Ovarian cancer is the leading cause of death in women with gynecological malignancies in the U.S., with an overall 5-year survival of <50%. Recurrences are often locoregional, involving peritoneum and abdominal organs, and especially for platinum-resistant disease which shows reduced response rates to chemotherapy, improved treatments are needed. We developed a novel strategy using retroviral replicating vectors (RRV) for highly efficient and tumor-selective delivery of prodrug activator (`suicide') genes such as cytosine deaminase (CD), and based on our preclinical findings, first-in-human multi-center clinical trials of RRV-CD suicide gene therapy delivered by local tumor site injection were initiated throughout the U.S. for recurrent high-grade glioma, and have shown highly promising signs of clinical benefit, radiographic responses, and increased survival, and a Phase IIB/III registrational trial is now on-going. Furthermore, based on our recent preclinical data showing that, even with systemic intravenous delivery, RRV is highly restricted to actively dividing tumor cells, without spread to normal tissues in immunocompetent hosts, and results in prolonged survival upon prodrug administration without systemic side effects, FDA recently approved a new clinical trial for intravenous delivery of RRV. Now that there is clinical precedent, it is timely to consider applying systemically or locoregionally administered RRV also to systemic malignancies such as ovarian cancer. Accordingly, here we propose the first preclinical studies to evaluate the feasibility, safety, and efficacy of RRV-mediated suicide gene therapy for treatment-refractory recurrent ovarian cancer. At the cellular level, we will perform gene expression profiling to evaluate whether primary patient-derived ovarian cancer cells express candidate anti-viral restriction factors that might impact future clinical application of this strategy, and empirically test whether RRV replication and suicide gene activity in ovarian cancer cells correlates with gene expression profiling results, using RRV pseudotyped with different envelopes to enable co-infection (Aim 1). Metastatic ovarian cancer will necessitate widespread vector delivery via intravenous or intraperitoneal administration, and both routes will be evaluated for transduction efficiency and biodistribution/safety in ovarian cancer peritoneal carcinomatosis models, using human xenografts in immunodeficient mice and syngeneic models in immunocompetent mice (Aim 2). Next, we will evaluate suicide gene therapy with both the clinical RRV-CD (Toca 511) vector and a newly developed RRV encoding bacterial nitroreductase (NTR), which generates a pro-immunogenic alkylating agent within infected cancer cells, individually and in combination (Aim 3). As tumor-localized prodrug conversion by RRV destroys immunosuppressive tumor stroma, while incurring minimal systemic myelotoxicity and maintaining an intact immune system, we will also test th...