Article Text

PDF

Liver free papers 108–123

Statistics from Altmetric.com

108 INTERLEUKIN 10 SECRETION DIFFERENTIATES BETWEEN INTERSTITIAL DENDRITIC CELLS FROM HUMAN LIVER AND SKIN

S. Goddard, D Adams. Liver Unit Laboratories, Queen Elizabeth Hospital, Birmingham, UK

Dendritic cells are thought to be the only cell capable of initiating primary immune responses, to produce both immunity and tolerance. They are also able to direct the kind of T cell response generated to specific antigen. Liver immune responses are relatively weak, for instance liver allografts are less susceptible to rejection. Human liver dendritic cells (DCs), which may orchestrate the liver's unique immunoregulatory functions, remain poorly characterised. We used a novel technique of overnight migration from tissue pieces of normal liver and skin to obtain human tissue-derived DCs with minimal manipulation and no additional cytokine treatment.

As presented at a previous BSG liver DCs have a monocyte-like morphology and a partially mature phenotype after migration overnight from tissue. We now show that liver DCs express CD123, a marker expressed by a subset of DCs associated with initiating Th2 T cell responses. In addition, a functional comparison was made between liver and skin DCs isolated the same way. ELISA measurement of cytokine in DC conditioned media showed that, liver DCs produced IL-10 whereas skin DCs failed to secrete IL-10 even after stimulation and neither skin nor liver-derived DCs secreted IL-12. The effect of DC stimulation on T cells was studied following coculture and T cell intracellular cytokine staining. Liver DCs stimulated T cells to secrete IL-10 whereas skin DCs stimulated IFNγ and IL-4 secretion in the absence of detectable IL-10.

We show for the first time clear tissue-specific differences in human non-lymphoid DCs. The ability of liver DCs to secrete IL-10, a cytokine implicated in down-regulation of immune responses, may explain how interstitial DCs from normal liver can maintain tolerance to gut derived Ag, by controlling the type of response generated in tissue or draining lymph node

109 SUSCEPTIBILITY TO PRIMARY SCLEROSING CHOLANGITIS IS ASSOCIATED WITH A POLYMORPHISM OF THE MMP-9 (GELATINASE B) GENE

S.N. Cullen, D.P. Jewell, R.W. Chapman. Gastroenterology Unit, University of Oxford

Background: Primary sclerosing cholangitis (PSC) is a disease of the intrahepatic and /or extrahepatic bile ducts which is characterized by concentric obliterative fibrosis and bile duct strictures eventually leading to biliary cirrhosis. The matrix metalloproteinase family of zinc-containing proteolytic enzymes is involved in mediating extracellular matrix degradation. An association between a functional polymorphism of MMP-3 (stromelysin) and susceptibility to PSC has recently been described. MMP-9 polymorphisms have been described in association with progression of coronary atherosclerosis and cancer metastasis. This study assessed carriage of MMP-9 polymorphisms in relation to susceptibility to PSC.

Method: DNA was extracted from 69 patients with well-documented PSC, 71 patients with ulcerative colitis, and 92 healthy controls. Primers were designed to examine 8 polymorphisms in the MMP gene using a SSP/PCR method. PCR products were run on 1% agarose gels and read under UV light. PSC and UC patients were compared with controls using 2x2 contingency tables and a χ2 test (with Yates correction). A Bonferroni correction for multiple comparisons was made using a factor of 8 (the number of polymorphisms tested).

Results: The R279Q polymorphism was significantly associated with susceptibility to PSC compared with controls. The frequency of the mutant allele was 32% in the PSC patients compared with 17% in controls (pc =0.008). There was a trend towards increased carriage in the ulcerative colitis group but this did not reach statistical significance after correction (pc =0.16). No associations were seen with any of the other polymorphisms tested.

Conclusion: There is increased carriage of the R279Q polymorphism in PSC patients. This polymorphism is in the catalytic region of the gene and may therefore influence the function of MMP-9. Studies are currently being undertaken to address the possible functional effects of this polymorphism.

