cancer world 51

LATE DIAGNOSISWhy does it happen, and

how can we do better?

CAN WE TRUST THETESTERS?

Quality control inEurope’s pathology labs

LUNGSCAPEA trials network fit for the

age of personalised medicine

November-December 2012 Number 51

CancerW

orld51

Novem

ber-Decem

ber2012

AndreasEngert

Learning from Hodgkin’s

November-December 2012 I CancerWorld I 1

3 EditorialPenalised for having cancer

4 Cover StoryAndreas Engert: Learning from Hodgkin’s

14 Cutting EdgeTesting the testers

23 CrosstalkLate diagnosis: Why does it happen? How can we do better?

30 Best ReporterThe origin of a special success

37 e-Grand RoundOECI accreditation: Is yours a top-class cancer centre?

44 Spotlight OnLungscape: a living lung laboratory

50 Impact FactorGastro-oesophageal cancer – is CROSSing over so hard to do?Cetuximab dosing by rash – is the scaling of EVEREST meaningful?

58 NewsroundSelected news reports

64 My WorldThe challenges and rewards of working in cancer

EditorKathy [email protected]

Assistant EditorAnna Wagstaff

Editorial AssistantCorinne Hall

Editorial AdvisorsJacques BernierFatima CardosoFranco CavalliAlberto CostaVincent T. DeVita

Contributing WritersJaffer Ajani, Marc BeishonMariela Blum, Clive CooksonSimon Crompton, Janet FrickerHeinz-Josef Lenz, Peter McIntyreSebastian Stintzing, Anna Wagstaff

Publishing AdvisorsGillian Griffith, Fedele Gubitosi

Website LiaisonAlexandra Zampetti

Art EditorJason Harris

ProductionHarrisDPIwww.harrisdpi.co.uk

Printed byGrafiche Porpora

Cover photographJorge Nogueira

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

contents

cancerworldShaping the future of cancer care

E D I T O R I A L


uch attention is given to thespiralling cost of treating can-cer, yet the financial cost tosurvivors is rarely given anyconsideration.A recentDutch

study published in the European Journal ofCancer (vol 48, pp 2037–42) throws somelight on the problem. Nearly a third of sur-vivors experienced a change in theirwork sit-uation as a result of their cancer. Somechoseto give up work, others were unable to workbecause of their disease or had to switch topart-timework.Afewweresacked.Peoplealsofacedproblemsobtaininghealthand life insur-ance as well as mortgages. While many sur-vivors did finally manage to obtain loans andinsurance, they had to pay a higher premiumor interest rate. This study mirrors findingsfromother countries,whichhave shown thathaving cancer often leads to a loss of incomeand/or increased expenditure.

Insomecountriespatientshave tobear thecosts, either fully or partially, of theirmedica-tion, tests,proceduresandphysicianvisits.Butthere are alsomany hidden costs. People canfind themselves out of pocket because ofhigherutilitybills, theneed tobuynewclothesbecause of weight changes, and the cost ofspecial diets, wigs and prostheses. Travellingto a treatment centre can be very expensive,especially for thosewho live a longway away,and familymembersmayalsobehit financially

if theyneed to accompany thepatient, or stayclose bywhile the patient is hospitalised.

As many cancers are becoming chronicconditions, these financialburdenscanpersistin the long term,andcanbehighly stressful forpatients and their families at a timewhen theyarealready struggling tocopewith thedisease.Finding solutionswon’t beeasy given thecur-rent economic climate inEurope,with risingunemployment, cuts inwelfare benefits, andrecent increases in living expenses.

Greater awareness of the social and eco-nomic burden of cancer could help ensurethat patients are routinely asked about howthey are coping financially, and are givenadviceabout their rightsandentitlements.Thepatient survey recently launchedby theEuro-peanCancerLeagues, tobuildabetterpictureof social and economic problems, could be avery helpful contribution. Policy initiativesaimed at reducing the financial burden ofcancer are alsoneeded.These should includeprotecting employment rights and removingobstacles to accessingwelfare benefits whenneeded. A number of organisations acrossEuropearecurrently addressing this issue,buttheseefforts are sporadicandnotdelivering forall cancer patients. It makes sense to joinforces at anEUlevel to raise awareness aboutthe financial consequences of cancer and toadvocate for changes that couldmake all thedifference to patients’ lives.

KAT H Y R E DMOND ED I T O R

M

Penalised forhaving cancer

hy should we bother with Hodgkinlymphoma? It accounts for only 1%of cancers, and the probability ofcure has improved dramatically overthe past 40 years, with 94% of

patients now expected to survive. Does it reallydeserve research funds and the attention of thewider cancer community?

The answer from theHodgkin’s research com-munity is an emphatic yes. In the 1970sHodgkin’sbecame one of the first curable diseases in oncol-ogy, and ever since there have been importantdebates about best treatment and long-termrisk/benefit balance that have real relevance to thewider world of cancer.

Hodgkin lymphoma, which most commonlyaffects young adults, focuses minds on patients’

lives: how to give themdecades of life free not onlyfrom the effects of cancer, but also from theeffects of treatments used to cure them. Theseissues can only become of wider importance asmore and more of the global population experi-ences cancer as a long-term condition. In thewords of a recent editorial by Joseph M Connorsin theNew England Journal of Medicine, Hodgkin’sis “The Great Teacher”.

A leader in the quest for more acceptable, lesstoxic, approaches to treating Hodgkin lymphomais Andreas Engert, chairman of the GermanHodgkin Study Group (GHSG) and professor ofinternal medicine, haematology and oncology attheUniversityHospital of Cologne.He knows onlytoo well that the problemwithHodgkin’s is not somuch how to cure it, but how not to kill with the

Hodgkin’s patients survive for longer – but with more serious and lasting

damage – than almost any other group of cancer patients. Finding ways

to address the problems of both current and future survivors can provide

valuable lessons for other cancers, argues Andreas Engert.

C O V E R S T O R Y

S IMON C R OMPTON

W

4 I CancerWorld I November-December 2012

Andreas Engert:Learning from Hodgkin’s

treatments. Engert was only 14 when his fatherwas diagnosed with Hodgkin lymphoma. He wascured with chemotherapy, but then died 16 yearslater – when Engert had already embarked on hiscareer in oncology – from cardiovascular problemsresulting from treatment.

Such outcomes are notuncommon. Studies indicatethat the risk of developingneoplasms after treatmentfor Hodgkin’s is 22% at 25years – an 18-fold increasedrisk compared to the rest ofthe population. Peoplewhohave survived Hodgkinlymphoma also have a sig-nificantly increased risk ofcoronary artery disease,valve disease, congestiveheart failure, pericardialdisease, stroke, arrhythmiaand sudden cardiac death.

Since his father’s death, Engert’s career has beendefined by searching for alternatives to currentchemotherapy and radiotherapy regimens. It hasbeen a quest that has seen him following andthen taking up themantle fromVolker Diehl – theman who treated his father and one of the Euro-

pean giants inHodgkin’s research – and thenfirmly establishing his centre as a worldleader in developing and evaluating newapproaches for haematological andsolid cancers.

In the process, Engert has some-times found himself cast as a youngradical, determined not to be boundby history, restless to move theagenda forward, and sometimesin profound disagreement withmany in the Hodgkin’s commu-nity about the right balancebetween effectiveness and toxicity

in treatments.But todayEngert, now assistant direc-

tor of the Department of Internal Medi-cine at the University Hospital of

Cologne, believes that the

JORGENOGUEIRA

C O V E R S T O R Y


long search for less lethal cures may have reacheda vital juncture. The days of tortuous debate aboutwhich chemotherapy regimens are best may benumbered.

Engert is keen to talk about a new targeted drug– brentuximab vedotin, codenamed SGN-35 –which he says may be about to change everythingin Hodgkin lymphoma. Results from trials in theUS presented atASCO last year showed that thedrug induced remission in 75% of patients withrelapsed or refractoryHodgkin’s, with 35% achiev-ing long-lasting and complete remission.

What is more, because it is an antibody-drug-conjugate – a combination of an antibody and adrug guided safely to its target – the dangers of sys-temic treatment and radiotherapy are avoided.The drug was registered in theUS last year, and isexpected to be registered in Europe soon.

