Tumor directed gene therapy for the purpose of destroying cancer cells through replicative "oncolysis" or by intratumoral expression of toxic or immunostimulatory genes requires an efficient, tumor targeted vector. Vectors are limited by inefficient replication in vivo, inefficient tumor targeting, and safety concerns. As a unique approach to addressing these limitations, our laboratory has studied poxviruses as tumor selective replicating vectors. The best in vivo antitumor results achieved to date have been with a mutated WR strain of vaccinia virus. The unique advantage of this strain of vaccinia over other vectors currently being explored for this purpose is the efficiency of in vivo replication. Intradermal injection of 10(6) pfu of the wild type (non-mutated) vaccinia in non-human primates leads to a 108 cm(2) zone of necrosis in 8 days - directly related to cellular destruction from viral replication. We have mutated the virus through insertional deletion of both the thymidine kinase (TK) gene and vaccinia growth factor (VGF) gene. The mutant virus no longer causes destruction of normal tissue, but has completely preserved replication efficiency in tumor tissue and can safely be delivered systematically to successfully treat subcutaneous tumors in mice. Plans are now underway for clinical trials.