- Email: [email protected]
P1091 : One year effectiveness of Baclofen treatment in 100 alcohol-dependent patients
POSTERS Figure (top) gives the probability of death using DF at day 0 (Yaxis) and Lille model at day 7 (X-axis). As an example, a patient admitted with a DF at 100 with a Lille model at day 7 of 0.25 has a probability of death of 27% at 6 months, which increases at 72.3% if Lille model is 0.7. The likelihood of the joint-effect model was improved in comparison to a model based solely on Lille model (p = 0.016). We conducted a second analysis Lille-MELD on a subgroup of 638 patients for whom MELD score was available (Figure, bottom). The efficacy of the joint-effect model was also better as compared to each model alone (p < 0.001). As an example, a patient admitted with a MELD score at 28 with a Lille model of 0.25 has a probability of dying of 27% at 6 months, which increases at 69.7% if Lille model is 0.7. The two combinations (Lille+Maddrey DF and Lille+MELD) showed good discrimination and calibration, the latter meaning that predicted survival fitted the observed survival. Using the Akaike information criterion, we found that the combination Lille+MELD was a better fitting prognostic model as compared to the combination Lille+Maddrey DF. Conclusions: We propose a new approach that predicts the outcome of patients with severe alcoholic hepatitis in terms of a continuum risk of mortality. This may be helpful not only to improve management of patients but also to develop trials for future strategies. P1088 MORPHOMETRICAL QUANTIFICATION OF FIBROSIS CORRELATES WITH CLINICAL CIRRHOSIS STAGE AND PREDICTS LONG-TERM SURVIVAL IN PATIENTS WITH ALCOHOLIC CIRRHOSIS L. Binder1 , F. Rainer1 , J. Haybaeck2 , S. Xu3 , W. Spindelboeck1 , A. Wee3 , R. Stauber1 , C. Lackner2 . 1 Department of Internal Medicine, 2 Institute of Pathology, Medical University of Graz, Graz, Austria; 3 Department of Pathology, National University of Singapore, Singapore, Singapore E-mail: [email protected] Background and Aims: Histological subclassification of cirrhosis may allow for the definition of severe fibrosis stages with different clinical severity and prognosis. Recently the Laennec fibrosis scoring system was found to correlate with the clinical cirrhosis stage and grade of portal hypertension. However, the value of histological staging is limited by intra- and interobserver variability. We therefore aimed to determine the relationship between morphometrically quantified fibrosis measured as collagen proportionate area (CPA) using digital assisted image analysis and clinical stage of septal fibrosis and cirrhosis as well as long-term survival in patients with alcoholic liver disease (ALD). Methods: Liver biopsies of 149 consecutive ALD-patients, without other cause of liver disease were studied. CPA was quantified using Tissue Studio ® software (Definiens™). The data were externally validated at the National University of Singapore. In addition fibrosis was staged semiquantitively using the clinical research network (CRN) and Laennec systems. Clinical stages at the the time of liver biopsy were determined according to D‘Amico. Complete follow up and outcome data were available from all patients. Results: Mean CPA values progressively increased with Laennec fibrosis stage (stage 0: 1.4±0.8%, stage 1: 3.5±2.6%, stage 2: 8.8±6.1%, stage 3: 14.6±9.2%, stage 4: 21.1±11.8%, stage 5: 34.8±14.0%, stage 6: 48.1±12.4%; p = 0.000 by ANOVA) and with clinical D‘Amico stage (stage 1: 12.5±13.7%, stage 2: 30.0±19.8%, stage 3: 40.2±18.2%, stage 4: 35.7±18.2%; p = 0.000 by ANOVA). On ROC analysis, a CPA of 35% was identified as best cut-off for discrimination of cirrhosis. In patients with septal fibrosis or cirrhosis (CRN stage >2), Kaplan–Meier analysis revealed an association of 5-year survival to CPA (cut-off 35%, p = 0.035) but not to Laennec fibrosis stage (p = 0.294). Conclusions: These findings suggest that morphometrical subclassification yields two fibrosis stages with distinct clinical S758
