368 Proffered Papers Poster presentations (Wed, 23 Sep, 14:00−17:00) Gastro-intestinal malignancies − Non-colorectal cancer 6520 POSTER Weekly intravenous and intraperitoneal paclitaxel combined with S-1 for advanced gastric cancer with peritoneal metastasis H. Ishigami 1 , J. Kitayama 1 , S. Kaisaki 1 , A. Hidemura 1 , M. Kato 1 , K. Otani 1 , T. Kamei 1 , D. Soma 1 , H. Miyato 1 , H. Nagawa 1 . 1 The University of Tokyo, Department of Surgical Oncology, Tokyo, Japan Background: A phase II study to evaluate the efficacy and tolerability of weekly intravenous and intraperitoneal paclitaxel combined with S-1 was performed in gastric cancer patients with peritoneal metastasis. Methods: Gastric cancer patients with peritoneal dissemination and/or can- cer cells on peritoneal cytology were enrolled. Paclitaxel was administered intravenously at 50 mg/m 2 and intraperitoneally at 20 mg/m 2 on days 1 and 8. S-1 was administered at 80 mg/m 2 /day for 14 consecutive days, followed by 7 days rest. The primary endpoint was the 1-year overall survival rate. Secondary endpoints were the response rate, efficacy against malignant ascites and safety. Results: Forty patients were enrolled, including 21 with primary tumors with peritoneal dissemination confirmed by staging laparoscopy, 13 with peritoneal recurrence, and 6 with positive peritoneal cytology only. The median number of courses administered was 7 (range 1−23). The 1-year overall survival rate was 78% (95% CI, 65−90%). The overall response rate was 56% (95% CI, 32−79%) in 18 patients with target lesions. Malignant ascites disappeared or decreased in 13 of 21 (62%) patients. The incidences of grade 3/4 hematological and non-hematological toxicities were 40% and 15%, respectively. The frequent grade 3/4 toxicities included neutropenia (38%), leukopenia (18%), anemia (10%) and nausea (8%). Catheter obstruction observed in one patient was the only complication related to the peritoneal access device or intraperitoneal infusion. There were no treatment-related deaths. Gastrectomy was performed in 16 patients after response to chemotherapy, and the 1-year overall survival rate was 94%. Conclusions: Combination chemotherapy of intravenous and intraperi- toneal paclitaxel with S-1 is well tolerated and active in gastric cancer patients with peritoneal metastasis. 6521 POSTER Analysis of patterns of failure after a study of interobserver variability in target volume delineation in postoperative radiochemotherapy for gastric cancer C. Moretones 1 , A. Navarro 1 , D. Leon 1 , A.M. Boladeras 1 , M. Maci` a 1 , M. Cambray 1 , V. Navarro 1 , F. Guedea 1 . 1 Institut Catala d’Oncologia Hospital Duran i Reynalds, Radiation Oncology, Barcelona, Spain Background and Purpose: In 2001, the INT0116 trial showed that adjuvant chemoradiotherapy has a significant role in reducing recurrence and increasing survival in gastric cancer. However, a previous study of interobserver variation between radiation oncologists in volume delineation reported a significant difference in standard deviation between observers, although this was clinically less evident (ESTRO 27 #4556). The aim of the present study was to analyze failure patterns by delineated fields and determine if there is a relation with radiation oncologists volume delineated. Materials and Methods: In 2008, four physicians from our hospital trained in delineating upper abdomen volumes were asked to delimitate the planning target volume (PTV) according to the MacDonald scheme on the same 3D CT-images in 9 postoperative radiochemotherapy gastric cancer cases. Instructions were given to include the tumor bed, the remaining stomach if partial surgery was performed, anastomosis, the duodenal loop and perigastric, celiac, local paraaortic, splenic, hepatoduodenal and pancreaticoduodenal lymph nodes. Enhanced preoperative CT images were available. None of the observers had knowledge of the volumes outlined by the others. One year later we analyzed the status of these patients by recording recurrences and reviewing the 3D-planning volumes, as either distant or locoregional (gastric or tumor bed, the anastomosis and regional lymph nodes), assuming that PTV included more than group 2 lymph nodes. Any lymph node recurrence outside the PTV was defined as distant metastasis. Results: The median follow up was 12.3 months (range 6−17 months). At this time 4 patients relapsed. The mean time to relapse was 9.32 months (range 6.23−6.18), and all of them presented distant recurrence while 1 also had locoregional recurrence. Distant metastases were commonly peritoneal seeding, but in two cases extra-abdominal metastases (CNS and lungs) were found. Only one case of locoregional failure and relapse was both within and outside of the PTV in all four delineated volumes. One case was lost of follow up and analysis excluded. Five patients remain free from disease. Conclusions: Locoregional recurrence rates and patterns of failure correlate to published results. No differences were found between observers in locoregional failure, confirming poor clinical impact of interobserver variation between radiation oncologists in volume delineation as reported in previous study. 