Article Text

Download PDFPDF

Gastroenterology, Hepatology, and Nutrition

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


M. Samyn, D. Bogdanos, Y. Ma, B. Portmann, D. Vergani, G. Mieli-Vergani.Institute of Liver Studies, King’s College Hospital, London, UK

Background: An overlap syndrome between autoimmune liver disease (AILD) and systemic lupus erythematosus (SLE) has been occasionally described in adults, but never in children.

Patients: Between 1973 and 2004, eight (7 female) children who presented to our centre with AILD were also diagnosed with SLE. Median age at presentation was 8.75 (range 2.7–12.4) years. Duration of symptoms ranged between 0.5 and 12 months (median 1 month), the most common being jaundice (7), fever (5), lethargy (5), vomiting (4), joint (3), and abdominal pain (3). All children had high IgG (20.8–68.9 g/l, median 40.1 g/l). Five children had antinuclear antibody (ANA) and smooth muscle antibody (SMA), one ANA only and one SMA only, while one was ANA, SMA, and liver kidney microsomal1 (LKM1) antibody positive. Liver histology showed severe inflammation in all four children biopsied at presentation, and moderate (2) and mild (2) inflammation in children biopsied after starting treatment. Fibrotic and cirrhotic changes were seen in six and one, respectively. Treatment with prednisolone was started at a median time of 1 day after presentation (range 1–4). Azathioprine was added in four patients. Mycophenolate mofetil replaced azathioprine in two and cyclosporin was used in one. SLE was diagnosed between 1 week and 6.9 years (median 7 month) after AILD. Eight children had skin, seven joint, four renal, four haematological, and three pleural/pericardial involvement. ANA became positive in all, anti-double stranded DNA and extractable nuclear antigens in four and two, respectively. C4 was low in five. Soluble liver antigen (SLA) antibodies were positive in 5/7, ribosomal P antibodies in 4/5 tested. SLE was treated with high dose pulse steroids (5), cyclophosphamide (4), hydroxychloroquine (3), methotrexate (2), and plasmapheresis (1). Median follow up is 3.7 (range 0.2–8.7) years. Six children have normal liver function tests while two continue to have mildly raised AST. Two patients required further pulse steroids because of SLE relapse.

Conclusion: AILD/SLE overlap syndrome does exist in children with AILD. SLE can be present at diagnosis of AILD or develop during follow up and must be timely and aggressively treated.


J. L. Hartley1, R. M. Brown1, A. Tybulewicz3, P. Hayes3, D. Wilson2, P. Gillett2, P. McKiernan1.1Birmingham Children’s Hospital, Birmingham, UK; 2Royal Hospital for Sick Children, Edinburgh, UK; 3The University of Edinburgh, Edinburgh, UK

Objectives: Hyaluronic acid is excreted by the liver via sinusoidal cell adhesion molecules. Movement of hyaluronic acid across sinusoidal cells is impeded in fibrosis leading to a rise in serum hyaluronic acid. As a marker of fibrosis, hyaluronic acid may obviate the need for liver biopsy, a procedure that has a significant morbidity risk and may be inaccurate due to sampling error. In paediatric practise hyaluronic acid has not previously been evaluated as part of a histologically controlled study in unselected patients.

Methods: Ninty three unselected consecutive children (median age 7.5 years, range 0.8 months to 19 years) undergoing a liver biopsy between April 2003 and March 2004 were prospectively recruited. Liver biopsy and fasting hyaluronic acid levels were taken simultaneously. The Ishak score was used to stage fibrosis and scores of three or greater were regarded as significant fibrosis. Hyaluronic acid levels were measured using an enzyme linked binding protein assay (© 2002 Corgenix, Inc)(adult reference range 0–75 ng/ml). Liver function tests and anthropometric data were recorded.

Results: 23/93 (25%) biopsies had significant fibrosis with 5 staged as cirrhotic (stage 6). The hyaluronic acid levels in this group was significantly higher than the children with mild fibrosis (less than 3), (median level 72 ng/ml v 30 ng/ml, Mann Whitney U test p<0.005). Using a stepwise logistic regression analysis of laboratory and anthropometric measurements hyaluronic acid was the only variable predictive of significant fibrosis (p = 0.007). Hyaluronic acid level >50 ng/ml had a sensitivity of 65% and specificity of 68% for significant fibrosis.

