c w 48 education & knowledge through people &...

� Mahasti Saghatchian: pioneering a quality mark for Europe’s cancer centres �Why prostate units deliver the best personalised care � Fixing the holes in the opioidsupply lines � The 250-year-old theorem that is offering hope to rare cancer patients

Mahasti Saghatchian

Education & knowledge through people & facts

Number 48, May-June 2012

CancerWorld

48M

AY-JUN

E2012

CANCER WORLD � MAY/JUNE 2012 � 1

Contents

3 EditorialFraming the argument over futile care

4 Cover StoryMahasti Saghatchian: pioneering a quality mark for Europe’s cancer centres

13 e-Grand RoundRecent trials in osteosarcoma and their implications for future studies

24 Cutting EdgeCan the Reverend Bayes help deliver proven therapies for patientswith rare cancers?

34 Spotlight OnFixing the holes in the opioid supply lines

44 Impact FactorLocalised non-bulky Hodgkin lymphoma – future questionsFirst-line bevacizumab for ovarian cancer – new standard of care?

52 NewsroundSelected news reports

58 Systems & ServicesProstate cancer units: it’s about options and quality

EditorKathy [email protected]

Assistant EditorAnna Wagstaff

Editorial AssistantAlexandra Zampetti

Editorial AdvisorsJacques BernierFatima CardosoFranco CavalliAlberto CostaVincent T. DeVita

Contributing WritersSusana Banerjee, Marc BeishonOlivier Casasnovas, Bertrand CoiffierSimon Crompton, Stan KayePeter McIntyre, Anna Wagstaff

Publishing AdvisorsGillian Griffith, Fedele Gubitosi

Website LiaisonAlexandra Zampetti

Art EditorJason Harris

ProductionHarrisDPIwww.harrisdpi.co.uk

Printed byGrafiche Porpora

Cover photographYoan Valat

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 ©2012 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

We“overdiagnose, overtreatand overpromise”. Thiswas the claim made by

numerous newspaper headlines in responseto theLancet Oncology report last Septemberon delivering affordable cancer care. Whilewedoneedopen and frankdiscussions abouthow to curb the spiralling costs of cancercare, it was unfortunate that the mediafocused so heavily on the cost of futile treat-ment in the last weeks of life, blaming it allon a culture of excess. This sparked reportsthat patients would be denied potentiallylife-prolonging treatments purely on thebasisof cost and generated fears that patientsmight be abandoned in their final months.

There isnodoubt thatwedoovertreat andoverpromise in the advanced cancer setting.We know, for instance, that many patientsreceive cancer treatments in the lastweeks oftheir life, and that some of these treatmentshave no reasonable chance of helping thepatient and are associated with severe side-effects that can lead to hospitalisation andeven death. We also know that receivingchemotherapy is associated with a delay inreferral to palliative care. But blaming this ona culture of excess is too simplistic.

Making the right decisions in later stagesof advancedcancer is difficult for doctors andpatients alike. It is often impossible to predicthow longapatientwill live, andwhilewehavean increasing number of therapies to choosefrom, we don’t yet know enough about whostands to benefit and by how much. Therecanbe ahugedisconnect between the expec-

� Kathy Redmond � EDITOR

tations of patients and families and those ofclinicians, adding to the difficulty of con-ducting honest conversations with patientsabout their prognosis, treatment options andend-of-life preferences.

Therearenoeasyanswers.Butcouldwebemaking things harder for ourselves by posingoptions in terms of a choice between eitherfightingcancer or optimisingquality of life?Anemerging body of evidence shows that inte-gratingpalliative care into themainstreamcareof cancerpatientsnotonly improves theirqual-ity of life, butmight evenhelp them live longer.Early involvement of palliative care specialistshasalsobeenshowntocutdownonfutilemed-ical interventions andhelp families copebetterwith their loss of a loved one.

ASCOisnowrecommending thatpatientsshould be offered concurrent palliative careand standard cancer treatments early in thecourse of their advancedcancer journey.Thisis in linewith efforts to stimulatemeaningfulinteraction between mainstream oncologyandpalliative care specialists thatESMOandotherEuropeanprofessionalbodieshavebeenpursuing for some time.However, progress sofar has been infuriatingly slow.

Greater integration of palliative carerequires changes in the way we organisecare and train clinicians. We need to get onwith this as a matter of urgency. If we fail totake a lead in addressing shortcomings in thewaywe care for patientswith advanced can-cer, the simplistic arguments about awaste-ful culture of excess couldwin, and patientswill be the losers.

Framing the argumentover futile care

CoverStory

4 � CANCER WORLD � MAY/JUNE 2012

Mahasti Saghatchian:pioneering a quality mark for

Europe’s cancer centres� Marc Beishon

She’s an internationalist, she believes in quality control and she’s not afraid of a bit of friction.

Who better, then, to lead the project to define standards for Europe’s cancer centres and sort

the centres that meet them from those that need to do better?

InEurope, if a hospital chooses to call itself acomprehensive cancer centre – either as astandalone oncology facility or a departmentwithin a general hospital – it is free todo so inmost countries. Like other terms that convey

quality and authority, such as ‘university hospital’ or‘institute’, the public might assume that rigorousstandards are applied by authorities to guaranteethat status. But while there will almost certainly bemany general hospital regulations about issues suchas infectioncontrol,wastemanagementandradiationexposure, patients would be hard pushed to findout just how good their cancer care is, or howmucha centre is contributing to education and research.

“It’s not enough for acancerhospital simply to saythey are one of the top centres – they need to showthey are,” says Mahasti Saghatchian, chair of theaccreditation and designation group at the Organi-sation ofEuropeanCancer Institutes (OECI). “Justbecause a centre hasmany top oncologists does notautomatically mean that patients are always gettingthe best treatments, or that they are participating aswell as theycouldbe in researchprogrammes.Amongthekey aimsof theOECIaccreditationproject is forcentres tobenchmark themselves against others and

addressweaknesses, andalso to recognisewhere theycanwork together in researchbybuilding trust in theircapabilities.Andnot leastwehope itwill alsobeasignof trust for funders.”

As Saghatchian acknowledges, the accreditationtool for cancer centreswas some time in gestation– itwas six years inpreparationbefore launching in2008,and the first roundof centres finally received accred-itation in 2010.A further aim – that of designation –has been added to categorise locations as a unit, cen-tre, research centre or comprehensive centre.

Founded in 1979, the OECI has been around along time,but it hadmainlybeena relatively informalnetworkinggroup for cancer centredirectors inwest-ern and eastern Europe, says Saghatchian. “Thatchanged when, in 2000, Ulrik Ringborg of theKarolinska in Stockholm, and Thomas Tursz, thendirector of the Institut Gustave Roussy in Villejuif,Paris, became OECI presidents and developed avision for comprehensive cancer centres inEurope,in particular to integrate research with care anddevelop translational research networks.”

The accreditation project is part of this vision,which is similar to the comprehensive cancer centrestructure in the US, but with more of a focus on all

CoverStory


aspectsof cancermanagement rather thana researchnetwork.But, aswith anynewmeasure, it has takena lot of ‘selling’, particularly as there is a substantialcommitment in timeand fees. “It’s certainlybeenoneof themost controversial projects I’ve been involvedwith in oncology,” says Saghatchian.And theOECIhas had to find the initial resources to develop thestandard, recruit auditors and so on, before feesfromcentres canmake theprogrammeself-funding.

“Whathas really helpedget it off the ground is itsincorporation as a work package in the Eurocan-Platform, theEC-funded7thFrameworkProgrammeproject that aims to unite 28 European institutes ina translational research effort,” says Saghatchian.“It’s one of the commission’s networks of excellencefor research andwemanaged to get accreditation in,verymuch as a cherry on the cake – and all the par-ticipating centreswill alsohave toundergo the auditto take part in Eurocan.”

Not all the participants inEurocan are hospitals

– someare research institutes –but three of the firstsixOECIaccreditedcentres areEurocanmembers,namely the Netherlands Cancer Institute (NKI),the Christie in Manchester, and Valencia’s cancercentre (the other three are the Portuguese oncologyinstitutes in Lisbon, Porto and Coimbra). Othercentres are inpeer review, andapplications arepend-ing for a major expansion, including heavyweightssuch as the Institut Gustave Roussy (IGR)(Saghatchian’s ownemployer),King’sHealthPartnersin London, Cambridge Cancer Centre, InstitutCurie in Paris and the German Cancer ResearchCentre (DKFZ) inHeidelberg.

But what could mark a major breakthrough is adecision by Italy’s ministry of health to fund all tenof the country’s comprehensive cancer centres togo through accreditation. “We are starting to seegovernments and healthministries interested in theproject. If they want to accredit their oncologyeffort, say as part of a national cancer plan, the

YOAN

VALAT

OECI is the only international organisation able todo it,” says Saghatchian.

As sheadds, theOECIandEurocanPlatformarealsonowpartners in the secondEuropeanForumonOncology, takingplace inMay inBerlin,where a keyaim is to discuss the ‘bottom up’ structural reformsthat the OECI is promoting in European oncology,including aworkshopon ‘organisational concepts forcomprehensive cancer centres’.

Although much of the initial impetus for theaccreditation is coming fromthe translational researchside, Saghatchian stresses that the role of cancer cen-tres in all aspects of health improvement throughoncology is verymuchpart of the vision.Clinical careand infrastructurehaveasmuchweightingas researchin the standard, which itself is not set in stone – it iscurrently being revised to focus on factors that canreallydifferentiatepractice. Inanycase, asSaghatchianadds, accreditation only lasts four years, after whichany centremust go through the process again.

There are nowmoves to extend the project as an‘umbrella’ to include accreditation for specificcancer centre departments such asbreast units (where there isprogress on a pan-Europeanscheme for certificationfrom EUSOMA andother parties), and alsoprostate cancer care,where there is cur-rently very little tospeak of in compara-tive tools. “We arediscussing the ideaof adding prostateunits as an annexe tothe OECI tool, whichwould take probably anextra day in the reviewprocess to carry out,” shesays. “But it’s important to notethat we are not going to duplicate

professional guidelines, such as how to carry outsurgery or apply systemic therapy. We are taking aglobal view of a centre and its activities, resourcesand outcomes.”

The accreditation work is one part ofSaghatchian’s role at IGR,where shecarries out twojobs: executive in charge of international and Euro-pean affairs, and a medical oncologist in the breastcancer unit. It’s more or less an equal split betweenthe two roles, and anunusual arrangement inEuro-peanoncology, especially for someone inmid-career.But suchportfoliopositionsare likely tobecomemoreprevalent in cancer centres precisely because of theneed tohave specialists andnot administrators in thefrontline of networking and benchmarking work, toimprove research collaboration and care outcomes.

Saghatchianwas born in Iran before the Islamicrevolution, and although her parents were notinvolved in politics they chose to leave for Francewith their two daughters when it became apparentthat opportunities for girls under the new regime

after 1979would be limited. “I chose to studymedicine partly because I had impor-

tant family figures who had beenin medicine – my grand-

mother was one of the firstIranian women physi-cians – and also becauseIwanted a profession Icould do anywhere inthe world.”

Her sights wereset firmly on enteringa specialty with a

Clinical care and infrastructure have as much

weighting as research in the standard

CoverStory


In the genes.Saghatchian’s grandmother,

Maliheh Dadgaran-Pessian,took a lead in driving through

changes in Iran’s medicalprofession when she became one of

the country’s first women physicians

gets for treatment. It’s almost finished–wehave iden-tified a set of 214genes that predict late recurrencesand a gene that is overexpressed.”

Saghatchian’sPhDsupervisor isLaura van ’t Veer,the pioneer in gene expression profiling, and thework is exactly the kind of translational researchthat demands more cooperation among Europeancentres, shesays. It iswhyadvocatesofTRANSBIG’sMINDACT adjuvant therapy profiling trial are soenthusiastic – not about the primary question somuch, but the ‘goldmine’ of frozen tissue samplesfrom 6000 patients and the collection of expressiondata from 44,000 arrays. “It’s why we participate inMINDACT at IGR, but it has been the other maincontroversial area forme, alongwith theaccreditationtool.There is almost a religiousdividebetween thosewho believe in the Mammaprint gene profile andthose who don’t, but for me it’s not about belief butabout science. Every day we use markers that havenotmet full approval in an evidence base – but thatshouldn’tpreventus fromgoingonwith the research.”

Saghatchian spent five years as amedical oncol-ogy fellowat IGR,beforemoving toanacademicgen-eral hospital in Paris, the Georges Pompidou

strong and growing research component, and shequickly rejected fields suchascardiology in favour firstof immunology, and thenoncology, but she candidlyadmits that, even relatively recently, she foundmed-ical oncology lacked much research promise, com-prisingas itdoesmainlychemotherapy. “If I’mhonest,really themost gainshavebeen in surgery and radio-therapy inmy field of breast cancer – it is only latelythatwehavepersonalisedmolecular therapies and Ithinkmedical oncology’s time is verymuch to comein breast cancer. In our tumour board meetings atIGR we have a lot of discussion about surgery andradiation choices, but it’s always the same adjuvanttherapies – there has not been much change, apartfromHerceptin.”

Acase inpoint for the future is her own researchfor a PhD. “I have been looking at breast cancerpatients who relapse late – half of the 30% whorelapse do so after five years, but all trial work is onshort-term rates, up to five years – no one is lookingathowtoprevent late relapses, aswedon’tunderstandthem and can’t select those patients and treat themaccordingly. I’vebeendoingmicroarrayprofiles to seeifwecan findpredictivemarkers for relapse and tar-

CoverStory


Clearing the hurdles – the Christie experienceThe Christie in Manchester, UK, one ofEurope’s largest cancer centres, wasamong the first to receive OECI accred-itation as ‘comprehensive’.ChrisHarrison,medical director, says the attraction oftaking part is thewide benchmark it offersto compare against others inEurope. “Wehave anational peer reviewprogramme forcancer, which focuses on care quality,but theOECI review covers not just clin-ical care but also our teaching andresearch, and thedegree towhich they areintegrated – that being the hallmark of acomprehensive cancer centre.“We had to assemble a portfolio of infor-mation for the OECI audit team, whichcomprised people such as another cancercentre director and a specialist nurse, andthey visited us for two days, meeting ourexecutive team, doctors and staff, and

they went into real depth on a number ofareas. They identified several things weneed tomove furtherwith, such as a fullyelectronic patient record system, and abetter ability to publish outcomes of care.We also had to explainwhywe don’t havedirect responsibility for screening andprevention, which we do though supportthrough our regional network.As a large,mostly single-site centre, including basicscience, and with links to a local aca-demic system, we were able to give thereviewers a good account of our teachingand research structure.”Harrison adds that, given the UK’s rela-tivelyworse outcomes,which are thoughtto result from later presentation, theOECI’smove to gathermore comparativedata on outcomes across European cen-treswill beparticularly valuable, and could

help develop the role of centres in pre-vention and early detection.He is also a co-opted member of theOECIboard, andnotes that a centre suchas the Christie needs to be involved at aEuropean level, given the increasingmovement of patients and staff acrossborders, the impact of European legisla-tion and the need to reach a critical massfor research through programmes such astheEurocanPlatform. “I have also chairedtwo reviews myself, at Jules Bordet inBrussels and theNKI inAmsterdam, andmade other visits on behalf of MahastiSaghatchian, toEstonia, for example. It isapparent that theOECI accreditation is acatalyst for centres that want to improvetheir care.” Saghatchian, he adds, hasbeen the driving force to getting theaccreditation group established.

European Hospital, where she looked after lungcancer patients, among other roles, for two years. “Ifound thatoncologyaway fromcancercentrescanbea really different job. There can be a fear of cancerpatients andamisunderstandingofwhat’s possible inthe emergency unit, for example. In day-to-day carewe didn’t have palliative care teams or pain special-ists, andnomolecular profiling – that had to be sentelsewhere–and it is impossible todo researchwhenyou don’t have enough patients. My own expertisesuffered because I didn’t see rare cases, and if I didImight not have known how to treat themwell.”

It is highly unlikely that such hospitals couldmeetOECIcriteria, but Saghatchian says that pub-licity for centres that do become accredited mayhelp patients and primary care doctors make moreinformed referral decisions. “There is little informa-tion for patients about where the best care and spe-cialists are. This isn’t just true for oncology of coursebut for all specialisms–youoftengo towhere youaretold to go or where your friends went.” In hospitalsthathaveacancerdepartment there is a tendencyalsofor surgeons to refer patients there rather than toexternal cancer centres, which is part of the long-standing discussion about the primacy of organ-basedpractitioners versusmultidisciplinaryoncology.

Many large general hospitals do have compre-hensive cancer centres, and Saghatchian acknowl-edges the extra resources that can be brought tobear from other specialists. She is keen to stressthat any hospital with a cancer centre is free to seekaccreditation, but concedes that some smaller oneswill be content with national systems, and are notseeking international recognition. Unicancer, theFrenchprogramme, andothernational initiatives arebeginning to apply rigorous audit – theNHS inEng-

land, forexample,has started local auditofcolorectal,lung, oesophago-gastric and head and neck cancers,in some cases at the level of individual units.

“But it is also the case that national systems suchas ours in France are applying only basicminimumstandards for oncology in most smaller hospitals,such as the number of breast operations that needto be done. It’s why cancer plans tend to fail in myview– politicians oftenwon’tmake the tough deci-sions to close oncology departments that do notmeet higher standards.”

Saghatchian returned to IGR in2003, but askeddirector Thomas Tursz for a position that wouldnot be a full timeclinical post. “Itwas partly becauseoncologywas a bit dull and also toughwith somanydyingpatients – I didn’twant to suffer fromburnout– but it was also about my personal history as a for-eigner.Evenatmedical school I had runa small soci-ety for foreign students and had the feeling thatinternational exchange work was a great way tokeep yourself fresh and learnmore. Thomaswantedsomeone to develop international affairs and hecreated the job for me.”

As she points out, it is perhaps surprising thatmorecancer centresdonothave similar roles. “Noneof the other centres in France has someone like meIbelieve,but it is very important for IGRtohaveaunitto attract funds for research programmes and be avoice for the centre.”

Her half-time post relates directly to the aims ofthe OECI accreditation project, which is why shehas been so keen to champion it in Europe,although there is an element of competition. “AtIGRwewere finding it hard to get funding for aca-demic research, but nowwe aremuchmore organ-ised about the way we respond to ‘calls’ for

“There is little information for patients

about where the best care and specialists are”

“Politicians often won’t make the tough decisions to

close oncology units that do not meet higher standards”

CoverStory


madeIGRattractive for youngpeople todoPhDsandtowork abroad, andnot leastwehadamajor interiorrefurbishment six years ago–although theoutside isstill rather grim.” LexEggermont, the current direc-tor, was a brave appointment, she adds, as he isDutch,buthasmadean impactwithexcellent finan-cial management and has further boosted IGR’sinternational standing.