6522 POSTER Adjuvant chemoradiotherapy with continious infusion 5-fluorouracil and bi-weekly cisplatin and infusional 5-fluorouracil for operated locally advanced gastric cancer U. Dogan 1 , H. Abali 1 , F. Ozmen 1 , B. Oksuzoglu 1 , N. Aslan 2 , N. Ozdemir 1 , B. Budakoglu 1 , T. Guler 1 , M. Tumoz 2 , N. Zengin 1 . 1 Ankara Numune Educational and Research Hospital, Medical Oncology, Ankara, Turkey; 2 Ankara Numune Educational and Research Hospital, Radiation Oncology, Ankara, Turkey Background: Adjuvant chemoradiotherapy followed by surgery for gastric cancer is widely accepted in clinical practice, but optimal regimen is still debated. In this study was analysed the effectiveness and aplicability of adjuvant treatment protocol with cisplatin, infusional 5-fluorouracil (5FU) and folinic acid (CFF) plus infusional 5FU and concurrent radioterapy in patients with gastric adenocarcinoma. Material and Methods: Between May 2005 and Dec 2008, 65 curatively resected gastric adenocarcinoma were included in this restrospective study. Inclusion criteria were as follows: pathologic tumor category T4, or N2 or N3 or involved/resected lymph nodes with ratio greater than 1/3. Chemotherapy regimen consisted of folinic acid 200 mg/m 2 , cisplatin 50 mg/m 2 , 5FU 400 mg/m 2 bolus followed by 5FU 1600 mg/m 2 46 h-continuous infusion every 14 days. Chemoradiotherapy was administered after 2 cycles of CFF 14 (consisting of 4 courses chemotherapy) as a radiotherapy 4500cGy, 5 weeks and concurrently 5FU 200 mg/m 2 /day. Two more cycle of CFF were administered after chemoradiotherapy. The results was compared with those of 62 patients at similiar stages treated in accordance with Intergroup 0116 trial in our clinic. Results: Of patients, 48 were male and the median age was 55 years. Their clinical stages were stage II in 1 patient, stage III in 37 patients and stage IV M0 in 27 patients. D1 and D2/D3 lymphadenectomy was performed in 21.4% and 78.5% patients, respectively. Fifty seven (87.7%) patients could complete at least 90% of planned treatment. Grade 3/4 hematologic toxicity occured in 16.8% and grade III/IV non-hematologic toxicity occured in 13.7% of patients. The median follow-up for CFF and Intergroups arms was 15 (6−36) months and 15.5 (3−72) months, respectively. While two groups were equal for stage and prognostic parameters, patients with N3 were much more (24 vs 8 patients) in CFF group (p = 0.002). Median disease free survival (18 months, 95% CI: 13.9−22.0) was insignicantly (p = 0.51) longer in CFF 14 patients than INT0116 patients (14 months, 95% CI: 7.7−20.3). Median overal survival of CFF 14 patients (19 months, 95% CI: 15.2−22.8) was almost identical with INT0116 patients (20 months, 95% CI: 12.6−27.3) (p = 0.73). However, grade 5 toxicity was much more common in Intergroup arm (0 vs 5 patients, p = 0.02). Conclusion: Bi-weekly CFF 14 and concurrent radiotherapy protocol seems to be well tolerated (87.7% of patients could receive the whole set) and it provided equivalent survival rates with INT0116 protocol for locally advanced gastric adenocarcinoma patients. 6523 POSTER Palliative chemotherapy does not improve survival in metastatic oesophageal cancer S. Horn 1 , S. Dominguez 1 , M. Vanhuyse 1 , E. Amela 1 , S. Kanoun Balajouza 1 , M. Jasserand 1 , X. Mirabel 1 , N. Penel 1 , A. Adenis 1 . 1 Centre Oscar Lambret, Gastrointestinal Oncology, Lille, France Background: The role of chemotherapy in metastatic oesophageal carcinoma (MOEC) remains debated as randomised trials comparing chemotherapy to best supportive care are lacking. The objective of this retrospective study on one of the largest series of MOEC ever reported was to analyse the survival impact of chemotherapy after stratification to prognostic factors (PF). Material and Methods: All consecutive MOEC treated at the Northern France Cancer Center (Centre Oscar Lambret) from 1995 to 2008 were randomly split into a development (n = 171) and a validation cohorts (n = 113). We had first identified PF on development cohort and validated then on the validation cohort. Then, we analysed the impact of chemotherapy after stratification to these PF. Of 284 patients, 250 were men, and median age was 59 (range, 37−86). Main metastatic sites