Summary: Despite 75% of this study cohort having mild fibrosis, hyaluronic acid was a significant non-invasive predictor of histological fibrosis. In this unselected study group the clinical use of hyaluronic acid is limited by sensitivity and specificity.

Conclusion: In clinical practise hyaluronic acid can not yet replace histiological examination to evaluate hepatic fibrosis. Further evaluation is needed to ascertain the utility of serial hyaluronic acid levels in targeted patient groups.


P. Subramaniam, A. Knisely, G. Mieli-Vergani, A. Baker.Institute of Liver Studies, King’s College Hospital, London, UK

Background: Alagille’s syndrome is diagnosed in the presence of at least three major syndromic features. To study the clinical and histological features at presentation, we performed a retrospective analysis of the records of 121 children with Alagille’s syndrome.

Results: Cholestasis was seen in 118/121 (97.5%), pale stools in 102/121 (84.2%), characteristic facies in 94/121 (77.6%), posterior embryotoxon in 73/121 (60.3%), butterfly vertebrae in 45/121 (37.1%), heart disease (most often peripheral pulmonary stenosis), in 108/121 (89.2%), renal disease in 28/121 (23.1%), birth weight 2.5 kg or less in 55/108 (50.9%). Cholesterol of >3.5 mmol/l was seen in 36/48 (75%) children before 8 weeks of age (mean (SD) 5.37 (2.29) mmol/l), and in 51/63 (80.9%) below 16 weeks of age, (mean 5.40 (2.13) mmol/l). Between 16 weeks and 1 year serum cholesterol of >5 mmol/l was seen in 7/10 (70%) median 6.9 mmol/l and range 2.3–40 mmol/l, above 1 year of age in 12/13 (92%) median 6.3 mmol/l and range 4.6–17 mmol/l. Liver biopsy showed characteristic features of paucity of interlobular bile ducts (PILBD) in 59/77 (76.6%) children below 16 weeks of age, in 11/15 (73.3%) between 16 weeks and 1 year of age and in 8/12 (66.66%) above 1 year of age. Biopsy findings other than PILBD were seen in 18/77 (23.3%) children below 16 weeks of age of whom 12/77 (15.5%) showed features of non-specific hepatitis and 6/77 (7.79%) had biliary features. Biliary features on biopsy were also seen in 4/14 (28.5%) between 16 weeks and 1 year of age, and in 3/12 (25%) above 1 year of age. HIDA scans showed no excretion of isotope into the bowel after 24 hours in 21/35 (60%) and small/contracted/no gall bladder on ultrasound was seen in 30/105 (28.5%). 12/121 (9.91%) had a diagnostic laparotomy and operative cholangiography, two proceeding to Kasai porto-enterostomy prior to referral to our unit and 7/121 (5.78%) had PTC/ERCP because of diagnostic difficulty.

Conclusion: Clinical features of Alagille’s syndrome are not as consistently informative as suggested in the literature. Heart disease/murmur and facial features are the features most likely to be noted at presentation. Hypercholesterolaemia is an early feature and while non-specific may be a helpful pointer. Histology is not characteristic in 25%, HIDA scan and ultrasound may suggest a false diagnosis of biliary atresia in 60% and 28% respectively, supporting the concept that infants with liver disease warrant early specialist referral.


K. Bargiota, M. Samyn, A. Baker, S. Bansal, A. Dhawan, D. Hadzic, G. Mieli-Vergani.Institute of Liver Studies, King’s College Hospital, London, UK

Background: Standard treatment for autoimmune liver disease (AILD) in children consists of prednisolone +/− azathioprine. However, intolerance or failure to respond to this treatment has lead to the need for alternative therapies, including cyclosporin, tacrolimus, and mycophenolate mofetil (MMF). This is the first report on the use of MMF in the treatment of children with AILD.

Objective: To evaluate the efficacy and safety of MMF in children with AILD.

Patients: Between 1999 and 2004, 25 children (15 female) with AILD (autoimmune hepatitis (18), autoimmune hepatitis/sclerosing cholangitis overlap syndrome (7)) were treated with MMF. Twenty two were type I (ANA/SMA+), one was type II (LKM1+), one was both ANA/SMA and LKM1 positive, and one was LC 1 positive. All patients were treated with prednisolone (2 mg/kg/day, maximum: 60 mg/day) at the time of diagnosis. MMF was introduced at a median time of 1.4 (range 0–9) years after diagnosis at a dose of 10 mg/kg/day aiming for a maximum dose of 40 mg/kg/day (maximum 2 g/day). Median follow up was 2.2 (range 0.2–4.6) years after MMF had been commenced. The indications for the use of MMF were: intolerance to azathioprine (2), failure to respond to azathioprine (9), recurrent relapses of AILD (10), and intolerance to cyclosporin (1). In three children MMF was used with prednisolone as initial treatment because of severe jaundice (2) and diabetes mellitus (1).