The experience so farwithOECI accreditation,says Saghatchian, is that standards of care – such asthe percentage of patients seen bymultidisciplinaryteams – are relatively straightforward to compareamong centres. “It’s harder to look at research andeducation programmes, and also the integration ofresearchwith care. The cultural and organisationaldifferences between countries are also big chal-lenges of course, and we have no plans to work inany language other than English.”

Establishingdefinitions andquestions for collect-ingdata thatavoidmisunderstandingsandcompare likewith likehas taken a lot of effort, evenwith seeminglysimple factors suchas thenumber of patients treated,and the resources and infrastructure in place.

“And one of themain issues that the project hasrevealed is just how difficult it is for centres to col-lect data about themselves –we’ve realised that sen-iormanagement often do not have a clear picture of

European framework projects, for example. In the7th Framework Programme we are involved inmore than 20 calls that are now a major source ofincome. Before it was just an ad hoc effort by a fewstaff who knew what to do.”

That may be competing with other centres tosome extent, but Saghatchian adds that new part-nerships are forged within programmes such asTRANSBIGandCHEMORES. “In theCHEMORESlungcancer andmelanomaFP6project, for example,wedidn’t knowsomeof theotherpartnerswell at all.Now it’s finished, a lung project has emerged that’sindependent and wouldn’t have happened withoutthe original programme. Basic scientists tend toknow each other around the world, but in transla-tional research, clinicians oftendon’t knowwhobestto workwith andwho has the best infrastructure.”

Saghatchianconsiders that IGRnowprettymuchmeets the OECI criteria for a true comprehensivecancercentre,but it’s takena lot ofwork,drivenespe-cially by previous director Tursz. “We have nationalquality assessment and benchmarking of Frenchcentres through Unicancer, which checks aspectssuch as multidisciplinary care. Five years ago, only70% of breast cancer patients were discussed bymultidisciplinary teams–nowit’s 100%.Thomasalsochanged department headswhoweren’t doingwell,

CoverStory


YOAN

VALAT

what exactly is going on in their organisations.Onevery tough question is, ‘What is your researchbudget?’But the data are often not centralised andyou do wonder how they manage without crucialinformation like this.And the bigger the institution,the more difficult it can get.”

A case in point is King’sHealth Partners in Lon-don, which is currently in progress with its OECIapplication. “It’s definitely harder for centres suchasKing’s to collect data because it has multiple sites,wherepeoplemaynotbemeasuring the same things,or in the sameway.”Meanwhile anexample sheciteswhere reviewers foundresearchandclinical care inte-gration is not as strong as it could be is at HelsinkiUniversity Hospital. “They did not find a specificorganisation for translational cancer research.Butweare finding that centreswelcome the reviewprocessbecause itdoeshelp themtohighlight areas thatneeddevelopment and gives them evidence to ask formore resources.”

For the timebeing, a country thatwill benotablefor its absence in theOECIaccreditationprogrammeis Germany, except for DKFZ in Heidelberg, whichis aEurocanPlatformmember. Saghatchianexplainsthat ismainly because ofGermany’s history of treat-ing cancer by organ specialists, with all the contro-versy thathas created. “TheGermanCancerSocietyhas its own certification strategy and organisation,OnkoZert, for progressively addressing the issuesrather than tackling themheadon.TheGermanprob-lem is specific to the country andwewon’t domuchthere in next few years except for a pilot.”

In fact, following a move to establish secondopinion services for testicular cancer, an increasingnumber of prostate cancer units have been certifiedinGermany – asmany as 68 by last year. This expe-rience is feeding into work by ESO and OECI onestablishingaprostateunit standard (seealsoSystems& Services, p 58). “There is certainly a huge need.Evenat IGRwedon’t havea formalprostateunit andwewouldwelcomeguidelines andcarepathways forprostate cancer.”

“We’ve realised that senior management often do

not have clear picture of what exactly is going on”

CoverStory


Party time. With nine-year-old son Olivier,all dressed up andready to dance thenight away, Dubai,

New Year’s Eve 2011

THE ACCREDITATION PROCESS

The accreditation component of the OECI’s programme is a quality standard forcancer centres, andwas developed in two rounds of pilots before the first versionwas launched in 2008. Apart frommembership of the OECI, cancer centres areasked tomake a considerable commitment to the programme, including a fee of€30,000 for the larger organisations. Accreditation lasts four years.The accreditation process involves completing a self-assessment questionnaireand attaching a long list of supporting information, before the audit team con-ducts its review.The designation part of the programme is a spin-off designed to help organi-sations of similar standing form networks with each other, and was also pilotedand validated. The four categories are: cancer unit, clinical cancer centre, can-cer research centre and comprehensive cancer centre (an accreditation pack-age for the third category, research centres, is yet to be developed).The quality standards and a usermanual, as well as newsletters and other infor-mation, are available at oeci.selfassessment.nu

Another factor slowingprogress, sheadds, is a chronicunder-use of IT– “I’mamazedwedon’t domorewithtools suchas iPhones andemail. Sometimes I get thefeelingpeoplearehappy toslowdownthepaceofworkbecause of fear of overload.” That applies at IGR aselsewhere – and the use of modern IT is one of theOECI standards – but otherwise her centre is nowdoing better than ever, she says,with its recent refur-bishment and improved efficiency leading to morefunding. “TheFrenchhealth systemthough is slippinginqualityandaccessandweare facingevenmorepres-sure from the pharmaceutical industry. We’ve hadsome drug scandals, such as with a diabetes pill,which is causingmistrust towards doctors.”

Some less pressurised aspects of her work atIGR includedhelping toproduceabookofpaintingsof breast cancer patients, and a study on the impactofusingbeauty treatmentson self-imageanddepres-sion, carried out with l’Oréal. And she has not for-gotten her roots, setting up a link between IGR andMAHAK, anorganisation in Iran thathelps childrenin the country receive cancer treatment.

Above all, the theme that best sums up herapproach is networking and movement. “I love theEuropeanwork–you learn somuchwhenyoumovearound and people should definitely aim to work inother countries.”

Perhaps the OECI accreditation process willintroduce a measure of foreign personnel andmovers in future.

In the current revision of the accreditation,Saghatchian says some basic standards will beremoved because they are common to all. ‘We arefine-tuning thequantitativedata todevelop indicatorsthat showdifferences.For example, oneof theambi-tious indicators we want is to compare survivorshipbetweencentres– theoutcomedata.Thatmeanscol-lecting the samedataonpatients at the same time fortheir disease, including follow-up. At present wecan only look at country registry data across Europe–but that doesn’t showwhere apatientwas treated.”

Saghatchian feels the OECI has taken a lead indriving forward the benchmarks for improving out-comes in Europe, and she has certainly brought agreat deal of passion to her European work. Sheexpresses frustration that other organisations donotseem to have the same focus. She would like to seethe EORTC, which organises international cancertrials, continue its modernisation towards transla-tional research;ECCO, she says, needs to articulateits vision better; and advocacy groups should bepushing much harder for breaking down regulatorybarriers, suchas in tumour collection. “Weare tryingto launch a neo-adjuvant trial where we want tocollect samples before and after giving Herceptin –butas there isno immediatebenefitwecan’t do it. It’sone reasonwhyprogress inpersonalisedmedicine isslow.” She would also like to see the research com-munity become much more imaginative in usingthe talents in other fields, such asmathematics.

CoverStory


“I’m an oncologist, I work with patients whohave breast cancer”“Ah, that must be hard.”“No, no, no, people should know the truth!No it’s not hard, it’s incredible!We need to open a window onto theseamazing lives that I mix with every day”

Extract from the catalogue to an exhibitionof paintings of women touched by breastcancer, ‘La Vie en Plus’, which was acollaboration between artist ThierryDussac and Mahasti Saghatchian withthe Institut Gustave Roussy. This isthe portrait of Mahasti

e-GrandRound


Recent trials in osteosarcomaand their implications for future studies

Osteosarcoma is a rare cancer and most oncologists will not come across it very often. In this

overview a sarcoma specialist presents current evidence on the best way tomanage osteosarcoma

patients, and looks at what has to be done to improve prognosis in this disease, where survival

rates have not changed since the early 1980s.

Osteosarcoma is a malignantmesenchymal tumour pro-ducing osteoid. It is a rare type

of cancer, with an incidence of only 2–3 permillion per year, occurringmostlyin adolescence and affecting moremales than females (in a ratio of 1.4:1).In adolescence it usually occurs in themetaphyses of long bones, usuallyaround the knee. The problemwe faceis with metastases, which occur inabout 90% of patients. Both primaryand later metastases usually occur inthe lungs and sometimes in bones, butrarely elsewhere.In terms of imaging methods, the

osseous compartment it still best visu-alised by conventional X-ray. This is themethod of choice for bony alterations.MRI is also needed to look at theprimarytumour, showing the amount of mar-row involvement, the soft tissues, andthe relationship to vessels and nerves,providing essentially all the informationthe surgeonneeds. Imaging for systemicspreaddepends on location: X-ray for thechest and bone scans for bones. But themost valuable imaging method is a CTscan of the chest.

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, Stefan Bielack, fromStuttgart’s Olgahospital, in Germany, pro-vides an update on recent clinical trials inosteosarcomaand the implications of thefindings for future studies. BruceMorland,from theBirminghamChildren’sHospital inthe UK, poses questions arising duringthe e-grandround live presentation.The presentation was summarised bySusanMayor.

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

This CT scan shows ini-tial staging in a patientwho has a very tinylesion; it is theonly lesionand the rest of the CTscan is normal. At theend of chemotherapy,there was no change.Howshouldwe treat thispatient? We will comeback to this later.The table below

shows results of surgeryin several series ofpatients with primarymetastatic osteosarcoma,comparing those whoachieved surgical remis-sion with those who didnot. In most series,patients who did notachieve a surgical remis-sion, or whose metastases were notremoved, did not usually survive for fiveyears. The situation is similar at recur-rence.Patientswhodonot achieve a sur-gical remission rarely become long-termsurvivors; patients who do achieve a sur-gical remissionhave a reasonable chanceof long-termsurvival.Complete surgery isnecessary for long-term survival.

IDENTIFYING METASTASESHowwell canCT tell uswhethermetas-tases are present?Avery interesting jointItalian–Scandinavian study looked at 51osteosarcoma patients with suspectedmetastases onCTscan.At surgery itwasfoundthat39hadmetastaseswhile22didnot (AnnOncol 2001; 12:1601–04).How canwe tell whether small nod-

ulesmight bemetastases? Radiologistsalways tell us that small nodules that donot change after chemotherapy cannotbemetastases, or that if they disappearthey are not metastases.What the Ital-ian–Scandinavian study foundwas thatchanges in nodule number and sizeduring chemotherapy did not indicate

whether patients really hadmetastases.The only factor that was significant wasthat a lesion smaller than 5mmwas lesslikely to be a metastasis than largerlesions (P=0.035). But ten of 25patients with nodules <5 mm hadmetastases. So if a patient has a smallnodule on CT scan, even if it is onlyone, and even if it does not change


e-GrandRound

during therapy, it can be a metastasis.Another study fromNewYork looked

at 28 patients who underwent 54 thora-cotomies. Preoperative CT scans in allthe patients showed183 suspectednod-ules.At surgery, the surgeons found 329nodules, 209 of which were osteosar-coma. This means that a CT scan hadunderestimated thenumber of lesions in19 of the 54 patients whowere referredfor scanning – about one third (J PediatrSurg 2006; 41:200–206).My take home message for lung

metastases, both primary and secondary,is that there is no perfect imagingmethodology.CT is the best, but it is notperfect.Youwill often findmore lesionsthan expected and should lookbilaterally,even if a CT scan has shown metas-tases only on one side. It is essential toremove allmetastases or the patientwillnot survive.Surgery is usually open thoracotomy.

Video-assisted thoracoscopic surgery(VATS) is not recommended becausesurgeons should palpate the lungs. Wedo aCT scan after thoracotomy andwesend patients back to the thoracic sur-geon if there are stillmetastases. Iwouldeven send them back a third time ifthere were still no remission.

TREATING A PATIENT WITH A SMALL LESION

A single small lesion in the chest visible on CT scan beforechemotherapy was still there after the treatment ended. What now?Source: Slides courtesy of Prof P Winkler, Olgahospital Stuttgart,

Prof I Arlat, Katharinenthospital, Stuttgart and Dr M Schilling,

Radiologische Praxis, Stuttgart-Bad Canstatt

SURVIVAL OF PATIENTS WITH VS WITHOUT REMISSION AFTER SURGERY

These five-year results show that complete resection is a prerequisite for survivalYes – remission; No – no remission

Initial staging End of chemo

Primary Yes No pMeyers et al. 1993 surg. remission 20% 0% <0.001Kager et al. 2003 surg. remission 40% ca. 0% <0.001Tsuchiya et al. 2002 surg. resection 31% 5% <0.0001Daw et al. 2006 surg. remission 40% 6% 0.001

RecurrenceTsuchiya et al. 2002 early 10% 5% <0.0001

late 48% 8% <0.0001Kempf-Bielack et al. 2005 38% 0% <0.0001Ferrari et al. 2003 39% 0% <0.0001

seen byCT.Even larger lungmetastases –up to about 1cm in diameter – are some-times not picked up by PET. Bonemetas-tases are rather infrequent as primarymetastases. You usually see them by bonescan and there are no data that indicatePETwould bemore sensitive.WherePETmay have role is in following the osteosar-coma during preoperative chemotherapy,predicting response to preoperative chemo.Question:What is the role of radiotherapyin lung metastases?Answer: We have a very old study withwhole-lung adjuvant chemotherapy – arandomised trial from the EORTC. Thisstudy, performed in the late 1970s, showedthat if you do not give chemotherapy, thenadjuvant radiotherapy to the lungs willreduce the risk of recurrence by a smallamount. Adding adjuvant radiotherapyto effective chemotherapy will not addanything significant. The only role that Isee for radiotherapy of lungmetastases is for

Going back to our discussion of thepatient with the very tiny lesion that didnot change during chemotherapy (figureopposite): when she underwent thora-cotomy the lesion was found to be ametastasis and she had three more thatwere not evident on CT scan.

Question:Could you give some indicationof the potential role of adjuvantchemotherapy in patients with recurrentmetastatic disease in the lungs? You haveclearly outlined the role for surgery.Whatis the role of chemotherapy?Answer:There are two situations. One iswhen lungmetastases are inoperable andyou cannot remove them by surgery. Thelargest series in inoperable metastases allhave the same message: if you givechemotherapy, you can prolong survival byseveral months but you will not cure thepatient. As a group, patients who receivechemotherapy live a fewmonths longer.

The second situation is whereyou do achieve surgical remission.Here the role of second-line adju-vant chemotherapy is not as clear.The two largest series are fromItaly, and from our German/Austrian/Swiss group. The Ital-ians did not find any benefit fromsecond-line adjuvant chemother-apy. We found that freedom fromsecond recurrence was increasedby about 5–6%, whichwas statis-tically significant but not atremendous improvement. In apatient with pulmonary metas-tases at recurrence, Iwould discussand offer adjuvant chemotherapy,but I would never try to convincesomeone against their will.Question:Does PET have a rolein defining metastatic disease?Answer:No, there is currently norole for PET in defining metasta-tic disease in osteosarcoma. PETwill not usually pick up lungmetastases that are too small to be

a limited lesion that you cannot resect. Iwould discuss with the radiotherapistwhether it could be irradiated.

SURGERY AFTER PRIMARY TUMOURThe bar chart below shows the types ofsurgery used in a largemulticentre studyby five-year intervals for 2000 extremityosteosarcomas. In the 1980s, limb sal-vage (shown in yellow)wasused in aboutone-third of patients, and the othertwo-thirds had either amputation or rota-tionplasty. The proportion of amputa-tions has since dropped dramatically.Most patients with extremity osteosar-comas treated since 2000 went on tohave limb salvage (Cancer Treat Res2009; 152:289–308).

LOCAL RECURRENCEOur group looked at recurrence in 1702osteosarcoma patients; 576 developedrecurrences and 75 of these had local

recurrence. Forty-four had localrecurrence only and their five-year survival was 26%; 31 hadlocal recurrence combinedwithmetastases and their five-yearsurvival rate was only 7% (JCO2005; 23:559–568). The resultsshow the importance of avoidinglocal recurrence.A study conducted by our

groupusingdata from theCOSS(Cooperative OsteosarcomaStudyGroup) registry (AnnOncol2011; 22:1228–35) shows thereare three predictive factors forlocal recurrence.Tumour response to chemo-therapy. The local recurrencerate for patients who had poorresponse to chemotherapy was10% while the local recurrencerate for patients who had a goodresponse (less than 10% viabilityof the tumour) was only 3%. So,chemotherapyclearlyhasa role inlocal control.

e-GrandRound


TYPES OF SURGERY FOR OSTEOSARCOMA

This study of data from 2000 patients with osteosarcoma of theextremities, taken from the COSS (Cooperative OsteosarcomaStudy Group) registry, shows that the use of limb sparing surgeryhas increased significantly since 1980, while amputation is usedfar less frequentlySource: S Bielack, H Jürgens, G Jundt et al. (2009) Osteosarcoma: the

COSS experience. Cancer Treat Res 152:289–308, reprinted with kind

permission. © Springer Verlag 2009

Rotationplasty

Limb salvage surgery. There was aslightly higher local recurrence rate forpatients who had limb salvage. In ourexperience, the local recurrence ratewas2.8% in470patientswhohadampu-tations and 7.5% in 885 patients havinglimb salvage. So, there is a higher localrecurrence rate for those patients havinglimb salvage.Location for the biopsy and thesurgery.Patientswhohad tomove fromthe centre that did the biopsy to anotherfor surgery had significantly higher localfailure rates. Thismay be because oftenthe biopsywas not performed in aman-ner conducive to definitive surgery andmayhave contaminated thewoundmorethan desired.Surgical volumes. I would haveexpected that patients who had surgeryat a centre doing more surgery wouldhave a lower local recurrence rate thansmall centres not doing as many opera-tions, but in our series they had identicallocal recurrence rates, so surgical volumeis not a predictive factor. However, ourfigures did show that the large centres

did limb salvage in 70% of patients,while the small centres did this in only50% of patients. So, while the localrecurrence rates are the same at largeand small centres, there is a higher rateof amputation if the surgeon is not veryfamiliar with sarcoma surgery.

