Animal models for glioma drug discovery

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>1. Introduction</p><p>2. Historical account of</p><p>high-grade glioma therapy</p><p>3. Histological and molecular</p><p>characteristics of</p><p>high-grade glioma</p><p>4. Mouse models of gliomas</p><p>5. Preclinical studies using</p><p>GEMMs of glioma</p><p>6. Conclusions</p><p>7. Expert opinion</p><p>Review</p><p>Animal models for glioma drugdiscoveryTerreia S Jones* &amp; Eric C Holland*University of Tennessee Health Science Center, Department of Clinical Pharmacy, Memphis, TN,</p><p>USA and Memorial Sloan-Kettering Cancer Center, Department of Neurosurgery, New York,</p><p>NY, USA</p><p>Introduction: High-grade gliomas are among the most deadly of all cancer</p><p>types and are also the most common malignant primary tumors of the CNS.</p><p>Large-scale studies that have analyzed the transcriptional and translational</p><p>expression patterns of glioma have found that the majority of these tumors</p><p>can be categorized based on specific genomic anomalies. Genetically engi-</p><p>neered mouse models (GEMMs) that represent the molecular subgroups of</p><p>the human disease harbor a variety of molecular alterations that have been</p><p>proven to drive gliomagenesis. These models provide an opportunity to assess</p><p>the effects of novel therapies in the presence of specific molecular defects.</p><p>Research using GEMMs, which are associated with these subclasses, allow</p><p>researchers to assess drug efficacy by subclass.</p><p>Areas covered: In this review, the authors discuss the histological and molec-</p><p>ular characteristics of malignant gliomas, the therapies used to treat them</p><p>and the animal models that closely recapitulate them.</p><p>Expert opinion: It is likely that GEMMs that recapitulate the molecular charac-</p><p>ter of human tumors will provide a more accurate prediction of individuals</p><p>who may be more or less likely to benefit from specific therapies. This know-</p><p>ledge can be then used to drive clinical trial design and this, in turn, could</p><p>lead to better therapeutic outcomes.</p><p>Keywords: gliomas, molecular classification, mouse models, preclinical trials</p><p>Expert Opin. Drug Discov. (2011) 6(12):1271-1283</p><p>1. Introduction</p><p>It is estimated that ~ 24,000 individuals will be diagnosed with a primary CNStumor this year in the US [1]. Diffuse gliomas (glial tumors that infiltrate into nor-mal brain tissue) are among the most difficult to treat neoplasms and mortality ratesare directly proportional to tumor grade [2]. Diffuse gliomas account for the vastmajority of primary brain tumors with grade IV glioblastoma (GBM) being themost frequent and the most malignant and deadly overall. High-grade gliomas arealmost always rapidly fatal regardless of the therapeutic management used. Mostpatients diagnosed with a GBM will die within the first 2 years from diagnosisdespite aggressive therapy. Indeed, even the low-grade tumors can undergo malig-nant transformation into a GBM over time, usually within 5 -- 10 years fromdiagnosis of the primary tumor [3].</p><p>Despite &gt; 40 years of research, only modest improvements in survival for high-grade glioma have been made. The challenge in providing effective therapies forthe treatment of these tumors is evidenced in a 5-year survival rate of &lt; 10% forGBM [4]. Today, the standard of care for treating the primary disease includestemozolomide and radiation and now bevacizumab is used commonly at tumorrecurrence. A number of primary papers and reviews have been published discussingthe role of signal transduction pathways in gliomagenesis and mouse models havevalidated the importance of some of these pathways [5]. There are several signal</p><p>10.1517/17460441.2011.632628 2011 Informa UK, Ltd. ISSN 1746-0441 1271All rights reserved: reproduction in whole or in part not permitted</p><p>Exp</p><p>ert O</p><p>pin.