The laboratory is developing viral vectors which can be administered systemically to a tumor-bearing host, and selectively target tumor cells for the expression of genes for enzyme/prodrug therapy. We have shown that a vaccinia virus which has a deletion in the thymidine kinase gene specifically replicates in tumor tissue in vivo and expresses high levels of a marker gene selectively in tumor cells. This selective expression can be translated into improved survival and cures when the virus expresses either the cytosine deaminase gene (converts nontoxic 5-FC to toxic 5-FU) or purine nucleoside phosphorylase gene (converts nontoxic 6-MPDR to toxic 6-MP) in a murine model of hepatic metastases from colorectal cancer. In addition, treatment with the prodrug can rescue the mice from toxic effects of the virus. We have now isolated a vaccinia mutant which has a deletion in both the thymidine kinase and vaccinia growth factor (VGF) genes. It has previously been shown that a deletion in VGF markedly attenuates the virus in normal tissues, but replication in transformed cells is unaffected. This double mutant may allow for improved tumor selectivity and improved therapeutic results. In parallel with the study of new virus is the development of more potent and specific enzyme/prodrug systems. We are constructing a virus which expresses a fusion gene combining a secreted B-lactamase and the scFv against a melanoma antigen. B-lactamase can convert nontoxic cephalosporin conjugated chemotherapeutic agents into toxic agents. This protein may expand the toxic effect beyond the cells expressing the gene, while maintaining specificity for melanoma. Because neutralizing antibodies to vaccinia in patients immunized against small pox may prevent viral infection and gene expression, we are exploring other pox vectors which may have similar selectivity without cross reactivity. We have shown that the yaba like disease (YLD) virus can be grown in high titres in CV-1 cells, similar to vaccinia, and that it infects human cell lines. We have shown that it does not cross react with vaccinia antibodies. We have sequenced the YLD thymidine kinase gene and created a recombinant virus expressing a marker gene (green fluorescent protein). We have found that this recombinant YLD infects, replicates, and efficiently expresses genes in certain human tumors. We are in the process of testing this virus as a gene delivery vector in vivo with human tumors in nude mice. We have also investigated transcriptional targeting of melanoma cells using a human tyrosinase promoter combined with a tandem murine melanocyte enhancer element, and have demonstrated high levels of gene expression specifically in human melanoma cells. When this system is used to express the purine nucleoside phosphorylase gene, specific melanoma cell killing was demonstrated. This transcriptional control has been incorporated into an adenovirus vector expressing luciferase or purine nucleoside phosphorylase, and melanoma specific expression is accomplished. We have now constructed a virus which uses the melanoma specific transcriptional elements to express the essential adenoviral E1A gene (in order to achieve melanoma specific viral replication) followed by a major late promoter (requiring E1A for activation) expressing luciferase or purine nucleoside phosphorylase. We hope that this will allow for tumor specific replication and specific expression of the gene of interest. We have identified 5 human melanoma cell lines that infect well with adenovirus, and will test our new virus in vivo against these tumors. As part of the metabolism sections effort in regional therapy for metastatic cancer, we have developed hyperthermic peritoneal perfusion for peritoneal carcinomatosis. This has been developed in a phase I trial with cisplatin delivered intraoperatively in the context of surgical debulking and hyperthermia. Extensive cisplatin pharmacokinetics have been performed, describing this system. This has been extended to include post-operative dwell chemotherapy with paclitaxel and 5-FU, and a phase I trial escalating these agents in the early post-operative period is nearing completion. We have included histologies ranging from gastric cancer to peritoneal mesothelioma. The most promising results have been seen with peritoneal mesothelioma, where many patients have had complete resolution of ascites, prolonged disease free intervals, and responses to re-treatment. A phase II trial using these agents at the maximum tolerated dose is now under review. We provide surgical consultation to the endocrine and urology services, and independently developed and prospectively maintain databases on parathyroidectomy and pancreatic neuroendocrine tumor resections. We are investigating the utility of rapid intraoperative parathyroid hormone and insulin levels in the surgical decision making during parathyroidectomy and enucleation for insulinoma, respectively. We have also developed complex laparoscopic procedures for endocrine surgery, including laparoscopic adrenalectomy and pancreatectomy. These procedures markedly shorten hospital stay and recovery time compared to the more standard open procedures. Along with this, we have utilized laparoscopic ultrasound to screen for pancreatic neuroendocrine tumors and liver metastases.