LOCAL THERAPYFOR INOPERABLE SITESInoperable sites include many axialosteosarcomas andmetastases that can’tbe reached by surgery. It is usually saidthat radiotherapy does not work forosteosarcoma. Is that really true? Thereare a couple of interesting publicationson this. DeLaney and colleagues fromthe US looked at 41 patients withosteosarcoma and inoperable lesions:27 primaries, 10 local recurrences andfourmetastases. Theywere given radia-tion and some chemotherapy. The localcontrol rate was 68% at five years, withrelatively highdoses of radiation (median66Gy). It was 78% for patientswho hadgross or subtotal resection togetherwithradiotherapy and40% for thosewhohad

biopsy only (Int JRadiatOncolBiol Phys2005; 61:492–498).We did a retrospective

analysis of theCOSS registrydata for 100 patients with 66primary tumours, 11 localrecurrences and 23 metas-tases. Radiation doses werealso relatively high (median56 Gy) and all patients hadchemotherapy. The local con-trol ratewas 30% at five yearsin this highly heterogeneousmulticentre cohort: 48% forsurgery plus radiotherapy,22% for radiotherapy alone,40% for primary tumours,17% for local recurrences and0% for metastases (CancerTreatRes2009;152:147–164).The information we can

take out of this is that radio-therapy can work for selected osteosar-coma lesions that are not operable. Thefinding that itworks better if you take outlarge parts of the tumour – debulking –is interesting, as is the finding that itworks better for primary tumours thanfor recurrences or metastases. I thinkthis is because radiotherapy works bestif you give it together with effectivechemotherapy, and we have effectivechemotherapy for primary osteosarcomabut not for recurrences.The take home message for local

therapy is: operate, operate, operate.Surgery is themost important local ther-apy.Youneed a good surgeonwhoknowshow to achieve adequatemargins. Limbsalvage is often feasible. The risk of localrecurrence can be reduced by gettingfour things right:� good imaging, because the surgeonneeds to knowwhere to cut

� smartplanningbetweenall disciplines� good chemotherapy to devitalise thetumour as much as possible, and

� good surgery.Radiotherapy may be an option for


e-GrandRound

PREDICTORS OF LOCAL RECURRENCE

The results of thisstudy of data on localrecurrence (LR) ratesamong 1820 patientswith osteosarcoma ofthe extremities, takenfrom the COSS registry,shows four factorspredicting for localrecurrence, three ofthem significantSource: Adapted from

D Andreou et al.

(2011) Ann Oncol

22:1228–35

n Limb P % LR Psalvage (at 5 yrs)

ResponseGood 826 67% 0.045 3.1% <0.0001Poor 515 62% 10.2%

Limb salvagesurgeryYes 885 100% - 7.5% 0.001No 470 0% 2.8%

Centre performingbiopsy & surgerySame 882 66% 0.572 4.2% 0.001Different 406 64% 10.1%

Centre volume(</> 1/year)Large 1034 70% <0.0001 6.1% 0.761Small 321 51% 5.4%

primarymetastases.Histologic responsesto chemotherapy were graded as eithergood (>90% destroyed) or poor (<90%destroyed; >10% viable). Results (seetable below) showed a significant dif-ference in five-year survival and event-free survival across a six-grade histologicresponse system used in German-speaking countries, ranging from five-year survival of 84.5% in grade 1 (verygood response) to 40.5% in grade 6 (verypoor response), with a gradual decline inprognosis along the scale (JCO 2002;20:776–790). If you put histologicresponse to preoperative chemotherapyinto amultivariate model of overall sur-vival it beats even primarymetastases asa prognostic factor, so chemotherapyresponse is very important.A summaryof osteosarcoma trials from

theearly 1980s to the late1990s (see tableoverleaf) shows that everybodyuses essen-tially the samedrugs.A randomised trial ofchemotherapy versus no chemotherapy,theMIOStrial, showed that chemotherapyis efficacious (NEJM1986;314:1600–06).Quite a fewtrialshave tried tousemodifiedpostoperative chemotherapy in poorresponders but none have shown it works.

selected inoperable lesions. There aresome studies with proton or heavy ionradiotherapy that may help to definewhether these innovative radiation tech-niques can be more effective than con-ventional radiotherapy.

Question: What do you do if you have aresidual disease following surgery?What isthe role for amputation in themodern era?Answer: If you do not achieve adequatemarginswith limb salvage, then you shouldamputate. I think you also need to thinkabout limb amputation and particularlyrotationplasty for tumours in very smallchildren, because growing endoprosthesesare tedious for young children in their lives.

CHEMOTHERAPYChemotherapy for osteosarcoma wasstarted and evaluated in the 1970swiththree drugs: high-dose methotrexate,doxorubicin, and cisplatin. Ifosfamidewas added in the 1980s. These drugs arestill being used today. Adjuvant combi-nation regimenswere introduced in thelate 1970s and early 1980s and preop-erative (neo-adjuvant) chemotherapyhas been used since the early 1980s.

Timing of chemotherapy in relationto surgeryWe all think that you have to give pre-operative chemotherapy, but is this true?Results of a randomised study by thePediatric Oncology Group in 100patients (JCO 2003; 21:1574–80) didnot show a significant difference in five-year event-free survival betweenpatientswho had immediate surgery and thosewho had their surgery after a course ofpreoperative chemotherapy (69% vs61%). Similarly our experience for theCOSS group, looked at retrospectivelybut with larger patient numbers, showsthat patients who had delayed surgeryafter preoperative chemotherapy hadthe same prognosis (54.4%) as thosewhohad surgery first and chemotherapyafter (59.9%) (JCO 2002; 20:776–790).We can conclude that, if you give thesame total amount of chemotherapy, itprobably does not matter much whenyou perform surgery in terms of survival.For preoperative chemotherapy, we

confirmed that histologic response ofthe primary tumour is related to five-yearsurvival based on treating 1320 osteosar-comas of the extremitieswith orwithout

e-GrandRound


RESPONSE TO PREOPERATIVE CHEMOTHERAPY PREDICTS SURVIVAL

This study of the data from 1320 patients with osteosarcoma of the extremities,taken from the COSS registry, shows that response to preoperative chemotherapyis more important even than primary metastases in predicting survivalEFS – event-free survival

Source: Adapted from S Bielack et al. (2002) JCO 20:776–790

Response n 5-year survival 5-year EFSGood 734 77.8% 67.6% P<0.0001Poor 586 55.5% 38.6%

Grade 1 184 84.5 % 79.2%Grade 2 236 82.5 % 68.4%Grade 3 296 70.9 % 60.1%Grade 4 283 59.7 % 44.7%Grade 5 237 53.7% 34.3%Grade 6 40 40.5% 25.7%

Multivariate Cox model of overall survival

Variable Risk ratio (95% CI) PMacroscopic residual tumour 4.01 2.66–6.04 0.0001Poor response 2.44 1.98–3.01 0.0001Primary metastases 1.88 1.33–2.65 0.0003Axial site 1.87 1.25–1.80 0.002Age >40 years 1.41 0.70–2.85 0.340

prising, as we have been using the samedrugs since then.The take home messages for

osteosarcoma chemotherapy are:� Giving chemotherapy is much bet-ter than not; multicentre groupscan achieve similar results to singlecentres; and patients with localisedextremity osteosarcomas do betterthan those with other tumours.

� Almost everyone uses preoperativechemotherapy; however, survivaloutcomes are similar when youoperate immediately.

� Poor response to preoperativechemotherapy is a very bad thing.People try to improve resultsin poor responders by addingdrugs and increasing intensity post-operatively, but we do not knowwhether this works.

� Almost everyone uses the samedrugs (two to four of: high-dosemethotrexate, doxorubicin, cis-platin, ifosfamide).

� Nothing much has changed overthe years in terms of five-yearsurvival.


e-GrandRound

Prospective trials, which usually includeonlypatientswith localisedosteosarcomasof theextremities,may tell onlypart of thestory. The prognosis is much poorer forpatientswhopresentwithprimarymetas-tases or tumours of the axial skeletonthan for thosewith localised limbtumours(see figure opposite, top).Thebar chart opposite (bottom) illus-

trates the progress in Europe and inNorth America in the last 20 years intermsof five-year survival: basically, thereis none. The survival rates have been sta-ble since 1980, which is not really sur-

OSTEOSARCOMA TRIALS WITH CHEMOTHERAPY

Protocol Drugs n EFS (years)

COSS-80 (Winkler et al 1984) DOX, MTX, (DDP or BCD), +/- IFO 116 68% (2.5)

MIOS (Link et al 1986) MTX, DOX, DDP, BCD vs. 18+59 66% & 69% (2)no chemo 18+18 17% & 9%

COSS-82 (Winkler et al 1988) preop all MTX, DOX, DDP; postop GR: 59 68% (5)MTX, DOX, DDP; PR: DDP, IFO, CYC, ActoD vs.preop all MTX, BCD; postop GR: 60 45%MTX, BCD; PR: DOX, DDP

IOR/OS-1 (Bacci et al 1990) MTX (HD), DOX, DDP, BCD vs. 127 58% (5)MTX (ID), DOX, DDP, BCD 42%

SSG-T10 (Sæter et al 1991) preop MTX; postop GR: MTX, BCD; 97 54% (5)PR: MTX, DOX, DDP, BCD

EOI-80831 (Bramwell et al 1992) DOX, DDP vs. 99 57% (5)MTX, DOX, DDP 99 41%

MSKCC (Meyers et al 1992) MTX, DOX, DDP, BCD 279 65% (5)T4-T12

EOI-80861 (Souhami 1997) DOX, DDP vs. 199 44% (5)MTX, DOX, DDP, IFO, BCD, VCR 192 44%

CCG-782 (Provisor et al 1997) preop all MTX, BCD; postop GR: MTX, DOX; 268 53% (8)PR: DOX, DDP, BCD

COSS-86 (Fuchs et al 1998) MTX, DOX, DDP i.a. vs i.v., IFO 171 66% (10)

IOR/OS-2 (Bacci et al 2000) MTX, DOX, DDP; postop GR: MTX, DOX, DDP; 164 63%;58% (5;10)PR: MTX, DOX, DDP, IFO (HD), ETO

IOR/OS-4 (Bacci et al 2001) MTX, DOX, DDP, IFO 133 56% (5)

IOR/SSG (Bacci et al 2002) DOX, MTX, DDP, IFO (HD) 70 73% (5)PILOT

EFS – event-free survival; DOX – doxorubicin; MTX – methotrexate; DDP – cisplatin; BCD – bleomycin, cyclophosphamide, actinomycin D; IFO – ifosfamide; GR – goodhistologic response to preoperative chemotherapy; PR – poor response; Cyc – cyclophosphamide; ActoD – actinomycin D; VCR – vincristine; ETO – etoposide

Chemotherapydose intensityA retrospective analysisfrom Italy suggestedthat patients who hadchemotherapy given ata high dose intensityhad amuchbetter prog-nosis than patientswhoreceived less than theirallotted amount ofchemotherapy overtime (Oncol Rep 2001;8:883–888). Therewere two other retro-spective analyses, one from the Euro-pean Osteosarcoma Intergroup, whichlooked at doxorubicin and cisplatin (JCO2000; 18:4028–37) and one from ourgroup, which looked at methotrexate,doxorubicin, cisplatin and ifosfamidecombinations (Pediatr Blood Cancer2006; 47:42–50). Neither showed thatpatients who had higher dose intensitychemotherapy had better outcomes.The question of dose intensity was

looked at in a randomised trial by aBritish/Dutch/Belgian/Danish group.Patients were randomised to conven-tional chemotherapy (cisplatin plus dox-orubicin, two cycles preoperatively andfour postoperatively) or an intensifiedarm where G-CSF was added as sup-port, and dosing intervals werecompressed from three to twoweeks, which meant that threecycles were given before surgeryand three cycles after surgery.The response ratewas higher forthe compressed arm (50% vs36%). But five-year progression-free survival rates were identicalfor the compressed and the con-ventional arms (41% vs 39%)(JNCI 2007; 99:112–128).Retrospective and prospec-

tive analysis of high-dose chemo-therapy with stem cell rescue –the ultimate dose intensification

– did not show improved prognosis, sothe take home message is that we arenot likely to improve results by doseintensification.The question we need to ask is

whether we can improve outcomes forpoor responders, because their long-termsurvival rates arewell below50%.To lookat this we would need to randomiseapproximately700patients,whichmeansthatweneedabout1400patientswhoarewilling to be randomised after havingreceived10weeksofpreoperativechemo,which means we need far more than2000patients to go into sucha trial.Withadisease thatoccurs inonlyabout twopermillion people per year, no country cando that by itself, so you need an inter-

group collaboration.That is what we did

in theEURAMOStrial,which recruited 2260patients from326 insti-tutions in 17 countriesfrom 2005 to 2011.Figures presented byKatja Zils and col-leagues at SIOP 2011(SIOPabstracts,PediatrBlood Cancer 57:779)showed that complexinfrastructures span-ningmany institutions,

countries and even continents, areneeded for such a large trial, but that itcanbedone (seealsoCuttingEdgep24).

WHAT ELSE CAN IMPROVEPROGNOSIS?Having seenno improvement inprogno-sis resulting from modifications tochemotherapywemightneed todosome-thing else.Onepossibilitymaybe liposo-malmuramyl-triphosphate-ethanolamine(L-MTP-PE), a macrophage activatorderived frommycobacterial cellwall.Pre-clinical testingwas carriedoutmore than20 years ago in dogs, and in humansmacrophage infiltration into osteosar-coma lungmetastaseswasobservedwiththis drug. It was not clear whether

patients survivedbetter, but tox-icity was manageable, withmainly fever and chills (JCO1992; 13:1310–16).This drugwas taken forward

into aprospective trial in theUS– the INT0133 trial.However, atthe time the trial was designed,people thought that asking aquestiononly aboutMTPwouldbe too simple, so there was asecond randomisation of ifos-famide versus no ifosfamide.Thiswas added tocisplatinpost-operatively, but in the preopera-tive phase patients had either

e-GrandRound


PROSPECTIVE TRIALS MAY TELL ONLY PART OF THE STORY

Patients with primary metastasesor tumours of the axial skeletonhave a much worse prognosis thanthose with localised limb tumoursSource: S Bielack, H Jürgens,

G Jundt et al. (2009) Osteosarcoma:

the COSS experience. Cancer

Treat Res 152:289–308, reprinted

with kind permission. © Springer

Verlag 2009

FIVE-YEAR SURVIVAL RATES

Progress in survival for patients with osteosarcoma has beenminimal since the early 1980sEuropean data: Stiller et al. (2006) Eur J Cancer 42:2124–35,

North American data: Mirabello et al. (2009) Cancer 115:1531–43

Europe (EU & others) North America

ifosfamide or cisplatin. In the end, therewere four arms: onewith ifosfamide andMTP, onewithonly ifosfamide, onewithonly MTP and one with neither. MTPwas given 48 times.The results for 667 patients pub-

lished in 2005 (JCO 2005; 23:2004–11) showed that the addition ofifosfamide to standard chemotherapydid not enhance event-free survival. Thethree-year event-free survival rate was68% for patients receivingMTP but noifosfamide, compared to71% for patientswho receivedno ifosfamide andnoMTP.Overall, adding ifosfamide to standardchemotherapy did not improve event-free survival. The authors suggested thatadding MTP to chemotherapy mightimprove event-free survival, but therewas interaction between the two ran-domisations to ifosfamide and MTP,precluding definitive statements.A secondpublication showedsix-year

event-free survival of 64% for patientstreatedwithneither ifosfamidenorMTPand63% for those given additionalMTPbut still no ifosfamide (JCO 2008;26:633–638), so therewasnodifferencefavouringMTP observed in patients nottreatedwith ifosfamide.Patients seemedto do better with ifosfamide plus MTP(71%) compared to ifosfamide with noMTP (58%). Combining arms showedthe six-year event-free survival was 61%for non-MTP arms and 67% for MTParms,whichwasnot significant.However,overall survival for the combined MTParms was statistically better than for thecombined non-MTParms (78% vs 70%,P=0.03).Theauthors said theycouldnotprove interaction, soconcluded therewasno interaction.L-MTP-PE (mifamurtide) is now

licensed in Europe, but a license wasrefused in the US because the FDAconsidered there was not sufficient evi-dence of a survival advantage. In a letterto the JCO, several leaders of interna-tional osteosarcoma groups said they

considered it was an interesting agent,but that additional clinical evaluationsare required before it can be consideredfor routine use (JCO 2008; 26:3102–03). I think that the information is notsufficient to use this agent as a part ofroutine treatment today, but we shouldcontinue to study it, preferably in a ran-domised prospective trial. An interna-tional group thatmet inLondon in 2010concluded that anMTP trial should beperformed, comparing chemotherapywith and withoutMTP.Other options for trials include: opti-

mising chemotherapy schedules; inhaledGM-CSF to enhance immune responseto osteosarcoma cells (but thiswas triedand failed; Clin Cancer Res 2010;16:4024–30); IGF-1R inhibitors, whichshowed positive in vitro results, but nopositive phase II data so far in osteosar-coma; mTOR antagonists, with (some-what) positive phase II data in bonetumours (JCO 2012; 30:78–84); bis-phosphonates, which are being tried invarious trials; and a rank ligand inhibitor,denosumab. Most of these, apart frombisphosphonates, are not yet advancedenough to go into phase III trials.

IN SUMMARYThe take homemessages are:� Exact staging for osteosarcoma ismandatory.

� Cure requires good surgery and goodchemotherapy,which should includeseveral of the four standard agents,which are doxorubicin, cisplatin,methotrexate and ifosfamide.We donot know the value of additionaldrugs.

� Intergroup collaboration is helpfulto get to results and is also feasible,although not easy. For the future, weneed biology-driven questions andwe must work to ensure that inter-group studies come up with biologyresults that can lead to biology ques-tions for future trials.

