cancer world 41

� Ruth Ladenstein: raising standards of care for our young patients � Can tumoursignatures help guide treatment decisions? � A ‘simple’ doctor’s quest to find bettertherapies � Focus on advanced breast cancer: new conference tackles a neglected topic

Ruth Ladenstein

Education & knowledge through people & facts

Number 41, March-April 2011

CancerWorld

41M

ARCH-APRIL2011

CANCER WORLD � MARCH/APRIL 2011 � 1

Contents

3 EditorialA million voices against cancer

4 Cover StoryRuth Ladenstein: raising standards of care for our young patients

13 e-Grand RoundTreatment of triple negative breast cancer

22 Cutting EdgeReading the signs: the role of genomic signatures in guiding treatment decisions

30 Best Cancer ReporterThink before you ruin your patient’s chance of parenthood – prize for journalistwho highlighted poor practice in Poland

34 MasterpieceA simple doctor’s quest to improve on today’s treatments: how Stan Kayecame to lead drug development at the Royal Marsden

42 Spotlight on...Where are the consensus guidelines for women with metastatic disease? – Newconference will tackle this neglected topic head on

48 Impact FactorFailure of bevacizumab in early-stage colon cancerUltra-targeted accelerated partial breast irradiationusing TARGIT – a cautionary note

Newsround

62 Systems & ServicesAn advanced oncology degree for busy specialists across the globe

EditorKathy [email protected]

Assistant EditorAnna Wagstaff

Editorial AssistantCorinne Hall

Editorial AdvisorsJacques BernierFatima CardosoFranco CavalliAlberto CostaVincent T. DeVita

Contributing WritersMarc Beishon, Simon CromptonSusan Mayor, Peter McIntyreDaniel Sargent, Rajiv SarinAnna Wagstaff

Publishing AdvisorsGillian Griffith, Fedele Gubitosi

Website LiaisonCorinne Hall

Art EditorJason Harris

ProductionHarrisDPIwww.harrisdpi.co.uk

Printed byGrafiche Porpora

Cover photographRené Van Bakel

Published byEuropean School of Oncology

Direttore responsabileAlberto Costa

Registrazione Tribunale di RomaDecreto n. 436 del 8.11.2004

All enquiries about Cancer Worldshould be made to:ESO Editorial OfficeVia del Bollo 420123 Milan, Italye-mail: [email protected]: +39 02 8546 4522Fax: +39 02 8546 4545All correspondence should be sentto the Editor at [email protected]

Copyright ©2011 European School of Oncology.All rights reserved

Cancer World is published six times per year by the European School of Oncology.It is distributed at major conferences, mailed to subscribers and to Europeanopinion leaders, and is available online at www.cancerworld.org


Editorial

ThisSeptember, for only the sec-ond time in its history, theUnitedNationsGeneralAssem-

bly will hold a Special Session focused onhealth.World leaders will gather to discussthe global challenge posed by non-com-municable diseases (NCDs) and agree onconcise, action-oriented solutions. Forthose of us concerned about the rapidlygrowing toll of death and suffering fromcancer worldwide, it represents an oppor-tunity that we cannot afford to miss.

Chronic diseases, including cancer, killmore than twice asmany people as all infec-tious diseases, maternal and childhoodconditions andnutritional deficiencies com-bined. They are on the rise, with the largestincreases happening in low- and middle-income countries, which are ill equipped tocope with the human and economic toll.

Yet less than 1% of global funding forhealth goes to support low- and middle-income countries tackleNCDs.One reasonis that funds from the biggest global healthdonors are often linked to the MillenniumDevelopmentGoals, which don’t evenmen-tion chronic diseases. Securing a SpecialUNSession on this topic represents a huge vic-tory for the alliance of international NCDcivil society groups, providing a crucial plat-form to raise awareness about the scope ofthe problem and forceworld leaders to focuson identifyingworkable solutions.Civil soci-ety groupswill be consulted before the Sum-mit and they will be represented at theSummit. This means that the cancer com-

� Kathy Redmond � EDITOR

munity will have a voice and can help shapewhat comes out of it. Individuals, institu-tions or groups that want to help putNCDson the global health agenda can visitncdalliance.org to see how to get involved.

TheUnion for InternationalCancerCon-trol (UICC) – the lead cancer organisationwithin the NCD Alliance – aims to gatherone million World Cancer Declaration sig-natures to present to world leaders at theSummit. Cancer organisations across theworld are calling on their members to addtheir names to send a forceful message togovernments and international health policymakers about the cancer community’s com-mitment to take action to stem the tide ofcancer. The Maximise Life Global CancerCampaign initiative, run by the Max Foun-dation global patient advocacy group, hasalready secured 13,000 signatures frompeo-ple in almost 90 countries, showingwhat canbe achieved. InEurope, ESO is joiningwithother organisations, including ECCO andtheEuropeanSociety forMedicalOncology,to call on everyonewho cares about cancer tosign the Declaration.

The World Cancer Declaration detailskey, affordable actions any country can takethat could significantly reduce death andsuffering from cancer in a relatively shortperiod of time.Wehave aunique opportunityto help make sure the global health com-munity gets behind this Declaration andputs cancer at the top of the agenda, whereit belongs. So get online at www.world-cancerday.org andmake your voice heard!

A million voicesagainst cancer

CoverStory

4 � CANCER WORLD � MARCH/APRIL 2011

Ruth Ladenstein:raising standards ofcare for our young patients

� Marc Beishon

Paediatric oncologists have awell-deserved reputation for collaborating and treating patientswithin

trial protocols, but there are limits to what they can achieve alone. Leading practitioner Ruth

Ladenstein is now calling on the EU and member states to commit to improving paediatric

cancer care by providing specialist facilities and serious backing for research and data collection.

Ofall the cancer specialities, some ofthe most spectacular gains in out-comes have undoubtedly been inpaediatric oncology. Fewcould arguethat an increase to about an 80%

cure rate from less than 20% across the range ofchildhoodcancers, albeit over several decades, is nota cause for celebration. Although these cancersare rare, there could be several hundred thousandpeople inEurope alive todaywho survived a cancerdiagnosis when they were young.

However, as Ruth Ladenstein, president ofEurope’s Paediatric Oncology Society (SIOPE),points out, this success only highlights the need tomaintain and improve the rigorous research envi-ronment that led to the gains, as there can be norelaxation of standards, while there are major chal-lenges ahead.Around15,000 youngpeople, aged18andunder, arediagnosedwithcancer inEuropeeachyear, and at present cure rates more than 3000 will

die,making cancer thebiggest cause of death in thisage group for those above infancy. “We know thatwhen children are treated outside clinical trial set-tings their outcomes are not nearly as good; survivalis on average about 20% worse – a dramatic drop,”she says. “Andwith a new drug in a trial setting weare looking for a 5% to 10% improvement.”

This “life-saving factor” should be borne inmind, she says, when considering the impact onpatients of the obstaclesEuropeanUnion regulationhas placed on conducting trials, and indeed the lackof multidisciplinary paediatric oncology centresable to participate in this sort of research in manycountries, particularly in eastern Europe.

“Integration of research and care is a hallmarkof paediatric oncology,” says Ladenstein, with about80% of children now treated either in clinical trialsor with prospectively monitored therapeutic pro-tocols. But as childhood cancer treatments arefirmly in the ‘orphan’ (more rare) disease category,

CoverStory


thus attracting far less industry funding than themuch larger adult cancer field,most of the researchis reliant on investigators working in an academicsetting, in often complex protocols,mostlywith ‘off-label’drugs – i.e. drugs that have never been trialledand approved for use in children. The EuropeanClinical Trials Directive has had a dramatic effecton this already fragile research base – an impacteven greater than in the adult tumour area.

“We estimate that the number of new trials hasgone downby 70% since the implementation of thedirective as there is just somuchmore funding andtime now needed to deal with issues such as ethi-cal committees around Europe and insurance intrials deemed to be high risk. Meanwhile we havevirtually no funding or interest in running studies onexisting off-label drugs, and these are classed insome countries as investigational medical prod-ucts, which further adds to the administrative bur-den if we need to use them in trials.And of course

because paediatric cancers are uncommon we doneed multicentre, multicountry studies to accruesufficient patient numbers.”

As she adds, just as in adult cancer, the era ofchemotherapy has largely run out of steam at thepaediatric level, and the pursuit of translationalresearch and new biological therapies is especiallydemanding for academic investigators short offunds. Then there is the paradox that this branch ofoncology also contains – or should do – the mostneglected group in cancer, namely teenagers andyoung adults.And all childrenwith cancer need themajor commitment of follow up through much oftheir lives to monitor the effects of treatment.

There are also urgent needs for gatheringmuchbetter epidemiological data from various coun-tries, for fostering multidisciplinary standards ofcare and for getting patients and families moreinvolved in pressing for research. It’s a huge agendaby any measure.

RENEVANBAKEL

Ladenstein speaks from long-standing involvementin the paediatric oncology research community,specialising in neuroblastoma, andhaving spent hercareer working up to head the solid tumour unit atSt Anna Children’s Hospital in Vienna, Austria,and also holding an associate professorship at theUniversity of Vienna. The St Anna Kinderkrebs-forschung (children’s cancer research institute) haslong been a research hub in the German-speakingregion and internationally, but it has really been puton thewidermapwith two recentEU initiatives thatLadenstein hopeswill help to unravel the ‘red tape’that she believes could seriously hold back progressin her speciality.

The first, now ended, was ‘Overcoming cancerwith research’, a two-year communications projectthat aimed to raise public awareness of childhoodcancer research. This media project, for whichStAnnawas the coordinating organisation,workingwith theGermanChildhoodCancer Foundation asa partner, has produced a comprehensive website

(www.overcomingcancerwithresearch.eu), a film(Little Heroes – Great Opportunities), press con-ferences and other activities.

It also provides details of various paediatricresearch networks and other projects that haveEUfunding. One of these is ENCCA (EuropeanNet-work forCancer Research inChildren andAdoles-cents), amajor €12million initiativeunder theEU’s7th Framework Programme, for whichLadensteinis the coordinator. “It is a four-year project thatstarted this year and our aim is nothing short ofbuilding a sustainableEurope-wide virtual institutethat will unite the paediatric oncology commu-nity,” she says.

Meanwhile, she adds, disparities in care stan-dards are being addressed by SIOPE, which hasdrawnup ‘EuropeanStandards ofCare forChildrenwith Cancer’ for paediatric oncology, and a ‘seven-point plan’ for delivering the overall agenda (see p9),including a call for all member states to havenational cancer plans that contain specific

CoverStory


respondmuchbetter to chemotherapy and childrenare wonderful patients to be with. They under-stand a lot when you explain properly and it makesthemmature in a very short time. It’s a pleasure tobewith themand their families at a critical time, andnow I also see them as grown-ups with their ownchildren.” But of course there is great sadnesswhentreatment fails in some. “I especially feel forteenagers – you should never die when you havehope for the life ahead of you.”

As she adds, it is the right place to be for thosewhowant to be rewarded in terms of outcomes andscientifically. “We are on the edge of a fast-movingfield and there is so much research to be done.”

Needingmore research experience herself, andthe recipient of an Austrian award, Ladensteincast around for a project abroad, landing in Lyon,France, at the Léon Bérard Centre. France hasbeen a European cradle of paediatric oncology,and shewas quickly immersed in analysis of data onneuroblastoma transplants, and also on Ewingtumours and lymphoma patients around Europe.She also studied mechanisms in neuroblastomacells in the laboratory, and went on to work andstudy further in Paris.

Neuroblastoma is themost common childhoodsolid tumour outside of the brain, and themost fre-quent of all under the age of five – in fact it is thesecond most common cause of death in childrenafter domestic accidents.As a neuroendocrine dis-ease it often develops from the adrenal glands.Ladenstein explains that it also has awide spectrumof risk, and stemcell transplants are given after high-dose chemotherapy treatment for overcomingtumour cell resistance in themore severe cases. Butlow-risk disease often regresses to a benign statewithout any treatment, and identifying howbest toapply high-dose regimens became a particular goalfor her following her return toAustria.

“It is one of the most fascinating of cancersbecause it is completely drivenby tumourbiology, aswe have been discovering,” says Ladenstein. “We

standards for age-appropriate treatment andcare for children and adolescents with cancer.

It’s a familiar story inEuropean oncology:manyinterest groups have realised recently that there ismuch to be gained by combining the efforts ofnational societies and institutions to gain scale forresearch and to lobby and network more effec-tively at bothEuropean and state levels.As Laden-stein adds, “While there has been effectivenetworking in childhood cancer – indeed more sohistorically than inmost adult tumours – efforts havelargely been focused on specific diseases such asneuroblastoma.”

Now it’s vital to unite researchnetworks and lob-byingwork, she believes, andLadenstein finds her-self at the head not only of one of oncology’s mostimportant European societies, but also a key proj-ect in ENCCA, given the impact it could have onissues such as the Clinical Trials Directive.

Her path intomedicinewas almost preordained– “It was what I wanted as a little girl” – and shefound herself drawn to paediatrics despite a con-scious choice to resist it, as she felt women toooften find themselves earmarked for the speciality.“But I loved it,” she says, “and I did all my standardpaediatric training at StAnna. In fact despitemov-ing into oncology I’m still a practising general pae-diatrician, as when I’m on call in the hospital I seeall children, not just those with cancer.”

Itwas early inher careerwhenher chief,HelmutGadner – one of the pioneers of the BFM (Berlin-Frankfurt-Münster) leukaemia protocols devel-oped in Germany, who recently retired fromStAnna – pointed her in the direction of oncology.“We had just started to treat children with cancerthen, and he gave me a paper to study on sarcomapatients, and from thatwe started the firstAustriansarcoma study.”

Ladenstein says that far from childhood cancerbeing a daunting area, “it’s exciting because wehave a 40% better chance of curing them than wedo with adults. Some types of paediatric tumours

CoverStory


“Children are wonderful patients to be with –

they understand a lot when you explain properly”

have learnt that neuroblastoma in infants can evenregress from themetastatic stage anddoes not needchemotherapy, unless there are specific risk factorsthat do require intensive treatment. So farwe knowhalf of patients need little or no chemotherapy andcan be spared high-dose treatment, and we haveimproved outcomes in the high-risk group from20%–35% to more than 50%, which is quite anachievement. It is knowing early on who has anunfavourable profile that improves outcomes.”

Ladenstein has been at the centre of both localandEurope-wide neuroblastoma research that hasfound prognostic markers for risk and developednew treatments and protocols, and she is the co-ordinator of the SIOP European NeuroblastomaResearchNetwork (SIOPEN-R-NET) and chair ofSIOPE’s neuroblastoma group (the research net-work was funded by the EU’s Fifth FrameworkProgramme, but continues to operate today). “Thehistory of paediatric oncology is thatwe runonepro-tocol after another and make slight progress byoptimising treatment plans overmany years, butwehave learnt so much more now about prognosticmarkers, stratified treatment and biology in mostchild tumour types.

“Inneuroblastoma, a key focusnow is still on thehigh-risk groupandwehave ahuge trial running thathas accrued more than 1500 children across 20countries andweare getting exciting results from therandomisation,whichwewill be taking to theASCOconference in theUS this year.Wewill show that aEuropeanprotocolwehavedeveloped is performingbetter than the bestAmerican standard.”

She says the Children’s Oncology Group in theUShaddemonstrated a significant improvement forneuroblastoma immunotherapyusing amonoclonalmouse-human chimeric antibody (ch.14.18). TheSIOPENgroup thenundertook to provide access tothis antibody for neuroblastomapatients inEuropevia the trial, but this work illustrates well the diffi-culties that paediatric oncologists face in pursuingnew treatments.

It involves the production and distribution for clin-ical testing of this ‘chimeric’ (combination) anti-body for use in the high-risk trial – but so far this isan entirely academically driven effort, with all thatmeans for pressure on funds to bring a new drug tomarket. “It is very unusual for us to attempt drugdevelopment without industry support – but wehave obtained about €2 million through our ownfundraising efforts. Even so, we only have a limitedamount of the drug for the controlled trials, andweare hoping to find an industrial partner and alsogreater government support for drug production,especially in theUK, sowecanopenupmore trials.”

She also mentions another drug that couldimprove outcomes when given in combination, bypromoting white blood cell production, which isbeingused in trials in theUS, but is simply not avail-able in Europe. “We eventually tracked down apotential ‘importer’ supplier in Switzerland, but thepharmaceutical licence holderwasn’t interested inmaking it available,” says Ladenstein. “It’s very hardwhen parents read about our trials and ask whywecan’t give these drugs to all children. I have toexplain we are not a drug company, that the drugsaren’t mature enough yet to be on the market andthere canbe concerns about toxicity, and simply thatwe do not have enough of them, such as thechimeric antibody, and we are not allowed to offerthe drug outside a controlled trial setting.”

The antibody in question was first researchedsome20 years ago for adult and childhood cancers,but as Ladenstein points out, children in Europehave been “extremely poorly served” in access toinnovative drugs that have been investigated anddeveloped for adults. She is encouraged, however,by a recent initiative that could help children gainbetter access to newdrugs, namely the requirementfor pharmaceutical companies to develop paediatricinvestigation plans (PIPs) for new adult drugs,where appropriate, under the recent EU Paedi-atric Medicine regulation, which also aims to pro-mote safe and effective treatments in general.

“It’s very hard when parents read about our trials

and ask why we can’t give these drugs to all children”

CoverStory


require ‘expedited’ reporting to theEuropeanMed-icines Agency, EMA, and Ladenstein fears thatmuch of these data, which could be valuable forknowledge about say toxicities inmultiagent trials,are disappearing into a ‘black hole’.

There are funds available from the EU’sFramework Programmes for drug developmentthat could help investigate the pharmacoki-netic/dynamic behaviour of off-label drugs inchildren and so move towards approval, but asLadenstein points out, the only way for academ-ics to access these funds is to compete against oneanother, whichmeansmanywill simply waste a lotof effort writing applications.

“We need dedicated funding to investigate theolder drugswe use, and there is a feeling that someshould be entered into randomised trials, whichwould be very costly.Howeverwedoneed to chargeexperts to do thework on correlating drugs properlyin terms of their behaviourwith the course of a dis-ease and the dose, as clearly children are different

The benefits, however, won’t be felt for a longtime, she says. “We may see an impact from thecrossover from adult drugs in 10 to 20 years time.There are only a few that are ongoing at present, butit is a move in the right direction.” The regulationalso fails to resolve the major problem of gettingapproval for drugs that are already widely usedoff-label. “There is not yet any investigationalprocess or funds for us to do this.”

