Adenovirus-mediated gene therapy for experimental spinal cord tumors: tumoricidal efficacy and functional outcome

BRAIN RESEARCH Brain Research 691 (1995) 76-82

ELSEVIER

Research report

Adenovirus-mediated gene therapy for experimental spinal cord tumors: tumoricidal efficacy and functional outcome Ahmet t~olak a

a,1,

j. Clay Goodman

a,b,

Shu-Hsia Chen c,d, Savio L.C. Woo c,d, Robert G. Grossman a, H. David Shine a,c,*

Department ofNeurosurgery, Baylor College of Medicine, 6560 Fannin Street, Suite 944, Houston, TX 77030, USA b Department of Pathology, Baylor College of Medicine, Houston, TX 77030, USA c Department of Cell Biology, Baylor College of Medicine, Houston, TX 77030, USA d Howard Hughes MedicalInstitute, Baylor College of Medicine, Houston, 17( 77030, USA Accepted 2 May 1995

Abstract

We evaluated the efficacy of adenoviral-mediated gene therapy of experimental spinal cord tumors and the functional outcome after this treatment. Spinal cord tumors were generated in the thoracic region of the spinal cord in Fischer 344 rats by stereotaxic intramedullary injection of 1 × 104 9L gliosarcoma cells. Seven days after tumor cell injection, a replication-defective adenoviral vector carrying the herpes simplex virus thymidine kinase gene (ADV-tk) or a control adenoviral vector carrying the fl-galactosidase gene (ADV-flgal) was injected into the tumors. Beginning 12 h later the animals were treated with the antiviral drug ganciclovir (GCV; 50 mg/kg) or saline twice a day for 6 days. The neurological performance of the animals was assessed during and following treatment. Eighteen days after tumor cell injection, all of the control animals had paraplegia and large tumors. In contrast, no tumors were detected in animals treated with ADV-tk and GCV. In long-term studies, two of the 5 animals treated with ADV-tk and GCV remained tumor-free and remained neurologically intact at 6 months whereas all animals in the control groups became paraplegic within 18 days. Keywords: Adenovirus; Gene therapy; Herpes simplex virus; Thymidine kinase; Spinal cord tumor

I. Introduction

Gene therapy is currently under investigation as a treatment for CNS tumors [1,6,7,14-17,19]. Moolten [11,12] was the first to propose that the transduction of tumor cells with the herpes simplex virus thymidine kinase gene (HSV-tk) and subsequent treatment with ganciclovir (GCV), a guanosine analog, would have tumoricidal effects. HSV-tk phosphorylates GCV which is further phosphorylated by cellular enzymes and incorporated into the newly synthesized DNA of dividing cells, leading to cell death. Non-transduced tumor cells adjacent to transduced cells also die in the presence of GCV. This phenomenon is known as the 'bystander effect'. The mechanism underlying this effect is unknown [1,2,5-8,10,14,16,17,19].

* Corresponding author. Fax: (1) (713) 798-4643. E-mail: [email protected]. 1 Present address: Department of Neurosurgery, Inonu University School of Medicine, Malatya, Turkey. 0006-8993/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0006-8993(95)00616-8

Retroviral or adenoviral transduction of the HSV-tk gene into experimental glial brain tumors, followed by GCV treatment, has reduced and in some experiments has eradicated the tumors [1,6,7,14-16,19]. However, it may be difficult to attain similar therapeutic efficacy in human glial brain tumors because of their greater size and irregular margins when compared to the rodent model. A more suitable therapeutic target in humans may be intramedullary glial tumors of the spinal cord. These tumors are smaller and more circumscribed than are glial brain tumors. Astrocytomas and ependymomas are the most common intramedullary tumors of the spinal cord [4,13]. Patients with these tumors generally have a poor prognosis for complete functional recovery even with intensive surgical and radiation therapy [3,4,13]. Gene therapy has the potential to eradicate these tumors with minimal damage to the surrounding cord. Studies of the efficacy of gene therapy of brain tumors have tended to focus on tumor size and the length of survival after treatment [1,6,7,14,17]. However, the quality

A. ~olak et al. / Brain Research 691 (1995) 76-82

of neurological function after treatment is also of major importance. Therefore, in the present study therapeutic efficacy was assessed by measurement of neurological function, as well as by tumor size and length of survival.