Question:What is the current standard ofcare for osteosarcoma now that theEURAMOS trial is closed? This relatesparticularly to patients who have a poorhistological response to three-drugchemotherapywith cisplatin, doxorubicinandmethotrexate. There is always a temp-tation to give patients different chemother-apy postoperatively if they have had a poorhistological response, but would you saythat the standard of care should still bethree-drug chemotherapy in this subset?Answer: Yes. I think that whatever well-chosen chemotherapy you give pre-operatively should be the standard forpostoperative treatment as well. Nobodyhas been able to prove that changingchemotherapy by making it more aggres-sivewill ultimately alter the disease course.Wewill have towait awhile to seewhetherintensificationwith ifosfamide and etopo-side will result in a higher cure rate.Question: Is there a role at all for high-dosechemotherapy and stem cell support forosteosarcomas in an inoperable site.Answer:No!Question:RegardingMTP, do you thinkwe would be having the same degree ofsoul searching about its use if it was lessexpensive?Answer:Cost is one thing. But there aretwo other issues that concern me. One isthe additional burden for the patient.They will have to go to the hospital 48additional times for their infusions – thisis 48 days of their lives, which is a lot forpatients who are living their last days. Iwould rather they do other things than sitin a hospital being treatedwith a drug thatmay not work. The other issue is toxicities,whichwe need to considerwhen planningfuture trials with biologic agents. It will bemore difficult to add them to a standardregimen that includes MTP than to onethat does not. If we addMTP to the stan-dard regimen we should be quite surethat it is a drug that really benefits thepatient, otherwise it will bemore difficultto move forward.


e-GrandRound

patient advocates, statisticians, epidemi-ologists, pathologists and representativesfrom cooperative trials groups and phar-maceutical companies – spent two daystrying to find commonground onhow toconduct clinical research where patientpopulations are small.Organised byESMOandRareCan-

cers Europe, the conference had twoaims: to bring everyone involved in rarecancers behind a research strategy thatcould significantly speedup thedevelop-ment of an evidence base, and to build aunited front that can be used to seekagreement on how regulators and payerscanbettermeet theneedsof this groupofpatients, for whom traditional standardsof evidence are difficult to achieve.

Can the Reverend Bayes helpdeliver proven therapies for

patients with rare cancers?

� Anna Wagstaff

When you are diagnosed withcancer, the last thingyouwanttohear is thatmedical experts

have few treatment options to choosefromandnotmuchevidenceonwhich tomake that choice. Yet this remains thereality formanyof the fourmillioncancerpatients in theEU-27whoare livingwitha ‘rare cancer’.There are 186 of these rare cancers

(using the recently proposed definition ofacancerdiagnosed in fewer than6 inevery100,000 people per year), with the seem-inglycontradictory result that almostone infour cancer patients has a rare cancer.The consequences can be seen in

their markedly poorer prognosis. Five-year survival figures – which broadly

speaking reflect theefficacyof treatment– showpatientswith rare cancers do sig-nificantly worse, with fewer than halfsurviving for five years (47%), comparedwith almost two-thirds for patients withmorecommoncancers (65%).While thismay inpart reflect the inherent nature ofthese particular cancers, it is also proba-bly a result of thecomparativedifficultyoflearning about how to treat rare cancers.Howto advance thecauseof this dis-

parate group of patients, and unblockprogress in improving treatment strategiesand developing new therapies, is a ques-tion that is commanding the attention ofan increasingly coherent community spe-cialising in rarecancers. InearlyFebruary,around50ofthem–clinicians, researchers,


CuttingEdge

Conducting clinical trials in people with rare cancers is not easy when the numbers in a

small trial do not add up to convincing evidence. Now some researchers are pressing for

a new approach – using Bayesian trial designs tomake themost of available knowledge.

BREASTBREASTLUNGLUNG COLORECTALCOLORECTAL

CuttingEdge


show that the difference in outcomebetween the trial arms is big enough forit not to have come about by chance –the all-important P-value.The standard P-value required by

regulators – and payers – is P ≤ 0.05,which means the odds of the demon-strated differencehaving come about bymere chance is less than 1 in 20. Aschancewill alwaysplay a larger role in theoutcome where numbers are small(throwing a six two times out of four ismuchmore likely than throwing a six 20times out of 40), proving a difference isnot just chance requires largenumbers ofpatients: the smaller the difference, thelarger the number of patients required.

Where there are too few patients toprove that an observed differencecould not have come about by chance,the study is said to be ‘underpow-ered’. As a general rule of thumb, saidBruzzi, trials of therapies for patientswith early-stage disease require 500–5000 patients, while in advanced dis-ease the numbers are a bit lower, at300–1000.As a consequence, groups of

patients who cannot muster this levelof participation in a clinical trial riskbeing excluded from the world of evi-dence-based medicine: no research,no ‘standard of care’, no guidelines, noaccess to approved therapies.

A QUESTION OF NUMBERSPaolo Bruzzi, clinical epidemiologistat the National Cancer Institute inGenova, Italy, explained the nub of theproblem. According to the traditionalrules of medical evidence nothingcounts but the numbers – which isexactly what rare cancers don’t have.To demonstrate the value of a new

treatment or treatment strategy, youhave to show beyond doubt that it ben-efits the intended group of patients– better survival, better quality of life –more than a comparator, which is usu-ally the standard of care. Using thetraditional ‘frequentist’ approach, thisinvolves treating enough patients to

ILLUSTRATION:FREDVANDEELEN,WWW.ORGANISART.CO.UK

generate the scores, so as to minimisethe risk of bias. The credibility of theresult canbe further testedby subjectingthe model to a sensitivity analysis: con-troversial values can be changed, or partof the evidence can evenbe erased to seewhat impactmore (or indeed less) scep-tical assumptions would have on thefinal probability distribution.

PATHOLOGICAL PATHWAYS ANDPRIOR PROBABILITIESBuilding a consensus around the use ofBayesian approaches to clinical studiescould be key to giving patients with rarercancersaccess toawhole rangeofnewbio-logical therapies, theconferencewas told.This is becausebiologicals that target

mutational pathways, or combinations ofpathways, are rarely specific to a singlecancer,whichmeans that there is aprob-ability that a therapy developed and

and has never yet formed the basis forapproval of a new therapy by the regula-tors or indeed reimbursement.Oneof thetasks this conference set itself, there-fore, was to start building agreementaround rules for defining prior probabil-ities that can command confidence.The proposal put to the conference

suggested a scoring system for ratingstudies for their validity and pertinence,“so that the assumptions of all calcula-tions are explicit and can be criticised”.A study in an identical patient popula-tion would score higher on pertinencethan onewith the same cancer but at anearlier stage; a well-designed trial witha control arm would score better onvalidity than a study that had used his-toric controls (results from an earliersequence of patients) or none at all.The proposal suggests using a trans-

parent and open consensus process to


CuttingEdge

THE BAYESIAN ALTERNATIVEPrompted by the need to develop anevidence base for treating these smallergroups of patients, doctors, researchersand drug developers have begun to turnto an alternativemethodology.Originat-ing from a theorem developed by anEnglish priest and mathematician, theReverendThomasBayes, and first pub-lished in 1763, theBayesian approach tomodelling probability has one greatadvantage over frequentist approaches:it enables all types of relevant informa-tion to be fed into the probabilitymodel.Using a frequentist approach, you

may have awell-designed and rigorouslyexecuted randomised controlled trialthat comes up with impressive results,yet fails the test of significance becausetoo few patients were spread across thetrial arms to demonstrate that the resultcould not have come about by chance.End of story.Using a Bayesian approach, how-

ever, the results of that same trial couldbe looked at taking into account thestrength of ‘prior’ evidence – relevantinformation that could have been gath-ered from any number of sources: bio-logical and preclinical studies, casereports, uncontrolled studies, studieswith surrogate endpoints, studies onother similar cancers, or studies on thesame cancer in different stages.The advantages of this approach

when working with small groups ofpatients is clear. Thedisadvantage is thatthe process of defining the strength ofprior evidence is open to subjectivity andtherefore to potential bias, which is whyit has been regarded with scepticism bythe scientific andmedical establishment,

“As a rule of thumb trials of therapies for patients

with early-stage disease require 500–5000 patients”

Almost 200 types ofcancer are each diagnosedin fewer than 6 out ofevery 100,000 people,every year, and thus fit therecently proposeddefinition of a rare cancer– some of these areexceptionally rareSource: RARECAREproject on surveillance ofrare cancers in Europe(www.rarecare.eu).Slides courtesy of AnnalisaTrama, Istituto Nazionaledei Tumori, Milan

FAMILIES OF RARE CANCERS

approved to treat patients with one ofthese cancers could also be of benefit topatientswithcancers that share the samemutatedpathways.Bayesianmethodolo-gies allow drug developers to take intoaccountknowledgegained in trials inoneindication when investigating the samedrugusedagainst the samepathwaysbutin a different indication. If regulators,and indeedpayers, areprepared toacceptrigorouswell-designedBayesianstudiesasabasis for approving access to themarketand reimbursement, this could substan-tially reduce the number of patientsneeded toprovide thenecessaryevidence.This in turnwouldmake it commerciallymore feasible to develop the drug evenwhere the potential market is small, andwouldalso cut the time taken forpatientsto get access to the drug.This new paradigm for developing

drugs across tumour types that share amutational pathway has a number ofadvantages, commentsAndrasFehervary,head of Market Access for NovartisEurope. “By predicting response, it

mental changes to the cur-rent system thatwouldmakeit easy to conduct trials of tar-geted drugs in patients withcancers where the targetedpathway is known to play arole, particularly when thatdrug has already beenapproved in one indication,andwhere the disease is seri-ous and there are few or notherapeutic options. Such asystem, he suggests, could bebased on numerous, fairlysmall, investigator-initiatedtrials; aneffectivewayof iden-tifying eligible patients; and

agreement fromboth regulators andpay-ers on allowing information gatheredboth before and after the trials to formpart of the overall evaluation of the effi-cacy and value of thedrug in that setting.He paints this scenario:“Assume we are working within

INCa [the French cancer research net-work], and assume a patient has notbeen accurately diagnosed, but is clearlysuffering cancer-related symptoms. Thepatient is sent to the Institut GustaveRoussy, and is comprehensively screenedagainst a range of biomarkers. Thatpatient is identified as potentially suf-fering from a rare cancer, and there is aclinical trial running in oneof the21cen-tres linked to INCa specifically for thatform of cancer. The patient is veryquickly moved into that trial, and ismatched [confirmed to have the relevantdiagnosis for the trial]. The trial is set upto run on Bayesian principles.“Assume thepatient respondswell to

the treatment, and that these resultscontribute to the overall clinical trial

reduces the number of patients neededin clinical trials; by determining responseas early as possible, itmeans trials canbeconcluded faster; and by predicting notonly activitybut also adverseevents it pro-vides the basis for ‘companion diagnos-tics’ that can be used in routine clinicalpractice to see which patients wouldbenefit most, and which might sufferthe greatest toxic effects. It would alsoaccelerate development of new drugsand reduce attrition.”It’s awin–winscenario, saysFehervary,

which aims at getting the right therapy atthe right dose for the right patient at theright time–a goal sharedby the industry,healthauthorities,physiciansandpatientsand their associations alike.“For this tohappen inasustainableway,

we do need to cooperate to create a moreefficient system, whichmust be a patientcentric, andpatient outcomescentric, sys-tem,” he adds. “And this calls for newmodels of collaboration between industryand its partners on important steps.”Fehervary would like to see incre-

CuttingEdge


Building a consensus around Bayesian approaches could

be key to giving patients access to new biological therapies

SURVIVAL IS POORER FOR RARER CANCERS

Speeding up developmentof new therapies for rarercancers will be essentialto closing the starksurvival gapSource: RARECAREproject on surveillance ofrare cancers in Europe(www.rarecare.eu).Slides courtesy of AnnalisaTrama, Istituto Nazionaledei Tumori, Milan

results that show the drug is an activemolecule and effective in that setting. Inprinciple, assume there are30patients inthat group, we should be able to go toEMA[theEuropean regulators] and say:‘Wehave an activemolecule that shouldbemade available to patients, but hasn’tgone through the full safety tests thatcome through the larger trials.However,the risk profile of a patient suffering froma rare cancer is different, because of thepoorer outlook for rare cancers, andpatientswant access.’Hypothetically thiscould lead to approval.”Fehervary also mentions the need

for greater involvement ofpatients groups in clinicaltrial design and execution, astronger focus on patientadherence to their prescribedtreatment, better manage-ment of side-effects,and a more equitableaccess to drugs and tooptimal standards oftreatment as importantareas for improvement.

BIGGER IS ‘NOTALWAYS BETTER’Paolo Casali is a medicaloncologist who has spent most ofhis career trying to improve out-comes for patientswith sarcomas –rare cancers now thought to consistof more than 50 (even more rare)subtypes. One of the key organisersof the conference, he is an avowedBayesian enthusiast, and insists that justbecause a trial is small this does notmean it has to be eithermethodologicallyunsound or inconclusive. He doesaccept, however, that the smaller the

trial the more important is a rigorousmethodology: transparent, pre-agreed,open to sensitivity testing. And that isexactly why it is so important to build aconsensus around how this can beachieved.Casali points out that, over the years,

more than a few drugs have beenapproved for small populations on theback of a fairly poor evidence base – forinstance trials thathadnocontrol arm, orthat provide data only on tumour shrink-age. In these cases, where the regulatorsaccept that it would not

be possible to run a fully powered phaseIII randomised controlled trial, pharma-ceutical companies (or other sponsors ofnew drugs) are always uncertain abouthow much evidence regulators willdemand to back up the application forapproval, and this may deter them fromdeveloping drugs for rare indications.Howmuchbetterwould it be tohave

trials designed according to agreedBayesian principles from the outset,arguesCasali. Insteadofbringing inaddi-

tional informationat theendofthe trial, to be assessed, eval-uated and applied by theteamof regulators as they seefit, that same informationwould have to be submit-ted in advance, with a con-sensus over its validity andpertinence, transparencyabouthow that consensuswas reached, and a sensi-tivity analysis on moreuncertain assumptions.Casali also ques-

tions the received wis-dom that bigger trialsare always necessarilybetter. “In rare cancersin particular, large col-laborations inevitablylead to involving a large

number of institutionswhose skills in thedisease are limited, which has implica-tions for the quality of carewithin thesestudies,” he says.Standard quality control measures

used in large trials, such as centralpathology review and central review ofscans, are useful but introduce theirown problems, and will never cover allaspects of good-quality care saysCasali.


CuttingEdge

“Large collaborations inevitably lead to involving

institutions whose skills in the disease are limited”

The Reverend Bayes and his theorem. Thepotential for using our knowledge of cancerbiology to speed up the evaluation of newtherapies and treatment strategies is promptingrenewed interest in a theorem that was firstpublished almost 250 years ago in PhilosophicalTransactions under the title ‘An Essay TowardsSolving a Problem in the Doctrine of Chances’(Phil Trans 1763; 53:370)

academia has first a role towork togetherto assess the feasibility of conclusivetrials using the most robust methodol-ogy... Research groups should avoidapplying whenever possible what canpossibly be more debatable methodol-ogy,” he told the conference.Where patient numbers are small,

says Lacombe, the answer is to seekinternational collaboration, if necessarybetween collaborative groups.One suchcollaboration, which answered severalquestions about the best use of temo-zolomide in patients with the aggressivebrain tumourglioblastoma, involved threemajorNorthAmerican groups –RTOG,theNCCTGand theCanadianNCI– inaddition to the EORTC. Lacombe doesnot deny the challenges posed by thesesorts of collaborations – a lot can gowrongwithoutmeticulous planning andunrelenting efforts to keep everyone instep every step along the way. But thisshouldnot be anexcuse for not trying, heinsisted, and he cautioned against anyrecommendations that could be inter-preted as a green light to do “local, smalland inconclusive initiatives.”Matt Seymour, director of the UK’s

National Cancer Research Network(NCRN), and a specialist in gastro-intestinal cancers, took a similar line, butadded that there would be a lot lessneed for huge international collabora-tions ifmore countriesmademore con-sistent and concerted efforts to increase

“Every clinician knows that tumourresponse assessment involves complexclinical reasoning, as does every clini-cal decision. Clinically speaking, bydefinition, ‘blind’ central reviews,which skip clinical data, will lead to aworse tumour response assessment,not a better one.”“There has to be some trade-off

between the methodological require-ments and clinical quality, otherwise,clinicians will not believe in their ownstudies,” arguesCasali, adding that thisis precisely what has happened withseveral adjuvant randomised trials and“basically all” randomised trials com-paring adriamycin against adriamycinand ifosfamide in soft tissue sarcomas.“In fact, many sarcoma experts cur-rently rely on small uncontrolled stud-ies for their everyday decisions onmedical therapy more than on largerandomised trials, even though theseare in fact available.”

BAYESIAN IS ‘SECOND BEST’Denis Lacombe, director of the head-quarters ofEurope’s largest clinical trialsorganiser, theEORTC, is distinctly cau-tious about the use ofBayesianmethod-ologies for gettingnewdrugs approvedorextending theiruse tonew indications, onthe basis of results from small trials.“While alternative designs should be

investigated to allow therapeutic progressfor these patients [with rare indications],

the proportion of cancer patients treatedwithin trials,The UK tried it, he said, and as a

result “over the last 10 years the numberof cancer patients enrolled in trials hasincreased five fold, to one in every sixcancer patients –matching the trial pop-ulation of thewhole ofNorthAmerica.”One consequence of this, he says, isthat theNCRNwas able to recruit suf-ficient patients from theUKalone to runlarge randomised controlled trials insome very rare cancers, including onedemonstrating efficacy of palliativechemotherapy in glandular carcinomaand another in anal cancer demonstrat-ing the efficacy of chemoradiation.International collaboration will still

be essential particularly for very rarecancers, added Seymour, and indeed hewas instrumental in launching the Inter-national RareCancer Initiative last yearas a partnership between CancerResearch UK, the EORTC, the UK’sNCRN and the US National CancerInstitute. This group focuses principallyon trials in cancer indications with nomore than 3 new cases annually per100,000. Examples include trials in theadjuvant and advanced setting compar-ing treatment strategies for patientswithsmall bowel adenocarcimona – a cancerwith only around 6newcases diagnosedper 1,000,000 each year.“We need to be ambitious; we can

achieve randomisedcontrolled trials in an

CuttingEdge


He cautions against allowing Bayesian designs

to be used “as an excuse for underpowering”

“We need to be ambitious; we can achieve randomised

controlled trials in an international setting”

international setting. Where protocolsare well designed, well written and veryclear, you can achieve high quality evenfor rare cancerswherecentres areputtingsmaller numbers of patients.”Seymour accepts, however, that some

cancers are so rare that, evenwith inter-national cooperation, it is simplynot pos-sible to recruit enoughpatients to answera trial questionwithin a reasonable lengthof time using the traditional frequentistmethods and significance levels.It is only in these situations that he

would consider turning to Bayesianmethodologies, and even then, only

where some “genuinely credible” priorevidence is available, and only whereall prior evidence is agreed by everyonebefore the trial starts. LikeLacombe, hecautions against allowing Bayesiandesigns andprior probabilities to beused“as an excuse for underpowering”.

