Materials and Methods: Between January 2004 and De- cember 2008, 126 consecutive patients (mean age 58 years) were referred for intra-arterial hepatic chemotherapy (IAHC). Ports were percutaneously implanted through fem- oral access under local anesthesia by an interventional ra- diologist (IR) when no surgery was planed, or during lapa- rotopy in addition to a surgery (S) performed for other purpose (resection of liver metastases or primary tumor). Implantation success rate, primary permeability duration (P1), secondary permeability duration (P2)(after revision), the rate of complications, and dysfunctions that lead to IAHC discontinuation were reported. When dysfunction such as extra-hepatic perfusion, incomplete hepatic perfu- sion, thrombosis or migration of the catheter occurred, the revision was performed percutaneously on all patients. Results: 2 IR and 2 surgeons in a single institution im- planted 126 ports. The success rates of implantation for IR and S were 97% (n=65/67) and 98% (n=58/59), respec- tively. 111 patients received IAHC in our institution and could be followed, including 56 IR and 55 S. P1 and P2 were 4.81 and 7.58 courses of IAHC respectively. P1 was the same for IR and S (4.80 vs. 4.82 courses), but P2 was significantly higher for IR than S with 9.18 vs. 5.95 courses respectively (p=0,004). Revisions resulted in a significant increase of the permeability rate but this gain was signifi- cantly higher for IR than S: 4.38 vs. 1.13 courses (p=0.0005). Complications rates for IR and S were not statistically different: 9% (n=5/56) and 13% (n=7/55) re- spectively (p=0.52). Dysfunction for IR and S occurred in 59% (n=33/56) and 35% (n=19/55), respectively (p=0.01), and required 51 revisions in 39 patients. The rates of complications or dysfunctions that lead to IAHC discon- tinuation for IR and S were 19.6% (n=11/56) and 34.5% (n=19/55), respectively (p=0.08). Conclusion: Overall, better outcomes of IAHC occurred following percutaneous versus surgical implantation of the port catheter, including significantly longer permeability and a lower rate of IAHC discontinuation. 11:27 AM Abstract No. 77 Tumorgenesis and angiogenesis factors in hepatocellular carcinoma after locoregional therapy A.B. Farris 1 , R. Dhanasekaran 2,3 , N. Dursun 1 , E. McIntosh 1 , I. Coban 1 , V. Adsay 1 , H.S. Kim 2,3 ; 1 Pathol- ogy, Emory University School of Medicine, Atlanta, GA; 2 Division of Interventional Radiology and Image Guided Medicine, Emory University School of Medicine, Atlanta, GA; 3 Radiology, Emory University School of Medicine, Atlanta, GA. Purpose: Assess effects on tumorgenesis and angiogenesis factors on hepatocellular carcinoma (HCC) after locore- gional therapy (LRT). Materials and Methods: Liver explants of consecutive pa- tients receiving orthotopic liver transplantation for HCC were studied. Immunohistochemistry was performed for cytokeratin CK7, CK19, P53 and vascular endothelial growth factor (VEGF), and the extent of staining was ana- lyzed within and surrounding the HCC. Tumor burden, stage, grade, extent of necrosis, and effects on non-tumor liver parenchyma, such as steatosis were studied. Statistical analysis was performed using SAS JMP software. Results: 57 explants without pre-transplant LRT and 30 explants with pre-transplant LRT were studied. Pathologic and histologic factors of tumor burden, stage, steatosis, and grade were similar between LRT group and control group. Percent necrosis was higher in the LRT vs. control group (56% versus 4%, P  0.0001). By immunohistochemistry, there were no statistically significant differences in tumoral or peritumoral CK7, P53 or CK19, or in peritumoral VEGF. There was a statistically significant difference in tumoral VEGF with LRT-treated HCC demonstrating significantly lower levels of VEGF vs. control group (30% versus 65%, p = 0.03). Conclusion: Necrosis was significantly higher in the LRT group, and tumoral VEGF was lower in the LRT group, suggesting a connection between growth factors and treat- ment-induced necrosis; however, further study is needed on the microenvironment and pathogenesis in HCC. 11:39 AM Abstract No. 78 Evaluation of a novel percutaneous isolated liver perfu- sion (PILP) system in a porcine model G. Maleux 1 , D. Monbaliu 2 , C. Verslype 3 , M. Van De Velde 4 , Y.L. Hoogeveen 5 , E. Van Cutsem 3 ; 1 Department of Radiology, University Hospitals Leuven, Leuven, Belgium; 2 Transplant Surgery, University Hospi- tals Leuven, Leuven, Belgium; 3 Digestive Oncology, Uni- versity Hospitals Leuven, Leuven, Belgium; 4 Anesthesiol- ogy, University Hospitals Leuven, Leuven, Belgium; 5 UMC Sint-Radboud, Nijmegen, Netherlands. Purpose: To evaluate the safety, repeatability and perfor- mance of a novel percutaneous isolated liver perfusion (PILP) system (Medical Device Works, Brussels, Belgium) in a porcine model. Materials and Methods: In six consecutive Dutch Landrace pigs (44 - 58 kg) the PILP-device, consisting of a portal vein occlusion balloon, an hepatic artery occlusion balloon and an hourglass-shaped vena cava stent-graft was inserted in respectively the distal portal vein main branch, the proper hepatic artery and hepatic portion of the vena cava inferior. Additionally, a veno-venous bypass between the distal por- tal vein and the right femoral vein and an isolated perfusion between the proximal portal vein and the vena cava stent- graft were created with use of two perfusion pumps. Perfu- sion was performed retrogradely (from the hepatic veins to the portal vein) with a flow rate between 100-300 ml/min. Leak testing was performed with use of scintillation counter, measuring radioactivity in the systemic circulation, after injection of 70 MBq of Tc99 into the perfusion circuit. Afterwards, perfusion was performed without drug (n=2), with 50 mg of melphalan added in the perfusion fluid (n=2), and with 25 mg of oxaloplatin added in the perfusion fluid (n=2). At the end of the procedure, all devices were re- trieved. Finally, all pigs were euthanasized. Results: In all pigs it was feasible to successfully insert all components of the PILP-device. Leak testing showed a leakage of less than 6% to the systemic circulation over the course of the perfusion. All animals underwent perfusion without observable differences in vital signs. At the end, all components of the PILP-device could be succesfully re- trieved and angiographic control after retrieval could not demonstrate any vessel wall damage. Conclusion: The percutaneous PILP-device can be safely and accurately inserted and retrieved in a pig model without vessel wall damage. Additionally, during retrograde perfu- sion, leakage from the perfusion to the systemic circulation was less than 6%. Finally, in all pigs isolated liver perfusion Scientific Sessions MONDAY S31