“This is the single most effective drug we nowhave for Hodgkin’s,” says Engert, whose Germangroup has successfully trialled the drug in refrac-tory or relapsed disease. “It’s thrilling for me – andI’m not just saying this becausewe getmoney fromthe pharmaceutical company to conduct clinicaltrials. It’s thrilling because I worked on linking anti-bodies with plant toxins in the late 1980s when Iwas at the Imperial Cancer Research Fund. Trialseventually showed that these immunotoxins werenot good enough because they were immuno-genic [produced an immune response]. But now,20 years later, this company has produced a verywell-tolerated combination. So that story, for me,has now come full circle.”

The story has come full circle in more waysthan one. Six years ago, Volker Diehl – who led thedevelopment of the BEACOPP chemotherapyregimen for advanced Hodgkin lymphoma –described the regimen as a “great poison” in aninterview with Cancer World. “I would like tohave something better,” said Diehl, who had spentdecades working on improvements to chemother-apy. “Sometimes I wake up in the night and ask

what will happen tomy young patients in 10 to 15years.”

NowEngert,whowas co-ordinating editor of theCochraneHaematologicalMalignanciesGroup for10 years, believes that “something better”may havearrived. What is more, after 30 years when mono-clonal antibodies were being developed for othercancers, but therewere nonewdrugs forHodgkin’s,several new targeted drugs for the disease are nowin trials.

“The drug companies are now knocking on ourdoor,” says Engert. “Before, they always said it wastoo small a market – that since you cure most ofthese patients anyway, there’s no use looking at the10–20%who are not cured. Theywanted to investtheirmoney in other areas such as lung cancer.Wehad discussions for many years, with many drugcompanies. I remember in the 1990s we found abispecific molecule that was pretty effective andScheringwas interested in it, but the guyswho hadthe money did the calculations and said no.

“It has been frustrating, given progress on newtargeted drugs for other cancers. My interest hasalways been on antibodies, since I was a medicalstudent. I remember that when rituximab firstcame out for non-Hodgkin lymphoma, nobodywas interested. They said antibodies are fine fordiagnostics, for distinguishing between differenttumour types. But treatment?Noway.Wewere oneof the first European centres to take rituximaband use it in combinationwith chemotherapy, andfor diseases like chronic neutrophilic leukaemia.”

Born 1959 in Braunschweig, Germany, Engertrebelled against his father’s wishes that he shouldbecome a banker, and decided to study medicine.Whenhewas amedical student, his father kept urg-ing him to go and see Volker Diehl inHanover, themanwhohad curedhisHodgkin’swith chemother-apy. “I said no,” saysEngert,with a smile. “Whateveryou say, I do exactly the opposite – that’s whatmany youngmendo I guess.”As amedical student,he had initially focused on psychiatry and psycho-

C O V E R S T O R Y


“They said antibodies are fine for distinguishing

between tumour types. But treatment? No way”

somatics, then decided it wasn’t for him. He knewhe didn’t want to go into surgery, and he also knewthat it was internal medicine – immunology,nephrology and oncology – that he found mostinteresting. So he changed his mind about Diehl.

“I realised that he was a really interesting doc-tor. I worked for him, wrote my thesis with him,and then followed him. My father and VolkerDiehl were certainly very influential on the direc-tion I took.” IndeedDiehl –who famously culturednotoriously fragile Hodgkin cell lines for the firsttime, opening up new worlds of research possi-bilities – seems to have exerted an almost gravita-tional force on the trajectory of Engert’s career.

WhenDiehl moved fromHanover to theUni-versity of Cologne, to become director of theDepartment of Medicine and chairman of the

GermanHodgkin StudyGroup, Engert took up theoffer to come with him – despite other offers towork in oncology, nephrology and immunology inHanover. Heworked withDiehl for two years, butwhen offered a six year contract, he again rebelled.

“I said I didn’t want it – I want to go intoresearch.” That was when he spent two and a halfyears researching antibody-based immunother-apy for patients with malignant lymphoma underprofessor Philip Thorpe at the Imperial CancerResearch Fund in London: “It was a great time, tolearn all the techniques and have time for scienceonly.” But then Thorpe went to Dallas. Engertdidn’t want to follow him to America but he didwant to continue work on trialling the promisingimmunotoxin drugs he had helped develop inLondon. So in 1991 he decided to rejoin Diehl in

C O V E R S T O R Y


“It was a great time, to learn all the

techniques and have time for science only”

JORGENOGUEIRA

Cologne, where as head of the laboratory ofimmunotherapy he could conduct phase I and IIclinical trials on the new drugs – which ultimatelyproved disappointing.

Engert has remained there ever since, becom-ing deputy medical director in 1999, and takingover from Diehl as chairman of the GHSG in2007. He fell in love with Cologne, he says. Asense of the city’s scientific heritage hangs aroundthe vast atrium of theUniversityHospital where heworks, with photographs and memorials to thelikes of medieval experimenter Albertus Magnusand heart radiography pioneer FriedrichMoritz.

As for the shadow of his predecessor, Engerthas a pragmatic view – respectful, but clear that ifthings are to progress, past achievements have tobe viewed as inadequate. There is little room forsentiment in medicine.

“I had hadmy own interest in immunotherapyfor a long time and brought a lot of new treatmentaspects into the group,” he says. “I supportedVolker and worked with him, and when he left thehospital and I took over, there were many thingsI maintained – but it was good for me to havemore freedom to run things as I would haveloved to before.”

“Over these years the knowledge and experi-ence of this group has broadened substantially,” hesays. “Volker’s work allowed us to become one ofthe leading groups on this disease worldwide, and

we really appreciate that. But things have evolvedand adapted to modern aspects of treatment andhow studies have to be done.” The GHSG todayrecruits patients from more than 400 centres infive European countries.

The increasing influence of the group has ledit into intense exchanges with other groups –debates about best treatment that Engert believesare extremely important for clinical progress, andwhich he admits to finding enjoyable. “It would beboring if everyone had the same opinion and therewas no more development,” he says.

One of the most important has been aboutwhat should be the standard chemotherapyregimen for advanced Hodgkin lymphoma. Thedebate, conducted over the past five years on thepages of themost high-powered professional jour-nals, has seen the German group pitted againstgroups inAmerica, the UK and Italy.

It started in 2004 when Engert andDiehl pre-sented the findings of amajor studywhich showedthat theGerman group’s BEACOPP regimenwas20% better at tumour control and 11% better foroverall survival than current standard chemother-apy regimens revolving around ABVD (doxoru-bicin, bleomycin, vinblastine and dacarbazine).What is more, the benefits became clearer thelonger patients were followed up.

But American study groups took issue:BEACOPP was too toxic for standard use, they

C O V E R S T O R Y


Hodgkin lymphoma, also known as Hodgkin’s disease, is a cancer found in the lymph nodes. Most commonlyit starts in the lymph nodes of the neck. One in five lymphomas are Hodgkin.Hodgkin lymphoma can affect people of any age, but it most frequently affects two groups: those aged 15–35,and those aged over 55.It is characterised by the presence of Reed-Sternberg cells – B-lymphocytes that have become cancerous. Non-Hodgkin lymphomas do not have Reed-Sternberg cells.The malignant cells are large – but they are widely scattered compared to other cancers. While in lung can-cer or other lymphomas, pathology reveals a high density of malignant cells, Hodgkin cells are surrounded bybenign ‘bystander’ cells. Research by Engert showed that if just the malignant cells are destroyed, the bystandercells disappear as well.Until Volker Diehl first successfully cultured Reed-Sternberg cells in 1978, a significant barrier to researchingthe condition was the fragility of the malignant cells once removed from the body. Even today, when there havebeen over a quarter of a million attempts to culture Reed-Sternberg cells, there are only 14 cell lines in theworld. Five of these were cultured by Diehl.

HODGKIN LYMPHOMA

said. Engert acknowledges they had a point: itis more toxic than ABVD. But what theseother international groups – the “anti-BEA-COPP alliance” as Engert calls them – havefailed to acknowledge, he claims, is that it isalso much more effective.

Another study, published in theNew Eng-land Journal of Medicine last July by Vivianiand colleagues at the Istituto NazionaleTumori in Milan, indicated that BEACOPPwas 12% better at tumour control thanAVBD. But, according to Engert, the articleand accompanying editorial instead concen-trated on the fact that overall survival withBEACOPP was no better than with theless toxic ABVD – “which is not sur-prising given the small number ofpatients in the study.”

“They were basicallysaying this study showedthatABVDwas as goodas BEACOPP, whichwe think is unfair.We always said thatBEACOPP is moretoxic thanABVD. Butwe all have to stick tofacts and not overshootin our attempts to dis-prove one another.”

TheIstitutoNazionaleTumorimeanwhile is emphatic that “we havenot hidden or confused themessage.”