Results: Twenty children had normal transaminases after a median time of 3.4 (0.3–15.3) months of treatment with MMF and 20 achieved a maintenance dose ⩽5 mg/day of prednisolone. Twelve patients experienced side effects: leucopenia (7), hair loss (1), blood in stools (1), dizziness and nausea (1), diarrhoea (1), and abdominal pain (1). The dose of MMF was decreased in eight and discontinued in four. In two, MMF was restarted with no recurrence of the side effects. MMF was stopped in four because of persistent diarrhoea in two children with ulcerative colitis, persistent leucopenia (1) and abdominal pain (1).

Conclusion: MMF is effective and well tolerated in children with AILD who are intolerant or do not respond to standard treatment and can be considered when the use of azathioprine is contraindicated.


R. K. Russell1, H. E. Drummond1, L. Smith1, E. Nimmo1, N. Anderson1, D. C. Wilson2, P. M. Gillett2, P. McGrogan3, K. Hassan3, L. T. Weaver4, M. Bisset5, G. Mahdi5, J. Satsangi1.1Gastrointestinal Unit, Edinburgh University, Edinburgh, UK; 2Royal Hospital For Sick Children, Edinburgh, UK; 3Yorkhill Hospital, Glasgow, UK; 4Glasgow University, Glasgow, UK; 5Royal Aberdeen Children’s Hospital, Aberdeen, UK

Aims: The chronic inflammatory bowel diseases (IBD), Crohn’s disease, and ulcerative colitis are common causes of morbidity in children. The incidence and prevalence of early onset IBD, and Crohn’s disease in particular, continues to rise. Three single nucleotide polymorphisms R702W, G908R, and Leu1007finsC in the NOD2/CARD15 gene have been strongly implicated in susceptibility to adult onset Crohn’s disease. We investigated these NOD2/CARD15 variants in early onset IBD.

Methods: Patients <16 years at diagnosis were recruited with parents included where possible, with blood collected for DNA analysis. IBD was diagnosed if patients matched standard diagnostic criteria. R702W, G908R, and Leu1007finsC mutations were assessed by PCR/TaqMan analysis. Transmission disequilibrium testing (TDT) was used to assess linkage and association within family trios. Detailed genotype–phenotype analysis including multivariate analysis was undertaken. The Vienna classification was used to categorise CD location and behaviour.

Results: 247 IBD patients (167 Crohn’s disease, 60 ulcerative colitis, and 20 indeterminate colitis) and 414 parents were recruited. Allele frequencies in Crohn’s disease were 6.4%, 2.8%, and 4.5% for R702W, G908R, and Leu1007finsC, respectively. These combined allele frequencies were significantly higher than controls (p 0.007). However, the Crohn’s disease allele frequencies are lower than in the index populations studied, 87 (p 0.0001).1 Compared with ulcerative colitis patients, Leu1007finsC (4.5% cf. 0%) and overall NOD2/CARD15 carriage (19.7% v 6.8%) were significantly higher in Crohn’s disease patients (p 0.02 for both). TDT demonstrated preferential transmission of Leu1007finsC (p 0.006).19.7% of Crohn’s disease patients carried at least 1 NOD2/CARD15 mutation. The population attributable risk for the 3 NOD2/CARD15 mutations was only 7.7%. Univariate analysis in Crohn’s disease patients demonstrated NOD2 carriers were more likely to need surgery 8.3% v 2.7% (p 0.0001) and were less likely to form granulomas 7.7% v 2.6% (p 0.0005). At diagnosis NOD2/CARD15 carriers had less inflammatory disease behaviour 3.6% v 7.5% (p 0.03) and more stricturing disease (p 0.03). At 2 years follow the association with stricturing disease was stronger (p 0.01). Multivariate analysis demonstrated NOD2/CARD15 carrier status was an independent risk factor for surgery (p 0.001, OR 4.5(1.8–11.3)).

