glioblastoma multiforme: molecular biology and new perspectives

Gene Therapy and Molecular Biology Vol 1, page 133

133

Gene Ther Mol Biol Vol 3, 133-148. August 1999.

Glioblastoma multiforme: molecular biology andnew perspectives for therapyReview Article

Giorgio Palù, Luisa Barzon, and Roberta BonaguroInstitute of Microbiology, University of Padova Medical School, Padova, Italy

__________________________________________________________________________________________________Corresponding Author : Giorgio Palù, MD, Institute of Microbiology, Via A. Gabelli 63, 35121 Padova, Italy.Tel: +39-049-8272350; Fax: +39-049-8272355; E-mail: [email protected]

Key words : Therapy, gene therapy, brain tumors, gliomas, glioblastoma multiforme, molecular biology, pathogenesis,immunotherapy, neoangiogenesis, oncogenes

Abbreviat ion : GBM, glioblastoma multiforme

Received: 23 November 1998; accepted 30 November 1998

SummaryPathogenic features of glioblastoma multiforme and of other gliomas are reviewed in the presentarticle . Emphasis is given to those genetic alterations which are involved in oncogenesis , to theprocess of tumor neoangiogenes is and to the ro le p layed by the immune system in controllingneoplastic growth. Aspects which are relevant to therapeutic interventions are also dissected, andgene therapy in particular. A new gene therapy approach that combines tumor suicide, via enzyme-directed prodrug activation, and cytokine-promoted immune rejection i s reported, together withresults from the first application of this approach in humans.

I. IntroductionThe outcome of malignant gliomas remains extremely

poor, in spite of aggressive use of currently availabletherapies. Recent advances in elucidating the molecularbiology of gliomas have led to the development ofinnovative therapeutic strategies. The more promisingapproaches involve gene therapy, aiming at increasingtumor cell chemosensitivity and/or immunogenicity, bytransfer of genes expressing cytokines and prodrugactivating enzymes.

Glioblastoma multiforme (GBM) represents 15-20% ofall intracranial tumors and 50% of gliomas (Russel andRubistein, 1989). It affects 5,000 Americans and 1,000Italians every year, and typically occurs in adults, with apeak incidence in the fifth and sixth decades of life. It is avery aggressive tumor, with a uniform and profoundmorbidity. Because of its morbidity it contributes to thecost of cancer on a pro capite basis more than any othertumor. Despite surgery, radiotherapy and/or chemotherapy,the prognosis is extremely poor and has not substantiallychanged over the last two decades, death resulting in 80%

of patients for tumor recurrence within 6-12 months fromtreatment.

Glioblastoma multiforme (grade IV astrocytoma) isusually located in the cerebral hemispheres, though itoccasionally appears at other sites, such as the cerebellum,the brain stem and the spinal cord. Histology showsmarked cytological diversity, ranging from tumorscomposed of small cells with scant cytoplasm to thosecomposed of multinucleated giant cells. The World HealthOrganization (WHO) classification recognizes two distinctsubvariants of the tumor: (i ) giant cell glioblastoma,characterized by a predominance of enormous,multinucleated giant cells and, on occasion, an abundantstromal reticulin network; and (i i ) gliosarcoma, in whichhyperplastic vascular elements have undergone sarcomatoustransformation.

Current therapies for malignant gliomas includesurgical removal of the tumor mass, which is mandatoryfor precise diagnosis, and irradiation. Although surgeryimproves the prognosis (Levin et al, 1993), the infiltrativebehavior of malignant gliomas precludes their complete

Palù et al: Perspectives for therapy of glioblastoma multiforme

134

resection, and 90% of GBM recur within 2 cm of theprimary site. Postoperative radiotherapy is thereforecommonly administered, with a significant improvementin survival (Hochberg and Pruitt, 1980; Walker et al,1980). Despite surgery and irradiation, however, only a fewpatients are alive two years after diagnosis. Results ofchemotherapy trials are disappointing (Hosli et al, 1998).This is due both to the tumor intrinsic chemoresistance(Petersdorf and Berger, 1996) and to the tumor locationwithin the central nervous system, which limits thepenetration of drugs (Janzer and Raff, 1987; Mak et al,1995). Among malignant gliomas, GBM is the leastresponsive to medical treatment. Available protocolsinclude both monochemotherapy and polychemotherapyregimens. Nitrosoureas are the leading drugs in gliomachemotherapy, with response rates as single agents varyingfrom 10% to 40% (Young et al, 1973; Fewer et al, 1972;Hoogstraten et al, 1972). Other drugs, evaluated inmonochemotherapy (Forsyth and Cairncross, 1996),occasionally showed clinically and radiologically objectiveresponses. Among these are vincristine (Smart et al,1968), procarbazine (Rodriguez et al, 1989), paclitaxel(Chamberlain and Kormanik, 1995; Prados et al, 1996),and temozolomide (Newlands et al, 1996). However,methodological bias present in most studies raise doubtsabout the validity of these results. The most commonlyused polychemotherapy regimens for gliomas are PVC (i.e.a combination of CCNU, procarbazine, and vincristine) andMOP (i.e. a combination of procarbazine, vincristine, andmechlorethamine). Response rates (complete or partial) of17-37% have been reported for glioblastomas (Levin et al,1980; Coyle et al, 1990). More recently, interesting resultshave been obtained in GBM patients with the ICE regimen(ifosfamide, carboplatin and etoposide), although inassociation with severe hematological toxicity (Sanson etal, 1996). The role of PVC as adjuvant chemotherapy iscontroversial (Fine et al, 1993), and, overall, there is noclear-cut evidence that survival of glioblastoma patients isimproved by chemotherapy (Hosli et al, 1998).

II. Molecular biology of glioblastomamultiforme and corrective gene therapy

As for most cancers, brain tumors derive from a multi-step process of successive alterations, including loss of cellcycle control, neoangiogenesis and evasion of immunecontrol. Figure 1 summarizes the genetic alterationsassociated with the malignant transformation of astrocytes.Most of these changes involve the loss of putative tumorsuppressor genes or activation of proto-oncogenes.

Gene therapy of cancer, in its most direct form, shouldaim at replacing a mutated gene with its correct form, or atsuppressing the abnormal oncogenic function. At present,however, such a corrective gene therapy, faces the

insurmountable task that gene replacement, or genesuppression, should simultaneously involve a number ofdifferent genes, and should be applied to all tumor cells toreverse the malignant phenotype. Hence, corrective genetherapy seems to be quite difficult to propose as a singletherapeutic approach.

A. Genetic alterations1. Oncogenes

Several members of the protein-tyrosine kinase receptorfamily are over-expressed by gene amplification inmalignant gliomas, including the epidermal growth factorreceptor (EGFR), the platelet-derived growth factorreceptor-α (PDGFRα) and the c-met genes (Furnari et al,1995). A high percentage of glioblastomas also haveEGFR gene rearrangements that may lead to the expressionof a truncated, constitutively activated receptor. Thetransfer of a mutant human EGFR gene into glioblastomacells caused constitutive self phosphorylation and apronounced enhancement tumorigenicity in nude mice

o f

Figure 1 . Simplified representation of oncogenes and tumorsuppressor genes contributing to malignant progression ofastrocytic tumors.


135

(Ekstrand et al, 1994; Nishikawa et al, 1994). Numerousstrategies are currently being investigated to specificallyinhibit EGFR using antibodies, immunoconjugates orantisense technology. The selective inhibition of EGFR inhuman GBM cells with kinase-deficient mutants inhibitedcell proliferation and transforming efficiency in athymicmice (O'Rourke et al, 1997), and converted radioresistanthuman glioblastoma cells to a more sensitive phenotype(O'Rourke et al, 1998), providing a rationale for genetherapy applications.

Other dominant oncogenes, such as N-myc, fos, src, H-ras or N-ras, and mdm2 are amplified and highly expressedin gliomas (Collins, 1993). GBM produce high levels ofinsulin-like growth factor I (IGF-I). When this alterationhas been targeted by a vector expressing an antisense anti-IGF-I gene, rejection of genetically altered rat C6 gliomacells was observed. Injection, even at a site distal to thetumor, caused regression of established brain GBM.Destruction of the tumor was mediated by a glioma-specific T CD8+ (CTL) response (Trojan et al, 1993).

2. Genes associated to cell immortalization

A role in cell immortalization has been proposed fortelomerase, the RNA-protein complex that elongatestelomeric DNA. Telomerase is expressed almostexclusively in cancer cells, but not in normal cells,suggesting the possibility that gene therapy may beapplied to inhibit this function. A successful example oftreatment via antisense oligonucleotides directed againsthuman telomerase suppressed glioma cells growth andsurvival, both in vitro and in vivo, through the inductionof apoptosis (Kondo et al, 1998).

3. Tumor suppressor genes

Molecular and cytogenetic analyses of gliomas haveshown frequent losses of genetic material, suggesting theinactivation of putative tumor suppressor genes. Loss ofheterozygosity (LOH) has been described in chromosome1p, 9p, 10p, 10q, 11p, 13q, 17p, 19q, and 22q, and insome cases the tumor suppressor gene involved in LOHhas been identified. This is the case of the p53 tumorsuppressor gene, which maps in 17p. Wild-type p53protein is involved in G1 cell cycle arrest and apoptosis ofDNA-damaged cells and is therefore crucial in preventingmutation or deletion of functional genes. Mutations of p53seem to be an early event in glioma tumorigenesis, beingfrequently detected also in low grade astrocytomas. Alongwith p53 mutations, amplification of the mdm2 oncogene,whose product binds to and degrades p53, accounts for p53inactivation in gliomas. Since p53 plays a key role in thepathogenesis of most cancers, it has raised great interest asa target for cancer gene therapy. Transduction of malignant

cells with wild type p53 can significantly inhibit growthand neoangiogenesis, or can induce apoptosis in p53mutant cells in several tumor models in vitro, includinggliomas (Badie et al, 1995; Van Meir et al, 1995; Gomez-Manzano et al, 1996). The presence of functional p53 hasalso been shown to modulate chemoresistance.Consequently, another possible advantage of the restorationof wild type p53 may be sensitization to chemotherapy andradiotherapy. Indeed, the combination of p53 genetransduction with radiation or chemotherapy (Lowe et al,1994) has resulted in local tumor control superior to eithertherapy alone (Fujiwara T et al, 1994; Gjerset et al, 1995;Ngyuyen et al, 1996, Lang et al, 1998). This combinedtherapy is currently under investigation in clinical trials(Roth and Cristiano, 1997; Nielsen and Maneval, 1998).