</p><p> Dru</p><p>g D</p><p>isco</p><p>v. D</p><p>ownl</p><p>oade</p><p>d fr</p><p>om in</p><p>form</p><p>ahea</p><p>lthca</p><p>re.c</p><p>om b</p><p>y U</p><p>nive</p><p>rsity</p><p> of </p><p>Uls</p><p>ter </p><p>at J</p><p>orda</p><p>nsto</p><p>wn </p><p>on 1</p><p>1/13</p><p>/14</p><p>For </p><p>pers</p><p>onal</p><p> use</p><p> onl</p><p>y.</p></li><li><p>transduction inhibitors that have been developed and manyare currently being tested in clinical trials for glioma (referto Arko et al. for a review) [6,7]. Unfortunately, to date manyof these agents do not appear to significantly impact thecourse of the disease when compared to standard of care ther-apy. However, there are a few studies that have shown a mod-est benefit in primary disease when a signal transductioninhibitor is combined with temozolomide and radiation [8-10]and a number of these agents are being tried in combinationwith other targeted therapies at tumor recurrence [11-13]. Thelack of success in developing effective novel therapies can beattributed in part to the intra- and inter-tumoral heterogene-ity. Molecular studies have revealed the complexity of thesetumors and may provide a more detailed means of character-izing them than the traditional histological classification sys-tem used by clinicians over the past several decades. In thisreview, we discuss the histologic and molecular characteristicsof malignant gliomas, the therapies used to treat them and themouse models that can be useful in testing novel therapies inpreclinical studies.</p><p>2. Historical account of high-grade gliomatherapy</p><p>Since the mid-20th century, there have been few significantadvances in the therapeutic management of glioma. Figure 1</p><p>displays the limited number of significant advances inGBM therapy and is reflective of the modest improvementsin survival over time. In the early 1970s, the median sur-vival for patients diagnosed with a high-grade glioma wasonly 6 months [14]. Surgical removal of as much of thetumor mass as possible is directly correlated with sur-vival [15,16]; hence, surgery has remained the primary treat-ment modality for glioma. Historically, post-surgicalglioma management that provided the best opportunity atextending survival in these patients included lipophilicdrugs and radiation therapy. In the 1980s, results fromclinical studies solidified the role of radiation and chemo-therapy in improving survival in these patients and thistherapy subsequently became the standard of care [17-22].One of the biggest challenges to identifying effective drugtherapies for glial tumors was designing drugs that werecapable of penetrating the BBB to achieve adequate druglevels at the tumor site. The first drugs proven to be effec-tive against glioma were lipophilic drugs belonging to thenitrosourea drug class. These drugs function by generatinglethal DNA damage in tumor cells leading to cell death.However, the damage exerted by these drugs was not onlyrestricted to tumor cells, hence a significant amount ofadverse effects are associated with them. Drugs in this classinclude carmustine (BCNU) and lomustine (CCNU).BCNU was approved by the FDA in 1977 and becamethe most commonly used agent in this class. In a clinicalstudy that compared BCNU alone, radiation therapy aloneand radiation plus BCNU, it was found that by combiningBCNU with radiation the median survival can be extendedby as much as 12 months [19,23]. Later, in an effort todecrease the toxic side effects associated with systemicadministration of BCNU and increase drug concentrationsat the tumor site, a biodegradable wafer impregnated withBCNU was developed and approved by the FDA (1996)for implantation at post-surgical resection of recurrent dis-ease. However, this local delivery approach would onlyextend survival by ~ 2 months in GBM patients [24]. Stud-ies are being conducted that combine BCNU implants withstandard therapy [25].