325 18-FDG Pet as Factor in Predicting Survival of Patients with Pancreatic Carcinoma Cosimo Sperti, Claudio Pasquali. Franca Chierichetti, Domenico Decet, Andrea Ferronato, Sergio Pedrazzoli The prognosis of patients with pancreatic adenocarcinoma is usually very poor; the prediction of survival by clmicopathologic features is often difficult. Recently, the uptake of 18-FDG in PET scan has been found to be correlated with survival in patients with pancreatic cancer. Aim of this study was to evaluate the role of 18-FDG PET as prognostic factor for the patients with pancreatic adenocarcinoma. From 1996 to 2001 48 out of 106 patients who had a PET scan in the preoperative work-up for pancreatic cancer, received also an additional abdominal PET scan .'~ith semiqnantitative analysis of the uptake. The standardized uptake value (SUV: n.v <2.5) was obtained, and the patients were divided in two groups, according to pathological SUV < or > 4. Patients were evaluated according to TNM (UICC) classifica- tion, gradig of the tumor, and type of operation. Survival curves were prepared using the Kaplan-Meier method, and statistical comparisons were made by log-rank test. A Cox proportional-hazards analysis was used to determine the independent variables that influ- enced survival. Significance was considered as p< 0.05. Results: There were 25 males and 23 females, with mean age 66.3 years (range 48-82 years). Fifteen patients underwent pancreatic resection, 15 bypass surgery, and 18 were not operated. According to TNM staging the patients were found as follows: stage 1 in 2 patients, stage III in 14, stage IVa in 14, and stage Ivb in 18. Twelve patients showed a well differentiated tumor while 21 had a moderate and 15 a poor differentiated or undifferentiated tumor. Twenty-five cancers showed SUV values < 4, and 23 SUV > 4. Survival was significantly influenced by tumor stage (p <: 0.001) when divided in two groups (stage IVb vs. stage Iva, III and I), grading (p = 0.007) and by SUV value (p = 0.01). Multivariate analysis showed that stage (p = 0.004), grading (p = 0.005), and SUV (p = 0.032) were independent predictors of survival. When the patients analyzed for SUV value, were stratified according to TNM stage, FDG uptake was related to survival (p = 0.023) only in those patients without distant metastases. Conclu- sions: SUV determination in abdominal FDG-PET provides informations of prognostic rele- vance in patients with pancreatic cancer before surgery or medical treatment, and may therefore be useful in the selection of patients for therapeutic management. 326 In-Vitro and In-Vivo Effects of Trolmnin I, a Cartilage-Derived Angiogenesis Inhibitor Beatrice E. Kern, Carlos Feruaadez-Del Castillo, Bozena Antoniu, James H. Balcom IV, Andrew L Warahaw Introduction: Antiang~ogenesis agents have promise in initiating cancer implantation and growth. Troponin I is an antiangiogenic compound derived from cartilage. Its functional segment resides at amino-acid 96-116. We studied the effects of this peptide (pTnl) on a) endothelial cell tube (capillary progenitor) formation, b) endothelial cell division, c) induction of intercellular adhesion molecule-1 (1CAM-l) by pancreatic cancer cells (CAPAN-1) and d) growth of pancreatic cancer liver metastases in the mouse. Methods: a) Human vascular endothelial cells (HUVEC) were seeded on a basement membrane matrix, incubated with graded concentrations of pTnI, and tube formation was evaluated histologically, b) HUVEC were labeled with carboxyfluorescein diacetate succinimidyl ester (CFDA,SE), incubated +/- pTnl, and fluorescence measured at 96 hours by flow cytometry to deterrmne cell division. c) ICAM-1 expression was measured by flow cytometry on HUVEC incubated with CAPAN- 1 and treated +/- pTnl. d) CAPAN-1 cells were injected into the spleen of nude mice pre- treated with 3.5 mg/kg/day of pTnl, and body weight, liver weight and tumor burden were measured at 6 weeks. Results: a) pTnl inhibited endothelial cell tube formation (p<0.0001) at concentrations as low as lpg/mL b) Endothelial cell division was inhibited significantly at 96 hours by 3mg/ml pThl (p= 0.0001). c) Supernatant from CAPAN-1 cells upregnlated ICAM- 1 on HUVEC by 96%. This upregulation was diminished by pre-incubation of CAPAN- 1 cells with less than 10ng/ml pTni (p<0.01). d) Mice treated with pTnl had fewer liver metastases compared to controls (liver/bod~ weight 5.5% versus 11.1%, p<0.05). Conclu- sion: The peptide 94-123 of Troponin 1 has an antiangiogenic effect in pancreatic cancer. In-vitro it inhibits pre-vessel tube formation, endothelial cell division, and ICAM-1 upregula- tion by cancer cells, and in-vivo it reduces metastases from pancreatic cancer to the liver in a mouse model. 