The cell cycle regulator genes provide an additionaltarget for corrective gene therapy. The p105Rb product ofthe retinoblastoma tumor suppressor gene (Rb) is one ofthe most critical regulators of cellular proliferation. TheRb protein (pRb), when unphosphorylated, is responsiblefor arrest of cell cycle by inhibition of the activity of theE2F family of transcription factors. Normal cell cycleprogression requires inactivation of Rb throughphosphorylation by cyclin-dependent kinases (CDK). Thisprocess, in turn, is regulated by CDK inhibitors. Amongthese, p21 protein is induced directly by p53; p16 protein,and its homologue p15, specifically bind to and inhibitCDK4, and may therefore regulate Rb phosphorylation,and cell cycle progression. Dysregulation of cell cyclecontrol is a frequent finding in malignant gliomas, likedeletion or loss of expression of p16 and p15 tumorsuppressor genes (Jen et al, 1994; Nishikawa et al, 1995),amplification of CDK4 (He et al, 1994), and deletion ormutation of the Rb tumor suppressor gene (Henson et al,1994). Interestingly, both of the latter events take placewhen the p16 gene is intact and correctly expressed (He etal, 1995). Restoration of wild-type p16 gene in gliomacells through an adenoviral vector arrested cells in G0-G1phases of the cell cycle (Fueyo et al, 1996) and suppressedglioma cells invasion in vitro (Chintala et al, 1997).Overexpression of p21 increases the susceptibility ofglioblastoma cells to cisplatin-induced apoptosis (Kondo etal, 1996), whereas adenovirus-mediated transfer ofexogenous E2F-1 protein induced massive apoptosis andsuppressed glioma growth in vivo and in vitro (Fueyo etal, 1998). The possibility that E2F-responsive promotersmay be more active in tumor cells relative to normal cells,because of loss of pRb function, has been exploited todesign adenoviral vectors containing transgenes driven bythe E2F-1 promoter for gene therapy of gliomas (Parr et al,1997). These vectors showed tumor-selective geneexpression in vivo and reduced toxicity of the normaltissue with respect to standard adenoviral vectors.


136

Inhibition or inactivation of genes/factors involved inDNA repair and/or cellular SOS response could represent agene therapy approach that potentiates radiation therapy. Infact, inhibition of the RAD51 gene by antisenseoligonucleotides enhanced the radiosensitivity of mousemalignant gliomas, both in vitro and in vivo, improvingsurvival (Ohnishi et al, 1998). This gene, a homologue ofthe yeast RAD51 and E. coli RecA genes, is involved inrepair of DNA double-strand breaks, in recombinationrepair, and in various SOS responses to DNA damagecaused by gamma-irradiation and alkylating agents.

Deletions of large regions or even of the entire copy ofchromosome 10 are a genetic hallmark of GBM. At leasttwo tumor suppressor genes located on chromosome 10(one on each arm) have been demonstrated to participate toglial oncogenesis. A first candidate tumor suppressor gene,called PTEN (Phosphatase and tensin homologue deletedon chromosome TEN) was recently characterized (Li et al,1997). The DNA region encoding PTEN is altered inglioblastoma multiforme, but not in lower grade astrocytictumors (Tohma et al, 1998; Ichimura et al, 1998;Chiariello et al, 1998).

B. NeoangiogenesisTumors may remain in a state of dormancy until they

establish a blood supply for receiving oxygen andnutrients. The complex process of neoangiogenesis isregulated by numerous factors, some with angiogenicproperties, i.e. vascular endothelial growth factor (VEGF),platelet derived growth factor (PDGFα), basic fibroblastgrowth factor (bFGF), epidermal growth factor (EGF),interleukin-8, and by endogenous inhibitors ofangiogenesis, i.e. thrombospondin-1, platelet factor 4(PF4), angiostatin, endostatin. VEGF, which binds to twospecific tyrosine-kinase receptors, called Flk-1 and Flt-1,has been demonstrated to play a key role in angiogenesis ofgliomas. Indeed, VEGF and its receptors are downregulatedin the normal adult brain, whereas, VEGF is highlyproduced by GBM cells. Since both flt-1 and flk-1 genesare expressed by proliferating endothelial cells of gliomas,this leads to the establishment of a paracrine loop.Moreover, VEGF expression is higher around necroticareas and seems to be stimulated by hypoxia.

Glioblastoma multiforme is one of the most highlyvascularized solid neoplasms; therefore, treatments thattarget neoangiogenesis would be of great interest in clinicalpractice. Co-injection of rat C6 glioma cells, eithersubcutaneously or intracerebrally in nude mice, togetherwith cells producing retroviral vectors encoding adominant-negative mutant of the Flk-1 receptor showedinhibition of neoangiogenesis, reduction of tumor growth,and survival improvement (Millauer et al, 1994; 1996).Antisense VEGF oligonucleotides (Saleh et al, 1996) and

ribozymes against VEGF mRNA have been successfullyemployed to reduce VEGF expression in glioma cells (Keet al 1998), once more suggesting a potential role forantiangiogenic gene therapy. Similarly, bFGF antisensecDNA decreased C6 glioma cells proliferation (Redekopand Naus, 1995). Besides inhibiting the production ofangiogenic factors, a therapeutic intervention could alsoconsist of providing tumors with antiangiogenic factors.Indeed, retroviral and adeno-associated viral vectorsexpressing a modified PF4 were reported to inhibitendothelial cells proliferation in vitro and the growth ofintracerebrally implanted gliomas (Tanaka et al, 1997).Retroviral and adenoviral vectors transducing angiostatingene increased apoptotic death of glioma tumor cells(Tanaka et al, 1998). Additionally, the intratumoraldelivery of angiostatin gene by an adenoviral vectorproduced inhibition of tumor growth in vivo, suppressionof neovascularization, and a marked increase of tumor cellsapoptosis (Griscelli et al, 1998). Damage of tumormicrovasculature was reported also in human malignantglioma xenografts, after gene therapy followed byradiotherapy. The treatment consisted of intratumoralinjection of adenoviral vectors expressing tumor necrosisfactor-α (TNF-α), under control of the Egr-1 promoter(Staba et al, 1998). The use of viral vectors containingradiation-inducible promoters, such as Egr-1, has theadvantage of selectively, spatially, and temporally limitingthe effects of the therapeutic gene in the radiation field.Recently, this strategy has yielded interesting results in rat9L glioma cells (Manome et al, 1998).

III. Suicide gene therapySuicide gene therapy operates by tumor transduction of

genes converting a prodrug into a toxic substance;independently, the gene product and the prodrug are non-toxic. The prototype of this approach exploits the selectiveintracellular phosphorylation of ganciclovir (GCV), drivenby the herpes simplex virus thymidine kinase gene product(HSV-TK). This activation generates a toxic drugmetabolite that inhibits DNA synthesis, inducing celldeath. For in vivo gene transfer of the HSV-TK gene tomalignant cells, packaging cells that produces retroviralvectors expressing HSV-TK, have been injected directlyinto the tumor to transduce replicating cells. An interestingfeature of the HSV-TK/GCV system is the bystanderkilling of nontransduced cells.

The mechanisms that are responsible for this effecthave not been fully defined, but are likely to include: (i )transfer of non-diffusible, phosphorylated GCV toneighboring cells through gap junctions; (i i ) endocytosisby nontransduced cells of cellular debris containing toxicGCV; and (i i i ) stimulation of host antitumor immuneresponse. The therapeutic efficacy of the HSV-TK/GCV


137

system may be further increased by the use of adenoviralvectors, since these vectors can also transduce restingtumor cells. However, adenoviral vectors will infect alsonormal cells; hence, the inclusion of sequences able torestrict gene expression only in tumor cells can circumventthis problem. Selective tumor toxicity was obtainedpositioning the suicide gene under control of the E2F-responsive promoter elements which are de-repressed inglioma cells (Parr et al, 1997).

The effectiveness of suicide gene therapy has beenexplored for a variety of neoplasms, especially forrefractory and localized diseases such as GBM. The HSV-TK/GCV scheme has been demonstrated to be effective inanimal models, by ex vivo and in vivo transduction withretroviral, adenoviral, or adenoviral-associated vectorsexpressing HSV-TK (Ezzeddine et al, 1991; Culver et al,1992; Takamiya et al, 1992, 1993; Barba et al, 1993,1994; Ram et al, 1993; Kim et al, 1994; Vincent et al,1996; Maron et al, 1996; Mizuno et al, 1998). Novel genedelivery systems were also applied in a mouse model forgene therapy of meningeal gliomatosis. Liposomes coatedwith Sendai virus envelope protein were highly efficient indelivering the therapeutic gene in disseminated glioma cells(Mabuchi et al, 1997).

A factor that may limit the effectiveness of HSV-TK/GCV therapy is the GCV crossing through the blood-brain barrier. This can be circumvented by the use of thebradykinin analogue and potent blood-brain barrierpermeabilizer RMP-7, which, administered intravenously,increase the delivery of GCV into rat brain tumors,enhancing the cytotoxic and bystander effects of HSV-TK/GCV (LeMay et al, 1998).

Several clinical trials exploiting the HSV-TK/GCVsystem have been initiated. It is too early to estimate theeffectiveness of these therapeutic procedures.Notwithstanding the evidence for growth-suppressiveactivities of HSV-TK plus GCV, cure rates are low.Explanation for lack of complete response in humans mayreside in the different biological behavior of GBM cellswhen injected into animals (Sturtz et al, 1997).