110 PRIMARY BILIARY CIRRHOSIS (PBC): NO SPECIFIC ASSOCIATION WITH MICROCHIMERISM

R. Buckland, K.L.E. Dear1, A.C. Goodeve, J.C.E. Underwood2, D. Gleeson1. Division of Genomic Medicine and 2Dept of Pathology, University of Sheffield; 1Liver Unit, Sheffield Teaching Hospitals, Sheffield, UK

Background: Some diseases may result from immunocompetent fetal cells acquired during previous pregnancies (allo-immunity) and persisting in the mother for decades (microchimerism). In women with scleroderma who have had male children, male DNA is found in skin and peripheral blood more frequently than in control women. Primary biliary cirrhosis (PBC) mainly affects older women and has similarities to graft versus host disease, consistent with a role for alloimmunity. Studies on the association of PBC with microchimerism have been small (<20 PBC liver specimens) and results have been conflicting.

Aim: To address the association of PBC with microchimerism in a larger cohort.

Methods: We studied (a): blood (2 extractions x 2 PCR= 4 PCR) from 55 women with PBC and 49 normal control women (irritible bowel or G-O reflux; normal liver enzymes) and (b): archived needle liver biopsies (1 extraction x 2 PCR = 2 PCR) from 42 women with PBC, 21 women with normal liver histology or mild steatosis (normal controls), and 32 women with autoimmune hepatitis (AIH) (disease controls). All had had ≥1 male child and none met ARA criteria for scleroderma or had had a blood transfusion or a liver transplant when tissue was obtained. Male DNA was assayed by PCR using specific Y chromosome primers; lower detection limit = 1/3 of DNA in 1 cell (2pg).

Results: The table shows subject numbers (%) with 0, 1 and.≥2 PCRs positive for male DNA.

Abstract 110

Conclusion: PBC showed an unexpected negative association with microchimerism in blood and no significant association (compared to either control group) in liver. This, the largest study to date, does not support a role for microchimerism in the pathogenesis of PBC.

111 FASTING INSULIN, 31,32 SPLIT PRO-INSULIN AND INSULIN RESISTANCE IN PATIENTS WITH NON-ALCOHOLIC FATTY LIVER

G.Constable, D.Cumming, P.J.Wood, S.Wootton, M.A.Stroud. Institute of Human Nutrition, Southampton University Hospital, UK

Introduction: Nonalcoholic fatty liver (NAFL) is associated with insulin resistance and an increased risk of type II diabetes (T2DM). Development of T2DM results from either insulin resistance or defective insulin processing/secretion or both. Insulin resistance leads to increased lipolysis and flux of fatty acids (FFA) to the liver, which in turn may increase hepatic glucose output. Insulin processing involves enzymatic conversion of proinsulin to insulin through a series of site-specific cleavages. Recent development of specific assays allows different molecules of the proinsulin processing pathway to be measured separately. 32,33 split proinsulin is the predominant form of proinsulin, known to be elevated in T2DM and impaired glucose tolerance. Fasting levels have also been shown to predict development of T2DM.

Methods: We measured fasting insulin, intact proinsulin and 32,33 split proinsulin levels in patients with NAFL (n=24) compared to a healthy reference group (n=14). Insulin resistance index (IRI) was calculated from fasting plasma glucose and insulin levels using the well recognised mathematical model of glucose:insulin interactions - the `homeostatic model assessment' (HOMA). Body composition was assessed by BMI, waist:hip ratio and bioelectrical impedance analysis (BIA).

Results: Patients with NAFL were obese according to BMI: 31.19 +/- 0.81 vs 26.55 +/- 0.81, p<0.004. The NAFL group had higher insulin levels than reference: 21.44 +/- 2.64 mU/l vs 9.07 +/- 1.16 mU/l, p<0.0014 and higher 32,33 split proinsulin levels: 18.50 mU/l +/- 1.85 mU/l vs 8.55 +/- 0.73 mU/l, p<0.0003. The NAFL group were also significantly insulin resistant using HOMA, p<0.0015. These results were independent of BMI and body fat assessed by BIA.

Conclusions: NAFL is associated with hyperinsulinaemia, elevated 32,33 split proinsulin levels and insulin resistance. Patients with NAFL are at increased risk of developing T2DM. Measurement of 32,33 split proinsulin may help to select individuals with the highest risk for targeted intervention. Insulin resistance causes an increased flux of FFA to the liver leading to fat deposition if metabolic competence for disposal is exceeded. Increased FFA flux may also result in hepatic overproduction of glucose through mechanisms that are incompletely understood.

