APPENDIX C. 
Scientific Abstract 
GENE THERAPY FOR RECURRENT PEDIATRIC BRAIN TUMORS 
Brain tumors arc the second most common cancer in children. Supratentorial malignant 
gliomas account for 10-12% of childhood brain tumors. Although short-term survival 
appears to be superior to adults with malignant gliomas, no more than 20% of children with 
these tumors survive beyond 3 years. Associated supratentorial embryonal tumors 
(pincoblastomas, cerebral neuroblastomas, supratentorial PNET's) comprise an additional 
5% of brain neoplasms in children, demonstrating survival rates similar to the malignant 
gliomas. The current proposal will study the feasibility of utilizing a novel approach of 
gene transfer therapy in children with recurrent or progressive malignant supratentorial 
tumors. 
Building upon the prcclinieal and phase I adult studies at NIH (Dr. Edward Oldfield, 
N1NDS), a phase I trial will utilize NIH 3T3 producer cells to deliver a retroviral vector 
(GITKSVNa) to transfect proliferating brain tumor cells with the ganciclovir-susceptibility 
gene associated with the herpes simplex enzyme thymidine kinase (HS-tk). The brain is 
felt to be the ideal anatomic location for gene transfer based upon the selective proliferating 
capacity of the neoplastic cells (in contradistinction to the post-mitotic parenchymal brain 
ceils) and relative "immunologic privilege" allowing exposure of the producer cells for 
vector integration. Subsequent delivery of ganciclovir has been shown to produce selective 
cell kill of the transfected neoplastic ceils. 
A phase 1 trial in children (3-21 years old) will allow assessment of feasibility and potential 
unique toxicitics following the phase 1 adult study at NINDS. The producer cells will be 
instilled via multiple stereotactically placed cannulas, preplanned to homogeneously deliver 
a total of between 1 x 108 and 2 x 109 cells within the tumor volume. Seven days 
following the operative procedure, the patient will start a 14 day course of ganciclovir 
administration. Patients will be monitored for CNS, hematologic, renal, or other toxicities. 
Measurement of tumor response will include serial MRI studies, including an 
investigational analytical assessment of dynamic enhancement that appears to correlate with 
metabolic potential, and MR spectroscopy, an evolving technique correlating measured 
bicxhcmical changes with tumor viability. 
A total of 6 patients will be accessioned to the phase 1 study to allow adequate evaluation of 
potential toxicities. Guidelines include both parental consent and the child's assent. The 
investigators propose to follow the current trial with phase II studies including (1) 
traditional tumor type-specific evaluation of response in children with recurrent malignant 
supratentorial tumors, and (2) an "upfront" phase II study in children with supratentorial 
malignant gliomas. 
[302] 
Recombinant DNA Research, Volume 18 