Conclusions: The three NOD2/CARD 15 variants in the early onset Crohn’s disease population studied have a definite, but relatively small contribution to Crohn’s disease susceptibility, and behaviour. Despite a low NOD2/CARD15 carrier frequency compared with most other populations studied carriers are at a significantly higher risk of aggressive disease as shown by increased need for surgery. Priority should now be given to identifying novel genetic determinants within this high risk, high incidence population.



D. C. Wilson13, A. T. Tybulewicz1, L. Pieterse1, P. Rogers1, D. Hoole2, P. M. Gillett1.1Department of Paediatric Gastroenterology and Nutrition, RHSC, Edinburgh, UK; 2Pharmacy Department, RHSC, Edinburgh, UK; 3Child Life and Health, University of Edinburgh, Edinburgh, UK

Introduction and Aim: Use of methotrexate for induction of remission in children with chronic, non-remitting Crohn’s disease is poorly studied and there is minimal data for its use in maintenance of remission. We aimed to review our population of children who entered remission on methotrexate treatment.

Methods: Retrospective and prospective cohort study of patients with Crohn’s disease, resistant to or intolerant of azathioprine, who were treated with methotrexate (16 week induction course of s/c methotrexate (15 mg/m2), followed by the equivalent oral dose) from 1/8/97 to 1/11/04 at our regional centre. Paediatric Crohn’s Disease Activity Index (PCDAI) was used as a well validated score to objectively measure disease activity during treatment; PCDAI ⩽10 indicated full remission and ⩽15 indicated partial remission.

Results: 24 patients were treated with an induction course of s/c methotrexate; two were excluded where our protocol was not followed. Of the 22 patients treated, 21 (95%) entered remission (clinically and PCDAI ⩽15). Methotrexate was started at mean (range) age of 13.8 (10.6–18.5) years. Median (range) time to partial/complete remission was 15 (4–42) weeks, with median duration of follow up being 60 (29–208) weeks. Nine (43%) entered clinical remission with PCDAI ⩽15, and remained in long term remission (four full remission, five partial remission). 12 patients relapsed, with a mean (range) time to relapse of 26 weeks (4–64 weeks). Nine (43%) had a single relapse, of whom seven (32%) then re-entered remission, while one required infliximab and one surgery. Three (14%) had ⩾2 relapses; all received infliximab, two later required surgery, and one has subsequently entered remission again. On treatment, 10 (48%) patients had raised ALT, which subsequently resolved, and 7 (33%) had mild nausea/vomiting.

Conclusions: Methotrexate appears to be effective in maintaining remission in many who have failed to remit on azathioprine, and appears to be well tolerated and safe. However, an appropriate evidence base is needed from well designed RCTs.


L. Stewart1, N. Nwafor2, P. Jackson2, D. C. Wilson3.1Department of Dietetics, RHSC, Edinburgh, UK; 2Community Child Health, RHSC, Edinburgh, UK; 3Child Life and Health, University of Edinburgh, Edinburgh, UK

Background: The epidemic of childhood obesity is well documented in the UK. However, little attention has been paid to the prevalence of obesity within the specific population of children and teenagers with special needs.

Aim: To review the occurrence of obesity (body mass index (BMI) >95th centile) in a population of primary and secondary children attending special needs schools.

Methods: The school nursing records of 139 ambulatory children (age 10.6 (SD 3.3) years, 44 females) attending three special needs schools (one primary, one secondary, and one combined school) were audited for the most up to date weight and height measurement. Records were reviewed for the period 2000–2004. BMI was calculated, and BMI centile and SDS (UK 1990 reference data) noted.

Results: In the 71 secondary school children (mean (SD) age 13.3 (1.4) years, 25 females), 35% (25/71) were obese and 23% (16/71) were grossly obese (BM I>99.6th centile). In the 68 primary school children (mean (SD) age 7.7 (1.8) years, 19 females), 16% (11/68) were obese and 7% (5/68) were grossly obese. These results compare with figures in the general Scottish paediatric population of 9% obese (>95th centile) in Primary year 1 and 15% in Secondary year 3.1

Summary and Conclusions: A higher percentage of ambulant children with special needs in this audit were obese than in the general paediatric population. The prevalence of obesity more than doubles between the primary and secondary school cohorts. In secondary school, not only are more obese than the general population, but almost 25% are grossly obese (BMI >99.6th centile). These figures show the highest incidence of childhood obesity in any population reported in the UK. We recommend that further research is urgently undertaken in this specific population, particularly in terms of prevalence, causes, prevention, and treatment of obesity.



