mide, or fludarabine plus cyclophosphamide intraperitoneally on Day 7 after a palpable tumor was found on the hindquarter. Within each chemotherapy subgroup, mice were randomized to receive either irradiated B16 melanoma cells (1  10 6 ) or RPMI cell culture media on Day 14. Serial tumor measurements were taken until Day 32 or until the animal became moribund. Tumor volume was calculated by standard method. Blood was taken for lymphocyte measurement prior to vaccination on Day 14. Immu- noadjuvants (GM-CSF, IL-2, and CpG) were added to the vacci- nation regimen in subsequent experiments. Statistical analysis was performed using chi-square test and a Student’s t-test. Re- sults: Lymphodepletion was highest in the fludarabine plus cy- clophosphamide group: 1.1 k/ul absolute lymphocyte count com- pared with 3.6 k/ul (Cyclophosphamide), 9.1 k/ul (fludarabine), 13.1 k/ul (control), (p=0.0001). Interestingly, tumor growth in the irradiated B16 melanoma cells alone following lymphodepletion group was significantly greater on Day 32 than the group receiv- ing saline following lymphodepletion (p=0.007). However, tumor growth was significantly reduced following addition of either IL-2 or CpG but not GM-CSF to the vaccination regimen compared to controls; (p=0.019, p=0.002, and p=0.07, respectively). The com- bination of lymphodepletion plus irradiated B16 and IL2 or CpG had significantly stronger anti-tumor effect than irradiated B16 plus IL2 (p=0.013) or irradiated B16 plus CpG (p=0.02), respec- tively. Conclusions: Thus, the present study demonstrates that pre-vaccination lymphodepletion with fludarabine and cyclophos- phamide treatment can enhance the efficacy of active specific immunotherapy. However, this effect is dependent on the type of immune adjuvant used in the vaccination strategy. P238. MUCOSAL VACCINATION PREVENTS SPONTANE- OUS COLORECTAL CARCINOMA AND INDUCES LONG-TERM IMMUNITY IN MICE. S. Kim-Schulze, D. Moroziewicz, A. Wainstein, B. Elliott, H. L. Kaufman; Colum- bia University, New York, NY The goal of tumor vaccination is to induce a long-lived antigen- specific T cell response that can prevent the development of spon- taneous cancers or induce regression of established tumors. Pre- vious efforts targeting carcinoembryonic antigen (CEA) by recombinant poxvirus vectors have shown some promise in gen- erating CEA-specific T cell responses but this has not yet resulted in significant clinical responses in patients with cancer. The lack of therapeutic activity may be associated with anti-vector neutral- izing antibody titers, the weak antigenicity of CEA and the lack of priming in an appropriate microenvironment. To overcome these obstacles, we hypothesized that direct vaccination through the rectal mucosa might provide a more physiologic method of boost- ing CEA immunity and prevent vector-specific neutralization. We tested this hypothesis using direct intra-rectal vaccination with a vaccinia virus expressing CEA in CEA transgenic mice engineered for spontaneous development of colorectal tumors (CEA/Apc1638 mice). Vaccinia and CEA-specific antibody titers were measured by ELISA and T cell responses were determined through thymi- dine proliferation and standard chromium release assays using T cells derived from spleens and Peyer’s patches. The effect on tumor growth was determined by vaccinating CEA/APC1638 mice at 2 months of age and determining the number of adenomas and carcinomas at 6 months of age. We found that intra-rectal vacci- nation induced strong CEA-specific immunity as measured by antibody titers and T cell assays in both the systemic (spleen) and local (Peyer’s patches) compartments. These responses corre- sponded with a significant reduction in the number of adenomas and carcinomas in vaccinated mice. This data supports the impor- tance of the mucosal route for vaccinating against colorectal can- cer and suggests that mucosal immunization may be an effective strategy for boosting systemic immunity to CEA or other tumor- associated antigens. P239. CCR5 POLYMORPHISM AS A PREDICTOR OF CLINI- CAL RESONSE TO