TTHHEE NNIICCBB (National Institute for CellularBiotechnology) – located on Dublin CityUniversity campus – has a long track recordof cancer research and is one of the leadinginstitutes in Ireland for translational cancerresearch.While research has been focused mainly onnon-neuronal cancers – for example,breast, lung and more recently pancreasand skin – tumours arising in the centralnervous system, in particular in the brain,have not been studied at the NICB untilrecently.

Malignant brain tumours are one of themost virulent forms of cancer. They canresult in serious physical and cognitiveimpairments. Brain tumours are oftendescribed as cancer of the soul, since the

brain is the centre of thought, personalityand emotion. Brain tumours are cate-gorised into primary and secondary.Primary brain tumours originate in thebrain and rarely spread outside the brain,whereas secondary tumours start outsidethe brain, for example in the lung, breast orskin, and metastasize to the brain.Primary brain tumours

Primary brain tumours develop frombrain cells and give rise to gliomas, menin-giomas, central nervous systemlymphomas, medulloblastomas, andtumours of the nerve sheet.

The majority, however, arise from glialcells, hence their name gliomas andaccount for 42% of all primary braintumours and 71% of all malignant types.

Gliomas include oligodendrogliomas,ependymomas, and astrocytomas/glioblas-tomas, the latter representing the biggestgroup (about 80%). Astrocytomas aregrouped according to their grade of malig-nancy:l Pilocytic astrocytoma (grade I), which is

mainly found in childrenl Astrocytoma grade II, IIIl Glioblastoma (grade IV) – which is most

common in adults between the age of 30and 70.There are no socio-economic boundaries

for the development of a these tumours.Cause still unknown

The cause of primary brain tumours isstill unknown. Exposure to ionising radia-tion and head injuries may be risk factors,

?? cancerprofessional Winter 2007 Winter 2007 cancerprofessional ??

Researchers are looking for a molecular profilethat can predict in vitro responsiveness toEFGR inhibitors in patients with brain tumours

From the bedsideto the lab

Verena Amberger-Murphy

biotechnology biotechnology

?? cancerprofessional Winter 2007 Winter 2007 cancerprofessional ??

as well as some inherited conditions forexample neurofibromatosis, Von Hippel-Lindau syndrome, Li-Fraumeni syndrome,and Turcot’s syndrome. However, onlyabout 5% of all cases can be attributed tothese causes.

In Ireland the incidence of primary braintumours is 7/100,000 people per year(6.4/100.000 per year in the US).

There are about 316 new cases of primarybrain tumours diagnosed every year in theRepublic of Ireland and the mortality isaround 67%. Despite major improvementsin the surgical treatment the prognosis ispoor.

Survival depends on the tumour gradeand age. The five-year survival can be 65%for grade I and II and less than 10% forgrade III and IV (glioblastoma) (Cancer Atlasof the UK and Ireland 1991-2000).

The survival rate has improved very littlein the recent decade. The median survivalof glioblastoma (grade IV) is 189 days andhas not significantly changed within a 12-year period (1993-2004).1

Treatment options for patients withmalignant glioma were limited and usedaccording to the principle ‘one size fits all’;hence all subtypes were treated in thesame way.

Treatment included a surgical attempt toremove visible tumour or at least reducethe tumour size (removal of as much visibletumour as possible) and radiotherapy,whereas chemotherapy consisting of acombination of procarbazine, vincristine,and CCNU (PCV) played a marginal role. Inparticular, glioblastoma (grade IV) was con-sidered as highly chemoresistant.

Due to the infiltrative nature of theselesions, recurrence is inevitable, even fol-lowing an apparently complete surgicalremoval as checked by post-operative MRIor CT scan.Sub-groups show more sensitivity

Many clinical trials using older or newerdrugs show that there are often a smallnumber of responders; whereas the major-ity of patients enrolled would not respond,which leads to the conclusion that the clin-ical trial failed.

In 1988, however, Cairncross and McDon-ald discovered that a subgroup ofmalignant gliomas revealed a better sensi-tivity to the combination chemotherapyPCV. Investigating the tumour tissue theyfound out that these tumours were charac-terised by deletions at chromosomes 1pand 19q2,3 and that patients bearing thesetumours would greatly benefit from thetreatment with PCV.

