G. Auzinger

629 CEREBRAL OEDEMA IS RARE IN ACUTE-ON-CHRONIC LIVER FAILURE (AOCLF)

mediated immune response. These immune modulations, and the peripheral T-cell depletion known to occur in cirrhosis, may support development of autoimmunity. We therefore aimed to investigate whether cirrhosis confers an increased risk of autoimmune disease. Methods: We used the Danish National Patient Registry (NPR) to identify all Danish citizens diagnosed with cirrhosis in 1977– 2009. For each of them we sampled five genderand age-matched population controls. Cox regression was used to compare the rate of rheumatoid arthritis, multiple sclerosis, or inflammatory bowel disease development (identified through NPR) between cirrhosis patients and controls. Results: We included 38,394 cirrhotic patients. During 217,558 total years of follow-up 442 of them developed rheumatoid arthritis (N =206), multiple sclerosis (N=20), or inflammatory bowel disease (N=216). Cirrhosis patients were more likely than controls to develop autoimmune disease (hazard ratio 1.66, 95%CI 1.49–1.85), irrespective of gender. The association was strongest shortly after cirrhosis was diagnosed and stable at a hazard ratio of about 1.45 from two years on after diagnosis of cirrhosis. Conclusion: Liver cirrhosis associates with development of multiple sclerosis, inflammatory bowel disease and/or rheumatoid arthritis. This may indicate a general enhancement in susceptibility for autoimmune phenomena in cirrhosis, and the underlying mechanisms should be delineated.

Interlobar Artery Resistive Index predicts Acute-on-Chronic Liver Failure Syndrome in Cirrhotic Patients with Acute Decompensation

INTRODUCTION Patients with acutely decompensated (AD) cirrhosis are at risk for developing acute-on-chronic liver failure (ACLF) syndrome. This syndrome is associated with a high short-term mortality rate. The aim of our study was to identify reliable early predictors of developing ACLF in cirrhotic patients with AD. PATIENTS AND METHODS We assessed 84 cirrhotic patients admitted for AD without ACLF on admission. We performed routine blood testing and detailed ultrasound Doppler studies of systemic arteries and mayor abdominal veins and arteries. We also calculated liver-specific and intensive care unit predictive scores. The area under the ROC curve (AUROC) was calculated for all variables that were significantly different between patients who developed ACLF and those who did not. Sensitivity, specificity, positive and negative predictive values, as well as diagnostic accuracy predicting the short-term development of ACLF were determined. RESULTS of the 84 patients, 23 developed ACLF whereas 61 did not. In the univariate analysis, serum levels of creatinine and urea, prothrombin time ratio, MELD score, portal vein and femoral artery flow velocity as well as the renal and interlobar artery resistive indices (RI) were associated with the short-term development of ACLF. However, only interlobar artery RI had independent predictive value in the multivariate analysis. The AUROC value for RI of the interlobar arteries was 0.9971. CONCLUSION On the first day of admission, ultrasound measurement of the RI of the interlobar arteries recognizes with high predictive accuracy those cirrhotic patients admitted with AD who will develop ACLF during hospital admission.

P138 SERONEGATIVE ACUTE LIVER FAILURE REPRESENTS A MACROPHAGE–T CELL ACTIVATION SYNDROME

P136 DEFICIENCY OF IL-33 SENSITIZES TO SEVERE LIVER INJURY INDUCED BY ConA BUT NOT BY CCl4 IN MICE M.I. Arshad, A. Filliol, V. Genet, C. Lucas-Clerc, J.-P. Girard, C. Piquet-Pellorce, M. Samson. Inserm U 1085, Institut de Recherche Sante Environnement & Travail (IRSET), Rennes, France; Institute of Microbiology, University of Agriculture, Faisalabad, Pakistan; Service de Biochimie CHU Rennes, Universite de Rennes 1, Rennes, Institut de Pharmacologie et de Biologie Structurale, Centre National de la Recherche Scientifique (IPBS-CNRS), Universite de Toulouse, Toulouse, Structure Federative BioSit UMS 3480 CNRS-US 18 Inserm, Rennes, France E-mail: michel.samson@univ-rennes1.fr

