BACKGROUND: Cancer gene therapy expressing specific levels of multiple genes has not been adequately tested. This study investigates the antitumor effects of recombinant vaccinia viruses (recVVs) that express predictable levels of single and multiple cytokines in a passive immunization murine breast cancer (C3HBA) model. METHODS: Seventeen recVVs encoding different cytokine combinations under weak and strong VV promoters were constructed. Cytokine production was measured in vitro by enzyme-linked immunosorbent assay. Mice were immunized with infected C3HBA cells and rechallenged 10 days later with 10(6) parental tumor cells. Controls were treated with saline or recVVs not expressing cytokines. Tumors were measured for 24 days. Data were analyzed using Fisher's exact test and the Breslow-Gehan-Wilcoxon test. RESULTS: recVVs encoding multiple cytokines induced secretion of each cytokine at predictable levels corresponding to VV promoter strength. Ten of 10 controls developed tumors by day 11 after rechallenge. recVVs producing large but not small amounts of murine granulocyte-macrophage colony-stimulating factor (GM-CSF) alone resulted in complete protection in all immunized mice (p = 0.0003) for 24 days and disease-free survival (DFS) was significantly prolonged versus controls (p < 0.001). GM-CSF under a weak promoter was also effective in combination with human interleukin-1 beta (hIL-1 beta; p = 0.0220; DFS, p = 0.031) or interferon-gamma (IFN-gamma; p = 0.0037; DFS, p = 0.003). Effectiveness of hIL-1 beta and IFN-gamma vaccines depended on cytokine combinations and not the amount of cytokine produced. IL-2 or tumor necrosis factor-alpha vaccines had no significant antitumor effect. CONCLUSIONS: (a) recVVs can simultaneously express controllable levels of two cytokines; (b) tumor cells modified by recVVs generate systemic antitumor immunity; and (c) strength of immunity appears to be related to the amounts and specific cytokine(s) produced.
BACKGROUND: Cancer gene therapy expressing specific levels of multiple genes has not been adequately tested. This study investigates the antitumor effects of recombinant vaccinia viruses (recVVs) that express predictable levels of single and multiple cytokines in a passive immunization murinebreast cancer (C3HBA) model. METHODS: Seventeen recVVs encoding different cytokine combinations under weak and strong VV promoters were constructed. Cytokine production was measured in vitro by enzyme-linked immunosorbent assay. Mice were immunized with infected C3HBA cells and rechallenged 10 days later with 10(6) parental tumor cells. Controls were treated with saline or recVVs not expressing cytokines. Tumors were measured for 24 days. Data were analyzed using Fisher's exact test and the Breslow-Gehan-Wilcoxon test. RESULTS: recVVs encoding multiple cytokines induced secretion of each cytokine at predictable levels corresponding to VV promoter strength. Ten of 10 controls developed tumors by day 11 after rechallenge. recVVs producing large but not small amounts of murinegranulocyte-macrophage colony-stimulating factor (GM-CSF) alone resulted in complete protection in all immunized mice (p = 0.0003) for 24 days and disease-free survival (DFS) was significantly prolonged versus controls (p < 0.001). GM-CSF under a weak promoter was also effective in combination with humaninterleukin-1 beta (hIL-1 beta; p = 0.0220; DFS, p = 0.031) or interferon-gamma (IFN-gamma; p = 0.0037; DFS, p = 0.003). Effectiveness of hIL-1 beta and IFN-gamma vaccines depended on cytokine combinations and not the amount of cytokine produced. IL-2 or tumor necrosis factor-alpha vaccines had no significant antitumor effect. CONCLUSIONS: (a) recVVs can simultaneously express controllable levels of two cytokines; (b) tumor cells modified by recVVs generate systemic antitumor immunity; and (c) strength of immunity appears to be related to the amounts and specific cytokine(s) produced.