About 80%of drugs used in paediatric oncologyare used off-label, and even those that are approvedare not often labelled appropriately for certain agegroups in terms of dose calculation, for example.“Weneed to take steps to ensure all thedrugsweusefor children are safe and effective – but despite abackwash of 30 years of clinical trials we still havethis huge burden of off-label drug use and barriersto moving forward, such as the continued classifi-cation ofmany of our drugs as investigationalmed-ical products in some countries, despite their longuse.” If drugs are treated as investigational, they

CoverStory


“We could certainly aim now to get that 80%

off-label figure down to 40% in five years’ time”

HOW WELL DOES YOUR COUNTRY SERVE ITS YOUNG CANCER PATIENTS?

SIOPE has drawn up this seven-point plan as a guide forpolicy makers on how to upgrade paediatric cancerservices.1. Cancer plans. Every country should have a nationalcancer plan that contains specific standards for age-appropriate treatment and care for children and ado-lescents with cancer.2. Registries. Every country should support prospectiveregistration of new cases and outcomes of all casesusing the International Childhood Cancer Classificationscheme, extended to include adolescent cases.3. Access to specialists. Every country should havedefined referral pathways so that each patient is man-aged at an age-appropriate specialist treatment centrethat works within a national or cross-border networkstructure and can have access to innovative therapiesin development when needed.

4. Multiprofessional teams. Every child and adolescentwith cancer should be treated by a multiprofessionalteam which has a sufficient volume of activity tomaintain expertise and which participates in audit andaccreditation schemes.5. Specialist training. Specialist training in paediatrichaemato-oncology should be recognised in every Euro-pean country.6. Family support. The crucial role of parental/familysupport should be recognised as critical to treatment out-come and survival of the young cancer patient.7. Research.Greater EU and national support is neededfor investigator-led clinical and translational research, toreverse the recent decline in participation in clinicaltrials, which, to date, has greatly benefited the devel-opment and delivery of ‘best practice’ of care for youngpeople with cancer.

It seems from existing data that paediatric cancersare not increasing in overall incidence, but ifnational registries improve, Ladenstein says, trendsin certain tumours and leukaemias may becomeapparent at a European level.At present, she adds,only localised events such as Chernobyl and otherpollution in somecountries appear to have given riseto higher thanusual rates of childhood cancers. “Wealso need very long observation times concerningtreatment – for example it took 30 years for us to seethe higher incidence of breast cancer in thosewhohad been diagnosedwithHodgkin’s lymphoma andhad been given certain drugs and radiotherapy.”

The biggest work package, in terms of ‘personmonths’, is on networking among preclinicalresearch groups to create common data sharingand bio-information tools, and Ladenstein notesthat an overall aim of ENCCA is to bringresearchers together across the range of paediatriccancers, to possibly identify shared biologicalpathways, for example.

More funding is likely to come from pressurefrom advocacy groups, she believes. “SIOP [Inter-national] has a committee working with the Inter-nationalConfederation ofChildhoodCancerParentOrganisations, which can be a strong voice for us.I know from speaking to people at theUSNationalInstitutes of Health that funders are driven muchmore by parents than by doctors.”

She hopes that SIOPE will benefit fromincreased membership as a result of ENCCA, aspeople recognise the importance of integratedworking. The European branch of the society hasaround 900 members and could do with moreinterest fromnational organisations, but Ladensteinsays there is difficulty in getting people to committo additional membership fees. “But we have astrongEuropean agenda that ENCCAwill increasefurther – and I hope it will show people what theyare missing.”

SIOPE is a founder member of ECCO andruns a paediatric stream at the conference. It joins

“It took 30 years to see the higher incidence of breast

cancer in those treated for Hodgkin’s lymphoma”

CoverStory


from adults.We could certainly aimnow to get that80% off-label figure down to 40% in five years’time.” If standard chemotherapy protocols can beoptimised, Ladenstein says that adding new drugscould then add another 10% benefit in outcomes.

Lobbying from SIOPE will continue on thisissue, as it is not explicitly part of the work pro-gramme for ENCCA. “In the new project, though,we will be aiming to influence the Clinical TrialsDirective so it ismore feasible for paediatric oncol-ogy, such as bydeveloping a contract framework thatallows academic institutions to become coordinat-ing pan-European trial sponsors, delegating tasks tonational bodies sowe can share the burden.We arealso looking to use a not-for-profit insurance organ-isation, maybe insuring studies through nationalhealth services.” (BothLadenstein and the previousSIOPEpresident, KathyPritchard-Jones, havewrit-ten about the absurdities of the Clinical TrialsDirective, which include the stipulation that crush-ing tablets in trials to enable children to swallowdrugs is not allowed as it is deemed a ‘manufactur-ing’ process –EJC 44:2106–2111.)

The clinical trials work package of ENCCA(just one of 18 ‘work packages’ tackling variouschallenges in paediatric oncology) will also aim tostreamline childhood cancer trials by using standardtemplates and datasets; determining just what aninvestigational medicine should be; and cuttingduplication and fragmentation by promotingmoremultinational trials.

“In otherwork packageswewant to explore howwe can build better registry data for childhood can-cers, andhowwecan improve long-termsurvivorshipas children grow up to become adults, by followingup late-effects of treatment.One idea is for them tocarry a survivor’s passport that contains updatedinformation and is always with them. Anotherproject is PanCare, which focuses on long-termeffects.Thechallenge is that this is not like caring forthosewith just diabetes or aheart condition–wewillstill needmultidisciplinary teams throughout.”

children should be looked after by paediatriciansthroughout their childhood and not just referred tospecialists.”

She has been fortunate, she adds, to haveworkedwithmentors such asGadner inVienna andThierry Philip in Lyon, and notably Olivier Hart-mann at the InstitutGustaveRoussy inParis, “a fan-tastic personality”, who died in 2009.

Despite her workload, Ladenstein, who has ateenage daughter, has many hobbies, includingmountain hiking, sailing, diving andballroomdanc-ing. She may need all her nifty footwork skills indancing the tangowith thepowers that be to achievethe establishment of theEuropean virtual institutefor paediatric oncology, which is one of key aims forthe next few years.

the SIOP International congress when it convenesoutsideEurope (this year itwill be inNewZealand).But as Ladenstein adds, it is important that net-working continues in countries that aremost under-represented in membership and that have moretrouble meeting standards of care, such as in east-ern Europe. “SIOPE has a partnership with thenational society in Poland, and we jointly drew uptheEuropean Standards ofCare forChildrenwithCancer,” she says.

A survey of the state of regulations and standardsof children’s cancer care in 27European countriesconducted in 2008by JerzyKowalczyk, of theChil-dren’s Hospital in Lublin, Poland, revealed thatonlyAustria, Belgium, France, Germany and Italyhad officially recognised regulations in place, withthemost comprehensive in Germany.

SIOPE’s European Standards of Care, whichwere developed from this, are described by Laden-stein as guidelines on the minimum requirementsfor bringing children and families through intensivetreatment, “on factors such as access to drugs andprotocols, sufficient teammembers, how childrenare looked after in thewider context such as school-ing, and so on.”

Given the importance of trials and protocols forbest outcomes in paediatric oncology, having ded-icated cancer centres for children is also critical,adds Ladenstein, but it can mean travelling andstaying a longway fromhome.And at primary carelevel, those countries that have paediatricians whosee children as they grow up are also in a betterposition than those where children mainly seegeneral practitioners.

”You need a lot of expertisewith children to sus-pect the symptoms of cancer early on. For example,a child needs to be undressed completely to seepossible swelling associated with neuroblastoma,which would more commonly be attributed to acondition such as gastroenteritis. Similarly, tired-ness and swellings can be associated withleukaemia. These are things I teachmy universitystudents in basic oncology classes. I do feel that

CoverStory


“Having dedicated cancer centres for children is

critical, but it can mean staying a long way from home”

e-GrandRound


Treatment of triple negativebreast cancer

Triple negative breast cancers, as a subgroup, are associated with a poor prognosis. But different

subtypes within triple negative disease are associated with different outcomes, and they also

differ in the way they respond to different treatments. HereAngelo Di Leo provides an overview

of what is currently known and the questions being addressed by ongoing clinical trials.

In treating triple negative breast can-cer, the first thing is to ensure thediagnosis is correct – it is essential to

correctly evaluate oestrogen receptor(ER), progesterone receptor (PgR) andHER2 in primary tumour samples toeliminate ‘false’ triple negative tumours.TheALTTO trial, which is testing lapa-tinib and trastuzumab in the adjuvanttreatment ofHER2-positive breast can-cer patients, included central pathologyreview for ER, PgR and HER2 status.Results showed discordance in 4%–16% of cases between the evaluation ofER, PgR andHER2 in local laboratoriesand the central laboratory (personalcommunication,GiuseppeViale, Euro-pean Institute of Oncology, Milan).

This is a serious issue because adju-vant therapy decisions are largelyguided by these biomarkers. Good com-munication between oncologists andpathologists is essential to try to reducesuch discordance.

Another important consideration isthat not all triple negative tumours havea bad prognosis. For example, a retro-spective study of 13 InternationalBreast Cancer Study Group adjuvant

TheEuropeanSchool ofOncology presentsweekly e-grandrounds which offer partici-pants theopportunity to discussa rangeofcutting-edge issues, from controversialareas and the latest scientific develop-ments to challenging clinical cases, withleading Europeanexperts in the field. Oneof these is selected for publication in eachissue of Cancer World.In this issue, Angelo Di Leo of Sandro Pit-iglianiMedicalOncologyUnit, PratoHospital,Italy, reviews the challenges of managingtriple negative breast cancer, including thebiological heterogeneitywithin the subgroupthat can impact on clinical outcomes, theclinical trial evidence with cytotoxic agentsand emerging data with PARP inhibitors.Lisa Carey, of the Lineberger Comprehen-sive Cancer Center, University of North

Carolina, United States, poses questionssent in by participants during the live pres-entation. The e-grandround was sum-marised by SusanMayor.

The recorded version of this and other e-grandrounds is available at www.e-eso.net

trials reported atASCO2010 identifiedpatients with ER-negative, grade 3 earlybreast cancer and compared outcomesfor central pathology laboratory con-firmed medullary cancers (n=47) andductal infiltrating cancers (n=1407).Despite high grade and ER negativity,patients with medullary tumours hadless vascular invasion, better disease-free survival and better overall survivalcompared with those with the ductalsubtype (see above).The better prog-nosis for the medullary subtype shouldbe considered in treatment decisions.

SENSITIVITY TOCYTOTOXIC AGENTSAnthracyclinesAnthracyclines have been used to treatbreast cancer for many years. Triplenegative tumours may have prolifera-tion-driven overexpression of theanthracycline drug target topoisomeraseII (topoII) alpha and impaired DNArepair due to BRCA1/2 dysfunction.As such, this subgroup might be par-ticularly sensitive to treatment withanthracyclines.

Preclinical data support the conceptof increased activity for topoII inhibitorsin tumours carrying BRCA1/2 dys-function. In a cell linemodel, the topoIIinhibitor etoposidewas administered toBRCA1 wild type and BRCA1 defi-cient breast cancer cells, and BRCA2wild type and BCRA2 deficient fibrob-lasts. Differential cytotoxicity wasobserved based on BRCA status withgreater cytotoxicity in cells with BRCAloss (see opposite, top graphs). Thesame cell lines were pretreated withaclarubicin at low dose to block thetopoII binding site without causing celldeath, then re-treated with etoposide.This resulted in markedly reducedetoposide cytotoxicity and eliminationof any differential BRCA effect (seeopposite, bottom graphs). This high-lighted that topoII-inhibitor-inducedDNA damage is predominantly medi-ated by topoII binding rather than directDNA binding.

In addition, an exploratory plannedevent-free survival analysis by fourmolecular subgroups (defined usingER,PgR, grade, HER2) reported at ASCO


e-GrandRound

2010 looked at five adjuvant clinicaltrials comparing anthracyclines withCMF (cyclophosphamide, methotrex-ate, 5-fluorouracil) in the adjuvant treat-ment of early breast cancer. Resultsshowed no clear superiority of anthra-cyclines over CMF in the so-calledhighly hormone sensitive tumours,whileanthracyclines appeared to be betterthanCMF inmoderately hormone sen-sitive, HER2 amplified and triple nega-tive subgroups.Notably, anthracyclinesseemed to bemore active thanCMF inthe triple negative subgroup (see p16).

These data have not been confirmedby other groups, but this is the largestdataset to date looking at these differentsubtypes. At ASCO 2009, contrastingresultswere presented in the exploratorysubgroup overall survival analysis of theCanadian NCI-MA5 trial comparinganthracycline-based therapy versusCMF. In this analysis, CMF wasfavoured over anthracyclines in the corebasal triple negative subgroup.

Retrospective and underpoweredclinical data are interesting but con-flicting, and as such cannot be trans-lated to current clinical practice. Moredata are required from larger studies.

Summing up this section, cell linesand molecular pathology data suggestthat triple negative ductal infiltratingcarcinoma may have increased sensi-tivity to anthracyclines. Clinical datafrom retrospective studies with limitedstatistical power are controversial.Whilewe wait for more data – ideally fromprospective studies – anthracyclinesshould still be considered an importantcomponent of chemotherapy regimensfor triple negative tumours.

Question: You talked about medullarycancers, and what you called ‘false’ triplenegative tumours. Can you expand onthe different categories of prognosis sep-arate from medullary and other triplenegatives?

NOT ALL TRIPLE NEGATIVE TUMOURS ARE BAD

A pooled retrospective analysis of 13 adjuvant trials showed high-grade, ER-negative breast cancersvary markedly in outcome depending on whether they are of medullary or ductal subtypeSource: Huober et al. (2010) JCO 28 (suppl 15): abstract 630

showing a remarkably high, 72%, pCRrate with neoadjuvant cisplatin. Inter-estingly, this is the only study in whichpatients were selected according tothe presence of BRCA1 mutation.

This is an important signal, whichseems to suggest there can be extraor-dinary activity with cisplatin if patientsare selected according to the presenceof BRCA1 dysfunction.

Answer: This is an important point. Myimpression is that we have slightly ignoredthe heterogeneity in the histology of triplenegative tumours. Beyond ductal carci-noma, there are also medullary, apocrineand squamous cell carcinomas whichare also triple negative. Based on retro-spective data and not on prospectivelydesigned studies, my impression is thatthese different subtypes may need a dif-ferent treatment approach. Medullarytumours have better outcomes than infil-trating ductal triple negative carcino-mas. So I think we should make an effort– ideally a collaborative effort – to gainmore knowledge about the treatment andprognosis of the uncommon subtypes.Question: One of the things that strikesme about patients with triple negativebreast cancer when they come to theclinic is that they are convinced theyhave a poor prognosis, no matter what. Itmight be interesting to remind readersthat disease-free survival at five years fortriple negative breast cancer of conven-tional types is 85%, so many patients cando quite well.Answer: When patients know they havetriple negative status, they are often ter-rified, independently of whether theyhave medullary or ductal infiltrating car-cinoma. I think it is our task to explain topatients that this is a disease in whichthere is heterogeneity in terms of out-come, even though these tumours are allcharacterised by triple negative status.

PLATINUM COMPOUNDSFour neoadjuvant, single-arm trialshave been reported with platinumcompounds in triple negative breastcancer, in which patients were treatedwith cisplatin alone or in combina-tion. Results showed pathologicalcomplete response (pCR) rates rang-ing from 15% to 72%. It is particularlyimportant to note results from a smallstudy reported at ASCO 2009 (Gron-wald 2009, abstract 502) of 25 patients

e-GrandRound


Preclinical data show that DNA damage and cytotoxicity from topoII inhibitors is greater inBRCA deficient cells (top) and that cytotoxicity is mediated indirectly by topoisomerase binding,rather than direct DNA binding (bottom)Source: Treszezamsky et al. (2007) Cancer Res 67:7078–7081

BRCA1/2 DEFICIENT CELL LINES RESPOND MORE TO TOPOII α INHIBITORS

BRCA1 deficient breast cancer cells andBRCA2 deficient fibroblasts exposed to etoposide

Etoposide 25 µmol/L in presence of low dosetopoII α inhibitor (aclarubicin)

defined not from the archived primarytumour sample but in real time fromcirculating tumour cells. Non-triplenegative patients are also being treatedin this trial, providing a control arm todemonstrate that the superiority of aDNA damaging regimen occurs pre-dominantly in triple negative disease.

Predicting sensitivity to DNAdamaging regimensSeveral studies have been conducted toidentify molecular markers of sensitiv-ity to DNA damaging regimens. AJapanese study in 60 patients with earlydisease treatedwith neoadjuvant epiru-

bicin/cyclophosphamide followed bydocetaxel assessed DNA repair pro-teins on tumour samples at baselineand 18–24 hours after chemotherapy.The panel of DNA repair proteinsassessed by immunohistochemistryincluded BRCA1, Rad 51, γH2AX, andconjugated ubiquitin (all at baseline)and Rad 51 (post treatment). Theseproteins are all involved in homologousrecombination, which is a key mecha-nism of repair of double-strand DNAbreaks induced by anthracyclines.DNAdamage response score was assessedfrom 0 to 4, with the highest scorereflecting the greatest efficacy in DNA

A single-institute phase II trial with cis-platin in triple negative advanced breastcancer was reported at ECCO/ESMO2009, in which 126 patients pretreatedwith first-line chemotherapy were ran-domised tometronomic (repetitive, lowdoses) oral cyclophosphamide andmethotrexate with or without cisplatin.Results showed better outcomes inpatients treatedwith cisplatin: CR8% vs5%; PR 55% vs 28%; median time toprogression 13 vs 7 months; medianoverall survival 16 vs 12 months(Bhattacharyya et al. Eur J Cancer 7(3 Suppl):18). This gives further evi-dence for efficacy of cisplatin in triplenegative breast cancer, but more dataare needed.

Another study randomised patientswith metastatic triple negative breastcancer to cetuximab followed by cetux-imab plus carboplatin on progression, orto cetuximab–carboplatin (JCO 26:abstract 1009). Results showed higheroverall response rate when the combi-nation was used initially (17% vs 6%with cetuximab alone) and greater clin-ical benefit (31% vs 10% with cetux-imab, and 25% for the two drugs usedsequentially).

The first-line TNT trial, fromKing’sCollege London, in patients with cen-trally confirmed triple negativeadvanced breast cancer is underway tocompare four treatment arms: plat-inum, platinum plus PARP inhibitor,docetaxel, and docetaxel plus PARPinhibitor. Patients may cross over atprogression (personal correspondence,Andrew Tutt, King’s College London).I think this trial will fully address theissue of platinum compounds in triplenegative advanced disease.