2. Materials and methods

We performed three experiments (see protocols below). The virus vectors, tumor cell lines, tumor generation, adenoviral vector injection, GCV treatment, and neurological assessment were identical in all three experiments. 2.1. Viral vectors

Construction of the replication-defective adenovirus carrying the HSV-tk gene (ADV-tk) under control of the Rous sarcoma virus long-terminal-repeat promoter (RSVLTR) has been described previously [6]. Briefly, the HSVtk gene was inserted into the plasmid pADL.1/RSV that contained the RSV-LTR 1:o generate pADL.1/RSV-tk. Recombinant adenovirus was produced by co-transfecting 293 cells with pADL.1/RSV-tk and a plasmid, pJM17, containing the adenovirus genome [9]. A control adenovirus vector carrying the gene for E. coli /3-galactosidase (ADV-flgal) under control of the RSV-LTR was kindly provided by M. Perricaudet (Institut Gustave Roussy, Centre National de la Rechercher Scientifique, Villejuif, France). Virus titers were determined by plaque assay and stored at - 8 0 ° C in a solution of 10 mM Tris-HC1 (pH 7.0), 1 mM MgCI 2 and 10% glycerol at a titer of 2.8 X 101° pfu/ml. 2.2. Tumor cell culture

The 9L glioma cell line was derived from a tumor, described as a mixed glioblastoma multiforme and sarcoma or a gliosarcoma [20], initially generated by Nmethyl-nitrosourea induction in a CD Fischer rat. The 9L tumor cells (kindly provided by P.J. Tofilon of M.D. Anderson Cancer Center, Houston, TX) were maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal calf serum, 100 U / m l penicillin, and 100 /xg/ml streptomycin in 5% CO 2. Tumor cells were mobilized from the culture dish by adding 0.05% trypsin in ethylenediamine tetra-acetic acid (EDTA) for 3 min. The cells were then collected, centrifuged and suspended in Hank's balanced salt solution (HBSS) at a concentration of 6.6 X 103 cells//zl. Before and after implantation, the cells were counted in a hemocytometer and their viability was determined by trypan blue exclusion. 2.3. Tumor cell injection

Adult female Fischer 344. rats (155-175 g) were used as host animals. The rats were anesthetized with halothane

77

and an intramuscular injection (0.6 ml/kg) of an anesthetic solution consisting of ketamine (42.8 mg/ml), xylazine (8.6 mg/ml), and acepromazine (1.4 mg/ml). Thoracic laminectomies were performed at T4-T5 with the aid of a surgical microscope. The animals were then placed into a stereotaxic frame. A syringe with a 26-gauge needle attached to the manipulating arm of the stereotaxic apparatus was mechanically lowered under manual control until the tip of the needle was positioned 0.5 mm beyond the dura in the posterior medial sulcus of the spinal cord. Tumor cells (1 × 10 4) suspended in 1.5 /xl HBSS were injected into the sulcus in the middle of the cord section exposed by the T4-T5 laminectomies. The ceils were slowly injected over a 5-min period, and the needle was left in place for another 5 min before being slowly withdrawn. The paravertebral muscles and fascia were sutured with 6 / 0 silk sutures and the skin was closed with clips. 2.4. Virus injection and ganciclovir treatment