1008 CHARACTER AND TEMPORAL EVOLUTION OF APOPTOSIS IN ACETAMINOPHEN-INDUCED ACUTE LIVER FAILURE

Copyright

57 SECRETORY LEUKOCYTE PROTEASE INHIBITOR (SLPI) IS A PIVOTAL MEDIATOR OF ANTI-INFLAMMATORY RESPONSES IN ACUTE LIVER FAILURE

57 SECRETORY LEUKOCYTE PROTEASE INHIBITOR (SLPI) IS A PIVOTAL MEDIATOR OF ANTI-INFLAMMATORY RESPONSES IN ACUTE LIVER FAILURE C. Antoniades, W. Khamri, R. Abeles, A. O’Brien, L. Possamai, W. Bernal, C. Starling, R. Mitry, G. Auzinger, D. Gilroy, M. Heneghan, N. Heaton, A. Quaglia, D. Vergani, J. Wendon, M. Thursz. Section of Hepatology, Imperial College London, Intitute of Liver Sciences, King’s College London, Metabolism & Experimental Therapeutics, Faculty of Medical Sciences, UCL, London, UK E-mail: c.antoniades@imperial.ac.uk

92 THE HEPATIC INFLAMMATORY MICROENVIRONMENT INDUCES ENDOTOXIN TOLERANT MONOCYTES IN ACETAMINOPHEN-INDUCED LIVER FAILURE

by APAP (45±3.2) was attenuated following TLR4 antagonist (20±2.3) (p < 0.01). This was associated with a reduction in brain water (p < 0.01). Both the TLR4 antagonists significantly reduced pericentral necrosis of the liver. Both these interventions showed an improvement in the survival as using the log rank test (p< 0.02). TLR4 KO mice treated with APAP were protected from liver necrosis and had significantly better survival than wild type controls (p < 0.002). Conclusion: These data provides evidence for an important of TLR4 in APAP induced ALF and provide the rationale for a clinical trial of this strategy in ALF.

PTU-012 The impact of comorbidity on post liver transplant survival and resource utilisation in patients transplanted for acute liver failure utilising the Charlson comorbidity index

Introduction The presence of comorbidities negatively impact post liver transplant (LT) survival for those transplanted with chronic liver disease. Methods assess the impact of comorbidities on survival in patients transplanted for acute liver failure (ALF). Results 176 patients underwent LT for ALF over 9 years. Median follow-up was 92 months (range 35–142). Median age was 33 years (17–67) and 122 (69.3%) were females. Fifty-nine patients (33.5%) were transplanted for Paracetamol induced ALF. Ninety-six (54.6%) patients had ≥1 comorbidity. The commonest comorbidity was renal dysfunction in 84 (48%), pulmonary disease in 10 (6%), connective tissue disease in 5 (3%) and 2 (1%) had diabetes. Patients with ≥1 comorbidity had significantly increased 6 month (25% vs 13%, p=0.046), 12 month (27% vs 13%, p=0.023) and overall mortality (32% vs 17%, p=0.019). Similar results were demonstrated for graft survival. Recipient age ≥ 40 years (OR=1.37, 95% CI 1.02 to 1.86, p=0.039), the presence of comorbidity (OR=1.46, 95% CI 1.05 to 2.03, p=0.022) and renal dysfunction (OR=1.62, 95% CI 1.18 to 2.23, p=0.003) were associated with increased post LT mortality on univariate analysis. However, only the presence of comorbidity (OR=1.43, 95% CI 1.03 to 1.98, p=0.032) and renal dysfunction (OR=1.59, 95% CI 1.15 to 2.19, p=0.004) were independently associated with mortality. Other recipient related, donor, or graft variables were not associated with mortality. Patients with ≥1 comorbidity had significantly increased ICU length of stay (LoS) of 20 days (3 to 134) compared to those without comorbidities, 16 days (2–102), p=0.005. Conclusion Pre- LT comorbidity as defined by the presence of ≥1 comorbidity, significantly impairs overall post-LT patient and graft survival in patients transplanted for ALF. Patients with ≥1 comorbidity had significantly increased ICU LoS which may suggest increased resource utilisation. Competing interests None declared. Reference 1. Volk ML, et al. Modified Charlson comorbidity Index for predicting survival after liver transplantation. Liver Transplant 2007;13:1515–20.