We are also starting a phase II ran-domised study in Italy comparing a nonDNA damaging regimen (capecitabineplus oral vinorelbine) with aDNAdam-aging regimen (cisplatin plus cyclophos-phamide). The triple negative status is


e-GrandRound

ANTHRACYCLINES VS CMF IN DIFFERENT SUBTYPES

An exploratory (planned) analysis of four adjuvant clinical trials showed that certain subtypes of breastcancer, including triple negative tumours, are more sensitive to anthracyclines than to the CMF regimenSource: Di Leo, presented at ASCO 2010, abstract 519

repair (Asakawa et al. Breast CancerRes 12(2):R17). The DNA responsescore showed an inverse correlationwith tumour shrinkage and responserate. The lowest score was associatedwith the highest efficacy (see below).These data are very interesting, butvery preliminary.

TheComet assay is apotential tool for identi-fication of DNA dam-age and prediction ofsensitivity toDNAdam-aging therapy. Differentchemotherapy agentscan cause DNA strandbreaks leading to DNAfragmentation. This frag-mentation can bemeas-ured usingComet, a testcommonly used in toxi-cology, in which cellscrapings are taken fromfresh primary tumourand corresponding non-cancer tissue. Cells arelayered on agarose pre-coated slides. Single gelembedded cells are lysedto isolate nucleoids con-taining supercoiled loopsof DNA. LabileDNA atsites of damage is able tounwind andmigrate differentially out ofthe cell during electrophoresis. Whenobserved by fluorescence microscopy,cells with DNA damage resemblecometswith a nuclear head and ‘tails’offragmentation (see p18). The comettail in cancer cells are due toDNA frag-mentation, while healthy cells show noDNA fragmentation.

Software linked to the microscopemeasures comet tail length and tailintensity for each cell, and calculatesaverage values for each sample. Thetumour sample Comet score is thenadjusted by comparison to the non-tumour sample score (Mol Biotechnol

Summing up this section, data fromphase II non-randomised trials in theneo-adjuvant setting suggest that triplenegative tumours carrying BRCA1/2dysfunctionmight be highly sensitive toplatinum compounds. Data from phaseII and phase III randomised trials are

still preliminary but arenot against the hypothe-sis that triple negativetumours have increasedsensitivity to platinumcompounds. Ongoingstudies are attemptingto identify molecularprofiles and biologicalfeatures predictingresponse to DNA dam-aging cytotoxics.

PARP INHIBITORSPARP inhibitors poten-tially provide a targetedtherapy approach forpatients with triple neg-ative disease. A tumourwith dysfunction inBRCA and dysfunctionin PARP may lack twoimportant DNA repairmechanisms: the baseexcision repair mecha-nism, which is mainly

dependent on PARP, and homologousrecombination, which is mainlydependent onBRCA1/2. Inactivation ofthese two pathwaysmay lead to endoge-nous cell damage and increased sensi-tivity toDNAdamaging chemotherapy,radiotherapy and other agents that dam-age DNA.

In triple negative breast cancer,homologous recombination may beinactivated due to BRCA1 dysfunc-tion, with compensatory upregulation ofPARP activity and the base excisionrepair pathway. If you then give a PARPinhibitor, you will also inactivate thebase excision repair mechanism.

26:249–261). The non-tumour samplecan have some DNA damage that isinduced by tissue sampling, so it is veryimportant to adjust for this.

In our pilot study of 91 patientswith early breast cancer classified bymolecular subtypes (highly endocrine-

sensitive, moderately endocrine-sensi-tive, HER2 positive, triple negative),results showed no substantial differ-ences in Comet scores by subtype.However, the triple negative subgrouphad the largest inter-patient variation.Therewas also substantial intra-tumourheterogeneity in Comet scores, sug-gesting some areas of a tumour havesubstantial dysfunction in DNA repairwhile others do not. Intra-tumouralheterogeneity in DNA repair and pre-sumed heterogeneity in sensitivity toDNAdamaging agentsmay in part offera biological rationale for resistance toDNA damaging therapy.

e-GrandRound


TUMOUR SHRINKAGE BY DNA DAMAGE RESPONSE SCORE

Epirubicin–cyclophosphamide given neoadjuvantly led to greater shrinkage in tumourswith a low DNA damage response score (DDR)Source: Asakawa et al. (2010) Breast Cancer Res 12(2):R17

phase II randomised study was thebasis for moving to the phase III study,with the same therapy arms, that hasrecently completed accrual in the US.Results should become available in thecoming year.

Remaining questions withPARP inhibitorsWe have a lot of data on PARPinhibitors, but we also havemany ques-tions about their optimal use in breastcancer patients. It is important to under-stand whether there are differences inactivity between continuously adminis-tered oral agents and intermittentlyadministered IV agents. Thismight haveimplications in terms of the degree ofinhibition of the PARP enzyme.

Other questions include: What arethe best cytotoxic partners for these

compounds? What is their activity innon-triple negative disease, particularlyin those displaying some dysfunction inDNA repair? Are there molecular pre-dictors of response toPARP inhibitors aswell as toDNAdamaging agents?Whatis the long-term safety of these com-pounds? This last question is extremelyimportant, particularly because we arealready looking at evaluating these com-pounds in early disease.

Three studies further evaluatingPARP in triple negative disease are:� The UK-ledNeoBIG trial. This will

test iniparib in combination withdocetaxel or carboplatin/gemcit-abine in the neoadjuvant setting.

� The BIG trial, co-ordinated by theInternational Breast Cancer StudyGroup. Thiswill test iniparib in com-binationwith epirubicin in the first-


e-GrandRound

Cell death due to loss of these twopathways is called ‘synthetic lethality’.This is the rationale for the use of PARPinhibitors in cancers carrying BRCA1dysfunction, and also for the combina-tion of DNA damaging chemotherapywith PARP inhibitors.

Several PARP inhibitors are beingevaluated in clinical trials. The mostadvanced data on PARP inhibitorswere reported recently in The Lancetby Andrew Tutt and colleagues (vol376, pp 235–444). This internationalstudy was undertaken in BRCA-mutated patients with advanced, heav-ily pretreated breast cancer, in whichall patients were treated with the oralPARP1 inhibitor, olaparib (400 mgtwice daily or 100 mg twice daily).Both doses showed activity in thisheavily pretreated population, withthe 400 mg dose showing a higheroverall response rate of 41% comparedwith 22% for the 100 mg dose. Sometoxicity was reported, with the maingrade 3–4 side-effects being fatigue,nausea, vomiting and anaemia. Ola-parib appears to be tolerable, withmanageable toxicity.

Other data come from a randomisedphase II trial with the parenteral PARPinhibitor iniparib (NEJM 364:205–214).Patients had triple negative metastaticbreast cancer, but dysfunction inBRCA1 was not considered in the eli-gibility criteria, as it was for the ola-parib study. One hundred and twentypatients participated in this study.Mosthad received one or two prior lines oftreatment for metastatic disease. Thecontrol arm was gemcitabine plus car-boplatin and the experimental arm wasthe same chemotherapy plus iniparib.Results showed a higher rate of clinicalbenefit in the iniparib arm (55.6% vs33.9%, P=0.015), with longer progres-sion-free survival (5.9 months vs 3.6months, P=0.012) and overall survival(12.3 vs 7.7 months, P=0.014). This

PREDICTING SENSITIVITY TO DNA DAMAGING AGENTS

The Comet assay distinguishes healthy cells (a) from cancer cells, which have a comet-like tailindicating DNA fragmentation (b, c, d)Source: Galardi et al., presented at the IMPAKT Breast Cancer Conference, Brussels, 2010

Lisa Carey, from the Lineberger Comprehensive Cancer Center, University ofNorth Carolina, USA, hosted a question and answer session withAngelo Di Leo.

LC: The last section focusing on PARPinhibition is probably one of the mostexciting areas. We have one phase IItrial, but I think it is also worth notingthat there were triple negative cancersthat do not carry germline BRCA muta-tions, and there have been some phase IItrials using various agents that have notbeen so exciting. This is obviously anarea to be determined in a number oftrials moving forward.Could you expand on long-term safety?There is so much enthusiasm for thesedrugs that if they do turn out to be positivein the phase III setting, I think people mayuse them quite widely. You commented alittle about your concerns in the adju-vant setting.AD: I think this is an interesting field ofresearch. The concern is related to poten-tial secondary tumours, leukaemia andmyelodysplastic syndrome, particularlywhen these agents are used in combina-tion with DNA damaging agents, specif-ically anthracyclines. This is an issue thatneeds to be clearly explored. DifferentialPARP inhibitor effect in tumour andhealthy cellsmight suggest thatwewill notsee any long-termsafety concerns, but thisis an issue that will have to be carefullyevaluated in future-generation adjuvanttherapy trials with PARP inhibitors.

Q: At which point do you regard apatient triple negative in terms of ER andPR levels of negativity? The conventionin the US has been to use ASCO/CAPguidelines, which set the threshold verylow, so that essentially any ER or PRstaining is called positive. We all recog-nise that this may misclassify somepatients. The CALGB triple negativetrial allows up to 10%. What has beenyour experience?AD: I think there is no definite answer.We can say that you and I are co-ordi-nating efforts across theAtlantic, in thecontext of the BIG and the US inter-group collaborative effort. We are tryingto look at some neoadjuvant and adju-vant therapy trials in an attempt tounderstand if there is a threshold thatcan tell us when a tumour has to beconsidered triple negative versus nontriple negative. For the time being, myimpression is that if a patient is below10% for ER and PgR, and has HER2-negative status, I would feel comfortablein considering the tumour triple nega-tive, assuming there is ductal infiltratinghistology and other characteristics sug-gesting triple negative status, such ashigh proliferation.Q: Could you comment on the role ofbevacizumab in the treatment of triple

negative breastcancer? Arethere any data?AD: Some ret-rospective ana-lyses have beendone in the context of metastatic breastcancer trials where bevacizumab hasbeen tested, but I do not think thatthese data were strong enough to makeany definitive comment on its role intriple negative breast cancer.An ongoingtrial is testing bevacizumab in the adju-vant setting after chemotherapy, specif-ically in triple negative tumours. Myimpression is there is no strong rationalefor thinking that antiangiogenic treat-ment should be particularly active intriple negative disease.LC: I agree. I do not know of any reasonto think that bevacizumab works any bet-ter or worse in triple negative than in anyother subgroup of breast cancer. The for-est plot seems to show that the extent towhich it works is fairly uniform across allsubsets of breast cancer. It is a real chal-lenge for us as to how we select for thesedrugs as we go forward. The neo-adjuvantCALGB trial also has randomisation toreceive or not receive bevacizumab, sothat would also be directly tested there inthe neoadjuvant setting.

line treatment ofmetastatic disease,followed by cyclophosphamide, cis-platin plus iniparib. This study, aswell asNeoBIG, has been designedwith the aim of obtaining comple-mentary data that can lead to amorerational design of an adjuvant ther-apy trial in the context of BIG.

� The US NSABP neoadjuvant trial.This is an interestingdesignproposal,

addressing the role of PARP inhibi-tion and the role of anthracyclines,with the primary endpoint of pCR.

The take home message on PARP inhi-bition is that this is a targeted approach,particularly in tumours carrying DNArepair dysfunction suchas triplenegativecancers. Several PARP inhibitors (oraland intravenous formulations) are cur-rently being tested in clinical trials. Pre-

liminary data fromphase II trials suggestthat this new class of agentsmay be par-ticularly active in triplenegative tumourswithoutmajor safety concerns.Althoughcurrent data are very provocative, treat-ment with PARP inhibitors is not yet a‘standard’for daily practice, andnoPARPinhibitor has yet been approved. Furtherresults are awaited and these will proba-bly become available in the next year.

e-GrandRound


competing research groups scrambled tobe the first to identify and validate signa-tory gene sets. But a head-to-head racebetween teams in Amsterdam and Rot-terdamseekinga signature thatwoulddif-ferentiatebetweenbreast cancers athighrisk of recurrence and more indolenttumours gave the first indication thingsmight not be quite so simple. Both sig-naturesproved tobequiteeffectiveatpre-dictingprognosis, but only threeof the76genes used in the Rotterdam signaturewere also found among the 70 genesmaking up theAmsterdam signature.

Meanwhile, thegeneral conceptof thetumour signaturewasbrought into increas-ing disrepute as the technology becamemore widely available and hundreds ofdisparate studies were conducted in pop-ulationswhere the recurrence ratewithinthe studiedpopulationwas too lowand thenumbers of genes being studied – andconsequently thenumberofmultiplecom-parisons – was too high. Commentators

Reading the signsThe role of genomic signatures in guiding treatment decisions

� Anna Wagstaff

Read the signature, choose thetreatment. This was the enticingprospect that opened up with

the advent of technologies capableof reading the gene expression profileof tens of thousands of genes intumour tissue.

But more than a decade after thesetechniques became available, even thetwo best-known and tested multi-geneassays are not yet part of mainstreamclinical practice in Europe. AlthoughGenomic Health’s Oncotype DX andAgendia’s Mammaprint (both designedto sort high-risk from low-risk breastcancers), show that the technology hasgot better, quicker and considerablycheaper, thewhole field is lookingmorecomplex than many had hoped.

It was only two years ago that theSt Gallen conference, which has beensetting out consensus guidelines foradjuvant treatment of early breast can-cer since 1978, firstmentioned a role for

multi-gene assays, and then only forwhen traditional clinical and immuno-histochemistry tests are inconclusive.AsEurope’s breast cancer community gath-ers for the 11th StGallen conference inMarch 2011, the role of multi-geneassays in guiding adjuvant treatmentwill again be up for discussion. Theirdeliberations will be followed byspecialists in other fields – colorectal,lung and prostate cancer –who are alsolooking for guidance on adjuvant ther-apy, andwonder howhelpful the variousmulti-gene assays currently beingproposed and trialled in their areascould be.

NOT SO SIMPLEThe idea that every tumourcouldbeclas-sified into clinically relevant categoriesbasedon theexpressionof a signature setof genes led to bonanza time for special-ist biostatisticians in the 1990s. Theyfound their services in huge demand as


CuttingEdge

The concept of a genomic signature that can define any tumour according to its unique biology

has been central to the emerging paradigm of personalised therapy. But how helpful are

multi-gene assays in guiding treatment decisions in clinical practice, andwhat role canwe expect

them to play in the future?

CuttingEdge


ILLU

STRATIO

N:F

RED

VAN

DEELEN

,WW

W.O

RG

AN

ISART.

CO

.UK

talked about the literature being “repletewith rather unconvincing, small studies,whichareunderpowered, variable in theirmethodology and quality assurance andwhich are often linked to incompleteclinical data sets. Even meta-analysis ofthese sorts of studies is rather suspect…so perhaps it is not surprising that littleprogress has been made in defining auseful marker, which would help main-streamclinical decisionmaking,” (CancerJournal 16:210–213).

Also undermining the concept of asingle tumour signature was growing evi-dence that the biological make up canvary considerably from one region of a

tumour to another. One area might testhighlypositive for amplifiedHER2expres-sion, for instance,while another area testsnormal.Taken togetherwith theevidencethat biology changes over time and inresponse to treatment, and that thegeneticsignature of the primary tumour andmetastasis can differ greatly, the questionarises whether a tumour signature mayonly apply for a certain place and time.There is also a growing body of evidencethat thebiologyof thenon-canceroushosttissue surrounding the tumour plays amajor role in determining prognosis andpossibly also points to the need for differ-ent treatments.Perhapsweneedhost sig-natures as well as tumour signatures.

Some specialists also question thevalue of the information given by thegene signature. Ideally, they argue, tar-geted therapies should be directed at afunctional pathway towhich tumour sur-vival is addicted; however, gene statusdoes not reliably correlate with down-stream protein status or with pathwayfunction. Thus, the fact that a tumourshows amplified HER2 expression, forinstance, does not mean that HER2 isnecessarily the driving force, which iswhy not all patients withHER2-positivetumours benefit from HER2 inhibitorssuch as transtuzumab (Herceptin). Thepoint iswell illustratedbyacomparisonofthepathways involved in theAmsterdamand Rotterdam gene signatures, whichdemonstrate that, despite the minimaloverlap between the genes, they had 21pathways in common.

“BUT IT WORKS”Steve Shak, chief medical officer atGenomic Health, is well-equipped toargue the finerpoints of targets and func-tional mechanisms, having come fromGenentech, where he led Herceptinthrough the approval process. But hisanswer to those who question the use-fulness of OncotypeDX is a simple one:it does what it says on the tin – it assists

like John Ioannidis from theUniversity ofIoannina in Greece led calls for fewer,larger, better-designed studies, showingthat the huge data sets that were beinggathered from relatively small numbers ofpatientswereopen to almost any interpre-tation.As he commented at amedia ‘real-ity check’ in 2007: “with 30,000 genes tochoose from, anyone looking for a signifi-cant pattern is quite likely to find one,”(Cancer World Jan/Feb 2008).

His remarks were echoed last yearbyRachelMidgleywho, as leadclinicianin theQUASARtrials, has beenexamin-ing theevidence for amulti-geneassay toguide treatment in colorectal cancer. She

ever made it into clinical practice.”Buteverynewtechnologycomeswith

its own learningcurve, and thereare signsnow that the field is maturing. In theUS, theFDA-ledMACQproject is tryingtobringsome levelofqualitycontrol to thewhole field of microarray, RT-PCR andother ‘next-generation sequencing tech-nologies’ andpromote their proper appli-cation in discovery and development.Crucially this includes providing guide-lineson thecapabilities and limitationsofvarious data analysis methods in devel-oping and validating predictive models,and reaching a consensus on ‘best prac-tice’. Mammaprint (which came out ofthe Amsterdam research) has becomethe firstmulti-geneassay tohave its prog-nostic powers recognised by the FDA,although approval by theFDA(orEMA,its European counterpart), is not arequirement for these tests.

Lessonshavebeen learned too about

aremultiple studies to show it is fit for thespecific purpose for which I am going tobe using it,” he says. “We worked withleaders in oncology throughout theworldtodomultiple studies that not only iden-tified the best genes but also validatedtheiruse inmultiplewell-defined rigorousclinical studies.We’venowdone14stud-ies inmore than4000patients.” TheuseofOncotypeDX to guide adjuvant treat-mentdecisions inearlyER-positivebreastcancers is now included in thepublishedASCOclinical guidelines.

