AASLD Abstracts Figure 1. This graph compares the cumulative survival for patients who underwent TIPS revision versus no revision. Figure 2. This graph compares the likelihood of liver transplantation for patients who underwent TIPS revision versus no revision. Tu1685 TOWARD A MAGNETICALLY ACTUATED PUMP FOR THE TREATMENT OF REFRACTORY ASCITES Nicolo Garbin, Jesse Taylor, Qasim Khalil, Mubashir Khan Background: It has been estimated that by 2020, 100,000 patients will develop refractory ascites (RA) in the United States and Europe. The current standards of care lower patients’ quality of life and increase cost of care. Our team has developed an inexpensive magnetically actuated pump for the management of RA. The system is composed of: a catheter-like implantable pump (CIP), and a magnetic external controller (MEC). Performance of the system in-vivo is unknown. Method: The proposed solution exploits magnetic coupling of diametrically magnetized cylindrical magnets to transfer mechanical power across a physical barrier. In particular, a driving magnet, embedded in the MEC, couples with a driven magnet contained in the CIP to form a reciprocating positive displacement pump design. The CIP is intended to be implanted subcutaneously to collect ascitic fluid, and to pump ascitic fluid into the bladder for natural fluid removal. The MEC is intended to be placed over the implantation site to actuate the pump by the patient and/or care giver daily. The system flow rate can be easily regulated up to 3.65L/min as demonstrated with ex-vivo bench studies. Study: In-vivo validation of the system has been assessed on a 42-Kg Yorkshire-Landrace swine. The CIP was implanted subcutaneously with the inflow lead placed in the abdominal cavity, while the outflow lead was secured with a purse-string suture at the apex of the bladder. RA was simulated by introducing 3 L of warm saline in the doubled-sutured abdominal cavity via two intracath. A Foley catheter was used to empty the bladder prior to pump actuation and to measure the displaced liquid. The pump was actuated three times for 10 minutes with 10 minutes interval between each actuation. The amount of displaced fluid was measured at the end of each 10 minutes run. Results: 250 ml, 8ml, 5 ml of fluid were collected from the bladder at the end of each pump actuation run. The second and third results suggested pump malfunctioning as the amount of fluid would be reasonable for actual urine production over a period of 10 minutes. The first run however validated pump functioning as the amount of fluid could only be associated with active fluid delivery. When retrieving the CIP, the reason of malfunctioning was kinking of the inflow leads. The short length of the subcutaneous tunnel imposed a sharp turn to the soft silicone inflow and outflow leads which partially compromised the device functioning. Nevertheless, this study validated the device functioning in-vivo. Anti-kink solution on the inflow and outflow leads will be included in future design iterations. The proposed device has the ability to easily and actively manage RA fluid in a novel and low-cost manner. S-1452 AASLD Abstracts (a) Schematic of refractory ascites and pump’s placement; (b) CIP; (c) MEC bottom half. (a) Surgical placement of CIP; (b) insertion of CIP outflow lead into urinary bladder; (c) MEC activation over the subcutaneously implanted CIP. Tu1686 PRE-TIPS SARCOPENIA IS ASSOCIATED WITH SHORT-TERM BUT NOT LONG-TERM MORTALITY IN PATIENTS WITH CIRRHOSIS. Ahmed Ouni, wei zhang, Walid Khan, Blake A. Thompson, Nabeel Moon, Craig Meiers, Hugh Davis, Brian Geller, Andreas G. Zori, Giuseppe Morelli Background and Aims: Sarcopenia is common in patients with cirrhosis and is associated with increased mortality. Placement of transjugular intrahepatic portosystemic shunt (TIPS) is frequently utilized in cirrhotic patients with hepatic decompensation. No studies have evaluated the association between the presence of sarcopenia and its effect on mortality after TIPS. The purpose of this study was to investigate the post TIPS mortality in patients with sarcopenia. Method: We conducted a single-centered retrospective study of 305 patients with cirrhosis who underwent TIPS between 2010 and 2015. Psoas muscle index at the third lumbar vertebra (PMI) was measured by computed tomography adjusted by height. Sarcopenia was defined as a PMI less than the 25 th percentile of the study population stratified by gender, as previously described. We conducted univariate and multivariate analyses to identify the association between sarcopenia and post-TIPS mortality. Results: Overall, 184 (60.3%) patients received TIPS for refractory ascites, 97 (31.8%) patients received TIPS for oesophageal variceal (EV) bleeding, and 24 (7.9%) for other indications. The mean age was 63.88 ± 9.86 years old. Sarcopenia was present in 67 (22%) of patients. Median MELD scores for the Sarcopenic and non-Sarcopenic groups were 13.25±3.98 and 12.98±5.06 (p = 0.602), respectively. BMI was higher for the non-Sarcopenic patients compared to those with Sarcopenia, 29.13 ± 6.27 kg/m 2 and 25.36 ± 6.37 kg/m 2 (p <0.001), respectively. Average PMI was lower in the Sarcopenic group compared to those without Sarcopenia, 2.66 ± 0.47 cm 2 /m 2 vs. 4.94 ± 1.28 cm 2 /m 2 (p <0.001). Patients with sarcopenia had a higher mortality rate at one year when compared to patients without sarcopenia (38.8% vs. 25.2%, p < 0.05). Using multivariate analysis after adjusting MELD and BMI, patients with sarcopenia had higher odds of mortality compared to patients without sarco- penia (odds ratio [OR]: 1.88, 95%; confidence interval [CI]: 1.056-3.378). However, no statistically significant differences were found between sarcopenia and long term (2-year, 3- year, or 5-year) mortality. Conclusion: In cirrhotic patients with sarcopenia who underwent TIPS, there was increased mortality at one year but not beyond. Future studies are needed to investigate the impact of post-TIPS nutrition optimization in this patient population in preventing post-TIPS adverse outcomes.