The debate is moving on again, with the Ger-man group providing evidence in a Lancet articlethis May that less toxic doses of BEACOPP areequally effective as larger doses. Engert is all tooaware of the awkward balance between risk andbenefit, acknowledging that sometimes a highrisk is necessary whilst doing everything possibleto minimise it.

What suchdebates reveal toEngert is the impor-tance of following the patient’s story for as long as

possible after treatment,not jumping to conclu-sions after short-term

studies. “The stories of ter-rible side-effects aren’t over

yet,” he says. “Hodgkin’s is a greatteacherbecause youcan followpatients

carefully through to the end – the consequencesof their treatment may not be visible for 20 or 30years.Weare certainly keen to get rid of chemother-apy and radiotherapy.However, this certainly cannotbe done at the price of a very high relapse rate.”

It is natural enough that patients, oncologistsand indeed drug companies should want to con-centrate on cure. But people who have beentreated for Hodgkin’s form the biggest cancer sur-vivor group inWestern countries, with 70–80,000in Germany alone, so that researchers and

C O V E R S T O R Y


“We all have to stick to facts and not overshoot

in our attempts to disprove one another”

The Engertfamily circle

clinicians are able to look far beyond remission.Hodgkin’s, says Engert, can serve as a ‘model can-cer’, providing answers not just on side-effects, butalso on how to support patients long-term throughenduring problems such as infertility and fatigue.

This human element of the oncologist’s role isimportant to Engert, who has always tried to bal-ance laboratory workwith clinical work.One of thereasons he went into oncology was that it seemedless mechanistic, more based on long-term rela-tionships with people, than specialties such as car-diology. The relationship is not always a simple one,he acknowledges. Patients tend to focus solely oncure at the start of their journey, and oncologistshave to be careful to counsel them about side-effects such as infertility, hair loss and fatiguethat will become ofmajor importance to them later.The complexity of some decision-making requiresmeaningful collaboration.

“Pretty early on I realised that cancer patientsgive back a lot,” says Engert. “They realise theyhave to workwith you, andwant to co-operate. Yousee them develop in this respect, whereas in otherareas such as cardiology, people are hardly off theintensive care unit after heart attack and they’re offto work again, and there’s no long-term workingrelationship with the doctor.”

What about Engert’s human side? Given themotivating role played by his father’s experienceswith cancer, how much does his own young fam-ily know about his work andwhat he is trying to do?Aged 53, married to an artist, he has four childrenaged between 10 and 17, and their photographscover one wall of his office. He tries to devote asmuch time as possible to home life, but thedemands of work and international conferencesdon’t make this easy and sometimes he brings hischildren into work or takes them to conferenceswith him.

Sometimes he has his doubts about howmuchof his work his children really understand, but hedoes know that, among friends and family of the

children’s friends, there have been several withcancer who have askedEngert for advice. The chil-dren know about this and discuss it – there is notaboo around cancer. And Engert admits that hewould be “honoured” if any of them decided to fol-low him intomedicine. Perhaps that story too willcome full circle.

If it does, the challenges his children facemaybe very different than Engert’s. “I think this newdrug, and others to come after it, will changetreatment for Hodgkin’s fundamentally,” he says.Already well-tested in relapsed patients, brentux-imab vedotin is now undergoing trials as a main-tenance treatment, with the aim of reducing thenumber of relapses. Engert believes that in twoyears, therapy for relapsed patients will havechanged substantially, and within a decadechemotherapy and radiotherapy will no longer bemainstays of treatment.

Does that mean everything is rosy? Not quite.For those who specialise in Hodgkin’s, there isalways the nagging worry about the unforeseenfuture – the worries about treatment side-effectsthat gave Volker Diehl sleepless nights. Engert tooknows the risks that can come with new treat-ments. In 2006 he advised the biotech companyTe Genero not to test its antibody TGN1412 onhealthy volunteers – advice the company rejected.As a result, 10 volunteers in London sufferednear-fatal side effects. “I told them it was nonsenseto try this drug on healthy volunteers in this typeof trial – you can’t do this. But they did.”

It all goes to show that there is no room forcomplacency in treatment development. “Drugsaren’t like water,” says Engert. “If we had a shinynew side-effect-free drug, and then a year after itsintroduction patients started dying from second-ary cancer then, yes, I too would start havingnight sweats and nightmares.”

For the moment, he is cautiously confident.But he knows only toowell that the greatest arbiterof treatment success is time.

C O V E R S T O R Y


“Pretty early on I realised that cancer patients

give back a lot, they want to co-operate”

s the era of personalisedmedicine has developed,it has brought with itthe need to conduct anincreasing number of bio-

logical tests to determine the appropri-ate personalised therapy. Key examplesare HER2 status for response to Her-ceptin (trastuzumab) and the KRASmutation, which predicts for a poorresponse to EGFR inhibitors such asErbitux (cetuximab).

Now the floodgates are opening, asincreasingly sophisticated genetic testingof tumours is revealing a raft of genemutations, dislocations and fusions thatare the target of current or future drugs,notably in lungcancer,where in just a few

years possible treatment options havechangeddramatically. The sheernumberof patients where molecular testing willbe essential – such as the many withmetastatic colon or lung cancer – is a bigchallenge now.

Cancer centres are pressing aheadwith projects that carry out several assayson tumours to give more information topatients and their oncologists, and toidentify people eligible to join clinicaltrials.At national level in countries suchas France and the UK, cancer policymakers are rolling out andmaking prepa-rations for the routine testing of a rangeof genetic mutations in the major can-cers such as breast, colorectal and lung.

These projects highlight one of the

big challenges with molecular bio-markers – ensuring that testing is carriedout to a consistently high standard. AsCancer Research UK, which is organ-ising theUK’s StratifiedMedicine Pro-gramme for a panel of genetic tests,comments about the earlier introductionof Herceptin, “Having to routinely,reliably and accurately test a breasttumour’s HER2 levels, as part of ‘busi-ness-as-usual’, was uncharted territoryfor many pathologists.”

Indeed, Giuseppe Viale, professorof pathology at the University of Milan,has written extensively about contro-versies in HER2 testing, and has notedthat despite decades of experiencewith assays, and the availability of

Treating the right patient with the right therapy at the right time

depends on having the right picture of every patient’s key biological

markers. A number of initiatives are now springing up across

Europe to ensure that the testers are indeed getting it right.

C U T T I N G E D G E

MARC B E I S H O N


A

Testing the testers

standardised reagents, kits and guide-lines, “the accuracy and reproducibil-ity of the test results are still a majorconcern worldwide”. Results of severalquality-control schemes for pathologylaboratories have shown as many as15–20% false-positives with immuno-histochemistry testing, and there canalso be a high error rate in anotherkey testing method, fluorescence insitu hybridisation (FISH).

But it is also true that we still don’t

understand enough about the biologyof breast cancer to be certain abouthow HER2 testing should best be car-ried out and interpreted. Tumours arenot uniform in biology, and HER2 sta-tus may change over time. Uncertaintyalso remains about the right thresholdvalues for predicting who could bene-fit from Herceptin. “Many oncologistsseem to believe that pathologists arenow infallible in assessingHER2 statusof breast cancer,” Viale notes, adding

that it is frustrating that oncologistsare pushing for more precise identifi-cation of targets for new agents inbreast cancer while basic questionsremain about HER2 testing (see Con-troversies in testing for HER2, ASCO2011 Educational Book).

KRAS testingThe good news is that tests for otherpredictive molecular markers do notall suffer from the same complexity.



“Many oncologists seem to believe that pathologists

are now infallible in assessing HER2 status”

CAN

CER

RESE

ARC

HU

K

And although quality concerns are stillvery much to the fore, they have beentackled much more systematically inthe case of KRAS, for example, as Hanvan Krieken, professor of pathology atthe Radboud University NijmegenMedical Centre, explains. “Indeed,HER2 testing is still far from optimal.One important reason lies in the biol-ogy. It is not completely known whatfactor is the most important for pre-dicting trastuzumab response: theamount of protein or the regulation ofthe expression of the gene or the num-ber of copies of the gene. The otherimportant factor is that early on therewas limited interest in the qualityaspects. The situation with KRAS isvery different: it is clear that activatingmutations in the gene are the indicatorand within two years of this discoverythere was quality assurance.”