The antitumor effects elicited by HSV-TK/GCVprompted to explore other prodrug activating enzymes. Apromising system exploits the ability of E. coli cytosinedeaminase (CD) to convert the relatively non toxic 5-fluorocytosine (5-FC) to the chemotherapeutic agent 5-fluorouracil (5-FU). Significant antitumor effects of CD/5-FC were observed in nude mice bearing tumors derivedfrom C6 glioma cells and transduced with CD. A"bystander" effect could also be demonstrated (Ge et al,1997), suggesting a potential role for gene therapy ofglioblastoma.

As already reported for p53, both the CD and HSV-TKsystems sensitize cancer cells to radiation. Animal models

have shown encouraging results both in vitro and in vivo(Khil et al, 1995; Kim et al, 1997). 9L glioma cellstransduced with a retrovirus encoding a CD/HSV-TKfusion gene exhibited enhanced sensitivity to both GCVand 5-FC, as well as increased radiosensitivity (Rogulski etal, 1997). This experiment suggests the feasibility of acombined approach with two suicide genes associated withradiotherapy.

Another prodrug activation system is represented bycytochrome P450 2B1 (the liver enzyme catalyzingcyclophosphamide and ifosfamide activation) gene transferfollowed by cyclophosphamide or ifosfamideadministration. Metabolites of these drugs produce inter-strand DNA cross-linking in a cell cycle-independentfashion. C6 and 9L rat glioma cells, when stablytransfected with the P450 2B1 gene, become highlysensitive to cyclophosphamide in in vitro and in vivomodels (Wei et al, 1994; Manome et al, 1996). Rabbitcytochrome P450 isozyme CYP4B1, which converts theinert prodrugs 2-amonianthracene (2-AA) and 4-ipomeanol(4-IM) into highly toxic alkylating metabolites, alsoshowed high antitumor effects, both in vitro and in vivo.The treatment had relatively low toxicity and wasassociated with a bystander effect, not requiring cell-to-cellcontact (Rainov et al, 1998).

Another suicide gene system is based on E. coli purinenucleoside phosphorylase (PNP), which generates toxicpurine nucleoside analogues intracellularly, either from 6-methylpurine-2’-deoxyriboside or arabinofuranosyl-2-fluoroadenine monophosphate. Significant antitumoractivity and low systemic toxicity were reported in nudemice bearing human malignant D54MG glioma tumorsexpressing PNP (Parker et al, 1997).

Phosphorylation of the prodrug cytosine arabinoside(ara-C) by deoxycyticine kinase (dCK) is a limiting stepfor activation. Thus, ara-C, a potent antitumor agent forhematological malignancies, has only minimal activityagainst most solid tumors. Transduction of the dCK cDNAby retroviral and adenoviral vectors also resulted in markedsensitization of glioma cell lines to ara-C in vitro, and insignificant antitumor activity in vivo (Manome et al,1996).

Unlike other prodrug activating enzymes, E. coli gptsensitizes cells to the prodrugs 6-thioxanthine (6TX) and 6-thioguanine (6-TG), and confers resistance to differentregimens (mycophenolic acid, xantine, and hypoxanthine),providing a means to select for gpt-positive cells. Rat C6glioma cells transduced with a retroviral vector expressingthe gpt gene exhibited significant 6TX and 6GTsusceptibility and a "bystander" effect in vitro. Anantiproliferative effect was demonstrated in vitro and invivo (Tamiya et al, 1996; Ono et al, 1997).


138

Tumor suicide can also be achieved by direct infectionof tumor cells with a conditionally replicative virus, i.e. aninfectious agent able to replicate and to kill only dividingcells. In the case of tumors highly proliferating in thecontext of a completely post-mitotic tissue, such as thebrain, gene transfer can ideally be obtained by usingneurotropic herpes viral vectors, which are renderedconditionally replicative after deletion of non-essentialgenes (Lachmann and Efstathiou, 1997).

Tumor specific cell death has already been provided inanimal glioma models by HSV vectors, deleted inneurovirulence genes, such as γ 34.5, thymidine kinase andribonucleotide reductase (Chambers et al, 1995; Mineta etal, 1995; Boviatsis et al, 1994; Miyatake et al, 1997;McKie et al, 1996; Andreansky et al, 1996). Their directinjection into gliomas produced tumor regression withminimal bystander effects on surrounding normal tissue.Enhancement of replication of defective HSV vectorslacking γ 34.5 gene and a significant reduction of tumormass was observed combining ionizing radiation (Advaniet al, 1998).

Another way to treat malignant gliomas emerged fromthe discovery that these tumors often express functionalFas (CD95) (Weller et al, 1994). Fas is a transmembraneglycoprotein belonging to the nerve growth factor/TNFreceptor superfamily: when activated, Fas can transduce anapoptotic signal through its cytoplasmic domain.Apoptosis is triggered by the binding of Fas to its naturalligand (FasL) or by cross-linking with anti-Fas antibodies.A high proportion of human glioma cell lines are sensitiveto apoptosis mediated by anti-Fas antibodies in vitro.Some other cell lines are resistant, but may be renderedsensitive after stable transfection with human Fas cDNA(Weller et al, 1995). These results offer new possibilitiesfor treating gliomas with anti-Fas antibodies or solubleFasL. One possible drawback of such an approach is thatother Fas-positive cells may be affected, like infiltratingleukocytes. Thus their activity may be reduced, restrictingstrategies relying upon simultaneous immune responseenhancement.

IV. Immunotherapy of cancerGlioblastoma multiforme has the propensity to

microscopically infiltrate normal structures since its earlystage of development. This characteristic makes therapeuticsuccess rather difficult by any approach. Moreover, itbecomes virtually impossible to obtain specific targetingof all tumor cells and sparing of normal ones. Therefore,inducing an efficient immune response against malignantcells becomes an attractive and essential treatment strategy.In this perspective, the natural circulatory properties ofcells of the immune system offer also an importantsupport for the recognition of secondary lesions.

Despite the location in the central nervous system(CNS), a long-believed “immunologically privileged site”,glioblastoma cells may interact with immune cells. Theseinteractions are mediated by receptor-ligand recognitionduring cell to cell contact and by a plethora of cytokines.An imbalance in the tumor-host relationship, resulting indeficit in some components of the response, may explainthe aggressive growth of malignant gliomas. Beforediscussing the designed strategies to increase the immuneresponse against glioblastoma cells, we review the morerecent acquisitions in the “dialogue” between theseneoplastic cells and the immune system (Dietrich et al,1997).

A. Antigenicity of glioblastoma cellsThe presence of specific antigens at the surface of

tumor cells to be recognized by cells of the immunesystem is essential for the generation of a specificantitumor immune response. At present, no tumor antigenable to elicit an immune response has been identified inglioblastoma in vivo. MAGE-1 (melanoma antigen) wasthe first tumor-specific antigen to be identified (Van derBruggen et al, 1991), and MAGE family members areexpressed by some glioblastoma cell lines (Rimoldi et al,1993), but not in uncultured tumors (De Smet et al, 1994).This observation could be explained with different levels ofDNA methylation induced by culture, where MAGEexpression was regulated by methylation (De Smet et al,1995).

Proteins that are structurally altered during malignanttransformation, or that contribute to this process, arepossible tumor antigen candidates. The frequent alterationsof p53 in the early stages of carcinogenesis, for example,may provide new antigenic peptides that could trigger animmune response. Consistent with this possibility,specific cytotoxic T-cell clones were generated in vitroagainst mutated p53 protein (Houbiers et al, 1993).Additionally, in vivo immunization with mutated p53peptide was shown to induce specific CTL clones able tolyse MHC-matched tumor cells expressing the mutated p53gene (Noguchi et al, 1994).

The future identification of glioblastoma-specificantigens would aid new treatment strategies.

B. Tumor-induced immunosuppressionA high proportion of glioblastomas have shown to be

infiltrated by lymphocytes, mostly T lymphocytes, butalso B and NK cells. Differently from what reported forother tumors, the level of lymphocyte infiltration does notrelate with a better prognosis. In fact, tumor-infiltratinglymphocytes (TIL) of gliomas appear to be functionallydefective. Abnormalities range from abnormal


139

hypersensitivity responses, depressed response tomitogens, decreased humoral responses (CD4+ T helpercells deficit?), and impaired T cell-mediated cytotoxicity.These functional alterations may be explained, at least inpart, by a defective high-affinity IL-2 receptor (IL-2 R)(Roszman et al, 1991).

It has been demonstrated that glioblastoma cellsproduce and release soluble factors that are responsible for adepressed immune response. T lymphocytes from normalindividuals exhibit immunologic abnormalities whengrown in presence of supernatant obtained fromglioblastoma cell line cultures (Roszman et al, 1991). Themost important soluble suppressing factor seems to beTGF-β2. This cytokine acts as a growth-inhibitory factor(Sporn et al, 1986; Sporn et al, 1987), and has a definedvariety of immunoregulatory properties, including:inhibition of: (i ) T cell proliferation; (i i ) IL-2R induction(Kehrl et al, 1986); (i i i ) cytokine production (Espevik etal, 1987; Espevik et al, 1988); (iv ) natural killer cellactivity (Rook et al, 1986); (v ) cytotoxic T celldevelopment (Jin et al, 1989; Ranges et al, 1987); (vi )LAK cell generation (Espevik et al, 1987; Jin et al, 1989);and (vi i ) production of tumor-infiltrating lymphocytes(Kuppner et al, 1989). Most cells secrete TGF-β in a latentform (Sporn et al, 1986), but glioblastoma cells have alsothe capacity to convert it to an active form, throughproteolytic cleavage. This was demonstrated by anexperimental work in which T-cell suppression mediatedby TGF-β2 was inhibited when proteolytic enzymes wereblocked by protease inhibitors (Huber et al, 1992).

Other soluble factors, namely prostaglandin E2 (PGE2),IL-1 receptor (IL-1 R) antagonist, and interleukin-10 (IL-10) may be implicated in immunosuppression, although invivo intervention has not been fully defined. A potentialdown regulation of some immune functions was shown forIL-10 cytokine, which was produced both by GBM cellsand normal brain tissue (Nitta et al, 1994; Merlo et al,1993). Furthermore, in different animal models, human andmurine IL-10 was demonstrated to stimulate the acquisitionof a specific and efficient antitumor immune response(Berman et al, 1996).

