Intra-arterial delivery of p53-containing adenoviral vector into experimental brain tumors

Cancer Gene Therapy (2002) 9, 228 – 235 D 2002 Nature Publishing Group All rights reserved 0929-1903 / 02 $25.00 www.nature.com / cgt

Intra-arterial delivery of p53-containing adenoviral vector into experimental brain tumors Tatsuya Abe,1,4 Hiroaki Wakimoto,1 Robert Bookstein,2 Daniel C Maneval,2 E Antonio Chiocca,1 and James P Basilion1,3 1

Molecular Neuro-Oncology Laboratories, Neurosurgery Service, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts, USA; 2Canji Inc., San Diego, California, USA; 3Center for Molecular Imaging Research, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts, USA; and 4Department of Neurosurgery, Oita Medical University, Oita, Japan.

Human tumor xenografts established in athymic rat brains were used to determine the feasibility of intravascular delivery of tumor suppressor genes to brain tumors. Both tumor size and number were compared to characterize the effect of tumor burden on tumor transduction efficacy by a control LacZ - containing adenoviral vector. Experiments with tumors grown in vivo for either 3, 5, or 7 days demonstrated that 5 - day - old tumors provided the best target for vector infection and transgene expression by this mode of administration. Intra - arterial mannitol facilitated transduction efficiency. Tumor burden did not seem to affect transduction, while tumor location appeared to be an important factor. Based on these results, intra - arterial infusion of a p53 - containing adenoviral vector was carried out and resulted in significant retardation of brain tumor growth 3 days after administration. Effects at longer time points were not as significant. These findings indicate that intra - arterial administration of adenoviral vectors containing p53 is efficient and can result in changes in tumor size, but that long - term control of tumor growth may require multiple adenoviral treatments. Cancer Gene Therapy ( 2002 ) 9, 228 – 235 DOI: 10.1038 / sj / cgt / 7700437 Keywords: gliomas; gene therapy; p53; adenovirus; systemic administration

M

alignant gliomas are the most common and malignant form of primary brain tumors. In spite of significant advances in the understanding of the molecular mechanisms of oncogenesis and progression, treatment for this illness remains palliative. Classifications based on genetic mutations have been proposed: a subset of these tumors manifests with mutations in the p53 tumor suppressor gene, while another present with mutations in the p16 tumor suppressor pathway ( including deletions of the entire INK4A and INK4B locus ) and amplification of epidermal growth factor receptor ( EGFR ).1,2 However, the recent characterization of p19ARF (the product of INK4B ) as a positive modulator of p53 function suggests that ultimately, the majority of malignant gliomas will possess defects in p53 tumor suppressor pathway function.3 Replacement of wild -type p53 into p53 -defective tumor cells generally causes growth arrest ( for review, see Ref. [ 4] ). These findings have resulted in multitudes of preclinical studies assessing this tumor suppressor gene as a potential gene therapy agent including intrahepatic artery delivery of p53containing adenovirus for liver cancers,5 direct stereotactic injection of p53- expressing adenovirus into brain tumors,6 Received December 10, 2001. Address correspondence and reprint requests to: Dr James P Basilion, Center for Molecular Imaging Research, MGH - East, Building 149, 13th Street, 5406, Charlestown, MA 02129, USA. E - mail: [email protected]

and assessment of p53 therapy with head and neck cancer models in mice.7 A number of phase I clinical trials of gene therapy for recurrent malignant gliomas have been completed.8,9 Gene therapy vectors have been delivered into brain tumors by stereotactic or freehand injection into the neoplasm. Published results have revealed inefficient transduction with expression of the delivered cDNA localized in proximity to the injecting needle track.10,11 Strategies aimed at increasing the extent of anatomic distribution of gene therapy vectors would, therefore, be desirable. Intravascular delivery of the gene therapy vector to infect the tumor mass via its vasculature may provide a route of administration to increase the anatomic extent of adenovirus - mediated cDNA delivery. Several studies have demonstrated the potential of this strategy by combining vascular delivery with the use of pharmacologic disrupters of the blood – brain barrier ( BBB ).5,12 – 14 Recently, we have shown that the efficiency of this process is enhanced by depletion of complement or reduction in its activation, so much so that it allows for efficient infection of multiple foci of tumor within the brain.15,16 Knowing that cDNA can thus be delivered into brain tumors by intravascular administration of the vector has thus provoked our interest in testing this route of administration with p53- expressing adenoviral vectors. In the current study, we first defined the significance of tumor size and location as variables that could affect delivery of such vectors into human glioma xenografts in rats. We then

