Gene Therapy for Meningeal Carcinomatosis 
producer cells with or without GCV administration did not result in any observable toxicity, and 
histology showed no evidence of meningeal inflammation or parenchymal injury. Intraventricular 
injection of HStk and BAG producer cells resulted in transduction of the choroid plexus with some 
injury to the choroid plexus after GCV treatment. 
Intraventricular administration of HStk (both clone 53 and clone 7) and B-galactosidase 
retroviral vector-producer cells in rhesus monkeys did not produce adverse systemic or 
neurological effects, with or without GCV treatment. Necropsy data demonstrate transduction and 
injury of the choroid plexus in GCV-treated animals, however, this was not of clinical 
significance. Vector particles were detected along the whole length of the spinal canal (from the 
craniocervical region to the lumbar spine) 24 h after intraventricular cell injection, but not 
thereafter. These studies suggest that intraventricular and intrathecal injections of HStk vector- 
producer cells and subsequent GCV treatment are safe The limited period during which retroviral 
titers could have been detected in the CSF suggest that the vector-producer cell dose will need to be 
increased or multiple injections of producer cells given for optimal treatment 
1.5 Clinical Experience 
Experience with Retrovir al-mediated Gene Transfer Into Humans 
A number of retroviral-mediated gene-marking ( e.g. transduction of marrow cells in 
patients undergoing bone marrow transplantation) and therapy experiments (e.g. insertion of 
cytokine genes into autologous tumor in an effort to vaccinate patients) are currently being 
performed in humans. No untoward side effects related to retroviral-mediated gene transfer have 
been observed in any patient. 
Eight patients with malignant brain tumors (6 recurrent glioblastoma, 1 multifocal 
glioblastoma with intracerebral metastases, and 1 metastatic malignant melanoma) were enrolled 
in a trial of gene therapy to evaluate toxicity and antitumor efficacy in human tumors. All 
patients had failed standard therapy, which included surgery and radiotherapy, and, in some 
patients, chemotherapy. CT or MRI-guided stereotaxis was used to distribute injections of 
murine HStk vector-producer cells throughout the contrast-enhancing portion of the brain 
tumors. Each patient received 5x1 0^ to 1.0x10^ producer cells in 5-10ml (1.0x10^ cells/ml). 
Seven days after cell injection, ganciclovir (5 mg/kg, BID) was administered iv for 14 d. MR 
scans were performed frequently at the early stages of the treatment and at 2 wk intervals 
thereafter. PET scans, using FDG, were performed before and immediately after treatment in 
patients in whom PET identified a hypermetabolic tumor before treatment. Five of the 8 patients 
had evidence of an antitumor response. This was characterized by a decrease in tumor size 
(n=3) and/or acute changes in its consistency (micro and macrocysts, reduced gadolinium 
enhancement, n=5). After the initial response, a transient increase in size of the region of gd- 
enhancement occurred in 2 patients 2 wk after finishing treatment without any accompanying 
symptoms. Enhancement of the injection tracts was also evident. This spontaneously reversed 
2-3 wk later (4-5 wk after completion of treatment) and is believed to represent a transient, self- 
limiting, inflammatory response. Three patients with rapidly enlarging glioblastoma progressed 
despite treatment. Biopsies obtained from 2 of these patients showed viable tumor cells, focal 
necrosis, an inflammatory response, and mild lymphocytic infiltrates. The injection of the 
murine cells and ganciclovir therapy were well tolerated by all patients and no adverse effects 
were observed. 
[600] 
Recombinant DNA Research, Volume 18 
