Article Text

240. FAS-LIGAND PROTECTS ISLET CELLS FROM APOPTOSIS IN CHRONIC PANCREATITIS
  1. R.A. Al-Mufti,
  2. R.C.N. Williamson,
  3. R.T. Mathie
  1. Division of Surgery, Anaesthetics & Intensive Care, Hammersmith Hospital, Imperial College School of Medicine, London W12 0NN, UK
  1. S.H. Rahman,
  2. B.J. Ammori,
  3. G. Barclay2-1,
  4. I.G. Martin,
  5. M. Larvin,
  6. M.J. McMahon
  1. Academic Unit of Surgery, The General Infirmary, Great George Street, Leeds LS1 3EX, UK;2-1National Transfusion Science Laboratories, Edinburgh, UK
  1. T. Wong,
  2. H. Kocher,
  3. P. Kane,
  4. J. Karani,
  5. A. Patel,
  6. J. Devlin
  1. Institute of Liver Studies, Depts of Surgery, and Radiology, Kings College Hospital, Bessemer Road, London SE5 9RS, UK
  1. J. Meenan,
  2. G. Rottenberg
  1. Endosonography Centre, Guy's and St Thomas' Hospital, London, UK
  1. S.J. Forbes5-1,
  2. K.M. Hodivala-Dilke,
  3. R. Jeffery,
  4. T. Hunt,
  5. M.R. Alison5-2,
  6. R. Poulsom,
  7. N.A. Wright
  1. Histopathology Unit, Imperial Cancer Research Fund, London; 5-1Dept of Medicine at St Mary's Hospital; 5-2Dept of Histopathology at Hammersmith Hospital, Imperial College School of Science Technology and Medicine, London, UK
  1. J. Devlin,
  2. J. Wendon,
  3. G. Auzinger,
  4. B. Portmann,
  5. N. Heaton,
  6. M. Rela
  1. Institute of Liver Studies, Kings College Hospital, London SE5 9RS, UK
  1. N.C. Fisher8-1,
  2. C. Matthews8-2,
  3. G. Criddle8-3,
  4. J. Gwinnett8-4,
  5. J. Hanson8-3,
  6. A. Phillips8-4,
  7. J. Rao8-5,
  8. E.T. Swarbrick8-2
  1. Depts of Gastroenterology of8-1Dudley Hospitals and 8-2Wolverhampton Hospitals, and Public Health Depts of 8-3Dudley, 8-4Wolverhampton, and 8-5Sandwell, UK
  1. R. Malik,
  2. N. Mellor,
  3. C. Selden,
  4. H. Hodgson
  1. Centre for Hepatology, Royal Free Hospital, Rowland Hill St, London NW3, UK
  1. S.J. Gerred,
  2. P.T.F. Kennedy,
  3. N. Heaton,
  4. M. Rela,
  5. P. Muiesan,
  6. A. Bomford,
  7. B. Portmann,
  8. S. Norris,
  9. J.G. O'Grady
  1. Institute of Liver Studies, King's College Hospital, Denmark Hill, London, UK
  1. A.S. Austin,
  2. Y.R. Mahida,
  3. S.D. Ryder,
  4. J.G. Freeman12-1
  1. Division of Gastroenterology, University Hospital, Nottingham; 12-1Dept of Medicine, Derby City General Hospital, Derby, UK
  1. A.J. Hughes,
  2. J. Dyer,
  3. E. Elias,
  4. S. Olliff
  1. D. Tripathi14-1,
  2. H.F. Lui14-1,
  3. R. Jalan14-1,
  4. E. Forrest14-1,
  5. A.J. Stanley14-1,
  6. A. Helmy14-1,
  7. D.N. Redhead
  1. Dept of Radiology, Royal Infirmary of Edinburgh, Edinburgh; 14-1Liver Unit, Dept of Medicine, Royal Infirmary of Edinburgh, Edinburgh, UK

Statistics from Altmetric.com

Background: Exocrine (acinar) destruction and endocrine (islet) preservation in chronic pancreatitis (CP) is related to differential apoptotic indices within these cellular compartments.

Aim: To define the role of cell cycle regulatory proteins in the control of pancreatic epithelial cell apoptosis.