Intra - arterial administration of p53 - containing adenovirus T Abe et al

229 show that transient control of tumor growth is possible by intra - arterial administration of a p53 -containing adenoviral vector.

into parental glioma cell lines followed by selection. This manipulation has been used previously and has been demonstrated to enhance the tumorigenic capacity of several cell lines in the brain of nude rats (often, gliomas overexpress mutant EGFR ). In general, @EGFR -transduced tumor models are excellent models for in vivo work because of their rapid growth in rat brains.17 Gli36@EGFR cells were propagated at 378C with 5% carbon dioxide atmosphere in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum ( Sigma, St. Louis, MO ) containing 100 U /mL penicillin, 100 g/ mL streptomycin ( Gibco BRL, Gaithersburg, MD ), and 0.5 g/mL puromycin ( Sigma). For these studies, recombinant replication -

Methods Cell lines and adenoviruses

Human Gli36@EGFR glioma cells were established by transduction and clonal cell selection of human Gli36 cells with a retroviral vector expressing the @EGFR cDNA ( Dr T Ichikawa and Dr EA Chiocca, unpublished results ). This cell line was established by retroviral transfer of mutant EGFR

A. Cell survival (% of control)

150

FTCB (p53) GFCB (con)

100

50

0 104

105

106

107

108

109

(pn/ml)

B. 25

right anterior 20

left anterior 50

Tumor volume (mm3)

20

right posterior

40 15

15

30 10

10

20 5

5

10

0

Days:

0

2

5

BGCG (con)

2 FTCB (p53)

5

0

2

5

BGCG (con)

2

5 FTCB (p53)

2

5

BGCG (con)

2

5

FTCB (p53)

Figure 1 Effect of infection with p53 - containing adenovirus on Gli36@EGFR cell lines and tumors. A: Gli36@EGFR cells were treated with p53 containing adenovirus for 1 hour at the indicated concentrations. Seventy - two hours later, cells were washed, harvested, and adherent cells quantified. pn / mL: adenovirus particle number per milliliter. Error bars = SD. B: Rats were implanted with multiple Gli36@EGFR tumors ( control n = 4, treated n = 4 ) and 5 days later, tumors were directly injected with control or p53 - containing adenovirus. At the indicated times two animals were sacrificed and tumor volumes determined. Error bars = SD.

Cancer Gene Therapy

Intra - arterial administration of p53 - containing adenovirus T Abe et al

230

Table 1 Lack of difference in tumor transduction in single versus multiple tumor models

Tumor number Single Multiple

% X - gal – positive cell, right posterior tumor 4.54 ± 0.52 3.81 ± 1.43

Rats were implanted with either multiple brain tumors ( left anterior, right anterior, and right posterior, n = 4 ) or single brain tumors ( right posterior, n = 3 ). Five days after implantation, intra - arterial administration of LacZ - containing adenovirus was performed. Forty - eight hours later, animals were sacrificed and % LacZ - positive cells were measured after X - gal staining ( error bars represent ± SEM ).

deficient adenoviral vectors containing wild -type p53 ( FTCB ) and deleted for viral regions, E1, E3, and protein IX were used.18 Expression of the p53 transgene was directed by the human cytomegalovirus immediate -early promoter /enhancer. The control vector, GFCB, was constructed to match FTCB except for its transgene, which is enhanced green fluorescent protein (Clontech, Palo Alto, CA ). The BGCG control vector containing Escherichia coli LacZ was constructed as described in Cheney et al.19 All of the adenoviruses were grown in 293 cells and purified by DEAE column chromatography as described.20 Adenovirus particle concentrations were determined by Resource Q high -performance liquid chromatography,21 and the primary structure of all transgenes was verified by automated sequencing of viral DNA. Cells (3104 ) were plated in six - well dishes and treated with different concentrations of adenovirus in 100 L of culture medium for 1 hour. Two milliliters of replete medium was then added to supernatants. Seventy - two hours later, cells were harvested by trypsinization, and cells were counted by Coulter Counter (Coulter Electronics, Luton, UK ).