112 INHIBITION OF APOPTOSIS OF ACTIVATED HEPATIC STELLATE CELLS BY TIMP-1 IS MEDIATED VIA EFFECTS ON MMP INHIBITION: IMPLICATIONS FOR REVERSIBILITY OF LIVER FIBROSIS

F. Murphy, R. Issa, X. Zhou, H. Hussain, S. Ratnarajah, P. Soloway1, H. Nagase2, M.J.P. Arthur, R.C. Benyon, J.P. Iredale. Liver Group, Division of Infection, Inflammation & Repair, Southampton University, Southampton SO16 6YD, UK; 1Rosewell Park Cancer Institute, New York, USA; 2The Kennedy Institute, Imperial College, London, UK

Introduction: The activated hepatic stellate cell (HSC) is central to liver fibrosis as the major source of collagens I and III and the tissue inhibitors of metalloproteinase-1 and -2 (TIMPs). During spontaneous recovery from liver fibrosis there is a decrease of TIMP expression, an increase in collagenase activity and apoptosis of HSC, highlighting the potential role of TIMP-1 and -2 in HSC survival.

Aims: To determine if TIMP-1 and TIMP-2 directly inhibit HSC apoptosis in tissue culture and in models of liver fibrosis in vivo.

Methods: Effects of recombinant TIMPs and mutated TIMP-1 on cultured activated HSC were examined after induction of apoptosis by cycloheximide in vitro. Rat and murine models of experimental liver fibrosis induced by CCl4 were examined during spontaneous recovery. HSC number, TUNEL staining and TIMP-1 mRNA were assessed.

Results: TIMP-1 and 2 demonstrated a consistent, significant and dose dependent anti apoptotic effect on HSC activated in tissue culture. A non-functional mutated TIMP-1 (T2G mutant) did not inhibit apoptosis indicating that inhibition of apoptosis was mediated through MMP inhibition. Studies of experimental liver fibrosis in the rat demonstrated that loss of activated HSC correlated with a reduction in TIMP-1 mRNA expression determined by PCR. Persistence of HSCs in more advanced fibrosis correlated with persistent TIMP-1 mRNA expression. After induction of fibrosis in vivo, TIMP-1 knockout mice demonstrated significantly more HSC apoptosis relative to wild types at 3 and 7 days of spontaneous recovery.

Conclusion: TIMP-1 and -2 inhibit apoptosis of activated HSC. The anti apoptotic effect of TIMP-1 is mediated via MMP inhibition.

113 DISTRIBUTION OF THE CONSTITUTIVE (COX-1) AND THE INDUCIBEL (COX-2) CYCLOOXYGENASE IN HUMAN LIVER CIRRHOSIS: A POSSIBLE ROLE OF COX-2 IN PATHOGENESIS OF LIVER CIRRHOSIS

N. Mohammed, S. Abd El-Aleem, H. Abdel Hafiz, M. Maklouf, R.F.T. McMahon. Department of Histopathology, Clinical Sciences Building, Manchester Royal Infirmary, Oxford Road, Manchester M13 9WL, UK

Several mediators of systemic vasodilatation and inflammation in liver cirrhosis have been reported. Among these are prostaglandins (PGs), which have been proposed as one of the main mediators during inflammation. In this study, liver biopsies from fifteen patients with clinically and pathologically diagnosed liver cirrhosis secondary to hepatitis B and C, were taken. In addition 3 liver biopsies from healthy controls were used. The protein expression of the constitutive (COX-1) and the inducible (COX-2) cyclooxygenase was investigated using immunocytochemistry.

We have shown that COX-2 was completely absent from the control group but was highly expressed in the cirrhotic livers. COX-2 was seen mainly in the inflammatory cells infiltrating the liver, sinusoidal cells, vascular endothelial cells and epithelial lining of bile ducts. On other hand COX-1 was expressed in normal and cirrhotic livers. COX-1 was exclusively seen in sinusoidal cells and vascular endothelial lining cells. There were no significant differences in COX-1 expression between normal and cirrhotic livers.

It is therefore clear that COX-2 is induced in liver cirrhosis and this could contribute to the overproduction of prostaglandins which could be a major contributor to hyperdynamic circulation associated with liver cirrhosis. High production of COX-2 in cirrhotic liver could explain the occurrence of hepatocellular carcinoma since COX-2 is believed to be carcinogenic. Finding that COX-2 and not COX-1 is markedly upregulated in cirrhosis could provide a possible new line of treatment using selective COX-2 inhibitors to treat the inflammation and also to minimise the occurrence of HCC in cirrhotic patients.