Methods: Formalin-fixed, paraffin-embedded tissue from six cases of CP and eight normal controls (N) were studied using immunohistochemistry for BCl-2, Bax, retinoblastoma protein (Rb) and Fas ligand (Fas-L). Labelling patterns were assessed using semi-quantitative (intensity-proportion [I-P]) or fully quantitative (labelling index [LI]) analysis.

Results: Rb protein was expressed at a higher level by acinar cells in CP than N (mean acinar cell Rb LI 1.45% in CP and 0.437% in N; p<0.05) but this difference was not observed within islet cells. BCl-2 was more strongly expressed by acinar than islet cells in CP and N (e.g. N mean BCl-2 I-P score 4.88 in acini and 0.375 in islets; p<0.05) while Bax was strongly expressed by a subset of islet cells and weakly by centroacinar cells (e.g. N mean Bax I-P score 1.75 in acini and 4.31 in islets; p<0.05). Fas-L was more strongly expressed by islet than acinar cells in CP and N (e.g. CP mean Fas-L I-P score 5.67 in islets and 2.83 in acini; p<0.01).

Conclusion: Strong BCl-2 expression by acinar cells and of Bax by islet cells was an unexpected result and indicates complex control of apoptosis within these cell populations. Increased Rb protein expression by acinar cells in CP may differentially promote apoptosis within these cells. Fas-L expression by islet cells may protect islets from apoptosis by promoting apoptosis of cytotoxic T-lymphocytes.

241. THE EFFECT OF SELECTIVE INHIBITION OF INDUCIBLE NITRIC OXIDE SYNTHASE ON THE OUTCOME OF SEVERE ACUTE PANCREATITIS

Background: Nitric oxide (NO), overproduced by inducible NO synthase (iNOS), has been established as a key inflammatory mediator in systemic inflammatory response syndrome leading to multiple organ failure.

Aim: To examine the effect of treatment with the selective iNOS inhibitor L-N6-iminoethyl-lysine (L-NIL) on the course of severe acute pancreatitis (AP) in the rat, both local pancreatic and systemic organ failure.

Methods: Acute severe necrotising pancreatitis was induced in 10 adult male Wistar rats by a combination of biliopancreatic duct infusion of glycodeoxycholic acid and arterial infusion of caerulein. Data were compared to those in sham operated controls (n=10). The effect of L-NIL treatment was examined in a disease group (n=10) and in a control group (n=10). L-NIL was administered two hours after induction of pancreatitis. Blood pressure, PaO2, and serum creatinine were measured at 7 h. Pancreatic sections were examined histologically and morphometric measurement of areas of acinar cell necrosis as a percentage of total acinar tissue area was performed.

Results: Compared with controls at 7 h, the pancreatitis group displayed reduced mean arterial blood pressure [mean (s.e.m.) 77.8 (3.7) versus 108.2 (2.2) mmHg, P<0.0001], reduced PaO2 [66.7 (1.9) versus 115.2 (3.6) mmHg, P<0.0001], and raised serum creatinine [132.5 (20.6) versus 41.8 (2.5) μmol/L, P<0.0001]. Treatment with L-NIL corrected the reduction in blood pressure [mean (s.e.m.) 113.5 (3.4) mmHg, P<0.0001], and ameliorated the reduction of PaO2 levels [91.43 (2.2) mmHg, P<0.0001] and the rise in serum creatinine [78.9 (4) μmol/L, P=0.028]. There was reduction in morphometric measurement of areas of acinar necrosis as a percentage of total acinar tissue in histological sections in the treated group compared to pancreatitis [29.61 (0.79) versus 39.66 (1.47) %, P<0.0001].

Conclusion: In this experimental model of severe acute pancreatitis, treatment with L-NIL two hours after induction of pancreatitis corrected circulatory failure, ameliorated renal and respiratory functions, and limited the progression of pancreatic necrosis. L-NIL may prove to have a useful role in the supportive treatment of a severe attack.