D. Hawcutt, D. Casson, S. Williams, J. C. Blair.Royal Liverpool Childrens Hospital Alder Hey, Liverpool, UK

Background: The burden of childhood obesity on paediatric services is likely to increase. To assess current workload and project future demand, we recorded body mass index (BMI) standard deviation (SD) score of all children attending outpatient clinics during a 6 week period. The prevalence and age distribution of obesity (BMI >2SD) was calculated.

Results: 2277 children (51.7% M, 48.3% F, mean 8 year, range 0.1–19.5 years) were recorded. 13.8% patients were obese of whom 29.7% had a BMI >3SD. Prevalence of obesity increased with age (<2 years 5.4%, 2–5 years 10.2%, 5–10 years 14.4%, >10 years 18.6%) but was equal in males and females (13.8% M, 13.6% F). 7.6% of obese patients, and 19.4% of patients with BMI>3SD were referred specifically for obesity. However, an additional 16.9% of children with BMI>2SD had obesity related health problems. We estimate that obesity referrals and related health problems account for at least 667 outpatient episodes a year in this centre.

Discussion: National figures report a prevalence of obesity in 6–10 year olds of 11.2%, 10–15 year olds of 13.8%.1 Our data report a higher prevalence among paediatric outpatients. This may reflect a greater level of obesity in the local or national population. It is also possible that these children have greater health problems, potentially caused by their obesity. Achieving compliance with S.I.G.N. guidelines2 (referral of all children with BMI >3SD for medical assessment of their obesity) would result in increased workload.

Conclusion: The prevalence of obesity is high in the paediatric outpatient setting and exceeds that reported in the normal population. At present, this generates a significant workload in outpatients, the burden on inpatient services is unknown. If accepted guidelines2 are implemented, and the prevalence of obesity continues to rise, additional service provision will be required to meet this clinical demand.




C. Mehta1, G. Gupte2, K. Sharif2, S. Protheroe3, S. Beath2, J. DeGiovanni1.1Cardiology Department, Birmingham Children’s Hospital, Birmingham, UK; 2Liver Unit, Birmingham Children’s Hospital, Birmingham, UK; 3Gastroenterology Department, Birmingham Children’s Hospital, Birmingham, UK

Background: Children with irreversible intestinal failure require parenteral nutrition administered through central venous lines for long term survival. Multiple line insertions and infections can lead to thromboses in the major vessels of the neck and impaired venous access. We report on a new technique for central line insertion in children with occluded vessels in the neck, which is a useful alternative to transhepatic line and right atrial line placement requiring a sternotomy.

Aim: To describe our experience with the use of the stereotactic technique in children with impaired venous access.

Methods: A retrospective review of the intestinal failure database was conducted to identify children having stereotactic long needle puncture of the SVC stump using biplane fluoroscopy. Once the vessel was punctured, using the Seldinger technique, a sheath was placed in the vessel and a subcutaneous tunnel created for the Hickman line, which was placed at the RA/SVC junction. Review of the medical records was done to document the demographic details and complications.

Results: Five Hickman lines inserted stereotactically were identified in four patients. The median age at insertion was 7 years. Three children out of four are alive 6–24 months after stereotactic line insertion. There were two identified complications: death within 24 hours—no cause found in particular no mediastinal injury related to line placement (1); pericardium breached with guide wire-no sequelae (1). One patient had repeat line insertion 3 months apart due to inadvertent removal by the patient.

Conclusions: This is a new and relatively non-invasive technique, which allows safe insertion of central venous catheters even in clinically frail children with severely compromised venous access who are awaiting small bowel transplantation.


S. M. Hill, C. Coakes, P. Carmichael.Great Ormond Street Hospital for Children, London, UK

Background: It is now possible for children with chronic severe intestinal failure to survive throughout childhood on treatment with parenteral nutrition. In some cases intestinal function is so poor that little or no enteral nutrition is tolerated. However, it is not known whether such children achieve their potential for intellectual development. Early reports suggested that parenteral nutrition might be associated with central nervous system injury,1 while subsequent studies have found that most children attain adequate growth and development within the first few years of life.2 Little is known about the longer term impact of parenteral nutrition on cognitive development.

Aim: The aim of this study was to assess the cognitive development of children who had received parenteral nutrition for more than 5 years.

