- Email: [email protected]
Prevalence of Bariatric Surgery in Patients with Severe Alcoholic Hepatitis Undergoing Evaluation for Early Liver Transplantation
significantly blocked the LPS-induced mRNA expression of pro-inflammatory cytokines IL6, IL-1β and TNFα in a dose-dependent manner. Conclusions: CRAMP deficiency exacerbates ALD through multiple mechanisms, including dysbiosis and endotoxemia and increases hepatic inflammation. Synthetic CRAMP peptide reduces alcohol exposure-induced liver steatosis and injury. CRAMP pathway may represent a therapeutic approach for ALD.
Sa1578 CELLULAR BIOENERGETICS: PERSONALIZING TREATMENT IN ALCOHOLIC LIVER DISEASE Ashwani K. Singal, Balu Chacko, Sumant Arora, Degui Zhi, Victor Darley-Usmar Background and Aims: Alcoholic hepatitis (AH) is associated with 40-50% mortality at 1 month. Corticosteroids (CS) provide 50% survival benefit with their response evaluable only at 1 week. Defects in bioenergetics or mitochondrial oxygen consumption rate (OCR) in peripheral cells are shown in diseases associated with systemic inflammation like diabetes and sepsis. We tested the hypothesis that AH patients with severe bioenergetics defects will progress to liver failure and be non-responsive to CS (NRS). Methods: After informed consent, 20 mL blood was collected from ALD patients (with or without AH) and healthy controls. A second 20 mL sample was collected at 1 wk. from AH patients receiving CS. Monocytes and neutrophils were isolated within 30 min using CD14 and CD15 antibodies respectively. Cellular bioenergetics and OCR (pmol/min./mcg protein) were obtained using XF96 analyzer (Seahorse Biosciences, Figure). Results: 78 ALD patients without concomitant hepatitis C (37 with AH) and 40 healthy controls were recruited. ALD with AH vs. without AH were younger (44±11 vs. 53±9 yrs.) with higher white blood cell count (16±10 vs. 6±9) and MELD score (30±10 vs. 16±10), P<0.0001 for all. Compared to healthy controls, OCR differed among ALD patients for basal (3.1±1.6 vs. 2.5±0.8, P=0.002), proton leak (0.6±0.4 vs. 0.5±0.2, P=0.03), non-mitochondrial (1.5±0.6 vs. 1.4±0.4, P=0.03), oxidative burst in monocytes (8.5±5.2 vs. 6.3±3.7, P<0.05) and neutrophils (44±13 vs. 53±14, P=0.01). OCR among ALD with AH vs. without AH differed for basal (2.5±1.3 vs. 3.5±1.8, P=0.02), proton leak (0.4±0.3 vs. 0.7±0.5, P=0.02), and neutrophilic oxidative burst (44±13 vs. 51±12, P<0.05). Of 18 AH patients treated with CS, baseline neutrophilic burst was higher different among 12 responders to compared to 6 NRS (47±13 vs. 35±7, P=0.002). Monocyte bioenergetics and neutrophilic oxidative burst at 1 week of CS treatment, available in 8 patients (5 responders) showed improvement in all components of bioenergetics among 5 responders compared to 3 NRS. Conclusions: Baseline cellular bioenergetics seems a promising biomarker for personalized medicine in ALD patients for predicting response to CS and outcome on follow-up. Data in larger multicenter population are needed before accepting the use of this novel biomarker in clinical practice.