114 EXPRESSION OF NITRIC OXIDE SYNTHASE ISOFORMS IN HUMAN LIVER CIRRHOSIS

N.A. Mohammed, S.A. Abd El-Aleem, M.M. Maklouf, M. Said, R.F.T. McMahon. Department of Histopathology, Manchester Royal Infirmary, Oxford Road, Manchester M13 9WL, UK

Several mediators of systemic vasodilatation in liver cirrhosis have been reported. Among these is nitric oxide (NO), which has been proposed as one of the main mediators. In this study sera and liver biopsies from fifteen patients with clinically and pathologically diagnosed liver cirrhosis were taken. In addition sera from 7 and 3 liver biopsies from healthy controls were used. Serum levels of nitrite (the end product of nitric oxide) were measured using Griess reaction and the protein expression of the inducible nitric oxide synthase (iNOS) and constitutive nitric oxide synthase (ecNOS) was investigated using immunocytochemistry. We have shown that the serum nitrite levels (94 ± 9.8 μmol/L) in cirrhotic patients were significantly (P <0.05) increased by comparison to the control (36.6 ± 11.03 μmol/L). iNOS was completely absent from the control group but was highly expressed in the liver of the cirrhotic group. iNOS was seen mainly in the inflammatory cells infiltrating the portal tracts, blood monocytes, hepatocytes, sinusoidal cells and vascular endothelial lining. However, ecNOS was only seen in the vascular endothelial lining of both the control and cirrhotic groups but much higher in the latter. It is therefore clear that NO is augmented in cirrhotic patients and it is mainly produced by induction of iNOS. Moreover, NO upregulation is dependent on the inflammatory stage of liver cirrhosis. ecNOS production could be a normal chronic adaptation mechanism of the endothelium to the chronically increased splanchnic blood flow secondarily to portal hypertension. In the near future, the appropriate inhibition of NO synthesis by using selective iNOS inhibitors may provide a novel strategy for the treatment of patients with liver cirrhosis or at least improve the fate of cirrhosis.

115 SPLANCHNIC VASCULAR HYPOREACTIVITY IN HUMAN CIRRHOSIS IS RELATED TO DISEASE SEVERITY AND MEDIATED BY NITRIC OXIDE AND CARBON MONOXIDE

R. Harry, D. Grieve, M. Bowles, N. Heaton, P. Muiesan, M. Rela, J. Wendon, A. Shah1. Institute of Liver Studies and 1Department of Cardiology, King's College Hospital, Bessemer Road, London SE5 9RS, UK

Cardiovascular changes of cirrhosis correlate with disease severity1 and are associated with vascular hyporesponsiveness to vasoconstrictors2 and splanchnic vasodilation3. The effect of disease severity on in vitro vascular reactivity and the mechanisms involved in humans have not been studied.

Methods: We studied endothelial-denuded rings of human hepatic artery from patients undergoing orthotopic liver transplant for cirrhosis (n=9) and from organ donors and patients undergoing hepatic resection (controls: n=6). Decompensated cirrhosis was defined as Child Pugh score > 8 (n=5) and compensated cirrhosis <8 (n=4). The response to 80 mmol/L potassium chloride was recorded. Rings were then incubated with either 0.1 mM L-NMMA (a non-selective nitric oxide synthase inhibitor), 0.1 mM ZnPP (a non-selective haem oxygenase inhibitor) or vehicle control for 30 min. Cumulative dose response curves to phenylephrine (PHE) were constructed.

Results: Decreased maximal response to PHE was seen in decompensated cirrhosis compared with controls (P<0.002) and compensated cirrhosis (P<0.002) (Figure). In compensated cirrhosis, the maximal response was not different from controls. L-NMMA or ZnPP improved the maximal response in decompensated cirrhosis toward control values (1.11+/-0.19 mg/g and 0.93+/-0.10 mg/g respectively). Neither inhibitor affected the PHE response in compensated cirrhosis or controls.