242. INCREASED SYSTEMIC NITRIC OXIDE IN PATIENTS WITH SEVERE ACUTE PANCREATITIS IS ASSOCIATED WITH ALTERED GUT MACROMOLECULAR PERMEABILITY: AN ENDOTOXIN MEDIATED INFLAMMATORY RESPONSE?

Background: There is considerable evidence implicating nitric oxide (NO) as a critical mediator of the systemic inflammatory response to bacterial endotoxaemia. Severe acute pancreatitis (AP) is associated with alterations in intestinal permeability, and bacterial translocation may in part contribute to the local and systemic manifestations of this disease. However, the mechanisms remain speculative and the role of NO in humans with AP has not been studied.

Methods: Patients with a clinical and biochemical diagnosis of AP were studied within 72 hours of onset of abdominal pain. The 24-hr urinary nitrite excretion, reflecting NO production, was measured using the Greiss reaction. The ratio of renal excretion of the enterally administered polyethylene glycol (PEG) 3350/400 was measured to determine intestinal macromolecular permeability. The IgM:IgG EndoCAb ratio was used as a marker of systemic endotoxin exposure. Attacks were classified as mild or severe according to Atlanta criteria.

Results: Sixty-five patients with AP (severe 20) and 20 healthy control subjects were studied. Urinary nitrite excretion was significantly increased in patients with severe attacks (median 20.6) compared to mild attacks (median 15.7, p = 0.003), and the latter was significantly greater in healthy controls (median 6.3, p = 0.004). PEG excretion ratios were significantly increased in patients with severe attacks compared to mild attacks (r = 0.8, p < 0.01). In patients with severe disease, urine nitrite excretion demonstrated a positive and significant correlation with intestinal macromolecular permeability and the IgM:IgG EndoCAb ratio (r = 0.7, p = 0.006, and r = 0.8, p = 0.001 respectively).

Conclusion: Systemic NO is increased in AP, and is greatest amongst patients who develop a severe attack. The strong correlation with altered gut permeability and endotoxin exposure suggests that the mechanism may be mediated through bacterial up-regulation of inducible nitric oxide synthase activity.

243. EVALUATION OF ENDOSCOPIC ULTRASOUND VERSUS HELICAL COMPUTED TOMOGRAPHY IN THE ASSESSMENT OF PANCREATIC MASSES: RESULTS FROM A UK CENTRE

Introduction: Endoscopic ultrasound (EUS) is an emerging diagnostic modality which has been shown to be more accurate in the diagnosis, and locoregional staging of pancreatic carcinoma than conventional computed tomography (CT). The aim of this study was to determine the local staging accuracy of EUS, compared with helical computed tomography (HCT) in a United Kingdom centre.

Patients, materials and methods: Between September 1999 and October 2000, 43 consecutive patients underwent radial EUS (GF-UM20) and HCT for assessment of pancreatic masses. Of these, 17 patients underwent laparoscopy and proceeded to surgery. Surgery included radical lymph node clearance in all cases, and portal vein resection where necessary. Tumour size, portal venous invasion, and nodal involvement were assessed and compared with histopathological examination of the resected specimens as the reference standard. The χ2 test for paired data was used to compare groups.

Results: Of the 17 patients, who underwent surgery 15 had a pancreaticoduodenectomy (Whipples) procedure, one a distal pancreatectomy, and one a laparotomy without resection. The portal vein was infiltrated in three patients, and resected in two. 10 patients had adenocarcinoma, 2 neuroendocrine tumour, 4 focal pancreatitis, and one an anaplastic tumour on histological examination. EUS was more accurate than HCT in the accuracy of lymph node staging (82% vs 47%, p=0.03), and was as accurate as HCT with respect to size (76% vs 47% within 0.5cm, p=0.07), and portal venous involvement (94% vs 82%, p=NS). All patients with portal vein invasion were detected by EUS.

Conclusion: EUS is accurate in the local, and nodal staging of pancreatic masses, and is sensitive in the assessment of portal vein invasion.

244. DOES ENDOSCOPIC ULTRASOUND (EUS) IMAGING, IN THE ABSENCE OF BIOPSY, HAVE A ROLE IN STAGING CANCERS OF THE PANCREAS AND AMPULLA?