KRAS mutations are important,because they occur in around 40% ofpatients with colorectal cancer, andthey prevent EGFR inhibitors such asErbitux and Vectibix (panitumumab)from working. As the large majority(almost 85%) of patients with colo-rectal cancer have tumours that over-express EGFR, knowing who will andwho will not benefit from an EGFRinhibitor is essential to avoid subject-ing large numbers of patients to treat-ments that won’t do them any good.Guidelines for using EGFR inhibitorstherefore advise that testing for EGFRstatus alone is not enough; you alsoneed to test for a variety of mutationsin KRAS.

As van Krieken adds, not all of the60% who do not have KRAS activatingmutations will benefit, and there isnot enough evidence yet to further

separate out a target group from thismajority. “But KRAS testing is com-mon now – I don’t think anyone isgiving the anti-EGFR drugs in colo-rectal cancer without it.” What hasbeen critical, he says, is participationin quality assurance programmes atnational and international levels, suchas the European KRAS EQA (externalquality assessment) scheme – a Euro-pean programme he proposed withcolleagues at the European Society ofPathology in 2008, and which is now inits fourth round.

External quality assessmentElisabeth Dequeker, the KRAS EQAscheme coordinator, who is based atthe Biomedical Quality AssuranceResearch Unit in Leuven, Belgium,says the first step had been a pilot toestablish regional laboratories thatcould prepare slides for the scheme –a quality control exercise in itself.These regional laboratories now sendsamples for which the KRAS muta-tion status is known, but not revealed,to laboratories participating in thescheme, to check they are getting accu-rate results. Between 2009 and 2011,150 labs in Europe and in countriessuch as Indonesia and Israel reached agenotyping score of 90% or better inthe scheme. “Some countries such asthe UK and Germany also have theirown national assessment schemes, butmost smaller countries do not, andhere in Belgium for several years now itis a requirement that all molecularoncology tests must be accredited. Thebest way of validating tests is with anEQA scheme,” says Dequeker.



“Our scheme has found a significant number

of errors that could have had a big effect on therapy”

CAN

CER

RESE

ARC

HU

K

Given that there are several ways oftesting for KRAS mutations, and also achoice of commercial kits and home-grown systems, the EQA scheme hasnot attempted to distinguish betweenapproaches. “If you implement yourmethodology well and validate itaccording to standards your methodwill be stable,” she says. “But molecu-lar pathology is not easy and ourscheme has found a significant numberof errors that could have had a bigeffect on therapy.”

Van Krieken reports that, from thefirst round of the KRAS EQA scheme,85% of laboratories had a 100% scoreon the samples, but 15% still had a sub-stantial number of mistakes. “One mis-take is in selection of patient material– you need to have sufficient tumour inyour samples, and you may need toenrich them by cutting out normal tis-sue. The other main problem was ininterpretation of results – some labswere making mistakes even using com-mercial systems and were missingsome of the mutations.”

Some labs may have to pull out ofmolecular testing if they can’t improve,says van Krieken, but the indicationsare that with the right quality and edu-cational schemes in place there will befew that cannot make the grade. In2009, the Italian Association of Med-ical Oncology and the Italian Society ofPathology and Cytopathology ran aKRAS quality scheme similar to theEuropean one. It found that only twoparticipating centres out of 59 failed a100% pass rate, and those two alsofailed a retake, being unable to extractenough genomic DNA for the muta-tional analyses.

PREDICTIVE BIOMARKERS AND TARGETED THERAPIES

The Italian quality scheme sent sam-ples with a particularly high (>70%)content of tumour cells, but the organ-isers said that the impressive resultsmay also reflect the benefits of anongoing networking programme (calledKRAS aKtive).

Lung cancer mutationsThe European Society of PathologyEQA scheme has this year beenextended to non-small-cell lung cancer(NSCLC) in two phases, first just theALK mutation, which has generatedmuch interest in oncology, and then asecond round covering ALK, KRASand EGFR, says Dequeker. France,which has its own quality schemes, iscollaborating with Leuven on validatingEGFR and BRAF programmes – the

latter, for advanced melanoma, is turn-ing out to be quite a challenge.

“What we are finding is that eachtest has its own problems,” says vanKrieken. “For example, the quality ofDNA in a skin cancer biopsy can beaffected by melanin, and sensitivityissues are very important in lung canceras samples can be small. Each test andtumour needs its ownEQAprogramme– you cannot assume a lab doing wellwith KRAS in colon cancer will also dowell with say BRAF for melanoma.”

Rolf Stahel, professor of oncology atZurich University Hospital, Switzer-land, and a lung cancer specialist, com-ments that the diagnostic possibilitiesfor NSCLC are becoming very com-plex, although so far only EGFR muta-tions are recommended for testing at



“You cannot assume a lab doing well with KRAS in colon

cancer will also do well with BRAF for melanoma”

Biomarker Cancer type Drug EMA approval

BCR-ABL Chronic myeloid or acute Imatinib 2001translocation lymphoblastic leukaemia Dasatinib 2006

Nilotinib 2007

KIT and PDGFRA Gastrointestinal Imatinib 2002mutations stromal tumours

HER2 amplification Breast cancer Trastuzumab 2000Lapatinib 2008

HER2 amplification Gastric cancer Trastuzumab 2009

KRAS mutations Colorectal cancer Panitumumab 2007Cetuximab 2008

EGFR mutations Non-small-cell lung cancer Gefitinib 2009Erlotinib 2011

ALK translocation Non-small-cell lung cancer Crizotinib Not yet approved*

BRAF V600 mutation Melanoma Vemurafenib 2012

*USA FDA approval obtained in 2011. Abbreviation: EMA, European Medicines Agency.Source: F Nowak, J-C Soria and F Calvo (2012) Nat Rev Clin Oncol 9:479–486, reprinted withpermission, © Nature Reviews Clinical Oncology



the first-line stage by the ESMO con-sensus panel on NSCLC pathologyand molecular testing, on which heserves. The targeted drugs on offerhere are the EGFR blockers Iressa(gefitinib) and Tarceva (erlotinib), withothers in the pipeline, and there is asizeable population – 10–12% of thosewith NSCLC.

“Now ALK testing is also recom-mended, but in second-line treatment,”says Stahel. “But here the testing isnot easy as we don’t have a final wordyet on how best to test for it – will it beFISH, or immunohistochemistry andthen FISH, or genetic sequencing.Andit is in only about 5% of patients.”

WhileALK testing has now movedfrom the investigational stage into clin-ical practice, the key drug, Pfizer’sXalkori (crizotinib), was only recom-mended to receive conditional approvalin Europe a few months ago, with thefinal decision still pending. Othermutations of interest, such as HER2,ROS1 and RET, are lower still in fre-quency, at 1–2%. This is now raisingquestions such as whether to prescreenwith a cheap assay and then apply amore precise and expensive test on asubgroup. Stahel adds that testing is a“tremendously fluctuating field” as wemove up the hierarchy from lookingfor single mutations, with traditionalimmunohistochemistry testing nearthe bottom, to more sophisticated testssuch as FISH, to exome sequencing atthe top – and indeed whole genomesequencing ultimately.

“And if we test for a huge string ofmutations we may also get results wedon’t want to know,” he says. “The

pathologists here in Switzerland aregung ho about buying new sequencingmachines – but how are we going tocope with the information in the clinic?How do we communicate results thatare nothing to do with lung cancer?”

Stahel’s own focus is very much onidentifying the right clinical questionsand ensuring high-quality testing. Heleads the European Thoracic OncologyPlatform(ETOP),which in itsLungscapeproject is analysing a large cohort ofNSCLC tumours for a panel ofmolecu-lar characteristics such as ALK, andwhich is playing a part in the EuropeanEQA scheme, among other activities.(Formoreon this, seeLungscape: a livinglung laboratory, on p 44.)

A regional networkOne potential way forward with lungcancer is outlined by Thomas Zander,a lung oncologist at Cologne UniversityHospital in Germany, who attractedattention at this year’s ASCO with anabstract that described a regionalscreening network that has been set upfor molecular testing of lung cancerpatients at community hospitals in theCologne-Bonn area. “The primary aimis to provide both well-establishedmarkers such as EGFR and ALK andalso new diagnostics to communityhospitals, so each lung cancer patienthas the opportunity to know every-thing that is meaningful about theirtumour,” says Zander. He calls theprocess ‘spreading’ – running assayson a panel of mutations, so thatpatients have opportunities to partici-pate in trials and to obtain new drugsbefore approval.

“TakeALK – it is not yet in routine clin-ical testing or treatment with crizo-tinib, but there is a consensus that ifyou test and find it you should treat. InGermany, although you have to askspecifically for reimbursement, wehave not had a problem in a singlecase, which is a good indication thatthe health community agrees it shouldbe done. What we can do then is makedrugs such as crizotinib available a yearor so before approval.”