Conclusions: Hepatic artery hyporeactivity to PHE occurs only in patients with decompensated cirrhosis and not those with compensated cirrhosis. Restoration of PHE responsiveness by L-NMMA or ZnPP in these endothelial-denuded vessels suggests that smooth muscle derived nitric oxide and carbon monoxide may be important and induced only in advanced disease.

1Sherlock et al, 1958;2Iwao et al, 1997;3Lunzer et al, 1975.

116 ACUTE LIVER FAILURE SERUM CAUSES APOPTOTIC CELL DEATH BY DOWN-REGULATION OF β1-INTEGRIN ACTIVITY

P.N. Newsome, K. Humphreys, R. Buttery, T. Sethi, P.C. Hayes, J.N. Plevris. Department of Medicine, University of Edinburgh, UK

Integrity of the cytoskeletal axis is important for the maintenance of cellular differentiation, adhesion and viability. Disruption of this axis in acute liver injury may limit the role of stem cell/hepatocyte transplantation.

Aim: To study the effects (and mechanisms) of acute liver failure (ALF) serum on hepatocyte adhesion/cell death.

Methods and Results: HepG2 cells were cultured in media supplemented with 10/20% ALF or Normal human (NS) serum. Culture with 20% ALF serum led to significant increases in apoptotic cell death (Feulgen staining/TEM) after 24 (4%) and 48 (7.5%) hrs compared with negligible levels of apoptosis seen in culture with NS. Cellular adhesion (attachment to collagen coated plates after culture in either ALF or NS) was significantly decreased in cells grown in ALF serum. Of note this effect became pronounced after just 4 hours culture, well before apoptosis was observed (see table). Using Scanning Electron Microscopy cells cultured in ALF serum were strikingly more rounded in appearance and appeared far less adherent. Flow cytometric expression of the common integrins (β1,α2,α6) on HepG2 cells was carried out, and after 24 hours culture levels were seen to be higher on cells cultured in ALF serum (β1 log mean fluorescence: 2.61 in ALF vs 2.25 in NS). We then studied the activation level of the β1-integrin using a flow cytometric assay. After 24 hours culture, ALF serum significantly reduced the activity of the β1-integrin compared with control cultures (33.6 +/- 6.2 (ALF) vs 69 +/- 10.1 (NS)).

Abstract 116

Conclusion: Down-regulation of β1-integrin activity appears to be an early event in cells exposed to ALF serum, which precedes impaired cellular adhesion and apoptotic cell death, thus negating the possible therapeutic benefits of cell transplantation in liver injury. Modulation of integrin activity may be important in optimising cell transplantation.

117 IS THE PROTHROMBIN TIME STILL USEFUL FOR ASSESSING FULMINANT HEPATIC FAILURE?

P.N. Newsome, N.C. Henderson, L. Germain, J.N. Plevris, P.C. Hayes, C.A. Ludlam, K.J. Simpson. Scottish Liver Transplant Unit, Royal Infirmary of Edinburgh & Department of Haematology, University of Edinburgh, UK

Prognostic markers, such as prothrombin time, are important in guiding management in patients with Fulminant Hepatic Failure (FHF). The measurement of PT has recently changed from the Manchester method to the Innovin method, which utilises recombinant reagents and as such should provide less variability.

Aim: To study whether the two methods provide similar results when tested on samples taken from patients with FHF admitted to the Scottish Liver Transplantation Unit (SLTU).

Methods and Results: We prospectively studied 96 samples from 42 patients (35 paracetamol, 3 Non-A Non-B, 2 autoimmune, 1 Budd-Chiari Syndrome & 1 drug reaction) admitted to the SLTU from August 2000 to March 2001 with acute liver injury. Of these 17 developed encephalopathy (and thus FHF) and 9 met the King's College Poor Prognosis criteria. The equation derived by linear regression to relate the two values is Manchester prothrombin time = 0.61 + (1.72 x Innovin prothrombin time) (R2=0.766, p<0.001). Whilst there is a strong association between these methods it is clear that there is not equivalence. Significantly the Innovin measured PT consistently records shorter times than the Manchester measured PT, and thus usage of the Innovin PT could potentially lead to patients not being considered for liver transplantation as they would not meet KCPP criteria. Indeed in our cohort no patients had an Innovin PT of greater than 100 seconds, which is one of the requirements for consideration of liver transplantation for paracetamol poisoning. A time of 28.7 seconds and 57.8 seconds as measured by the Innovin corresponds to 50 and 100 seconds as measured by the Manchester method.

