Purpose: To assess the effect of partial ablation with IRE on T-cell and macrophage population in the tumor microenvironment. Materials: Murine urothelial carcinoma (MB49) were grown on the flanks of C57BL/6 mice. Animals were randomly assigned to receive partial IRE ablation (1500 V/cm, 90 pulses and 90 us pulse width) with non-invasive caliper electrodes or sham treatment. Five animals from each cohort were sacrificed at 2 and 9 days post-IRE. Tumor size was measured prior to treat- ment and at sacrifice. Harvested tumors were paraffin embedded and stained with H&E, and immunohistochemistry markers for helper (CD4), regulatory (FoxP3), and cytotoxic T-cells (CD8), and macrophages (F4/80). Stained slides were scanned at high resolution and 30 pictures (20 and 40X) were recorded from each tumor sample. Image processing software (ImageJ) was used to count cell populations staining positive for each marker. Student’s t-test was used to assess statistical significance of results. Results: Tumor volume measurements at day 2 (control: 51.4 ± 9.1 mm 3 ; IRE: 45.9 ± 17 mm 3 ) and day 9 (control: 76.9 ± 17.1 mm 3 ; IRE: 60.3 ± 18.7 mm 3 ) were not significantly different (p>0.31). Necrosis of 50% or more was observed on H&E staining in IRE treated tumors only. Compared to control, IRE treated tumors from day 2 had significantly fewer counts of helper (15.6 ± 12.6 vs. 6.2 ± 9.3), regulatory (16.2 ± 10.7 vs. 9.4 ± 9.3), cyto- toxic T-cells (135.5 ± 65.7 vs. 99.0 ± 66.3) and macrophages (11.0 ± 4.9 vs. 7.8 ± 5.1); (p 0.12)*. IRE treated tumors from day 9 had greater, but not significantly more cytotoxic T-cells (168.5 ± 95.9 vs. 223.4 ± 188.1, p>0.16). * Conclusions: Partial ablation with IRE can rapidly deplete T-cells and macrophages from the tumor microenvironment, which is followed by repopulation from systemic circulation. The signifi- cance of intratumoral change in immune cells following partial IRE, and its impact on tumor control require investigation. Abstract No. 451 Reversal of the Warburg phenotype by promotion of pyruvate dehydrogenase activity decreases hepatocellular carcinoma resistance to targeted anti-cancer therapies E. Fagbongbe 1 , D. Das 1 , G. McLennan 2 ; 1 Cleveland Clinic Foundation, Cleveland, OH; 2 Cleveland Clinic, Chagrin Falls, OH Purpose: To test the hypothesis that reversing the Warburg effect by sodium phenylbutyrate sensitizes HCC cells to MTOR inhibitor treatment. Materials: Patient derived primary HCC cells were cultured in Williams’ Media E supplemented with 10% fetal bovine serum, 1% PS, FGF, EGF, insulin and dexamethasone. Cells were treated with increasing concentrations of PhBA (0–15 mM) for 24 hours and 48 hours, and increasing concentrations of PhBA combined with fixed rapamycin concentration for 24- and 48-hours. Cell growth/cyto- toxicity was measured as OD at 490nm using a Microtiter kit from Promega. rapamycin, Na-PhBA, insulin, dexamethasone, and PS were purchased from Sigma. FBS was purchased from Denville. FGF and EGF were purchased from R&D Scientific. Results: Cytotoxicity assays showed dose-dependent and time- dependent toxicity in 2 primary HCC cell lines at 24- and 48-hours when exposed to PhBA. In HCC-11 at 24 hours, percent cyto- toxicity was 5%, 14%, 15%, 32%, 68% for 0.25-mM, 0.5-mM, 1-mM, 5-mM, 15-mM PhBA, respectively. In HCC-11 at 48 hours, percent cytotoxicity was 23%, 30%, 38%, 54%, 80% for 0.2-5mM, 0.5-mM, 1-mM, 5-mM, 15-mM PhBA, respectively. In HCC-37 at 24hours, percent cytoxicity was 13%, 13%, 18%, 30%, 47% for 0.25-mM, 0.5-mM, 1-mM, 5-mM, 15-mM PhBA mrespectively. In HCC-37 at 48 hours, percent cytotoxicity was 14%, 7%, 9%, 36%, 76% for 0.25-mM, 0.5-mM, 1-mM, 5-mM, 15-mM PhBA, respectively. Conclusions: Data shows a dose- and time-dependent effect of PhBA on primary HCC cells. Further studies will be needed to characterize the PhBA/rapamycin interaction. Abstract No. 452 Novel murine model of gastrointestinal microbleeding using irreversible electroporation P. Hangge 1 , Y. Pershad 1 , A. Witting 1 , P. Salmon 2 , H. Albadawi 1 , R. Oklu 1 ; 1 Mayo Clinic Arizona, Phoenix, AZ; 2 Bruker MicroCT, Kontich, N/A Purpose: Current treatment of gastrointestinal bleeding (GIB) involves prompt resuscitation and localization. However, approx- imately 5% of GIB cannot be identified on initial workup. On average, 7.3 diagnostic tests, 5 hospitalizations, and 46 units of blood per patient are required during workup. Improved diagnostic approaches of obscure GIB are needed. To evaluate emerging diagnostic agents, we describe a novel murine GIB model using irreversible electroporation (IRE) which can be detected with mi- cro-computed tomography (micro-CT). Materials: C57BL/6 mice received localized non-thermal IRE injury to the liver to induce hepatic microbleeding under anes- thesia. Prior to the procedure, 200U/kg heparin was injected fol- lowed by 300μl 1:1 v/v heparinized saline and 350 mg/ml iohexol. IRE was performed at 120V/mm using 1-cm 2 tweezer electrode plates applied to both sides of the right median lobe. Micro- perfusion was measured using Laser Speckle Contrast Analysis (LASCA), which was performed at baseline, immediately after IRE and at 10, 30, 45, 60 min post-IRE. Immediately prior to eutha- nasia, another 300μl of heparinized saline-ioxhexol was injected. Whole body micro-CT scan was performed with settings of: 32μm pixel size, 55kV, 181μA, rotation step 0.25º, frame average 3, with a 0.5-mm aluminum filter. Liver tissues were harvested for histo- logical evaluation. Results: Visual inspection of the IRE site showed evidence of hemorrhage within the tissue. Extrahepatic bleeding was not observed. LASCA imaging demonstrated 52% reduced perfusion compared to baseline immediately post-IRE. Perfusion increased to 60% of baseline after one hour post-IRE. Micro-CT analysis showed increased attenuation at the liver IRE site, suggestive of bleeding. Extravasation of erythrocytes within the hepatic paren- chyma was evident on microscopy. Conclusions: We demonstrate a novel, non-traumatic model of liver micro-hemorrhage which can be identified using non-invasive micro-CT imaging and confirmed by histology. Although trauma induced animal hemorrhage models exist, this is the first described model of micro-hemorrhage of the gastrointestinal system. This model can be useful to test emerging hemostatic and imaging agents. JVIR ▪ Scientific e-Posters S191 Scientific e-Posters