Crizotinib is being offered topatients via their primary oncologist,along with EGFR inhibitors, inZander’s area. Other mutations beingtested in Cologne include BRAF,KRAS, PIK3CA and ERBB2, with1750 samples being tested so far, esti-mated at 60–70% of NSCLC cases inthe region. Obtaining the right qualityof material for lung cancer diagnosticsis much more of problem than withcolon cancer, he says, given the diffi-culty of gaining enough tumour cellsfrom a lung biopsy.

Zander adds that the lung cancertesting represents a step-change in thenumber of assays being performed,using the array of technologies cur-rently available. “We started with fouror five assays, and we are already up to14, and in a few years it could be 40,”he says. What is needed are technolo-gies that ‘multiplex’ – measure simul-taneously – the various mutations, andwhich home in only on those that areallowed to be searched for.

Zander agrees this is a challenge.“That’s the great problem with newgene sequencing technologies – soon itwill be cheaper just to sequence every-

“And if we test for a huge string of mutations

we may also get results we don’t want to know”

thing, but for legal reasons we are notallowed to analyse anything else with-out asking the patient.” So far, assays atCologne do focus only on key lungmutations, he says, while data hand-ling has not been an issue.

Cost is amajor issue of course – out-side of approved EGFR drugs there isno reimbursement for the often expen-sive diagnostics being done inCologne,and Zander is reliant on grants.

National testing schemesMeanwhile national schemes wherecentralisation is the theme are under-way. Spain has taken early steps withdemonstrating how national EGFRscreening can work, as describedin a much-cited paper from 2009by Rafael Rosell and colleagues atthe Spanish Lung Cancer Group(see Screening for epidermal growthfactor receptor mutations in lung

cancer; NEJM 361:958–967).France probably has the most

advanced national molecular profilingprogramme in Europe. Running forfour years now, the French NationalCancer Institute (INCa) and the Min-istry of Health have set up a nationalnetwork of 28 regional moleculargenetics centres that carry out not justpredictive tests such as for HER2,EGFR andKRAS status, andmutations

MAP

OF

UK:C

AN

CER

RESE

ARC

HU

K



“Obtaining the right quality of material for diagnostics is

much more of problem with lung than with colon cancer”

Lille

AngersTours

Rennes

Nantes

Brest

CaenReims

Paris NancyStrasbourg

Dijon

Lyon Limoges

Poiters

Grenoble St.-Étienne

Bordeaux

MarseilleMontpellier Nice

Toulouse

Rouen

Besançon

St. Cloud

Villejuif

Clermont-Ferrand

NATIONAL TESTING PROGRAMMES

The French molecular profiling programme carried out55,000 tests across 28 regional centres in 2011. TheUK system is seeking to enrol 9000 patients withcancers that are treatable with targeted therapies, in thefirst stage of its Stratified Medicine Programme, as apilot for introducing molecular profiling throughout theNational Health Service



in leukaemia and GIST that predict forresponse to Glivec (imatinib), but alsodiagnostics for factors such as chro-mosomal abnormalities in sarcomasand lymphomas, and prognostic tests toguide treatment in tumours such asneuroblastoma.

An idea of the scale is given in apaper by Frédérique Nowak and col-leagues: in 2012, an estimated 40,500patients will be diagnosed with colo-rectal cancer, of whom 17,500 willneed to be tested for KRAS mutationstatus, which will be provided free ofcharge to patients and hospitals, withcompensation to local pathologists whohave to ship the tumour blocks to theregional centres. In 2011, more than55,000 predictive tests were carriedout in total, mainly in leukaemia,breast, colorectal and lung cancers.

The centres are also preparing theground for the fast introduction of newagents, as Zander’s project is doing forlung cancer.A new-therapy programmeis targeting biomarkers that are alreadyin use in clinical trials. It was launchedfirst for melanoma, lung and colorectalcancers – lung samples, for example,are being screened not just for EGFRbut also for BRAF, KRAS, P13KCA,ALK and HER2, while an INCa-funded lung cancer database aims tomake the most of linking molecularresults with clinical and follow-up data.

Nowak and colleagues note too that“molecular tests performed on solidtumours are plagued by various sample-related and methodological problems”and that testing labs face a “perma-nent evolution of their daily practice”,which is challenging for those that are

not research institutions. This is blur-ring the distinction between serviceprovision and translational develop-ment, and needs the joint input ofpathologists, molecular biologists andclinicians (for more on this see Tumourmolecular profiling for deciding therapy– the French initiative, Nature RevClin Oncol 2012, 9:470–486).

The UK also has an ambitious proj-ect in the Stratified Medicine Pro-gramme, headed by James Peach at themajor charity, Cancer Research UK,with industry support.Many thousandsof solid tumour tests are already beingcarried out in UK clinics, in particularEGFR and KRAS, but Peach says theproject will show how they can be donecollectively, mixing “say KRAS for colo-rectal cancer with BRAF and othermarkers to give a care choice to clini-cians or accrual to a clinical study forresearchers. We want to prove thatstandard tests such as KRAS can belinked to others in a single panel test,”he says.

In the first phase of the programme,surplus material is being taken athospitals with consent from about9000 people with breast, colorectal,lung, prostate, ovary or melanomatumours, whichwill be tested for about20mutations at three laboratories (over2600 patients had enrolled in the pro-gramme by June 2012). The data willalso be entered into a registry databasealongside clinical outcomes. The keyaim is to demonstrate to theUK’sNHShow quality-assured molecular testingcan be done, with associated bioinfor-matics systems, at a low cost – the sumof £300 (€380) or less is mentioned

for delivering a result that could haveseveral markers.

Again, cost in the UK, as in otherhealth services, is a sensitive issue –industry has stepped in to fund testingin the NHS and in other countries tosupport oncologists who were finding ithard to organise tests even for approveddrugs. Merck Serono, for example, hasbeen paying for KRAS tests to pavethe way for Erbitux, whileAstraZenecainitially funded EGFR tests for Iressain the UK.

Peach says the UK is going througha similar learning curve to other coun-tries in testing for mutations suchas KRAS, and quality assessmentschemes are currently underway fromthe UK National External QualityAssessment Service (NEQAS). ABRAF pilot for melanoma has alsostarted. Networking with other coun-tries, such as France, will be important,and also with industry.

While oncologists and pathologiststry to digest this already challengingagenda, the information about cancermutations just keeps on coming. InJuly this year, researchers at theCancerGenome Atlas in the US published inNature a ‘comprehensive molecularcharacterisation of colorectal cancer’,finding no fewer than 24 significantmutated genes.And in September, theCancer GenomeAtlas also found thathalf of squamous cell lung cancers, forwhich there are no current targetedtherapies, have mutations that may besusceptible to current or new drugs –some of these no doubt will soon beadded to Europe’s expanding molecu-lar testing programmes.

“The key aim is to show how quality-assured molecular

testing can be done, linked to bioinformatics, at a low cost”

Why does it happen? How can we do better?Late diagnosis


C R O S S T A L K

either through treatingpatients symptomaticallyorbyrelating symptoms to a health problem other thancancer…anda failure to fully or adequately examinepatients,useof inappropriateor inadequate tests, andreceivingor failing to followup inconclusivenegativeor false-negative test results” (Br J Cancer,101:S92–S101). But GPs see several hundred patients withpotential cancer symptomsevery year, ofwhomonlyahandfulwill turnout tohave thedisease.Theyoftenface pressure not to overload specialists and diag-nostic services unless there is a very strong basis forsuspicion. Is it fair to lay theblame for late diagnosison GPs? Cancer World’s Anna Wagstaff asked twoexperts – a general practitioner and a specialist ingastrointestinal cancer – to discuss the issue.

atients whose cancers are diagnosedlate are more likely to require treat-ments that aremoredebilitating,moreexpensive, andyet less likely to result ina cure.With experts in countries such

as Denmark and the UK now identifying earlierdiagnosis as key to improving their cancer outcomes– with the potential to prevent an estimated5000–10,000deaths a year in theUK– the spotlightis falling on general practitioners, and their capacitytoaccurately spot suspicious symptomsandfast-trackpatients for further investigations or specialist con-sultations. International studies have identified themost common themesassociatedwithdelayed refer-ral across cancer sites as “misdiagnosis, occurring

P

C R O S S T A L K


I agree that symptoms in cancer sometimesappear late andareoftenatypical,which iswhythey are recognisedby thepatient or thephysi-cian late. And when GPs are presented, forexample, with functional abdominal symp-toms– vague abdominal pain, a little anorexia,somedyspepsia– it canbedifficult for themtofilter those that might be related to a cancerfrom those that are not. However, sometimesthere is a problem with inadequate examina-tion. For instance, in the case of colon cancer,apatientmay report rectal bleeding; thedoctorsees the patient has haemorrhoids and con-cludes, “OK, it is probably linked to thehaem-orrhoids and it would not be appropriate toexamine the colon.” So theydecidenot to test.Or when tests are done, they don’t alwaysuse the most appropriate ones. For instance,in patients with dyspepsia and a little bit ofweight loss, theymay do anultrasound of the

pancreas; they find nothing, and they say thepancreas looks good.Butultrasounddoesnothaveperfect sensitivity–aCTscan isbetter inthis situation.With some tests and examina-tions, physicians sometimes do two or eventhree screening examinations where nothingis shown, because of the lack of sensitivity.

