through the Ommaya reservoir or taken directly from culture showed 
tumor regression with GCV treatment. These findings suggest that 
the passage of VPC's through the Ommaya reservoir had no adverse 
effect on their ability to transduce tumor in vivo , and sensitize 
them to GCV. 
G. Effects of CSF and serum on in vitro gene transfer. 
CSF was obtained from 3 patients with CNS malignancies. The CSF 
was placed on ice, spun and frozen. Serum was collected from a 
peripheral vein, placed on ice, spun and frozen. 
The human U251 glioblastoma cell line was placed at 50,000 
cells per well in 6 well plates. Fresh GlBgSvNa.29 supernate was 
harvest from a confluent T-175 flask. 1:1 and 1:10 mixtures of CSF 
(with or without heating for 30 minutes at 56°C) to supernate were 
placed on the U251 cells after aspiration of culture medium. Serum 
was added at 1:4 or a 1:20 dilution (with or without heating for 
30 minutes at 56°C) to supernate and placed on the U251 cells after 
aspiration of culture medium. Protamine (lOu/ml) was added to each 
well. After an 8 hour incubation, the supernate mixture was 
removed and replaced with cell culture medium. The following day, 
the cells were washed, fixed with f ormalin/gluteraldehyde and 
stained with X-GAL. The percent of cells staining blue was 
enumerated with light microscopy. The U251 cell line has no 
detectable background staining using this technique (see Appendix 
E: table 4) . 
H. Pre-Clinical Summary: The pre-clinical findings noted above 
suggest that the direct inoculation of a brain tumor with HS-tk 
retroviral vector-producing cells can mediate complete tumor 
rejection in animals. This in vivo gene transfer methodology 
selectively alters the sensitivity of a tumor cell to systemic 
chemotherapy. Treatment with GCV does not result in widespread 
damage to the host immune system like many forms of chemotherapy. 
Our animal studies have demonstrated no significant toxicity to 
normal brain or any of the proliferating non-CNS tissues, 
suggesting that the implantation of VPC in a brain tumor is not 
associated with non-specific systemic toxicity. 
II. Clinical Data 
Eight patients have been treated with the stereotaxic 
injection of HS-tk VPC and GCV infusion. No evidence of toxicity 
due to cell implantation or GCV administration has been observed 
as of April 1, 1993. 
4. Research Design and Methods 
I. Selection of Patients 
A. Inclusion Criteria 
1. Children with malignant supratentorial astrocytomas who failed 
standard therapy for their disease will be eligible to enter the 
study . 
2. Tumors will be divided into surgically resectable and 
surgically unresectable, as estimated from the pre-treatment 
radiological evaluation. These decisions will be made by the Pis 
in accordance with the standards of care of neurosurgical practice. 
3. Male or non-pregnant female patients 2-18 years of age. 
Patients of child bearing potential must practice strict birth 
Recombinant DNA Research, Volume 18 
[63] 