outcomes. CPA measurements may favorably substitute for semiquantitative histological fibrosis staging used in ALD while avoiding intra- and interobserver biases which limit the utility of conventional histological fibrosis staging. P1089 CELLULAR SENESCENCE AND LIPOPOLYSACCHARIDE/TOLL-LIKE RECEPTOR 4 SIGNALING IN ALCOHOLIC STEATOHEPATITIS G. Alpini1 , K. McDaniel2 , Y. Han3 , H. Francis2 , F. Meng2 . 1 Medicine, 2 Scott & White Healthcare, 3 Texas A&M HSC College of Medicine, Temple, United States E-mail: [email protected] Background and Aims: Cellular senescence is a stress-responsive program limiting the proliferation of damaged cells and leading to stable cell-cycle arrest. Accumulation of senescent hepatocytes may contribute to loss of functional hepatic mass and lead to liver decompensation and fibrosis. The current study aimed to characterize the functional role of cellular senescence during alcoholic-induced hepatitis. Methods: Senescence related gene expression was assessed using a Cellular Senescence PCR Array and/or real-time PCR analysis. Cellular senescence and viability were measured by SA-b-gal Activity and MTS assay. The upstream modulators of senescence were defined in ethanol/LPS treated hepatocytes and cholangiocytes in vitro, and in ethanol fed mice and TLR-4 knockout mice in vivo. Results: We identified that 5 week of ethanol feeding significantly increased total liver histopathology score and hepatocellular senescence by PCR array and SA-b-gal assay. The up-regulation of hepatic senescence initiators, PAI-1 and EGR1, were further verified by realtime PCR assay. Treatment of N-Heps and HiBECs with ethanol (86 mM) and LPS (20 ng/ml) for 72 hr significantly increased PAI-1 and EGR1 expression, along with the enhanced SA-b-gal activity. Silencing of PAI-1 decreased ethanol and LPS-induced senescence in both N-Heps and HiBECs, whereas inhibition of EGR1 only reduced the senescence and increased viability in N-Hep cells. Interestingly, silencing of PAI-1 and EGR1 also reduced the pro-fibrotic markers, a-SMA and TIMP-1 in N-Hep cells, whereas silencing of LPS receptor TLR4 decreased these markers in the same group of N-Hep cells, suggesting LPS-mediated cellular senescence during alcoholinduced liver fibrosis. Furthermore, the expression of TLR4 and the verified LPS related senescence markers, including E2F1, ID1 and IGFBP3 were significantly altered in ethanol-fed mice liver specimens compared to controls. TLR4 knockout mice displayed less sensitivity to alcoholic injury, along with reduced PAI-1 and EGR1 levels and recovered expressions of a-SMA and TIMP-1. Conclusions: Our results show that PAI-1 and EGR1 are the critical regulator of LPS induced cellular senescence and alcoholic hepatitis. The findings provide new insight into the function of LPS regulated cellular senescence and increase opportunities for the development of novel treatment paradigms for the management of alcoholic liver diseases. P1091 ONE YEAR EFFECTIVENESS OF BACLOFEN TREATMENT IN 100 ALCOHOL-DEPENDENT PATIENTS 1 C. Barrault1 , H. Lison2 , F. Roudot-Thoraval3 , A. Garioud2 , V. Behar ´ , 4 I. Rosa-Hezode ´ , D. Belloula2 , M. Medmoun2 , C. Fourny1 , G. Pulwermacher5 , H. Hagege ` 4 , J.-F. Cadranel2 . 1 Addiction and hepatology unit, Centre hospitalier intercommunal de Cr´eteil, Cr´eteil, 2 Hepatology unit, Groupe Hospitalier Public du Sud de l’Oise, Creil, 3 Public health unit, Henri Mondor, 4 Hepatology unit, Centre hospitalier intercommunal de Cr´eteil, Cr´eteil, 5 Addiction and hepatology unit, Groupe Hospitalier Public du Sud de l’Oise, Creil, France E-mail: [email protected] Background and Aims: Several studies have suggested efficacy of Baclofen (BAC) at low or high dose in reducing alcohol consumption. Since March 2014, Temporary Recommendation for Use of BAC has
Journal of Hepatology 2015 vol. 62 | S263–S864