Background: The clinical value of Endoscopic Ultrasound (EUS) in staging tumours of the pancreas and ampulla, like all other imaging modalities individually, is uncertain. Although EUS does provide a method for sampling tissue without the potential for coelomic spread, such biopsy equipment is not widely available.

Aim: To determine the clinical value of radial diagnostic imaging in tumours of the pancreas and ampulla.

Methods: All patients (n=36; m:20, f:16; Age range 42yrs to 78yrs, median 66yrs) with suspected tumours underwent EUS, dual-phase contrast helical CT and where appropriate ERCP. EUS was performed under conscious sedation (midazolam/fentanyl) using Olympus GFUM20 and GFUM200 radial echoendoscopes. Follow-up data of at least six months was available on all patients and surgical data on 15 cases.

Results: The patient cohort included pancreatic head adenocarcinomas (n=20; resection, n=8); ampullary carcinoma (n=10; resection, n=4) and pancreatic neuro-endocrine tumours (n=4; resection, n=4; insulinoma, n=3 and non-functioning, n=1). Duodenal stricturing limited EUS in two cases of pancreatic head carcinoma. In three-cases of adenocarcinoma, EUS identified an operable tumour not identified on CT, in one case EUS missed vascular invasion of the portal vein, although review of films identified this to be present. EUS identified vascular (SMA, SMV, PV) involvement seen on CT. All cases of ampullary carcinoma undergoing surgery were correctly staged by EUS. EUS identified 3 of 4 neuro-endocrine tumours, the fourth being identified by MRI. EUS identified enlarged nodes more readily than CT, but specificity and sensitivity were low (55% and 12% respectively).

Conclusion: Radial EUS provides additive information to that yielded by helical CT in the assessment of pancreatic/ampullary tumours. EUS is particularly helpful in identifying small adenocarcinomas undetectable by CT. Like other imaging modalities, EUS cannot differentiate between benign and inflammatory regional lymph nodes.

245. HEPATOCYTES DERIVED FROM BONE MARROW STEM CELLS SHOW POLYPLOIDIZATION IN HUMANS AND MICE

Background and aims: Using a technique to detect Y-chromosomes within human liver we have previously demonstrated in patients that have received sex mis-matched bone marrow or liver transplants that human hepatocytes can be derived from bone marrow stem cells. These cells may provide an alternative source of hepatocytes following liver damage. Using the technique of Y-chromosome detection we have examined these hepatocytes for polyploidization. This is of relevance as polyploidization is an integral feature of hepatocyte replication.

Method: Beta3-integrin-deficient mice were used as they develop centrilobular liver damage that results in a persistent low-grade liver regeneration. Six-week old female recipient mice underwent whole body gamma irradiation with 1000 rads to ablate their bone marrow and then received male wild type bone marrow transfer by tail vein injection. Animals were killed eight-weeks following bone marrow transplant and their livers were analysed using a murine Y chromosome probe for hepatocytes of male origin. In addition the livers of female patients who had received male bone marrow transplants and males who had received female liver transplants were also examined for evidence of Y-chromosome positive polyploid hepatocytes.

Results: We found evidence of Y-chromosome positive hepatocytes of both diploid and polyploid class in the livers of female mice that had received bone marrow transplants, female patients who had received male bone marrow transplants and male patients who had received female liver transplants.

Conclusion: The diverse ploidy status identified within hepatocytes derived from bone marrow stem cells suggests that they can undergo polyploidization and are capable of replicating within the liver. This indicates that these hepatocytes of bone marrow origin are able to contribute to hepatic regeneration following liver damage.

246. STEM CELL FACTOR LIMITS LIVER INJURY INDUCED BY PARACETAMOL POISONING

Background and aims: Paracetamol (acetaminophen) remains a common cause of acute liver failure and emergency liver transplantation in both the UK and USA. The cytokine networks involved in the pathogenesis of paracetamol induced acute liver injury are poorly understood. Stem cell factor (SCF) is a cytokine involved in cell growth and repair which was studied in a murine model of paracetamol poisoning.

Methods: CBA mice were injected IP with paracetamol solution following an 8 hour fast. SCF was quantified by ELISA, neutralised with anti-SCF antibodies and reconstituted with recombinant SCF (rSCF).

