Mo1428 Can Development of Acute Kidney Injury in Acute Pancreatitis Be Predicted?

AGA Abstracts

Mo1427 Ileus Is a Predictor of Local Infection in Patients With Acute Necrotizing Pancreatitis Robert A. Moran, Niloofar Yahyapourjalaly, Ayesha Kamal, Sandesh Rao, Theodore W. James, Swetha Kambhampati, Robert Klapheke, Vivek Kumbhari, Ellen M. Stein, Mouen A. Khashab, Anne Marie Lennon, Anthony N. Kalloo, Atif Zaheer, Ruben Hernaez, Vikesh K. Singh INTRODUCTION: Infected pancreatic necrosis (IPN) is associated with increased morbidity and mortality. Gut barrier dysfunction has been shown to increase the risk of bacterial translocation from the gut into the pancreatic bed. The utility of ileus, a clinical marker of gut barrier dysfunction, for predicting IPN is unknown. METHODS: A retrospective cohort study of patients with necrotizing pancreatitis (NP) was conducted from 2000-2014. Ileus was defined as ‡ 2 of the following criteria: nausea/vomiting; inability to tolerate a diet, absence of flatus, abdominal distension and features of ileus on imaging. Extensive necrosis was defined as >30% nonenhancing parenchyma on contrast-enhanced CT (CECT). Multivariable cox proportional hazard analysis was used to evaluate known and potential predictors of IPN. RESULTS: 142 patients were identified with NP, 61 with IPN and 81 with sterile necrosis. In comparison to a diagnosis of ileus documented in the medical chart, the ileus criteria had a sensitivity, specificity and positive and negative predictive value of 100%, 93%,78% and 100%, respectively. On multivariate cox proportional hazard analysis ileus [HR: 2.6, 95%CI:1.4-4.9] and extensive necrosis [HR:2.8, 95%CI:1.3-5.8] were independently associated with the development of IPN. Bacteremia [HR: 1.09,95%CI:0.6-2.1] was not associated with the development of IPN. CONCLUSION: Ileus in NP can be accurately defined using previously reported surgical criteria. Ileus is independently associated with the future development of IPN. Further studies will be needed to determine if ileus can serve as a clinical marker to direct therapeutic interventions aimed at reducing the incidence of IPN.

Mo1430 Clinical Characteristic of Hypertriglyceridemia Induced Acute Pancreatitis: A Single Center Study From Prospective Cohort Jae Hee Cho, Eui Joo Kim, Seong Han Choi, Su Young Kim, Yoon Jae Kim, YeonSuk Kim

Mo1428 Can Development of Acute Kidney Injury in Acute Pancreatitis Be Predicted? Raghavendra Prasad, Jayanta Samanta, Sukhwinder Singh, Narendra Dhaka, Yalaka R. Reddy, Munish Ashat, Sunil K. Arora, Raja Ramachandran, Vikas Gupta, Thakur D. Yadav, Saroj K. Sinha, Rakesh Kochhar

Background and Aims: Hypertriglyceridemia (HTG) is the 3rd most common cause of acute pancreatitis (AP). Elevation of serum triglyceride to ‡1000 mg/dl strongly indicates HTG induced AP and severe HTG is associated with persistent organ failure in AP. Because HTG induced AP is less known in the literature by this time, we aim to validate its clinical characteristics and the role of prediction for severity in AP. Methods: The analysis was performed on the basis of data from a prospective AP registry in Gachon university, Gil medical center between July 1, 2013 and March 31, 2015 Results: Among 365 patients with AP, HTG induced AP was present in 32 patients (8.7%). HTG induced AP group was younger (41.1±10.4 vs. 52.6±17.3 years, P<0.001) than the other group and associated with male (P=0.024), alcohol intake (P=0.002) and smoking (P<0.001). Onset to admission interval was shorter (1.2±1.4 vs. 3.2±6.8 days, P<0.001) and fever at admission was more frequently observed in HTG induced AP (71.9 vs. 47.4%, P=0.008). The recurrence rate of HTG induced AP was higher (46.9% vs. 23.7%, P=0.004) than the other group. On the multivariate analysis, younger age ( £40 years) (OR 4.4, 95% CI 1.8-10.4), smoking (OR 7.1, 95% CI 1.9-6.2), fever at admission (OR 4.1, 95% CI 1.7-9.7) were independently associated with HTG induced AP. In addition, the multivariate analysis showed that fever at admission was the only significant factor (OR 2.1, 95% CI 1.4- 3.3) related to moderately severe/severe AP compared to mild AP. Conclusions: HTG induced AP is associated with younger age (£40 years), smoking, fever at admission and its recurrence rate is higher than other etiology of AP. However, the level of TG is not related to severity of AP.