C. Costimulatory moleculesA complete T-cell effector function needs not only

antigen presentation, but also the delivering of activationsignals to the T cell, which is mediated by the so-calledcostimulatory molecules. Unresponsiveness of T cells(anergy) may be due to absence of the second signal,essentially given by the B7-CD28 interaction (June et al,1994). Two other members of the B7 family have beencloned, B7.1 and B7.2; their counter-receptors on T cellsare CD28 and CTLA-4, respectively. CD28 mediatesstimulatory effects, while CTLA-4 appears to be a negative

regulator of T cell responses. Glioblastoma cells andmonocytes that infiltrate the tumor are not expressing B7costimulatory molecules, while monocytes in the normaltissue that surrounds the tumor are B7-positive (Tada et al,1996). This suggests the possible intervention of localmechanisms able to down-regulate B7 expression inglioblastomas, impeding efficient T-cell priming andfavoring T-cell anergy. Moreover, B7-CD28 interactions inthe CNS have been shown to be essential to generate avalid CTL response towards viral antigens (Kuendig et al,1996). Hence, restoring B7 expression by gene transfermight become an interesting task to elicit a properimmune recognition of glioblastoma cells, and anappropriate immune response.

V. Restoring a proper immune responseMany approaches have been tented to restore a proper

immune response towards malignant gliomas. Aspreviously stated, T lymphocytes play a major role in theantitumor response, and priming of T lymphocytes requiresantigen recognition, with or without help from APC.Since no specific antigens have yet been identified forglioblastomas, a vaccination approach has been proposedby administration of genetically modified tumor cells.Moreover, tumor cells transfected to produce variouscytokines have been used to enhance lymphocyteresponsiveness in animal models.

The most interesting results were obtained with cells ofmurine glioma transfected with an expression vectorcontaining the murine interleukin 7 cDNA (Aoki et al,1992). IL-7 transfected glioma cells were vigorouslyrejected by a CD8+ T-cell-mediated immune response, thatwas proportional to the level of IL-7 production.Moreover, the response was tumor-specific, since no effectwas observed against other syngeneic tumor cells(melanoma and fibrosarcoma cells). IL-7 is a veryinteresting cytokine being able to increase IL-2R α chainexpression on CD4+ T lymphocytes and to inhibit TGF-βmRNA expression and production by murine macrophages(Dubinett et al, 1993). IL-12 can also be considered apromising agent to enhance the antitumor response, sinceit augments T-cell and natural killer-cell activities, inducesIFN-γ production, and promotes the differentiation ofuncommitted T cells to Th1 cells (Hendrzak and Brunda,1995). Vector-mediated delivery of IL-12 into establishedtumors suppresses tumor growth (Caruso et al, 1996) andcan induce immune responses against challenge tumors(Bramson et al, 1996). Moreover, IL-12 has other non-immune properties such as anti-angiogenic effects (Voestet al, 1995).

IL-4, another cytokine with pleiotropic functions,increases T cell proliferation and cytotoxicity, and enhanceseosinophil and B cell proliferation and differentiation. It


140

also exerts direct anti-proliferative effects in vitro againstmany tumor cell lines (Tepper, 1993). Antitumor effectsinduced by IL-4-transfected cells were reported in nudemice, suggesting T cell-independent mechanisms (Tepper,1993; Yu et al, 1993). A strong recruitment of eosinophilsand subsequent inhibition of the tumor growth was noted.Eosinophil depletion was not performed as a control; hencea direct anti-proliferative effect mediated by IL-4 cannot beexcluded. In any respect, a possible non-T-cell-dependentmechanism could be an advantage in the glioblastomasetting, considering the various abnormalities of T cellfunction.

In an attempt to overcome the localimmunosuppression mediated by TGF-β2 Fakhrai et alconducted an experimental work on rats by antisense genetherapy. 9L gliosarcoma transfected cells inoculated sub-cutaneously became highly immunogenic and were able toinduce eradication of an established wild-type tumor(Fakhrai et al, 1996).

The enhancement of antigen presentation and T cell co-stimulation has also been considered and may be achievedwith genes coding for cytokines, like GM-CSF, co-stimulatory molecules, like B7, or CIITA, a transcriptionfactor playing a critical role in the regulation of MHCclass II molecules. Encouraging results have been reportedin a murine melanoma model located in the CNS, wherebyan efficient antitumor response was induced by sub-cutaneous vaccination with irradiated, GM-CSF-producingtumor cells (Sampson et al, 1996). Vaccination with cellsco-transfected with B7 and IL-2 was able to mediaterejection of established tumors (Gaken et al, 1997),suggesting a possible application of such an interventionfor the treatment of glioblastomas.

VI. Combined gene therapy approachin humans

The strict localization of glioblastomas in the CNS,with only exceptional metastases, makes these tumorscandidates for approaches of direct intra-tumoral genedelivery. Retrovirus-mediated gene therapy of GBM isparticularly attractive, since these viral vectors transduceonly mitotically active cells, sparing the normal neuronaltissue composed of non-replicating cells.

Gene therapy of brain tumors by intra-tumoralinjection of retroviral vector producing cells (RVPC) inhuman patients was initiated by Oldfield and colleagues in1993 (Oldfield et al, 1993). The gene being transferred wasthat expressing the herpes simplex thymidine kinase(HSV-TK), which conferred sensitivity to the anti-herpesdrug ganciclovir. This treatment has proved free of toxicityand safe for there was no evidence of systemic spread of theretroviral vector (Long et al, 1998). However, a clinicalbenefit was limited to very small tumors (1.5 ml),probably because only malignant cells adjacent to theRVPC were transfected (Ram et al, 1997).

A new treatment strategy combining two differentmodalities, enzyme-directed prodrug activation (tumorsuicide) along with cytokine-promoted tumor rejection, hasbeen recently devised to amplify the antitumor response,and proved to be efficacious in animal models (Castleden etal, 1997) (Figure 2). A bicistronic retroviral vector co-expressing HSV-TK and human interleukin-2 genes hasbeen designed to pursue this new approach of cancer genetherapy in humans (Pizzato et al, 1998).

F i g u r e 2 . Structure of a bicistronic retroviral vector for transduction of genes coding for a cytokine and a prodrug-activating enzyme, expressed via a cap-dependent and an internal ribosome entry site (IRES)-dependent mechanism,respectively. A selectable marker (neomycin phosphotransferase gene, neo) is expressed under the control of a SV40 promoter.


141

Figure 3 . Gene therapy approach for treatment of glioblastoma multiforme, via intratumoral stereotactic injection of cellsproducing a triple gene retroviral vector.

F i g u r e 4 . Contrast-enhanced MRI sagittal images of left parietal GBM lesion before (l e f t ), and one month aftercompletion of GCV treatment (to the r ight ).


142

Figure 5 . Histology of stereotactic biopsies from patients treated by HSV/TK-IL-2 combined gene therapy. A) Toluidine bluestaining - Evidence of a large number of infiltrating inflammatory cells; Immunostaining with B ) monoclonal antibodiesmarking CD3+ cells; C) Mac 387 antibodies recognizing young/activated macrophages; D) monoclonal antibodies markingCD1+ infiltrating cells.

After in vitro characterization of efficacy and safety(Pizzato et al, 1998), the vector was employed in a pilotstudy to treat four patients with recurrent glioblastomamultiforme (Colombo et al, 1997; Palù et al, 1998)(Figure 3). A significant and sustained reduction (>50%of the initial volume) of the tumor mass (80 ml) wasdemonstrated by magnetic resonance imaging (MRI) andcomputerized tomography (CT) in one patient (Figure 4).In this case, the objective response was associated with adramatic clinical improvement. The other three patientsshowed areas of tumor necrosis (2 ml) around the site ofstereotactic RVPC injection and stabilized disease for along period of time (11-12 months).

In stereotactic biopsies taken before gancicloviradministration, large tumor infiltrates of immune-inflammatory cells (T lymphocytes, mostly CD4+ but alsoCD8+ granzyme B-positive cells, activated macrophages,NK cells, neutrophils) were present, notwithstanding thestandard steroid therapy (Figure 5 ). The observedinflammatory response has never been reported in previous

trials with thymidine kinase (Ram et al, 1997; Ostertagand Chiocca, personal communications).

Interestingly, endothelial cells stained positive for TKby in situ hybridization, indicating that the vector hadtargeted the neo-vascular component, a highly replicativepopulation in glioblastomas. This is consistent with ananti-angiogenic effect of this therapeutic approach, that, inaddition to direct tumor suicide and immune activation,may be relevant to the bystander phenomenon and to theclinical response. It is noteworthy that IL-2 wasmeasurable in the cerebral-spinal fluid, even after GCVtreatment. This cytokine might have derived from anautocrine-paracrine secretion of recruited infiltratingimmune-inflammatory cells, after primary expression intransduced cells.

Efforts to achieve more efficient gene transfer systemsare being sought for. These include the development ofnew generation retroviral vectors, produced at higher titresand characterized by higher transduction efficiency.Strategies involving envelope pseudotyping, use of new


143

packaging cell lines of human origin, and substitution ofpromoter elements will contribute to the improvement ofcurrent available vectors.

New therapeutic gene combinations should also beaccomplished in order to promote a more generalizedimmune response. Genes for cytokines other than IL-2(i.e., IL-4, IL-7, IL-12, GM-CSF) as well as genestargeting neoangiogenesis deserve further consideration forcombined treatment approaches.

AcknowledgementsThe authors wish to acknowledge Fondazione Cassa di

Risparmio di Padova e Rovigo and Regione Veneto forfinancial support.

ReferencesAdvani SJ, Sibley GS, Song PY, Hallahan DE, Kataoka Y,

Roizman B, Weichselbaum RR. (1 9 9 8 ) Enhancement ofreplication of genetically engineered herpes simplexviruses by ionizing radiation: a new paradigm fordestruction of therapeutically intractable tumors. GeneTher 5, 160-165.

Andreansky SS, He B, Gillespie GY, Soroceanu L, Markert J,Chou J, Roizman B, Whitley RJ. (1 9 9 6 ) The applicationof genetically engineered herpes simplex viruses to thetreatment of experimental brain tumors. Proc Natl AcadSci USA 93, 11313-11318.