Shak readily concedes that the track-record of research into molecular diag-nostics in general has not been a gloriousone. “Whenwe startedGenomicHealthin 2000,we looked across a landscape ofbiomarker research and development,and it was easy at that time to get verydepressed. There were tens of thou-sands of articles on thismarker and thatmarker, and very few if any of those


CuttingEdge

patients anddoctors todecidewhether toopt forchemotherapy incasesofearlyER-positive breast cancers, by indicating thelikelihood of recurrence and of responseto treatment.

OncotypeDXusesa techniquecalledreal-timepolymerase chain reaction (RT-PCR) tomeasure levelsofmRNAexpres-sion from21genes in samples of tumourtissue taken fromastandardparaffinbloc.Using these results, it then calculates a‘RecurrenceScore’of between0and100,whichcorresponds to the likelihoodof thecancer returningwithin 10 years. Scoresbetween 0 and 17 are defined as ‘lowrisk’; 18 to30as ‘intermediate risk’and31and above as ‘high risk’.

Shak says that the Oncotype DXbreast cancer assayhas successfully comethrough thekindof close scrutiny thatheandhis colleagueswelcome. “As aphysi-cian or patient I want a test that hasdemonstrated it really works; that there

� Gene signatures, or multi-gene assays, can be captured in avariety of ways. Agendia uses the Agilent microarray platformto generate theMammaprint, which classifies early ER+ breastcancers into high risk and low risk, based on the 70-gene sig-nature originally identified by the Amsterdam team. Alternativemicroarray platforms include Affymetrix, which was used for the76-gene Rotterdamassay and has also been used to generatea putative 23-gene signature to predict for recurrence in coloncancer (JCO 27:1564 –1571). This technology requires frozentumour specimens, although new techniques are being devel-oped to enable it to be usedwith paraffin-embedded specimensas well.

� Genomic Health uses an alternative technology, real-time poly-merase chain reaction (RT-PCR), to read the 21-gene signaturethat forms the basis of the Oncotype DX multi-gene assay forbreast cancer. This assay generates a Recurrence Score foreach tumour of between 1 and 100. This can be done usingparaffin-embedded tissue.

� Agendia argues thatMammaprint ismore helpful, in that it ratestumours as either ‘high risk’ or ‘low risk’, thus avoiding the

chance that readingsmay comeback as inconclusive. GenomicHealth counters that all the evidence shows the biology of ER+breast cancers to be a continuous variable, and that giving riskaccording to a continuous recurrence score is thereforemoreaccurate.

� There is disagreement also over which of the two assays canclaim greater validity.Mammaprint is the only assaywhose pow-ers to predict outcome have been approved by the FDA. On thestrength of this and other evidence, it is now reimbursed byMedicare in the US. However, Oncotype DX is the only assayto be included in the ASCO clinical guidelines as a tool for decid-ing who will benefit from chemotherapy (JCO 25:5287–5312),and it too is reimbursed by Medicare as well as Medicaid andthe major US health insurers. The evidence levels behind thedecision to include only OncotypeDX in the guidelines has beena point of debate (JCO 25:2057–2058; JCO 2058–2059;JNCI 101:1456–1452). More information about their relativemerits will be available with the results of the large Tailor X andMINDACT trials, but these are not due to report final results formany years yet.

Multi-gene assays – the story so far

the need for a more collaborativeapproach to secure the patient numbersnecessary to provide reliable results.Trials thesizeofMINDACTandTailorX,which seek to establish how effectivelyMammaprint andOnctopyeDX predictwho will benefit from adjuvantchemotherapy, haveenrolledmany thou-sandsofpatientswhowill be followed formanyyears.Whentheyweresetup, itwashard to envisage studies of a similar scaleoutside of the highly organised andwell-funded field of breast cancer. Yet we arenow beginning to see sizeable trials formolecularmarkers andmulti-geneassaysbeing carried out in colorectal cancer bysome of the big European and US col-laborative groups, including a variant ofOncotype DX for predicting recurrencerisk in colon cancer, being tested by theUS NSABP (National Surgical Breast

revealed themselves in terms of whichonesworked, theyactuallydidgroup them-selves in away thatwas very relevant toourunderstanding of breast cancer biology.”

This remains true, arguesShak, evenif you lookat someof themost recent evi-denceabout thebiologyof thehost tissue.CD68, amonocytemacrophage-relatedgene involved in the body’s immune sys-tem is one of the 21 signature genes.“That gene relates to thehost response tothe tumour. So our test captures infor-mation about the tumour and hostresponse and integrates it to give singlescore that is relevant for selection of theright treatment.”

As for theargument thatmRNAmeas-ures do not provide the most accuratereflection of the signalling pathways thataredriving the tumour, Shak is pragmatic,describinghimself as ‘agnostic’onwhetherDNA, RNA or protein offers the mostuseful information.GenomicHealthoptedfor measuring mRNA using the RT-PCRtechnique,he says, becausewith the tech-nology currently available, that is whatworks best. “Proteins are complicated,they undergo clipping and post-transla-tional modifications, the assays to look atthem are complicated and challengingand I think the issue is not oneofwhich ismore important, but which one can bepractically harnessed to serve patients.”

Indeed, the practicality of the Onco-typeDXtest iswidely regardedasone if itsbig advantages, because it can be doneusing paraffin-embedded tumour blocsthat are routinely collected for pathologyreports, it canbedeliveredbyanyovernightcourier serviceandGenomicHealthpridesitself on ahigh level of quality control andcan point to an excellent track record ofreproducibility and accuracy.

CuttingEdge


Do I have to do this? Finding more accurate ways to predict who will benefit from chemotherapy andwho will not would save countless cancer patients unnecessary suffering and cut healthcare costs

and Bowel Project) using tumour tissuefrom the QUASAR trials (see p 27,Beyond breast cancer).

THE SCIENTIFIC RATIONALEThe validity of these multi-gene assaysrests on more than just the statistical evi-dence that theywork, saysShak.Theset of21 genes used for theOncotypeDXassay– 16 ‘cancer genes’ and five referencegenes – make sense in terms of what isalready known about the biology of breastcancer. “We picked the genes becausewhen we looked across the studies theyprovided evidence that they really work.But the [cancer] genes actually turnedoutto be in four groups and then three inde-pendent genes.” Those four groups, heexplains, relate to proliferation, oestrogenreceptors,HER2and invasion. “Itwas veryreassuring to us that, when the genes

We are beginning to see sizeable trials for

multi-gene assays carried out in colorectal cancer

CO

RBIS

/ED

KASH

I

VIEW FROM THE CLINICCatherineOakman, amedical oncologistwith a special interest in breast cancerbased at the Sandro Pitigliani MedicalOncology Unit, in Prato Hospital, Italy,will be among themanymaking the jour-ney toStGallen thisMarch to learnmoreabout how evidence that has emergedsince March 2009 might impact onguidelines for clinical practice.

Lead author of an overview piece onbreast cancer assessment tools and opti-mising adjuvant therapy, publishedat theend of last year in Nature ReviewsClinical Oncology (vol 7, pp 725–732),Oakmanfollows the recommendationsofthe St Gallen 2009 consensus, whichpose the roleofmulti-geneassays as a lastrather thana first step in considerationofadjuvantchemotherapy for awomanwithER-positive disease.

Mammaprint and Oncotype DX canboth provide a molecular picture of bio-logical features, saysOakman,but tumoursare already assessed by clinical features –tumour size and lymphnode involvement–andbyhistopathological features–mor-phology, and ER, PgR, HER2 and Ki-67(proliferation)status.Theseareall capturedby the traditional St Gallen criteria, shesays, so the issue becomes one of howmuch additional information the multi-gene assay can supply.

“St Gallen criteria are very practicalbecause they incorporate tumourparam-eters which are readily assessed in dailyclinical practice. With that informationyoucanmakea therapydecision formostpatients.For individualswithER-positivedisease in whom you are still unclearabout additional benefit of chemotherapyover endocrine therapy alone, the St

Gallenconsensusadvisesconsiderationofa genomic test. The genomic test comesat the end and only if you are unsure.”

She does concede that the reliabilityof immunohistochemical markers hasbeen a problem. In trials where localpathology reports arequality controlledbyacentral laboratory, discrepanciesofup to20% have been found for ER, PgR andHER2.All the tests involve some level ofreader evaluation, anduntil recently therewasno standard agreement aboutwhereto set the diagnostic thresholds.

Improvementsmay come, saysOak-man, from close working relationshipsbetween oncologists and pathologists,andstandardisationofpathology testing–something she believes is already hap-pening in a number of ways.

Greater involvement inmultidiscipli-nary teams means pathologists areincreasingly awareofhowthe results theyreport are used to guide treatment deci-sions, and this gives them an enormousincentive to get themost accurate resultsthey can. There are also new step-by-step guidelines, published by ASCO inconjunction with the US College ofAmericanPathologists (CAP), for testingER,PgRandHER2.They canbe imple-mented in any hospital laboratory, andthough they are not binding, they set aquality standardbywhichanyhospital canbe evaluated.

ASCO-CAPguidelineshavenowsetthe threshold forERandPgRpositivity asat least 1%of tumournuclei stainingpos-itive.TheKi-67proliferationmarker,how-ever, remains problematic both in termsof an agreed threshold and standards fortesting. Oakman accepts that genomicmarkers are currently superior in robustly

and reproducibly measuring prolifera-tion, but argues that given time, stan-dards and guidelines may bring morereliability to thismarker as well.

For some patients, treatment deci-sions are fairly clear. Evidence showsthat, in general, tumourswith a veryhighER and PgR status are sensitive toendocrine therapy but derive little addi-tional benefit from chemotherapy. Indi-vidualswithHER2-positive tumours arelikely toderivebenefit fromtrastuzumab,and this is always given alongsidechemotherapy. For triple negative dis-ease, chemotherapy is the only currentlyavailable systemic option.

The problem arises where tumoursstain positive for ER, but show somesigns of a more aggressive cancer, forinstance low PgR, grade 2, intermediateKi-67 and/or some nodal involvement.This represents a sizablegroupofwomen,as about one third of the 75%of patientswith anER+ tumour fall into this uncer-taincategory, and thequestionofwhetherthese patients might derive additionalbenefit from chemotherapy overendocrine therapy alone is unclear. “Thisis the group of patients where genomictestsmight help,” saysOakman, “That isof course if, firstly, the patient can affordor has insurance to cover the substantialcost, and secondly, if shewouldbeagree-able to chemotherapy if the resultsshowed her genomic risk to be high.”

Even then, there is no guarantee thatthegenomic testwill provide theguidancedoctor and patient are looking for. “Ifsuch a patient is assessed as high or lowriskbyOncotypeDX, thishelps treatmentdecisions. However if such a patient isassessed as intermediate risk, I am still


CuttingEdge

The issue becomes one of how much additional

information the multi-gene assay can supply

chemotherapy, 40%ofwhomwill suffersignificant toxicity. The future forgenomic tests like Oncotype DX andMammaprint will lie in how far theyoutperform traditional clinical andpathological criteria and how the costsstack up against the benefits for patientsand the savings for health services.

However, even if the trials show

unclear about the estimated additionalchemotherapy benefit,” says Oakman.Shak, fromGenomicHealth, cites a sur-vey indicating that in 30% of cases, test-ingwithOncotypeDXdoes in fact lead toa change in the initial decision on treat-ment. This is not something Oakman isable to confirm fromherownexperience.

Currently both Oncotype DX andMammaprint are reimbursed in someEuropean states, but so far only by asmall minority of social insuranceproviders. Oakman suggests it would bebest to wait for the results of the TailorXand MINDACT trials to get a clear pic-ture of their powers to predict benefitfrom chemotherapy in this patient sub-group, before deciding whether publichealth insurances shouldcover thecostofgenomic tests, at least when traditionalmarkers are inconclusive.

ARE GENOMIC TESTS THE FUTURE?Tailoring treatments topatients is a strate-gic goal for oncology. For many people,this paradigm is all about tracking downtargets and designing biological thera-pies that can block them, and thisremains the great hope for the future.

For the moment, however, chemo-therapy remains amainstay of treatmentformany cancers, and as early detectionstrategies ensuremore andmore cancersare caught at an early stage, the ability tosort the patients whowill derive benefitfrom these toxic treatments from thosewho will not is by far the greatest‘personalisation’ issue in terms of thenumbers of patients affected.

Using current regimens in stage IIcolorectal cancer, for instance, curing anadditional three to four patients requiresputting one hundred patients through

beyond doubt that these genomic testshave clinical value, they still predict only‘general chemosensitivity’ rather thanspecifying a treatment. The quest forpredictive biomarkers for specificchemotherapies is another level ofsophistication for the future (see alsoTreatment of Triple Negative BreastCancer, p 13).

CuttingEdge


Genomic tests are reimbursed in some European states,

but only by a small minority of social health insurers

Beyond breast cancerIn January last year Genomic Health launched an Oncotype DX (RT-PCR) assay forpredicting the risk of recurrence in patients with stage II colon cancer (defined asinvolvement of the bowel wall without lymph-node involvement). The 12-gene assaywas shown to be significantly predictive in sorting those with a ‘low’ risk of recurrence(8%–10%within three years) from those with a ‘high’ risk (20%–25% risk (JCO 27 (15s):abstract 4000).

This January, results from a trial of Agendia’s experimental 18-gene Coloprintmicroarray assay were published (JCO 29:17–24), showing it too was significantlypredictive at sorting stage II and III patients at ‘low’ risk of recurrence (around 12.5%recurrence at five years) from those at ‘high’ risk (around 33% at five years).

Neither study claimed to provide statistical proof that the assays could predict whowould benefit from chemotherapy. However, David Kerr, one of the leaders of theQUASARtrial that collaborated in the study of the Oncotype DX assay, said, “When you look at thetotality of the data, we think that the new assay provides a clinically useful tool for[identifying] which stage II colon cancer patients should be selected for chemotherapy,”(Medscape 19 May 2010).

Roberto Labianca, lead author of ESMO’s clinical guidelines for diagnosis, adjuvanttreatment and follow-up of primary colon cancer (Ann Oncol21Supp5: v70–v77), insists,however, that while these data are interesting in the context of the wider search for bio-logical markers that can predict response to treatment in this patient group, they are notsufficiently robust to merit being included in guidelines for adjuvant treatment. For themoment, he says, decisions on adjuvant treatment should continue to be guided by clin-ical and pathological markers which define stage II disease as high risk according toinvolvement of the bowel wall, invasion of vascular, neural or lymphatic vessels insidethe tumour tissue, obstruction or perforation of the tumour during surgery, or where fewerthan 12 nodes have been examined.

The search is also on for multi-gene assays that might predict recurrence and ben-efit from treatment in many other cancers. Tests for prostate cancer and non-small-celllung cancer both feature in the Genomic Health pipeline.

Think before you ruin yourpatient’s chance of parenthoodPrize for journalist who highlighted poor practice in Poland

InPoland, being cured of cancer can oftenmeanbeing sentenced to infertility. The price paid forintensive cancer therapy hits young people juststarting out in life especially hard. But it doesn’t

have to be that way.“These days, patients treated for cancer can in

most cases conceiveandgivebirth tohealthychildren,”insists Professor Mariusz Bidziński, Chairman of thePolish Association for Oncological Gynaecology andHeadof theGynaecologicalCancerClinic at theWar-sawCancerCentre, “There is, however, one condition–actionneeds tobe takenbefore, rather thanafter, start-ing cancer treatment,” explains the specialist.But inPoland fewpeoplegive

it any thought. Patients are terri-fied by the diagnosis, and theirmain concern is to save them-selves. According to Bidziński,theyare rarely told that treatmentcan result in infertility. Doctorstend to focuson treatingpatients,rather than worrying about whatmighthappen to themafewyearsdown the line.“Not a single one of my col-

leagues at the Warsaw CancerCentrehas ever come to seekmyopinion on this matter,” says Bi-dziński. “Someof themare treat-

ing youngwomenwho, after therapy, can forget aboutbecomingmothers,” he adds.“InGermany,France, theBenelux countries or the

USA, a discussion is held with the patient aboutwhether, after successful cancer treatment, heor she isgoing to want to have children – this is standard pro-cedure,” says Dr Kamil Zalewski, who works at Bi-dziński’s clinic. “Not only that, but if a doctor in one ofourwesternneighbours should forget to raise the sub-ject in talkswith thepatient, heor shemaywell have toanswer for it before the courts,” Zalewski points out.“And here in Poland? Forget it!”Maybe Polish specialists would be willing to refer

at least younger patients for con-sultation, but where to? In ourcountry we don’t have a singlecentrewhere a frightened youngman or woman can seek com-prehensive help, where they cantalk about becoming a parent,not just with a cancer specialistbut also, for example,with a spe-cialist in infertility treatment or apsychologist.A fewwell-informed– and well-off – patients seekhelp of their own accord fromprivate in-vitro fertilisationclinics.


BestReporter

The plight of cancer patients for whom treatmentmeans risking losing the opportunity to have

children was taken up in Poland by the leading national dailyGazeta Wyborcza, in a piece that

earned reporter Sławomir Zagórski a Best Cancer ReporterAward. The article, titled “I had

cancer, I’m going to be a Dad,” is reprinted below.

Sławomir Zagórski

Many techniques are available to enable cancer patients to keep open theoption of having children even if the treatment renders them infertile. Thisarticle looks at what must change in Poland to ensure patients get the

chance to make use of these techniques in time

PREGNANCYFROM THE FREEZERWhat can be done, given that the cancer hasto be treated?The first thing to do is talk to the patient. Ask

whether, after successful treatment, he or she is plan-ning – even if only tentatively – to have children. “Atleast half the young people [up to the age of 35] inPolandwho suffer fromcancer– i.e. 3000 a year –willanswer such a question affirmatively,” says Bidziński.If a patient expresses thewish tobecomeaparent,

then action needs to be taken – and without delay,because cancer doesn’t wait.In thecaseofmen, thesituation is relatively straight-

forward. All that is required is an andrological exami-nation to check that the semen contains sperm cells,after which the semen is frozen. This can be safelystored in liquid nitrogen for years.With women, it is more difficult, but in this area

medicine is constantly comingupwithnewmethodsofdealing with the situation. One method relies on

BestReporter


A WORD ABOUT BONE-MARROW TRANSPLANTSDiseases whose treatments lead particularly fre-quently to infertility in both sexes include cancersrequiringbone-marrowtransplantation (certain formsof leukaemia or lymphomas, e.g.Hodgkin’s disease)andalso somesarcomas.Thechances that a twenty-year-oldwoman, after undergoing cancer treatment,will getpregnant spontaneously (i.e.,withoutmedicalhelp) are as a rule no more than five per cent. “Thecancer itself doesn’t have to affect the reproductivesystem – the ovaries or uterus,” explains Zalewski.“Aggressive chemotherapy, in cases of breast cancerfor example, is frequently all it takes for theovaries tostopworking.Andwhen theydo stopworking, it notonly means the end of all dreams of motherhood, italso means a fall in oestrogen production and earlyonset of themenopause,” he adds.Men are not spared either. If, for example, a

twenty-year-old man has to have a bone-marrowtransplant, doctorsmust first completely destroyhisowndiseasedbonemarrow,usingdrugs or radiation.In fourcasesoutof five, suchpre-transplant radiationconditioning fatally damages the delicate Leydigcells in thepatient’s testicles. If thepatientdoesman-age to survive the cancer, in 80%of cases his semenwill never again contain even a single sperm cell.In patients with testicular cancer (a disease that

affects mainly young men) chemotherapy and spe-cialist surgicalproceduresonabdominal lymphnodescan reduce the ability to procreate.