POSTERS been allowed by the French drug agency (ANSM) in this indication. The aim of the study was to assess effectiveness and safety of BAC at 12 months in alcohol-dependent patients with or without liver cirrhosis. Methods: Between June 2010 and September 2013, 100 consecutive patients from 2 liver and alcohology units were included in this prospective open label study. Patients provided written consent before treatment initiation. BAC was orally administered at a dose of 15 mg/day and weekly increased until alcohol indifference was obtained. The treatment was associated to social-psychological support and medical care. Results: BAC was started in 100 patients (75 males, mean age 53±9 years): 65 were cirrhotic and 16 had a chronic pancreatitis. After 1 year, 86 patients were still involved in the follow up, 83 were treated with BAC, 9 were lost of follow-up, 4 were dead and 1 had been transplanted. At a mean BAC dosage of 40 mg/day [30– 210], mean daily alcohol consumption (DAC) was reduced from 106 to 18 g/day (p < 0.001). A decrease of the DAC >50% was observed in 77 patients. Among them, a “low consumption group” of 64 patients was identified: 44 were completely abstinent and 20 drunk less than 30 g/day. No predictive factor of response was identified. In this group, a significant improvement of consumption biomarkers was observed: decrease of mean gGT activity from 4.8N to 2N (p < 0.001), mean ASAT activity from 2.6N to 1.1N (p < 0.001) and mean erythrocyte globular volume from 100.6 to 92.8 m3 and increase of mean platelets count from 171000 to 193000/mm3 (p = 0.032). In the 39 cirrhotic patients of the “low consumption group”, total bilirubin serum concentration significantly decreased from 34.2 to 19.5 mmol/L (p = 0.026), prothrombin time increased from 69 to 77% (p < 0.001) and albuminemia increased from 34.2 to 37.2 g/L (p = 0.07). Twenty patients (20%) reported minor side effects leading to a treatment withdrawal in 2 cases. No liver or renal function deterioration occurred in cirrhotic patients. Conclusions: In our cohort, baclofen treatment associated to a global care led to a dramatic reduction of alcohol consumption. This effective treatment is well tolerated and associated with a significant improvement of consumption biomarkers and of liver function tests in cirrhotic patients. P1092 SPECIFIC-SIZED HYALURONAN FRAGMENTS DIFFERENTIALLY REGULATE LPS SIGNAL TRANSDUCTION IN KUPFFER CELLS AFTER ETHANOL FEEDING TO RATS P. Saikia1 , K.A. Pollard1 , M.R. McMullen1 , L.E. Nagy1,2 . 1 Center For Liver Disease Research, Pathobiology, Cleveland Clinic, 2 Molecular Medicine, Case Western Reserve University, Cleveland, United States E-mail: [email protected] Background and Aims: Hyaluronan (HA), a major component of extracellular matrix, is ubiquitously present in many tissues. During acute and chronic inflammation or tissue injury, reactive oxygen species and matrix metalloproteinases increase HA turnover, resulting in the local and systemic accumulation of HA fragments of different molecular weights. Depending on their size, HA fragments can have either anti-inflammatory, anti-angiogenic and immunosuppressive function or pro-inflammatory function. HA is recognized as an important damage associated pattern molecule (DAMP) that regulates innate immunity. HA has been used as an indicator of liver injury for decades; however, it is not known if HA contributes to the pathophysiology of chronic ethanol-induced liver injury. Chronic alcohol consumption is associated with an increase in the sensitivity of macrophages to signaling via TLR4. Here we have tested the hypothesis that specific-sized HA fragments differentially regulate LPS-mediated signaling via TLR4 in Kupffer cells, the resident macrophage in the liver. Methods: Primary cultures of Kupffer cells were isolated from rats chronically exposed to ethanol via the Lieber-DeCarli diet or pairfed control diets.