Seven days following the tumor cell injection, the animals were anesthetized, placed in the stereotaxic frame and their spinal cords re-exposed at the T4-T5 level. In each animal that received tumor cells, the induced glial tumor was easily visualized on the dorsal surface of the spinal cord as a well-demarcated ovoid mass. Either 5.6 X 10 7 pfu of ADV-tk or 5.6 X 107 pfu of ADV-flgal in 2/~1 of a solution consisting of 10 mM Tris-HC1 (pH 7.0), 1 mM MgC12 and 10% glycerol were injected into the middle of the tumor over a 5-min period using the same procedure as was used to inject the tumor cells. Again, the needle was left in place for an additional 5 min, and then was slowly withdrawn. Before the injection, the needle was coated with a small number of carbon particles < 30 /xm in size to mark the needle tract and the site of virus injection within the tumor. The paravertebral muscles and fascia were closed with 6 / 0 silk sutures and the skin was closed with clips. Beginning approximately 12 h after virus injection (day 8), the animals were given intraperitoneal injections of GCV (50 m g / k g ) or saline twice daily for 6 consecutive days. 2.5. Neurologic examination

Motor and sensory function of the hind limbs was measured before the tumor cell injection, before the virus injection, and 18 days following the tumor injection. In a long-term survival experiment, measurements were repeated weekly for 80 days. Limb movement was scored as: 0, no response; 1, reflex extension only; 2, reflex withdrawal only; 3, appropriate complex movement, but weak; or 4, normal. Sensory examination was rated as: 0, no response; 1, withdrawal only; or 2, appropriate movement. Functional assessment was measured with the inclined plane technique described by Rivlin and Tator [18]. The

A. ~olak et al. /Brain Research 691 (1995) 76-82

78

10

maximal angle of a plane to the horizontal at which the rat could maintain a stable stance for 5 s was recorded.

8 A ¢,1

2.6. Histologic and morphometric analysis

E

The animals were anesthetized and fixed by cardiac perfusion with 100 ml of phosphate-buffered saline, pH 7.4 (PBS) containing 10 U / m l heparin, followed by 200 ml of 4% paraformaldehyde in PBS. The spinal cords at the treatment site were removed, placed in 4% paraformaldehyde in PBS for 24 h, then cryoprotected at 4°C in 21% sucrose in PBS for 2 4 - 4 8 h, mounted in OCT, frozen and sectioned at 10 /~m on a cryostat. The sections were mounted on gelatin-subbed slides and stained with hematoxylin and eosin. Computer-assisted morphometry was used to measure the tumors at the site that contained the largest cross-sectional area. Treatment effects were compared using the S T A T A ® statistical package (Stata Corp., College Station, TX).

3. Results

3.1. Experimental spinal cord tumors: characteristics, growth and effects on neurological function Eight days after injecting 1 X 104 9L cells into the posterior medial sulcus, a discrete tumor mass invariably formed that deflected the structures within the cord (Fig. 1). At 8 days the borders of the tumors were well-defined and the tumors resided entirely within the cord parenchyma. In some instances, 9L cells appeared to infiltrate the adjacent normal tissue along the perivascular spaces. No tumor cells were observed distal to the primary site in the parenchyma, central canal or subdural space. As the tumor

.....

i;iii~i=!!i?i i :i ,i

/::~i~:....

4~'::~1i!~]]!~!~;~:~::~:~ ; ~?%:T

.:;::~:M:~ ~

.~-

: :,: ~

0

4

E

=1

I-,- 2

0

~ 4

8

12

16

Days

Fig. 2. Spinal cord tumor size at 4, 8, 12, and 16 days after 9L tumor cell injection. The data represent largest cross-sectional tumor size and were pooled from 3 animals at each day. Error bars = S.E.M.

increased in size it grew out of the hole formed by the laminectomy to form a cauliflower-shaped mass outside the vertebral column that remained connected to the primary tumor by a narrow pedicle (Fig. 3c). The progression of these experimental tumors was observed by measuring their cross-sectional area measured 4, 8, 12, and 16 days after tumor cell injection (3 animals each time point; Fig. 2). Hind limb function was measured using the inclined plane technique as the tumors progressed. At 8 days after tumor cell injection, although the tumors were approximately 0.7 mm 2 in cross-sectional area, decrement in neurologic function was observed. However, by 16 days after 9L injection the tumors were approximately 5.4 mm 2 in cross-sectional area and all animals were paraplegic with both motor and the sensory grades equal to 0. The inclined plane angle for a stable