“Ask whether, after successful treatment, he or she is

planning – even if only tentatively – to have children”

BestReporter


freezing egg cells, though unfortunately this is stillmeeting with considerable resistance (barely a fewhundred successful pregnancies have been achievedthiswayacross theworld). Incomparably simpler is thebusiness of freezing readyembryos, but obtaining suchanembryoalso requires a spermcell. If a youngwomanalreadyhas ahusbandorpartner, thecouplecandecideto have embryos prepared and freeze them. If shedoesn’t have a partner, she can elect to be fertilised bysperm fromananonymousdonor, though such a solu-tion is, of course, not for everyone.For some years now, doctors have been (success-

fully) testing a new method of preserving fertility inwomen undergoing chemo- or radiotherapy. Beforethe treatment, the surgeon cuts out one of the ovaries(or a fragment of one) through a small incision in theabdomen, and removes tissue from it,which is cut intothin strips and then frozen.After the anti-cancer treat-ment is completed, and once it is certain that thepatient has completely recovered from the disease,the tissuecanbe thawed, checked toconfirmthat thereare no cancer cells present, and grafted onto theremainingovary. It transpires that the thawed tissuegetsto work and is soon able to produce normal egg cells.This technique was pioneered by Professor

Jacques Donnez of the Catholic University in Lou-vain (Belgium).Anarticle six years ago inThe Lancetbroughtnewsof the first baby tobeborn inhis clinicfollowing freezing and transplantation of fragmentsof ovary. Jacques Donnez’s patient had been suffer-ing fromHodgkin’s disease.Fourmonths after graft-ing the fragments of preserved ovary, the patient’speriods returned, and a few months later she wentinto labour perfectly naturally, without having toresort to in-vitro techniques.Since 2004manymore babies have come into the

world fromfrozenand thawedovary tissue– inPoland,no-one has tried it yet.

WHAT ABOUT THE YOUNGEST PATIENTS?Freezing ovarian tissue should in future offer hope ofmotherhoodeven for young girlswhohave yet to enterpuberty (children get cancer too – in Poland there areabout 1200 new patients a year). Doctors are hopingthat immatureeggcells recovered fromsuch tissuewillone day serve for in-vitro fertilisation.Specialists are also lookingatways tohelpboyswho

arenot yetproducing spermcells.Researchers in Israelare freezing testicular tissue fragments collected fromvery youngcancerpatients. It is not yet knownwhetherthis techniquewill work, but trials are continuing.One thing that canbedonewithoutdifficulty, even

now is the fairly simple procedure ofmovingwomen’sovaries away from irradiated areas– in cases of cervicalcancer, among others (such procedures are being per-formed inPoland;noneedhere for anymajoroperation,as access to the ovaries can be obtained through twosmall openings in the abdomen).Doctors are also experimentingwith removingonly

the cancerous tissue rather than the entire uteruswhere the cancer is caught early, again with a view topreserving patients’ future ability to parent children.Suchconservative procedureshavemadepossible thebirth of almost 500 babies around the world (two ofthem in Poland).

CANCER AND IN-VITRO TECHNIQUESOne of Bidziński’s dreams is the creation of a centreinPoland thatwouldbe responsible for the complexcareof youngcancerpatientswanting tobecomepar-ents in the future. “All itwould takewouldbeameas-ure of goodwill on thepart of the authorities, a bit ofequipment and some funding,” the specialist argues.“Patientswouldbe treated for cancerwhere they live,and theywouldcomeherepurely forconsultationandpossibly to have their semen, egg cells or fragmentsof ovary collected.”The question arises, however, as to who is going

to pay for such services and for the cost of storingpatients’tissues. “Despite the fact thatwehavebeentrying for years to persuade patients to make use ofsperm banks, very few do,” maintains Dr IwonaSkoneczna of theWarsawCancerCentre.At present, Poland is still not providing any kind

of help to its infertile citizens. Yet, as if not enoughmoneywerealreadybeing spentoncancer treatment,cancer patients can still demandhelp for later infer-tility treatment.“Paradoxically, it could be cancer patients who are

eventually the first tobenefit fromsuchhelp,”Bidzińskispeculates. “After all, hasn’t fate already given them arawdeal?Perhapswewhoarehealthy andhappy inourlives could devote a little of our money and energy tohelping them out.”

This article was first published in Gazeta Wyborcza on 27 August 2009, and is republished with permission

A simple doctor’s quest toimprove on today’s treatmentsHow Stan Kaye came to lead drug development at the Royal Marsden

Treating patients who are urgently searching for new therapies is tough and doesn’t get any

easier, saysStan Kaye, headofdrugdevelopment at theRoyalMarsden.He treasures themoments

whenanewdrug turnsup that shows real value.Thebig challenge formedical oncology is tounder-

standwhobenefits fromwhat; success depends on attracting to the field thebrightest and thebest.

Byhis ownassessment, he is a “simple doctor”,not a scientist. StanKayeworks in an officeof few pretensions, displaying pictures of

past players fromhis belovedLeedsUnitedFootballClub rather than professional accolades. He isquiet but avuncular, and says he was never partic-ularly ambitious. Yet it was a simple doctor’s moti-vation – to end the suffering brought on patients byfrankly inadequate cancer treatments – that broughthim to lead one of themost important early clinicaltrials units in Europe.

As head of the Drug Development Unit at theRoyalMarsdenHospital inSutton,England, and thechairmanof the Institute ofCancerResearch’s Sec-tionofMedicine,Kaye isnowoverseeing the first usein humans of a range of new targeted therapies thatcould transformcancer treatment in thenextdecade.

He finds it hard to conceal his excitement, forinstance, about the potential of PARP inhibitors – a

class of drugwhosepossibilitieswere first recognisedby scientists at the Institute ofCancerResearch sixyears ago, andwere subject to their first extendedsin-gle-agent clinical trials at the Royal Marsden threeyears ago. The drugs have yet to reach the market,but there is accumulating evidence that they willbring significant benefits to many patients withbreast andovariancancer, andpossibly others includ-ing prostate, endometrial and colon cancer.

“We are genuinely moving from chemotherapyto much smarter treatments based on a betterunderstanding of what causes cancer, and whatdistinguishes cancer cells from normal cells,” hesays. “This knowledge is being turned into newtreatments, often given as tablets, thatmakepeople’stumours shrink, and it’s just terrific. So in the nextfew years there’s going to be an increasing under-standing that you don’t treat all people with aparticular type of cancer in the sameway, and indi-


Masterpiece

� Simon Crompton

vidualised treatment will increasingly take over.”It’s all a far cry fromhis early days as a researcher

in the 1980s, when he was driven by seeing how“lousy” current treatmentswere. The past five yearsis bringing something truly different, and he canafford to be optimistic.

Kaye’s excitement about PARP inhibitors isrelated to his own special interest, ovarian cancer,where the drug is showing particular promise inhelping both common types and rarer familial formslinked to BRCAmutation. During phase I trials attheDrugDevelopmentUnit, Kaye andhis teamonthe clinical sideworked closelywith scientists whohaddeveloped the drug. In the first trial, 19 patientswith advanced ovarian, breast and prostate cancer,who were all BRCA gene mutation carriers, weregiven the newmedication, having exhausted otherforms of treatment.

“Early on, even in the first patients treated, we

Masterpiece


INSTITUTEOFCANCERRESEARCH

knew we were on to something. It only happensrarely that you see a new treatment really making adifference at this early stage.Wecouldn’t believe thefirst data because, in half the patients, the tumoursshrank. The drug had very modest toxicity andpatients said, ‘This is nothing like chemotherapy.’Sowe were delighted when an international trialshowed the same thing – roughly 50% responserate in recurrent ovariancancer.”Thedrugcouldcon-stitute a significant leap forward for ovarian cancer,Kaye reflects, where developments have been slowin recent decades and largely focused on improvedmultidisciplinary care rather than good drugs.

TheDrugDevelopmentUnit at theRoyalMars-den, one of the biggest in theworld, is buried in thehospital’s sprawling site in Sutton – Lego-styleblocks, puddle-strewn car parks, amuddy buildingsite and the fadedVictorianmonumentwhereKayehas his office.Around600patientswith cancer that

has progressed despite conventional treatmentcome here every year to enter phase I trials ofexperimental cancer drugs, knowing that the oddsare that the drug won’t help, but that there is achance that otherswill benefit from the knowledgegained, even if they do not. The centre takes thegenetic discoveriesmade at the Institute ofCancerResearch’s facilities in Chelsea, London, convertsthem into drugs, and offers them to patients withadvanced cancer. Funding comes partly from thedrug companies that are jointly developing thedrugs, aswell as fromCancerResearchUKand theDepartment of Health.

TRANSLATING GOOD SCIENCEINTO GOOD TREATMENTSIn the 10 years that Kaye has been in charge of thecentre, he’s closed the gapbetween thehospital andthe Institute of Cancer Research, distilling a dis-parate phase I operation scattered across the hos-pital into a single specialist wardwith 25nurses, 15doctors and 80 staff all working exclusively on thetrials. These changes involved major restructuringat the RoyalMarsdenHospital site at Sutton.

It produced a centre that is leading theway, notonly in drug development, but in the way that earlytrials are performed.Phase I trials, believesKaye, areabout tobecomemore important than theyhaveeverbeen, and theRoyalMarsden’s dedication to speed-

ingup the translationof good science into good treat-ments is showing the way ahead.

“The nature of phase I trials has changed a lotsince we conducted our first one in the mid-1980sin Glasgow. For instance, I did a trial with a drugcompany of a green compound. It had some strikingeffects in theexperimentalmodel, and it seemedsafeenough, but it went into the clinic withoutme hav-ing any idea of how itworked. This is somethingwewould rarelydonow.Weneed toknow if adrug ishit-ting its target, and theway they are assessed involvessophisticated biomarker work.You just wouldn’t dotrialsnowwithout all that, but in theolddayswedid.”


Masterpiece

Towards better treatments. Catching up with fellow specialistsat the 4th Ovarian Consensus Conference in Germany, 2004

“It will continue to get better if we get smarter at finding

out which patients should get these targeted drugs”

WHO BENEFITS FROM PARP INHIBITORS?

Mutation of BRCA1andBRCA2genes has been linked tohereditary breast, ovarian and prostate cancer. In 2005,the Institute of Cancer Research team in Chelsea, led byAlan Ashworth, discovered that, because these geneswere not functioning, DNA repair was defective, andthese cancers were therefore exquisitely sensitive todrugsblockinganenzymecalledPARP. Early trials of thesePARP-inhibitors at theDrugDevelopmentUnit showed thataround50%of tumours shrank. This result was replicated

in an international trial. Last year, it was announced thatAshworth’s teamhaddiscovered that thedrugwould alsokill cancer cells with other types of gene faults apart fromBRCA inherited cancers, so its potential ismuch broaderthan originally thought. The drug is now moving intophase III trials. According to Kaye, the beauty of PARPinhibitors is that doctors should be able to predict exactlywhowill benefit from them,because thegeneticdefect thatmakes people responsive will have a biomarker.

The result is likely to be safer treatments and amorerational basis on which to decide whether or not adrugmight go forward. Themajority of drugs fail –less than 10% of the drugs that go into phase I trialssucceed. “But its getting better,” says Kaye. “It willonly continue to get better if we get smarter at thiswhole business of understanding which are thepatients who should get these targeted drugs.”

Nowadays, phase I studies are farmore than theexperiments in safety they used to be. “They aim toestablish correct dosages, indicate substantial activ-ity, and to expand the number of patients involvedwith a real expectation of benefit.”

Kaye continually emphasises that patients needto be realistic about the prospect of significantimprovements in phase I trials. But the chance ofbenefit is still there; for instance, in a studypublishedin 2010, potential clinical benefit was seen in up to20%ofwomenwithmetastatic breast cancer takingpart in trials of novel approaches at the Marsdenbetween2002and2009 (Br J Cancer103:607–612).

Thoughphase I trials for cancer drugs inevitablyinvolve only those with advanced cancer who haveexhausted all other options, the findings from thesetrials often have implications for treating cancer atmuch earlier stages. So for individual patients, andthe greater cause of developing better cancer drugsquicker, Kaye believes that an expansion of phase Istudies – not just at theMarsden but nationally andinternationally – is the way forward.

“First-in-man studies are very demanding. Youare having to give small numbers of patients esca-lating doses, and work out over the course of sixmonths or sowhatmight be an appropriate dose andwhat the ups anddowns of that drug could be.Hav-ing got to thatpoint, youneed tounderstandwhetherthatdrug is going tobe successful – andyoumaywellwant to treat a number of patients with a particulartumour type or particularmolecular characteristicsto understand more. For that kind of work, whereyou’ve gone beyond the first-in-man part, we needmore facilities. Cancer Research UK is setting up

networks across theUK to help with that process.”“So the distinction between phase I and phase

II studies is blurring to such an extent that con-ventional phase II studiesmay not in some cases beappropriate. Having established a significant levelof activity in an expandedphase I study, thenext stepmight be a randomised phase II study.” The days ofsingle-arm non-randomised phase II trials may benumbered, he says.

AN ACCIDENTAL ONCOLOGISTKaye is what you might call an accidental oncolo-gist. Born and bred in Leeds, his father was a Pol-ish leatherworker (his name was originallyKrakowski) who came to Britain with the Polisharmy during the last war.

Kaye ended up going tomedical school becausehewas good at biology and chemistry at school, andhisparentswantedhim to followaprofession.Thosehe studied with at Charing Cross Hospital have

“Oncology requires no specialist manual skills, so it

stands or falls on the quality of people who come into it”

Masterpiece


Vital support. Meg Morrison, of the UK Cancer ResearchCampaign (forerunner of CRUK) presents a cheque, datedJune 1988, to support the research work being carried outby Kaye and his colleagues at the University of Glasgow

remained lifelong friends. One, BobLeonard, a football and squashpal, had starteda job inoncologyatCharingCrossHospital andurgedKaye to join him, a year after qualification. Kayeknew little of oncology, but was looking for an inter-esting hospital job at that time.

Hecameunder thesupervisionofKenBagshawe,oneof the founding fathersofmedical oncology in theUK, andwas bowled over by the lifesaving potentialof chemotherapy for youngwomenwith gestationaltrophoblastic tumours, who came from all over thecountry to be treated by Bagshawe and his team.“Their recovery was remarkable, and I thought: thisis great. I knew Iwanted to be an oncologist.”

The experience has been influential throughouthis career.Oncology, hepoints out, is anewspecialty,which has changed immensely over its 40 years.Unlike specialties like surgery, oncology requires nospecialistmanual skills. “Thatmeans it standsor fallson the quality of people who come into it,” saysKaye.Hebelieves that inspiringothers to follow ispartand parcel of what oncologists should do, and likeBagshawehehas tried to openother doctors’eyes tothe potential of the specialty.

He was able to do this during 20 happy years inGlasgow,wherehe tookup thepost of senior lecturerat theUniversity’sDepartment ofClinicalOncologyin 1981, becoming professor and head of theMed-ical Oncology Department in 1985. It was a small

unit, which grew substantially over 15years, andKayewasbusy covering awiderange of clinical areas. The students anddoctorshe taught therehavebecomepartof a “Glasgow mafia” that he has main-tained contact with. One is Johann deBono,who joinedKaye at the Institute ofCancer Research eight years ago, anduncovered the exciting potential of thenew hormonal therapy abiraterone totreat cases of advanced prostate cancer.

“Therearepeopleherewhomayprovetobemore successful thanme.Oneofthe things I enjoymost is encouragingand watching young folk who areinvolved in this area, andbecauseofthe reputation of the Drug Devel-opment Unit, we get super doctorsfrom all over theworld.”Kaye’s clubbability has served him

and those he has inspired well, and it’s amatter of satisfaction tohim to seewhat such

informal networks achieved. He became involved inthe European Organisation for Research and Treat-ment of Cancer in 1980, chairing the Early ClinicalTrials Group and the Scientific Audit Committee.“Drug development then was a completely differentworld. We worked in a small group of just 12 or soEuropean colleagues,” says Kaye. Drug companiesdidn’t seemsocommitted todrugdevelopment incan-cer. “What’s changed is there’s been somuch expan-sion, so much to work with, so much to learn, thatwe’ve lost theconceptofa relatively small groupofpeo-ple working together, and EORTC doesn’t need anEarly Clinical Trials Group because all the majorcentres aredoing their own thing.”While that is awel-come development in some ways, Kaye emphasisesthat maintaining links across Europe between like-minded clinical researchers remains essential.

DRIVEN BY THE TOUGH END OF CANCERKaye’s stories of drugdevelopment often incorporatetales of patients’ faces lighting up, or their generouscomments.His relationshipwith them–whether itbe inspiring or problematic – has always been fore-most in his motivations. “My research was alwaysfrom the point of view of someonewhosemain jobwas seeing patients with cancer. If I have any kindof mission, it’s that I’ve been seeing the tough end


Masterpiece

Three generations. With his wife Anna at theirdaughter Sarah’s wedding, and (right) withhis first grandchild Hollie, December 2010

withCancer ResearchUK’s initiative to introduce“stratified medicine”, where patients with cancerhave their tissue sampled with state of the artmolecular techniques, to diagnose and target treat-ment. This is already part of what theDrugDevel-opmentUnit does, but now, with the newCentre,there’s the potential tomake their pioneering tech-niques more widely available to cancer patients.And ending “lousy” cancer treatments for good.

“What’s most exciting now is the potential fortargeted treatment on a routine basis – patients hav-ing their tumour DNA sequence recorded regu-larly,” says Kaye. “It needs expensive equipment,and the Centre for Molecular Pathology will helpus with that hardware. I’m hoping that, over thenext few years, that will become a standard proce-dure here, and if we can show the way, I think it’sfeasible that it will become available nationally.”