So the problems are not doing the appro-priate test, not going far enough with somepatients, and inotherpatientsdoinga first testwith a lower sensitivity and then taking thedecision – OK let’s wait a couple of months,and if it deteriorates wewill go further.

It’s true that there can be psychologicaldownsides to testing. But not being testedwhen you have symptoms can cause greateranxiety.And failing to rule out cancer as a pos-sible cause of symptoms could have greaterconsequences than creating anxiety.Youhaveto weigh the risks against the benefits.

Finding the needle – the single patient atcancer risk – in the haystack of patients withtemporary and benign symptoms, is at thecoreofgeneralpractice.So it is fair toholdGPsresponsible for delayed referral anddiagnosisof cancer patients. But the task is not easy. InDenmark, a GP with an average list size has7500patientcontactsper yearbutwill seeonly10 new cases of cancer yearly covering allcancer types. Moreover, patients who turnout to be new cancer cases often present fewandnon-specific symptoms– only about halfpresent an alarm symptom in the initial con-

sultation. So you have to investigate a lot ofpatients to find the ones with cancer.

Further complicating this picture is thatmanyof thepatientswho seekhelpmost oftenfrom their GP are unjustifiably anxious abouthaving a seriousdisease, andmaywant tohavetests donewhen there is no good reason. Thisanxiety can seriously affect their wellbeing,and their fears are often heightened ratherthan allayed by being tested. But of coursethesepatientsmay suffer fromcancer just likeanybodyelse. SoGPsaswell as thehealth caresystems are facedwith real dilemmas here.

Tina ErikssonGeneral Practitioner, Denmark, and member of the European Executive Boardof theWorld Organization of Family Doctors (WONCA)

Eric Van CutsemHead of the Digestive Oncology Unit at the University Hospital Gasthuisberg,Leuven, Belgium

C R O S S T A L K


GPs can of course make mistakes likethe one you mention and fail to investigatefurther the cause of blood in the faecesbecause they assume it must be explainedby the presence of haemorrhoids. But itmay also be that the patient is reluctant tobe tested.

I would have to recommend a colon-oscopy, which for the patient involves aday of liquid food and laxatives to empty thecolon, and then being investigated quite farup through the rectum. It is not alwayseasy to get patients to accept this unlessthey really feel there is a serious risk.

With patients whose intestines usuallyfunction like clockwork, it can be obviouswhen something is wrong. But a lot ofpeople have benign colon symptoms mostof their life, which fluctuate up and down,and that can hide more serious changes.You have to look for things like weightloss, or maybe new forms of dyspepsiathat the patient didn’t have before. Some-times watchful waiting is the only thingyou can do if a patient is not willing to gofurther and your suspicion is quite weak.You can take a blood test to check haemo-

globin levels and then ask them to moni-tor their weight and their symptoms andcome back for further testing.

The lack of specific symptoms can alsobe a problem when it comes to referringpatients for CT scans of the pancreas.Pancreatic cancer can be difficult to spotbecause there are so few specific alarmsymptoms unless the patient looks jaun-diced. There may be unspecific pain in theupper abdomen, which could be due tomany reasons, some of them quite com-mon. In Denmark, GPs cannot referpatients directly for a CT scan. We canrefer them using a new fast-track cancerdiagnosis scheme, but you would need astrong enough suspicion to use that option.

In my practice we see patients withsymptoms that could be suspicious for can-cer 10 times a week. If I were to refer all ofthem to the fast cancer track and all my col-leagues did the same, the fast cancer trackwould break down next week. Already weare seeing long waiting lists, for instance,among patients with prolapsed discs,because the fast-track cancer diagnosisprocedure is using up so much CT time.

Access to CT scanning will depend on theresources and system in each country. It’strue that in some countries access can bedifficult, but in other countries CT scansare overused. In Belgium it is not difficultto get patients referred for a CT scan,but there is still a judgement problem –people can make the call too late. It isessential to correctly evaluate the severityof the symptoms.

Regarding patients’ reluctance toundergo a colonoscopy, there is a greatproblem with the public perception ofthis examination. There is no alternative

and it needs to be demystified so thatpeople are less resistant.

Doctors need to take time to explainwhat is involved and that the patient hasa lot more to gain than they have to lose.The most difficult part is the preparation,and doctors need to explain carefullyhow this should be done, because thecolonoscopy itself will be much moredifficult if the colon has not been ade-quately cleared. If it is well done thepatient experience is not too bad. Infor-mation and communication are key, aswell as good quality control.


C R O S S T A L K

I agree that good information and commu-nication are essential, and you need to taketime to build up a good picture of the symp-toms and risk factors. This in itself can be aproblem, particularly in countries like theUKwhereGPs aremeant to spend nomorethan 10 minutes with each patient. It alsotakes time to talk to patients about whyyou think they may need further tests.When we first started referring patients tothe fast-track cancer diagnosis, it took awhile to get used to those conversations – atfirst it seemed a little harsh. Before, youwould only raise the possibility of cancerwith patients where there were strongerreasons for suspicion. Now we have to say,“We have this suspicion, which is not veryserious, but we want to make sure, so I willsend you now.”You need to adapt your com-munication, so you don’t scare people, buthelp them understand that it is a fast way tomost probably find out that there is nothingwrong. If it is a cancer, it will be found ear-lier. It is also important to make sure that,

after they have been through the fast cancertrack, whatever the outcome is, they eithercall or come back for an evaluation or chat.

IT systemsmay also have a role to play inreducing late diagnosis. Theremaybe a com-binationofdiagnosis coding, data capture andaids to diagnosis that in the quite near futuremayhelpGPs. InFinland theyareworkingona decision-making software that could helpGPs be more systematic in reaching a diag-nosis. While you are in consultation with apatient you can enter the symptoms on yourcomputer andwill get suggestions about pos-sible diagnoses and other questions or inves-tigations you can do. So, for instance, if youhave unspecific pain in the upper part of theabdomen, you could be prompted to think:pancreatic cancer. IT systems can also makeit easier to refer patients for diagnostic tests.Linkeddata systemscanprovide informationabout the local options for referrals, they canenable you tomake the referral electronically,and theuseof standardisedcriteria canensurethat only people who need referral will get it.

GPs have a crucial role in diagnosing can-cers early. If GPs feel they don’t have enoughtime, they need to raise this within theirhealth system and get changes. I under-stand it is very difficult for them. Specialistshave very in-depth knowledge, but aboutmany fewer things. GPs, in contrast, have toknow about awide range of things – not onlyall the different cancers, but everything elselike diabetes and heart disease. Good inter-action between specialists and GPs mayhelp – I spoke to 200 GPs recently on theissue of familial risk, screening and othertopics at an evening meeting and theyclearly welcomed the chance to learnmore.

Well-organised health services with ade-quate resources are also important to

ensure appropriate access to diagnosticprocedures and prompt referral – it is clearthat some countries have a poorer recordthan others on late diagnosis.

Full implementation of EU recom-mendations for quality-controlled popula-tion-based screening programmes wouldalso result in more cancers being picked upearly, and GPs have an important role toplay here too.And society also has a role toplay in raising general awareness aboutcancer: if people have a better knowledge ofsymptoms and understanding of their riskand of the importance of early detection,they will go to their GPs earlier, and bemore open to undergoing further testingwhen it is recommended.