Aoki T, Tashiro K, Miyatake S, Kinashi T, Nakano T, Oda Y,Kikuchi H, Honjo T. (1 9 9 2 ) Expression of murineinterleukin 7 in a murine glioma cell line results in reducedtumorigenicity in vivo Proc Nat l Acad Sc i USA 89,3850-3854.

Badie B, Drazan KE, Kramar MH, Shaked A, Black KL. (1 9 9 5 )Adenovirus-mediated p53 gene delivery inhibits 9L gliomagrowth in rats. Neurol Res 17, 209-216.

Barba D, Hardin J, Ray J, Gage FH. (1 9 9 3 ) Thymidinekinase-mediated killing of rat brain tumors. J Neurosurg70, 175-82,.

Barba D, Hardin J, Sadelain M, Gage FH. (1 9 9 4 )Development of antitumor immunity following thymidinekinase-mediated killing of experimental brain tumors.Proc Natl Acad Sci USA 91, 4348-4352.

Berman RM, Suzuki T, Tahara H, Robbins PD, Narula SK,Lotze MT. (1 9 9 6 ) Systemic administration of cellular IL-10 induces an effective, specific, and long-lived immuneresponse against established tumors in mice. J Immunol157, 231-238

Boviatsis EJ, Scharf JM, Chase M, Harrington K, Kowall NY,Breakefield XO, Chiocca EA. (1 9 9 4 ) Antitumor activityand reporter gene transfer into rat brain neoplasmsinoculated with herpes simplex virus vectors defective inthymidine kinase or ribonucleotide reductase. Gene Ther1, 323-31.

Bramson JL, Hitt M, Addison CL, Muller WJ, Gauldie J,Graham FL. (1 9 9 6 ) Direct intratumoral injection of anadenovirus expressing interleukin-12 induces regressionand long-lasting immunity that is associated with highlylocalized expression of interleukin-12. Hum Gene Ther7, 1995-2002

Caruso M, Pham-Nguyen K, Kwong Y-L, Xu B, Kosai K-I,Finegold M, Woo SLC, Chen SH. (1 9 9 6 ) Adenovirus-mediated interleukin-12 gene therapy for metastatic coloncancer. Proc Natl Acad Sci USA 93, 11302-11306

Castleden SA, Chong H, Garcia-Ribas I, Melcher AA,Hutchinson G, Roberts B, Hart IR, Vile RG. (1 9 9 7 ) Afamily of bicistronic vectors to enhance both local andsystemic antitumor effects of HSVTK or cytokineexpression in a murine melanoma model. Hum GeneTher 8, 2087-102.

Chamberlain MC, Kormanik P. (1 9 9 6 ) Salvagechemotherapy with paclitaxel for recurrent primary braintumors. J Cl in Oncol 13, 2316-21.

Chambers R, Gillespie GY, Soroceanu L, Andreansky S,Chatterjee S, Chou J, Roizman B, Whitley RJ. (1 9 9 5 )Comparison of genetically engineered herpes simplexviruses for the treatment of brain tumors in a scid mousemodel of human malignant glioma. Proc Natl Acad SciUSA 92, 1411-1415.

Chen CY, Chang YN, Ryan P, Linscott M, McGarrity GJ,Chiang YL. (1 9 9 5 ) Effect of herpes simplex virusthymidine kinase expression levels on ganciclovir-mediated cytotoxicity and the "bystander effect". HumGene Ther 6, 1467-1476.

Cheney IW, Johnson DE, Vaillancourt MT, Avanzini J,Morimoto A, Demers GW, Wills KN, Shabram PW, BolenJB, Tavtigian SV, Bookstein R (1 9 9 8 ) Suppression oftumorigenicity of glioblastoma cells by adenovirus-mediated MMAC1/PTEN gene transfer. Cancer Res 58,2331-2334.

Chiariello E, Roz L, Albarosa R, Magnani I, Finocchiaro G.(1 9 9 8 ) PTEN/MMAC1 mutations in primaryglioblastomas and short-term cultures of malignantgliomas. Oncogene 16, 541-545.

Chintala SK, Fueyo J, Gomez Manzano C, Venkaiah B,Bjerkvig R, Yung WK, Sawaya R, Kyritsis AP, Rao JS.(1 9 9 7 ) Adenovirus-mediated p16/CDKN2 gene transfersuppresses glioma invasion in vitro. Oncogene 15, 17,2049-57.

Chiocca EA. (1 9 9 5 ) Brain tumor gene therapy in mice with anovel "suicide" gene, the cyclophosphamide-activatingCYP2B1 gene. Clin Neurosurg 42, 370-82.

Collins VP. (1 9 9 3 ) Amplified genes in human gliomas.Semin Cancer Biol 4, 27-32.

Colombo F, Zanusso M, Casentini L, Cavaggioni A, FranchinE, Calvi P, Palù G. (1 9 9 7 ) Gene stereotactic neurosurgeryfor recurrent malignant gliomas. Stereotact FunctNeurosurg 68, 245-251.


144

Coyle T, Baptista J, Winfield J. (1 9 9 0 ) Mechlorethamine,vincristine and procarbazine chemotherapy for recurrenthigh-grade glioma in adults, A phase II study. J C l inOncol 8, 2014-2018.

Culver KW, Ram Z, Wallbridge S, Ishii I, Oldfield EH, BlaeseRM. (1 9 9 2 ) In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors.Sc ience 256, 1550-1552.

De Smet C, Courtois SJ, Faraoni I, Lurquin C, Szikora JP, DeBacker O, Boon T. (1 9 9 5 ) Involvement of two Etsbinding sites in the transcriptional activation of theMAGE-1 gene. Immunonogenet ics 42, 282-290.

De Smet C, Lurquin C, Van Der Bruggen P, De Plaen E,Brasseur F, Boon T. (1 9 9 4 ) Sequence and expressionpattern of the human MAGE 2 gene. Immunogenet ics39, 121-129.

Dietrich P-Y, Walker PR, Saas P, de Tribolet N. (1 9 9 7 )Immunobiology of Gliomas, New perspectives for therapy.Ann NY Acad Sci 824, 124-140.

Dubinett SM, Huang M, Dhanani S, Wang J, Beroiza T.(1 9 9 3 ) Down-regulation of macrophage transforminggrowth factor-beta messenger RNA expression by IL-7. JImmunol 151, 6670-6680.

Ekstrand AJ, Longo N, Hamid ML, Olson JJ, Liu L, CollinsVP, James CD. (1 9 9 4 ) Functional characterization of anEGF receptor with a truncated extracellular domainexpressed in glioblastomas with EGFR geneamplifacation. Oncogene 9, 2313-20.

Espevik T, Figari IS, Ranges GE, Palladino MA Jr. (1 9 8 8 )Transforming growth factor-β1 (TGF-β1) and recombinant

human tumor necrosis factor α reciprocally regulate thegeneration of lymphokine-activated killer cell activity.Comparison between natural porcine platelet-derived TGF-β1 and TGF-β2, and recombinant human TGF-β1. JImmunol 140, 2312-2316.

Espevik T, Figari IS, Shalaby MR, Lackides GA, Lewis GD,Shepard HM, Palladino MA Jr. (1 9 8 7 ) Inhibition ofcytokine production by cyclosporin A and transforminggrowth factor . J Exp Med 166, 571-576.

Ezzeddine ZD, Martuza RL, Platika D, Short MP, Malick A,Choi B, Breakefield XO. (1 9 9 1 ) Selective killing ofglioma cells in colture and in vivo by retrovirus transfer ofthe herpes simplex virus thymidine kinase gene. NewB i o l 3, 608-614.

Fakhrai H, Dorigo O, Shawler DL, Lin H, Mercola D, BlackKL, Royston I, Sobol RE. (1 9 9 6 ) Eradication ofestablished intracranial rat gliomas by transforminggrowth factor beta antisense gene therapy. Proc NatlAcad Sci USA 93, 2909-2914.

Fewer D, Wilson CB, Boldrey EB. (1 9 7 2 ) Phase II study ofCCNU in the treatment of brain tumors. Cancer ChemRep , 56, 421-7.

Fine HA, Dear KBG, Loeffler JS. (1 9 9 3 ) Meta-analysis ofradiation therapy with an without adjuvant chemotherapyfor malignant glionas in adults. Cancer 71, 2585-97.

Forsyth PA, Cairncross JG. (1 9 9 6 ) Chemotherapy ofmalignant gliomas. In Cerebral Gliomas. Bai l lere 'sClin Neurol 5, 371-93.

Fueyo J, Gomez Manzano C, Yung WK, Clayman GL, Liu TJ,Bruner J, Levin VA, Kyritsis AP. (1 9 9 6 ) Adenovirus-mediated p16/CDKN2 gene transfer induces growth arrestand modifies the transformed phenotype of glioma cells.Oncogene 12, 103-110.

Fueyo J, Gomez Manzano C, Yung WK, Liu TJ, Alemany R,McDonnell TJ, Shi X, Rao JS, Levin VA, Kyritsis AP.(1 9 9 8 ) Overexpression of E2F-1 in glioma triggersapoptosis and suppresses tumor growth in vitro and invivo. Nat Med 4, 685-690.

Fujiwara T, Grimm EA, Mukhopadhayay T, Zang WW, Owen-Schaub LB, Roth JA. (1 9 9 4 ) Induction ofchemosensitivity in human lung cancer in vivo byadenoviral mediated tranfer of the wild type p53 gene.Cancer Res 54, 2287-2291.

Furnari FB, Su Huang HJ, Cavenee WK. (1 9 9 5 ) Genetics andmalignant progression of human brain tumors. CancerSurv 25, 233-275.

Gaken JA, Hollingsworth SJ, Hirst WJR, Buggins AGS, GaleaJ, Peakman M, Kuiper M, Patel P, Towner P, Patel PM,Collins MK, Mufti GJ, Farzaneh F, Darling DC. (1 9 9 7 )Irradiated NC adenocarcinoma cells transduced with bothB7.1 and interleukin-2 induce CD4+-mediated rejection ofestablished tumors. Hum Gene Ther 8, 477-488.