WHERE NEXT?The challenge in the coming monthswill be to amend the draft recom-mendation to reach a consensus onthe way forward for methods for clin-ical research into rare cancers thatsatisfies the needs expressed by

Casali, while addressing the concernsexpressed by Lacombe and Seymour.The hope is to publish a consensusdocument in the Autumn.Roger Wilson, who is currently in

treatment for recurrent myxofibrosar-coma, and is a former chair of the UK’sNational Cancer Research InstituteConsumer Liaison Group (and formerNCRI board member), says it will be aquestion of striking the right balance.“What I as a rare cancer patient

need is for my scientists to recognisethat there are benefits in bothapproaches and that they need to findthe balance which delivers patientbenefit. The scientist who drivesthrough a 10-year study in a rare can-cer to deliver a result that has beenovertaken by a clinical developmentwhich he knew nothing about when hedesigned the study is not a badscientist, just a brave and unlucky one.The scientist who used a Bayesianapproach and was able to adapt hisstudy to account for the new develop-ment and then delivered a result whichis more relevant clinically at the timeit is completed is not just a lucky sci-entist – he made his luck.”Doing nothing is not an option, he

insists. “We need our researchers andscientists to start to use Bayesian tech-niques today in rare tumours so we canassess what issues it raises, if any.”Casali agrees, and he told the confer-ence that during the course of thisyear, together with colleagues fromthe worldwide sarcoma community,he will be starting some Bayesiandesigned studies on new agents insarcomas.

TARGETING PATHWAYS NOT LOCATIONS

Glivec (imatinib), the first tyrosine kinase inhibitor, blocks the activity of abl, c-Kit and theplatelet-derived growth factor receptor PDGFR. It was initially approved to treat patients withchronic myeloid leukaemia. Later this was extended to patients with Kit-positive GIST, andthen in 2006 Novartis submitted a single study for approval in five other rare indications.The company is now pursuing a similar strategy with its mTor inhibitor everolimus (Afinitor/Votubia), which has been approved for renal cell carcinomaand subependymal giant-cell astro-cytoma on both sides of the Atlantic, and additionally for pancreatic NET in the US. The drugis currently in phase II trials for four additional rare cancers.Xalkori (crizotinib) inhibits the tyrosine kinases ALK, ROS1 and c-Met, and was recentlyapproved in the US for treating patients with non-small-cell lung cancer with the ALK muta-tion. Pfizer is currently running trials of the drug in anaplastic large-cell lymphoma and neu-roblastoma. In Europe, crizotinib is also being investigated in the EORTCCREATE trial for usein anaplastic large-cell lymphoma, inflammatorymyofibroblastic tumour, papillary renal cellcarcinoma type 1, alveolar soft part sarcoma, clear cell sarcoma, and alveolar rhabdo-myosarcoma – all of them rare cancers.GlaxoSmithKline is now taking this paradigm one step further by proposing to study its owninvestigational BRAF inhibitor with an investigational MEK inhibitor in patients who expressrelevant mutations, regardless of the location of their tumour. According to an article byMichael McCaughan (Elsevier Business Intelligence, November 16, 2011), the expectationseems to be that the US regulators, the FDA, would not oppose this approach in principlein the case of applications for approval acrossmultiple rare cancers, but would be less opento the same approach across more common cancers.


CuttingEdge

“We need to start to use Bayesian techniques today in

rare tumours so we can assess what issues it raises”

Fixing the holes in theopioid supply lines� Simon Crompton

Patients are still dying in agony despite concerted efforts over many years to change attitudes

towards the use and control of opiates. Could a new initiative, which works with NGOs,

governments and policy makers to address practical problems, finally hit the spot?

When his cancer paingrew so great he couldsee no other means ofescape, former police

officer BernardNg fromSingapore con-sidered killing himself. But that changedwhen ahospice providedhimwith effec-tive pain relief.“I once gave up living,” he says, fight-

ing back the emotionduring an interview

for the feature film ‘Life Before Death’,releasedworldwide thisFebruary. “I oncetold my wife, and myself, ‘I don't havequality of life –what is the use of living?’But today I see it the otherway. Iwant togo on living. With medication and thedoctor’s help I'm okay."“I'mnot asking for a lot. I justwant to

live a normal life without pain. And ifpossible I can do all the basic things

like take a bath, changemy own clothes,you know, without bothering my wife."According to the Union for Inter-

national CancerControl (UICC),morethan 3.3 million people with cancerare dying in pain, sometimes in agony,each year.Altogether, tens ofmillions ofpeople are needlessly suffering pain,and a WHO estimate says that 600million are going to suffer from


Spotlighton...

Spotlighton...


easily available and physicians hadnothing to give patients, they had toprovide a story to explain it. That storywas that drugs like morphine are dan-gerous and addictive.And that story hascontinued to this day. Some cliniciansfind it easier to build a wall betweenthemselves and the patient, saying theycan’t do anything for the pain. Theybreed an acceptance that HIV hurts, orcancer hurts.”

THE INVISIBLE PROBLEMAccording to Meg O’Brien, GAPRIdirector, untreated pain has long beenthe invisible world health problem,always falling between the cracks ofother initiatives. It is common even incountries like the US and UK, butthe problem is most dramatic in sub-Saharan Africa, which has 20% ofpainful deaths in the world, and just1% of the morphine.

GAPRI has identified themain barriersto people accessing pain relief as legaland regulatory restrictions, weak healthsystems, and concerns about drug diver-sion and addiction.All of these interact.ButMeg O’Brien says that if she couldwave a wand to change one thing, itwould be to fix clinicians’ attitudes tomorphine.“There are so many myths and mis-

understandings about the drugs,” shesays. “The reality is that, inmany coun-tries, even if we were to fix regulationsthat restrict access to the drugs, we’d seethe drugs expire on the shelves becausethey were under-used.”“The reason is partly historical.

When, in the past, pain relief was not

untreated pain in their lifetime. Thereason is simple: lack of access to basic,cheap and highly effective drugs,notablymorphine.Around 70% of can-cer deaths occur in low and middle-income countries, where just 6% of theopioid analgesics are consumed.Behind that bald fact lies a complex

tale of global over-regulation, corpo-rate indifference, burdensome red tape,professional fear and pervasive misun-

derstanding. Drugs aren’t getting topeople, not because they are expen-sive or hard to administer but becausesystems are failing.Now an initiative from the Union

for International Cancer Control(UICC) and the American CancerSociety (ACS) aims to end this needlesssuffering. The Global Access to PainRelief Initiative (GAPRI), launched in2010, is spearheading a range of actionsto make effective pain relief a globalreality by 2020. It aims to improve themarket for pain medicines, empowergovernments to expand pain relief,strengthen health systems and main-stream the issue of pain treatment inthe global health agenda.

Practical solutions. Resolving an acute morphineshortage in Uganda came down to ensuring thata hospice with the knowhow also had a licenceand an incentive to supply the health service

MORGANAWINGARD

This is the experience in Zimbabwe,where fear and ignorance about mor-phine conspire with highly restrictivedrug control legislation to make goodpain control a rarity.According toDick-sonChifamba, executive director at theIsland Hospice Service, only 30% ofpeople who need pain relief get it inZimbabwe. “Adoctor has to report to theSecretary of Health any patient whohas been on morphine for more thanfour months,” he says. “So because ofthe bureaucracy a number of doctors donot want be bothered with prescribingmorphine.” Resistance is particularlygreat from older doctors, who have notbeen trained in palliative care.

SUPPLY-SIDE ISSUESThere are supply problems with mor-phine too. Chifamba points out thatUN conventions designed to controlthe illegal use of drugs have ingrainedhighly restrictive systems. Supply prob-lems are exacerbated by the fact thatthe big players in the pharmaceuticalindustry show little interest in produc-ing morphine because profits comefrom patented medicines. That leavesit to small companies, with more lim-ited capacity, to supply middle- andlower-income countries.David Lee, an expert on opiates

and senior strategic advisor for EndoPharmaceuticals, acknowledges that


Spotlighton...

“There are also the usual fears of addic-tion and respiratory depression, whichmeans that both doctors and patientsseemorphine as amedicine of very lastresort. They hold off until the patient isin agonising pain.” Chifamba recallsrecently visiting a cancer patient whowas screaming in agony.Hewas alarmedbecause on his last visit the patient’spain had beenwell controlled.Whenheasked the relatives what had happened,they said they had not collected themorphine doses because they believedit was slowly killing the patient. Manyfamilies, he says, believe thatmorphineshould not be given continuously – onlywhen pain is extreme.

“Because of the bureaucracy a number of doctors

do not want be bothered with prescribing morphine”

MOONSHINEAGENCY

On top of this, national or local legalframeworks to get morphine where it isneededare eithernon-existent or so cum-bersome that health professionals arenot prepared to take themon. “There’s areluctance toprescribebecause of fear oflegal sanction,” says Lee.

if there were more profit in morphinethe global dynamics of supply anddemand would be very different. “It’scertainly a lowmargin business for thepharmaceutical industry. But I don’tthink you can say the whole situation istheir fault.”“In rich countries, there are many

types of opiates, but for many low- tomiddle-income countries morphine isthe only option for severe pain,” saysLee, who is CEO of the LAMBPainFoundation and leads a projectmanaged by the Foundation forHospices in Sub-SaharanAfrica and the African Pal-liative Care Association toeducate hospital adminis-trators and healthcareworkers in good painmanagement and treat-ment. “It’s cheap in prin-ciple, and the process toconvert the raw materialis relatively easy. Butmovement of it is heavilycontrolled by internationaland national laws.”Influential governments

waging wars on illegal drugsare putting political pressure ongovernments around the world notto loosen laws controlling distribu-tion. “There is a lot of emphasis oncontrol, and less on controlled avail-ability,” says Lee.

If there were more profit in morphine the global

dynamics of supply and demand would be different

Spotlighton...


IMAGESFROMMOONSHINEAGENCY

70%

Around 70% of cancer deaths are in poorer countries,

where just 6% of the opioid analgesics are consumed

A NEW DETERMINATIONBut if the global obstructions to goodpainmanagement are intricate anddeep-rooted, there are at least now signs of agrowing international determination toovercome them.First and foremost was the establish-

ment of GAPRI itself – the brainchild ofDavidHill,UICCPresident (2008–2010),

6%

FROM WORDS TO OUTCOMESDavid Lee welcomes the new politicalwill to act. Butwhatwill turnwords intoactions, and actions into outcomes?“All this high-level, professional noise

is helpful because there’s a referencepoint, an anchor,” he says. “It nowneedspeople who have the ability to do some-thing, to do something. There are a lot ofwell-intentioned people wanting tochange things, but one of the frustrationsof the not-for-profit sector is that some-times people work against each other,rather than working together. There aresigns now that it’s beginning to happen.I think that the GAPRI initiative is par-ticularly important in this respect,because of the way it interacts with somanydifferentNGOs, governments andother bodies.”Meg O’Brien too believes that

GAPRI is getting things moving. “It’s abit like a leaky pipe – you have to fix allthe holes at the same time, or else thewater starts coming out somewhereelse,” she says. Issues of availabilityand training of health professionals inpain control have to be addressedsimultaneously. GAPRI is doing thisby developing good-practice modelsthat will prove what can be done andwill spread knowhow.The pain-free hospital initiative, for

example, is a one-year programme tochange clinical practices at key hospitalsworldwide during 2012. It providestraining for physicians and nurses toevaluate and treat pain, using targetedadvocacy campaignswithin the hospitalto increase awareness of treatable pain.Simultaneously, it works with govern-ment officials to ensure that essential

who pushed the idea ahead in the light ofincreasing concerns about pain control inthe international cancer community.GAPRIhas takenshapealongsideanotherproject which has helped give voice,images and publicity to its campaign – afeature film called ‘Life Before Death’about the global crisis in pain, producedand directed by David Hill’s son, Mike.The award-winning documentary, filmedin 11 countries and released this year,interviews health professionals battlingthe pain epidemic and patients who haveexperienced extreme pain, and argues foreffectivepain treatment as abasichumanright (www.lifebeforedeath.com).Both initiatives have taken place in

the context of newexpressions of a globalwill to address palliative care issues. Lastyear’s UN General Assembly politicaldeclaration on non-communicable dis-ease stated that national policies shouldimprove access to palliative care and

foster partnerships between governmentand civil society to support palliativecare services.TheWorldCancerDeclaration,drawn

up by UICC in 2008, is more specific.Oneof its targets for 2020 is that effectivepain control measures will be availableuniversally to all cancer patients in pain.The declaration calls for action whereover-regulationhinders goodpaincontrol,andstipulates that governments and inter-national organisations should not allowglobal implementation of theUN’s inter-national drugcontrol conventions to inter-fere with legitimate efforts to advanceaccess to painmedicines.A hard-hitting report from theGlobal

TaskForceonExpandedAccess toCancerCareandControl inDevelopingCountrieshashighlighted the fact that “paincontrol,an issue for all cancers andmanyotherdis-eases, offers themostdistressingand insid-ious example of the cancer divide.”


Spotlighton...

“It now needs people who have the ability to do

something, to do something”

The Drugs Control Office, Kerala, India. Doctors often choose not to prescribe morphine because theycan’t face filling out all the permissions forms. This particular state recently overhauled its procedures,but elsewhere in India, as in most countries across the globe, the bureaucracy remains prohibitive

MOONSHINEAGENCY

HOST A SCREENING EVENT

Filmed in 11 countries across North America, Europe, Asia, and Africa, the documentaryfilm ‘Life Before Death’ tells the stories of the health professionals battling the sweep-ing epidemic of pain that threatens to condemn one in every ten people to an agonisingand shameful death.It shows how their struggle pits them against indifferent governments, dysfunctional bureau-cracies, over-zealous law enforcement agencies and, above all, the deep-seated attitudesof those around them. Their mission is to change the culture of medicine to becomemorefocused on care, rather than exclusively on cure.Through the eyes of patients and their families the film reveals the humanity that empow-ers people to care for those beyond cure. It uncovers hard truths about the torture occur-ring every day in hospitals around the world but also of the immense hope that comes fromhealthcare pioneers who accompany terminal patients on their journeys to dignified end-of-life experiences. It is a film about livingwell and dying better,making themost of everymomentin life before death.Health professionals and advocacy groups around the world have organised publicscreenings of this film to raise awareness of the issue and galvanise support forchange. If you are interested in hosting such a screening, copies of the film are avail-able from GAPRI, together with promotional materials and ideas for preparing dis-cussions and further action.For details on this and other ways you can help, go to http://www.treatthepain.com/do-something

all who needed it. Not only that, butthe Ugandan government was saving40% on what it had previously had topay for morphine.“It wasn’t rocket science to sort

out,” says O’Brien. The problem wasgetting existing stores of morphinepowder turned into liquid. So O’Briennegotiated new arrangements wherebya local hospice that already manufac-tured liquid morphine was licensed tosell to government. The hospice bene-fits by making a profit, and the supplychain is safe.“InUganda, we did the contracting,

pricing and forecasting –we even founda plastics manufacturer for the mor-

painmedicines are available at all times.GAPRI has launched the project inthree hospitals in India, and is workingto find funding for similar projects inVietnam, Haiti, Tanzania, Ethiopia,Turkey and Egypt.“Wemake filmswith patient stories,

have teeshirts for patients to wear, andthe idea is that everyone gets a feelingthat pain can be controlled,” saysO’Brien,who before starting full-time atGAPRI in 2010 had been involvedwithpain management initiatives as part ofthe Clinton Foundation. “Once we canshowhowpain scores are droppingwith-out addiction, we have a good demon-strative model that can spread throughthe whole health system.”GAPRI is addressing theproblemsof

global supply head-on, and will be cre-ating a new analysis ofmorphine supplyand demand to take direct to pharma-ceutical companies to show themwherethe potential markets can be found.With the aid of this, says O’Brien,

GAPRIwill be able to facilitate “bundlednegotiations”with pharmaceutical com-panies onbehalf of several governments.“We’ll act as the middle man,” she says.Recently, GAPRI has been working

to improve pain relief inUganda, wheremorphine shortages resulted inrationing andmany people in pain turn-ing up at health centres literally beg-ging for the drug.“I went there with colleagues for

four days, because I couldn’t get a goodunderstanding of what the problemwas by email. We went from place toplace, meeting 25 different people andorganisations, mapped out the problemand put together a plan.” Threemonthslater, morphine was easily available to

phine bottles. Finding the time and per-sonnel for those sorts of jobs isn’t alwayseasy.” GAPRI is recognising this by run-ning a fellowship programme, placingpeople in health ministries for threeyears to try and prevent pain manage-ment falling between the policy cracks.With the GAPRI initiative, there’s

finally the hope that good intentionsare being turned into practicalmeasuresand good outcomes. It’s a matter ofactually demonstrating what changecan achieve, and then encouraging goodpractice to spread. “A lot of campaignsstop at raising awareness, but some-times people needmore help than that,”says O’Brien.

GAPRI is running a fellowship programme,

placing people in health ministries for three years

Spotlighton...


I M P A C T F A C T O R

ImpactFactor

Localised non-bulky Hodgkinlymphoma – future questions

� Bertrand Coiffier and Olivier Casasnovas

Late toxicities from radiation therapy are frequent in patients withHodgkinlymphoma and can hamper survival. These late toxicities should decreasewith modern radiation therapy, but results are not mature and so theimportance of this decrease is still unknown.Hence, all studies inHodgkinlymphomamust report long-term outcome.