Sa1576
AASLD Abstracts
LEVELS OF CIRCULATING FOLLICULAR HELPER T CELLS AND THE PROGNOSTIC SIGNIFICANCE OF CD40 LIGAND ON SURVIVAL IN PATIENTS WITH ALCOHOLIC LIVER DISEASE Suthat Liangpunsakul, Kristin Hollister, Ruth Ann Ross, Laura Heathers, Kristina Chandler, Alexander Dent BACKGROUND: Excessive alcohol drinking (ED) is strongly associated with immune dysfunction. ED and patients with alcoholic liver disease (ALD) are several times more susceptible to bacterial infections and have a decrease in antibody responses to vaccinations. Follicular helper T cells (TFH) are a CD4 T cell lineage uniquely found in the germinal center. The specific function is to select B cells in the germinal center that produce high-affinity antibodies, the process generally requires the CD40 ligand (CD40L). We hypothesized that ED interferes with TFH cell differentiation, leading to the alteration in the circulating level of these cells and their ligand, CD40L. AIMS: (1) To determine the levels of TFH cells in ED and those with ALD when compared to healthy controls and (2) To determine the prognostic significance of serum CD40L in a cohort of patients with ALD. METHODS: ED is defined as men who drink > 4 standard drinks in a day (or > 14 per week) and women who drink >3 drinks in a day (or > 7 per week), per NIAAA definition. Subjects with ALD were those with clinical/radiographic evidence of cirrhosis with the history of alcohol consumption averaging at least 4 drinks/day (for men) or 3 drinks/day (for women), for at least 10 years. Baseline demographic and laboratory tests were obtained. Peripheral blood mononuclear cells were isolated from healthy controls(n=4), ED(n=20), and ALD(n=8) and subjected for flow cytometry for TFH (CXCR5+CXCR3+PD1+)cells. Serum levels of CD40L were determined in controls (n=30, age32±9yrs, 90%male, 83% White), ED (n=30, age38±12yrs, 63%male, 83% White), and ALD (n=83, age50±9yrs,77%male,92%White)using ELISA (R&D System). Subjects with ALD were prospectively followed from the time of enrollment till the study end date (July 1, 2016) or death. The prognostic significance of serum CD40L in ALD patients was determined using Cox Hazard model controlling for covariates. RESULTS: Subjects with ALD had average MELD scores of 11. 58 were Child A and 25 were Child B/C. The level of TFH (CXCR5+CXCR3+PD1+)cells was significantly lower in peripheral blood of subjects with ED and ALD by ~ 5 fold compared to controls (p<0.05). The serum level of CD40L was significantly lower in ED (511±302pg/ml) and ALD (452±142pg/ml) when compared to that in controls (915±579pg/ml, p<0.0005). Its level was significantly lower in Child B/ C ALD patients (336±72pg/ml) compared to those in Child A (502±135pg/ml). During the follow up (median follow up of 3.2 yrs), 15 (18%) patients with ALD died. In the multivariate Cox model, MELD scores (HR 1.14,95%CI 1.016-1.221,p=0.02) and the level of CD40L (HR 0.98, 95%CI 0.97-0.98,p=0.001) were the independent predictors of mortality. CONCLUSIONS: Patients with ALD had significantly lower level of circulating TFH and CD40L. The level of CD40L was an independent predictor of mortality in patients with ALD.
Sa1579 ACUTE KIDNEY INJURY IN PATIENTS WITH ALCOHOLIC HEPATITIS: A MULTICENTER STUDY ON 773 PATIENTS Sujan Ravi, Vivek Desai, Douglas Simonetto, Joaquin Cabezas, Rakhi Maiwall, Page D. Axley, Shiv Kumar Sarin, Jose Altamirano, Ramon Bataller, Vijay Shah, Ashwani K. Singal Background and Aims:Alcoholic hepatitis (AH) is a clinical syndrome with high mortality. Acute kidney injury (AKI) occurs frequently in these patients and is associated with poor prognosis. We performed this study to evaluate the occurrence of AKI, its predictors and impact on patient outcomes in hospitalized patients with AH. Methods: AH was defined in patients with heavy alcohol use (>40g/day in women & 60 g/day in men)for more than 6 months, with less than 60 days of abstinence, bilirubin > 3mg/dL, onset of jaundice within 8 weeks, AST >50IU L, AST:ALT ratio >1.5 and both values <400 IU/L. Data was collected from patients from four tertiary care centers, including patient demographics, SIRS criteria, presence of cirrhosis, laboratory values, occurrence of infection, ICU care, 90-d mortality, and length of stay (LOS). Results: 773 patients (median age 47 yrs., 82%males, 46% Caucasian and 64% cirrhosis) were analyzed: 117 from center A (median age 47 yrs., 66%males, 80% white, 52% cirrhosis), 365 from center B (median age 45 yrs., 98%males, 100% Asian, 85 % cirrhosis), 224 from center C (median age 