Human Gene Therapy Subcommittee - July 29-30, 1991 
achieve high-level, long-term TK expression in vivo; (2) to kill the tumor cells, thereby 
reducing the tumor burden; and (3) to immunize the patients against their own tumors. 
The retroviral vector to be used is the STK. This vector contains both the neomycin 
resistance gene and the HSV-TK gene. Ganciclovir is a nucleoside analog that is 
preferentially phosphorylated by the HSV-TK gene. The triphosphate form is toxic to 
the cells and once phosphorylated, it cannot cross cell membranes. 
Dr. Freeman presented data that showed a dose-response of the murine fibrosarcoma 
cell line, k-balb, to ganciclovir. As the concentration of ganciclovir was increased, the 
TK+ cells were killed and the control cells were unaffected. To determine the effects 
of TK+ tumor cells in vivo, mice were injected with cells and then treated with 
ganciclovir. Subcutaneous ganciclovir treatment prevented tumor development. Day 
five treatment resulted in tumor regression. When mixtures of TK+ and TK- cells 
were injected subcutaneously into mice and at least 50% of the injected cells were 
TK+, there was complete tumor regression. Intraperitoneal injections of the cell 
mixtures had the same effect, suggesting that these cells do not have to be in direct 
contact. In vitro experiments showed that when TK+ and TK- tumor cells were mixed 
in culture and subsequently treated with ganciclovir, all tumor cells were eliminated. 
Data indicate a killing effect of the TK+ cells. This effect is observed with a relatively 
low number of TK+ cells. 
Dr. Freeman presented data concerning the effects of TK+ cells on TK- tumor cells in 
the presence of ganciclovir. The cells died by either necrosis or apoptosis. 
Dr. Freeman proposed that dying TK+ cells break into vesicles that carry either toxic 
metabolites of ganciclovir or the TK enzyme. These vesicles are phagocytized by 
neighboring cells. Fluorescent activated cell sorting (FACS) and microscopy data show 
that the killing effect of TK+ cells on TK- cells may be due to a transfer mechanism 
such as cell vesicles. 
Dr. Freeman described murine experiments designed to determine the effects of the 
treatment on pre-existing tumor. TK+ k-balb murine fibrosarcoma cells were injected 
intraperitoneally into mice with pre-existing tumor, and increased survival was 
observed. When the animals died, tumors were located at the subcutaneous sites 
because these cells were not exposed to the intraperitoneally injected TK^ cells. In 
patients attempts will be made to inject cells that come in contact with all ovarian or 
intraperitoneal tumor cells. Patients will be selected who have tumor confined to the 
peritoneal cavity. The purpose of these experiments and the Phase I study is to 
establish a dose for the treatment. It is difficult to extrapolate from an animal to a 
human model. The mice may only survive 20 days, whereas patients may survive many 
months or years. 
Dr. Freeman said that these experiments were performed using a TK-modified ovarian 
[900] 
Recombinant DNA Research, Volume 14 