Another more recent finding came fromStupp and his colleagues. They found that

the methylguanine methyltransferase(MGMT) gene, which gives rise to a DNArepair enzyme, is correlated with theresponsiveness to the newly developeddrug temozolomide (Temodar, Temodal).4,5

Patients, who expressed the methylatedform of the MGMT gene showed betteroverall survival than patients with the non-methylated form. The median survival was22 months (versus 15 months) and the two-year survival rate was 46% (versus 23%).Tyrosine kinase inhibitors

In 2005, Mellinghoff et al published theresults of a clinical trial using the tyrosinekinase inhibitors erlotinib (Tarceva) andgefitinib (Iressa) in the treatment of malig-nant glioma. They reported 20%responders within the patient cohort, whoshowed a reduction in tumour size mea-sured by MRI.6

Responsiveness to these tyrosine kinaseinhibitors was associated with co-expres-sion of EGFRvIII, a mutant variant ofepidermal growth factor receptor (EGFR),and phosphatase and tensin homologue(PTEN), a tumour suppressor, which is oftendeleted in malignant gliomas. However,others have shown that responsiveness toEGFR blockade occurs independently of the

presence of mutated EGFRvIII.7

The results of a small number of respon-ders reflect the cellular and molecularheterogeneity of malignant gliomas. Thereis a need for further in vitro studies onpatient tissue to be able to predict respon-siveness to these types of targetedtherapies.Irish Cancer Society funding

In 2006, we were able to secure fundingfrom the Irish Cancer Society for a three-year translational research project, which isperformed in collaboration with the Neu-ropathology Department in BeaumontHospital.

Together with consultant neuropatholo-gist Michael Farrell we are trying to find amolecular profile, which is predictive for invitro responsiveness to the EGFR inhibitorserlotinib (Tarceva), gefitinib (Iressa), andthe platelet-derived growth factor receptor(PDGFR) inhibitor imatinib (Gleevec).

EGFR and PDGFR and their downstreameffectors, for example Ras, Akt, and mTOR(mammalian target of rapamycin) are criti-cally involved in the regulation of gliomasurvival, proliferation, invasion, and angio-genesis.

EGFR and PDGFR are attractive targets,

biotechnology biotechnology

From our collection of results we hope to be able to find aspecific expression profile, which can be attributed toresponsiveness to tyrosine kinase inhibition and can beused as predictors for individual treatment plans

?? cancerprofessional Winter 2007 Winter 2007 cancerprofessional ??

because they arefrequentlyamplifiedand/or over-expressed inmalignantgliomas.Amplificationof EGFR is oftenassociated withthe expressionof a mutatedvariantEGFRvIII, whichis constitutivelyactive andgrowth factorindependent.The expressionof EGFRvIII islinked toincreasedexpression ofmetallopro-teinases andextra cellularmatrix mole-cules, whichresults inincreased inva-siveness of thetumour.Recent clinicaltrials with

erlotinib and gefitinib have shown thatthere is a large variability in patientresponse. It became obvious that there wasa subset of patients who responded muchbetter to these inhibitors, however, thecause for this variability is not yet fullyunderstood.

As yet there is very little data available onthe in-vitro effects of EGFR and PDGFR/c-kitblockade on proliferation, invasion andexpression of proangiogenic factors inglioma cells.In vitro responsiveness

With our project we go from bed-side(patient) back to the lab by investigatingpatient tissue prior to any treatmentexcept resection. We determine in vitroresponsiveness to tyrosine kinaseinhibitors by measuring their effects onproliferation, invasion and angiogenesissignalling, which represent the three cardi-nal biological characteristics of malignantgliomas. Ethical approval for this study wasgained from Beaumont Hospital EthicsBoard in 2005.

Candidate patients are identified on thebasis of clinical presentation and imaging.With the help of a patient informationleaflet the project is explained to these

patients by a senior house officer of theNeuropathology Department and inter-ested patients sign a written consent formprior to surgery. Patients, who havereceived tumour treatment before, are notconsidered for this study, except caseswhere tissue from the first interventionwas already included in the study.