PTU-002 The hepatic inflammatory milieu of acetaminophen induced acute liver falure induces alternatively activated, M2-like, macrophages

Introduction Human acetaminophen-induced acute liver injury (AALF) is characterised by areas of hepatic necrosis that are infiltrated by macrophages (mϕ) and elevated concentrations of anti-inflammatory cytokines. Anti-inflammatory mediators, such as secretory leucoprotease inhibitor (SLPI) and IL-10 contribute to the functional switching of mϕ from a proinflammatory (M1) type to an alternatively activated (M2) phenotype that promotes tissue repair processes. We sought to determine the effects of the hepatic inflammatory milieu on peripheral monocytes (MO) and MO derived mϕ in AALF. Methods Phosphoflow technique was used to evaluate NF-kBp65, STAT-3 signalling pathways in ex-vivo MO in 10 AALF patients and 10 healthy controls (HC). Results expressed as MFI and ratio of activation. Regional changes (portal vein [PV]), hepatic vein [HV])) were assessed in 5 AALF patients at time of transplantation. Serum (AALF [n=34]; HC [n=15]), hepatic (AALF [n=7]; HC [n=8]) and regional levels of TNF-α, IL-10 and SLPI were measured. The effect of the hepatic microenvironment was assessed in five cell culture experiments: purified CD14+ MO from HC were incubated with homogenates from AALF and culture medium (CM). Results In MO, TLR-4 stimulation reduced NF-kBp65 expression in AALF compared to HC (0.8 vs 1.6; p=0.001). Ex-vivo STAT-3 expression was significantly elevated in AALF patients compared to HC (600 vs 232; p=0.02). AALF patients had higher serum concentrations of IL-10 (170 vs 40; p<0.02), SLPI (71200 vs 43310; p<0.0001) compared to HC. A trans-hepatic (HV>PV) gradient was seen for IL-10 and SLPI but not for TNF-α. Hepatic levels of IL-10 (2 vs 0.6; p<0.02) and SLPI (442 vs 116; p<0.01) were significantly elevated in AALF compared to HC tissue, with peak concentrations of IL-10 detected in necrotic areas while SPLI was highest in areas of hepatic regeneration. In vitro exposure to AALF milieu induced a unique CD14+CD16+ macrophage phenotype characterised increased expression of alternative (M2) activation markers- CD36 (78 vs 55%; p=0.01) and CD163 (58 vs 32.5%; p=0.04), reduced LPS-induced TNF-a (19.8 vs 40.5%; p=0.02), IL-6 (13.2 vs 22%; p=0.04) and enhanced phagocytosis (80.5 vs 70%; p=0.03) when compared to MO incubated CM. Conclusion In AALF, circulating monocytes show modulations in intracellular signalling pathways compatible with anti-inflammatory responses. Our data also indicate that the anti-inflammatory hepatic microenvironment preferentially induces alternatively activated (M2)-like macrophages that may be implicated in resolution of acute liver injury. Competing interests None declared.

PWE-275 Multiple organ dysfunction syndrome (MODS), rather than the need for renal replacement therapy per se, dictates poor prognosis in patients with cirrhosis admitted to ICU