Why does it happen? How can we do better?Late diagnosis


C R O S S T A L K

either through treatingpatients symptomaticallyorbyrelating symptoms to a health problem other thancancer…anda failure to fully or adequately examinepatients,useof inappropriateor inadequate tests, andreceivingor failing to followup inconclusivenegativeor false-negative test results” (Br J Cancer,101:S92–S101). But GPs see several hundred patients withpotential cancer symptomsevery year, ofwhomonlyahandfulwill turnout tohave thedisease.Theyoftenface pressure not to overload specialists and diag-nostic services unless there is a very strong basis forsuspicion. Is it fair to lay theblame for late diagnosison GPs? Cancer World’s Anna Wagstaff asked twoexperts – a general practitioner and a specialist ingastrointestinal cancer – to discuss the issue.

atients whose cancers are diagnosedlate are more likely to require treat-ments that aremoredebilitating,moreexpensive, andyet less likely to result ina cure.With experts in countries such

as Denmark and the UK now identifying earlierdiagnosis as key to improving their cancer outcomes– with the potential to prevent an estimated5000–10,000deaths a year in theUK– the spotlightis falling on general practitioners, and their capacitytoaccurately spot suspicious symptomsandfast-trackpatients for further investigations or specialist con-sultations. International studies have identified themost common themesassociatedwithdelayed refer-ral across cancer sites as “misdiagnosis, occurring

P

C R O S S T A L K


I agree that symptoms in cancer sometimesappear late andareoftenatypical,which iswhythey are recognisedby thepatient or thephysi-cian late. And when GPs are presented, forexample, with functional abdominal symp-toms– vague abdominal pain, a little anorexia,somedyspepsia– it canbedifficult for themtofilter those that might be related to a cancerfrom those that are not. However, sometimesthere is a problem with inadequate examina-tion. For instance, in the case of colon cancer,apatientmay report rectal bleeding; thedoctorsees the patient has haemorrhoids and con-cludes, “OK, it is probably linked to thehaem-orrhoids and it would not be appropriate toexamine the colon.” So theydecidenot to test.Or when tests are done, they don’t alwaysuse the most appropriate ones. For instance,in patients with dyspepsia and a little bit ofweight loss, theymay do anultrasound of the

pancreas; they find nothing, and they say thepancreas looks good.Butultrasounddoesnothaveperfect sensitivity–aCTscan isbetter inthis situation.With some tests and examina-tions, physicians sometimes do two or eventhree screening examinations where nothingis shown, because of the lack of sensitivity.

So the problems are not doing the appro-priate test, not going far enough with somepatients, and inotherpatientsdoinga first testwith a lower sensitivity and then taking thedecision – OK let’s wait a couple of months,and if it deteriorates wewill go further.

It’s true that there can be psychologicaldownsides to testing. But not being testedwhen you have symptoms can cause greateranxiety.And failing to rule out cancer as a pos-sible cause of symptoms could have greaterconsequences than creating anxiety.Youhaveto weigh the risks against the benefits.

Finding the needle – the single patient atcancer risk – in the haystack of patients withtemporary and benign symptoms, is at thecoreofgeneralpractice.So it is fair toholdGPsresponsible for delayed referral anddiagnosisof cancer patients. But the task is not easy. InDenmark, a GP with an average list size has7500patientcontactsper yearbutwill seeonly10 new cases of cancer yearly covering allcancer types. Moreover, patients who turnout to be new cancer cases often present fewandnon-specific symptoms– only about halfpresent an alarm symptom in the initial con-

sultation. So you have to investigate a lot ofpatients to find the ones with cancer.

Further complicating this picture is thatmanyof thepatientswho seekhelpmost oftenfrom their GP are unjustifiably anxious abouthaving a seriousdisease, andmaywant tohavetests donewhen there is no good reason. Thisanxiety can seriously affect their wellbeing,and their fears are often heightened ratherthan allayed by being tested. But of coursethesepatientsmay suffer fromcancer just likeanybodyelse. SoGPsaswell as thehealth caresystems are facedwith real dilemmas here.

Tina ErikssonGeneral Practitioner, Denmark, and member of the European Executive Boardof theWorld Organization of Family Doctors (WONCA)

Eric Van CutsemHead of the Digestive Oncology Unit at the University Hospital Gasthuisberg,Leuven, Belgium

C R O S S T A L K


GPs can of course make mistakes likethe one you mention and fail to investigatefurther the cause of blood in the faecesbecause they assume it must be explainedby the presence of haemorrhoids. But itmay also be that the patient is reluctant tobe tested.

I would have to recommend a colon-oscopy, which for the patient involves aday of liquid food and laxatives to empty thecolon, and then being investigated quite farup through the rectum. It is not alwayseasy to get patients to accept this unlessthey really feel there is a serious risk.

With patients whose intestines usuallyfunction like clockwork, it can be obviouswhen something is wrong. But a lot ofpeople have benign colon symptoms mostof their life, which fluctuate up and down,and that can hide more serious changes.You have to look for things like weightloss, or maybe new forms of dyspepsiathat the patient didn’t have before. Some-times watchful waiting is the only thingyou can do if a patient is not willing to gofurther and your suspicion is quite weak.You can take a blood test to check haemo-

globin levels and then ask them to moni-tor their weight and their symptoms andcome back for further testing.

The lack of specific symptoms can alsobe a problem when it comes to referringpatients for CT scans of the pancreas.Pancreatic cancer can be difficult to spotbecause there are so few specific alarmsymptoms unless the patient looks jaun-diced. There may be unspecific pain in theupper abdomen, which could be due tomany reasons, some of them quite com-mon. In Denmark, GPs cannot referpatients directly for a CT scan. We canrefer them using a new fast-track cancerdiagnosis scheme, but you would need astrong enough suspicion to use that option.

In my practice we see patients withsymptoms that could be suspicious for can-cer 10 times a week. If I were to refer all ofthem to the fast cancer track and all my col-leagues did the same, the fast cancer trackwould break down next week. Already weare seeing long waiting lists, for instance,among patients with prolapsed discs,because the fast-track cancer diagnosisprocedure is using up so much CT time.

Access to CT scanning will depend on theresources and system in each country. It’strue that in some countries access can bedifficult, but in other countries CT scansare overused. In Belgium it is not difficultto get patients referred for a CT scan,but there is still a judgement problem –people can make the call too late. It isessential to correctly evaluate the severityof the symptoms.

Regarding patients’ reluctance toundergo a colonoscopy, there is a greatproblem with the public perception ofthis examination. There is no alternative

and it needs to be demystified so thatpeople are less resistant.

Doctors need to take time to explainwhat is involved and that the patient hasa lot more to gain than they have to lose.The most difficult part is the preparation,and doctors need to explain carefullyhow this should be done, because thecolonoscopy itself will be much moredifficult if the colon has not been ade-quately cleared. If it is well done thepatient experience is not too bad. Infor-mation and communication are key, aswell as good quality control.


C R O S S T A L K

I agree that good information and commu-nication are essential, and you need to taketime to build up a good picture of the symp-toms and risk factors. This in itself can be aproblem, particularly in countries like theUKwhereGPs aremeant to spend nomorethan 10 minutes with each patient. It alsotakes time to talk to patients about whyyou think they may need further tests.When we first started referring patients tothe fast-track cancer diagnosis, it took awhile to get used to those conversations – atfirst it seemed a little harsh. Before, youwould only raise the possibility of cancerwith patients where there were strongerreasons for suspicion. Now we have to say,“We have this suspicion, which is not veryserious, but we want to make sure, so I willsend you now.”You need to adapt your com-munication, so you don’t scare people, buthelp them understand that it is a fast way tomost probably find out that there is nothingwrong. If it is a cancer, it will be found ear-lier. It is also important to make sure that,

after they have been through the fast cancertrack, whatever the outcome is, they eithercall or come back for an evaluation or chat.

IT systemsmay also have a role to play inreducing late diagnosis. Theremaybe a com-binationofdiagnosis coding, data capture andaids to diagnosis that in the quite near futuremayhelpGPs. InFinland theyareworkingona decision-making software that could helpGPs be more systematic in reaching a diag-nosis. While you are in consultation with apatient you can enter the symptoms on yourcomputer andwill get suggestions about pos-sible diagnoses and other questions or inves-tigations you can do. So, for instance, if youhave unspecific pain in the upper part of theabdomen, you could be prompted to think:pancreatic cancer. IT systems can also makeit easier to refer patients for diagnostic tests.Linkeddata systemscanprovide informationabout the local options for referrals, they canenable you tomake the referral electronically,and theuseof standardisedcriteria canensurethat only people who need referral will get it.