Introduction: Cytokines play an important role in the pathophysiology of acute pancreatitis (AP) and acute kidney injury (AKI). Aims & Methods: To study the predictive role of inflammatory cytokines in development of AKI in patients with AP. In this prospective study between July 2013 and December 2014 consecutive eligible patients of AP were enrolled and organ failure was defined according to modified Marshall Score. Serum levels of interleukin (IL)-6, IL-8, IL-10, IL-1b and TNF a were measured at baseline (day 1) for all patients and on day 3 in those who had AKI. For comparative analysis patients were divided into 2 groups: with and without AKI. The AKI cohort was further subdivided into persistent AKI (P-AKI) and transient AKI (T-AKI). Statistical analysis was done using SPSSv20.0 to study the predictive value of different cytokines for development of AKI. Results: Of the 107 patients, AKI developed in 20 (18.7%). T-AKI was seen in 7 (35%) while 13 (65%) had P-AKI. Patients with AKI had significantly higher IL-6 (p=0.004), IL-8 (p<0.0001) and TNFa (P=0.05) levels on day 1 when compared to non-AKI group. In the AKI group, day 3 levels of TNFa (p=0.010) were significantly higher than day 1 levels whereas IL-10 (p= 0.04) levels were significantly lower than day 1 levels. Significant rise on day 3 of TNF a (p=0.004) was observed in the P-AKI group. Day 1 levels of IL-6 and IL-8 had strong positive correlation with severity indices such as SIRS (p<0.001), BISAP (p<0.001), CTSI (p<0.0001) and with outcome measures like need for intervention (p<0.0001), hospital stay (p<0.0001) and intensive care stay (p<0.0001). Conclusion: IL-6 and IL-8 levels at admission were significantly associated with development of AKI in AP. Rising levels of TNF asuggested development of persistent AKI.

Mo1431 Day of the Week and Acute Pancreatitis: Is There a Negative Weekend Effect on Outcomes? Paul T. Kroner, Marwan S. Abougergi, Christopher C. Thompson

Mo1429

Introduction Differences in treatment outcomes based on whether hospital admission occurs on a weekend versus weekday have been reported for several medical conditions including myocardial infarction and cerebrovascular accidents. We sought to determine if in- hospital mortality, morbidity or resource utilization vary based on day of admission among patients admitted to the hospital with acute pancreatitis. Materials and Methods This was a retrospective cohort study using the 2012 National inpatient sample, the largest publically available inpatient database in the United States. The inclusion criteria were age older than 18 years and an ICD-9 CM code for a principal diagnosis of acute pancreatitis. There were no exclusion criteria. The primary outcome was in-hospital mortality. The secondary outcomes were morbidity as measured by intensive care unit (ICU) admission, shock, at least one organ failure and multi-organ dysfunction, and resource utilization as measured by use of abdominal ultrasound, abdominal CT scan, and total parenteral nutrition (TPN), length of stay (LOS) and total hospitalization charges. Weekend was defined as midnight Friday through midnight Sunday. Odds ratios and means were adjusted for the following confounders using multivariate regression analysis models: Age, sex, race, median income in the patient's zip code, Charlson Comorbidity Index, hospital region, urban location, size and teaching status. Results 71,310 patients with acute pancreatitis were admitted on a weekend, while 203,855 patients were admitted on a weekday. The mean age was 51 years and 48% of patients were female. The in-hospital mortality rate was 0.8%. All adjusted odds ratios, adjusted means and p values are shown in Table 1. On multivariate analysis, similar inhospital mortality rates were found for patients admitted on a weekend compared with a weekday. In addition, morbidity as measured by rate of ICU admission, shock, single or multi-organ failure did not differ between the two groups either. When resource utilization was examined, patients admitted on a weekend and a weekday had similar rate of abdominal CT scan, abdominal ultrasound or TPN use as well as similar length of hospital stay and total hospitalization charges. Conclusion Unlike for other medical conditions, day of admission is not a predictor of in-hospital mortality among patients admitted to the hospital with acute pancreatitis. In addition, day of admission does not impact either morbidity (as measured by ICU admission, shock, single or multi-organ dysfunction) or resource utilization

Validation of the BISAP Score as a Predictor of Severe Acute Pancreatitis, InHospital Mortality and 30 Day Mortality in an Academic Hospital System Kunjam Modha, Tiffany Chua, Daniel Jang, Peter Junwoo Lee, Rocio Lopez, Tyler Stevens Introduction and aim: The bedside index for severity in acute pancreatitis (BISAP) was derived and validated in 2008 as a prognostic scoring system to identify patients at an increased risk of mortality in acute pancreatitis (AP). It was originally validated using a large administrative database, and few subsequent studies have been done in research databases for external validation. We assessed the accuracy of the BISAP score in predicting additional adverse outcomes, including severe AP, multi-system organ failure (MSOF), in-hospital mortality, and 30 day mortality. Methods: A retrospective study was conducted using our IRB-approved acute pancreatitis database. Consecutive patients admitted either to our tertiary care center or to 8 affiliated hospitals who had a primary discharge diagnosis of AP (ACG definition, ICD9 577.0) were included. Patients with chronic pancreatitis were excluded. Receiver Operating Characteristics (ROC) analysis was done to assess the ability of admission BISAP to predict severe pancreatitis (revised Atlanta, Marshall Score), MSOF (>1 organ systems), in-hospital mortality, and 30-day mortality. All analyses were done using SAS version 9.4. Results: 536 subjects were included in the final analysis, including 53 (9.9%) with severe AP, 27 (5%) with MSOF, 14 (2.6%) who died in the hospital, and 13 (2.4%) who died within 30 days of hospitalization. The ROC area under the curve (AUC) for prediction of severe AP, MSOF, in-hospital mortality and 3-day mortality are presented in table 1. The AUC was ‡0.80 (very good discrimination) for all 4 outcomes. Conclusion: The admission BISAP score not only accurately predicts mortality as originally proposed, but demonstrates similar prediction of severe AP and organ failure. Although the score is simple and reliable compared to more complex scoring systems, the AUC is not excellent (>0.90), indicating the need for further biomarkers and prediction tools. Area under ROC curve for all outcomes

AGA Abstracts

S-710