48 yrs., 67%males, 95% white, 35% cirrhosis), and 67 from center D (median age 51 yrs., 73%males, 76% white, 61% cirrhosis). AKI during hospital stay occurred in 32% of patients (35%, 33% 30% and 27% at centers A, B, C, and D respectively). Among 372 (48%) patients with SIRS, AKI was more frequent among those with infection (44% vs. 34%, p <0.001). Patients developing AKI had more frequent encephalopathy (HE), SIRS, tachycardia, and had higher MELD score, ABIC score, and mDF score, P <0.001 for all analyses. Variables at admission predicting in-hospital AKI were MELD score [1.08 (1.06-1.11)], HE [2.37 (1.57-3.56)], and SIRS [1.87 (1.28-2.75)]. Among components of SIRS, tachycardia [1.64 (1.12-2.41)] and WBC count [1.67 (1.11-2.52)] predicted AKI. 194 (25%) patients died within 90 days of index hospitalization. Patients with AKI had higher 90-d mortality (47% vs. 23%,), need for ICU stay (57% vs. 15%,), mean LOS (14.8 vs. 8.5 d) and infection (45% vs. 19%); P<0.001 for all. On cox-regression analysis, in-hospital AKI increased risk for 90-d mortality by 38%. SIRS and infection were other predictors of 90-d mortality after adjusting for age, sex and MELD score. 48 (6.2%) patients required renal replacement therapy (RRT), with no impact of RRT on 90-d mortality (59% vs. 56%, log rank p= 0.11). Conclusion: AKI occurs in about a third of AH patients during hospitalization. Presence of SIRS, specifically WBC count and tachycardia at admission, predict occurrence of AKI. AKI negatively impacts patient outcomes with higher short-term mortality and increased utilization of resources. Further studies are needed to develop measures to prevent occurrence of AKI among hospitalized patients with AH.
Sa1577 PREVALENCE OF BARIATRIC SURGERY IN PATIENTS WITH SEVERE ALCOHOLIC HEPATITIS UNDERGOING EVALUATION FOR EARLY LIVER TRANSPLANTATION Meera P. Bhardwaj, Jonathan Nahas, Gene Y. Im Background: After bariatric surgery, patients experience accelerated alcohol absorption and are at increased risk for alcohol use disorder (AUD). Anecdotally, there has been an increase in bariatric surgery patients presenting with severe alcoholic hepatitis (AH) at our center. The purpose of this study was to identify the prevalence and outcomes of patients with prior bariatric surgery undergoing evaluation for early liver transplantation (LT) for severe AH. Methods: Using a prospectively maintained database from January 2012 to October 2016, consecutively hospitalized patients with AH at our center were retrospectively examined. Patient demographics, clinical and psychosocial characteristics, laboratory and radiological data were analyzed. Results: In a nearly 5-year study period, a total of 213 patients hospitalized with AH were identified. The prevalence of bariatric surgery was 8.5% (n=18), with 17 having a history of Roux-en-Y gastric bypass and 1 with lap band prior to presentation. Overall, excluding the bariatric cohort, the median age was 48 years and a slight male predominance, with high-risk of mortality demonstrated by median discriminant function (DF), Lille, and MELD scores of 72, 0.98, and 31, respectively. In the bariatric cohort, the median age was 44 years and all were women (p<0.01), with median DF, Lille, and MELD scores of 61, 0.927, and 27 respectively. Slightly more than half (10/18) of the bariatric cohort received AH-specific treatment, but were universally nonresponders complicated by infection (11/18, 61%) and gastrointestinal bleeding (4/18, 22%). Hepatorenal syndrome occurred in 55%, with 6 (33%) patients requiring hemodialysis. More than half of the cohort also required ICU admission (10/18). Two-thirds of the cohort presented with their first liver-decompensating event. The median time from bariatric surgery to presentation at our center was 10 years. The median time of alcohol use prior to presentation was 10 years, and 4 patients developed new use after surgery. The 6-month mortality rate in the bariatric cohort was 44% (8/18), compared to 43% (n=83/195) in the remainder of the study group. Four bariatric patients were lost to follow-up. Only 2 bariatric patients had favorable psychosocial profiles acceptable for early LT. Of these, 1 underwent early LT during the index admission, and the other improved significantly without LT. Conclusions: The prevalence of bariatric surgery is a higher-than-expected 8.5% in a single-center cohort of severe AH patients undergoing evaluation for early LT. This adds further evidence to the negative impacts of alcohol on patients after bariatric surgery.