Results: Kupffer cells from ethanol-fed rats were more sensitive to stimulation with LPS, resulting in increased expression of mRNA for the pro-inflammatory cytokine, TNFa. Challenge of Kupffer cells with specific-sized HA fragments (from 7 kDa to 2000 kDa) had no effect on TNFa mRNA expression. However, when Kupffer cells were pre-treated with HA fragments for 5 h prior to LPS challenge for 60 min, specific-sized HA fragments differentially suppressed or stimulated TNFa mRNA expression. When Kupffer cells were pretreated with HA fragments of 35kDa, the sensitization of Kupffer cells from ethanol-fed rats to LPS returned to pair-fed values. In contrast, pre-treatment with 74kDa HA had a synergistic effect in production of TNFa, increasing expression in Kupffer cells from both ethanol- and pair-fed rats. HA fragments, of 7kDa or 2000kDa, had no effect on TNFa production. Conclusions: In summary, these data suggest that specific-sized HA fragments differentially regulate TLR4-mediated signaling in Kupffer cells and that HA 35kDa fragments may play a protective role in alcoholic liver disease by decreasing the production of inflammatory mediators by Kupffer cells. This work was supported by NIH Grant RO1-AA011975 to LEN. P1093 CHARACTERIZATION OF ACETAMINOPHEN THERAPEUTIC MISADVENTURE: A SINGLE CENTRE EXPERIENCE A. Louvet1 , C. Vanveuren1 , S. Dharancy1 , A. Cannesson-Leroy1 , F. Artru1 , G. Lassailly1 , V. Canva-Delcambre1 , P. Mathurin1 . 1 Maladies de l’appareil digestif, Hˆ opital Huriez, Lille, France E-mail: [email protected] Background and Aims: Therapeutic doses of acetaminophen may provoke acute liver failure, the so-called “therapeutic misadventure with acetaminophen” but this entity has never been evaluated with a prospective design. Such study would help determining prerequisites of this condition and assessing outcome. Methods: Single-centre prospective study in which patients were included with severe acute liver failure related to acetaminophen as follows: acetaminophen overdose (AO) or intake at a therapeutic dose (≤6 g/day), referred to as acetaminophen therapeutic misadventure (ATM). Chronic drinkers were defined by an alcohol consumption ≥30 g/day. Results: From 2002 to 2014, 271 patients were included, 205 AO and 66 ATM. ATM occurred only in chronic drinkers (89.4%) or in patients who had starved for several days (10.6%). By definition, acetaminophen intake was higher in AO as compared to ATM: 16 g vs. 3.15 g (p < 0.001). In 70% of patients, intake was below the daily 4 g dose classically regarded as safe. No patient developed ATM after a single intake of acetaminophen as confirmed by a median time of intake at 4 days vs. 1 in AO (p < 0.001). At admission, patients with ATM were older than AO (age 44 vs. 30.1 years; p < 0.001) and had a deeper impairment of liver function (prothrombin time 30.7 vs. 22.1 s, p < 0.01; albumin 32.7 vs. 38.1 g/l, p < 0.001; bilirubin 47.1 vs. 25 mg/l, p < 0.001; lactate 3.31 vs. 1.95 mmol/l, p = 0.009). At admission, transaminases were not different between ATM and AO: AST 4138 vs. 3983 IU/l (p = 0.1), ALT 2416 vs. 3831 IU/l (p = 0.22). In ATM and AO, we found a predominance of females: 63.6% and 56.6% (p = 0.31) and percentage of chronic drinkers was 89.4 vs. 36.6% (p < 0.001), respectively. As expected in acetaminophen poisoning, we observed a dramatic drop after 3 days by 67% vs. 50.2% (p < 0.001) in ALT and by 93.6 vs. 89.4% (p = 0.02) in AST in ATM and AO respectively. One-month overall survival was 84.6% in ATM and 92.6% in AO (p = 0.05). On the global cohort, only the number of King’s College criteria (risk ratio 3.15, 95% CI: 2.04–4.88, p < 0.001) independently predicted 1-month mortality, whereas albumin, age and alcohol consumption did not (p = 0.13, 0.41 and 0.75 respectively). Conclusions: Therapeutic misadventure with acetaminophen is not a rare condition and occurs after several days of acetaminophen