Introduction Critically ill cirrhotics admitted to ICU who receive renal replacement therapy (RRT) have a poor prognosis. It is unclear whether prognosis relates to renal dysfunction per se or whether RRT is one facet of MODS. We report the 7-year experience of outcomes and physiological disturbances in cirrhotics admitted to ICU who received RRT. Methods We analysed physiological and biochemical variables on admission for 478 cirrhotic patients admitted to ICU, excluding those transplanted during that hospital admission. Patients were cohorted on RRT requirement (RRT+/RRT−) at any point during ICU admission. Outcome, organ failure scores and utilisation of organ support were recorded. Results Of 253 RRT+ patients, 22% survived ICU and 13% survived hospital. Of 225 RRT−, 81% survived ICU and 59% survived hospital (p. On day 1 of admission, RRT+ had a higher prevalence of systemic inflammatory response syndrome (76% vs 62%), Child-Pugh (12 vs 11), MELD (32 vs 17) and SOFA scores (13 vs 11) (pEven when the renal component was removed from the SOFA score (SOFAminusRENAL), RRT+ had higher scores than RRT- (9 vs 7) (p. 23% of RRT+ commenced RRT after day 3 of admission; this did not affect ICU or hospital survival compared to those that commenced RRT before day 3. RRT+ survivors required less ventilation (39% vs 93%) and vasopressors (52% vs 89%) than RRT+ non-survivors and had lower Child-Pugh (12 vs 13), SOFA (12 vs 14) and SOFAminusRENAL (8 vs 10) scores (p=1 of admission. 23 patients required RRT but no other organ support, 18/23 (78%) survived to ICU discharge and 11/23 (48%) to hospital discharge. Conclusion The extent of MODS, rather than requirement of RRT per se, dictates poor prognosis in cirrhotics needing RRT in ICU. Requirement for RRT should not preclude admission to ICU, rather, prognostication should take into account other elements of MODS; in particular a concomitant requirement for circulatory and respiratory support. Competing interests None declared.

OP05 The impact of comorbidities on outcome of patients assessed for liver transplantation, waiting list mortality and post liver transplantation survival using the Charlson comorbidity index

Introduction Severity of liver disease determines accurately the outcome of patients on liver transplant (LT) waiting list (WL). Comorbidities are known to affect post-LT outcomes but their effect on outcome of transplant assessment (TA) or WL mortality have not been fully explored in previous studies. Aim To study the impact of comorbidities on TA, WL mortality and post LT survival. Method Retrospective study of all patients assessed for LT at our centre between 1 January 2000 and 31 December 2007 (n=1484). Patients with acute liver failure (175), amyloid (43), those assessed for re-LT (149) and 24 with incomplete information were also excluded. Nine comorbidities (Charlson Comorbidity Index - CCI) were prospectively defined according to Volk et al (Liver Transplant 2007;13:1515–20). Kaplan–Meier analysis was performed to determine impact of comorbidity on outcome. Cox regression hazard analysis was used to determine predictors of outcome and presented as (OR, 95% CI, p value). Results We analysed 1093 patients. Median age was 54 years (17–84), 67.5% were men (738). There were 192 (17.6%) patients with hepatocellular carcinoma (HCC). Patients with ≥1 comorbidity were 499 (46.6%) with most common comorbidities being diabetes (23.2%) and renal dysfunction (12.1%). Of 1093 assessed patients, 826 (75.6%) were listed. Patients with ≥1 comorbidity had significantly decreased LT free survival (log rank =33.586, p<0.001). Multivariate analysis showed CCI (1.79, 1.52 to 2.11, p<0.001), age (1.03, 1.00 to 1.05, p=0.035), Na (0.93, 0.89 to 0.97, p=0.001), MELD (1.10, 1.06 to 1.14, p<0.001) as being predictive. Of those listed for LT (826), 600 (72.6%) were transplanted, 161 (19.5%) died on WL and 65 (7.9%) were delisted. Listed patients with ≥1 comorbidity had significantly decreased LT free survival (log rank =9.045, p=0.003). Multivariate analysis showed CCI (1.79, 1.52 to 2.11, p<0.001), age (1.02, 1.01 to 1.03, p=0.006), pre-LT Hb level (0.87, 0.80 to 0.95, p=0.003), Na (0.96, 0.93 to 0.99, p=0.014) and MELD (1.14, 1.11 to 1.18, p<0.001) were predictors of listing outcome. Transplanted patients with ≥1 comorbidity had significantly decreased post LT survival (Log rank =7.645, p=0.006). Multivariate analysis showed that only CCI (1.36, 1.13 to 1.64, p=0.001) and HCC (1.74, CI 1.21–2.51, p=0.003) were independently associated with post-LT mortality. Similar post LT survival results seen when CCI was divided into 0, 1 and ≥2 comorbidities (Log rank =11.342, p=0.003). Conclusion We demonstrate that comorbidities significantly impact on the outcome of patients with chronic liver disease at TA, on wait-list and post LT survival. Adding CCI to known liver prognostic models may improve their prediction ability.