AASLD Abstracts Risk factors of hepatobiliary dysfunction IVLE, intravenous lipid emulsion Sa1569 HERBAL AND DIETARY SUPPLEMENT USE AMONG U. S. ADULTS WITH CHRONIC LIVER DISEASE Matthew Zheng, Silpa Yalamanchili, Mital Shah, Sheela S. Reddy, Victor Navarro, Dina Halegoua-De Marzio Introduction Herbal and dietary supplements (HDS) are used by over 50% of Americans, but their use is not commonly reported to healthcare providers. Due to its role in drug metabolism, the liver is often the target of HDS toxicity making HDS-induced liver injury a common cause of acute liver failure in the United States. Patients with decompensated liver disease and impaired drug metabolism may be at increased risk of HDS induced hepatotoxicity. The aim of this study is to determine the prevalence of herbal and dietary supplement use among patients with chronic liver disease (CLD). Methods Data from National Health and Nutrition Examination Surveys (NHANES) from 2001 to 2011 were queried for people who self-identified as having CLD. Based on prior studies (source: https:// livertox.nih.gov/Herbals_and_Dietary_Supplements.htm), we compiled a list of 22 single and multiple ingredient supplements that have been reported to cause hepatotoxicity. Of those patients who self-identified as having liver disease, we then recorded their supplement intake, and specifically looked for any of these known hepatotoxic agents. Results Of the 55,388 individuals polled in the NHANES data, 568 (1%, 55% male, mean age 53.8 years old) of respondents reported that they had a history of CLD. Of these, 313 individuals (55%, 55% male, mean age 54.8 years old) admitted to using HDS. The most common HDS taken was vitamin C (24, 7.6%), Vitamin E (14, 4.4%), Vitamin D (13, 4.1%). Only 7 individuals (2.2%) reported taking HDS that were potentially hepatotoxic. These included 3 individuals who took supplements with saw palmetto, and 4 individuals taking supplements with ingredients that included green tea extract, Hydroxycut, Herbalife, or black cohosh. There was a trend toward increased HDS use with only 31 respondents reporting use in 2001 compared to 71 in 2011. Discussion The majority of patients with CLD are taking HDS with trends similar to the general population. However, only a small minority of patients with liver disease are taking supplements with known hepatotoxicity. Further research needs to be done to elucidate the risks and reasons for supplement use in patients with underlying liver disease. Moreover, our study also demonstrates the need for better counseling of patients with liver disease about the dangers of herbal and dietary supplementa- tion. Sa1570 LIVER INJURY IN CHILDREN UNDERGOING TREATMENT FOR CANCER: ANALYSIS OF A LARGE SINGLE-CENTER COHORT Vanessa Cardenas, Nikhil P. Mankuzhy, Lili Zhao, Bailey Anderson, Kevin Frank, Maclovio J. Lopez, Robert J. Fontana, Rajen Mody, Frank DiPaola Introduction: Drug-induced liver injury (DILI) is an important cause of morbidity and mortality in both adults and children. Conventional agents used to treat pediatric cancers carry risk for liver toxicity, and there is growing interest in new targeted agents, many of which also appear to be hepatotoxic in pediatric clinical trials. We analyzed a large pediatric oncology database to retrospectively identify all cases of liver injury, and to describe the presenting features. Our aim is to identify etiologies over liver injury, including drug-induced liver injury and to characterize associations and risk factors. Methods: This study was IRB approved. We retrospectively reviewed the University of Michigan Pediatric Hematology Oncology (PHO) REDcap database from 01/01/2004 to 07/31/2016. Inclusion criteria were age ≤ 25 years plus any one of the following: ALT or AST > 5xULN or alkaline phosphatase (AP) > 2xULN on two consecutive lab draws within two months, or total bilirubin ≥ 2.5 mg/dl plus any elevation of AST, ALT or AP > ULN on any single lab draw. We excluded liver injury after solid organ or bone marrow transplantation. EMR data mining software (DataDirect) was utilized to identify cases; for each case, pattern of liver injury (hepatocellular, cholestatic or mixed) was assessed. Data for each case was extracted including age, cancer diagnosis, liver-related morbidity, and mortality. Results: There are 1916 patients in the REDcap database. 