Ge K, Xu L, Zheng Z, Xu D, Sun L, Liu X. (1 9 9 7 ) Transductionof cytosine deaminase gene makes rat glioma cells highlysensitive to 5-fluorocytosine. Int J Cancer 71, 675-9.

Gjerset RA, Turla ST, Sobol RE, Scalise JJ, Mercola D,Collins H, Hopkins PJ. (1 9 9 5 ) Use of wild type p53 toachieve complete treatment sensitization of tumor cellsexpressing endogenous mutant p53. Mol Carc inog 14,275-285.

Gomez-Manzano C, Fueyo J, Kyritsis AP, Steck PA, Roth JA,McDouwell TJ, Steck KD, Levin VA, Yung WK. (1 9 9 6 )Adenovirus-mediated tranfer of the p53 gene produces rapidand generalized death of human glioma cells via apoptosis.Cancer Res 56, 694-699.

Griscelli F, Li H, Bennaceur Griscelli A, Soria J, Opolon P,Soria C, Perricaudet M, Yeh P, Lu H. (1 9 9 8 ), Angiostatingene transfer, inhibition of tumor growth in vivo byblockage of endothelial cell proliferation associated with amitosis arrest. Proc Natl Acad Sci USA 95, 11, 6367-6372.

Hamel W, Magnelli L, Chiarugi VP, Israel MA. (1 9 9 6 ) Herpessimplex virus thymidine kinase/ganciclovir-mediatedapoptotic death of bystander cells. Cancer Res 56,2697-2702.

He J, Allen JR, Collins VP, Allalunis-Turner MJ, Godbout R,Day RS 3rd, James CD. (1 9 9 4 ) CDK4 amplification is analternative mechanism to p16 gene homozygous deletionin glioma cell lines. Cancer Res 54, 5804-5807.


145

He J, Olson JJ, James CD. (1 9 9 5 ) Lack of p16ink4 orretinoblastoma protein (pRb) or amplification-associatedoverexpression of cdk4 is observed in dinstinct subsets ofmalignant glial tumors and cell lines. Cancer Res 55,4833-4836.

Hendrzak JA and Brunda MJ. (1 9 9 6 ) Interleukin-12, biologicactivity, therapeutic utility, and role in disease. LabInves t 72, 619-637.

Henson JW, SchniTKer BL, Correa KM, von Dimling A,Fassbender F, Xu HJ, Benedict WF, Yandell DW, Louis DN.(1 9 9 4 ) The retinoblastoma gene is involved in malignatprogression of astrocytomas. Ann Neurol 36, 714-721.

Hochberg FH, Pruitt A. (1 9 8 0 ) Assumptions in theradiotherapy of glioblastoma. Neurology 30, 907-911.

Hoostraten B, Gottlieb JA, Caoili A. 1 9 7 3 CCNU in thetreatment of cancer, A phase II study. Cancer 32, 38-43.

Hosli P, Sappino AP, de Tribolet N, Dietrich PY. (1 9 9 8 )Malignant glioma, Should chemotherapy be overthrownby experimental treatments? Ann Oncol 9, 589-600.

Houbiers JG, Nijman HW, Van Der Burg SH, Drijfhout JW,Kenemans P, Van Der Velde CJ, Brand A, Momburg F, KastWM, Melief CJ. (1 9 9 3 ) In vitro induction of humancytotoxic T lymphocyte responses against peptides ofmutant and wild-type p53. Eur J Immunol 23, 2072-2077.

Huber D, Philipp J, Fontana A. (1 9 9 2 ) Protease inhibitorsinterfere with the transforming growth factor-(-indipendentpathway of tumor cell-mediated immunosuppression. JImmunol 148, 277-284.

Ichimura K, Schmidt EE, Miyakawa A, Goike HM, Collins VP.(1 9 9 8 ) Distinct patterns of deletion on 10p and 10qsuggest involvement of multiple tumor suppressor genesin the development of astrocytic gliomas of differentmalignancy grades. Genes Chromosomes Cancer 22,9-15.

Janzer RC, Raff MC. (1 9 8 7 ) Astrocytes induce blood brainbarrier properties in eÿdotelial cells. Nature 325, 253-257.

Jen J, Harper JW, Bigner SH, Bigner DD, Papadopoulos N,Markowitz S, Willson JK, Kinzler KW, Vogelstein B.(1 9 9 4 ) Deletion of p16 and p15 genes in brain tumors.Cancer Res 54, 6353-6358.

Jin B, Scott JL, Vadas MA, Burns GF. (1 9 8 9 ) TGF-β downregulates TLiSA1 expression and ihibits the differentiationof precursor lymphocytes into CTL and LAK cells.Immunology 66, 570-576

June CH, Bluestone JA, Nadler LM, Thompson CB. (1 9 9 4 )The B7 and CD28 receptor families. Immunol Today 15,321-331

Ke LD, Fueyo J, Chen X, Steck PA, Shi YX, Im SA, Yung WK.(1 9 9 8 ) A novel approach to glioma gene therapy, down-regulation of the vascular endothelial growth factor inglioma cells using ribozymes. Int J Oncol 12, 1391-6.

Kerhl JH, Wakefield LM, Roberts AB, Jakowlew S, Alvarez-Mon M, Derynck R, Sporn MB, Fauci AS. (1 9 8 6 )Production of transforming growth factor ( by human Tlymphocytes and its potential role in the regulation of Tcell growth. J Exp Med 163, 1037-1050.

Kim JH, Kim SH, Brown SL, Freitag SO. (1 9 9 4 ) Selectiveenhancement by an antiviral agent of the radiation inducedcell killing of human glioma cells transduced with HSV-TKgene. Cancer Res 54, 6053-6056.

Kim JH, Kim SH, Kolozsvary A,Brown SL, Kim OB, FreytagSO. (1 9 9 5 ) Selective enhancement of radiation responseof herpes simplex virus thymidine kinase transduced 9Lgliosarcoma cells in vitro and in vivo by antiviral agents.Int J Radiat Oncol Biol Phys 33, 861-868.

Kim SH, Kim JH, Kolozsvary A, Brown SL, Freytag SO.(1 9 9 7 ) Preferential radiosensitization of 9L glioma cellstransduced with HSV-TK gene by acyclovir. JNeurooncol 33, 189-194.

Kondo S, Barna BP, Kondo Y, Tanaka Y, Casey G, Liu J,Morimura T, Kaakaji R, Peterson JW, Werbel B, BarnettGH. (1 9 9 6 ) WAF1/CIP1 increases the susceptibility ofp53 non-functional malignant glioma cells to cisplatin-induced apoptosis. Oncogene 13, 1279-85.

Kondo S, Kondo Y, Li G, Silverman RH, Cowell JK. (1 9 9 8 )Targeted therapy of human malignant glioma in a mousemodel by 2-5A antisense directed against telomerase RNA.Oncogene , 16, 25, 3323-30.

Kundig TM, Shahinian K, Kawai K, Mittruecker HW, Sebzda E,Bachmann MF, Mak TW, Ohashi PS. (1 9 9 6 ) Duration ofTCR stimulation determines costimulatory requirement of Tcells. Immunity , 5, 41-52

Kuppner MC, Hamou M-F, Sawamura Y, Bodmer S, de TriboletN. (1 9 8 9 ) Inhibition of lymphocyte function byglioblastoma-derived transforming growth factor β2. JNeurosurg 71, 211-217

Lachmann RH, Efstathiou S. (1 9 9 7 ) The use of herpessimplex virus-based vectors for gene delivery to thenervous system. Mol Med Today 3, 404-411.

Lang FF, Yung WK, Raju U, Libunao F, Terry NH, Tofilon PJ.(1 9 9 8 ) Enhancement of radiosensitivity of wild-type p53human glioma cells by adenovirus-mediated delivery of thep53 gene. J Neurosurg 89, 1, 125-132.

LeMay DR, Kittaka M, Gordon EM, Gray B, Stins MF,McComb JG, Jovanovic S, Tabrizi P, Weiss MH, Bartus R,Anderson WF, Zlokovic BV. (1 9 9 8 ) Intravenous RMP-7increases delivery of ganciclovir into rat brain tumors andenhances the effects of herpes simplex virus thymidinekinase gene therapy. Hum Gene Ther 9, 989-995.

Levin VA, Edwards MS, Wright DC. (1 9 8 0 ) Modifiedprocarbazine, CCNU and vincristine (PCV3) combinationchomtherapy in the treatment of malignant brain tumors.Cancer Treat Rep 64, 237-241.

Levin VA, Gutin PH, Leibel S. Cancer, Principles and Practiceof Oncology. In De Vita VT Jr, Hellman S, Rosemberg SA


146

(eds) (1 9 9 3 ) , Neoplasms o f the Centra l NervousSys tem , Philadelphia, JB Lippincot, 1679-737.

Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J,Miliaresis C, Rodgers L, McCombie R, Bigner SH,Giovanella BC, Ittmann M, Tycko B, Hibshoosh H, WiglerMH, Parsons R. (1 9 9 7 ) PTEN, a putative protein tyrosinephosphatase gene mutated in human brain, breast, andprostate cancer. S c i e n c e , 275, 1943-1947.

Libermann TA, Nusbaum HR, Razon N, Kris R, Lax I, Soreq H,Whittle N, Waterfield MD, Ullrich A, Schlessinger J.(1 9 8 5 ) Amplification, enhanced expression and possiblerearrangement of EGF receptor gene in primary humanbrain tumors of glial origin. Nature 313, 144-147.

Long Z, Li LP, Grooms T, Lockey C, Nader K, Mychkovsky I,Mueller S, Burimski I, Ryan P, Kikuchi G, Ennist D,Marcus S, Otto E, McGarrity G. (1 9 9 8 ) Biosafetymonitoring of patients receiving intracerebral injectionsof murine retroviral vector producer cells. Hum GeneTher 9, 1165-1172.

Lowe SW, Bodis S, McClatchey A, Remington L, Ruley HE,Fisher DE, Housman DE, Jacks T. (1 9 9 4 ) p53 status andthe efficacy of cancer therapy in vivo. Sc ience 266, 807-810.