AASLD Abstracts
S-1112
Sa1578 CELLULAR BIOENERGETICS: PERSONALIZING TREATMENT IN ALCOHOLIC LIVER DISEASE Ashwani K. Singal, Balu Chacko, Sumant Arora, Degui Zhi, Victor Darley-Usmar Background and Aims: Alcoholic hepatitis (AH) is associated with 40-50% mortality at 1 month. Corticosteroids (CS) provide 50% survival benefit with their response evaluable only at 1 week. Defects in bioenergetics or mitochondrial oxygen consumption rate (OCR) in peripheral cells are shown in diseases associated with systemic inflammation like diabetes and sepsis. We tested the hypothesis that AH patients with severe bioenergetics defects will progress to liver failure and be non-responsive to CS (NRS). Methods: After informed consent, 20 mL blood was collected from ALD patients (with or without AH) and healthy controls. A second 20 mL sample was collected at 1 wk. from AH patients receiving CS. Monocytes and neutrophils were isolated within 30 min using CD14 and CD15 antibodies respectively. Cellular bioenergetics and OCR (pmol/min./mcg protein) were obtained using XF96 analyzer (Seahorse Biosciences, Figure). Results: 78 ALD patients without concomitant hepatitis C (37 with AH) and 40 healthy controls were recruited. ALD with AH vs. without AH were younger (44±11 vs. 53±9 yrs.) with higher white blood cell count (16±10 vs. 6±9) and MELD score (30±10 vs. 16±10), P<0.0001 for all. Compared to healthy controls, OCR differed among ALD patients for basal (3.1±1.6 vs. 2.5±0.8, P=0.002), proton leak (0.6±0.4 vs. 0.5±0.2, P=0.03), non-mitochondrial (1.5±0.6 vs. 1.4±0.4, P=0.03), oxidative burst in monocytes (8.5±5.2 vs. 6.3±3.7, P<0.05) and neutrophils (44±13 vs. 53±14, P=0.01). OCR among ALD with AH vs. without AH differed for basal (2.5±1.3 vs. 3.5±1.8, P=0.02), proton leak (0.4±0.3 vs. 0.7±0.5, P=0.02), and neutrophilic oxidative burst (44±13 vs. 51±12, P<0.05). Of 18 AH patients treated with CS, baseline neutrophilic burst was higher different among 12 responders to compared to 6 NRS (47±13 vs. 35±7, P=0.002). Monocyte bioenergetics and neutrophilic oxidative burst at 1 week of CS treatment, available in 8 patients (5 responders) showed improvement in all components of bioenergetics among 5 responders compared to 3 NRS. Conclusions: Baseline cellular bioenergetics seems a promising biomarker for personalized medicine in ALD patients for predicting response to CS and outcome on follow-up. Data in larger multicenter population are needed before accepting the use of this novel biomarker in clinical practice.