:: =:~=== ii;!~!:i:!¸= !=!!i ~==

/:=:: = : jill!i/~:~ :~

Z: ~ •

Fig. 1. Photomicrograph of an experimental spinal cord tumor 8 days after 9L tumor cell injection. The spinal cord was sectioned longitudinaly approximately 0.1 mm lateral to the central canal. The tumor (t) is well defined and is deflecting white mater (w) and gray mater (g) of the spinal cord. Bar = 0.5 mm.

A. ~olak et al. / Brain Research 691 (1995) 76-82

79

Fig. 3. Gross morphology of rat spinal cors 18 days after 9L tumor cell injection, a: Sham-operated spinal cord. b: Spinal cord after tumor cell injection and subsequent treatment with ADV-tk + GCv. c: Control spinal cord with a large tumor that has grown out of the spinal cord. Injection sites in a and b are marked by a charcoal spot (arraws). Bar = 0.5 cm.

stance had decreased f r o m a control v a l u e o f 60 degrees to 2 5 - 3 0 degrees.

3.2. Tumoricidal efficacy o]'ADV-tk + G C V treatment T u m o r i c i d a l efficacy o f a d e n o v i r u s - m e d i a t e d transduction o f H S V - t k and G C V administration was tested in tumor bearing rats that w e r e treated with either A D V - t k

and G C V (experimental group) or A D V - f l g a l and G C V (control group). A t u m o r - f r e e s h a m group w a s included in the analysis that w e r e not injected w i t h 9 L cells but w e r e treated w i t h A D V - t k and G C V . Each group consisted o f 5 animals. A l l animals w e r e sacrificed at 18 days after t u m o r cell injection. The spinal cords o f all animals w e r e e x a m ined, and the tumors in the e x p e r i m e n t a l and control groups w e r e measured. Gross e x a m i n a t i o n o f the spinal

Fig. 4. Photomicrographs of cross-:;ections of spinal cords 18 days after 9L tumor cell injection, a: Section through tumor site in an animal that received a control treatment (ADV-/3gal + GCV). Note the large tumor that has compressed the spinal cord and extended out of the spinal canal through the hole form by the laminectomy procedure to form a large tumor outside the vertebral column, b: Section through the site of the former tumor in an animal that received the experimental treatmevt (ADV-tk + GCV). No tumor is visible. Bars = 0.5 ram.

80

A. ~olak et al. / Brain Research691 (1995) 76-82

Table 1 Treatments and neurologic performance of animals in long-term survival experiments 18 days after tumor cell implantation Group Virus Drug Motorscore Sensory score Incline degrees 1 2 3 4

ADV-flgal Saline 0.3+0.4 ADV-flgalGCV 0.3+0.4 A D V - t k Saline 0.6+0.6 A D V - t k GCV 2.8+1.1

0 0 0 1.75+1.0

23-t-3 24-1-4 25+5 53+8

Each value represents the mean and S.D. of 5 animals.

1.0 .~o.8

i

--

"~e 0 6 II.

T

3.3. Long-term survival and neurological performance after ADV-tk + GCV treatment of spinal cord tumors The long-term effects of ADV-tk + GCV treatment of spinal cord tumors on survival and neurological performance was tested. Four groups of tumor-bearing rats (5 each group) were treated as described in Table 1 and tested for neurological performance as described in Section 2. The rats were monitored and if they presented signs of morbidity or paraplegia they were sacrificed and their spinal cords inspected for tumors. Animals in the control groups (groups 1, 2 and 3) became paraplegic within 20 days after 9L cell injection while animals that received ADV-tk + GCV (group 4) survived longer (Fig. 5). One of the group 4 rats died at 24 days after the tumor cell implantation but no tumor was present in the spinal cord of this animal. This animal had a urinary tract infection; the probable cause of death was sepsis. Two rats in group 4 became paraplegic at 31 and 44 days at which time they were sacrificed. Tumors were visible in the epidural and supravertebral spaces of these animals, but not within the parenchyma of their spinal cords. The remaining two animals in group 4 are still alive ( > 270 days) and are neurologically normal. Logrank statistics showed significant ( P < 0.048) difference in the survival of the groups.