307 (16%) of the enrolled patients had at least one episode of liver injury. There was no difference in prevalence of liver injury between males and females (16% and 17%; p = 0.606). A majority of the cases occurred in young children (< 1 year = 5%, 1-10 years = 51%, 11-18 years = 39%, 19-25 years = 5%). The most common liver injury pattern was hepatocellular (74%), followed by cholestatic (14%) and mixed (12%) (Table S-1110 AASLD Abstracts 1). Regarding cancer group, leukemia appeared to be overrepresented versus solid tumors among cases (56% vs 44%) as compared to the entire cohort (26% vs 74%). Among the cases, 3 patients developed liver failure and 4 venoocclusive disease, and there were 57 deaths, including one death attributed to acute liver failure. Conclusions: Liver injury is common among children treated for cancer. Our data suggests that liver injury is equally prevalent among males and females and more common in younger children (< 11 years). Most cases demonstrated a hepatocellular pattern of liver injury with relatively high amino- transferases. There may be an increased risk for liver injury among children with leukemia vs those with solid tumors. Serious liver-related morbidity was noted among the cases including venoocclusive disease and liver failure, and 19% of cases died. We plan further investigation to describe the etiologies of liver injury, including drugs, as well as associations and risk factors. Sa1571 EFFICACY OF KERATIN 18 (CK18) AS A BIOMARKER OF SEVERITY OF LIVER INJURY IN ACUTE ALCOHOLIC HEPATITIS Vatsalya Vatsalya, Heather Clair, Keith C. Falkner, Matthew C. Cave, Craig J. McClain Introduction: Acute alcoholic hepatitis (AAH) continues to be a major cause of liver related morbidity and mortality for which there are no good blood biomarkers of cell death or prognosis. Keratin 18 (CK18) is found in epithelial cells and is released from hepatocytes upon cell death. It also is a substrate for Caspase 3; thus, the full epithelial cell component (M65) generally indicates necrosis and the cleaved fragment (M30) indicates apoptosis. We postulated that CK18 levels would be a better indicator of liver injury/prognosis compared to standard biomarkers such as AST/ALT Patients: Fifty-five male and female patients aged 21-70 years old with AAH were included in this investigation. They were recruited from the NIAAA-funded U01 DASH consortium. Patients were classified as moderate (n=15) or severe (n = 33) based on MELD score (≤ 19 and >19); seven from the overall recruited patients were categorized as mild (MELD<12) and not sick enough for hospitalization or medical therapy. Data for demographics, drinking history (AUDIT) and MELD, Maddrey DF, and other liver-injury markers were collected. Methods: Blood samples were collected at enrollment and serum CK18 protein was analyzed using ELISA and APOPTOSENSE kit. Comparison of levels CK18 proteins between the AAH patients (sub-classified by severity) was performed using ANOVA. Levels of the CK18 fragments were also correlated with calculated biomarkers such as MELD, DF, etc. Other tests were performed by the hospital clinical laboratory or the DASH Consortium core laboratory. Results: Severe AAH patients had significantly higher MELD, Maddrey DF; however AST and ALT were actually higher in moderate AAH patients. Both CK18-M65 and CK18-M30 showed clinically relevant levels in moderate and higher levels in severe AAH patients compared to AST/ALT. (Fig 1,2). Both M65 and M30 showed a correlation with liver disease severity as assessed by markers, such as MELD and DF. Conclusions: The blood biomarker, Keratin 18, appears to reflect severity of liver injury and it performs much better than traditional blood biomarkers such as AST/ ALT. Future studies will evaluate where admission CK18 accurately predicts prognosis.