Sa1576
AASLD Abstracts
LEVELS OF CIRCULATING FOLLICULAR HELPER T CELLS AND THE PROGNOSTIC SIGNIFICANCE OF CD40 LIGAND ON SURVIVAL IN PATIENTS WITH ALCOHOLIC LIVER DISEASE Suthat Liangpunsakul, Kristin Hollister, Ruth Ann Ross, Laura Heathers, Kristina Chandler, Alexander Dent BACKGROUND: Excessive alcohol drinking (ED) is strongly associated with immune dysfunction. ED and patients with alcoholic liver disease (ALD) are several times more susceptible to bacterial infections and have a decrease in antibody responses to vaccinations. Follicular helper T cells (TFH) are a CD4 T cell lineage uniquely found in the germinal center. The specific function is to select B cells in the germinal center that produce high-affinity antibodies, the process generally requires the CD40 ligand (CD40L). We hypothesized that ED interferes with TFH cell differentiation, leading to the alteration in the circulating level of these cells and their ligand, CD40L. AIMS: (1) To determine the levels of TFH cells in ED and those with ALD when compared to healthy controls and (2) To determine the prognostic significance of serum CD40L in a cohort of patients with ALD. METHODS: ED is defined as men who drink > 4 standard drinks in a day (or > 14 per week) and women who drink >3 drinks in a day (or > 7 per week), per NIAAA definition. Subjects with ALD were those with clinical/radiographic evidence of cirrhosis with the history of alcohol consumption averaging at least 4 drinks/day (for men) or 3 drinks/day (for women), for at least 10 years. Baseline demographic and laboratory tests were obtained. Peripheral blood mononuclear cells were isolated from healthy controls(n=4), ED(n=20), and ALD(n=8) and subjected for flow cytometry for TFH (CXCR5+CXCR3+PD1+)cells. Serum levels of CD40L were determined in controls (n=30, age32±9yrs, 90%male, 83% White), ED (n=30, age38±12yrs, 63%male, 83% White), and ALD (n=83, age50±9yrs,77%male,92%White)using ELISA (R&D System). Subjects with ALD were prospectively followed from the time of enrollment till the study end date (July 1, 2016) or death. The prognostic significance of serum CD40L in ALD patients was determined using Cox Hazard model controlling for covariates. RESULTS: Subjects with ALD had average MELD scores of 11. 58 were Child A and 25 were Child B/C. The level of TFH (CXCR5+CXCR3+PD1+)cells was significantly lower in peripheral blood of subjects with ED and ALD by ~ 5 fold compared to controls (p<0.05). The serum level of CD40L was significantly lower in ED (511±302pg/ml) and ALD (452±142pg/ml) when compared to that in controls (915±579pg/ml, p<0.0005). Its level was significantly lower in Child B/ C ALD patients (336±72pg/ml) compared to those in Child A (502±135pg/ml). During the follow up (median follow up of 3.2 yrs), 15 (18%) patients with ALD died. In the multivariate Cox model, MELD scores (HR 1.14,95%CI 1.016-1.221,p=0.02) and the level of CD40L (HR 0.98, 95%CI 0.97-0.98,p=0.001) were the independent predictors of mortality. CONCLUSIONS: Patients with ALD had significantly lower level of circulating TFH and CD40L. The level of CD40L was an independent predictor of mortality in patients with ALD.
Sa1579 ACUTE KIDNEY INJURY IN PATIENTS WITH ALCOHOLIC HEPATITIS: A MULTICENTER STUDY ON 773 PATIENTS Sujan Ravi, Vivek Desai, Douglas Simonetto, Joaquin Cabezas, Rakhi Maiwall, Page D. Axley, Shiv Kumar Sarin, Jose Altamirano, Ramon Bataller, Vijay Shah, Ashwani K. Singal Background and Aims:Alcoholic hepatitis (AH) is a clinical syndrome with high mortality. Acute kidney injury (AKI) occurs frequently in these patients and is associated with poor prognosis. We performed this study to evaluate the occurrence of AKI, its predictors and impact on patient outcomes in hospitalized patients with AH. Methods: AH was defined in patients with heavy alcohol use (>40g/day in women & 60 g/day in men)for more than 6 months, with less than 60 days of abstinence, bilirubin > 3mg/dL, onset of jaundice within 8 weeks, AST >50IU L, AST:ALT ratio >1.5 and both values <400 IU/L. Data was collected from patients from four tertiary care centers, including patient demographics, SIRS criteria, presence of cirrhosis, laboratory values, occurrence of infection, ICU care, 90-d mortality, and length of stay (LOS). Results: 773 patients (median age 47 yrs., 82%males, 46% Caucasian and 64% cirrhosis) were analyzed: 117 from center A (median age 47 yrs., 66%males, 80% white, 52% cirrhosis), 365 from center B (median age 45 yrs., 98%males, 100% Asian, 85 % cirrhosis), 224 from center C (median age 48 yrs., 