</p><p>After almost 10 years of practically no advances in survivalor therapeutic management for glioma, temozolomide (aderivative of the alkylating agent dacarbazine) was approvedby the FDA for refractory high-grade disease in 1999 andwas later approved for primary disease in 2005. Today, thepost-surgical standard of care for treating the primary diseaseconsists of concurrent temozolomide and radiation followedby adjuvant temozolomide. This combination has providedthe most significant improvement in survival since the1980s (median survival of 14.6 months) [4,26]. Morerecently, bevacizumab is being used more often at tumorrecurrence [27-30] and clinical studies are being performed toinvestigate bevacizumabs utility in treating primary dis-ease [31]. However, it remains to be determined whether thereis a survival benefit with bevacizumab therapy.</p><p>Article highlights.</p><p>. The lack of success in developing more effectivetherapies against glioma can be attributed to the intra-and inter-tumoral heterogeneity and expression studieshave confirmed the molecular complexity ofthese tumors.</p><p>. Genetically engineered mouse models (GEMMs) ofglioma have been developed by overexpressingcomponents of signal transduction pathways thatpromote cell proliferation and survival.</p><p>. The Cre-Lox system allows for targeted deletion of agene flanked by loxP sites within a specific cell typewhen Cre recombinase is expressed under the control ofthe tissue-specific promoter.</p><p>. The RCAS/tv-a system allows for the delivery of anoncogene of interest into brain cells by expressing theTVA receptor under the GFAP astrocytic promoter or thenestin neural progenitor promoter.</p><p>. The GEMMs that are associated with the molecularcharacter of human glioma can be used to screen novelcompounds and therapeutic approaches for efficacy andtoxicity prior to being used in the clinical setting.</p><p>. The use of imaging modalities such as MRI andbioluminescence imaging provides the opportunity toconfirm tumor presence and monitor the tumorsresponse to therapy in vivo over time without sacrificingthe animal.</p><p>This box summarizes key points contained in the article.</p><p>Animal models for glioma drug discovery</p><p>1272 Expert Opin. Drug Discov. (2011) 6(12)</p><p>Exp</p><p>ert O</p><p>pin.</p><p> Dru</p><p>g D</p><p>isco</p><p>v. D</p><p>ownl</p><p>oade</p><p>d fr</p><p>om in</p><p>form</p><p>ahea</p><p>lthca</p><p>re.c</p><p>om b</p><p>y U</p><p>nive</p><p>rsity</p><p> of </p><p>Uls</p><p>ter </p><p>at J</p><p>orda</p><p>nsto</p><p>wn </p><p>on 1</p><p>1/13</p><p>/14</p><p>For </p><p>pers</p><p>onal</p><p> use</p><p> onl</p><p>y.</p></li><li><p>3. Histological and molecular characteristicsof high-grade glioma</p><p>In 1979, the WHO formalized a system to classify CNStumors using criteria based primarily on tumor histopathol-ogy. In this system, tumors are grouped based on the resem-blance of tumor cells to normal glial cells and the relativedegree of malignancy. In 2007, the classification system wasrevised to reflect changes in tumor grade, new entities and var-iants of some CNS tumors [32,33]. This system groups braintumors by grade (WHO grades I -- IV) (Table 1), with lowergrades reflecting the more benign and treatable tumors andhigher grades reflecting the more aggressive and difficult totreat tumors. Grade I gliomas are slow growing well-circumscribed tumors that are curable by complete surgicalresection. Grade II tumors are relatively slow growing tumorsexhibiting cellular differentiation but can diffusely infiltrateinto normal brain tissue and have a modest proliferativecharacter. Grade III tumors exhibit regions of anaplasia,</p><p>diffuse infiltration and are highly proliferative. And grade IVgliomas have additional characteristics including microvascu-lar proliferation and pseudopallisading necrosis [34]. Thesetumors are further divided into subcategories based on glialcell morphology (i.e., astrocytic, oligodendroglial, mixed oli-goastrocytic and ependymal). The most common high-grade (or malignant) gliomas include anaplastic astrocytomas(grade III), anaplastic oligodendrogliomas (grade III), ana-plastic oligoastrocytoma (grade III), anaplastic ependymoma(grade III) and GBM (grade IV).