placing a linear skin incision over the bregma, burr holes ( 1 mm in diameter ) were drilled in the skull approximately 1 mm anterior and 2.5 mm lateral to the bregma on both sides, and 3 mm posterior and 2.5 mm lateral to bregma on the right side. The location for implantation of the tumors was based on established models and designed to minimize surgical side effects to the animals. Two hundred thousand Gli36@EGFR cells ( in a 2 -L volume ) were injected at a depth of 4.5 mm from the dura, using a 5- L Hamilton syringe. Three to 7 days after tumor implantation, animals were reanesthesized for intracarotid catheterization. The right common carotid artery ( CCA ) was exposed through a 3 -cm midline incision, as previously described.15 Twenty percent mannitol (Abbott Laboratories, North Chicago, IL ) was warmed to 378C and infused at a rate of 0.36 mL /kg /sec for 50 seconds. Immediately following mannitol infusion, 1 mL of adenovirus was infused for 10 minutes using a microsyringe pump ( Harvard Apparatus, Natick, MA ). In animals where adenovirus was directly injected into the tumors, the same burr holes and stereotactic coordinates were used. For tumor transduction assays, animals were sacrificed 2 days after adenovirus treatment and perfused by intracardiac infusion of a solution containing 4% neutral paraformaldehyde ( PFA ) in 0.9% sodium chloride and 10 mM sodium phosphate, pH = 7 ( PBS ). After harvesting, brains were kept in 4% PFA for 1 day, transferred to 30% sucrose in PBS for 5 days, then frozen and stored at 808C until analyzed. Preparation of other tissues was performed similarly. Histological studies

Frozen sections of brains and other organs were prepared on a cryostat and then stained with hematoxylin for anatomical detail and 5 -bromo- 4- chloro- 3- indolyl -  - D - galactopyranosidase ( X -gal; Fisher Biotech, Pittsburg, PA ) to reveal expressed LacZ activity. To avoid ‘‘false -positive’’ LacZ results, sections were stained with X - gal for only 3 hours prior to analysis.16

Animal studies

Quantitative analysis of tumor sections

Adult female nude rats (nu /nu, 200 – 250 g ) were anesthesized with an intraperitoneal injection of 0.5 mL of 0.9% NaCl containing 12.5 mg of ketamine and 2.5 mg of xylazine. After immobilizing the rats in a stereotactic apparatus and

Following staining with X -gal, tumor sections (n =4 ), randomly selected from each animal, were analyzed using an Olympus BX60 microscope. The sections were scanned

Table 2 Tumor volumes in rat brains

Table 3 LacZ cDNA expression in brain tumors following intra arterial BGCG administration

Days of tumor growth 3 - Day growth, treat, sac day 5 5 - Day growth, treat, sac day 7 7 - Day growth, treat, sac day 9

Volume ( mm3 ), right posterior tumor 1.93 ± 0.49 6.94 ± 2.62 32.5 ± 10.99

Volume of tumors was determined using three radial measurements using the formula 4 / 3r 1r 2r 3 ( where r 1 and r 2 are the perpendicular radii of the center cross - section of the tumor and where r 3 is the longitudinal radius of the tumor ). For 3 - day growth calculations, n = 6 animals; for 5 - day growth, n = 5 animals; and for 7 - day growth, n = 3 animals. No significant difference in tumor size was measured with or without mannitol; therefore, all animals were used in volume calculations. Mean ± SD is also reported.

Cancer Gene Therapy

Days of tumor growth

% X - gal – positive cell, right posterior tumor*

Mannitol treatment

No

Yes

3 - Day growth, treat, sac day 5 5 - Day growth, treat, sac day 7 7 - Day growth, treat, sac day 9

0.38 ± 0.493

3.13 ± 2.963

0.11 ± 0.082

20.0 ± 29.33

0.181

1.31 ± 0.162

*Values presented ± SD and the number of animals used to generate each data point included in superscript. For n = 1, the SD was not reported.