Journal of Hepatology 2015 vol. 62 | S263–S864
S759
outcomes. CPA measurements may favorably substitute for semiquantitative histological fibrosis staging used in ALD while avoiding intra- and interobserver biases which limit the utility of conventional histological fibrosis staging. P1089 CELLULAR SENESCENCE AND LIPOPOLYSACCHARIDE/TOLL-LIKE RECEPTOR 4 SIGNALING IN ALCOHOLIC STEATOHEPATITIS G. Alpini1 , K. McDaniel2 , Y. Han3 , H. Francis2 , F. Meng2 . 1 Medicine, 2 Scott & White Healthcare, 3 Texas A&M HSC College of Medicine, Temple, United States E-mail: [email protected] Background and Aims: Cellular senescence is a stress-responsive program limiting the proliferation of damaged cells and leading to stable cell-cycle arrest. Accumulation of senescent hepatocytes may contribute to loss of functional hepatic mass and lead to liver decompensation and fibrosis. The current study aimed to characterize the functional role of cellular senescence during alcoholic-induced hepatitis. Methods: Senescence related gene expression was assessed using a Cellular Senescence PCR Array and/or real-time PCR analysis. Cellular senescence and viability were measured by SA-b-gal Activity and MTS assay. The upstream modulators of senescence were defined in ethanol/LPS treated hepatocytes and cholangiocytes in vitro, and in ethanol fed mice and TLR-4 knockout mice in vivo. Results: We identified that 5 week of ethanol feeding significantly increased total liver histopathology score and hepatocellular senescence by PCR array and SA-b-gal assay. The up-regulation of hepatic senescence initiators, PAI-1 and EGR1, were further verified by realtime PCR assay. Treatment of N-Heps and HiBECs with ethanol (86 mM) and LPS (20 ng/ml) for 72 hr significantly increased PAI-1 and EGR1 expression, along with the enhanced SA-b-gal activity. Silencing of PAI-1 decreased ethanol and LPS-induced senescence in both N-Heps and HiBECs, whereas inhibition of EGR1 only reduced the senescence and increased viability in N-Hep cells. Interestingly, silencing of PAI-1 and EGR1 also reduced the pro-fibrotic markers, a-SMA and TIMP-1 in N-Hep cells, whereas silencing of LPS receptor TLR4 decreased these markers in the same group of N-Hep cells, suggesting LPS-mediated cellular senescence during alcoholinduced liver fibrosis. Furthermore, the expression of TLR4 and the verified LPS related senescence markers, including E2F1, ID1 and IGFBP3 were significantly altered in ethanol-fed mice liver specimens compared to controls. TLR4 knockout mice displayed less sensitivity to alcoholic injury, along with reduced PAI-1 and EGR1 levels and recovered expressions of a-SMA and TIMP-1. Conclusions: Our results show that PAI-1 and EGR1 are the critical regulator of LPS induced cellular senescence and alcoholic hepatitis. The findings provide new insight into the function of LPS regulated cellular senescence and increase opportunities for the development of novel treatment paradigms for the management of alcoholic liver diseases. P1091 ONE YEAR EFFECTIVENESS OF BACLOFEN TREATMENT IN 100 ALCOHOL-DEPENDENT PATIENTS 1 C. Barrault1 , H. Lison2 , F. Roudot-Thoraval3 , A. Garioud2 , V. Behar ´ , 4 I. Rosa-Hezode ´ , D. Belloula2 , M. Medmoun2 , C. Fourny1 , G. Pulwermacher5 , H. Hagege ` 4 , J.-F. Cadranel2 . 1 Addiction and hepatology unit, Centre hospitalier intercommunal de Cr´eteil, Cr´eteil, 2 Hepatology unit, Groupe Hospitalier Public du Sud de l’Oise, Creil, 3 Public health unit, Henri Mondor, 4 Hepatology unit, Centre hospitalier intercommunal de Cr´eteil, Cr´eteil, 5 Addiction and hepatology unit, Groupe Hospitalier Public du Sud de l’Oise, Creil, France E-mail: [email protected] Background and Aims: Several studies have suggested efficacy of Baclofen (BAC) at low or high dose in reducing alcohol consumption. Since March 2014, Temporary Recommendation for Use of BAC has
Journal of Hepatology 2015 vol. 62 | S263–S864