Masterpiece


Meet the team. All phase I trials at the Royal Marsden arecarried out in a single specialist ward by this dedicated teamof medical, nursing and support staff and research fellows

of cancer treatment for a long time, and I don’t likeit really. It’s tough. It doesn’t get easier lookingafter young patientswho are desperately looking fornew treatments because their disease is going to befatal. I think that links together with my sense ofwhat the new science offers, to drive me.”

He’s now 62 and revelling in the recent birthof his first granddaughter. His wife and eldest sonare bothGPs, and he has two other grown up chil-dren who have flown the nest (one a teacher andone an embryo filmmaker). Though he admits towanting – one day – to see a bit more of theworld than he manages to glimpse from airportsand international conference centres, he’s notbanking on retiring just yet. There are too manyplans to put in place.

Out of thosemuddy puddles between theRoyalMarsden buildings is about to spring a newCentreforMolecular Pathology – a research centre, fundedthrough Department of Health research support,that will speed up the process of introducing per-sonalised medicine in daily patient care. It links

ROYALMARSD

ENHOSPITAL

“One of the things I enjoy most is encouraging and

watching young folk who are involved in this area”

Where are the consensus guidelinesfor women with metastatic disease?New conference will tackle this neglected topic head on

� Marc Beishon

Women with advanced breast cancer can live a full and active life. But median survival is still

hovering around three years and the pace of progress is frustratingly slow.A new conference now

seeks to develop a more evidence-based approach for treating and caring for women with

metastatic disease, so they can benefit from the progress in knowledge and technology that has

done so much to improve outcomes in early breast cancer.

This year, a new, regular con-ference will convene thatwill challenge a long-stand-ing and often forgottenissue in oncology – that

there is little we can do to greatlyimprove the outcome for advancedbreast cancer. Despite the plethora ofmeetings, research and experts alreadyfocusing on breast cancer, metastaticdisease has been neglected for treat-ment and management guidelines infavour of the early stages because of itsdifficulty and complexity.That is simply not good enough, say

the clinicians behind theFirstAdvancedBreastCancer InternationalConsensusConference (ABC1), to be held in Lis-bon this November. Not only are theremany questions unanswered about how

to manage advanced disease, they say,but also the many women faced withmetastatic illness deserve amuchmorepositive, evidence-based approach, andsupport systems that lessen the all toofrequent isolation felt by people withan incurable condition.“It is metastatic disease that kills

patients so we will never cure breastcancer unless we focus much more onits advanced stage,” says Fatima Car-doso, who will co-chair and host theconference as head of the breast cancerunit at the newChampalimaudCancerCenter in Lisbon. “Already a small sub-set of thosewithmetastatic cancer showpromise for a cure if we identify andmanage them correctly,” she adds. “Forthe majority of patients, however, theaim is to improve length and, especially,

quality of life. If we could transform itinto a chronic condition it would be amajor step forward. But we cannot justgive up on aiming for awider cure – andto do that we have to convince investi-gators and the industry that it is worth-while to invest in high quality clinicaland translational research that couldlead to major gains, as we have done inearly-stage disease.”Shepoints out that results fromwork

in early-stage breast cancer are seen asmeaningful when they translate intoyears or even decades of survival, but inthe metastatic setting gains are mostlyweeks or months at best. “The mediansurvival of advanced disease hasimproved from two to three years in adecade, and that is not acceptable inmyview. But we have made far more sub-


Spotlighton...

Spotlighton...


following years, with consensus sessionsat every EBCC. “At the session onadvanced disease at the 2010 EBCCwe had 1000 people who came on thelast day of themeeting – but a fewhoursevery two years when we bring peopletogether is just not enough for what weneed to do,” says Cardoso.Now the aim is for a panel at thenew

conference to develop consensus rec-

stantial improvements in supportive andpalliative care that benefit the patient’squality of life.”The concept behind this new addi-

tion to the cancer conference calendararose from a taskforce on metastaticbreast cancer set up in2006byESOandthe European Breast Cancer Confer-ence (EBCC). In 2007 the taskforcepublished a set of recommendationstatements in The Breast (vol 16, pp 9–10) onmanagingmetastatic breast can-cer, with a view to developing detailedguidelines and supporting papers in the

ommendations that will take the publi-cation of international managementguidelines closer still.As Cardoso explains, the idea is to

establish a conference similar to the StGallen meeting held every two years inSwitzerland, which publishes a consen-sus paper on early-stage breast cancertreatment. “Studies show that countriesthat have applied the guidelines devel-oped at St Gallen have improved theirsurvival, and that’s what we want foradvanced disease too.”But she recognises that it will be a

“Survival in advanced disease has improved from two

to three years in a decade, and that is not acceptable”

challenge. “There are of course already afew national and regional guidelines formetastatic disease, but they are notwellfollowed and toomany oncologists havegiven up on the idea that they can help.”And there are limitations to existingguidelines, such as a lack of depth onspecific needs of advancedbreast cancerpatients in the light of recent knowl-edge, she adds. These limitations mustbe addressed, and doing so at an inter-national level will greatly improve thechances of the ‘sceptics’ in the cancercommunity taking guidelines seriously,Cardoso feels.The sceptics’case ismainly that there

are too many variables in advancedbreast cancer patients for guidelines tobe worthwhile, particularly after theusual first-line treatments have beenapplied. Then it becomesmore ‘art’ thanscience. Cardoso disagrees. “It is notdifferent fromearly stage disease,wherein any case you need to adapt guidelinesto the patient in front of you by, forexample, balancing the side-effectsaccording to the chances of a cure.“In metastatic disease you have to

add quality of life factors and possiblyprolongation of life, but not a cure in themajority of cases, so the balance is dif-ferent and more difficult. But if we aretalking about increasingly personalisedtreatment in the early stages,more thanever now we must also apply the sameapproach in themetastatic setting.”A marker for personalised manage-

ment in metastatic disease is the initialset of 12 statements published in TheBreast (vol 16, pp 9–10), which includesnot only brief notes on treatment optionsapplicable thenbut also theneed for psy-chosocial support, informed decision

making, quality of life assessments andenrolment in well-designed trials. Notsurprisingly, given the complexity, amulti/interdisciplinary team is crucial(and this is the first recommendation).All theseareas, andothers,needmore

research, saysCardoso. Themany ques-tionsaboutdrugs, inparticular, stemfromanother major obstacle to progress. Asshe comments, “In drug development,industry and the cooperative researchgroups tend to see themetastatic researchsetting only as a bridge to reach the adju-vant stage. This often leaves importantmanagement questions for metastaticbreast cancer patients unanswered.”It means that even after many years

oncologists still have doubts onwhetherto use certain drugs in sequence or incombination, how many lines to con-

sider,whethermaintenance therapy is anoption, and so on. With trials linkedmainly to a particular single use of adrug, there are huge problems gettingfunding to do more complex trials, sheadds. “I can understand that companiesare not interested in supporting thiswork, but it’s harder to accept that eventhe cooperative groups, which shouldfocus on academic research, are unwill-ing to do the trials.”The ESO taskforce has, however,

nowpublished several review papers onthe available data, for example on com-bination versus sequential single-agentchemotherapy, and on a patient subsetwho potentially have a chance of a curebecause they have only one or a fewmetastatic lesions, usually on one organ.But as Cardoso adds, these papers

also serve to identify much moreresearch that needs to done, such asquality of life andpredictive factorswhenusing chemotherapy regimens, and inthe ‘curative’ paper, crucial questionssuch as whether to remove a primarytumour in a patient withmetastatic dis-ease.A study addressing the latter issueis currently underway in the US withacademic funding, and was initially setup as a cooperative study between USand the rest of the world (under theBreast InternationalGroup). “But it hasbeen impossible so far to obtain thefunds – andperhaps also thewillingness– to run this purely academic trial out-side the US,” she says.“Another reason we need better


Spotlighton...

“Countries using St Gallen guidelines have improved

survival – that’s what we want for advanced disease too”

A life worth living. The needs of women living withadvanced breast cancer – for a longer life and abetter quality of life – have tended to be over-shadowed by the heavy focus on early disease

A lack of participation intrialswashighlightedby anproject alliedto the MBCAdvocacy Working Group,the Bridge survey of 950 women in ninecountries withmetastatic disease.In Lisbon, national advocacy groups

suchas theBreastCancerNetworkAus-tralia and AdvancedBC.org in the USwill behighlightingwork theyaredoing tosupport those livingwithmetastaticbreastcancer.AdvancedBC.org is runbyMusaMayer, an advocate who has writtenextensivelyonbreast cancer andhasbeenapioneer in raisingawarenessofadvanceddisease. In a paper published last year(Seminars in Oncology Nursing 26:195–202), she examines key lessons learnedfrom surveys of need, such as theBridgesurvey. “Patients and familieswant, need

guidelines is that we can now offer pro-cedures we couldn’t do before, such asstereotactic radiotherapy onbrainmetastases,” addsCar-doso. “Technology like this ischanging howwecanmanagethese patients.”Alongwith a lack of interest

by industry and someclinicians,Cardoso notes that until a fewyears ago the patient advocacygroups were not the force theycouldhave been for advanceddis-ease. As she writes in an editorialinThe Breast (vol 18, pp271–271),patient groups have focusedmainlyon the prevention, diagnosis andtreatment of early breast cancer,which is “totally understandable”given the larger number of womeninvolved and the difficulty of con-fronting the invariably fatal side of thedisease in the advanced stage. But thishas left many women with metastaticdisease isolated as “forgotten heroes”, asshe puts it.Several initiatives are helping to

change this. Stella Kyriakides fromCyprus Europa Donna has been on theESO taskforce from the start, while aninternational group formetastatic breastcancer has been set up, theMBCAdvo-cacy Working Group, which has pub-lished its own consensus report(‘Bridging gaps, expanding outreach’ –The Breast 18:273–275). This briefreport identifies three priorities: improv-ing access to information, resources andsupport services; raising the profile ofmetastatic disease within the widerbreast cancer community and with thepublic; and increasing understanding ofand access to clinical trials.

and deserve better services,” she says.In the leadup to theNovembercon-

sensus meeting, Cardoso says morework is being done on themes such aswhether it ishelpful todetectmetasta-tic diseasebefore it becomes sympto-matic, andhowto followupand treatpatients, given that tests can bedemanding and time consuming.“Wearealso lookingmoreat the roleofmaintenance therapy, the num-ber of treatment lines to give andwewould like todomuchmoreonpsychosocial support for patientsand their families.”As she concludes, the pio-

neers in early-stage breast cancer hadtheir sceptics too. “Just look at GianniBonadonna’sworkonadjuvantchemother-apy in the1970s–half the scientificworlddid not believe it would work,” she says.“Ourworknowmayseemverydifficultbutit doesn’t scareme.”

Spotlighton...


� A webcast of the workshop onmetastatic breast cancer guidelinesat the2010EBCCisat tinyurl.com/32txp8c (on the ECCOwebsite).

� Both theMBCAdvocacyWorkingGroup and the Bridge survey aresupported by Pfizer Oncology. Seewww.bridgembc.com for the con-sensus report and survey and alsoPfizer’smedia room formore infor-mation at tinyurl.com/32wdx7m

See you in Lisbon. Up until now, discussionsabout treatment of advanced breast cancerhave had to be slotted into a single sessionat broader conferences

Advanced

Breast Cancer

3-5 November

2011 � Lisbon,Portuga

l

First Consensus Conference

FURTHER INFORM

ATION AVAILABLE AT

www.ABC-lisbo

n.net

dedicated med

ia partner

exclusive partne

r journal

A B 1C

B r i d gi n g t h e G a p

Chairs:

F.Cardoso, PT � L.Norton,US � E.Winer,US � A.Costa, IT/CH

“We can now offer procedures we couldn’t do before,

such as stereotactic radiotherapy on brain metastases”

ImpactFactor

Failure of bevacizumabin early-stage colon cancer

� Daniel Sargent

A randomised phase III trial of patients with stage II and III colon cancer showed no benefit of

adding bevacizumab to standard adjuvant oxaliplatin plus fluorouracil and leucovorin. Despite

suggestive evidence of a short-term benefit, these data and other similar findings dictate that

adjuvant bevacizumab should not be used in colon cancer.

Thefirst definitive evidence of clin-ical benefit of bevacizumab inmetastatic colon cancer was

reported in 2001.1 Since then, this agenthas become a standard component ofthe treatment of multiple tumour typesin the setting of advanced cancer.Although bevacizumab has producedvariable success across disease entities,its clear activity in patientswith stage IVcolon cancer logicallywarranted an eval-uation of its efficacy in patients withearlier-stage disease. Adjuvant therapywith fluorouracil-based regimens fol-lowing surgical resection of stage IIIcolon cancer has been the standard ofcare for approximately 20 years;2 in 2003the addition of oxaliplatin to fluorouraciland leucovorin (a combination calledFOLFOX) became the current standardof care in adjuvant colon cancer.3 Now,Allegra et al.4 report the first trial testing

the addition of bevacizumab to theFOLFOX regimen for patients withstage II and III colon cancer.

In the randomised,multicentre trialof 2672 evaluable patients, no benefitwas observed for the addition of beva-cizumab to standard FOLFOX for theprimary endpoint of disease-free sur-vival.4 The endpoint chosen was appro-priate, as results based on disease-freesurvival have been demonstrated to behighly predictive of later overall survivalfindings.5 The overall trial results aredefinitive; the possible benefit suggestedby the hazard ratio of 0.89 is clearly non-significant (P=0.15).Anymodest possi-ble benefit in the initial phase of the trialwas attenuated over time; in the periodfrom 2 to 3.5 years of follow-up, therecurrence rate in the bevacizumab armwas in fact higher than in the controlarm. In addition, the results of a second

trial of bevacizumab in the setting ofstage III colon cancer were announcedin September 2010.6 In that trial, theresults at three years numericallyfavoured the control arm (FOLFOXalone). Thus, on the basis of these twotrials, bevacizumab is clearly not rec-ommended for the adjuvant treatment ofstage II and III colon cancer.

In an exploratory analysis,Allegra etal.4 identified a possible short-term dis-ease-free survival benefit of bevacizumabin the first 15 months following ran-domisation. This finding is intriguing,given that the bevacizumab treatmentwas administered for 12months.Owingto the potential bias induced by unequaltime to imaging between the two arms ofthe trial (imaging frequencywas not pro-tocol mandated), a sensitivity analysiswas conducted attempting to adjust forthe differential time to first imaging that


This article was first published inNature Reviews Clinical Oncology 2011 vol.8 no.1, and is published withpermission. © 2011Nature Publishing Group. doi:10.1038/nrclinonc.2010.205, www.nature.com/nrclinonc

ImpactFactor

was observed. In this analysis, the short-term benefit of bevacizumab remained,with some attenuation of the first 15month hazard ratio, from0.61 to 0.71.Allegra etal.4 conclude that, allfactors considered,this finding likely rep-resents a true biologicaleffect of bevacizumab in reducingrecurrence risk while it is beingadministered. I agree with thisconclusion that indeed there likelyis a short-term benefit while beva-cizumab is being delivered. Whetherlonger-term bevacizumab exposurewould further delay recurrence, eventu-ally eradicate tumour cells and thus pre-vent recurrence, or haveno further effectcan only be tested through a subsequentrandomised trial. However, given thatmost stage II and III patients are curedby surgery alone, and considering theadverse effects and inconvenience asso-ciatedwith bevacizumab and the cost, itis unclear whether such a trial couldsucceed (or even be appropriate) in thestage III setting. A trial of extended-duration bevacizumab in the alternativesetting of maintenance therapy forpatients with resected stage IV disease,where the recurrence risk ismuchhigher(50%–70% risk of recurrence withintwo years) may be a more promisingalternative.

On the basis of the results reportedby Allegra et al.,4 FOLFOX followingsurgical resection remains the standardof care for stage III colon cancer. Thisfinding is clearly disappointing, as it rep-resents the third agent with demon-strated activity in stage IV disease thathas failed to improve outcomes in earlier-stage disease. Specifically, in addition tobevacizumab, the proven activity of bothirinotecan and cetuximab in patients

with advanced disease has not trans-lated into benefit in patients with stageIII disease.7,8As such, it seems the stan-dard paradigm for drug development in

this setting is broken:activity in metastaticdisease is not a reli-able predictor of adju-vant therapy benefit.

The biological reasonsfor this discordance arethe subject of intensediscussion, possibilities

include a different biologybetween existing visible versus

micrometastatic disease (including theconcept of epithelial–mesenchymaltransition of tumour cells) and the pres-ence of therapy-resistant cancer stemcells. However, the fact remains thatwe currently do not have an accuratepredictor of efficacy for a newproposedadjuvant therapy. A possible alternativeapproach, the use of neoadjuvantchemotherapy in early-stage disease,seemsworthy of exploration.Rectal can-cer therapy has moved primarily to theneoadjuvant paradigm (as has muchresearch in breast cancer). The ability totest a therapy’s impact in an intacttumour (as well as obtaining pre-treat-ment andpost-treatment biospecimens)is very attractive.Clinical trials of neoad-juvant therapy for colon cancer are ongo-ing.9 In the setting of stage II disease,given the high cure rate (approximately20% recurrence risk), verymodest ben-efit of fluorouracil10 and no benefit ofoxaliplatin,3 research is focusedon strate-gies for risk assessment to identifypatients who are at high risk of recur-rence and thus may be considered foradjuvant therapy.

Even if the paradigm of advanceddisease testing before adjuvant trialsremained appropriate, at the present

time there is a dearth of agents in later-stage (phase III) testing in advancedcolon cancer.Currently, themost press-ing adjuvant therapy question seems tobe that of the optimal duration oftherapy. On the basis of the cumulativeneurological toxic effects of oxaliplatin,reducing the treatment time to threemonths (from the current six months)would be highly advantageous. Thisquestion is being tested in four ongoingtrials and one planned randomised trialthat are being conducted throughouttheworld (including theTOSCA trial inItaly, the SCOT trial headquartered inthe UK, the C80702 trial in the USAand the PRODIGE/GERCOR trial inFrance). These trials have prospectivelyagreed to pool their data for a definitivenoninferiority analysis with at least10,500 patients through the Interna-tional Duration Evaluation of AdjuvantChemotherapy (IDEA) collaboration.

In conclusion, the primary clinicalimplications of the study byAllegra et al.4

are clear – bevacizumab should not beused in the adjuvant setting in coloncancer outside clinical trials – and atthe same time raise many new ques-tions of how to best develop new agentsbefore adjuvant testing. Innovativestrategies are needed to assess newagents and treatment strategies, as coloncancer remains amajor cause of cancerdeath worldwide.