Methods: Sixteen children, 6 male and 10 female aged 6.8–16.4 years who were dependant on parenteral nutrition and had been on treatment from 6–16 years (mean 10 years), were reviewed. Seven children were of British origin and for 9 English was their second language. All children had undergone annual cognitive assessments with age appropriate tests. These were: The Griffiths Mental Development Scale (birth to 3 years); the Wechsler Preschool and Primary Scale of Intelligence (WPPSI; 3–6 years), and the Wechsler Intelligence Scale for Children (WISC-III; 6–16 years).

Results: The mean full scale IQ score was 87.5 (SD 18.8). All the scores fell within the average to low average range. There was no significant change in IQ with age and no significant difference between verbal and non-verbal IQ.

Conclusion: The cognitive ability of children on progressively longer periods of parenteral nutrition does not significantly change over time and remains within the average to low average range. Children can develop normally on long term treatment with parenteral nutrition.




M. G. Gnanalingham, A. Mostyn, G. S. Gopalakrishnan, J. Dandrea, M. E. Symonds, T. Stephenson.Centre for Reproduction and Early Life, University of Nottingham, Nottingham, UK

Background: Uncoupling protein 2, an inner mitochondrial protein, has postulated roles in solute exchange and macrophage mediated immunity within the neonatal lung. The effect of maternal nutritional deprivation in mid (maximal placental growth) and late (maximal lung growth) gestation on uncoupling protein 2, glucocorticoid receptor, and 11B-hydroxysteroid dehydrogenase type1 (11BHSD1, activates cortisol) mRNA abundance in the ovine neonatal and adolescent lung has not been determined.

Methods: Fourteen twin and 12 singleton bearing ewes were entered into the study. The singleton bearing ewes were either nutrient restricted (NR), receiving 60% of their energy requirements from 28 to 80 days gestation, or fed 100% requirements throughout pregnancy (Controls-C). All mothers gave birth spontaneously at term and the offspring were tissue sampled at 6 months. The offspring of the twin bearing ewes were sampled at 6 hours or 30 days after birth. Their mothers also received either 60% (NR) or 100% (C) of their energy requirements from 110 days gestation up to term. Total lung RNA was extracted and uncoupling protein 2, glucocorticoid receptor, and 11BHSD1 mRNA abundance measured by RT-PCR. Results are given as means (SD) relative to 18S rRNA (see table). Lung weights and total protein concentration were similar between groups and increased with postnatal age. Uncoupling protein 2, glucocorticoid receptor, and 11BHSD1 mRNA peaked at 6 hours and then declined with postnatal age. NR increased uncoupling protein 2, glucocorticoid receptor, and 11BHSD1 mRNA at all ages in the lung.

Abstract G26 Demographic details of children undergoing Hickman line insertion

Abstract G28

Abstract G30 Treatment response and outcome for each sub-group

Conclusion: Irrespective of the timing of maternal nutritional deprivation, uncoupling protein 2, glucocorticoid receptor, and 11BHSD1 mRNA are increased, indicating increased cortisol sensitivity and possible underlying susceptibility to infection.


J. S. Gordon1, P. Reid2, C. Thompson2.1Napier University, Edinburgh, UK; 2Royal Hospital for Sick Children, Edinburgh, UK

Aims: To determine the effectiveness of the regional Childhood Idiopathic Constipation Management Pathway “Tough Going” by investigating through an audit how it is being used and whether it has facilitated change in practice. This audit was as far as possible a repeat of the audit completed in 1999 prior to the launch of “Tough Going”.

Methods: “Tough Going” was introduced to a region of 650 000 total population in January 2001. A retrospective audit had been planned to survey implementation, so data were collected between February 2002 and January 2003. The sample comprised 173 health professionals who were issued copies of “Tough Going” and 18 who had purchased copies; 35 parents of children with constipation referred to specialist nursing and child and family mental health services; 66 children with constipation attending our regional paediatric Accident and Emergency (A&E) department between June and August 2003. These data were compared with audit data collected between February1998- January 1999. The data collection tools were developed, piloted, and implemented by the group. We used (a) a postal questionnaire for health professionals; (b) a telephone questionnaire for parents/carers; (c) an audit pro forma of A&E attendees sourced from case notes and hospital patient information system (4D); (d) an audit pro forma of specialist nurse and child and family mental health services referrals from individual case load data; and (e) a survey of children referred to paediatric surgery.