Conclusions: Our study demonstrates that Innovin measured PT is different from that measured with the Manchester reagent, and therefore its usage in FHF, as per the KCPP criteria, cannot be recommended. Widespread usage of the Innovin PT in FHF will require validation in larger studies before it can be recommended. This finding has important implications in the management of FHF.

118 PENTOXIFYLLINE IMPROVES SHORT TERM SURVIVAL IN SEVERE ACUTE ALCOHOLIC HEPATITIS

J. Diamond, R. Krisciunaite, H.H. Mohamed, A. Bathgate, P.C. Hayes. Liver Unit, Department of Medicine, Royal Infirmary of Edinburgh, EH3 9YW, UK

Background and Aims: Pentoxifylline(PTX), an inhibitor of Tumor Necrosis Factor has been reported to improve the outcome of acute alcoholic hepatitis1.The aim of the current study was to review our experience with this drug and to compare the results with a large control population.

Methods: The treatment group comprised of 8 consecutive patients with severe acute alcoholic hepatitis with Maddrey discrimination factor (MDF) >32 who were consequently treated with PTX (400 mg orally 3 times/day) for 4 weeks. A group of 35 patients who were admitted before PTX was used in our unit with similar MDF score served as the control group. There were no statistical significant differences between the two groups as regards the demographic and clinical criteria or laboratory values, with the exception of an elevated serum creatinine, which was significantly more common in the control group (p < 0.05).

Results: The four week mortality in the treatment group was 0% as compared with the control group with a 4 week mortality of 77.1% (p < 0.0001). Hepato-renal syndrome developed in 50% of patients in the treatment group compared with 80% of the control group. Serum creatinine on admission, development of hepato-renal syndrome and serum alkaline phosphatase were three independent factors associated with mortality. Serum bilirubin and MDF showed significant improvement over 4 weeks in the PTX treated group (p < 0.03 and 0.02 respectively). The drug was well tolerated in all the patients and all received 4 weeks therapy. 2 of the patients in the treated group died subsequently on follow up 2 and 3 months later.

Conclusions: Treatment with pentoxifylline was well tolerated and appears promising in improving short-term survival in patients with severe acute alcoholic hepatitis. Such improvement in survival rate appeared to be related to the significant decrease in the risk of development of hepato-renal syndrome.

1Akriviadis E, Botla R, Briggs W, et al. Pentoxifylline improves short-term survival in severe acute alcoholic hepatitis: a double-blind, placebo-controlled trial. Gastroenterology 2000;119:1637–48.

119 ACUTE SICKLE CELL HEPATOPATHY: A NEW CONTRAINDICATION TO PERCUTANEOUS LIVER BIOPSY

N. Zakaria, A. Knisely, B. Portmann, J. Wendon, R. Arya, J. Devlin. Institute of Liver Studies, King's College Hospital, London SE5 9RS, UK

Background: Percutaneous biopsy of the liver is an invaluable diagnostic tool in the investigation of liver pathology and is associated with a low incidence of complications. The mortality rate post-liver biopsy is approximately 0.1%. Following recent experience of complications including mortality following liver biopsy in patients with sickle cell disease, we undertook a systematic review of our clinical experience in this population.

Methods: Sixteen patients with sickle cell disease who underwent a percutaneous liver biopsy were identified. Clinical records, post-mortem reports, and the Coroner's death register were reviewed. Demographics, duration of disease, frequency of sickling crises, sickle cell genotype, haematological and biochemical indices, and histopathological findings were correlated with course after biopsy and clinical value of data yielded by biopsy.

Results: Five of 16 patients (31%) suffered serious haemorrhage, and four died (80%/25%). None of the eleven patients without biopsy complications was in acute sickling crisis at the time of biopsy; four of the five patients with complications were. Four of these five patients underwent biopsy for an emergency indication. Chronic venous outflow obstruction, marked hepatic sequestration of erythrocytes, and sinusoidal dilatation were strongly associated with complications. Data obtained by biopsy in patients who suffered complications were assessed as not of substantial value in clinical management, nor likely to have been of value had patients survived, as most of these patients (80%) had hepatic sequestration crisis. In most of those (8/11; 73%) who did not suffer complications, however, the biopsy data were useful.