Details of the references cited in this article can be

accessed at www.cancerworld.org


Practice pointThe use of adjuvant bevacizumab inthe settingof stage II or III coloncanceris not recommended.

Author affiliations: Division of Biomedical Statistics and Informatics, Mayo Clinic Cancer Center, Rochester, Michigan. Acknowledgements: The author would like to thank Axel Grotheyfor helpful discussions. Competing interests: The author declares that he has previously worked as a consultant for Genentech

ImpactFactor

Ultra-targeted accelerated partialbreast irradiation usingTARGIT – a cautionary note

� Rajiv Sarin

Oneof the seven ongoing trials of accelerated partial breast irradiation (APBI) has concluded that

single-dose intraoperative radiotherapy should be considered as an alternative to protractedwhole-

breast irradiation.With amedian followup of two years, such conclusions seempremature.Until

the risk andpattern of recurrence is reported at longer followup,APBI should remain experimental.

Accelerated partial breast irradi-ation (APBI) is an abbreviatedradiotherapy course delivered

to the tissue surrounding the excisioncavity. It is under intense clinical inves-tigation as a therapeutic approach forlow-risk early-stage breast cancers.1

Phase II studies of APBI using multi-catheter brachytherapy or external-beam radiotherapy show highlocal-control rates at 7–12 years

median follow up.2–4 These studieshighlight the long natural history ofthe low-risk early-stage breast cancersfor which APBI is being proposed. Ittherefore seemed prudent to wait forthe 5–10 year follow-up data on~16,000 women enrolled in seven ran-domised controlled trials comparingAPBI with whole-breast irradiation2 tomake definitive conclusions on thesafety and efficacy of APBI and to

establish the clinical, pathological ortechnical contexts where cautionshould be exercised with differentAPBI techniques.2 However, the inves-tigators of one of these trials, theTARGIT trial, thought otherwise.

In their report,5 the TARGIT inves-tigators state that their trial “providesrobust andmature evidence... showingthat targeted intraoperative radiother-apy is safe,” and concluded that, “for


This article was first published inNature Reviews Clinical Oncology 2010 vol.7 no.12, and is published withpermission. © 2010Nature Publishing Group. doi:10.1038/nrclinonc.2010.187, www.nature.com/nrclinonc

ImpactFactor

selected patients with early breast can-cer, a single dose of radiotherapy deliv-ered at the time of surgery by use oftargeted intraoperative radiotherapyshould be considered as an alternativeto external beam radiotherapy deliv-ered over several weeks.” In the accom-panying commentary, Azria andBourgier6 are convinced that, for elderlypatients, “APBI is the new standardand TARGIT is an excellent approach.”Furthermore, a lead TARGIT investi-gator has highlighted that the UKNational Health Servicewastes £4 million annuallyon homeopathy, which heargues could be used to pro-vide TARGIT.7 He goes on toquestion whether the intro-duction of TARGITAPBI willhave to wait for financialapproval while homeopathycontinues to creep under thehurdle.Against this backdrop,they argue, the oncology com-munity should take a standon endorsing the wider use ofTARGITAPBI as standard ofcare for women fulfilling theselection criteria of theTARGIT trial.While this viewis endorsed by the TARGITtriallists and other reasearchers,5–7

Iwould consider this premature becauseof issues highlighted in this article.

In the randomised, multicentreTARGIT trial, 2232 women aged ≥45years with low-risk early-stage breastcancer received intraoperativeTARGITAPBI or whole-breast irradiation for5–7 weeks.5 At its initiation in 2000,the trial aimed to publish the results in2006 with median follow up of fiveyears.8 However, accrual was highlyextended and two thirds of the patientswere recruited in the last three years ofthe decade-long accrual. For the wholecohort of randomised patients, theminimum follow-up period is not spec-

ified and the median follow up is 25months. Fewer than 20% of theenrolled patients were followed upbeyond four years before publication ofthe data. The authors have, therefore,restricted the display of local recur-rence rates to four years, which was notsignificantly different between the twoarms – local recurrence rate of 1.2%with TARGIT and 0.95% with whole-breast irradiation (P=0.41). From thisreport, it is not clear if there were anyincidences of local recurrence beyondthe four-year period.

The authors have used two argu-ments to support their definitiveconclusions, despite such shortfollow up. First they argue thatthe background five-yearbreast cancer recurrencerate in their cohort of low-risk breast cancerpatients, which theyhad projected 10years ago to be 6%and formed thebasis of recruiting

2232 women, is nowexpected to be around

1.5% (based on data from the con-trol whole-breast irradiation groupin the study) and, therefore, only

585 cases are sufficient to prove non-inferiority. In contrast, other prospec-tive APBI studies, such as theMammoSite study,9 that have dis-cussed results of a subset of theirpatients having a longer follow up thanthe whole cohort, have showed thecharacteristics and treatment variablesof this subset, and have refrained fromdrawing practice-changing conclu-sions. With a decade-long accrual inthe TARGIT trial,5 the characteristicsof patients, disease and treatment mayhave changed during the trial periodand it should be seen how representa-tive they are of the entire enrolled pop-ulation of 2232 women. Vaidya et al.5

argue that a quarter of the 2232enrolled patients are sufficient to drawconclusions on noninferiority; how-ever, no mention is made about theinternational steering committee ofthe TARGIT trial increasing the samplesize from 2232 to 3432 women inMarch 2010.10 The second argument isthat the short follow-up period coversthe peak hazard of local recurrencethat occurs between two and threeyears after surgery, allowing them todraw cautious yet reasonable conclu-sions about efficacy. Yet prospectivestudies on similar low-risk patientstreated with quality-assured APBIhave shown that actual breast cancerrecurrence rates increase with time.For example, the German–AustrianESTRO phase II trial4 that assessed273 patients with a median 63monthsfollow up showed that a negligible four-year breast cancer-recurrence rate,similar to the present TARGIT report,equated to a recurrence rate of 2.3% atfive years and 5% at eight years. More-over, in women with a nonhomoge-neous dose of radiation, the eight-yearrecurrence rate was 7.5%. If theTARGIT trial with just 212 women atrisk at four years after intraoperativeAPBI can draw reasonable conclusionson efficacy, many other ongoing orrecently concluded randomised trials ofAPBI would be better placed to drawsimilar conclusions without anyfurther wait.

When mature data are presentedfrom the TARGIT study, interpreta-tion should acknowledge that 234 outof 1113 women (21%) in the TARGITAPBI cohort received treatment in theform of mastectomy or whole-breastradiotherapy either because of protocolviolation or adverse pathology.5 Theprimary analysis has been performedon an intention-to-treat basis, as rec-ommended by the CONSORT guide-lines. However, with one in five women


ImpactFactor

in the TARGIT arm undergoing stan-dard treatment, the effect of TARGITAPBImay have been overestimated. Itwould be important to know themedian follow up in the women whoreceived TARGIT APBI withoutwhole-breast radiotherapy and thenumber and site of breast cancer recur-rences in this cohort at clinically appro-priate time points.

Based on radiobiologicalmodelling,Vaidya et al.5 suggest that the biologicaldose fromTARGIT50 kVX-rays will be20%–30% higher than the physicaldose. Assuming this is true, it meansthat at 1 cm from the excision cavity thephysical dose equivalent is only 7Gy. Ifthe long-term results of TARGIT showthat a single dose of 20 Gy at the sur-face and 5–7 Gy at a depth of 1 cm isable to control breast cancer, it wouldimply that either the volume of tissuethat needs full-dose irradiation is a fewmillimetres or that the low dose ofradiation at 1 cm, which we otherwiseconsider subtherapeutic, is sufficient tocontrol cancer. Such findings wouldhave far-reaching implications for can-cer treatment with adjuvant radiother-apy. If the low physical dose of 5 Gy at1 cmwith TARGIT is radiobiologicallya much higher dose and sufficient tocontrol cancer then it can also beexpected to correlate with late toxiceffects, especially in the 142 womenwho received full-dose conventionalwhole-breast irradiation following full-dose TARGIT. The absence of reportedincidences of fat necrosis in the Vaidyaet al.5 study is unusual for an APBIseries. Without knowing the compli-ance with annual mammography dur-ing follow up, site of recurrence and itsdistance from the lumpectomy site,the possibility of under-reporting ofbreast cancer recurrence cannot beruled out. Imaging of another spherical

device placed intraoperatively in theexcision cavity has shown that in someinstances there could be a significantgap between the surface of the applica-tor and the breast tissue.1As TARGIT isan ultra-targeted form of radiotherapywith very sharp fall off of the radiationdose and is delivered in a single sitting,placement precision and its verificationis crucial. Even a few millimetres offluid between the applicator surfaceand excision cavity wall would seri-ously compromise the absorbed dosewith 50 kV X-ray.

Trials evaluating new adjuvant ther-apy in early-stage breast cancer refrainfrom reporting results at very shortmedian follow up, without clearly indi-cating them as interim results, anddo not make practice-changing con-clusions at these interim time points.Leading publications with definitivetherapeutic recommendations based onselective discussion on a subgroup ofless than 20% of patients with a four-year follow up is a new phenomenon inearly-stage breast cancer. This requiresall triallists, reviewers and editors totake a clear stand. I fear that thepremature report5 with definitive con-clusions, accompanying supportivecommentary6 and correspondence in aleadingmedical journal, and associatedmedia coverage, may trigger a race toreport the remaining six APBI trialsprematurely.

Most of these trials already havemuch longermedian follow up than theTARGIT trial and if the statistical ration-ale proposed for early reporting anddrawing definitive conclusions areaccepted for the TARGIT trial, theywill be applicable to almost all theremaining trials.

There is no doubt that APBI isapproaching a very exciting phase andhas real potential to offer safe, conven-

ient and cost-effective breast conserva-tion in the coming decade. But whendealingwith a low-risk diseasewith longnatural history, premature conclusionscan sometimes be counterproductive.

References

1. R Sarin (2005) Partial-breast treatment for

early breast cancer: emergence of a new

paradigm. Nat Clin Pract Oncol 2:40–47

2. M Mannino and J Yarnold (2009) Accelerated

partial breast irradiation trials: diversity in

rationale and design. Radiother Oncol 91:16–22

3. BV Offersen, M Overgaard, N Kroman et al.

(2009) Accelerated partial breast irradiation

as part of breast conserving therapy of early

breast carcinoma: a systematic review.

Radiother Oncol 90:1–13

4. V Strnad et al. (2010) Accelerated partial

breast irradiation: 5-year results of the

German–Austrian multicenter phase II trial

using interstitial multicatheter brachytherapy

alone after breast-conserving surgery. Int J

Radiat Oncol Biol Phys.

doi: 10.1016/ijrobp.2010.01.020

5. JS Vaidya et al. (2010) Targeted

intraoperative radiotherapy versus whole breast

radiotherapy for breast cancer (TARGIT-A trial):

an international, prospective, randomised, non-

inferiority phase 3 trial. Lancet 376:91–102

6. D Azria and C Bourgier. (2010) Partial breast

irradiation: new standard for selected patients.

Lancet 376:71–72

7. M Baum. (2010) Homoeopathy waives the

rules. Lancet 376:577

8. Lancet. (2010) Protocol 99PRT/47 targeted

intraoperative radiotherapy (Targit) for breast

cancer. www.thelancet.com/protocol-

reviews/99PRT-47

9. JC Nelson et al. (2009) Four-year clinical

update from the American Society of Breast

Surgeons MammoSite brachytherapy trial. Am J

Surg 198:83–91

10. NIHR Health Technology Assessment

Programme. (2010) TARGIT protocol

version 5.0, April 2010. http://www.hta.ac.uk/

protocols/200700600049.pdf

Author affiliations: Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Centre, Kharghar, Navi Mumbai, India


ImpactFactor


N E W S R O U N DSelec ted repo r t s ed i t ed by Jane t F r i cke r

Low-dose aspirinreduces death fromseveral cancers� The Lancet

Daily aspirin reduced deaths due to anycancer by 20%, reports a meta-analysis

study, but the benefits only really becameapparent after patients had been taking thedrug for five years or more.

While earlier studies have suggested thatlong-term aspirin therapymay protect againstcolon cancer, the current study led by PeterRothwell, of the Department of Clinical Neu-rology at Oxford University, is the first toshow that aspirin protects against other can-cers, such as oesophageal, gastrointestinal,lung, brain and pancreatic cancers.

Several lines of evidence have suggestedthat long-term use of aspirinmight reduce therisk of some cancers, particularly gastro-intestinal tumours. In animal models aspirinreduces incidence or growth rate or both ofseveral cancers, mediated by inhibition of thecyclo-oxygenase enzymes and production ofprostaglandins and other inflammatorymedi-ators. Observational studies in humans havealso suggested that aspirin reduces the risk of

three UK-based trials that had continued toobtain data for deaths due to cancer aftercompletion of the trials via the national deathcertification and cancer registration systems.

At 20 years follow-up the three trialsshowed that the risk of cancer deathremained 20% lower for all solid cancers(P<0.0001) and 35% lower for gastrointesti-nal cancers (P<0.0001) among the partici-pants taking aspirin. When the fall in risk ofdeath was broken down according to indi-vidual types of cancer, it was 60% foroesophageal cancers, 40% for colorectal can-cer, 30% for lung cancer and 10% for prostatecancer. Reductions in pancreatic, stomachand brain cancers were difficult to quantifydue to the small number of deaths. Takinglarger doses of aspirin and smoking and gen-der had no effect on the results.

The authors, from Oxford, Edinburgh andJapan, conclude, “These findings provide thefirst proof in man that aspirin reduces deathsdue to several common cancers. Benefit wasconsistent across the different trial popula-tions, suggesting that the findings are likely tobe generalisable.”

Whenweighing up the risk and benefits oftaking aspirin, they add, clinicians will nowneed to consider the protective effects againstcancer. “Although the reduction in risk of

certain cancers. In an earlier study, publishedin the Lancet in October 2010, Rothwell andcolleagues showed that long-term low-doseaspirin (75 mg per day) reduced death ratesfrom colorectal cancer bymore than a third. Inthe current research the team studied deathsdue to all cancers.

The meta-analysis identified eight ran-domised trials of daily aspirin versus no aspirin,including 25,570 patients, that had originallybeen undertaken to look at primary or sec-ondary prevention of vascular events. Doses ofaspirin in the eight trials ranged from 75 mgto 500 mg per day. Altogether 674 patientsdied from cancer in the course of the studies.

Results showed that during the periodof the clinical trials, which lasted for aboutfour years, allocation to the aspirin groupreduced the risk of death from cancer by 21%(P=0.003).

On analysis of individual patient data,available from seven trials involving 23,535patients, it became apparent that benefitsincreasedwith time. After five years the risk ofall cancers was reduced by 34% (P=0.003)and the risk of gastrointestinal cancers by54% (P=0.003).

The researchers also wanted to determinewhether the benefits from aspirin continuedover time, and this was made possible by the

ImpactFactor


ischaemic vascular events on aspirin in healthyindividuals is partly offset by a small increasein risk of non-fatal bleeding complications, thebalance of risk and benefit will now be alteredby the reduction in cancer deaths after fiveyears’ treatment. Our analyses show that tak-ing aspirin daily for 5 to 10 years would reduceall-causemortality (including any fatal bleeds)during that time by about 10%.”

Limitations of the study, write the authors,include the fact that it only used trials ofdaily aspirin, that too few women had beenrecruited to allow the investigators to deter-mine the effects of aspirin on gynaecologicalcancers, and that they were unable to deter-mine the effect of continued aspirin use after20 years.

The next step, say the authors, will beto explore whether there is any protectiveeffect of aspirin on the incidence or progres-sion of cancer.� P Rothwell, F Fowkes, J Belch et al. Effect of

daily aspirin on long-term risk of death due to

cancer: analysis of individual patient data from

randomised trials. Lancet 7 December 2010,

doi:10.1016/S0140-6736(10)62110-1

Radiation therapy forhead and neck cancerleads to hearing loss� Arch Otolaryngol Head Neck Surg

Patients who have undergone radiotherapyfor head and neck cancers experience a

higher incidence of hearing loss and moresevere hearing handicaps than age-matchedcontrols who do not have cancer, reports aBrazilian study.

Treatments for head and neck cancer,including surgery, chemotherapy and radio-therapy, either alone or in combination, are allknown to affect the auditory system andcause temporary or permanent hearing loss.With radiotherapy, when the inner ear isincluded in the irradiation field, permanentsensorineural hearing loss may result from the

loss of ciliated cells in the cochlea, withlatency periods ranging from 1.5 to 5 years.

In the current study, Christiane Schultzand colleagues, from the Hospital de Câncerde Barretos (Barretos, Brazil), set out to inves-tigate hearing difficulties among 141 patientswith head and neck cancer who had under-gone radiotherapy alone or in associationwith chemotherapy or surgery. The patients,together with 141 age-matched controls (whohad never undergone oncological treatmentplacing their hearing at risk), underwent hear-ing evaluations and completed the HearingHandicap Inventory for the Elderly (HHIE)questionnaire, which assessed the effect ofhearing loss on their lives. The degree of hear-ing handicap was divided into three cate-gories according to severity.

Results show that hearing loss wasdetected in 103 (73.3%) of participantsexposed to radiation therapy versus 69(48.9%) of age-matched controls (P<0.001).Severe or profound hearing loss occurred in6.4% of right ears and 8.5% of left ears in theradiation-treated group, as compared with0.7% of right ears and 1.4% left ears of con-trol participants.

Hearing loss was mostly sensorineural(resulting from disorders or damage involvingthe nerves or the inner ear) as opposed to con-ductive (resulting from interference in soundtransmission, usually involving the outer ormiddle ear).

Furthermore, 19.1% of patients in theradiation treatment group had suffered asevere hearing handicap versus 2.8% in thecontrol group (P<0.001). “This indicates that,when present, hearing losses were substan-tially greater and more incapacitating afterthe radiotherapy,” write the authors.

There was also found to be a correlationbetween the degree of hearing loss and scoreon the HHIE questionnaire, with participantswhose hearing loss went untreated beingmore likely to report feeling lonely, depressed,worried, anxious or paranoid, and to havefewer social activities and be less able toprocess information about their environments.

“This is extremely important because

behavioural patterns that are more depressiveor that present greater tendencies for socialisolation can sometimes be attributed to thecancer or to the functional sequelae of thetreatment. Nonetheless, one must rememberthat hearing loss and hearing handicap mayalso lead to such behaviour,” write the authors.