67%males, 95% white, 35% cirrhosis), and 67 from center D (median age 51 yrs., 73%males, 76% white, 61% cirrhosis). AKI during hospital stay occurred in 32% of patients (35%, 33% 30% and 27% at centers A, B, C, and D respectively). Among 372 (48%) patients with SIRS, AKI was more frequent among those with infection (44% vs. 34%, p <0.001). Patients developing AKI had more frequent encephalopathy (HE), SIRS, tachycardia, and had higher MELD score, ABIC score, and mDF score, P <0.001 for all analyses. Variables at admission predicting in-hospital AKI were MELD score [1.08 (1.06-1.11)], HE [2.37 (1.57-3.56)], and SIRS [1.87 (1.28-2.75)]. Among components of SIRS, tachycardia [1.64 (1.12-2.41)] and WBC count [1.67 (1.11-2.52)] predicted AKI. 194 (25%) patients died within 90 days of index hospitalization. Patients with AKI had higher 90-d mortality (47% vs. 23%,), need for ICU stay (57% vs. 15%,), mean LOS (14.8 vs. 8.5 d) and infection (45% vs. 19%); P<0.001 for all. On cox-regression analysis, in-hospital AKI increased risk for 90-d mortality by 38%. SIRS and infection were other predictors of 90-d mortality after adjusting for age, sex and MELD score. 48 (6.2%) patients required renal replacement therapy (RRT), with no impact of RRT on 90-d mortality (59% vs. 56%, log rank p= 0.11). Conclusion: AKI occurs in about a third of AH patients during hospitalization. Presence of SIRS, specifically WBC count and tachycardia at admission, predict occurrence of AKI. AKI negatively impacts patient outcomes with higher short-term mortality and increased utilization of resources. Further studies are needed to develop measures to prevent occurrence of AKI among hospitalized patients with AH.
Sa1577 PREVALENCE OF BARIATRIC SURGERY IN PATIENTS WITH SEVERE ALCOHOLIC HEPATITIS UNDERGOING EVALUATION FOR EARLY LIVER TRANSPLANTATION Meera P. Bhardwaj, Jonathan Nahas, Gene Y. Im Background: After bariatric surgery, patients experience accelerated alcohol absorption and are at increased risk for alcohol use disorder (AUD). Anecdotally, there has been an increase in bariatric surgery patients presenting with severe alcoholic hepatitis (AH) at our center. The purpose of this study was to identify the prevalence and outcomes of patients with prior bariatric surgery undergoing evaluation for early liver transplantation (LT) for severe AH. Methods: Using a prospectively maintained database from January 2012 to October 2016, consecutively hospitalized patients with AH at our center were retrospectively examined. Patient demographics, clinical and psychosocial characteristics, laboratory and radiological data were analyzed. Results: In a nearly 5-year study period, a total of 213 patients hospitalized with AH were identified. The prevalence of bariatric surgery was 8.5% (n=18), with 17 having a history of Roux-en-Y gastric bypass and 1 with lap band prior to presentation. Overall, excluding the bariatric cohort, the median age was 48 years and a slight male predominance, with high-risk of mortality demonstrated by median discriminant function (DF), Lille, and MELD scores of 72, 0.98, and 31, respectively. In the bariatric cohort, the median age was 44 years and all were women (p<0.01), with median DF, Lille, and MELD scores of 61, 0.927, and 27 respectively. Slightly more than half (10/18) of the bariatric cohort received AH-specific treatment, but were universally nonresponders complicated by infection (11/18, 61%) and gastrointestinal bleeding (4/18, 22%). Hepatorenal syndrome occurred in 55%, with 6 (33%) patients requiring hemodialysis. More than half of the cohort also required ICU admission (10/18). Two-thirds of the cohort presented with their first liver-decompensating event. The median time from bariatric surgery to presentation at our center was 10 years. The median time of alcohol use prior to presentation was 10 years, and 4 patients developed new use after surgery. The 6-month mortality rate in the bariatric cohort was 44% (8/18), compared to 43% (n=83/195) in the remainder of the study group. Four bariatric patients were lost to follow-up. Only 2 bariatric patients had favorable psychosocial profiles acceptable for early LT. Of these, 1 underwent early LT during the index admission, and the other improved significantly without LT. Conclusions: The prevalence of bariatric surgery is a higher-than-expected 8.5% in a single-center cohort of severe AH patients undergoing evaluation for early LT. This adds further evidence to the negative impacts of alcohol on patients after bariatric surgery.
AASLD Abstracts
S-1112