Results: Paracetamol injection lead to significant reduction in hepatic SCF concentration measured by ELISA (control 2100+300ng/gm liver, mean+sem, paracetamol 970+170, n=6, p<0.05) at 24 hours. Injection of 200mg/kg paracetamol was associated with 80% survival at 96 hours, but inhibition of SCF in this “sublethal” model was associated with survival of only 40% at 96 hours (p<0.05, n=20 in each group): this was associated with significantly increased areas of liver necrosis at 96 hours in the survivors (control 11.6+5.4 % total area of liver, mean+sem, anti-SCF treated 25.7+8.5, n=5, p<0.05). Conversely administration of rSCF (1ug) improved survival of mice injected with 300mg/kg paracetamol improved survival from 30% at 96 hours to 90% (p<0.05, n=20 in each group): this was associated with improved hepatic histology at days 1, 2, 4 and 6 in the rSCF treated group. SCF treatment was correlated with significant reduction in cytochrome P4502E1 expression in both murine liver and culture murine cell lines.

Conclusions: Hepatic SCF plays an important role in modulating paracetamol induced liver injury in a murine model. A potential mechanism for this effect is the inhibition of paracetamol activation by cytochrome P4502E1. Other mechanisms such as an effect of SCF on hepatocyte proliferation are currently under investigation.

247. FATAL ASSOCIATION BETWEEN HAEMOPHAGOCYTIC LYMPHOHISTIOCYTOSIS AND LIVER DYSFUNCTION

An association between Haemophagocytic Lymphohistiocytosis (HLH) and liver disease is poorly defined. In this study, we review our experience of this condition when presenting in association with liver dysfunction including that following liver transplantation.

Thirty six patients with hepatic dysfunction and histopathologically confirmed haemophagocytosis who fulfilled the Henter criteria were reviewed. In all cases a bone marrow aspirate/trephine was examined.

This syndrome complicated three clinical groups: 1) immunodeficient patients including liver transplant recipients, SLE etc; 2) haematological malignancy; and 3) acute liver failure (ALF) of unknown origin (NANB hepatitis). In the liver recipients, HLH complicated both early graft dysfunction and late cytomegalovirus infection and was associated with multi-organ failure and death in all cases. HLH was evident in 80% of adult transplant deaths in the last year (8 of 10). Fever was evident in the majority of the ALF cases with profound thrombocytopaenia invariable at presentation in all cases. Less than 10% of cases were diagnosed prior to referral. In groups 2 and 3, jaundice (mean 154 umoll), fever, splenomegaly and either pleural effusions or alveolar infiltrates were common at presentation. The AST activity (372 (iu/l) was modest. Haemophagocytosis was also present in post-mortem liver biopsies when available.

Management protocols attempted in these patients included immunosuppression, chemotherapy, high-volume hamofiltration, GCSF, intravenous immunoglobulin and anti-viral therapy. The response to these treatments was very poor with only four patients surviving. Disease progression was characterised by severe pancytopaenia, acute lung injury and circulatory failure.

The appalling outcome of HLH in liver disease merits urgent clinical and research attention. This report draws attention to the very unfavourable prognosis of this poorly recognised association and the difficulties encountered in management

248. LIVER DISEASE MORTALITY IN THE BLACK COUNTRY 1993–1999

Background: Treatment of advanced liver failure is a major burden on healthcare resources. To determine incidence of liver disease mortality and any underlying trends we analysed data in three boroughs of the West Midlands “Black Country” (Wolverhampton, Dudley and Sandwell, total population 845,000) from 1993–1999.

Methods: Public health mortality files were analysed for liver-related deaths using ICD reference codes and keyword searches. Case notes were analysed in cases of liver disease of unspecified cause. In-patient episode data for 1995–1999 were obtained from health authority records.