In a recent publication, Meyer etal.1 presented a 12-year follow-upof patients with localised non-

bulky Hodgkin lymphoma included ina study that compared chemotherapyto a radiation-based treatment.1 Inclu-sion criteria in this study – previouslypublished with a short follow-upperiod2 – were not-too-low risk patients(stage IA with one involved node anderythrocyte sedimentation rate [ESR]<50 mm were excluded) but not-too-high risk (patients with tumour diam-eter >9 cm, a tumour larger thanone-third of the chest wall diameter or

with intra-abdominal disease wereexcluded).2 The study design was quitecomplicated and divided the patientsinto a chemotherapy arm and a radia-tion arm.After randomisation, patientsin the chemotherapy arm received doxo-rubicin, bleomycin, vinblastine anddacarbazine (ABVD); patients with acomplete remission or unconfirmedcomplete remission after two cyclesreceived four cycles in total, and theremaining patients received six cyclesin total. Patients assigned to the radia-tion armwith at least one unfavourablerisk factor (>39 years old,ESR>49mm,

more than three disease sites, or mixedcellularity or lymphocyte-depleted his-tology) received two cycles of ABVDbefore radiotherapy, whereas thosepatients with no risk factor receivedonly radiotherapy (subtotal nodal radi-ation therapy). The study was openedto enrolment in January 1994 and ter-minated inApril 2002, but only 405 ofthe 450 patients had completed enrol-ment. The decision to terminate enrol-ment was taken by theData and SafetyMonitoring Board because by that timethe radiation protocol was outdated.This trial was a complicated study withtoo many possible biases and difficult-to-interpret results that would likelyhave had little effect on the existingpool of Hodgkin lymphoma trial data.The first results with a 4.2-yearmedianfollow-up period showed a significantlybetter progression-free survival (PFS;or freedom-from-disease progressionas it was called in the study) forpatients randomised to the radiationarm, with a similar overall survival inboth arms but a slight increase of deathfrom causes other than Hodgkin lym-phoma in the radiation arm.2

This study was saved by the lateanalysis, even though 14% of the


This article was first published in Nature Reviews Clinical Oncology vol. 9 no.3, and is published with permission.© 2012 Nature Publishing Group. doi:10.1038/nrclinonc.2012.7

CLINICALONCOLOGY

ImpactFactor

patients were lost to follow up;1 lateresults (median follow-up period 11.3years) showed a lower 12-year PFS(87% vs 92%; hazard ratio [HR] 1.91;P=0.05) but an improved 12-year over-all survival rate (94% vs 87%; HR 0.50;P=0.04) for the patients in thechemotherapy arm compared with theradiotherapy arm. This longer overallsurvival was related to a lower numberof patients dying from causes otherthanHodgkin lymphoma (12 deaths inthe ABVD arm versus 24 in the radia-tion arm). These numbers will likelycontinue to increase because the num-ber of secondary cancers is muchhigher in the radiation-based arm thanthe chemotherapy arm (23 vs 10).These results raise several questions:first, what is a good balance betweenchemotherapy and radiotherapy inpatients with localised Hodgkin lym-phoma? Second, is it possible to reducethe intensity of therapy in somepatients? Third, when can results froma randomised study be considereddefinitive in patients with Hodgkinlymphoma? Finally, what is the bestendpoint for future studies in patientswith Hodgkin lymphoma?The current treatment for localised

Hodgkin lymphoma – a short courseof chemotherapy plus low-doseinvolved-field radiotherapy – curesover 90% of patients.3,4 To increasethis cure rate, deaths after relapse orfrom other causes need to bedecreased or avoided. The treatmentof relapsed patients has improvedrecently with the use of high-dosetherapy with stem-cell transplants andnew drugs. The ABVD regimen wasassociated with few severe late com-plications; secondary myelodysplas-tic syndrome (MDS) or acute myeloidleukaemia (AML) are rare and thedose of doxorubicin is usually too lowto induce cardiac failure.5 By con-trast, MDS and AML are more fre-

quent with combined therapy, andsecondary solid tumours increase overtime after radiotherapy.6 The 30-yearincidence of secondary cancers withmantle radiation therapy is around30% but decreases by 60% to 12%with involved-field radiation therapy.The long term follow-up of anothertrial – EORTC/GELAH10 – will giveinsights on the risk of a secondarycancer after involved-node radiationtherapy. Cardiovascular complicationsare also more frequent after radio-therapy, even though they are lesscommon now with the standard use ofinvolved-field radiation therapy.7

Is it possible to reduce the use ofradiotherapy or to reserve it for a sub-group of patients with localisedHodgkin lymphoma? To date, two stud-ies comparing results of chemotherapyalone versus the combined modalityhave been reported with a short fol-low-up period; these studies demon-strated either no PFS benefit8 or amarginally better PFS9 for the com-bined modality and the same overallsurvival for both treatmentmodalities.8,9

To a certain extent, the Meyer et al.1,2

study also compared bothmodalities, as73% of the patients included in theradiation arm received a combination ofchemotherapy plus radiation.However,long-term outcome favours the chemo-therapy arm, the extended radiationarm being hampered by an excess ofdeath due to late toxic effects.1

On the basis of these three stud-ies,1,2,8,9 there is no clear evidence thatwe can safely omit a modern radio-therapy treatment in all patients withlocalised non-bulky Hodgkin lym-phoma because PFS results are con-troversial and data on long-term overallsurvival with current combined treat-ments are unavailable.Recently, response-adapted therapy

has emerged as a new concept that issupported by the development of func-

tional imaging. In this therapy design,patients achieving complete remissionas determined by 18FDG PET assess-ment after two chemotherapy cycleswill not receive radiotherapy, but thosewithout a complete remission will. Togeneralise this idea, randomised stud-ies must show that these patients withearly complete remission will not havea shorter survival than those receivingradiotherapy. Preliminary results ofPET relevance to identify patients eli-gible for radiotherapy are in favour ofthis hypothesis, at least in advanced-stage Hodgkin lymphoma.10 However,the majority of these studies are ongo-ing and definitive results have not yetbeen published. Involved-field radia-tion therapy remains the standard treat-ment for these patients until suchresults demonstrate that radiotherapyis not necessary in early responders.Furthermore, an additional issue toaddress is establishing suitable rules forinterpreting interim PET scan results.The trial published by Meyer et

al.1,2 is also remarkable because resultswere modified from the early2 to thelater1 report.Although PFS results didnot change, overall survival changedfrom the same in both arms to beingbetter in the chemotherapy arm,because of late toxic events in theradiotherapy arm. Clearly, for diseasesin which overall survival is very good,such as localised Hodgkin lymphoma,


Practice points� Radiotherapy is associated withlate toxic effects

� Long-term follow up (>10 years)should bemandatory inHodgkinlymphoma trials

� Chemotherapy alone might besufficient treatment for selectedpatients

ImpactFactor

results must not be reported early onand a minimum of 10 years is neces-sary to allow the analysis of the lateeffects and deaths caused by late toxiceffects.It is too frequently the case that

study reports from trials in patientswith Hodgkin lymphoma or non-Hodgkin lymphomas are publishedwith less than five years of follow up.These early results are important, par-ticularly if there is a difference in over-all survival, or if a potential change forclinical practice is reported, but theymust be called ‘preliminary’ and fol-lowed by the publication of matureresults.This recommendation for the pub-

lication of mature results leads to theevaluation of endpoints of studies thatassess the first-line treatment of treat-ment-naive patients. Assessment ofPFS allows the evaluation of the effi-cacy of the tested therapy, but not latetoxicity. When there is a large differ-ence between the two arms (larger

than 20%), the early results are usuallyconfirmed by late results; however,when the difference is small (less than10%) resultsmust be called preliminaryand need to be confirmed by otherstudies and/or by mature results.In summary, our first goal is to

cure patients with cancer, but whenlong-term survival is over 90%, weneed to look at the possible toxiceffects of treatment on survival. Allrandomised studies showing a benefitin the experimental arm must bereported with a median follow-uplonger than 10 years to allow thisassessment to be completed.

References1. RM Meyer et al. ABVD alone versus radiation-basedtherapy in limited-stage Hodgkin’s lymphoma. NEJMhttp://dx.doi.org/10.1056/NEJMoa11119612. RM Meyer et al. (2005) Randomized comparisonof ABVD chemotherapy with a strategy that includesradiation therapy in patients with limited-stageHodgkin’s lymphoma: National Cancer Institute ofCanada Clinical Trials Group and the EasternCooperative Oncology Group. JCO 23, 4634–423. A Engert et al. (2010) Reduced treatment intensityin patients with early-stage Hodgkin’s lymphoma.

NEJM 363:640–6524. JO Armitage. (2010) Early-stage Hodgkin’slymphoma. NEJM 363:653–6625. D Hodgson. (2011) Late effects in the era ofmodern therapy for Hodgkin lymphoma. HematologyAm Soc Hematol Educ Program 2011:323–3296. AJ Swerdlow et al. (2011) Second cancer risk afterchemotherapy for Hodgkin’s lymphoma: a collaborativeBritish cohort study. JCO 29:4096–41047. BM Aleman et al. (2003) Long-term cause-specificmortality of patients treated for Hodgkin’s disease.JCO 21:3431–398. DJ Straus et al. (2004) Results of a prospectiverandomized clinical trial of doxorubicin, bleomycin,vinblastine, and dacarbazine (ABVD) followed byradiation therapy (RT) versus ABVD alone for stages I,II, and IIIA nonbulky Hodgkin disease. Blood104:3483–899. H Eghbali et al. (2005) Comparison of threeradiation dose levels after EBVP regimen in favorablesupradiaphragmatic clinical stages I-II Hodgkin’slymphoma: preliminary results of the EORTC-GELAH9-F trial [abstract]. Blood 106:a24010. A Engert et al. (2011) Reduced intensity ofchemotherapy and PET-guided radiotherapy in patientswith advanced stage Hodgkin lymphoma: The GHSGHD15 final results [abstract]. Ann Meeting Am SocHematol a589

Author affiliationsBertrand Coiffier, Department of Haematology, CentreHospitalier Lyon-Sud, Pierre-Bénite, France; OlivierCasasnovas, Department of Haematology, Hôpital leBocage, Dijon, France


First-line bevacizumab for ovariancancer – new standard of care?

� Susana Banerjee and Stan Kaye

Demonstrationof theclinically significant activityofbevacizumab inadvanced-stage ovarian cancer has attracted a great deal of interest. Here, we summa-rize the twopositive phase III trials that led toEMAapproval of bevacizumabas first-line therapy and discuss the optimumuse of the drug in this disease.

In December 2011, two positivephase III trials1, 2 that assessed beva-cizumab inpatientswithovariancan-

cer were reported in the New EnglandJournal of Medicine; these results led to

theEMAapproval of thedrugas first-linetreatment in combination with carbo-platin and paclitaxel for this disease.3

Bevacizumab is currently themostwidelytested antiangiogenic agent for the treat-

ment of cancer. Bevacizumab is amono-clonal antibody that targets the VEGFpathway,whichhas a critical role in ovar-ian function as well as in the spread ofovariancancer.4 Therefore,positive resultsfromclinical trials assessingbevacizumabin thisnotoriouslydifficult-to-treatdiseasehave been eagerly anticipated.The first study (GOG-0218) was

reported byBurger et al.1 andwas a dou-ble-blind, three-arm, placebo-controlledstudy in 1873 patients with newly diag-nosed stage III (incompletely resectedwith residual disease >1 cm) or stageIV epithelial ovarian cancer. Patientswere randomly assigned to one of threetreatments: combination chemotherapy(carboplatin–paclitaxel), carboplatin–paclitaxel chemotherapy plus concurrentbevacizumab, or carboplatin–paclitaxel

This article was first published inNature Reviews Clinical Oncology on 28 February 2012, and is published with per-mission. © 2012 Nature Publishing Group. doi:10.1038/nrclinonc.2012.28

chemotherapy plus concurrent andmain-tenance bevacizumab. The bevacizumabdose was 15 mg/kg for up to 22 cycles(15months total).After a protocol amend-ment, stage III patients withmacroscopicresidual disease of ≤1 cm were alsoincluded.Nevertheless, all patientsenrolledhadadvanced-stagedisease and their over-all outlook was worse than those patientsassessed in the second study, ICON7.2

Perren et al.2 published the resultsfrom the ICON7 study. The trial ran-domly assigned patients to one of twoarms:1528patients receivedcarboplatin–paclitaxel chemotherapywith orwithoutconcurrent and maintenance beva-cizumab. Bevacizumab was given at7.5 mg/kg (half the dose used in GOG-0218) for a total of 18 cycles (12monthstotal). In this trial, 9%ofpatientshadhigh-risk, early-stage disease (FIGOstage I orIIA, clear cell or grade 3 histology)whereas 30%were at the highest risk forprogression (FIGO stage IV, or stage IIIand >1 cm residual disease).The primary endpoint in both trials

was progression-free survival (PFS),whichwas evaluatedusingRECISTandGynecologicCancer Intergroup (GCIG)CA125 criteria in GOG-0218; onlyRECIST criteria were included in theassessment in ICON7. Despite key dif-ferences, for both studies the primaryendpoint was met for concurrent andmaintenance bevacizumab. In GOG-0218, median PFS was extended by3.8months (14.1monthsvs10.3months;P<0.001).1 In the ICON7 trial, themedian PFS was 17.3 months in thechemotherapy-alone arm compared to19.0 months with the addition of beva-cizumab (HR0.81;P=0.004).2

In GOG-0218, an additional analysiswas carried out that did not take accountof CA125 progression (that is, only inter-preting the response based on RECISTcriteria); in this analysis, themedian PFSwas sixmonths longer in the group receiv-ingbevacizumab(concurrent andasmain-

tenance) compared to thechemotherapy-alone control arm (12 months vs18 months; HR 0.645; P<0.001).1 How-ever, this analysis, which was required bythe regulatory agencies, has been criti-cised owing to the bias associated withunequal censoring in the two arms.In ICON7, the magnitude of PFS

improvement is relatively modest(1.7 months);2 however, a preplannedanalysis demonstrated that the benefitof bevacizumab is greater in patientsdefined to be at the highest risk of pro-gression. The 3.6-month improvementin PFS seen in this subgroup usingrestricted means analysis (restrictedmeans14.5months vs 18.1months;HR0.73;P=0.002) is similar to thedifferencein PFS reported in GOG-0218 for theequivalent arms (3.8months).For the assessment of the effects of

bevacizumab treatment on overall sur-vival, finalmaturedata areawaited.How-ever, in ICON7, an improvement inoverall survival with bevacizumab in thehigh-risk group was particularly note-worthy (28.8 months vs 36.6 months;HR 0.64, 95%CI 0.48–0.85; P=0.002).2

Thedemonstrationof a survival benefit ofalmost eight months in patients with apoor prognosis is very encouraging.Toxic effects were as expected, with

hypertension grade ≥2 being common(23%ofpatients in theGOG-0218 study;18%of patients in the ICON7 study) butgenerally well controlled. Overall, beva-cizumab treatment was well tolerated.Although bowel perforations had beenreported in earlier bevacizumab trials,5

these perforations were rare events inGOG-0218 (<3% of the patients) andICON7(1%of thepatients).However, theincidence was higher with bevacizumabtherapy compared to control arms.Based on these new trial results, is it

possible to say that bevacizumab is thenew standard of care? To answer this,several questions need to be addressed.First, which patients should be offered

bevacizumab?Althoughboth studiesmettheir primary endpoints for the wholetrial population, it could be argued thatgiven the overall survival benefit seen inhigh-risk patients in ICON7,2 womenwith stage IV or stage III >1 cm residualdisease should be considered for first-line treatment. The OCEANS study,6

in which patients with recurrent plat-inum-sensitive disease were treatedwith bevacizumab in combination withchemotherapy (carboplatin with gem-citabine), provides a new dimension tothis issue.This study reportedasignificantimprovement in PFS with the additionof bevacizumab (8.4 months vs 12.4months; HR 0.48; P<0.0001) andstrongly suggests a role for bevacizumabin this settingof recurrentdisease.5 There-fore, a reasonable proposal for patientsoptimally debulked and thus at a lowerrisk of early relapse would be to reservebevacizumab until first recurrence.The second question is what is the

optimaldoseofbevacizumab?The licenseddose of bevacizumab, based on the PFSdataofGOG-0218, is15mg/kg.3However,when comparing PFS improvement in asimilar patient population (high-risk) inICON7, there is no difference in PFSimprovement between the groups receiv-ing15mg/kgand7.5mg/kg.The7.5mg/kgdose is likely tobemorecost-effective and,so far, this is the dose which is associatedwith an overall survival benefit.Based on the available data, should

bevacizumabmaintenance be extended

ImpactFactor


Practice pointsThe addition of bevacizumab givenconcurrently with chemotherapy andcontinued asmaintenance treatmentsignificantly increases progression-free survival as first-line therapy forovarian cancer, in particular for thosepatients at high risk of progression.

ImpactFactor

until disease progression? The maxi-mal treatment effect, as indicated bythe greatest separation of PFS curves inGOG-0218 and ICON7, coincidedwith the end of planned bevacizumabtreatment. When bevacizumab isdiscontinued, the impression is thatthe disease returns promptly and thisis in keeping with observations inother cancers.7 Results from theOCEANS study,6 seemingly superiorto theGOG-0218 and ICON7 results,were achieved when bevacizumab wascontinued until disease progression.Taken together, these findings suggestthat bevacizumab therapy until diseaseprogression is warranted.A fourth question is: should beva-

cizumab be given in combination withchemotherapy (in addition to mainte-nance) for first-line therapy? The lack ofPFSdifferencebetween thechemother-apy-alonecontrol armand theconcurrent

bevacizumab arm in GOG-0218 wouldsuggest that the main impact of beva-cizumab is as maintenance treatmentpost chemotherapy.However, the signif-icantly increased response rates (48% vs67%;P<0.0001) in the subset of patientswith measurable disease followingdebulking surgery in the bevacizumabarm of the ICON7 trial, and in theOCEANS study (57% vs 79%;P<0.0001), indicates clearly that beva-cizumabenhances chemosensitivity, andits omission from concurrent treatmentmay be unwise.Finally, does the extent of benefit

reported so far justify the cost? For thosepatients with theworst initial outlook, aPFS improvement of fourmonths trans-lates into almost double the timewithoutchemotherapy before the first recur-rence. This improvement does repre-sent an important clinical benefit andpatient selection is therefore paramount.

The identification of a group of patientslikely to benefitmost frombevacizumabtreatment could tip the balance towardsa cost-effective therapy.These important studies by Burger

et al.1 andPerren et al.2 demonstrate thatthe anti-VEGFstrategyhas real potentialin ovarian cancer. In addition to beva-cizumab, other agents targeting thispath-wayare inactivedevelopment4 and futuretrials will undoubtedly clarify the beststrategy touseall theseapproaches for thebenefit of our patients.