Conclusions: Percutaneous liver biopsy in patients with acute sickle cell hepatopathy complicating sickle cell anaemia carries a high risk. To recognise acute sickle cell hepatopathy is important; such a condition may represent a newly identified contraindication to percutaneous liver biopsy.

120 COST EFFECTIVENESS OF HISTOACRYL GLUE VS. TIPS IN THE MANAGEMENT OF ACUTE GASTRIC VARICEAL BLEEDING

S. Mahadeva, M.C. Bellamy, D. Kessel, M.H. Davies, C.E. Millson. Liver Unit, St James's University Hospital, Leeds, UK

Introduction: The management of bleeding gastric varices (GV) has not been standardised. Although TIPS is used in most centres, endoscopic treatment with histoacryl glue has been shown to be effective recently. Cost-effective analyses of these methods are lacking.

Methods: Review of results of patients who were treated for bleeding gastric varices in this institution-initially by TIPS and later with histoacryl glue injection. Cost analysis, based on hospital charges, for a fixed financial year and comparison between the two groups were made for up to a period of 6 months, liver transplantation or death for each patient.

Results: 20 patients with bleeding GV had TIPS from January 1995 to December 1999 whilst 23 patients had histoacryl glue injection from January 2000 to October 2001. There were no significant differences in age, sex, diagnosis, Child-Pugh classification and transfusion requirement between the two groups. The TIPS group had 13 cases of GOV 1 and 7 with GOV 2, with 15 cases of GOV 1 and 8 with GOV 2 in the glue group (p=NS). In the TIPS group, 15/20 patients had the procedure within 24 hours of haemorrhage and 90% of stent insertions were successful. Complications consisted of three cases of pulmonary oedema, two cases of severe encephalopathy and a 15% stenosis rate at 6 months. In the glue group, there were 3±1.5 endoscopies and 2±1 injections per patient with a 96% haemostasis. There was one case of fatal (glue) pulmonary embolism and one blocked front endoscope lens which required repair. The re-bleed rate was 35%(TIPS) vs 40%(glue) (p=0.53). The inpatient stay was shorter in the glue group (13±1 vs 18±2, p=0.05), but there was no difference in the early mortality rate (17% vs 15%). The median cost within six months of initial GV bleeding was £2685 (1924–5717) for glue vs. £8211 (5517–12,001) for TIPS (p<0.0001).

Conclusion: In this comparable group of patients, histoacryl glue injection is more cost effective than TIPS in managing acute GV bleeding. A prospective, randomised trial is required to subtantiate our analysis.

121 GENETIC HAEMOCHROMATOSIS (GH): WHERE ARE ALL THE PATIENTS?

D. Thorburn1, A. Cook2, R. Spooner3, G. Reid1, A.J. Stanley1. 1Gastroenterology Unit, Glasgow Royal Infirmary; 2Department of Genetics, Yorkhill Hospital; 3Department of Biochemistry, Gartnavel General Hospital, Glasgow, UK

Background: Genetic haemochromatosis (GH) is the most common genetic disorder of Caucasian populations. A single autosomal recessive gene mutation (C282Y) accounts for over 90% of cases of GH in the UK. The highest carrier frequency reported for this mutation (approximately 1 in 10) is observed in North European or Celtic populations. Therefore there is likely to be a high prevalence of patients with GH in Scotland, but data are scarce.

Aims: To (1) estimate the gene frequency of the C282Y mutation in healthy control populations in Glasgow; (2) establish the prevalence of known cases of GH in Glasgow; (3) estimate the number of patients with undiagnosed GH in Glasgow.

Methods: C282Y mutation frequency was established anonymously in two healthy control populations from Glasgow: umbilical cord blood samples from consecutive newborn infants and randomly selected healthy elderly controls. All patients in the Greater Glasgow Health Board (GGHB) area homozygous for the C282Y mutation up to 1st August 2001 were identified. The prevalence of GH was estimated from the frequency of GH in controls and the known GGHB population. The number of patients with undiagnosed GH was estimated from this figure and the number of known cases of GH.

Results: 340 controls (163 infants, 177 elderly controls) underwent C282Y testing. The C282Y mutation gene frequency was 7.4% (6.4% and 8.2% respectively) with a carrier rate of 1 in 7. An estimated 1 in 183 (5.5 per 1000) were homozygous, which equates to 4942 cases of GH within the 904,400 population of the GGHB area. Only 240 (0.26 per 1000) C282Y homozygotes are recognised. This represents 4.9% (240 / 4942) of the estimated number of C282Y homozygotes in the area.