327 Is Extended Resection For Adenocarcinoma of the Body or Tail of the Pancreas Justified? Margn Shoup, Kevin C. Conlon, David Khmstra, Murray F. Brennan Introduction: Surgical resection is the primary treatment of pancreatic adenocarcinoma. Patients with body or tail tumors may have contiguous organ involvement or portal-splenic confluence adherence requiring extended operation to obtain complete gross resection. The aim of this study was to determine survival following contiguous organ or portal vein resection in patients with body or tail adenocarcinoma. Methods: Between 1983-2000, 513 patients with adenocareinoma of the body or tail of the pancreas were admitted for treatment and entered into a prospective database. Distal pancreatectomy with or without splenectomy was performed in all resected patients. Patients who had portal vein resection or contiguous organ resection including gastrectomy, adrena- lectomy or colectomy were considered to have extended resection. Patients not undergomg pancreatectomy were deemed unresectabte either because of metastatic disease (M1) or locally advanced dtsease ('1"4) Pathology was reconfirmed in all long term survivors. Results: Among all patients, 320 had M1 disease and 136 had T4 disease and did not undergo pancreatectomy. Of the 57 (11%) who were resected, extended resection was necessary in 22 (39%) patients; 14 (64%) for contiguous organ involvement and 8 (36%) for portal vein resection. Estimated blood loss, blood transfused and length of hospital stay were significantly more in patients requiring extended resection (p = 0.02., p = .01., p = .02 respectively). Median follow-up for patients still alive (n = 6) was 84 months (range 40-189 months). Median survival following resection was 15.9 months compared to 5.8 months in patients who were not resected (p<O.0001). Actual 5 and 10-year survival was 19% and 10% following extended resection, 9% and 3% following standard resection, and 0% and 0% if no resection was attempted due to T4 disease. Conclusion: Patients undergoing extended resection for adenocareinoma of the pancreatic body or tail have similar long-term survival as those undergoing standard resection, but markedly improved long-term survival compared to those not considered resectable second- ary to locally advanced disease. Extended distal pancreatectomy is justified in this group of patients. 328 Specific Gene Expression and Therapy for Pancreatic Cancer Using ]'he Cytosine Deaminase Gene Directed by The Rat Insulin Promoter Xiao-Ping Wang, Kaanyuki Yazawa, William Fisher, Charles Brunicardi Background: Suicide gene therapy has been shown to be an effective means of destroying pancreatic cancer cells, but cell specific delivery of the gene is required to limit host toxicity. The objective of this study is to determine whether the rat insulin promoter (RIP) will permit cell specific gene delivery and subsequent cell death in human pancreatic cancer cells. Methods: The RIP DNA was amplified using PCR method and the purified fragment was inserted into pCR-Bhint II-TOPO plasmid at Spel site which contains the coding sequence of yeast cytosine deaminase (CD). Transfection assays were carried out using both RIP-lacZ and RIP-CD DNA constructs.into two human pancreatic cancer cell lines, Panc-1 and MIA PaCa-2. Reporter assays using X-gal staining were performed and the in vitro cytotoxicity was examined in RIP-CD transfected cells which was treated by 5-FC (5-flucytosine) for 5 days. The expression levels of CD protein in the trsnsfected cells were determined 2 days post transfection by western blot analysis. The