POSTERS been allowed by the French drug agency (ANSM) in this indication. The aim of the study was to assess effectiveness and safety of BAC at 12 months in alcohol-dependent patients with or without liver cirrhosis. Methods: Between June 2010 and September 2013, 100 consecutive patients from 2 liver and alcohology units were included in this prospective open label study. Patients provided written consent before treatment initiation. BAC was orally administered at a dose of 15 mg/day and weekly increased until alcohol indifference was obtained. The treatment was associated to social-psychological support and medical care. Results: BAC was started in 100 patients (75 males, mean age 53±9 years): 65 were cirrhotic and 16 had a chronic pancreatitis. After 1 year, 86 patients were still involved in the follow up, 83 were treated with BAC, 9 were lost of follow-up, 4 were dead and 1 had been transplanted. At a mean BAC dosage of 40 mg/day [30– 210], mean daily alcohol consumption (DAC) was reduced from 106 to 18 g/day (p < 0.001). A decrease of the DAC >50% was observed in 77 patients. Among them, a “low consumption group” of 64 patients was identified: 44 were completely abstinent and 20 drunk less than 30 g/day. No predictive factor of response was identified. In this group, a significant improvement of consumption biomarkers was observed: decrease of mean gGT activity from 4.8N to 2N (p < 0.001), mean ASAT activity from 2.6N to 1.1N (p < 0.001) and mean erythrocyte globular volume from 100.6 to 92.8 m3 and increase of mean platelets count from 171000 to 193000/mm3 (p = 0.032). In the 39 cirrhotic patients of the “low consumption group”, total bilirubin serum concentration significantly decreased from 34.2 to 19.5 mmol/L (p = 0.026), prothrombin time increased from 69 to 77% (p < 0.001) and albuminemia increased from 34.2 to 37.2 g/L (p = 0.07). Twenty patients (20%) reported minor side effects leading to a treatment withdrawal in 2 cases. No liver or renal function deterioration occurred in cirrhotic patients. Conclusions: In our cohort, baclofen treatment associated to a global care led to a dramatic reduction of alcohol consumption. This effective treatment is well tolerated and associated with a significant improvement of consumption biomarkers and of liver function tests in cirrhotic patients. P1092 SPECIFIC-SIZED HYALURONAN FRAGMENTS DIFFERENTIALLY REGULATE LPS SIGNAL TRANSDUCTION IN KUPFFER CELLS AFTER ETHANOL FEEDING TO RATS P. Saikia1 , K.A. Pollard1 , M.R. McMullen1 , L.E. Nagy1,2 . 1 Center For Liver Disease Research, Pathobiology, Cleveland Clinic, 2 Molecular Medicine, Case Western Reserve University, Cleveland, United States E-mail: [email protected] Background and Aims: Hyaluronan (HA), a major component of extracellular matrix, is ubiquitously present in many tissues. During acute and chronic inflammation or tissue injury, reactive oxygen species and matrix metalloproteinases increase HA turnover, resulting in the local and systemic accumulation of HA fragments of different molecular weights. Depending on their size, HA fragments can have either anti-inflammatory, anti-angiogenic and immunosuppressive function or pro-inflammatory function. HA is recognized as an important damage associated pattern molecule (DAMP) that regulates innate immunity. HA has been used as an indicator of liver injury for decades; however, it is not known if HA contributes to the pathophysiology of chronic ethanol-induced liver injury. Chronic alcohol consumption is associated with an increase in the sensitivity of macrophages to signaling via TLR4. Here we have tested the hypothesis that specific-sized HA fragments differentially regulate LPS-mediated signaling via TLR4 in Kupffer cells, the resident macrophage in the liver. Methods: Primary cultures of Kupffer cells were isolated from rats chronically exposed to ethanol via the Lieber-DeCarli diet or pairfed control diets.
