Results: The CBD defect was successfully repaired with a gallbladder wall free flap. The patient remains well 2.5 years later on follow up with normalization of liver en- zymes and without sign of a bile leak. Conclusion: Mirizzi’s syndrome and laparoscopic CBD injury can happen in patients with significant comorbidity. Usually hepaticojejunostomy is required to repair the bile duct, which has a high complication and mortality rates in this subgroup of patients. Gallbladder wall free flap is a reasonable choice to repair the defects with good result and free flap survival like our case. The technique should be considered as an alternative to avoid morbidity and mortality in high-risk patients. Two giant gallstones found within the gallbladder causing compression of the common bile duct and Mirizzi’s Syndrome with cholecholedochal fistula. P 67. INDOCYANINE GREEN FLUORESCENCE COUNTERSTAINING FOR ANATOMIC LAPAROSCOPIC PARTIAL SPLENECTOMY T. Mizuno, C. Goumard, M. Okuno, E. A. Asare, G. W. Krampitz, T. A. Aloia, C. W. D. Tzeng, Y. S. Chun, J. E. Lee, J. N. Vauthey, M. Nagino and C. Conrad * Corresponding author. Takashi Mizuno, The University of Texas MD Anderson Cancer Center, USA Background: Laparoscopic partial splenectomy (LPS) is a valuable option in preserving immunologic function and achieving early recovery. While in LPS, demarca- tion following control of inflow vessels can define the transection plane, the precise demarcation line and adequate perfusion of the splenic remnant may be unclear. Methods: A 50-year old healthy male patient presents with an enlarging splenic mass (2.8e3.7 cm over 4 years). Core needle biopsy was non-diagnostic, howev- er, a low-grade lymphoproliferative malignancy could not be excluded. Following multidisciplinary tumor board discussion LPS to provide definitive tissue diag- nosis and remove this potentially malignant tumor was recommended. Results: With the patient in reversed modified French po- sition the firstly encountered splenic vein was exposed and displaced for exposure of the splenic artery. The inferior pole splenic artery was clamped before the vein to allow for autotransfusion. ICG was administered systemically. Rapid ICG wash-in and was-washed out confirmed preserved perfusion to the splenic remnant. Parenchymal transection was performed with minimal blood loss. Conclusion: LPS is a viable diagnostic and therapeutic option for potentially malignant splenic tumors allowing for preservation of splenic immunologic function and early recovery. Splenic arterial prior to venous control may technically be more challenging but allows for autotransfusion and minimizes blood loss. Intra- operative ICG fluorescence imaging facilitates visuali- zation of the demarcation line and perfusion of the splenic remnant. P 68. BMI AND PANCREATIC NECROSIS...DO OBESE PATIENTS DO WORSE? S. Quigley, N. J. Zyromski, M. G. House, E. P. Ceppa, A. Nakeeb, C. M. Schmidt, S. Mahajan, C. Budden and A. Roch * Corresponding author. Stephen Quigley, Indiana University, USA Background: Obesity is a risk factor for developing severe acute pancreatitis. However, the correlation between out- comes in obesity and Necrotizing Pancreatitis (NP) is un- known. We hypothesized that obesity negatively impacts NP outcomes. Methods: A retrospective review of database was performed on all patients readmitted at a level 1 center between January 2005 and July 2015. Number of Read- missions (NoRA) and Length Of Stay (LOS) were used as outcome measures and analyzed using parametric tests based on defined BMI classification. Results: A total number of 540 patients were identified. The median age of all patients included in the study (n = 399) was 69.6 years with 64% being males. The mean admission BMI was 29.3 Kg/m 2 . Mean NoRA was calcu- lated at 2.21 and LOS was 25.1 days. Four groups were separated based on BMI and compared for the above var- iables (see Table 1). Overall, patients with class I- III obesity had higher readmission rates and LOS. Forty-two deaths were identified representing a 12.8% mortality rate. The mean BMI of all patients that died was 30.9 kg/m 2 . Twenty-two of these patients (52%) had a BMI > 30 kg/ m 2 . One-way ANOVA was performed in standard statisti- cal fashion. Post-hoc testing demonstrated statistical sig- nificance (p = .043) occurs when comparing NoRA in Class III and non-obese patients. LOS did not reach statistical significance across the groups. Conclusion: Class III obese NP patients experience an increased number of hospital readmissions. Mortality rates appear to be even across obese and non-obese patients. HPB 2018, 20 (S1), S119eS172 ePoster Abstracts S141