Conclusions: As expected the C282Y mutation gene frequency is high in the control groups studied. Only a small minority (4.9%) of the estimated 4942 individuals with GH in Glasgow are recognised. Whether this reflects lack of biochemical or clinical penetrance of the C282Y mutation, or a failure of diagnosis requires futher study.

122 HEPATITIS C: WHY HAVE SO FEW PATIENTS BEEN TREATED?

J. Smart, T. Jones, R.G. Batey (introduced by A.E. Duggan). GE Department, John Hunter Hospital, Newcastle, NSW, Australia

Hepatitis C (HCV) is a chronic illness for which a relatively effective anti-viral therapy exists. Many sufferers are referred to liver clinics with long waiting lists but a significant number do not receive active therapy.

A retrospective review of all HCV positive patients attending an outpatient liver clinic was conducted to determine the number who commenced anti-viral therapy and the reasons why patients did not. From Oct 1994-Dec 2000 all those attending clinics were documented. Those seen from Oct 1994 - Oct 1997 were reviewed in detail to determine the reasons why patients did not receive treatment (see table).

Abstract 122

Of 490 HCV patients (298M/192F) seen from Oct 1994–Dec 1997, 174 received anti-viral therapy (interferon alone or with ribavirin) and 316 did not. 27% of females were treated compared to 41% males. Predominant risk factors for HCV in either group were 310 IVDU (63%) and 85 blood products (17%). Primary reasons for not receiving therapy: 105 (33%) did not meet Government criteria, 67 (21%) lost to follow up after visit 1 or 2, 54 (17%) health, 45 (14%) herbal, 22 (7%) social and 23 (7%) chose no treatment. Of the 1104 patients seen over the 7 years, only 405 (37%) received anti-viral therapy.

Only 37% of HCV patients attending our clinic received anti-viral therapy. The major reasons for non-treatment were related to inadequate evaluation and education pre-referral. This did not change with time. We have now provided referral checklists to general practitioners, established a new patient review clinic run by our clinical nurse consultant and we will evaluate the effect of these changes on treatment uptake rates. As HCV will continue to place major demands on busy liver clinics there remains a need to optimise use of clinic time.

123 USING THE JOINT BRITISH SOCIETIES CORONARY RISK PREDICTION CHARTS TO CALCULATE CORONARY HEART DISEASE RISK AFTER LIVER TRANSPLANTATION

D.A.J. Neal1, P. Flynn1, A.E.S. Gimson1, P. Gibbs2, G.J.M. Alexander1. 1Department of Medicine and 2University Department of Surgery, Addenbrooke's NHS Trust, Cambridge, UK

Background: Hypertension, hypercholesterolaemia and weight gain are common after liver transplantation. It is not known whether development of these complications alters the cardiovascular risk profile after liver transplant.

Methods: The case notes of 110 consecutive adult liver transplant recipients surviving beyond one year were reviewed.

Results: Median follow-up was 52 months (range 6–90 months). 74 % of patients developed hypertension compared with 3 % being hypertensive before transplant (P<0.001). Hypercholesterolaemia was present in 16 % before and 60 % after transplant (P<0.001). 29 % were overweight at the time of transplant compared with 58 % after transplant (P<0.001). Diabetes mellitus was present in 8 % before and 12 % of patients after transplant. There were 3 non-fatal cardiovascular events: 1 myocardial infarct, 1 heart failure and 1 cerebellar infarct. The Joint British Societies Coronary Risk Prediction Charts categorise 10-year coronary risk (on the basis of total cholesterol:high density lipoprotein cholesterol ratio, systolic blood pressure, smoking, diabetes mellitus, age and gender) as < 15 %, 15–30 % and > 30 %. Using these charts we categorised coronary heart disease risk before and after transplant (see table). If we assume patients require therapeutic intervention aimed at reducing risk when risk is 15 – 30 % or greater, 20 patients (18 %) would require treatment before transplant compared with 52 (47 %) after transplant.

Abstract 123

Conclusions: Coronary heart disease risk increases after liver transplant. The number of observed cardiovascular events is low. We would expect more events than we have seen.

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.