Results: There was a stepwise increase in liver-related mortality from 6.6 per 105 population in 1993 to 13.9 per 105 in 1999. This increase was exclusively due to alcoholic liver disease (ALD); incidence 3.1 per 105 in 1993 rising threefold to 9.3 per 105 in 1999, whilst mortality due to other defined liver diseases was stable at 0.5 per 105. In Wolverhampton and Sandwell (which have large Asian communities) ALD mortality rates in Asian and white populations were similar but Asian subjects died at an earlier age (median age at death 46 years vs. 55 yrs in whites, p<0.001). These data probably underestimate true ALD mortality since unspecified liver disease (around 20% of all liver mortality) was due to ALD in 62% of cases as judged by case note analysis; also ALD deaths increased by 8% when accounting for ALD “misclassified” as other diseases. Finally, in-patient liver disease episodes increased stepwise from 41 per 105 population in 1995 to 57 per 105 in 1999 which correlated significantly with mortality rates (r=0.74, p<0.01).

Conclusions: Liver disease mortality due to alcohol has increased dramatically in recent years, and if sustained this trend has important implications for gastroenterologists and public health specialists.

249. CORRELATION BETWEEN CEREBRAL PROTON MAGNETIC RESONANCE SPECTROSCOPY (1H MRS) AND NEUROPSYCHOMETRIC ABNORMALITIES IN PATIENTS WITH CHRONIC HEPATITIS C (HCV) INFECTION

Patients with chronic HCV score lower on quality of life scales compared to normal individuals and patients with chronic hepatitis B (HBV). We examine the hypothesis that a direct cerebral effect of HCV underlies this. Computerised and paper-based psychometric tests batteries were administered to 26 patients with histologically mild HCV hepatitis and 10 HCV antibody+ve, PCR-ve age, sex and intelligence-matched controls. 12/26 patients vs 0/10 controls (p=0.01) and 8/24 patients vs 0/10 controls (p=0.07) were impaired on computer and on paper-based batteries respectively. There were no differences according to a history of intravenous drug abuse (IVDA). In vivo 1H MR spectrawere acquired from voxels in the basal ganglia (BG), white matter (WM) and occipital grey matter (GM) with a 1.5T spectroscopy system in 3 patient groups: 1) 30 patients with biopsy-proven mild HCV infection (mean age 44 yr, 47% M, mean liver necroinflammatory score 2.4/18, mean fibrosis score 1.6/6). 67% had a history of IVDA; 2) 12 hepatitis B eAg+ve patients without cirrhosis, none IVDA+ve; 3) 29 healthy controls (mean age 42 yr, 52% M), none IVDA+ve.

Results: A significant elevation in BG and WM choline/creatine (Cho/cr) was seen in the HCV group compared to the other groups (*p<0.005).

Abstract 249, Table 1

No difference was seen in the HCV group according to IVDA+ve status. Fourteen HCV patients also had psychometric testing, showing correlations between BG Cho/Cr and indices of sustained attention (r=0.699, p=0.005) and quality of working memory (r=0.544, p=0.04).

Conclusion: Both cognitive impairment and cerebral metabolite abnormalities are seen in patients with mild HCV hepatitis. This may be due to a cerebral effect of systemic cytokines or direct infection of the CNS by HCV (as in HIV infection where similar 1H MRS abnormalities are seen).

250. CHARACTERISING THE EFFECTS OF THYROID HORMONE ON THE LIVER: A NOVEL APPROACH TO INCREASING LIVER MASS

Background: Liver growth can occur as part of two distinct mechanisms: 1) Compensatory Regeneration: Following resection, viral/drug injury; 2) Direct Hyperplasia: A primary mitogen increasing liver mass directly.

Aim: a) To assess if thyroid hormone acts as a primary mitogen. b) To characterise the effects of thyroid hormone when administered prior to a 70% partial hepatectomy (PH)—testing a primary mitogen with an established model of regeneration.

Method: a) Male sprague-dawley rats (n=7 per group) were injected with a single dose of tri-iodothyronine (T3) and sacrificed at intervals of 1,2,4,7,10 and 14 days. A control group received vehicle only twenty four hours prior to sacrifice. b) A 70% partial hepatectomy was performed on sprague-dawley rats (n=5 per group) on days 1,3 and 10 following a single dose of T3. A control group received vehicle only twenty-four hours prior to PH. All animals were sacrificed 24 hrs after partial hepatectomy.Cell Proliferation – Assessed by bromodeoxyuridine (BRDU) incorporation into nuclei and immunohistochemical recognition.