Mabuchi E, Shimizu K, Miyao Y, Kaneda Y, Kishima H,Tamura M, Ikenaka K, Hayakawa T. (1 9 9 7 ) Gene deliveryby HVJ-liposome in the experimental gene therapy ofmurine glioma. Gene Ther 4, 768-72.

Mak M, Fung L, Strasser JF, Saltzman WM. (1 9 9 5 )Distribution of drugs following controlled delivery to thebrain interstitium. J Neurooncol 26, 91-102.

Manome Y, Kunieda T, Wen PY, Koga T, Kufe DW, Ohno T.(1 9 9 8 ) Transgene expression in malignant glioma usinga replication-defective adenoviral vector containing theEgr-1 promoter, activation by ionizing radiation or uptakeof radioactive iododeoxyuridine. Hum Gene Ther 9,1409-1417.

Manome Y, Wen PY, Chen L, Tanaka T, Dong Y, Yamazoe M,Hirshowitz A, Kufe DW, Fine HA. (1 9 9 6 ) Gene therapyfor malignant gliomas using replication incompetentretroviral and adenoviral vectors encoding the cytochromeP450 2B1 gene together with cyclophosphamide. GeneTher 3, 513-520.

Manome Y, Wen PY, Dong Y, Tanaka T, Mitchell BS, KufeDW, Fine HA. (1 9 9 6 ) Viral vector transduction of thehuman deoxycytidine kinase cDNA sensitizes glioma cellsto the cytotoxic effects of cytosine arabinoside in vitroand in vivo. Nat Med 2, 567-573.

McKie EA, MacLean AR, Lewis AD, Cruickshank G, RamplingR, Barnett SC, Kennedy PG, Brown SM. (1 9 9 6 ) Selectivein vitro replication of herpes simplex virus type 1 (HSV-1)ICP34.5 null mutants in primary human CNS tumours--evaluation of a potentially effective clinical therapy. Br JCancer 74, 745-752.

Merlo A, Juretic A, Zuber M, Filgueira L, Luscher U, CaetanoV, Ulrich J, Gratzl O, Heberer M, Spagnoli G. (1 9 9 3 )Cytokine gene expression in primary brain tumors,

metastases and meningiomas suggests specifictranscription patterns. Eur J Cancer 29A, 2118-2125.

Millauer B, Longhi MP, Plate KH, Shawer LK, Risau W, ulrichA, Strawn LM. (1 9 9 6 ) Dominant-negative inhibition ofFlk-1 suppress the growth of many tumor types in vivo.Cancer Res 56, 1615-1620.

Millauer B, Shawer LK, Plate KH, Risau W, Ulrich A. (1 9 9 4 )Glioblastoma growth inhibited in vivo by a dominant-negative Flk-1 mutant. Nature, 367, 576-579.

Mineta T, Rabkin SD, Yazaki T, Hunter WD, Martuza RL.(1 9 9 5 ) Attenuated multi-mutated herpes simplex virus-1for the treatment of malignant gliomas. Nat Med, 1 , 9 ,938-943.

Miyatake S, Martuza RL, Rabkin SD. (1 9 9 7 ) Defectiveherpes simplex virus vectors expressing thymidine kinasefor the treatment of malignant glioma. Cancer GeneTher 4, 222-228.

Mizuno M, Yoshida J, Colosi P, Kurtzman G. (1 9 9 8 ) Adeno-associated virus vector containing the herpes simplexvirus thymidine kinase gene causes complete regression ofintracerebrally implanted human gliomas in mice, inconjunction with ganciclovir administration. Jpn JCancer Res, 89, 1, 76-80.

Newlands ES, O'Reilly SM, Glaser MG. (1 9 9 6 ) The CharingCross hospital experience with Temozolomide in patientswith gliomas. Eur J Cancer, 32A, 2236-41.

Nguyen DM, Spitz FR, Yen N, Cristiano RJ, Roth JA. (1 9 9 6 )Gene therapy for lung cancer, enhancement of tumorsuppression by a combination of sequential systemiccisplatin and adenovirus-mediated p53 gene transfer. JTorac Cardiovasc Surg, 112, 1372-1377.

Nielsen LL, Maneval DC. (1 9 9 8 ) p53 tumor suppressor genetherapy for cancer. Cancer Gene Ther 5, 52-63.

Nishikawa R, Furnari FB, Lin H, Arap W, Berger MS, CaveneeWK, Su Huang HJ. (1 9 9 5 ) Loss of p16ink4 expression isfrequent in high-grade gliomas. Cancer Res 55, 1941-1945.

Nishikawa R, Ji XD, Harmon RC, Armon RC, Lazar CS, GillGN, Cavenee WK, Huang HG. (1 9 9 4 ) A mutant epidermalgrowth factor receptor common in human gliomas confersenhanced tumorigenicity. Proc Natl Acad Sc i USA91, 7727-7731.

Nitta T, Hishii M, Sato K, Okumura K. (1 9 9 4 ) Selectiveexpression of interleukin-10 gene within glioblastomamultiforme. Brain Res , 649, 122-128.

Noguchi Y, Chen YT, Old LJ. (1 9 9 4 ) A mouse mutant p53product recognized by CD4+ and CD8+ T cells. Proc NatlAcad Sci 91, 3171-3175.

Ohnishi T, Taki T, Hiraga S, Arita N, Morita T. (1 9 9 8 ) Invitro and in vivo potentiation of radiosensitivity ofmalignant gliomas by antisense inhibition of the RAD51gene. B i o c h e m B i o p h y s Res Commun 245, 319-324.

Oldfield EH, Ram Z, Culver KW, Blaese RM, De Vroom HL,Anderson WF. (1 9 9 3 ) Clinical Protocol, Gene therapy for


147

the treatment of brain tumors using intra-tumoraltransduction with the thymidine kinase and intravenousganciclovir. Hum Gene Ther 4, 39-69.

Ono Y, Ikeda K, Wei MX, Harsh GR 4th, Tamiya T, ChioccaEA. (1 9 9 7 ) Regression of experimental brain tumors with6-thioxanthine and Escherichia coli gpt gene therapy.Hum Gene Ther 8, 2043-2055.

O'Rourke DM, Kao GD, Singh N, Park BW, Muschel RJ, WuCJ, Greene MI. (1 9 9 8 ) Conversion of a radioresistantphenotype to a more sensitive one by disabling erbBreceptor signaling in human cancer cells. Proc NatlAcad Sci USA 95, 10842-10847.

O'Rourke DM, Qian X, Zhang HT, Davis JG, Nute E, MeinkothJ, Greene MI. (1 9 9 7 ) Trans receptor inhibition of humanglioblastoma cells by erbB family ectodomains. ProcNatl Acad Sci USA 94, 3250-3255.

Palù G, Cavaggioni A, Calvi P, Franchin E, Pizzato M,Boschetto R, Parolin C, Chilosi M, Ferrini S, Zanusso A,Colombo F. (1 9 9 8 ) Gene therapy of glioblastomamultiforme via combined expression of suicide andcytokine genes, a pilot study in humans. Gene Ther (inpress)

Parr MJ, Manome Y, Tanaka T, Wen P, Kufe D, Kaelin WG Jr,Fine HA. (1 9 9 7 ) Tumor-selective transgene expression invivo mediated by an E2F-responsive adenoviral vector.Nature Med, 3, 1145-9.

Petersdorf SH, Berger MS. (1 9 9 6 ) Concepts in Neurosurgery,The molecular basis of neurosurgical disease. In Raffel C,Harsh IV GR (eds), Molecular Bas i s o fChemotherapy for Brain Tumors, Baltimore,Williams and Wilkins, chapter 12 (Vol 8), 198-210.

Pizzato M, Franchin E, Calvi P, Boschetto R, Ferrini S,Colombo M, Palù G. (1 9 9 8 ) Production andcharacterization of a bicistronic Moloney-based retroviralvector expressing human interleukin 2 and herpes simplexvirus thymidine kinase for gene therapy of cancer. GeneTher 5, 1003-1007.

Prados MD, Schold SC, Spence AM, Berger MS, Mc AllisterLD, Mehta MP, Gilbert MR, Fulton D, Kuhn J, Lamborn K,Rector DJ, Chang SM. (1 9 9 6 ) Phase II study of paclitaxelin patients with recurrent malignant glioma. J C l inOncol 14, 2316-2321.

Rainov NG, Dobberstein KU, Bahn H, Holzhausen HJ,Lautenschläger C, Heidecke V, Burkert W. (1 9 9 7 )Prognostic factors in malignant glioma, influence of theoverexpression of oncogene and tumor-suppressor geneproducts on survival. J Neurooncol 35, 13-28.

Rainov NG, Dobberstein KU, Sena-Esteves M, Herrlinger U,Kramm CM, Philpot RM, Hilton J, Chiocca EA,Breakefield XO. (1 9 9 8 ) New prodrug activation genetherapy for cancer using cytochrome P450 4B1 and 2-aminoanthracene/4-ipomeanol. Hum Gene Ther 9,1261-73.

Ram Z, Culver KW, Oshiro EM, Viola JJ, De Vroom HL, OttoE, Long Z, Chiang Y, McGarrity GJ, Muul LM, Katz D,Blaese RM, Oldfield EH. (1 9 9 7 ) Therapy of malignant

brain tumors by intratumoral implantation of retroviralvector-producing cells. Nature Med 3, 1354-1361

Ram Z, Culver KW, Wallbridge S, Blaese RM, Oldfield EH.(1 9 9 3 ) In situ retroviral-mediated gene transfer for thetreatment of brain tumors in rats. Cancer Res 53, 83-88.

Ranges GE, Figari IS, Espevik T, Palladino MA Jr. (1 9 8 7 )Inhibition of cytotoxic T cell development bytransforming growth factor β and reversal by recombinant

tumor necrosis factor α. J Exp Med 166, 991-998.

Redekop GJ, Naus CC. (1 9 9 5 ) Transfection with bFGF senseand antisense cDNA resulting in modification of malignantglioma growth. J Neurosurg 82, 83-90.