They conclude that in order to enablebetter rehabilitation of patients with headand neck cancer, hearing loss should formpart of the investigations.

� S Schultz, M Goffi-Gomez, P Liberman et al.

Hearing loss and complaint in patients with head

and neck cancer treated with radiotherapy. Arch

Otolaryngol Head Neck Surg November 2010,

136:1065–1069

Combination therapyshows promise inbiliary tract cancer� Lancet Oncology

Cetuximab in combination with gem-citabine and oxaliplatin produced encour-

aging results as a first-line palliative caretreatment in patients with biliary cancer. TheAustrian single-centre study found that theaddition of cetuximab was associated withincreased response, substantial tumour shrink-age and the potential for secondary resection.

Patients with biliary tract cancer have apoor prognosis, with an overall survival that isless than 15% at five years. The only curativetreatment is surgical resection, but even aftersurgical resection recurrence is frequentlyreported and until recently no standard pal-liative chemotherapy had been defined. Twophase III trials have recently shown that gem-citabine plus cisplatin (GEMCIS) or oxaliplatin(GEMOX) are superior in terms of overall sur-vival to gemcitabine alone.

The current phase II study by Birgit Gruen-berger and colleagues, from BarmherzigerBrüder Hospital Vienna, Austria, set out toinvestigate the efficacy and safety of adding

ImpactFactor


cetuximab to the GEMOX combination. Cetux-imab is a targeted therapy directed against theepithelial growth factor receptor (EGFR),which has been associated with improvedoutcome for malignancies including colorec-tal, lung and head and neck cancer.

Between October 2006 and July 2008, 30patients with unresectable biliary tract cancerwere enrolled from one centre in Austria. Allpatients received 500 mg/m2 cetuximab as atwo-hour intravenous infusion on day 1, and100 mg/m2 oxaliplatin on day 2, every twoweeks for 12 cycles. The primary outcomewas overall response rate.

The investigators found that 19 (63%) ofthe patients experienced objective response –three (10%) achieved complete response and16 (53%) achieved partial response. Followingmajor response to therapy, nine patients (30%)were able to undergo secondary curative resec-tion. Of this subgroup, five had intrahepaticcholangiocarcinoma that had initially not beenamenable to secondary resection, and fourpresented with locally advanced extrahepatictumours that had been unresectable due tovascular involvement. Grade 3 adverse eventssuch as skin rash, peripheral neuropathy andthrombocytopenia occurred in 13 patients,but none reported grade 4 events.

The authors write that comparisons ofthese results with response rates achieved inother studies verified that cetuximab plusGEMOX has a better overall response ratethan gemcitabine alone, GEMOX alone, orother chemotherapy combinations.

Following findings in studies of metasta-tic colorectal cancer, the association betweenKRAS mutation status and response to cetux-imab was also investigated. KRAS mutationswere detected in three out of 30 patients,but did not appear to preclude benefit fromcombined cetuximab and GEMOX. Patientswith KRAS mutated tumours had a shortermedian survival than did those with wild-type tumours (1.67 vs 7.67 months, P=0.071).

The authors conclude, “This combinationtreatment had an acceptable toxicity profileand resulted in potentially curative secondaryresection in a third of patients, which signif-

University of Bologna, Italy, concluded thattriple therapy induction represents a newstandard of care for patients with multiplemyeloma eligible for transplant.

Thalidomide, bortezomib and lenalido-mide have greatly advanced myeloma treat-ment during the past decade, with thethalidomide plus dexamethasone inductionregimen (TD) showing the highest activityand being acknowledged as the new standardof care for induction therapy in the US. How-ever, small studies have suggested that theaddition of bortezomib to TD (VTD) may resultin increased rates of high-quality responsesfor all phases of myeloma.

Cavo and colleagues undertook a phase IIIstudy to assess the efficacy and safety of VTDversus TD as induction therapy in preparationfor double autologous stem-cell transplanta-tion in newly diagnosed multiple myelomapatients. Altogether 480 patients aged 18 to65 from 73 sites in Italy with previouslyuntreated symptomatic myeloma were ran-domly assigned to receive VTD (n=241) or TD(n=239). The intention to treat analysisincluded 236 patients in the VTD arm and238 in the TD arm.

Results showed that 31% of patients inthe VTD arm achieved complete or near-com-plete response comparedwith 11%of patientsin the TD arm (P<0.0001). Furthermore, themedian time to best complete response was9 months in the VTD arm versus 14 months inthe TD arm (P<0.0001).

Patients in the VTD arm had a 29% three-year probability of progression or relapse com-pared with 39% for patients in the TD arm(P=0.0061).

However, on the down side, 56% (n=132)of patients in the VTD arm experienced grade3 or 4 adverse events compared with 33%(n=79) in the TD arm (P<0.000). Additionally,10% of patients on VTD experienced periph-eral neuropathy compared with 2% on TD(P=0.0004). The study showed no significantdifferences in stem-cell collection betweenthe two arms.

Despite the high levels of peripheral neu-ropathy, only one patient experienced grade 4

icantly lengthened progression-free survival.These findings provide justification for furtherstudies of this treatment combination in arandomised study of a large cohort.”

In an accompanying commentary, DavidMalka, Valérie Bogie and Michel Ducreux,from the Institut Gustave Roussy (Villejuif,France), cautioned that small, single-centrestudies can be prone to selection bias, givingthe example that 97% of patients in the cur-rent study had an Eastern Cooperative Oncol-ogy Group performance status of 0–1.Furthermore, they added, several previousreports have found KRAS mutation rates inbiliary cancers to be higher than recorded inthe current study. “Hence, data from largeprospective cohorts are needed to specify theactual prevalence of KRAS mutations – andBRAF mutations . . . and establish whetherthese mutations are predictive for inefficacyof anti-EGFR antibodies in patients withadvanced biliary cancers.”

� B Gruenberger, J Schueller, U Heubrandtner

et al. Cetuximab, gemcitabine, and oxaliplatin in

patients with unresectable advanced or

metastatic biliary tract cancer: a phase 2 study.

Lancet Oncol December 2010, 11:1142–1148

� D Malka, V Boige and M Ducreux. Biliary

cancers, chemotherapy, and cetuximab. ibid

pp 1110–1111

New standardof care defined inmultiple myeloma� The Lancet

The addition of bortezomib to standardinduction therapy (thalidomide plus dexa-

methasone) prior to double autologous stemcell transplantation in patients with newlydiagnosed multiple myeloma improved therate of complete or near-complete responsealmost three fold, a landmark study from theGIMEMA Myeloma Network has found. Theresearchers, led by Michele Cavo from the

ImpactFactor


peripheral neuropathy and only two patientsdiscontinued treatment due to toxic effects.

“Induction therapy with VTD was associ-ated with a significantly higher rate of com-plete or near complete response than wasinduction therapy with TD. Therefore, VTDrepresents a new standard of care tomaximisethe degree and speed of tumour reduction inpatients with myeloma who are eligible fortransplant,” conclude the authors.

Commenting on the finding that no dif-ference in overall survival was found betweenthe two groups, the authors speculate that thefollow-up period may have been too short todetect differences, that the sample size mayhave been too small and that the increasingavailability of effective treatments at time ofrelapse may have confounded any meaning-ful analysis of studies in first-line treatment.

In an accompanying commentary, PaulRichardson, from Harvard Medical School,writes, “The unprecedented high quality ofresponses engendered by these combinationswith a generally favourable safety profile bodeswell for continued benefit to patients, with yetfurther improvements in outcome still neededfor this otherwise incurable malignancy.”

The significant neurotoxicity encounteredwith VTD, he adds, contrasts with an “other-wise promising picture”. Strategies to reducetoxicity include use of less neurotoxic butactive combinations of drugs such as lenalido-mide and dexamethasone, or lowering dosesof bortezomib and thalidomide.

The question of whether additional drugsshould be added to the three-drug inductionstrategy (such as monoclonal antibodies, his-tone deacetylase inhibitors, heat shock pro-tein-90 inhibitors) also requires furtherconsideration, Richardson writes.

A study is currently underway by theEuropean Myeloma Network to addresswhether novel agents might delay or chal-lenge the need for autologous stem cell trans-plantation in myeloma.

� M Cavo, P Tacchhetti, F Patriarca et al.

Bortezomib with thalidomide plus dexamethasone

compared with thalidomide plus dexamethasone as

induction therapy before, and consolidation therapy

after, double autologous stem-cell transplantation in

newly diagnosed multiple myeloma: a randomised

phase 3 study. Lancet 18 December 2010,

376:2075–2085

� P Richardson. A new standard of care in newly

diagnosed multiple myeloma. ibid pp 2043–2044

Docetaxel sets newstandard of care inoperable, high-risk node-negative breast cancer� New England Journal of Medicine

For women with operable, high-risk node-negative, early-stage breast cancer, adju-

vant treatment with the drug combination ofdocetaxel, doxorubicin and cyclophosphamide(TAC) reduced the risk of recurrence by 32%compared with the traditional treatmentapproach of fluorouracil, doxorubicin andcyclophosphamide (FAC), reports a study fromthe Spanish Breast Cancer Research Group(GEICAM).

“Our findings. . . show that TAC is effectiveboth in patients with node-positive and inthose with high-risk node-negative early-stage breast cancer,” write the authors, led byMiguel Martin from the Hospital General Uni-versitario Gregorio Marañón in Madrid.

Although adjuvant taxane-based regi-mens are now the standard of care for patientswith node-positive early-stage breast can-cer, their efficacy in patients with high-risknode-negative disease has not been defined.The benefits of adjuvant chemotherapy inthe node-negative populations have, however,been well established.

In the open-label phase III GEICAM 9805study, between June 1999 and March 2003,1060 women with axillary-node-negativebreast cancer and at least one high-risk factorfor recurrence were randomised to treatmentwith either TAC (n=539) or FAC (n=521) for sixcycles every three weeks, following surgery.The studywas funded by Sanofi-Aventis (mak-

ers of docetaxel) and involved 40 centres inSpain, four in Germany and two in Poland.

Results show that at a median follow-up of 77 months, 87.8% of the women werealive and disease-free in the TAC group com-pared to 81.8% in the FAC group (HR 0.68,95% CI 0.49–0.93; P=0.01). The benefit wasconsistent across subgroups regardless ofhormone receptor status, HER2 status,menopausal status, age, tumour size or his-tologic grade. On the basis of the trial, thenumber of patients who would need to betreated to prevent recurrence in one patientis 17, write the authors.

The difference in survival – 95.2% for TACand 93.5% for FAC – was not significant(HR 0.76; 95% CI 0.45–1.26; P=0.29). This,the authors suggest, was because the numberof deaths was small (26 vs 34). The rates ofgrade 3 or 4 adverse events were 28.2% withTAC and 17.0% with FAC (P<0.0001), and seri-ous adverse events were recorded in 4.7% ofthe women in the TAC group compared to0.8% in the FAC group.

“The small number of deaths occurring atthe time of the analysis indicates that alonger follow-up period will be needed toassess survival among patients with node-negative breast cancer as compared withthose who have node-positive disease,” writethe authors.

“The GEICAM 9805 trial shows the effec-tiveness of an adjuvant taxane-based regimenover a non-taxane-based regimen in a popu-lation of patients with axillary, lymph-node-negative, early-stage breast cancer,” theyconclude, adding that the acute toxic effectsassociated with TAC are manageable whentreatment is combined with granulocytecolony-stimulating factor introduced as pri-mary prophylaxis.

The benefit of TAC over FAC in pre-menopausal women, speculate the authors,may be partly due to its ability to induceamenorrhoea in more women.

� M Martín, M Seguí, A Antó et al. Adjuvant

docetaxel for high-risk, node-negative breast

cancer.NEJM 2 December 2010, 363:2200–2210

An advanced oncology degreefor busy specialists across the globe� Peter McIntyre

Oncologists who want to develop and work to their full potential can find it hard to balance

the competing priorities of clinical practice and academic study. For some of them, wherever

in the world they may be based, this new online MSc in Advanced Oncology run by the

University of Ulm may be the answer.

It is always early morning some-where in the world! And some-where a busy cancer doctor orscientist is up extra early and log-ging on to a site in Germany

before they head off to their own clinicor laboratory. As they drink their tea orcoffee, they view a lecture by one of theexperts from theUniversity of Ulm andthey look over their notes as they take thebus or train to work.These are the first students tounder-

take an online Master of Science inAdvancedOncology–nowwell into theirsecondsemester at theUniversityofUlm,butphysically somewhereelsecompletely.The 18 students are doctors or scien-

tists specialising inhaematologyandoncol-ogy and working in cancer centres fromBrazil, Egypt, Iraq, Moldova, Romania,Nigeria and South Africa, to Germany,Italy and USA. Numbered among themare some scientists working for pharma-ceutical companies who need to learn

more about cancer therapy and trials.The two-year course is run from the

InternationalCenter forAdvancedStud-ies in Health Sciences and Services(ICAS) at Ulm University. MedicaldirectorManuela Bergmann, an oncol-ogist and haematologist, designed thecourse to fill a gap in postgraduate train-ing, to “export knowledge” and to helpstudents become leaders of cancer serv-ices in their own countries.“It is a study programme for people

who are very engaged in their professionand are all experts. Themajor problem insuch a group is timemanagement. Theyneed to be up to date but they cannotafford to go to seminars andbe out of theoffice formore than oneweek. So far asI know, no country in the world offers acurriculum online where you get anorganised structure with the latest levelof knowledge.We tried to find a solutionfor timemanagement and a structure fororganised knowledge.”

Each semester has 60–70 lectures,which the students follow at their ownpace. In addition they attend one-weekseminars (‘summer schools’) five timeseach year, where theymeet their lectur-ers and fellow students.There are six modules in all. The

four onlinemodules cover interdiscipli-nary oncology (including cellular andmolecular biology), clinical research(including ethical aspects andmanage-ment of trials), advanced therapies andmanagement. Themanagementmoduleprepares doctors to run a department,using lecturers from the Institute ofBusiness at Ulm University and fromMcKinsey consultants. Katarina Janus,professor of healthcare management atUlmand a pioneer of research intowhatmotivates doctors to work most effec-tively, also teaches on this module.The lectures themselves are interac-

tive – with opportunities to leave ques-tions for the lecturer and get a response,


Systems&Services


tology, is the youngest at just 29years old.“They have been encouraging older

people inhigherpositions todo thecourse.I was amazed and really happy that I wasaccepted,” he says.Ahmedwas still moredelighted to be awarded an ESObursary.“I would not be able to afford this courseif I were not supported byESO. They arepaying foreverything, thecourse, travelling,accommodation, everything.”“I believewe can change thewaywe

treat the patients with a better out-come if we have been taught well with

Systems&Services

UNIVERSITIYOFULM

The class of 2010. The first intake of 18doctors and scientists come from 10 countriesacross the globe, and are pictured here at oneof the week-long seminars that supplement thepredominantly online course work

usually within 24 hours. There is also alink up with the European School ofOncology (ESO) – which is also thepublisher of Cancer World – underwhich students can attend the weeklyonline live e-grandround sessions(www.e-eso.net).At the endof their two-year course they also attend the week-longMasterclass inOncology run jointlyby ESO andESMO (European Societyfor Medical Oncology), which isdesigned for future leaders of oncology.The two-year course costs almost

€20,000 plus travel and accommoda-tion for the five attendanceweeks.Whilea few students cover their own fees,most are supported by their employers(in the case of the pharmaceutical com-panies) or by charitable foundations.ESO supports two of the 2010–2011intake andwill support one of the dozenwho will join them inOctober 2011.Most of those on the course are in

their40s.AhmedRabea, assistant lecturerat theNationalCancer Institute inCairo,wherehespecialises inmalignanthaema-

the newest modalities and try to applythat in our job. I know there is a prob-lem in Egypt with financial support,but if we are taught to conduct properexperiments and trials we can supplyour patient with the proper medica-tion. It will give me an idea about thebest way to deliver knowledge andupdate junior residents. It has alsoopened a lot of opportunities for contactwith physicians from all over the world,and this may be a seed for futurecollaboration.”For the first semester, when he was

completing a fellowship atPrincess Margaret Hospital,Toronto, Canada, Ahmedwould listen to lecturesbefore he went to work andmake up time at the week-end. In Cairo things aretougher, since he works sixdays a week until all thepatients have been seen,which can be as late as mid-night. He is confident, how-ever, that he will finish thecourse successfully. “I amreally happy with this pro-gramme and I am eager tofinish it because it will do alot forme. I know I cando it.”The other ESO bursary

went to Zeinab El-Sayed,assistant professor at theDepartment of ClinicalOncology atAin ShamsUni-versity,Cairo, where she spe-cialises in head and neckcancers, sarcomas, paediatriconcology and cancer ofunknown origin.

“I obtainedmyMD in1999, but clinicaloncology is highly dynamic and requiresalmost continuousmedical education. Itis not the same as it was 10 years ago.When I saw the modules and the cur-riculum, I found this course very attrac-tive.” A visit to Ulm convinced her thatthe coursewas right. “Itwas very impres-sive. I found the staff very cooperative.The other students are a very nice group.Iwas very happy and felt I hadmade theright choice.”She rises early to doher studying, and

makes up time at the weekend, as she

ensures that her two boys, aged 16 and11, are also doing their homework!Course organisers can seehowmuch

time each student spends on line, butthe acid test is not the hours ofwork butpassing the exams in each module.There is no requirement to study a lec-ture if this is an area where a student isalready strong.ManuelaBergmann says,“The students can decide themselveswhere they have difficulties and have todo more, and where they have less tolearn. It is up to them.”ChatrinaMelcher, coordinator at the

European School of Oncol-ogy, said that ESOwas back-ing the course because of itssound structure and innova-tive approach. “The MasterOnline Advanced Oncologyprogramme stands out for itsimpressive and very wellstructured contents andbecause it provides academicrecognition. It is an excellentexample of blending tradi-tional attendance seminarswith e-learning,whichwe feltis an innovative approach.”This course does not

admit nurses or youngerdoctors who have just quali-fied. However, ManuelaBergmann says that addingnew courses will be consid-ered when this one hasproved itsworth.Applicationsfor the 2011 intake can beobtained from her at Ulm bywriting to [email protected] closing date for applica-tions is 15May 2011.


Systems&Services

“Clinical oncology is highly dynamic and requires

almost continuous medical education”

A working doctor. ZeinabEl-Sayed can combine her

study for the Masters degreewith a demanding job as a

clinical oncologist at AinShams University, Cairo

cancer world 41

Documents

advanced breast cancer

world leaders

cancer worldshould

standards of care

rights reservedcancer

chronic diseases

infectious diseases

people factsnumber