Twelve hours after viral injection the animals were treated intraperitoneally twice daily 
with either GCV or PBS for 6 consecutive days. Four treatment groups of 5 animals each 
were established: (1) ADV-bgal and PBS (TK-G-); (2) ADV-bgal plus GCV (TK-G+); (3) 
ADV-tk plus PBS (TK+G-); and (4) ADV-tk plus GCV (TK+G+). The GCV treatment groups 
were treated intraperitoneally with GCV at a dose of 125 mg/kg body weight twice daily for 3 
days and then at 100 mg/kg for another 3 days. The other groups were treated with PBS. 
One animal in the TK+G+ group died 15 days after C 6 injection and 2 days after the 
completion of the GCV treatment. The remaining animals were sacrificed 20 days after 
tumor implantation and fixed by intracardial perfusion of 4% paraformaldehyde in PBS. The 
brains were removed, cryoprotected in 21% sucrose in PBS, and frozen in OCT. Ten micron 
coronal sections were taken from the tumor implantation site and stained with hematoxylin 
and eosin. Image analysis was performed using Bioscan Optimas software (Edmonds, 
WS). All of the animals in the TK+G-, TK-G+, and TK-G- groups had large cerebral tumors 
(Fig. 5A) that had compressed and infiltrated adjacent brain parenchyma. The tumors were 
characterized by hypercellularity, nuclear pleomorphism, and focal vascular proliferation 
(Fig. 5B). Necrosis was not observed in any of the tumors but some had microhemorrhages. 
In contrast, 2 of the 4 animals in the TK+G+ group were tumor free (Fig. 5E) and 2 animals 
had small residual gliomas surrounding the injection tract (Fig. 5C). In animals without 
residual tumor, a small collection of macrophages and erythrocytes was present at the 
tumor injection site (Fig. 5F). Computerized morphometric analysis of the maximal cross- 
sectional area of the tumor revealed a 23-fold difference between the mean of the 
experimental group (TK+G+) and the mean of the 3 control groups (Fig. 6). There was a 
slight, but statistically significant (p=0.008), reduction in the mean cross-sectional area of 
the TK-G+ group when compared to the TK-G- and TK+G- groups, suggesting that GCV 
treatment may have had an inhibitory on tumor growth (Fig. 6). 
Assuming that the tumors were spherical, the TK+G+ animals had tumors with a mean 
volume of 0.06 mm 3 , whereas the control groups had a mean tumor volume of 32.1 mm 3 . 
Based upon counting nuclei in representative sections and computing mean nuclear density, 
the mean number of tumor cells in the TK+G+ group was 6 x 10 4 , and the number of tumor 
cells in the control groups was 3.2 x 1 0 7 , indicating that the gene therapy resulted in a 2.73 
log tumor cell reduction. 
HSV-tk Transduced Normal Brain Cells are Refractory to GCV Toxicity. 
To determine if ADV-tk and GCV had any deleterious effect on normal brain tissue, we 
examined coronal sections remote from the tumor site in each animal. In addition, ADV-tk 
and GCV was administered to 2 nude mice without tumors and their brains were examined. 
No necrosis, demyelination, loss of neurons, ependymal damage or inflammatory response 
was observed remote from the tumors and/or needle tracts. 
IV. ADENOVIRAL-MEDIATED TREATMENT OF 9L EXPERIMENTAL GLIOMAS IN 
FISCHER RATS 
In the previous experiments using C 6 gliomas in nude mice we demonstrated 
significant tumoricidal effect of ADV-tk/GCV treatment in the absence of a complete cellular 
immune system. However, despite a 500-fold reduction in tumor volume, 80% of the ADV- 
tk/GCV treated mice had residual tumors. Hence, we repeated these experiments usina an 
Recombinant DNA Research, Volume 20 
[725] 