During operation, fresh tumour tissue issent to the Neuropathology Departmentfor intra-operative diagnosis. Havingensured that sufficient tissue has been har-vested to allow for a definitive diagnosis,residual tumour will be split with 50% ofthe tissue snap frozen in liquid nitrogenand the other half placed into cold saltsolution and transferred to the NICB labo-ratory.Beaumont Hospital and NICB

This project is a two arm study. At Beau-mont Hospital, Rachel Howley, a PhDstudent from the Royal College of Sur-geons, performs immunohistochemicaland molecular analyses on formalin fixedand on fresh frozen tissue, respectively.Antibody staining provides informationabout the expression status of EGFR,EGFRvIII, PDGFRα and β, PTEN, and anumber of down-stream molecules withinthe signalling pathways.

Molecular analysis of genomic tissueDNA includes DNA sequencing, FISH (fluo-rescence in situ hybridisation), and realtime PCR (polymerase chain reaction) pro-ducing information about gene expressionlevels of EGFR, EGFRvIII, PDGFR, and PTEN.

At the NICB, Paula Kinsella, a PhD studentat the NICB, and I develop primary cell linesfrom the tissue samples. These cell lines aretested for their proliferative and invasiveactivity as well as for their sensitivity to thetyrosine kinase inhibitors erlotinib, gefi-tinib, and imatinib and to temozolomide(the chemotherapy drug currently used inthe treatment of gliomas) and docetacel(Taxotere), a powerful chemotherapy drugused for a variety of other cancer types.

Using Western Blotting technique wedetermine protein expression of EGFR,EGFRvIII, PDGFR, and the same down-stream molecules, which are analysed inthe tissue samples in Beaumont.DCU grant

With an additional grant, secured in early2007 from DCU’s Faculty of Science andHealth, we will analyse expression ofmicroRNAs, which are small regulatory RNAfragments, in two groups of our primarytumour cell lines chosen according to theirproliferative activity.

MicroRNAs have been identified to play acritical role in developmental and physio-logical processes and are implicated in the

pathogenesis of cancer.They represent small non-coding RNA

molecules which can either act as onco-genes or as suppressor genes. In particular,miRNA-21 has been found to be involved ininhibiting apoptosis through PI3-kinaseactivation, which is dependent on the dele-tion of phosphatase and tensin homolog(PTEN) from chromosome 10.

Using low density arrays we will deter-mine differences in microRNA expressionbetween fast and low proliferating celllines and functionally validate identifiedcandidates in each cell line. This analysisadds another dimension to our transla-tional study.

Results from both arms and themicroRNA study are collected in a databaseand are compared to maximise the validityof in vitro glioma response data. The data-base also includes key clinical, radiologic,pathologic and – if available- follow-upinformation of each patient, which can beincluded into the analysis of the data.Specific expression profile

From this collection of results we hope tobe able to find a specific expression profile,which can be attributed to responsivenessto tyrosine kinase inhibition and can beused as predictors for individual treatmentplans.

A secondary but important long-termgoal is the development of reverse phasemicroarrays, which represents a rapid, highthroughput technique for the characterisa-tion of signalling pathways in humangliomas, which would allow a more accu-rate diagnosis and a personalisedtreatment plan.Verena Amberger-Murphy is seniorresearch scientist at the National Institutefor Cellular Biotechnology (NICB), DublinCity University

References1. Tait MJ, Petrik V, Loosemore A, Bell BA, PapadopoulosMC. Survival of patients with glioblastoma multiformehas not improved between 1993 and 2004: analysis of625 cases. Br J Neurosurg 2007: 1-5.2. Cairncross JG et al. Specific genetic predictors ofchemotherapeutic response and survival in patientswith anaplastic oligodendrogliomas. J Natl Cancer Inst1998; 90: 1473-1479.3. Cairncross JG, Macdonald DR. Successful chemother-apy for recurrent malignant oligodendroglioma. AnnNeurol 1988; 23: 360-364.4. Stupp R et al. Radiotherapy plus concomitant andadjuvant temozolomide for glioblastoma. N Engl J Med2005; 352: 987-996.5. Hegi ME et al. MGMT gene silencing and benefit fromtemozolomide in glioblastoma. N Engl J Med, 352: 997-1003, 2005.6. Mellinghoff IK et al. Molecular determinants of theresponse of glioblastomas to EGFR kinase inhibitors. NEngl J Med 2005; 353: 2012-2024.7. Haas-Kogan DA et al. Epidermal growth factor recep-tor, protein kinase B/Akt, and glioma response toerlotinib. J Natl Cancer Inst 2005; 97: 880-887.

biotechnology biotechnology