Intra - arterial administration of p53 - containing adenovirus T Abe et al

by Sony three - chip Color Video Camera at 20 magnification. Images were quantified using Image -Pro PLUS software (Media Cybernetics, Silver Spring MD ) and color cut -offs were selected to discriminate blue -stained LacZ positive cells from nonblue brain tumor cells and from regions of the tumor that did not contain cells. Blue -stained cells were considered to express LacZ and were quantified as a percentage of all cells within the tumor perimeter. Tumor volume measurements were calculated from tumor length and width measured using Image -Pro PLUS software. Tumor volume was calculated using the formula 4/3r 1r 2r 3 ( where r 1 and r 2 are the perpendicular radii of the center cross - section of the tumor and where r 3 is the longitudinal radius of the tumor ). Results Effect of p53 expression on glioma cells and tumors

The human glioblastoma cell line used for these studies, Gli36@EGFR, possesses a missense mutation in the p53 gene

231 ( Arg248Leu ), rendering the protein inactive. To confirm that this defect could be corrected by delivery of a wild -type p53 cDNA, Gli36@EGFR cells were evaluated for cell survival after infection with an adenovirus containing wild - type p53 cDNA (FTCB ). Consistent with previously reported data in other cells, FTCB infection of Gli36@EGFR cells in culture resulted in significant cell killing compared to infections with control adenovirus (Fig 1A ). To test this effect in vivo, Gli36@EGFR cells were used to establish distinct glioma xenografts within the brains of athymic rats. Tumors were then directly injected with either FTCB or a control adenoviral vector (BGCG ) and tumor growth was monitored 2 and 5 days after treatment. Tumors that were injected with FTCB were significantly smaller in volume than tumors injected with control BGCG (Fig 1B ). It was evident that the effect consisted of retardation of proliferation rather than tumor regression because continued growth between days 2 and 5 was noted. Taken together, these results showed that FTCB produced an anticancer effect against Gli36@EGFR glioma cells in vitro and in vivo, in agreement with previous studies that employed other cell lines.5,6

B

A

right posterior tumor 60

infectivity (%)

50

40

30

20

10

0

Rat #:

1

2

3

4

5

BGCG Alone BGCG w/ mannitol

Figure 2 Effect of mannitol on viral infection of brain tumors following intra - arterial administration of adenovirus. a: Rats ( n = 5 ) were implanted with brain tumors, and 5 days later, LacZ - containing adenovirus was administered with or without mannitol via the right external carotid artery. Two days after viral administration, animals were sacrificed and LacZ expression visualized with X - gal. The data for each animal are individually displayed. Error bars represent standard deviation derived from quantification of four tumor sections. b: Histopathological sections derived for right posterior tumors from animal 4 ( A ). Sections were stained with H&E to reveal anatomical detail and with X - gal to reveal LacZ activity. Note lack of LacZ activity in normal brain tissue. c: LacZ activity in peripheral organs from rats that had been treated either with 7.111011 viral particles in the absence of ( A, C, E, G, and I ) or in the presence of mannitol ( B, D, F, H, and J ). Control kidney from rats without any treatment was included to demonstrate the high background staining in this tissue ( K ). Studies on liver from treated animals ( G and H ) showed significant staining with X - gal prior to mannitol administration and substantially less liver staining when mannitol was coadministered with adenovirus. Substantial staining was observed for both treatment groups in spleen and kidney with minor effects of mannitol. All sections were stained with H&E and with X - gal.

Cancer Gene Therapy

Intra - arterial administration of p53 - containing adenovirus T Abe et al

232 C

Figure 2 ( continued )

Parameters effecting viral transduction of brain tumors in vivo

Our goal for these studies was to deliver an effective dose of adenovirus to brain tumors via intra -arterial administration. We therefore tested in combination two parameters that we hypothesized might alter transgene expression in tumors via intra - arterial administration: (a ) the effect of tumor burden and (b ) the effect of BBB -disrupting agents. Effect of tumor burden on neoplastic transduction by intra arterial administration of adenoviral vectors. We first tested

the effect of tumor burden on transgene expression by varying tumor number. Rats were implanted with either a single right posterior tumor or three tumors (one right anterior, one left anterior, and one right posterior tumor), and tumors were allowed to grow for 5 days. After intra - arterial infusion of mannitol and LacZ -containing adenovirus, comparison of X -gal –stained tumor sections indicated that the right posterior tumors showed similar growth and