Details of the references cited in this article can beaccessed at www.cancerworld.org

Competing interests:Stan Kaye declares an association with Roche. See thearticle online for full details of the relationship. SusanaBanerjee declares no competing interests

Author affiliationsSusana. Banerjee, the Royal Marsden NHS FoundationTrust, London, UK; Stan Kaye, the Institute of CancerResearch and Royal Marsden Hospital, Sutton, UK


� Are trials being stopped too early?� Are patient groups skewing

the research agenda?� Are you getting the career

breaks you need?� Which is better? Medical

oncologist or organ specialist,robot or surgeon?

The new, redesigned CancerWorldwebsite invites you to contribute tocurrent debates by using its commentfacility. You can also suggest topics forcoverage and find links to related sites.Get online and take a look

Cancerworld It’s your world. Get online and have your say

www.cancerworld.org

NewsRound


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

Complications from roboticprostatectomy no betterthan conventional surgery� Journal of Clinical Oncology

Problems with continence and sexual func-tion are common following both robot-

assisted laparoscopic radical prostatectomy(RALRP) and open retropubic radical prostatec-tomy (ORRP), a US study has found.

Conventional wisdom holds that menundergoing the robotic procedure experienceless post-surgical urinary incontinence anderectile dysfunction compared to those under-going the traditional surgical approach. In thecurrent study Michael Barry and colleagues,from Massachusetts General Hospital, Boston,compared the continence and sexual functionof Medicare enrolees following treatment witheither ORRP or RALRP. Investigators used apopulation-based random sample drawn from20% of Medicare prostatectomy claims filedbetween August and December 2008. At amedian of 14 months following surgery, par-ticipants were asked to complete a mailed sur-vey that included self-ratings of problems withurinary continence and sexual function.

Completed surveys were obtained from 685participants, with 406 reporting having under-gone RALRP and 220 ORRP. When results were“dichotomized” 27.1% of men who had under-gone ORRP reported a moderate or big problemwith continence compared with 33.3% whohad undergone RALRP (P=0.113). For sexualfunction, 89% of men who underwent ORRPreported a moderate or big problem comparedwith 87.5% who had undergone RALRP (P=0.57).

robot-assisted surgery remains unclear, it hasbeen estimated that high proficiency in thistechnique may require that more than 200surgeries be performed,” they write.

� MJ Barry, PM Gallagher et al. Adverse effects

of robotic-assisted laparoscopic versus open

retropubic radical prostatectomy among a

nationwide random sample of Medicare-age men.

JCO 10 February 2012, 30:513–518

� MR Cooperberg, AY Odisho and PR Carroll.

Outcomes for radical prostatectomy: is it the

singer, the song, or both? ibid, pp 476–478

Semuloparin reducesthromboembolic eventsduring chemotherapy� New England Journal of Medicine

Semuloparin – the ultra-low molecularweight heparin – reduces the incidence of

thromboembolic events in cancer patientsundergoing chemotherapy, the SAVE-ONCOstudy has concluded.

It is well known that cancer patients receiv-ing chemotherapy are at increased risk of venousthromboembolism, with complications includingotherwise unnecessary hospitalisations, inter-ruptions of chemotherapy and anticoagulanttreatmentsor insertionsofavenacava filter.Cur-rent guidelines recommend antithrombotic pro-phylaxis for patients with cancer admitted tohospital for medical illness (administered for theduration of their hospital stay) and for patientsundergoing surgery for cancer, but not for rou-tine use in ambulatory chemotherapy patients.

“Our findings demonstrate the risks patientsactually face with these two procedures in thecontemporary national surgical experience inMedicare. Low case volumes likely contribute tothe high risk of adverse effects with both pro-cedures in the general population,” write theauthors. Whether the risk of adverse effectswill be lower over time with RALRP, they add,remains to be seen, but in the interim, there is aquestion about value for money.

“The apparent lack of better outcomes asso-ciated with RALRP also calls into questionwhether Medicare should pay more for thisprocedure until prospective large-scale out-come studies from the typical sites performingthese procedures demonstrate better results interms of adverse effects and cancer control,”they conclude.

In an accompanying commentary, MatthewCooperberg and colleagues, from the Universityof California, San Francisco, write, “Althoughmethodologically much more sound than anearlier analysis that tried to determine health-related quality-of-life outcomes on the basis ofclaims data alone, the study… still has significantlimitations.”

These limitation, the say, include the factthat all the subjects were aged 65 or older,which means there are no data to show whetherresults might have differed in younger patients.As baseline function was not measured, it is notpossible to say whether these were the same forthe two study groups. Furthermore, the surveyinstrument assessed only “bother” and notfunction. They also point out that all the oper-ations were performed in 2008, when many sur-geons may have still been “learning” the robot-assisted technique.

“Although the exact learning curve for

NewsRound


In the double-blind multicentre trial, Gian-carlo Agnelli, from the University of Perugia,Italy, and international colleagues, randomised3212 patients with a wide range of metasta-tic or locally advanced solid tumours to receivesubcutaneous semuloparin 20 mg once daily,or placebo, until there was a change ofchemotherapy. Patients in the study, who hadjust commenced a number of differentchemotherapy regimens, were recruited from395 centres in 47 countries.

Results show that, at a median treatmentduration of 3.5 months, venous thromboem-bolism occurred in 1.2% of patients receivingsemuloparin, compared with 3.4% receivingplacebo (HR 0.36, 95%CI 0.21–0.60; P<0.001).The incidence of clinically relevant bleedingwas 2.8% in the semuloparin group versus 2.0%in the placebo group (HR 1.40, 95%CI 0.89–2.21), with major bleeding occurring in 1.2%receiving semuloparin versus 1.1 % receivingplacebo (HR 1.05, 95%CI 0.55–1.99). The inci-dence of other adverse events was similar in thetwo treatment arms.

“The results of this study show that throm-boprophylaxis with the ultra-low-molecular-weight heparin semuloparin, as compared withplacebo, reduces the risk of venous thromboem-bolism among patients receiving chemotherapyfor cancer, with no apparent increase in the inci-dence of major bleeding,” conclude the authors.

Several criteria have been proposed to iden-tifycancerpatientsathighrisk forvenous throm-boembolism, they add, including specific cancertypes, chemotherapy regimens, levels of serumtissue-factor microparticles or P-selectin andpredictive scores for chemotherapy-relatedthrombosis. They suggest that stratification forthe risk of venous thromboembolism amongpatients with cancer might be clinically useful.

In the accompanying commentary, Elie Akland Holger Schünemann, from the State Uni-versity of New York, undertake a new pooledanalysis of low molecular weight heparin use incancer patients including data from their ear-lier Cochrane review of nine trials, the SAVE-ONCO trial and a recent third study including503 patients. When these studies are com-bined, the relative risk for symptomatic venous

thromboembolism is 0.57 and for death 0.94.“The key questions that are not answered

conclusively relate to the effect of treatmentwith low-molecular-weight heparin on qualityof life and whether such treatment affectstumor growth or dissemination,” write theauthors. At time of publication, they add, at leastsix additional low molecular weight heparintrials in cancer patients, aiming to enrol around3500 patients in total, are ongoing.

� G Agnelli, D George, A Kakkar et al.

Semuloparin for thromboprophylaxis in patients

receiving chemotherapy for cancer. NEJM 16

February 2012, 366:501–509

� E Akl and H Schünemann. Routine heparin for

patients with cancer? One answer, more questions.

ibid, pp 661–662

BRCA1/2 mutationspredict ovariancancer survival� JAMA

Among women with invasive epithelial ovar-ian cancer, mutations in the BRCA1 or

BRCA2 genes are associated with improved five-year survival in comparison to women who donot carry the mutations. The UK study revealingthat BRCA2 carriers show the best prognosis rep-resents the largest BRCA-associated ovariancancer outcomes study reported to date.

Approximately 10% of women with invasiveepithelial ovarian cancer carry deleteriousgermline mutations in BRCA1 or BRCA2. Thegoalof the studybyPaulPharoahandcolleagues,fromtheUniversityofCambridge,UK,was togaina better understanding of the effect on survivalof BRCA1/2 mutations compared to wild-typeBRCA1/2 from a multiple case series of epithelialovarian cancer.

In the pooled analysis, participants weredrawn from 26 international studies that hadenrolled participants between 1987 and 2010(10 studies from the US, six from Europe, twofrom Israel, one from Hong Kong, one from

Canada, one from Australia and five from theUK). Altogether data were obtained from 3879women with ovarian cancer – 909 with patho-genic germline mutations in BRCA1, 304 withgermline mutations in BRCA2 and 2666 who didnot carry BRCA1 or BRCA2 mutations.

Results show that the five-year overallsurvival was 36% (95%CI 34%–38%) for non-carriers, 44% (95% CI 40%–48%) for BRCA1carriers, and 52% (95%CI 46%–58%) forBRCA2 carriers.

The study also showed that the clinicalcharacteristics of epithelial ovarian canceramong BRCA1/2 carriers differed from that ofnon-carriers. Tumours with serous histology,high grade and advanced stage were all morelikely among carriers of both mutations.

In a secondary analysis, the investigatorsfound that the survival advantage conferred byBRCA1 mutations was partially mitigated as themutation site moved from the 5' to 3' end. Thissuggests, they write, that the site of the BRCA1mutation may be of individual prognosticimportance.

“BRCA1 and BRCA2 carriers with EOC[epithelial ovarian cancer] respond better thannon carriers to platinum-based chemotherapiesand have improved survival despite the factthat the disease is generally diagnosed at alater stage and higher grade,” write the authors.The findings, they add, could be used by health-care professionals for patient counsellingregarding expected survival.

“Given the important prognostic informa-tion provided by BRCA1 and BRCA2 status andthe potential for personalized treatment in car-riers, the routine testing of women presentingwith high grade serous EOC may now be war-ranted,” they write.

In an accompanying commentary, DavidHyman and David Spriggs, from MemorialSloan-Kettering Cancer Center, New York, writethat the findings represent the latest evidencethat ovarian cancer is a much more geneticallyand biologically heterogeneous disease thanpreviously appreciated. “Further studies in sim-ilarly large data sets are needed to better under-stand the effects of somatic and epigeneticalterations in BRCA gene function as well as

NewsRound


complex interactions with other inherited alle-les. The accelerating availability of detailedsomatic and germline genetic information willchallenge all physicians who stand at the bed-side of patients with cancer and struggle todeliver compassionate, individualized care.”

� K Bolton, G Chenevix-Trench, C Goh et al.

Association between BRCA1 and BRCA2

mutations and survival in women with invasive

epithelial ovarian cancer. JAMA. 25 January 2012,

307:382–390

� D Hyman, D Spriggs. Unwrapping the

implications of BRCA1 and BRCA2 in ovarian

cancer. ibid, pp 408–409

Adjuvant chemotherapyimproves outcomesfollowing D2 gastrectomy� The Lancet

Adjuvant chemotherapy should be consid-ered as a treatment option after curative

D2 gastrectomy, the phase 3 CLASSIC study hasconcluded.

In Eastern Asian countries (especially Japanand Korea) D2 lymph node dissection (definedas dissection of group 1 and 2 lymph nodes) isregularly performed as a standard procedure forgastric cancer over D1 lymph node dissection(dissection of group 1 lymph nodes only). Inwestern countries, D2 dissection has been asso-ciated with higher morbidity and mortality,although recent studies demonstrate thatwestern surgeons can be trained to perform D2gastrectomy for selected patients with lowmorbidity and mortality.

With guidelines now advocating D2 dis-section in centres with specialist expertise,increased acceptance of D2 gastrectomy raisesquestions about the optimum adjuvant therapyfor patients with operable gastric cancer. TheCapecitabine and Oxaliplatin Adjuvant Study inStomach Cancer (CLASSIC) study, was designedto compare the effect of adjuvant capecitabineplus oxaliplatin after D2 gastrectomy with sur-

� Y Bang, Y Kim, H Yang et al. Adjuvant

capecitabine and oxaliplatin for gastric cancer after

D2 gastrectomy (CLASSIC): a phase 3 open-label,

randomised controlled trial. Lancet 28 January

2012, 379:315–321

� T Nishida. Adjuvant therapy for gastric cancer

after D2 gastrectomy. ibid, pp 291–292

Everolimus overcomesresistance tohormone therapy� New England Journal of Medicine

Everolimus combined with the aromataseinhibitor exemestane extended progres-

sion-free survival in postmenopausal womenwith advanced hormone-receptor-positivebreast cancer, the BOLERO-2 study has found.The international phase III study, first pub-lished online to coincide with presentation atthe 2011 San Antonio Breast Cancer Sympo-sium, showed that combination treatmentmore than doubled progression-free survivalcompared to exemestane alone.

Resistance to endocrine therapy in breastcancer has been associated with activation of themammalian target of rapamycin (mTOR) intra-cellular signalling pathway. Everolimus, animmunosuppressant agent used to preventorgan transplant rejection, is known to inhibit themTOR protein. In preclinical studies, everolimusin combination with aromatase inhibitorsresulted in both the synergistic inhibition ofproliferation and the induction of apoptosis.

In the Breast Cancer Trials of OralEverolimus-2 (BOLERO-2) study, José Baselga,from Massachusetts General Hospital CancerCenter, Boston, and international colleagues,randomly assigned 724 women with hormone-receptor-positive advanced breast cancer toreceive either the combination of everolimusand exemestane (n=485; combination-ther-apy group) or exemestane and placebo (n=239;exemestane only group). The patients, whowere recruited from 189 centres in 24 countries,had experienced either recurrence or disease

gery alone on disease-free survival in patientswith stage II or III gastric cancer.

In the study, led by Yung-Jue Bang fromSeoul National University College of Medicine,in Jongno-gu, Seoul, 1035 patients with stageII or III gastric cancer were randomly assigned ina 1:1 ratio to adjuvant chemotherapy (n=520)or surgery alone (n=515). The study was under-taken in 37 centres in South Korea, China andTaiwan.

Results show that the three-year disease-free survival was 74% in the chemotherapy andsurgery group versus 69% in the surgery alonegroup (HR 0.56, 95%CI 0.44–0.72; P<0.0001).Grade 3 or 4 adverse events were reported in56% of patients in the chemotherapy and sur-gery group versus 6% in the surgery only group.

“This study shows that a 6 month course ofchemotherapy after D2 gastrectomy improves3–year disease-free survival compared withsurgery only,” conclude the authors.

Although overall survival data from theCLASSIC trial are not yet mature, the resultssuggest an improvement with chemotherapycompared with surgery alone. “An analysisafter a median follow-up of 5 years is plannedto conclusively establish the overall survivalbenefit of capecitabine and oxaliplatin in thissetting,” write the authors. A key question forthe trial, (as with any trial undertaken in onegeographical region), they add, is whether find-ings can be generalised to other regions wheredisease management practices might differ.

In an accompanying commentary, ToshirouNishida, from Osaka Police Hospital, Japan,raises the issue of adherence and safety. Withmore than half of patients in the CLASSIC studywho were treated with chemotherapy experi-encing grade 3 or 4 adverse events, nearly 10%withdrawing due to adverse events and 20%refusing treatment, he writes, non-adherencecould be considered a risk of compromisingdisease outcomes.

“Identification of higher-risk patients andprediction of drug efficacy by biomarkers, andintroduction of targeted agents such astrastuzumab, should be considered for adju-vant therapy of gastric cancer in the future,”writes Nishida.

NewsRound


progression while receiving previous therapywith a nonsteroidal aromatase inhibitor (anas-trozole or letrozole) in the adjuvant setting or totreat advanced disease.

At interim analysis the median progres-sion-free survival was 6.9 months in the com-bination therapy group versus 2.8 months in theexemestane-alone group (HR 0.43, 95%CI 0.35–0.54; P<0.001). According to central assess-ment, the median progression-free survival was10.6 months in the combination therapy groupversus 4.1 months in the exemestane-alonegroup (HR 0.36, 95%CI 0.27–0.47; P<0.001).

Serious adverse events were reported by23% of patients in the combination therapygroup versus 11% in the exemestane-alonegroup. Stomatitis was the most common grade3 or 4 adverse event, occurring in 8% of patientsin the combination group versus 1% in theexemestane-alone group. This was followed byanaemia (6% vs >1%), dyspnoea (4% vs 1%)and hyperglycaemia (4% vs >1%).

The positive results in the study, write theauthors, are consistent with the outcomes of twoearlier studies of everolimus and anti-oestrogentherapy in hormone-receptor-positive breastcancer patients. In one study, neoadjuvanteverolimus combined with letrozole improvedthe clinical response rate and decreased tumourcell proliferation in patients with newly diag-nosed breast cancer; while in a second study thecombination of everolimus and tamoxifenincreased progression-free survival in womenwith oestrogen-positive advanced breast cancerpreviously treated with an aromatase inhibitor.

“Taken together, these studies suggest thateverolimus adds to the anticancer activity ofantiestrogen therapy in a variety of clinical set-tings and with different classes of endocrineagents,” write the authors.

But benefits should be weighed againstthe side-effects observed with everolimus, theyadd. “The potential of everolimus to benefitpatient survival is not yet known,” they caution.

� J Baselga, M Campone, M Piccart et al.

Everolimus in postmenopausal hormone-receptor-

positive advanced breast cancer. NEJM 9 February

2012, 366:520–529

Adjuvant chemotherapyadds no benefit overchemoradiation alone innasopharyngeal cancer� Lancet Oncology

Adding adjuvant cisplatin and fluorouracilchemotherapy to concurrent chemo-

radiotherapy in patients with nasopharyngealcarcinoma confers no survival benefit, reportsa Chinese study.

In recent years, seven randomised phase IIIstudies comparing chemoradiation with radio-therapy alone have confirmed the value ofchemotherapy on survival for advancednasopharyngeal carcinoma. However, withthree of these trials adding concurrentchemotherapy to radiotherapy only and fourusing the regimen of concurrent chemoradio-therapy plus adjuvant chemotherapy, investi-gators have been “unclear” as to whetheradjuvant chemotherapy might deliver addi-tional survival benefits over concurrentchemoradiotherapy.

In the current study, Jun Ma and colleagues,from Sun Yat-Sen University Cancer Centre,Guangzhou, China, set out to investigatewhether addition of adjuvant chemotherapy toconcurrent chemoradiotherapy delivered fur-ther benefits. Between June 2006 and March2010, 508 patients with non-metastatic stage IIIor IV nasopharyngeal carcinoma were ran-domised to receive concurrent chemoradio-therapy plus adjuvant chemotherapy (n=251) orconcurrent chemoradiotherapy alone (n=257).

Patients in both groups received40 mg/m² cisplatin weekly for up to sevenweeks, given concurrently with radiotherapyat 2.0–2.27 Gy per fraction, with five dailyfractions per week for six to seven weeks. Inaddition, the chemotherapy adjuvant groupreceived 80 mg/m² adjuvant cisplatin and800 mg/m² per day fluorouracil for 120 hevery four weeks for three cycles. The studywas conducted in seven institutions in China.

