Comparative studies of a retrovirus versus a poxvirus vector in whole tumor-cell vaccines.

A number of experimental and clinical studies have used retroviral vectors to express transgenes in whole tumor-cell vaccines. Recently, poxvirus vectors such as vaccinia or avipox have been used toward this goal. The studies reported here compare for the first time the use of a retroviral vector versus a poxvirus vector (vaccinia) in whole tumor-cell vaccines. The transgene used was the T-cell costimulatory molecule B7-1, and the tumor was the weakly or nonimmunogenic MC38 murine colon adenocarcinoma. Recombinant retrovirus (R-B7) and the recombinant vaccinia (V-B7) induced equivalent expression of B7 on the surface of the carcinoma cell. Using live whole-tumor cells as vaccine, cells transduced via recombinant retrovirus (MC38/R-B7) and recombinant vaccinia (MC38/V-B7) equally induced protection against challenge by native MC38 cells 14 days later. Upon rechallenge with native MC38 cells 40 days later, however, the MC38/R-B7 vaccine was shown to be less effective than the MC38/V-B7 vaccine. Similar results were obtained when the tumor cells were irradiated prior to administration. When comparative studies were conducted in which X-irradiated tumor-cell vaccines were given to mice bearing experimental lung metastases, the MC38/V-B7 vaccine was shown to be significantly (P = 0.0351) more effective than the MC38/R-B7 vaccine. Additional studies were carried out in mice that had received vaccinia virus previously. Again, the X-irradiated MC38/V-B7 vaccine was statistically (P = 0.024) more effective than the MC38/R-B7 vaccine in the elimination of metastases. When the naïve and vaccinia-immune mice for each vaccination group were combined for meta-analysis (n = 16), the MC38/V-B7 was significantly more effective than the MC38/R-B7 in the treatment of pulmonary metastases (P = 0.0014) in this model. These studies thus demonstrate for the first time that a whole tumor-cell vaccine (either live or X-irradiated) containing a vaccinia transgene is at least as efficient, and sometimes more efficient, in inducing antitumor effects compared with the same vaccine using a retrovirus to express the transgene. The implications for the clinical applications of such approaches are discussed.