Cancer Gene Therapy

transgene expression in both single and multiple tumor models ( Table 1 ). This result thus showed that the right posterior tumor transduction efficacy was not effected by the presence of tumors in the two anterior positions. We next tested the effect of tumor size on adenoviral transduction. Tumors were established in rat brains and different sizes were generated by allowing implanted tumor cells to grow for 3, 5, or 7 days prior to administration of adenovirus. Animals harboring brain tumors of varying age then received an intra -arterial infusion of BGCG. They were then sacrificed 2 days later for tumor volume measurements and examination of LacZ expression (Tables 2 and 3). Varying tumor age affected tumor size. On average, when compared to the tumors transduced after 3 days of growth, there was a 16.8- and 3.6- fold increase in tumor volume for the 7- and 5 -day transduced tumors, respectively ( Table 2). The percentage of tumor cells expressing the LacZ cDNA, however, was not significantly different for the tumors transduced after 3, 5, or 7 days of growth in the absence of

Intra - arterial administration of p53 - containing adenovirus T Abe et al

233 mannitol (Table 3 ). These results seem to suggest that tumor size / age had little effect on the transduction efficiency of intra - arterially administered adenovirus vector in the absence of mannitol. Mannitol increases tumor transduction after intra - arterial delivery of adenoviral vector. Several reports have demonstrated the efficacy of the BBB -disrupting agent, mannitol, for increasing infection of brain tumors.22 Therefore, we used mannitol in an attempt to improve delivery of adenovirus to intracranial tumors. Infusion of BGCG ( 7.111011 pn in 1 mL ) was accomplished through the right external carotid artery after a bolus of mannitol. This dose of vector was selected empirically because it provided the best infection with tolerable toxicity in rats. Without mannitol, tumor transduction efficiency was less than 1%. Mannitol infusion increased the average extent of transgene expression to 20%, with one animal showing an increase to approximately 50% transduction ( Fig 2A and Table 3 ).

80

Tumor volume (mm3)

60

40

20

0

GFCB (con)

Mannitol infusion alters the apparent distribution of LacZ expression in rat organs. To assay for effects of mannitol on

the distribution of LacZ cDNA expression after intra -arterial administration of vector, we examined heart, lungs, spleen, liver, and kidneys of treated animals. Figure 2C shows representative sections for several rat tissues. Little evidence for LacZ cDNA expression was detected in heart and lungs from animals treated with LacZ - containing adenovirus with or without mannitol. Spleen and kidney showed varying degrees of LacZ -positive cells. Significantly, liver stained intensely for LacZ cDNA expression. Coadministration of mannitol decreased LacZ cDNA expression in liver, and, to a lesser degree, in kidneys and spleen. Mannitol thus appeared to alter the profile of cDNA expression in animals. Effect of intra - arterial FTCB on brain tumor growth

. . .. ..

Representative sections of tumors are shown in Figure 2B. LacZ expression was confined to the tumor in both the mannitol -treated and control animals with no expression observed in normal brain tissue. These findings confirmed that mannitol infusion aided the infection of glioma xenografts in rat brains.

. .. .. ..

FTCB (p53)

Figure 3 Effect of intracarotid artery administration of p53 - containing adenovirus on tumor growth in vivo. Rats ( control n = 9; treated n = 9 ) were stereotactically implanted with tumors and after 5 days infused with mannitol followed by either control ( GFCB ) or p53 - containing adenovirus ( FTCB ) via the intracarotid artery. The horizontal lines represent the average tumor sizes for each treatment group. The change in tumor size due to treatment was highly significant. Nonparametric Wilcoxon rank sum test, P = .0047.