Results: a) Liver mass was increased in animals treated with T3 as compared with controls. Maximum effect was seen on day 10 with a 20% increase in liver mass (p<0.05). A corresponding increase in total DNA (p<0.05) and liver protein (p<0.05) was seen at this time point.Proliferation – Peaked at day 1: 7% hepatocytes labelled compared to <1% in controls (p<0.01). b) In animals treated with T3 three and ten days prior to partial hepatectomy the liver weight at sacrifice was greater (p<0.05) than controls. There was a corresponding increase in liver protein (p<0.05) and total DNA (p<0.05) at these time points. Proliferation – Peaked when T3 was administered 24 hrs prior to PH 36% hepatocytes labelled compared to 26% in controls (p<0.01).

Conclusions: Thyroid hormone acts as a mitogen and synergistically with a 70 % partial hepatectomy. This raises the possibility of preoperative T3 administration to perform larger liver resections.

251. HEPATIC CO-FACTORS IN END-STAGE, ALCOHOL RELATED, LIVER DISEASE

Alcohol consumption is well recognised as an independent cause of cirrhosis and as an accentuating factor in the pathogenesis of advanced liver disease in hepatitis C carriers. However, a history of alcohol consumption may contribute to erroneous diagnoses or the failure to recognise significant co-factors. We undertook a retrospective analysis of these variables in 159 patients referred for liver transplantation with a primary diagnosis of alcoholic liver disease between 1996 and 2000.

Of the 159 patients, 31 (20%) had recognised chronic viral hepatitis (24 anti-HCV positive, 5 HBsAg positive, 2 anti-HCV & HBsAg positive). These 31 patients were younger than the remaining 128 patients without viral hepatitis (mean 49 vs 52 years, P = 0.03) and younger than a separate cohort of 79 non-alcohol abusing hepatitis C carriers with end-stage liver disease (mean 49 vs 53 years, P = 0.01). In 12 out of the 128 non-viral alcohol abusers the primary diagnosis was not alcoholic liver disease: 2 Autoimmune hepatitis, 2 Caroli's disease, 2 Nodular Regenerative Hyperplasia, 2 PBC, 2 PSC, 1 Sarcoidosis, 1 Haemochromatosis (C282Y homozygous). In a further 16 patients co-factors were identified: 14 hepatic siderosis (grade 3/4), 2 alpha-1 antitrypsin deficiency (MZ phenotype). Of the siderotic patients only 2 had HFE gene mutations detected, 1 had aceruloplasminaemia. Hepatocellular cancer was identified in 23 patients, 7 were discovered incidentally in the explanted liver.

Mean serum IgA was significantly higher in the alcohol cohort than the separate hepatitis C cohort (5 vs 7 g/l, P = 0,0002), but no other discriminatory laboratory variables were identified.

This study reveals a 10% erroneous diagnosis rate in patients referred with a diagnosis of alcoholic liver disease. It confirms the accelerated progression to cirrhosis in hepatitis C carriers abusing alcohol. Significant co-factors were present in a further 13%; this went unrecognised in most cases before transplantation. Despite advances in screening for hepatocellular carcinoma, 30% of the tumours were undiagnosed before transplantation. These data suggest a need for a more circumspective assessment of alcohol consuming patients with end-stage liver disease.

252. THALIDOMIDE BUT NOT PENTOXIFYLLINE LOWERS PORTAL PRESSURE IN HUMAN ALCOHOLIC CIRRHOSIS

Introduction: Blockade of tumour necrosis factor alpha (TNFα) activity reduces portal pressure in portal vein-ligated rats. We investigated the ability of two inhibitors of TNFα production, thalidomide and pentoxifylline, to reduce portal pressure in humans.

Methods: Abstinent patients with stable alcoholic cirrhosis and oesophageal varices were treated for two weeks with open-label pentoxifylline 1800mg daily (n=9) or thalidomide 200mg daily (n=10). Portal and systemic haemodynamics were measured invasively. The hepatic venous pressure gradient (HVPG-mmHg) was calculated by subtracting free from wedged hepatic venous pressures.