I

>~ 0.4

_E

T

t/J 0,2 cords of the rats in the sham-operated group and in the ADV-tk + GCV experimental group showed no pathological findings except a charcoal spot at the entrance of the needle on the dorsal surface of the spinal cord (Fig. 3a and b). The rats in the control group that were treated with ADV-/3gal and GCV had very large tumors (Fig. 3c). On microscopic examination 18 days after tumor cell injection, the tumors observed in the control group were characterized by hypercellularity, nuclear pleomorphism, and mitoses without inflammatory cell infiltration. These tumors had nearly destroyed the spinal cords, and extended out of the spinal canal through the opening in the vertebral column formed by the laminectomy procedure (Fig. 4a). The mean cross-sectional area of the tumors was 6.32 mm 2. In contrast, no tumor cells were seen in the spinal cords of the experimental group that received ADV-tk + GCV (Fig. 4b). Instead, macrophages, lymphocytes, neutrophils, necrosis and hemorrhage were apparent in the area where the tumor had been.

ADV-tk / GCV (n=5)

- ADV-tk I PBS (n=5) . . . . ADV-pgal / GCV (n=5) - - - ADV-I~gal I PBS (n=5)

0.0 2O

30

40

50

260

Days Surviving Fig. 5. Kaplan-Meier survival curve of rats with spinal cord tumors that received control or experimental treatments. One rat in the ADV-tk+ GCV group died from sepsis (S) and 2 that died had tumors (T).

Comparing each control group (groups 1, 2, or 3) individually to the experimental group (group 4) by the WilcoxonGehan test demonstrated that the experimental group's survival was significantly ( P < 0.009) longer than the control groups'. There was no significant difference in the survival of the control groups. The motor and the sensory grades of the animals in groups 1, 2, and 3 gradually decreased as did the inclined plane measurements (Table 1, Fig. 6). In these 3 groups, the spinal cord tumors produced paraparesis and finally paraplegia by approximately 18 days. They all had large spinal cord tumors when they were inspected at necropsy. There were no statistically significant differences in neuro-

,-- 70 ¢1

0

~ 50 "0

~ 4o C

-- 30 0

~

ADV-tk I GCV (n=5) ADV-tk / PBS (n=5) ADV-~gal / GCV (n=5) ADV-~gal / PBS (n=5)

~ 2O

<

0 Tumor Implantation

20

40

Days

60

80

Virus Injection (day 7)

Fig. 6. Assessment of hind limb function of rats with spinal cord tumors after control and experimental treatments. Tumor-bearing animals that received control treatments were neurologicallyintact at day 9 after viral injection but were paraplegic by day 18. The group that received the experimental treatment showed degradation of neurological function on day 18 that progressively got worse until the 3 animals were censored from the analysis because of death (see Fig. 5). Bars = S.D. n = 5 for each group.

A. (olak et al. / Brain Research 691 (1995) 76-82

logical measurements or tumor sizes between these 3 control groups. In contrast, the rats treated with ADV-tk and GCV (group 4) remained neurologically intact at 18 days after tumor cell injection. The rats in group 4 were able to maintain position on an inclined plane angle of 53 + 8.4 degrees and their average motor and sensory scores were 2.8 ___1.1 and 1.75 + 0.96, respectively, at 18 days. The differences in neurologic function between the group 4 and groups 1, 2, and 3 was statistically significant ( P < 0.0001 by ANOVA). At 19 day,; after tumor cell injection the aeurologic performance ot group 4 decreased and became variable and then returned to near normal value at day 52 IFig. 6). This degradation '.and recovery of the neurological function of group 4 reflected the onset of paraplegia and then censorship, respectively, of 2 rats that had spinal cord tumors (Fig. 5) which were removed from the analyses upon death.