Based on the above findings, we tested the effect of intra arterial FTCB on tumor growth in vivo. Single right posterior brain tumors were permitted to establish for 5 days and then FTCB ( or GFCB as a control ) was administered intra arterially immediately following mannitol infusion. Tumor size was then assayed 3 days later. There was a statistically significant difference in the size of tumors treated with FTCB ( mean = 19.2 mm3 ) compared to tumors treated with GFCB ( mean = 43.6 mm3 ) (Fig 3). These results thus confirmed that intravascular FTCB could result in a significant anticancer effect against brain tumors in animals. Discussion

The major objective of this study was to determine if an anticancer effect could be achieved with intravascular administration of a p53 -containing adenoviral vector in a model of human malignant glioma. We show that such an effect is possible when combined with infusion of mannitol. We also show that tumor size and location, but not tumor number, have a dramatic effect on adenoviral transduction in vivo, and that mannitol infusion in the carotid artery alters the apparent whole- body distribution of adenovirus transgene expression. Multiple brain tumors in rats are often used to replicate the different foci present in human GBM. In order to determine the optimal adenovirus -to - tumor tissue ratio for these studies, we varied tumor size 16- fold by allowing tumors to grow for different periods of time in vivo prior to treatment with adenovirus. The tumors transduced in vivo in the presence of mannitol after 5 days of growth were the most efficiently transduced tumors followed by tumors transduced after 3 or 7 days of growth. It could be hypothesized that the tumors transduced after 3 or 7 days of growth have inappropriate hemodynamics for viral delivery, presumably resulting from tumor mass, vascularization, and /or BTB permeability. Indeed, we and others have observed that smaller tumors are generally poorly vascularized, and as they

Cancer Gene Therapy

Intra - arterial administration of p53 - containing adenovirus T Abe et al

234

increase in size, perfusion of the tumor center diminishes.23 Interestingly, we observed rapid tumor growth in the posterior lesion compared with anterior lesions. This was associated with more prevalent vascularization in posterior tumor (data not shown ) and may partially explain its enhanced growth and infectability. To monitor the efficiency of in vivo delivery and gene expression in this system, we used a LacZ -containing adenovirus and detection of the protein product was assessed by X - gal staining of tumor sections. These studies suggested that under optimum conditions ( intracarotid administration of mannitol and LacZ- containing adenovirus to tumors grown for 5 days in vivo ), up to 50% of the tumor cells could be transduced. It is not clear, however, why there was substantial variability between animals. Experiments are under way to examine this in more detail. When these transduction conditions were used to deliver p53 -containing adenovirus to rats bearing brain tumors, we observed significant inhibition of tumor growth 3 days after adenovirus administration ( Fig 3 ). However, inhibition of tumor growth was less pronounced by 6 days (data not shown). The effectiveness of the p53- containing adenovirus to suppress tumor growth was greater than the extent of tumor transduction observed with LacZ -containing adenovirus. One possible explanation for this apparent discrepancy could be that X -gal staining does not detect tumor cells expressing low levels of LacZ,24 underestimating transduction efficiency. Therefore, the LacZ marker gene is useful for optimizing adenoviral delivery, but may not be as useful for predicting the therapeutic outcome of p53 delivery and expression. Additionally, the efficacy of the therapy may be partially augmented by a p53 -dependent bystander effect that has been observed in mixing experiments.25,26 Another interesting observation was made by staining for LacZ activity in different organs from treated animals. In the absence of treatment with mannitol, liver stained intensely for LacZ activity following intracarotid administration of LacZ - containing adenovirus. However, if adenoviral administration were performed immediately after mannitol infusion, the level of LacZ activity observed in the liver was dramatically reduced (see Fig 2C ). This also occurred for other tissues. The apparent redistribution in adenoviral gene expression is possibly due to increased uptake by the tumor or mannitol - induced vascular changes resulting in altered biodistribution of the adenovirus, and studies are currently underway to assess viral distribution in vivo.27 The studies presented here demonstrate the utility of intraarterial administration of p53- containing adenovirus for reduction of tumor size in a model of human glioblastoma. We have demonstrated that 3 days after intracarotid adenovirus administration, there is a significant inhibition of tumor growth. However, by 6 days, the difference between p53 -treated and control - treated tumors was not significant ( data not shown ). We and others have demonstrated that the level of transgene expression is decreased between 5 and 7 days after treatment of tumors in vivo, suggesting that multiple dosing with therapeutic adenovirus may increase the inhibitory effect of p53 on tumor growth.16,28 Li et al25 have demonstrated that multiple direct injections of p53expressing adenovirus into tumors in an in vivo mouse

Cancer Gene Therapy

system result in tumor cure. Additionally, combination therapy may augment the effects seen here. Several studies have demonstrated increased efficacy of conventional radiotherapy and chemotherapeutic drugs when combined with p53 gene therapy of cancers ( for review, see 29 –35 ). The work presented here enables one to determine if intracarotid delivery of the p53 transgene to brain tumors in vivo would be more effective when combined with conventional anticancer therapies for brain tumors.