At a median follow-up of 37.8 months,the estimated two–year failure-free survival

was 86% in the chemoradiotherapy plus adju-vant chemotherapy group, versus 84% in thechemoradiotherapy only group (P=0.13).Adverse events were similar in both groups,with the most common being stomatitis, whichoccurred in 31% of patients receiving chemora-diotherapy plus adjuvant treatment and 32%receiving chemoradiotherapy alone.

“In our trial, adjuvant cisplatin and fluoro-uracil chemotherapy did not improve outcome,with no significant effect on the risk of treat-ment failure, or estimated 2 year failure-freesurvival, overall survival, distant failure-freesurvival, or locoregional failure-free survival,”write the authors.

One possibility, they add, is that adjuvantcisplatin and fluorouracil does not represent aneffective combination for eradication ofmicrometastases in nasopharyngeal carci-noma. “New combinations of more tolerabledrugs that might improve efficacy ofchemotherapy as an adjunct in advancednasopharyngeal chemotherapy should beinvestigated,” they write.

In an accompanying commentary, JosephWee, from Duke-NUS Graduate MedicalSchool, Singapore, writes that two recentreports suggest that chemoradiation withfirst-generation drugs appears to work only inpatients with earlier stage disease with lowerdistant tumour burden.

This raises the question, he adds, ofwhether the addition of a third or fourth agentmight make a difference to outcomes. “Thisstrategy is being investigated by several groupsin the phase 3 setting, and are being done in theneoadjuvant setting to overcome the poorcompliance if chemotherapy is given afterradiotherapy,” he writes.

� L Chen, C Hu, X Zhong et al. Concurrent

chemoradiotherapy plus adjuvant chemotherapy

versus concurrent chemoradiotherapy alone in

patients with locoregionally advanced naso-

pharyngeal carcinoma: a phase 3 multicentre

randomised controlled trial. Lancet Oncol

February 2012, 13:163–171

� J Wee. Nasopharyngeal cancer: a promising

future. ibid, pp 116–117

Prostate cancer units:it’s about optionsand quality� Peter McIntyre

Men diagnosed with prostate cancer have a wealth of options on how to proceed, but they can

pay ahighprice if thequality of treatment is substandard.Coulddelivering all prostate care through

specialist multiprofessional units be the answer to safeguarding both standards and choice?

Themost common cancer inmenhas seen a rapid increase in casesand treatments over the past two

decades.About 382,000men inEuropeare diagnosedwith prostate cancer eachyear, and there are 89,000 deaths.How-ever, there has been a lack of clarity overthe best way to match the right treat-ment to the right patient.

There are at least three treatments, eachabout as good as the other in skilledhands: surgery (radical prostatectomy),radiotherapy and brachytherapy, in somecases combined with hormonal ther-apy.After a rapid increase in the numberof surgical cases in the 1990s and earlyyears of this century there has been apull back from aggressive treatment

in early or indolent cancers.The after-effects of treatment, par-

ticularly impotence or incontinence,scare a lot of men and there is increas-ing recognition that quality of life is ofgreat concern, alongside a desire to berid of the cancer.

Patients who are newly diagnosedwith prostate cancer have choices to


Systems&Services

YELLOWDOGPRODUCTIONS/WWW.GETTYIMAGES.COM

“We have three equally effective therapies,

so we cannot decide, as doctors, which is best”

make. Do they opt for immediate treat-mentor for active surveillance? If theyoptfor treatment, which is best for them?

There is a third choice, of whichpatients may not be sufficiently aware,and that concerns the kind of centrewhere their treatment and care willtake place.

It is widely accepted that cancer isbest treated in amultidisciplinary settingby specialists with expertise in the par-ticular disease, backed by a multipro-fessional team. While this is becomingthe norm for breast cancer, it is notwidely practised for prostate cancer.

Last year the European Journal ofCancerpublishedadiscussionpaper fromtheEuropeanSchool ofOncology (ESO)promoting specialist prostatecancerunitsand setting out proposals for what thatmightmean in termsof professional staffandexperience (seebox,p62).Thepaperwas reminiscent of proposals for breastcancer specialist centrespublished in the1990saspart of aEuropeanmovement toimprove treatment andpreventovertreat-ment. It meant in effect that unless asurgeon or radiotherapist was going tospecialise in this disease, they had nobusiness dabbling in it.

The same thing may happen, even-tually, for prostate cancer, but themovement is slow to gather momen-tum. The Deutsche Krebsgesellschaft(German Cancer Society) has takenthe lead by setting up a network ofcertified prostate cancer units. TheUK National Institute for Health andClinical Excellence (NICE) has setminimum standards, under which, forexample, specialist urology teamsshould undertake a minimum of 50radical operations per year.


SPEAKING THE SAME LANGUAGERiccardo Valdagni is director of theProstate Cancer Programme at theIstituto Nazionale Tumori, in Milan,coordinator of ESO’s Prostate CancerProgramme and lead author of the ESOpaper. He says that moving to a multi-disciplinary approach is a challenge.“Urologists, radiation oncologists andmedical oncologists have differentapproaches to the disease and speak dif-ferent languages. Themost ambitious –though necessary – step when theProstate Cancer Programmewas estab-lished was to share evidence-based aswell as institution-adapted guidelines forthe diagnosis, therapy, observation, andfollow-up of prostate cancer patients.

“The general worldwide approach isthat the patient has a biopsy, he receivesa prostate cancer diagnosis from theurol-ogist, and then the urologist generallymakes the first proposal of therapy. Weprefer to have aurologist, radiation oncol-ogist and psychologist (with a medicaloncologist on demand) meet with thepatient, discuss the therapeutic andobser-vational options, and offer support fordecisionmaking. International guidelinesall over the world say we have threeequally effective therapies, so we cannotdecide, as doctors, which is the best.”

Patients are then encouraged tochoose the treatment, weighting theirvalues and priorities, says Valdagni. Iserectile dysfunction a major issue forthem?What abouturinary incontinence?“Onepatientmaysay, ‘yes, very important’.Anothermaysay, ‘I don’tmindabout side-effects, Iwant the cancer out ofmybodyas soon as possible.’”

In Valdagni’s centre, few patients askthe clinician what he would do in their

shoes. He thinks this is because theyhave enough time and information tomake a decision, with psychological sup-port if necessary.

“In general, the problem of saying‘Heydoctor,whatwould youdo?’is relatedto the psychological effect of beingdiagnosed with cancer. Patients mayprefer at first to have someone take thedecision for them. Offering exhaustiveinformation on all his options and sup-portinghimpsychologically,we try tohelpthe patient find his way. The patient,instead of being an object of physiciancare, can become the subject of his care,decidingwhat is best forhis quality of life.”

At the Milan Prostate Cancer Pro-gramme,most patientswith small or clin-ically indolent disease choose activesurveillance.Of thosewho drop out fromactive surveillance and have treatment,about 45% choose surgery, 50% radio-therapy and 5%brachytherapy.

Thepattern inmonospecialist centresis quitedifferent, and it seems that unlessthey work together, specialists, perhapsunconsciously, influencepatients in favourof their speciality. One paper suggeststhat, if thepatient seesonlyaurologist, 70–80%opt for surgery. If they also see a radi-ation oncologist, 70% choose radiation.

LouisDenis speaksasa foundingmem-berof theEuropeanprostatecancerpatientgroupEuropaUomo, which advocates forpatient-centredcarewherequality of life isas important as survival. He is also thedirector of theAntwerpOncologyCentre,and says that, whilemultidisciplinary careis widely accepted in theory and is a legalrequirement in Belgium, it is not widelyimplemented. “We still face the dilemmabetween the traditional freedom of treat-mentchoice for the individual specialist and

Systems&Services

thebetter outcomesof cancer treatmentbymultidisciplinarymanagement.

“There is known overtreatment forpatientswithprostatecancer for anumberof reasons.Among theseweshould recog-nise lack of correct evaluation of thepatient’s health status, ignorance of theclinical course of low-risk, low-volumeprostate cancer and the availability ofadvanced technology that cannot rest idle.The slogan of Europa Uomo remains:‘First the Patient, then hisCancer’.”

UNDERSTANDING THE OPTIONSLawrenceDrudge-Coates, clinical nursespecialist in urological oncology at King’sCollegeHospital, in London, agrees.Hisis a specialist unit in all but name, with260 new prostate cancer patients a year.As one of two keyworkers for patients, heruns his own clinics and encouragespatients to take their time in making aninformed decision.

“One of the key roles that the clinicalnurse specialist plays is to take patients

through theprosandcons inmoredetail inlaymen’s terms. I explain what we havefound from the biopsies and scans andwhether it is an aggressive tumour. I gothrough the treatment options, but not intoomuchdetail. If a patient is being givena diagnosis, their ability to take in infor-mation is very greatly reduced. You give abit of information and supplement itwithgood literature, and I give the patient mycontact details as key worker, and theopportunity to discuss issues further.


Systems&Services

“IT WAS A PERSONAL DECISION TO DEFEND MY QUALITY OF LIFE”

Enrico Rambaldi is Pro-fessor of Philosophy at theUniversity of Milan, anexpert in bioethics andeditor of the Italian Journalof History of Philosophy.He found that his skills andtraining had not preparedhim for making a decisionabout his prostate cancer.Hewas diagnosed in 2008,

at the age of 72. He not only had prostate cancer but also anassociated sepsis that nearly killed him. “For some days I wasbetween death and life,” he says.When he recovered he visited some of the best specialistsin Milan to ask their advice. “I was just going from one toanother asking what they suggested. The choice whichwas offered to me was between a surgical operation andradiotherapy.“My reaction was not very good. I was worried about the dan-gers in relation to sexual activity and incontinence. I wasreally very, very unhappy.

“I was changing my mind from one day to another day. Oneday I’d say, ‘I have to have radiotherapy,’ talking with my wife.Then I decided to go for surgical intervention. I fixed thedate for my operation and every two or three days I changedmy mind.“As a philosopher I don’t know anything about my body. I amdependent on external information.”In the end Rambaldi opted for active surveillance.“I decided to start the active surveillance after several talkswith Valdagni. I thought that it would be unwise to put thequality of my life in danger. It really was a personal deci-sion to defend my quality of life. I didn’t want to get intoproblems with incontinence or no sexual activity becauseI was afraid. I feel well supported from the psychologicalpoint of view.“It was a very good decision. I don’t even take any medication.I go for a PSA check every three or four months, a consultationtwice a year and one biopsy in the four years.”Rambaldi says that the quality of his life has actually improvedsince he was diagnosed. “I appreciate more than before thepace of time. I am more careful not to waste my time and toproduce as much as I can in my philosophy.”

Making the choice

“Offering exhaustive information on all his options,

we try to help the patient find his way”

“Not all cancers have to be treated and Ithink this is still quite an alien concept formost patients. If active surveillance is anoption I would explain why. It may be acancer that isnotparticularly aggressive. Inmanycasesweadvise themtohave furtherbiopsies. There is a contract betweenmyself and the patient.”

Drudge-Coates takes time to talkabout the major possible effects of treat-ment – erectile dysfunction and inconti-nence. “You have to be very upfront and

state that thesearekey issues in relation tosurgery and radiotherapy. I don’t call thema side-effect because it belittles them.However,manypatients alreadyhaveerec-tiledysfunctionprior to treatmentbecauseofprostate cancer or othermedical issues,whichwealways assessprior to treatment.

“You can treat erectile dysfunction,and what we do here is actually beginpatients on PDE5 inhibitors such asViagra after theurethral catheterhasbeenremoved following surgery. There is evi-

dence to suggest that theearlier you intro-duce treatment, the more effective it islikely to be.”

The specialist nurse advises patientsthat incontinenceshouldgradually improveover time if theyundergoacourseofpelvicfloorexercises. “Asmallnumberofpatientsare never going to be completely conti-nent, related to anumberof issues, includ-ing the complexity of the surgery.”

Drudge-Coates advises patients to beupfront in questioning surgeons about

“Not all cancers have to be treated and I think this

is still quite an alien concept for most patients”

Systems&Services


For Daniel Sencier, whowas diagnosedwith prostatecancer in theUK at the ageof 58, choosing surgery feltlike cleansing himself ofthe cancer.After some bad experienceswith lost notes and otherproblemsathis localhospitalin Cumbria, he opted totravel to a specialist urology

centre at Addenbrooke’s in Cambridge, where he had a robot-assisted laparoscopic prostatectomy inNovember 2010.“I imaginedhaving radiationandmyprostatebeing fried insideofmeandendingupas thisball ofdead junkyouwouldbecarryingaroundinside you for the rest of your life.You could never be sure. Surgeryseemedveryclinical tome. Itmeant somebody lookingat something,seeing that it was bad, cutting it out and throwing it in the bin. Sothatwhen Ihadsurgerymyprostatewas inAddenbrooke’s and Iwasback in Penrith, a longway removed from it.”In a blog, Sencier described the contrast between his local hospi-tal and a specialist centre. “At [my local hospital] they are all lovely

well-meaning people and they all want to do the best by youwithevery bone in their bodies. They are just drowning and don'thave the quality people or the facilities to cope. My UrologyNurse, Jill at [the district hospital], is not just theUrologyNurse.She is a secretary, counsellor, cleaner, teamaker andmulti taskednurse. I sawabout 12different people atAddenbrookeswho all didjust a small part of Jill's job, but did it to perfection, because theysimply had the time to.”Speaking toCancer World, Sencier said, “At hospitals likeAdden-brooke’s,where theyhave the robotic surgerymachine, the surgeonshaveall beenout inAmerica for training.All thesupport staff go thereand theyhave amentorwhocomes over from theStates andworkswith the teams until they have done about 50 operations. There isahugeprogrammegoingon.While inahospital like [thedistricthos-pital]where theguydoesmaybe20operationsayear,hedoeshisbestwith the knowledge he has got. It is a lonely place for him I guess.“It is not just the removal of the cancer thatmakes thedifference atthe specialist hospitals, it is the continence and the erectile dys-function. You are more or less guaranteed if you go to Adden-brooke’s that youarenot going tobe incontinent formore thana fewweeksafterwards, but Ihear terrible stories in thechat roomsof guyswho are still incontinent several years afterwards.”

FEELING CLEANSED BY SURGERY – AT A WORLD CLASS CENTRE

incidence rates for incontinenceanderec-tile dysfunction. “I openly tell patientsthese are things you have got to be awareof because these are life-changing events.In the UK we are seeing patients cherry-pickingwhere theygo for surgerybasedontheexperienceof thesurgeonandbasedontheoutcome,whichmakesperfect sense.

“I think thiswill evolve as cancer cen-trespublish their results. Inmyexperiencepatients are asking surgeons more directquestions about complication rates andincontinence. ‘Howgoodare youas a sur-geon?’ ‘How many of these procedureshave you done?’”

Just as urologists and radiotherapyoncologists have to specialise in prostatecancer, so toodonurses.Drudge-Coates ison theboardof theEuropeanAssociationofUrologyNurses,which is in theprocessof defining the core competencies of thespecialist nurse and their training needs.

WHAT OUTCOMES SHOULD ACENTRE ACHIEVE?The ESO discussion paper publishedonline inDecember 2010 did not attemptto specifywhatoutcomesspecialistprostatecancer centres should achieve, andmaybecriticised for advocating somethingwithoutclear evidence of improved outcomes.

However,Valdagni isconfident that theevidencewill come. “We know that case-load is strongly related to thequalityof rad-ical prostatectomy andwe also know thatcaseload in radiation therapy is related toless use of secondary treatment. Thatmeans that if the centre has a high case-load andworks with a lot of prostate can-cer patients, radiation will be better andresultswill be better and secondary treat-ment for failure will be less.”


Systems&Services

“We are seeing patients cherry-picking where they

go for surgery based on the experience of the surgeon”

A NETWORK OF CERTIFIED UNITS ACROSS EUROPE

The ESO discussion paper, ‘The requirements of a specialist Prostate Cancer Unit’,1 arguesthat prostate cancer units are the most suitable structures for organising specialist multi-disciplinary care for patients at all stages of the disease, and that themultidisciplinary approachoffers patients the best chance of receiving high-quality medical procedures administered bya team of specialists, which is able to tailor treatment and observational strategies to theirneeds, and ensure access to specialist counselling, supportive care and rehabilitation. Thepaper proposes general recommendations andmandatory requirements for prostate cancerunits, with a view to laying the basis for a network of certified units across Europe.� Prostate Cancer Units are best established in large or medium sized hospitals covering

populations of at least 300,000 people and seeingmore than 100 newly diagnosed casesof prostate cancer each year, and within amultiprofessional team, where supportive careas well as clinical excellence can develop.

� Unitsmust have written protocols for diagnosis and themanagement record and on diag-nosis pathology, treatment clinical outcomes and follow-up, including side-effects and com-plications. The data must be available for audit.

� Uro-pathologists specialising in prostate disease should see at least 150 sets of prostatebiopsies a year and spend 50% of their time working in this field. Each centre should havetwoormore urologists trained in prostate cancer, each carrying out at least 25 radical prosta-tectomies a year and spending 30%of their time on prostate disease. Radiation oncologistsshould treat at least 25 prostate cancer patients a year or 15 prostate cancer brachyther-apy procedures. Similar levels of caseload and time are set for medical oncologists.

� In addition a centre should have one or more nurse specialists in prostate care, as wellas specialist radiologists, medical physicists, radiation therapy technicians, physiother-apists with special training, palliative care specialists, and professionals who can offer psy-chological support and counselling about changes in sexual function.

� Members of the Prostate Cancer Unit core teammust attendweeklymultidisciplinarymeet-ings where 90% of cases would be discussed for audit and for external verification.

� The patients’ right to information and self-determination should be respected and menoffered clear and easy-to-understand written and oral information. Patient advocatesshould be part of the network and every patient should be provided with a copy of histreatment and follow-up plan.

� Services may need to be reconfigured to staff specialist units. However, the paper saysthat such changes can provide financial savings and avoid multiple consultations.

The paper concludes that European countries “should consider the certification of ProstateCancer Units as a necessary way forward to ensure that men with prostate cancer receiveoptimal treatment and care.”1. R Valdagni et al. (2011) The requirements of a specialist Prostate Cancer Unit: a discussion paper

from the European School of Oncology. Eur J Cancer 47:1–7

c w 48 education & knowledge through people &...

Documents