expression levels of insulin promoter factor (IPF-1/PDX-1) in these human pancreatic cell lines as well as freshly isolated human speci- mens were determined using western blot analysis. Results: Western blot analysis demon- strated that sigmficantly increased levels of PDX-1 were found in Panc-1 but not Mia PaCa- 2 cells. X-gal staining showed that only PANC-1 cells, not M1A PaCa-2 cells, were positive for RlP-lacZ expression indicating that the insulin promoter directed reporter gene expression in Panc-1 cells, but not in M1A PaCa-2. RIP-CD transfected and 5-FC treated Panc-1 cells had significantly increased cell death compared with that of MiaPaca-2 cells, suggesting RIP directed suicide gene expression occurred in Panc-1 but not Mia PaCa-2 cells. Western blot analysis demonstrated that only Panc-1 cells, not MIA PaCa-2 cells, were able to express the CD protein. Furthermore, six freshly isolated human pancreatic cancer specimens and two pancreatic cancer liver metastases also had significantly increased levels of PDX-1. Conclusion: The rat insulin promoter is activated in selected human pancreatic cancer cells that express PDX-1. Significantly increased levels of PDX-1 have been found in 8/8 human pancreatic cancer specimens. Pancreatic cancer-specific cytotoxicity can be achieved using RIP-CD and 5-FC treatment in vitro. These results suggest that RIP could be used for cell specific gene therapy to target human pancreatic tumors. 329 Laparoscopic lntrahiminal Cystgastrostomy Ken Eto, Haroon Anwar, Kingsway Lira, Kathleen M. Manning, Toshiyuki Mori, Lawrence W. Way Objective: This report analyzes the results of 10 years' experience using a novel laparoscopic technique for performing cystgastrostomy (LCG) for pancreatic pseudocysts. Methods: In 1991 we devised a unique radially-expanding trocar that could be inserted percutaneonsly into the gastric lumen and used as an access tool for performing an intragastric LCG. Once the setup is established, it is easy to make the opening between the stomach and an adherent posterior pseudocyst, and the puncture wounds in the anterior wall of the stomach from the trocars are simple to close. We report the results in 46 consecutive operations for pancreatic pseudocysts, which were thought possibly to be amenable to LCG. Treatment indications were a symptomatic retrogastric pseudocyst > 4cm on imaging studies. Results: Forty.six LCG were attempted in 45 patients. The etiology was gallstones in 22 (49%), alcohol in 11 (24%), trauma in 3 (7%), post ERCP pancreatitis in 3 (7%), and other ni 6 (13%). The longest diameter of the pseudocyst was 11 +_5 cm (range 4.5-24). Success for the LCG was defined as a completely laparoscopic operation with permanent resolution of the pseudocyst. Thirty-six (76%) of the operations were successfully completed laparoscopi- cally, and ten were converted. The principal reason for conversion was anatomy unsuitable for cystgastrnstomy (8 cases). The operative time was 167 + 52minutes (range 96-295) for laparoscoptc cases and 233+90 minutes (range 125-3905 for converted cases. Operative blood loss for LCG cases was negligible in each. Median hospital stay for laparoscopic cases was 4-+ 4.6 days (range 1-25). Postoperative bleeding, treated endoscopically, occurred in two LCG cases, one from the cystgastrostomy margin, and one from inside the cyst. The former was the only complication directly attributable to the laparoscopic technique itself. The cyst recurred in one (4%) laparoscopically treated patient, but this recurrent collection responded to reopening the anastomosis with a push from the tip of an endoscope. Conclu- sion: Laparoscopic intraluminal cystgastrostomy is a technically straightforward, highly suc- cessful, and much less invasive method than laparotomy for treating retrogastric pseudocysts. Currently, the average LCG takes less than 2-h. Most patients are eating a full diet within 48-h of surgery and are ready to be discharged shortly thereafter. The technique has few complications and a minimal recurrence rate. We recommend laparoscopic intragastric cystgastrostomy as the preferred surgical approach for retrogastric pseudocysts. Ao787 SSAT Abstracts