The goal of this proposal is to examine the potential for a new generation of regional treatments for intrahepatic cancers using improved methods of delivery and forms of cytosine deaminase/5-flucytosine (CD/FC) in enzyme/prodrug gene therapy. Current gene therapy approaches have at least two critical limitations. The first concerns the ability to generate sufficient cytotoxicity and radiosensitization because of an inadequate ability to kill adjacent cultured tumor cells that are not transduced (""""""""bystander effect""""""""). FU generated from FC has been demonstrated to have a significant bystander effect. Unfortunately, FC is an inefficient substrate for bacterial CD compared to cytosine, the natural substrate. To begin to overcome this limitation we will develop the use of yeast (y) CD for which FC is a far more efficient substrate compared to the previously utilized bacterial (b) CD (Aim 1) and, thus will work at FC levels lower than the bCD in gut bacteria (for lowered systemic toxicity). A second limitation of the current versions of CD/FC system is selectivity. Preliminary data using a clinically relevant animal model for intrahepatic cancer (BD9 rats bearing intrahepatic K12 adenocarcinoma tumors) suggest that even (regional) hepatic arterial infusions of bCD adenovirus produce significant FU levels in the normal liver, high systemic concentrations of PU, and potential normal tissue toxicity. To overcome this limitation, we will construct an adenoviral vector using a tumor selective promoter for yeast CD (Aim 2). Preliminary data suggest that the use of CEA as a promoter significantly enhances the specificity of expression. In addition, we will enhance regional delivery of virus using vascular manipulations (Aim 3), including osmotic (mannitol), mechanical (microspheres) and hormonal (epinephrine) methods, and pretreatment with focal ionizing radiation. Preliminary data show that pretreatment with mannitol or with ionizing radiation can increase tumor CD expression. These improvements will be developed and then tested in therapy trials comparing the most aggressive standard therapy, i.e., hepatic arterial FUdR, with the best """"""""enhanced"""""""" CD/FC gene therapy (Aim 4) using MRI to monitor antitumor effects in vivo. Nude rats will be used in therapy trials and in all longer term studies where immunogenicity could affect results. Although the goal of this project is to understand how to design better regional chemoradiation treatments, our long term objective is to generate information and principles relevant and applicable to effective, systemic, tumor-selective gene therapy.