Results: The health professional response rate was 53% (92). 69 (75%) respondents use “Tough Going” with the sections most used being assessment, pharmacy guidelines, and care packages. 65 (71%) stated that this document had facilitated specifically identified changes in practice in their management of childhood constipation. Monthly A&E attendance by children with constipation has fallen by 29%, from 17 to 12 per month, and acute admissions by 36%, from 9 to 2 per month. This suggests that health professionals in the community are following the pathway effectively and referring patients to A&E who may require differential diagnosis. The data from families showed that they delayed seeking help and highlighted the need for a coordinated, supportive, multi-disciplinary approach.

Conclusions: The audit results suggest that “Tough Going” is meeting the expressed needs of healthcare professionals who use the document by providing a comprehensive resource. The successful implementation provides a less traumatic experience for children and families undergoing treatment. To sustain the momentum, it is necessary to both support users and to encourage non-users by regularly updating and promoting “Tough Going”.


P. F. Chai1, R. M. Brown2, J. L. McPortland2, K. Foster3, P. Davies1, D. A. Kelly1.1Liver Unit, 2Histopathology Department, 3Radiology Department, 4Statistical Advisory Dept, Birmingham Children’s Hospital, Birmingham, UK

Background: Autoimmune liver disease (AILD) includes sclerosing cholangitis, autoimmune hepatitis types 1 and 2 (AIHT1 & AIHT2), and overlap syndrome. The aim of the study is to review the indications and outcome of liver transplant in children with AILD.

Methods: A retrospective review of all children with AILD presenting from 1981 to 2004.

Results: 102 patients had confirmed AILD (62 F). 66 had AIHT1, 18 AIHT2, 9 sclerosing cholangitis, and 9 overlap syndrome. Median age at onset of disease was 11.9 years (range1.2–18.3). 98/102 were treated with prednisolone and azathioprine. Two had liver transplant for acute liver failure and two had mild disease, hence were not treated.

Median age at transplant was 14.3 years (range 1.7 to 21.8). One child with AIHT1 and three with AIHT2 patients had emergency liver transplant. Median duration of treatment to liver transplant for the remaining 15 children was 29 months (range 8–92). Indications for liver transplant include acute liver failure (22%), progression of liver disease despite treatment (39%), poor quality of life (22%), complications of portal hypertension (22%), and hepatopulmonary syndrome (5%). Main postoperative complications included biliary complications (4), upper GI bleed (1), and steroid induced psychosis (1). 11 patients (57%) had 19 episodes of acute rejection and two (11%) developed chronic rejection. Five patients (26%) had recurrence of disease (1 AIHT1, 3 AIHT2, 1 ASC). Median time from liver transplant to recurrence of disease was 27 months (range 5–51). Four had repeat liver transplant (2 for chronic rejection, 2 for recurrence of disease) and one had a third liver transplant (hepatic artery thrombosis). There were four deaths due to sepsis (1), surgery for ulcerative colitis (1), and recurrence of disease (2).

Conclusion: Most children with AILD respond to medical therapy and only a minority require liver transplant. Counselling for liver transplant should indicate the morbidity and mortality related to graft rejection and disease recurrence.


J. A. J. Raiman2, N. Hadzic1, M. P. Champion2, A. Dhawan1, A. J. Baker1, C. S. Ball1, G. Mieli-Vergani1.1King’s College Hospital, London, UK; 2Guy’s Hospital, London, UK

Background: Mitochondrial cytopathies may present in a variety clinical forms including neurological, hepatic, gastrointestinal, cardiac, and renal. Distinction between neurological signs and encephalopathy related to liver failure in an unwell child may be difficult in the absence of a specific diagnostic test.

Methods: A retrospective review of case notes was performed at tertiary liver referral centre. Clinical and biochemical data were collected from patients, fulfilling our diagnostic criteria, either (a) muscle enzyme activity <20% (of age adjusted mean), (b) evidence of a Mt DNA mutation, (c) evidence of a recognised mitochondrial syndrome, or (d) a combination of suspicious clinical and biochemical findings, family history, and abnormal enzyme activity between 20–40%.