4. Discussion These experiments demonstrate that recombinant adenaviral vectors can be used successfully to treat experimental malignant spinal cord tumors. An important feature of HSV-tk transduction and GCV treatment is the 'bystander :ffect' in which neighboring tumor cells that are not :ransduced are nevertheles.s killed. The underlying mechaaism of this phenomenon is not fully understood, but the Following mechanisms have been proposed by investiga:ors: metabolic cooperation via gap junctions [2]; endocy:osis or apoptosis of toxic remnants of cells destroyed by M)V-tk + GCV [8]; cell-mediated immune response 17,12,15]; and disruption of tumor blood supply [16,17]. Ram et al. [16,17] conclude from their observations that ~ISV-tk transduction and subsequent killing of endothelial .'ells of the rapidly growing tumor vasculature by GCV nay reduce the tumor blood supply and thereby lead to :umor regression. In our steady, animals treated with ADV:k and GCV had large hemorrhagic areas in the area where :he tumor had been. Simii[ar hemorrhagic areas were not )bserved in control animals. These observations lend sup)ort to the suggestion that one mechanism of the bystander •.ffect is the destruction of capillaries in the tumor. The use of such a small animal as the rat for studying :he treatment of spinal cord tumors results in a technical )roblem which affected the long-term outcome after treatnent. The spinal cords of tats are so small that tumor cells :an leak from the injection site to the epidural space and :scape ADV-tk transduction. In the short-term tumoricidal ,'xperiments no tumor cell,,; were found in animals treated vith ADV-tk and GCV at 18 days after tumor cell injecion. However, in the long-term experiments, two animals reated with ADV-tk and GCV died at 31 and 44 days after umor implantation and were found to have tumor growth n the epidural and supravertebral spaces away from the ;ite of tumor cell injection.

81

The present experiments provide support for future studies of gene therapy of malignant spinal cord tumors in animals in which the size of the spinal cord and the intramedullary tumor are equivalent to those in humans.

Acknowledgements We thank Mr. Grant Louie and Mr. Guenther Feigl for their expert technical assistance, Dr. M. Perricaudet for the use of the ADV-/3gal vector, and Dr. F. Graham for the cloning vector used to create the ADV construct. We also thank Dr. W. Hamilton for editorial assistance and Dr. C. Contant for statistical assistance. A.C. is supported by an Educational Program Project of Turgut Ozal Medical Center, Turkey. S.L.C.W. is an investigator and S-H.C. is an associate of the Howard Hughes Medical Institute. A portion of this research was supported by a grant from the Texas Higher Education Coordinating Board Advanced Technology Program.