HIGH-DOSE IL-2 THERAPY IN MELANOMA AND RENAL CELL CARCINOMA. C. Petrulio 1 , S. Kim-Schulze 1 , G. Deraffele 1 , D. Morozowiecz 1 , J. Mitcham 1 , D. Schrama 2 , J. Becker 2 , H. L. Kaufman 1 ; 1 Colum- bia University, New York, NY, 2 Julius Maximillus Univer- sity, Wurzburg, Germany Introduction: High-dose Interleukin-2 (IL-2) therapy confers a 16-20% clinical response in patients with malignant melanoma (MM) and renal cell carcinoma (RCC) through activation and proliferation of tumor-specific T lymphocytes. IL-2 has also been shown to induce regulatory T cells (Tregs), whose function is to suppress T cell acti- vation and proliferation, which may explain, in part, why the clinical response rate to IL-2 is not better. We have previously shown that patients who respond clinically to high dose IL-2 demonstrate sig- nificant decreases in Treg levels (9.0% to 1.2%) compared to non- responders (7.72% to 16.7%). The reasons for the heterogeneity in clinical response to high-dose IL-2 might be related to polymor- phisms in particular immune response genes. The chemokine recep- tor 5 gene (CCR5) is expressed on T cells and induces migration and activation of lymphocytes. A single amino acid substitution at posi- tion 32 of the CCR5 polypeptide sequence has been identified (CCR532) as a predictor of outcome for patients with non-Hodgkin’s lymphoma.. The purpose of this study was to correlate the presence of the CCR532 polymorphism with clinical responses to high-dose IL-2 therapy. Methods: Serum from 57 patients with MM or RCC treated with high-dose IL-2 was analyzed using nested PCR to iden- tify CCR5 polymorphisms. Patients were grouped into those with the with wildtype CCR5 and compared to heterozygote and homozygous CCR532 carriers. Clinical responses were determined by standard RECIST criteria. The presence of the CCR5 mutation was correlated with clinical response using Log-rank test and relative risk. Results: There was no effect of CCR532 on overall survival. The presence of the CCR532 mutation was associated with a trend toward de- creased survival in patients treated with high-dose IL-2 (P=0.158). The relative risk of wild type to CCR532 is 0.60 with 95% CI (0.28, 1.26). Conclusion: These data suggest that the presence of the CCR532 polymorphism trended toward a worse response to high- dose IL-2 therapy. Given the small sample size in the current study, a larger study may be required to better detect a significant differ- ence and we are pursuing this with a larger multi-institutional cohort. The detection of polymorphisms in CCR5 and other immune responsive genes may be molecular predictors of response to IL-2 and other forms of immunotherapy in patients with MM and RCC. P240. GENETICALLY MODIFIED MICROVASCULAR FREE FLAPS EXPRESSING IL-12 RESULT IN LOCALIZED TUMOR REGRESSION WITHOUT SYSTEMIC TOXIC- ITY. C. Hamou, M. P. Dempsey, J. Michaels, V, S. Ghali, G. C. Gurtner; Stanford, Stanford, CA Introduction: The widespread clinical use of cancer gene therapy has been prevented by the inability to deliver high levels of local transgene expression while limiting the systemic host response. Our lab has developed an ex vivo technique to transduce microvascular free flaps with a recombinant adenoviral vector which expresses the anti-tumor cytokine Interleukin 12 (IL-12). Using this approach, we are able to target treatment specifically to the required area without eliciting the known toxic effects associated with systemic adminis- tration of IL-12. Methods: Superficial inferior epigastric flaps were dissected in male Fischer rats and perfused with an adenoviral vector expressing the cytokine IL-12 (Ad.IL-12) following division of the pedicle. After a one hour incubation period, the flap was flushed with saline to remove unincorporated virions and a microvascular re-anastomosis performed. 110 6 beta-hCG secreting MADB-106 tumor cells were injected into the underlying muscle. Tumor growth was monitored directly with calipers and indirectly by rising levels of 327 ASSOCIATION FOR ACADEMIC SURGERY AND SOCIETY OF UNIVERSITY SURGEONS—ABSTRACTS