In order to develop a feasible model system for initial approaches to cancer gene therapy, our laboratories have used an estrogen-dependent nude mouse breast cancer model as preclinical data for this proposed Phase I trial infusing a retroviral vector into malignant effusions from breast cancer patients. This strategy employs an MMTV promoter-based retroviral vector which produces antisense RNA designed to inhibit the c-fos gene. The preclinical animal data demonstrates that: 1) XM6:antifos retroviral vector infects 30-70% of ex vivo transduced MCF7 human breast cancer cells, 2) results in cell-specific induction of anti-fos RND (cancer cells express the MMTV-regulated anti-fos RNA but peritoneal cells do not), 3) inhibits c-fos mRNA 70-90% within tumor cells, 4) inhibits tumor formation and tumor growth rate in nude mice by 90%, 5) intraperitoneal infusion of XM6:antifos vector into established intraperitoneal tumors results in inhibition of tumor growth with induction of apoptosis and differentiation. Preclinical studies of human malignant effusions demonstrate that we are able to ex vivo transduce 30-70% of cancer cells within an effusion and that the antisense transcripts are expressed in these transduced cells. These studies taken together indicate that we have the tools and model systems to evaluate mechanisms of tumor inhibition and to initiate patient studies. We hypothesize that XM6:antifos retroviral vector inhibits tumor growth by interfering with c- fos transcriptional target genes which disrupt paracrine factors and/or estrogenic signalling events which are important for tumor maintenance. This vector may provide a non-toxic breast cell specific antitumor strategy in patients with mammary carcinoma. To test this hypothesis we propose these aims: 1. To perform a Phase I Clinical Trial of XM:antifos retroviral vector administered into malignant ascites or cytology-positive pleural effusions. To examine specificity and rate of XM6:antifos transduction, local and systemic toxicities, presence of circulating retroviral vector, and clinical efficacy. 2. To study specific cellular effects of XM6:antifos on breast tumor cells in situ, namely proliferative rate, apoptosis, and levels of antisense and endogenous c-fos mRNA. In addition, to examine the effect of exogenous steroids on MMTV regulation and antisense fos expression in tumor cells as well as the stability of XM6:antifos in body fluids and serum and the patient's immune response to it.