References [1] Barba, D., Hardin, J., Ray, J. and Gage, F.H., Thymidine kinasemediated killing of rat brain tumors, J. Neurosurg., 79 (1993) 729-735. [2] Bi, W.L., Parysek, L.M., Wamick, R. and Stambrook, P.J., In vitro evidence that metabolic cooperation is responsible for the bystander effect observed with HSV tk retroviral gene therapy. Hum. Gene Ther., 4 (1993) 725-731. [3] Boring, C.C., Squires, T.S., Tong, T. and Montgomery, S., Cancer Statistics, 1994, Cancer J. Clin., 44 (1994) 7-26. [4] Cristante, L. and Herrmann, H.D., Surgical management of intramedullary spinal cord tumors: Functional outcome and sources of morbidity, Neurosurgery, 35 (1994) 69-74. [5] Caruso, M., Panis, Y., Gagandeep, S., Houssin, D., Salzmann, J.-L. and Klatzmann, D., Regression of established macroscopic liver metastases after in situ transduction of a suicide gene, Proc. Natl. Acad. Sci. USA, 90 (1993) 7024-7028. [6] Chen, S.-H., Shine, H.D., Goodman, J.C., Grossman, R.G. and Woo, S.L.C., Gene therapy for brain tumors: Regression of experimental gliomas by adenovirus-mediated gene transfer in vivo, Proc. Natl. Acad. Sci, USA, 91 (1994) 3054-3057. [7] Culver, K.W., Ram, Z., Walbridge, S., Ishii, H., Oldfield, E.H. and Blaese, R.M., In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors, Science, 256 (1992) 1550-1552. [8] Freeman, S.M., Abboud, C.N., Whartenby, K.A., Packman, C.H., Koeplin, D.S., Moolten, F.L. and Abraham, G.N., The 'bystander effect': Tumor regression when a fraction of the tumor mass is genetically modified, Cancer Res., 53 (1993) 5274-5283. [9] Graham, F. and Prevec, L., Manipulation of adenovirus vectors. In Murray, E.J. (Ed.),Methods in Molecular Biology: Gene transfer and Expression Protocols, Vol. 7, Humana Press, Clifton, 1991, pp. 109-128. [10] Kolberg, R., The bystander effect in gene therapy: Great, but how does it work?, J. NIH Res., 6 (1994) 62-64. [11] Moolten, F.L., Tumor chemosensitivity conferred by inserted herpes thymidine kinase genes: Paradigm for a prospective cancer control strategy, Cancer Res., 46 (1986) 5276-5281. [12] Moolten, F.L. and Wells, J.M., Curability of tumors bearing herpes

82

[13]

[14]

[15]

[16]

A. C~olak et al. / Brain Research 691 (1995) 76-82

thymidine kinase genes transferred by retroviral vectors, J. Natl. Cancer. Inst., 82 (1990) 297-300. Nelson, D.F., McDonald, J.V., Lapham, L.W., Qazi, R. and Rubin, P., Central Nervous System Tumors. In Rubin, P. (Ed.), Clinical Oncology, 7th edn., W.B. Saunders, Philadelphia, 1993, pp. 617644. Oldfield, E.H., Ram, Z., Culver, K.W., Blaese, R.M., DeVroom, H.L. and Anderson, W.F., Gene therapy for the treatment of brain tumors using intra-tumoral transduction with the thymidine kinase gene and intravenous ganciclovir, Hum. Gene Ther., 4 (1993) 39-69. Perez-Cruet, M.J., Trask, T.W., Chen, S.-H., Goodman, J.C., Woo, S.L.C., Grossman, R.G. and Shine, H.D., Adenovirus-mediated gene therapy of experimental gliomas, J. Neurosci. Res., 39 (1994) 506-511. Ram, Z., Culver, K.W., Walbridge, S., Blaese, R.M. and Oldfield,

[17]

[18]

[19]

[20]

E.H., In situ retroviral-mediated gene transfer for the treatment of brain tumors in rats, Cancer Res., 53 (1993) 83-88. Ram, Z., Waibridge, S., Shawker, T., Culver, K.W., Blaese, R.M. and Oldfield, E.H., The effect of thymidine kinase and ganciclovir therapy on tumor vasculature and growth of 9L gliomas in rats, J. Neurosurg., 81 (1994) 256-260. Rivlin, A.S. and Tator, C.H., Objective clinical assessment of motor function after experimental spinal cord injury in the rat, J. Neurosurg., 47 (1977) 577-581. Takamiya, Y., Short, M.P., Moolten, F.L., Fleet, C., Mineta, T., Breakefield, X.O. and Martuza, R.L., An experimental model of retrovims gene therapy for malignant brain tumors, J. Neurosurg., 79 (1993) 104-110. Weizsaecker, M., Dean, D.F., Rosenblum, M.L., Hoshino, T., Gutin, P.H. and Barker, M., The 9L rat brain tumor: Description and application of an animal model, J. Neurol., 224 (1981) 183-192,