinuous data were compared using the t test. Bicêtre Cedex, France Survival rates were calculated according to the M Hadchouel Keywords: Alagille syndrome; cholestasis; end stage Kaplan-Meier method, the results being ex- liver disease; liver transplantation pressed as value (SEM) and compared using Service de Chirurgie the log rank (Mantel-Cox) test. Multivariate Pédiatrique, Hôpital analysis was performed using Cox’s propor- de Bicêtre, 94275 Le Syndromic paucity of interlobular bile ducts tional hazard model, and relative risks (RR) Kremlin Bicêtre (Alagille syndrome, AGS) is an autosomal Cedex, France were calculated with 95% confidence intervals C Chardot dominant disorder defined clinically by the (CI). All significance tests were two tailed and association of at least three of five major diVerences were considered to be statistically Correspondence to: features (chronic cholestasis, congenital heart significant at p<0.05. All analyses were per- P Lykavieris, Service disease, “butterfly-like” vertebrae, posterior d’Hépatologie Pédiatrique, formed with the Statview 4.51.1 software Hôpital de Bicêtre, 78 rue du embryotoxon, and peculiar facies) and geneti- 1–3 (Abacus concepts Inc., Berkeley, California, Général Leclerc, F-94275 Le cally by mutations in the Jagged 1 gene. USA). Kremlin Bicêtre Cedex, Expression of AGS varies from a mild pheno- France type to severe diseases of the heart or kidney Accepted for publication and to the consequences of chronic cholestasis, Abbreviations used in this paper: AGS, Alagille 12 February 2001 including liver failure. AGS has long been said syndrome; LT, liver transplantation; RR, relative risk. www.gutjnl.com 432 Lykavieris, Hadchouel, Chardot, et al Table 1 Clinical signs and outcome at the study end point in 163 patients with Alagille (range 4–23 years). In all other patients pruri- syndrome and chronic liver disease. The study end point was defined as either death, liver tus persisted throughout life with varying transplantation, or last follow up Gut: first published as 10.1136/gut.49.3.431 on 1 September 2001. Downloaded from degrees depending on the type of drug used Patients with no and on compliance with therapy. Xanthomas Patients with neonatal clinically patent were noted at a median age of one year nine cholestatic jaundice neonatal jaundice No patients 132 31 months (range 9 months to 10 years) in 69 Median age (range) 9 y (2 mos–29 y) 14 y (3–44) p Value patients. They persisted until liver transplanta- tion (LT), death, or last follow up in 40 No positive/No patients No positive/No patients patients. In the remaining 29 children, xantho- studied studied mas disappeared at a median age of seven years Jaundice 102/132 5/31 <0.0001 (range 3–18 years). Hepatomegaly remained Pruritus 112/132* 17/28‡ 0.0003 Xanthomas 40/132 0/31 0.0004 present in 38 patients who died and in all Hepatomegaly 115/132 11/31 <0.0001 patients in whom LT was performed. In Splenomegaly 62/132 6/29 0.0095 patients surviving without LT, hepatomegaly Cirrhosis 35/76† 6/18¶ NS Oesophageal varices 25/71§ 4/11§ NS decreased progressively with age and was no GI bleeding 5/132 1/29 NS longer detectable in 17 of 21 patients after age Cholelithiasis 6/132 1/29 NS 15 years. Splenomegaly was found during Hepacellular carcinoma 0/132 1/31 NS Liver transplantation 44/132 0/31 0.0002 follow up in 84 patients; it progressively Death related to liver disease 17/132 2/31 NS regressed in 21 patients and one patient under- Dead 48/132 9/31 NS went splenectomy. Results of biochemical, his- Median (range)/No Median (range)/No patients studied patients studied tological, and endoscopic tests carried out at the study end point are indicated in table 1. Six ASAT (IU/l) 162 (23–650)/106 12 (25–186)/23 0.0004 ãGT (IU/) 408 (13–2288)/92 332 (28–787)/18 NS patients presented with cholelithiasis at a Serum cholesterol (mmol/l) 9.9 (1.3–50.5)/126 7.4 (4.8–8)/26 NS median age of four years (range 3–8 years). Gastrointestinal bleeding from oesophageal *Variably controlled under medical treatment. ‡Well controlled under medical treatment. varices occurred in five patients (including one †32 patients who underwent liver biopsy after one year of age and 44 patients whose liver was with associated portal vein obstruction) at a removed at LT. median age of three years nine months (range ¶18 patients who underwent liver biopsy after one year of age. §Number of patients who underwent endoscopy during follow up. 13 months to 23 years) and was the cause of death in three patients. Twenty three patients had one or more bone fractures and three Results patients developed large joint arthropathy after Of the 174 patients studied, 81 displayed all five age nine years. A partial improvement in jaun- major clinical features of the syndrome, 68 dice, pruritus, and liver tests was observed in displayed four major features, and 25 three four of 30 patients who underwent cholecysto- major features. In 132 children, 38 being small stomy and/or cholecystojejunostomy after fol- for gestational age (with a weight below the 10th low up periods ranging from three weeks to http://gut.bmj.com/ percentile for gestational age), the presenting four months; no significant improvement was symptom was neonatal cholestatic jaundice. In observed in the remaining 26 children. 42 children, two being small for gestational age Forty four patients underwent liver trans- (significantly less than in the other group, plantation in various centres at a median age of p<0.01), there was no history of clinically six years nine months (range 2 years 9 months evident neonatal jaundice and presenting symp- to 28 years), including 31 children in our toms occurred at a median age of 33 months group. Nine of the 44 transplanted patients had (range 4 months to 10 years) consisting of previously undergone a Kasai procedure. At on September 27, 2021 by guest. Protected copyright.