</p><p>From a global standpoint, elevations in platelet derivedgrowth factor (PDGF) and epidermal growth factor receptor(EGFR) signaling collectively are the most commonly observedaberrations across all GBMs. Approximately 30% of GBMsexhibit increased PDGF signaling and ~ 45% exhibit increasedsignaling of EGFR and these aberrations can occur together insome GBMs [35]. Elevations in PDGF signaling can occurthrough PDGFRa or PDGFb amplification or activatingmutations. PDGF binding to its receptor (PDGFR) can lead</p><p>6</p><p>4</p><p>2</p><p>0</p><p>1980</p><p>1977:FDAapproves</p><p>BCNU</p><p>1980: Walker et al.,shows ~12 month</p><p>median survival withradiation plus BCNU</p><p>Med</p><p>ian</p><p> su</p><p>rviv</p><p>al fo</p><p>r G</p><p>BM</p><p> (ye</p><p>ars)</p><p>1990 2000</p><p>Mouse modeling of glioma</p><p>Year2010</p><p>2009: Stupp et al.,shows 14.6 month</p><p>median survival withradiation plus TMZ</p><p>1999: FDA approvesTMZ for refractoryhigh-grade glioma</p><p>2005: FDA approves TMZ forprimary glioma</p><p>Molecular sub-grouping of GBMs</p><p>1996: FDA approvesGliadel wafers for</p><p>recurrent GBM</p><p>2009: FDA approvesbevacizumab</p><p>Figure 1. Historical account of survival and standard of care therapy for glioblastoma (GBM). The first drug approved for</p><p>treating GBM was carmustine (BCNU) in 1977. Combining BCNU with radiation therapy increased the median survival from</p><p>6 to ~ 12 months [19,23]. BCNU was approved in 1996 in the form of a biodegradable wafer for local delivery to the tumor bed.It was not until 2005 that a new systemic drug was approved for treating GBM (temozolomide, TMZ) and, in 2009,</p><p>bevacizumab was approved for recurrent disease. The median survival for primary disease remains suboptimal at</p><p>14.6 months [4]. In 1998, Uhrbom et al. developed the first molecularly and histologically relevant platelet derived growth</p><p>factor (PDGF)-driven genetically engineered mouse model (GEMM) of glioma [80]. Later, several other human disease relevant</p><p>GEMMs were developed and results from expression data from The Cancer Genome Atlas (TCGA) and others suggest that</p><p>these models may mimic the molecular character of the human disease [45-47,77,104].</p><p>Jones &amp; Holland</p><p>Expert Opin. Drug Discov. (2011) 6(12) 1273</p><p>Exp</p><p>ert O</p><p>pin.</p><p> Dru</p><p>g D</p><p>isco</p><p>v. D</p><p>ownl</p><p>oade</p><p>d fr</p><p>om in</p><p>form</p><p>ahea</p><p>lthca</p><p>re.c</p><p>om b</p><p>y U</p><p>nive</p><p>rsity</p><p> of </p><p>Uls</p><p>ter </p><p>at J</p><p>orda</p><p>nsto</p><p>wn </p><p>on 1</p><p>1/13</p><p>/14</p><p>For </p><p>pers</p><p>onal</p><p> use</p><p> onl</p><p>y.</p></li><li><p>to cell proliferation and survival by stimulating downstream sig-nal transduction pathways such as P13K/AKT and rat sarcomaoncogene (RAS)/MAPK (Figure 2) [36,37]. These pathways canbe activated by PIK3CA and by phophastase and tensin homo-log (PTEN) or neurofibromin 1 (NF1) loss. PTEN andNF1 function by regulating AKT and RAS activation, respec-tively, preventing inappropriate cell signaling for cell growthand proliferation. EGFR overexpression and hyperactivity canbe a consequence of gene amplification or caused by mutantvariants such as EGFRvIII, a constitutively active variant ofEGFR [38-40]. EGFR is a member of the erythroblastic leukemiaviral oncogene receptor tyrosine kinase family and, similar toPDGF and its receptor, ligand binding stimulates down-stream signaling pathways involved in tumor cell growth andsurvival (Figure 2).Loss of function of the tumor suppressors P53 and retino-</p><p>blastoma (RB) is also a common occurrence in GBM; how-ever, these lesions are more often associated with tumorsthat are believed to have progressed from a lower grade [5].P53 loss of function can be precipitated by either directgene mutation or deletion or indirectly byMDM2 copy num-ber alteration [35,41,42]. Similarly, RB function can be compro-mised through increased CDK4 copy number causing RBsuppression. And both P53 and RB are vulnerable to thecyclin-dependent kinase inhibitor 2A (CDKN2A) gene dele-tion as this gene encodes for the proteins INK4A and ARFwhich are res...</p></li></ul>


View more >