Acknowledgments

This work was funded by a grant from Schering -Plough to EAC. Special thanks to Dr P Rioux for his statistical analysis of the data, and Dr W Robert Bishop for his support.

References 1. Duncan EL, Reddel RR. Genetic changes associated with immortalization. A review. Biochemistry ( Moscow ). 1997;62 1263 – 1274. 2. Chin L, Pomerantz J, DePinho RA. The INK4a /ARF tumor suppressor: one gene – two products – two pathways. Trends Biochem Sci. 1998;23:291 – 296. 3. Carnero A, Hudson JD, Price CM, Beach DH. p16INK4A and p19ARF act in overlapping pathways in cellular immortalization [see comments]. Nat Cell Biol. 2000;2:148 – 155. 4. Nielsen LL, Maneval DC. P53 tumor suppressor gene therapy for cancer. Cancer Gene Ther. 1998;5:52 – 63. 5. Anderson SC, Johnson DE, Harris MP, et al. p53 gene therapy in a rat model of hepatocellular carcinoma: intra - arterial delivery of a recombinant adenovirus. Clin Cancer Res. 1998;4:1649 – 1659. 6. Badie B, Drazan KE, Kramar MH, Shaked A, Black KL. Adenovirus - mediated p53 gene delivery inhibits 9L glioma growth in rats. Neurol Res. 1995;17:209 – 216. 7. Clayman GL, el - Naggar AK, Roth JA, et al. In vivo molecular therapy with p53 adenovirus for microscopic residual head and neck squamous carcinoma. Cancer Res. 1995;55:1 – 6. 8. Markert JM, Medlock MD, Rabkin SD, et al. Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial [see comments]. Gene Ther. 2000;7:867 – 874. 9. Sandmair AM, Loimas S, Puranen P, et al. Thymidine kinase gene therapy for human malignant glioma, using replication deficient retroviruses or adenoviruses conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Hum Gene Ther. 2000;11:2197 – 2205. 10. Harsh GR, Deisboeck TS, Louis DN, et al. Thymidine kinase activation of ganciclovir in recurrent malignant gliomas: a gene - marking and neuropathological study. J Neurosurg. 2000;92:804 – 811. 11. Ram Z, Culver KW, Oshiro EM, et al. Therapy of malignant brain tumors by intratumoral implantation of retroviral vector producing cells [see comments]. Nat Med. 1997;3:1354 – 1361. 12. Muldoon LL, Nilaver G, Kroll RA, et al. Comparison of intracerebral inoculation and osmotic BBB disruption for delivery of adenovirus, herpesvirus, and iron oxide particles to normal rat brain. Am J Pathol. 1995;147:1840 – 1851. 13. Carew JF, Federoff H, Halterman M, et al. Efficient gene