Rimoldi D, Romero P, Carrel S. (1 9 9 3 ) The human melanomaantigen/encoding gene, MAGE-1, is expressed by othertumor cells of neuroectodermal origin such asglioblastomas and neuroblastomas. Int J Cancer 54,527-528.

Rodriguez LA, Prados M, Silver P. (1 9 8 9 ) Reevaluation ofprocarbazione for the treatmentof recurrent malignantcentral nervous system tumors. Cancer 64, 2420-2423.

Rogulski KR, Kim JH, Kim SH, Freytag SO. (1 9 9 7 ) Gliomacells transduced with an Escherichia coli CD/HSV-1 TKfusion gene exhibit enhanced metabolic suicide andradiosensitivity. Hum Gene Ther 8, 73-85.

Rook AH, Kerhl JH, Wakefield LM, Roberts AB, Sporn MB,Burlington DB, Lane HC, Fauci AS. (1 9 8 6 ) Effects oftransforming growth factor ( on the functions of naturalkiller cells, depressed cytolytic activity and blunting ofinterferon responsiveness. J Immunol 136, 3916-3920.

Rosemberg SA, Lotze MT, Muul LM, Chang AE, Avis FP,Leitman S, Linehan WM, Robertson CN, Lee RE, Rubin JT,Seipp CA, Simpson CG, White DE. (1 9 8 7 ) A progressreport on the treatment of 157 patients with advancedcancer using lymphokine-activated killer cells andinterleukin-2 or high dose of interleukin-2 alone. NewEngl J Med , 316, 889-905.

Roszman T, Elliott L, Brooks W. (1 9 9 1 ) Modulation of T-cell function by gliomas. Immunol Today 12, 370-374.

Roth JA, Cristiano R. (1 9 9 7 ) Gene therapy for cancer, Whatwe done and where we are going? J Nat l Cancer Ins t ,89, 21-39.

Russell DS, Rubistein LJ. (1 9 8 9 ) Pa tho logy o f Tumorsof the Central Nervous System , 5th Edition, EdwardArnold, London.

Saleh M, Stacker SA, Wilks AF. (1 9 9 6 ) Inhibition of growthof C6 glioma cells in vivo by expression of antisensevascular endothelial growth factor sequence. Cancer Res56, 393-401.

Sampson JH, Archer GE, Ashley DM, Fuchs HE, Hale LP,Dranoff G, Bigner DD. (1 9 9 6 ) Subcutaneous vaccinationwith irradiated, cytokine-producing tumor cells stimulatesCD8+ cell mediated immunity against tumors located in theimmunologically privileged central nervous system. ProcNatl Acad Sci USA 93, 10399-10404.


148

Sanson M, Ameri A, Monjour A, Sahmoud T, Ronchin P,Poisson M, Delattre JY. (1 9 9 6 ) Treatment of recurrentmalignant supratentorial gliomas with ifosfamide,carboplatin, and etoposide, A phase II study. Eur JCancer 32A, 2229-2235.

Smart CR, Ottomoan RE, Rochlin DB. (1 9 6 8 ) Clinicalexperience with vincristine with tumors of the centralnervous system and other malignant diseases. CancerChem Rep 52, 733-741.

Sporn MB, Roberts AB, Wakefield LM, Assoian RK. (1 9 8 6 )Transforming growth factor-(, biological fusion andchemical structure. Sc ience 233, 532-534

Sporn MB, Roberts AB, Wakefield LM, de Crombrugghe B.(1 9 8 7 ) Some recent advances in the chemistry andbiology of transforming growth factor-beta. J C e l l B i o l105, 1039-1045.

Staba MJ, Mauceri HJ, Kufe DW, Hallahan DE, WeichselbaumRR. (1 9 9 8 ) Adenoviral TNF-alpha gene therapy andradiation damage tumor vasculature in a human malignantglioma xenograft. Gene Ther 5, 293-300.

Sturtz FG, Waddell K, Shulok J, Chen X, Caruso M, Sanson M,Snodgrass HR, Platika D. (1 9 9 7 ) Variable efficiency ofthe thymidine kinase/ganciclovir system in humanglioblastoma cell lines, implications for gene therapy.Hum Gene Ther 8, 1945-53.

Tada M, Diserens AC, Hamou MF, Jaufeerally R, Van Meir E,de Tribolet N. (1 9 9 6 ) Brain Tumor Res Ther , 327-337.

Takamiya Y, Short MP, Ezzeddine ZD, Moolten FL,Breakefield XO, Martuza RL. (1 9 9 2 ) Gene therapy ofmalignant brain tumors, a rat glioma line bearing theherpes simplex virus type 1-thymidine kinase gene andwild type retrovirus kills other tumor cells. J NeuroscRes 33, 493-503.

Takamiya Y, Short MP, Moolten FL, Fleet C, Mineta T,Breakefield XO, Martuza RL. (1 9 9 3 ) An experimentalmodel of retrovirus gene therapy for malignant braintumors. J Neurosurg 79, 104-110.

Tamiya T, Ono Y, Wei MX, Mroz PJ, Moolten FL, ChioccaEA. (1 9 9 6 ) Escherichia coli gpt gene sensitizes ratglioma cells to killing by 6-thioxanthine or 6-thioguanine. Cancer Gene Ther 3, 3, 155-162.

Tamura M, Shimizu K, Yamada M, Miyao Y, Hayakawa T,Ikenaka K. (1 9 9 7 ) Targeted killing of migrating gliomacells by injection of HTK-modified glioma cells. HumGene Ther 8, 381-91.

Tanaka T, Cao Y, Folkman J, Fine HA. (1 9 9 8 ) Viral vector-targeted antiangiogenic gene therapy utilizing anangiostatin complementary DNA. Cancer Res 58, 3362-3369.

Tanaka T, Manome Y, Wen P, Kufe DW, Fine HA. (1 9 9 7 )Viral vector-mediated transduction of a modified plateletfactor 4 cDNA inhibits angiogenesis and tumor growth.Nat Med 3, 437-442.

Tepper RI. (1 9 9 3 ) The antitumor and proinflammatoryactions of IL-4. Res Immunol 144, 633-637.

Tohma Y, Gratas C, Biernat W, Peraud A, Fukuda M, YonekawaY, Kleihues P, Ohgaki H. (1 9 9 8 ) PTEN (MMAC1)mutations are frequent in primary glioblastomas (de novo)but not in secondary glioblastomas. J NeuropatholExp Neurol 57, 684-689.

Trojan J, Johnson TR, Rudin SD, Ilan J, Tykocinski ML, IlanJ. (1 9 9 3 ) Treatment and prevention of rat glioblastomaby immunogenic C6 cells expressing antisense insulin-like growth factor I RNA. Sc ience 259, 94-97.

Van Der Bruggen P, Traversari C, Chomez P, Lurquin C, DePlaen E, Van Den Eynde B, Knuth A, Boon T. (1 9 9 1 ) Agene encoding an antigen recognized by cytolityc Tlymphocytes on a human melanoma. Sc ience 254, 1643-1647

Van Meir E, Roemer K, Diserens AC, Kikuchi T, Rempel SA,Haas M, Huang HG, Friedmann T, de Tribolet N, CaveneeWK. (1 9 9 5 ) Single cell monitoring of growth arrest andmorphological changes induced by transfer of wild-typep53 alleles to glioblastoma cells. Proc Nat l Acad Sc iUSA 92, 1008-1012.

Vincent AJ, Vogels R, Someren GV, Esandi MC, NoteboomJL, Avezaat CJ, Vecht C, Bekkum DW, Valerio D, Bout A,Hoogerbrugge PM. (1 9 9 6 ) Herpes simplex virusthymidine kinase gene therapy for rat malignant braintumors. Hum Gene Ther 7, 197-205.

Voest EE, Kenyon BM, O’Reilly MS, Truitt G, D’Amato RJ,Folkman J. (1 9 9 5 ) Inhibition ofangiogenesis in vivo byinterleukin-12. J Natl Cancer Inst 87, 581-586.

Walker MD, Green SB, Byar DP. (1 9 8 0 ) Randomizedcomparisons of radiotherapy and nitrosoureas for thetreatment of malignant glioma after surgery. N E n g l JMed 303, 1323-1329.

Wei M, Tamiya T, Chase M, Bobiatsis EJ, Chang TK, KowallNW, Hochberg FH, Waxman DJ, Breakefield XO, ChioccaEA. (1 9 9 4 ) Experimental tumor therapy in mice using thecyclophosphamide-activating cytochrome P450 2B1 gene.Hum Gene Ther 5, 969-978.

Weller M, Malipiero U, Rensing Ehl A, Barr PJ, Fontana A.(1 9 9 5 ) Fas/APO-1 gene transfer for human malignantglioma. Cancer Res 55, 2936-2944.

Young RC, Walker MD, Canellos GP. 1 9 7 3 Initial clinicaltrials with methyl-CCNU 1-(2-chloroethyl)-3-(4-methylcycloexyl)-I-nitrosourea (MeCCNU). Cancer 31, 1164-1169.

Yu JS, Wei MX, Chiocca EA, Martuza RL, Tepper RI. (1 9 9 3 )Treatment of glioma by engineered interleukin 4-secretingcells. Cancer Res 53, 3125-3128.

Zerrouqi A, Rixe O, Ghoumari AM, Yarovoi SV, Mouawad R,Khayat D, Soubrane C. (1 9 9 6 ) Liposomal delivery of theherpes simplex virus thymidine kinase gene in glioma,improvement of cell sensitization to ganciclovir. CancerGene Ther 3, 385-392.


149


150

LTR IL-2 IRES TK SV NEO LTR LTR IL-2 IRES TK SV NEO LTR

TIL recruitment

bystander effect

tumor cell

nucleus

producer cells inoculation

mRNA

IL-2TK

GCV GCV-P GCV-3P

integrated viral vectortransduction


151

IRESIRES LTRLTRLTRLTR cytokinecytokine TKTK SVSV NEONEOΨ

regulatory elementsregulatory elements

therapeutic genestherapeutic genes

gene for positive selectiongene for positive selection

glioblastoma multiforme: molecular biology and new perspectives

Documents