Results: 24 cases (17 male) referred between 1991 and 2004. Median age of presentation was 0.33 years (range birth to 5 years). Half the cases were consanguineous. Presenting features included acute liver failure (ALF) (10), liver dysfunction (12), and hepatosplenomegaly (2). Extrahepatic involvement included neurological (18), haematological (3), renal (5), pancreatic (2), and failure to thrive (8). Median blood lactate 3.45 (range 0.9–17), CSF lactate 1.5 (1–6.45). Brain MRI was abnormal in 11/15 with either basal ganglia, or brainstem changes or atrophy. Diagnosis was confirmed on muscle respiratory chain enzymology in 20 (deficiency of complexes IV (9), I (5), I/IV (2) and II/III (4)) and mitochondrial DNA analysis in 1. A diagnosis of Alpers (2), Leigh’s (1), and Pearson syndrome (1) was made in 4 cases. Fourteen children died (median 1.79 years, range 0–5.41 years). Two children underwent transplantation due to deepening liver failure prior to the establishment of a diagnosis. Median follow up 3.88 years (range 0.5 to 13.8 years). One boy died from progressive neurological disease and respiratory failure 10 months after transplant. The other is now well with unremarkable liver graft function and normal neurodevelopment 3 years post transplant. Of the 10 survivors, two have progressive cirrhosis, and in eight the liver function stabilised or improved with medical treatment.

Conclusions: Mitochondrial disease should be sought in children presenting with cryptogenic liver disease, including acute liver failure, particularly with extrahepatic involvement, elevated blood, and CSF lactate. Diagnosis can be confirmed on muscle biopsy. Overall prognosis is poor and transplantation should be confined to those without neurological involvement.


C. Kortsalioudaki, S. Bansal, R. M. Taylor, G. Mieli-Vergani, A. Dhawan.Institute of Liver Studies, King’s College Hospital, London, UK

Background: Acute liver failure (ALF) carries a high mortality in children. N-acetylcysteine (NAC), an antioxidant agent that replenishes mitochondrial and cytosolic glutathione stores, has been used successfully in the treatment of acetaminophen induced ALF. In our unit NAC was introduced as the standard treatment of ALF since 1995.

Aim: To evaluate the safety and efficacy of N-acetylcysteine in children with acute liver failure not caused by acetaminophen poisoning.

Method: A retrospective review of medical records of 123 children presenting with non-acetaminophen induced ALF between 1989 and 2000 was made. ALF was defined as either INR>2 and abnormal liver function or INR>1.5 with encephalopathy and abnormal liver function. Children were divided into: group 1 (1989–1994): standard care (n = 49, 24 (49%) male, median age 2.4 (range 0.003–15.8 years)) and group 2 (1995–2000): NAC administration (n = 74, 41 (55%) males, 3.38 (range 0.005–17.4 years)). Diagnoses were idiopathic in 17 (35%) in group 1 v 30 (41%) in group 2, infectious causes in 6 (12%) v 9 (12%), metabolic causes in 6 (12%) v 7 (10%) neonatal haemochromatosis in 6 (12%) v 6 (8%), autoimmune liver disease in 5 (10%) v 4 (5%), Wilson’s disease in 5 (10%) v 3 (4%), miscellaneous in 4 (8%) v 15 (20%), nsd. NAC was administered as a continuous infusion (150 mg/kg/24 hour) until normal INR or transplant.

Results: Presenting features were: jaundice in 46 (94%) in group 1 v 58 (78.%) in group 2, p = 0.02, encephalopathy in 29 (59%) v 46 (62.%), p = ns, hepatomegaly in 36 (74%) v 56 (76%), p = ns, splenomegaly in 23 (47%) v 27 (37%), p = ns, ascites in 11 (22%) v 8 (11%), p = ns. Median duration of NAC administration in group 2 was 5 (range 1–77) days. Complications were noted in 8 (10.8%), rash in 3, bradycardia in 3, and dizziness and peripheral oedema in 1. One child had an allergic reaction (bronchospasm) and NAC was stopped. The length of hospital stay was 36 (6–264) days in group 1 v 25 (1–211) days in group 2, p = ns and length of PICU stay was 2 (1–23) days v 3 (1–68) days, p = ns. In group 1, 11 (22%) developed renal failure v 18 (25%) in group 2, p = ns, and 14 (29%) v 32 (44%) developed heart failure, p = ns. Survival with own liver occurred in 10 (20%) in group 1 v 23 (31%) in group 2, p = ns; death without transplant, 9 (18%) v 17 (23%), p = ns; and liver transplant was performed in 30 (61%) v 34 (46%), p = ns. Death after transplantation occurred in 14 (47%) in group 1 v 8 (24%) in group 2, p = 0.05.

Conclusion: NAC can be given safely to children with ALF with a low incidence of side effects. The NAC treated patients had better post transplant survival.