Intra - arterial administration of p53 - containing adenovirus T Abe et al

235 transfer to human squamous cell carcinomas by the herpes simplex virus type 1 amplicon vector. Am J Surg. 1998;176: 404 – 408. 14. Rainov NG, Dobberstein KU, Heidecke V, et al. Long - term survival in a rodent brain tumor model by bradykinin enhanced intra - arterial delivery of a therapeutic herpes simplex virus vector. Cancer Gene Ther. 1998;5:158 – 162. 15. Ikeda K, Ichikawa T, Wakimoto H, et al. Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses. Nat Med. 1999;5:881 – 887. 16. Ikeda K, Wakimoto H, Ichikawa T, et al. Complement depletion facilitates the infection of multiple brain tumors by an intravascular, replication - conditional herpes simplex virus mutant. J Virol. 2000;74:4765 – 4775. 17. Nishikawa R, Ji XD, Harmon RC, et al. A mutant epidermal growth factor receptor common in human glioma confers enhanced tumorigenicity. Proc Natl Acad Sci USA. 1994;91:7727 – 7731. 18. Wills KN, Maneval DC, Menzel P, et al. Development and characterization of recombinant adenoviruses encoding human p53 for gene therapy of cancer. Hum Gene Ther. 1994;5:1079 – 1088. 19. Cheney IW, Johnson DE, Vaillancourt MT, et al. Suppression of tumorigenicity of glioblastoma cells by adenovirus mediated MMAC1 / PTEN gene transfer. Cancer Res. 1998 58:2331 – 2334. 20. Huyghe BG, Liu X, Sutjipto S, et al. Purification of a type 5 recombinant adenovirus encoding human p53 by column chromatography. Hum Gene Ther. 1995;6:1403 – 1416. 21. Shabram PW, Giroux DD, Goudreau AM, et al. Analytical anion - exchange HPLC of recombinant type - 5 adenoviral particles. Hum Gene Ther. 1997;8:453 – 465. 22. Kroll RA, Neuwelt EA. Outwitting the blood – brain barrier for therapeutic purposes: osmotic opening and other means. Neurosurgery. 1998;42:1083 – 99; discussion 1099 – 1100. 23. Vajkoczy P, Schilling L, Ullrich A, Schmiedek P, Menger MD. Characterization of angiogenesis and microcirculation of high grade glioma: an intravital multifluorescence microscopic approach in the athymic nude mouse. J Cereb Blood Flow Metab. 1998;18:510 – 520. 24. Couffinhal T, Kearney M, Sullivan A, Silver M, Tsurumi Y, Isner JM. Histochemical staining following LacZ gene transfer

underestimates transfection efficiency. Hum Gene Ther. 1997;8:929 – 934. 25. Li Z, Rakkar A, Katayose Y, et al. Efficacy of multiple administrations of a recombinant adenovirus expressing wild type p53 in an immune - competent mouse tumor model. Gene Ther. 1998;5:605 – 613. 26. Nishizaki M, Fujiwara T, Tanida T, et al. Recombinant adenovirus expressing wild - type p53 is antiangiogenic: a proposed mechanism for bystander effect. Clin Cancer Res. 1999;5:1015 – 1023. 27. Schellingerhout D, Bogdanov A Jr, Marecos E, Spear M, Breakefield X, Weissleder R. Mapping the in vivo distribution of herpes simplex virions. Hum Gene Ther. 1998;9:1543 – 1549. 28. Tsujie M, Isaka Y, Nakamura H, Kaneda Y, Imai E, Hori M. Prolonged transgene expression in glomeruli using an EBV replicon vector system combined with HVJ liposomes. Kidney Int. 2001;59:1390 – 1396. 29. Badie B, Kramar MH, Lau R, Boothman DA, Economou JS, Black KL. Adenovirus - mediated p53 gene delivery potentiates the radiation - induced growth inhibition of experimental brain tumors. J Neurooncol. 1998;37:217 – 222. 30. Kataoka M, Schumacher G, Cristiano RJ, Atkinson EN, Roth JA, Mukhopadhyay T. An agent that increases tumor suppressor transgene product coupled with systemic transgene delivery inhibits growth of metastatic lung cancer in vivo. Cancer Res. 1998;58:4761 – 4765. 31. Roth JA, Swisher SG, Meyn RE. p53 tumor suppressor gene therapy for cancer. Oncology ( Huntington ). 1999;13:148 – 154. 32. Gjerset RA, Mercola D. Sensitization of tumors to chemotherapy through gene therapy. Adv Exp Med Biol. 2000;465 273 – 291. 33. Lang FF, Yung WK, Sawaya R, Tofilon PJ. Adenovirus mediated p53 gene therapy for human gliomas. Neurosurgery. 1999;45:1093 – 1104. 34. Trepel M, Groscurth P, Malipiero U, Gulbins E, Dichgans J, Weller M. Chemosensitivity of human malignant glioma: modulation by p53 gene transfer. J Neurooncol. 1998;39 19 – 32. 35. Dorigo O, Turla ST, Lebedeva S, Gjerset RA. Sensitization of rat glioblastoma multiforme to cisplatin in vivo following restoration of wild - type p53 function. J Neurosurg. 1998;88: 535 – 540.

Cancer Gene Therapy