Results: Thalidomide (T) consistently reduced HVPG (Table 1) and increased hepatic blood flow 1164 (738–2260) ml/min to 1505 (1054–2943) ml/min. Hepatic vascular resistance fell from 1498 (460–2546) to 559 (269–1039) dynes/sec/cm5 (p=0.028). Pentoxifylline (P) had no effect. Side effects led to dose reduction/ withdrawal in 4 (T) and 5 (P). Two patients (T) withdrew without reason. Data for patients completing two weeks treatment are expressed as median (range) and analysed using Wilcoxon signed ranks test (*p=0.028).

Abstract 252, Table 1

There was no significant change in systemic haemodynamics, bilirubin, albumin, INR or creatinine in either group.

Conclusion: Anti-TNFα therapy with thalidomide effectively reduces portal pressure and hepatic vascular resistance supporting a role for TNFα in the pathogenesis of human portal hypertension.

253. TIPS and BUDD–CHIARI SYNDROME

Background/aims: Patients with Budd–Chiari who are not amenable to hepatic vein recanalisation are usually treated either medically, with surgical shunt formation or liver transplantation. We reviewed the outcome of 17 patients who were treated with TIPS.

Methods: A retrospective case note analysis was undertaken of all patients with Budd–Chiari Syndrome who had a TIPS attempted.

Results: TIPS was attempted in 17 patients with a median age of 43 years (range 20–59) and median length of history 9 weeks (range 2–364 weeks). An underlying aetiology was found in 16 patients. 15 patients had a technically successful TIPS performed, neither of the 2 patients developed adverse sequelae following failed TIPS attempt. Of the 15 patients, one died soon after TIPS and one died of a subdural haematoma 6 weeks later. The remaining 13 have been followed up for a median of 31 months and have had excellent symptom resolution and have required an average of 0.92 re-interventions per patient.

Conclusion: In appropriately selected patients with Budd–Chiari, TIPS can provide a definitive treatment at least in the medium term.

254. RANDOMISED CONTROLLED TRIAL OF TRANSJUGULAR INTRAHEPATIC PORTOSYSTEMIC STENT-SHUNT (TIPSS) VERSUS TIPSS AND VARICEAL BAND LIGATION (VBL) IN THE PREVENTION OF VARICEAL REBLEEDING

TIPSS is highly successful in treating variceal haemorrhage and preventing variceal rebleeding. We have previously shown TIPSS to be superior to VBL in the prevention of variceal rebleeding. However patients with a TIPSS require regular radiological surveillance in order to ensure patency. It is not known whether the combination of TIPSS and VBL can prevent the need for regular surveillance without compromising efficacy.

Aims: To compare the efficacy of TIPSS against that of TIPSS in combination with VBL in the prevention of variceal haemorrhage.

Methods: Between 1996–2000 eligible patients who required a TIPSS following an oesophageal variceal bleed were randomised to the TIPSS only arm (n=36, group 1) or TIPSS and banding arm (n=31, group 2). Average follow up was 24 and 26 months respectively. In group 1 patients underwent regular portography and stent dilatation or parallel shunt insertion if required. In group 2 TIPSS surveillance was performed for the first year only with VBL to achieve variceal eradication following the TIPSS.

Results: There was no difference in the average age, and aetiology of liver disease. Average Child-Pugh score was 9.25±0.34 and 8.47±0.40 in groups 1 and 2 respectively. Three patients in group 1 and 4 patients in group 2 rebled from oesophageal varices. However 1 patient bled from gastric varices and 3 from portal hypertensive gastropathy in group 2. In all cases of variceal bleeds the TIPSS was occluded or narrowed. 17 and 11 patients died in groups 1 and 2 respectively. There was no significant difference in the cumulative risk of rebleeding or mortality, and in the incidence of encephalopathy or ascites.

Conclusions: These interim results demonstrate that the combination of TIPSS and banding is as effective as TIPSS plus surveillance in preventing variceal rebleeding. Combining TIPSS and banding without long term angiographic surveillance is a treatment option. This may be especially relevant to patients with less advanced liver disease, since they are not committed to lifelong invasive TIPSS checks.

View Abstract

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.