Results: Kupffer cells from ethanol-fed rats were more sensitive to stimulation with LPS, resulting in increased expression of mRNA for the pro-inflammatory cytokine, TNFa. Challenge of Kupffer cells with specific-sized HA fragments (from 7 kDa to 2000 kDa) had no effect on TNFa mRNA expression. However, when Kupffer cells were pre-treated with HA fragments for 5 h prior to LPS challenge for 60 min, specific-sized HA fragments differentially suppressed or stimulated TNFa mRNA expression. When Kupffer cells were pretreated with HA fragments of 35kDa, the sensitization of Kupffer cells from ethanol-fed rats to LPS returned to pair-fed values. In contrast, pre-treatment with 74kDa HA had a synergistic effect in production of TNFa, increasing expression in Kupffer cells from both ethanol- and pair-fed rats. HA fragments, of 7kDa or 2000kDa, had no effect on TNFa production. Conclusions: In summary, these data suggest that specific-sized HA fragments differentially regulate TLR4-mediated signaling in Kupffer cells and that HA 35kDa fragments may play a protective role in alcoholic liver disease by decreasing the production of inflammatory mediators by Kupffer cells. This work was supported by NIH Grant RO1-AA011975 to LEN. P1093 CHARACTERIZATION OF ACETAMINOPHEN THERAPEUTIC MISADVENTURE: A SINGLE CENTRE EXPERIENCE A. Louvet1 , C. Vanveuren1 , S. Dharancy1 , A. Cannesson-Leroy1 , F. Artru1 , G. Lassailly1 , V. Canva-Delcambre1 , P. Mathurin1 . 1 Maladies de l’appareil digestif, Hˆ opital Huriez, Lille, France E-mail: [email protected] Background and Aims: Therapeutic doses of acetaminophen may provoke acute liver failure, the so-called “therapeutic misadventure with acetaminophen” but this entity has never been evaluated with a prospective design. Such study would help determining prerequisites of this condition and assessing outcome. Methods: Single-centre prospective study in which patients were included with severe acute liver failure related to acetaminophen as follows: acetaminophen overdose (AO) or intake at a therapeutic dose (≤6 g/day), referred to as acetaminophen therapeutic misadventure (ATM). Chronic drinkers were defined by an alcohol consumption ≥30 g/day. Results: From 2002 to 2014, 271 patients were included, 205 AO and 66 ATM. ATM occurred only in chronic drinkers (89.4%) or in patients who had starved for several days (10.6%). By definition, acetaminophen intake was higher in AO as compared to ATM: 16 g vs. 3.15 g (p < 0.001). In 70% of patients, intake was below the daily 4 g dose classically regarded as safe. No patient developed ATM after a single intake of acetaminophen as confirmed by a median time of intake at 4 days vs. 1 in AO (p < 0.001). At admission, patients with ATM were older than AO (age 44 vs. 30.1 years; p < 0.001) and had a deeper impairment of liver function (prothrombin time 30.7 vs. 22.1 s, p < 0.01; albumin 32.7 vs. 38.1 g/l, p < 0.001; bilirubin 47.1 vs. 25 mg/l, p < 0.001; lactate 3.31 vs. 1.95 mmol/l, p = 0.009). At admission, transaminases were not different between ATM and AO: AST 4138 vs. 3983 IU/l (p = 0.1), ALT 2416 vs. 3831 IU/l (p = 0.22). In ATM and AO, we found a predominance of females: 63.6% and 56.6% (p = 0.31) and percentage of chronic drinkers was 89.4 vs. 36.6% (p < 0.001), respectively. As expected in acetaminophen poisoning, we observed a dramatic drop after 3 days by 67% vs. 50.2% (p < 0.001) in ALT and by 93.6 vs. 89.4% (p = 0.02) in AST in ATM and AO respectively. One-month overall survival was 84.6% in ATM and 92.6% in AO (p = 0.05). On the global cohort, only the number of King’s College criteria (risk ratio 3.15, 95% CI: 2.04–4.88, p < 0.001) independently predicted 1-month mortality, whereas albumin, age and alcohol consumption did not (p = 0.13, 0.41 and 0.75 respectively). Conclusions: Therapeutic misadventure with acetaminophen is not a rare condition and occurs after several days of acetaminophen
Journal of Hepatology 2015 vol. 62 | S263–S864
S759