GPs have a crucial role in diagnosing can-cers early. If GPs feel they don’t have enoughtime, they need to raise this within theirhealth system and get changes. I under-stand it is very difficult for them. Specialistshave very in-depth knowledge, but aboutmany fewer things. GPs, in contrast, have toknow about awide range of things – not onlyall the different cancers, but everything elselike diabetes and heart disease. Good inter-action between specialists and GPs mayhelp – I spoke to 200 GPs recently on theissue of familial risk, screening and othertopics at an evening meeting and theyclearly welcomed the chance to learnmore.

Well-organised health services with ade-quate resources are also important to

ensure appropriate access to diagnosticprocedures and prompt referral – it is clearthat some countries have a poorer recordthan others on late diagnosis.

Full implementation of EU recom-mendations for quality-controlled popula-tion-based screening programmes wouldalso result in more cancers being picked upearly, and GPs have an important role toplay here too.And society also has a role toplay in raising general awareness aboutcancer: if people have a better knowledge ofsymptoms and understanding of their riskand of the importance of early detection,they will go to their GPs earlier, and bemore open to undergoing further testingwhen it is recommended.

ancer scientists have a newpatron saint: Charles Dar-win. Research is showingthat the only way to curecancer in its many forms

will be tounderstand theDarwinian evo-lution that drives thediseasewithin eachpatient, as natural selection works ongenetic mutations within tumour cells.

Knowledge of cancer is advancingmore rapidly than any other field ofmedical science, because cancer is adisorder of DNA – and now, with thecoming of new technology to read themutations in individual tumours,researchers can find out for the firsttime what is really going on.

The outcome, say researchers, will be anewgeneration of tests andpersonalisedtreatments aimed at the particularchanges taking place in each tumour.Many cancers will be treated in a simi-lar way to infections by a fast-evolvingvirus such as HIV, where drug combi-nations can keep the emergence ofresistance in check.

“Today, despite the best efforts ofthe drug companies, we still usemedieval methods to attack most can-cers: cutting [surgery], burning [radio-therapy] and poisoning [non-specificchemotherapy],” says Alan Ashworth,head of the Institute ofCancerResearchin London. “Once we understand the

Darwinianprocess bywhich cancer cellsevolve in the body, we should be able tocontrol even advanced disease throughcombinations of specific drugs.”

Peter Johnson, chief clinician atCan-cer ResearchUK, the specialist charity,adds: “We aremoving into an era wherewewill have the tools to back cancer intoan evolutionary dead-end.”

Cancer starts when a single cellundergoes a mutation that takes thebiological brakes off and lets it divideout of control. The trigger for thatmay be damage to DNA caused by acarcinogen such as ultraviolet lightor cigarette smoke, it may result froman inherited genetic weakness, or it

Cancer genetics is intriguing, exciting and offers hope for real

progress in treatment, but try telling that to non-scientists.

Clive Cookson, science editor at the Financial Times, did just

that. His article, explaining the evolutionary process that

drives cancer, won him a Best Cancer Reporter Award.

B E S T R E P O R T E R

C L I V E CO O K S O N


The origin of aspecial success

C

Clear, accurate andstimulating stories likethis one help sustainthe public’s trust andbelief in scientific andmedical research

may just be random bad luck.Given that the average person con-

tainsmore than10 trillion cells – and theDNAcopying process is far fromperfect– it is amazing how rarely cancer getsgoing.Butonceitdoes, thedisease throwsmany rules of human biology out of thewindow. Inparticular, cancer cellsmutatefar more rapidly than healthy ones.

Although most of these randommutations are harmful and kill the cells,they occur so frequently that occasion-ally the processwill help the cancer pro-liferate, for example by building the

blood supply that a solid tumour needsto grow and, most importantly, bybecoming resistant to drugs prescribed totreat the disease.

“Cancer is a perversion of what thebody originally produces – and the lessa tumour looks like the tissue thatoriginally gave rise to it, the worse it isfor the patient,” says Prof Ashworth.“Cancers typically contain 10,000 to100,000 genetic changes compared tonormal cells, of which perhaps 10 areabsolutely critical for the tumour’sgrowth and survival.

“It is important to realise that cancerhas no ‘purpose’ – no drive to grow andspread,” he adds. “It is blind evolution,based on random genetic changes andnatural selection.”

Cancer genomics – a new disciplinebasedonreadingall3billionchemical ‘let-ters’ ofDNA in tumour cells – is openingresearchers’eyes to theastonishing rateofevolution. “The game-changer when itcomes to our understanding cancer isthe new generation of DNA sequencingmachines,” says Charles Swanton, pro-fessor of personalisedcancermedicine atthe University College London CancerInstitute. “It is likemoving from a black-and-white televisionwith four pixels to acolour TV with thousands of pixels.”Instruments to read the sequenceof four‘letters’that encodeDNA(abbreviatedasG,A, T andC) are doubling in perform-ance per dollar spent every fewmonths.This year, it will become possible to reada whole human genome in a day for lessthan$1,000.Eventually costswill fall farenough to make cancer genomics anaffordable tool not just for researchbut inroutinemedical practice, to monitor themutations takingplace inpatients as theyare treated.

Although oncologists have longrecognised that cancer is a genetic dis-ease, the conventional view is that itprogresses in a linear fashion, with alltumour cells genetically the same at anygiven point. But very recent researchhas knocked that idea on its head andadded another layer of complexity todiagnosis and treatment.

A genomic study of advanced kidneycancer, carried out by Cancer ResearchUK and published this month [March2012] in the New England Journal of


“We are moving into an era where we will have the

tools to back cancer into an evolutionary dead-end”



CANCER’S DEADLY TREE OF LIFE Medicine, shows what Prof Swanton,its lead author, calls “an extraordinaryamount of genetic diversity”within indi-vidual tumours. It confirms the findingsof smaller genetic studies of leukaemia,brain and breast cancers. The kidneycancer patient examined most closelyhad 118 different mutations – of which40were present in all biopsies, 53 werepresent inmost but not all samples and25 were detected in just one. Byanalysing the location of mutations, theresearchers traced the origins of particu-lar sub-types of cancer cells back in time– creating a map of how the pattern offaultsmighthaveevolved. “Weare seeingbranchedevolution that is very similar tothe ‘tree of life’ diagramCharles Darwinproduced in 1837 to postulate specia-tion [theprocessbywhichnewbiological

“We are seeing branched evolution very similar to the

‘tree of life’ diagram Charles Darwin produced in 1837”

A patient with advancedkidney cancer had 118genetically distinct cancer‘species’. This diversity intumours echoes CharlesDarwin’s depiction ofthe way speciesthemselves evolve.

The first knownsketch by Darwin of anevolutionary tree (1837)

Cancer can evolve and resist treatments that targetonly one mutation

The generation of DNA sequenceBillions of letters of genetic code

per day per machine

SYNDICSOFCAMBRIDGEUNIVERSITYLIBRARY(575(15)),FINANCIALTIMES

CANCERRESEARCHUK/WELLCOMETRUST

1. Cancer

starts with one

cell mutating

2. This then develops

into a Darwinian

ecosystem of

genetic diversity


species arise],” says Prof Swanton.On the faceof it, theheterogeneity of

cancerwithina singlepatient is badnewsfordiagnosis and treatment. It shows thata single biopsy may not reveal what isgoing onwith someoneandexplainswhymany cancers are so hard to eradicateespecially once they have spreadbeyondtheir original organ to remote ‘metastatic’sites around the body, where differentconditions drive their evolution.

But on the basis that knowledge ispower, researchers can see a way for-ward. “This underscores the importanceof targeting commonmutations found inthe ‘trunk’ of the evolutionary tree, asopposed to those found in the ‘branches’whichmay only be present in a relativelysmallnumberofcells,” saysProfSwanton.

Scientists will need to pin down the‘drivermutations’present in the ‘trunk’ofeach cancer’s Darwinian tree – andmarry themup to thenew targeteddrugsemerging from clinical trials worldwide.Effective newdrugs are coming on to themarket. The latest is vemurafenib,marketed by Roche of Switzerland asZelboraf. It targets amutation in a genecalled B-Raf, which is present in morethan half of malignant melanomas andtells the cancer cells to keep on growingbeyond the point that healthy onesshould stop.

Many patients who have this muta-tion respond spectacularly well tovemurafenib; even advancedmetastaticmelanomamaymelt away. But unfortu-nately a few vemurafenib-resistant cellsare almost always left behind, and theyseed a resurgence of the cancer, whichthis time is untreatable. Clinical trialsshow that a patientwill typically gain fourto eight months of life – which may be

beyond price but is far from a cure.“Drug combinationsmust be theway

forward now that we know cancers areevolving in the body, just like viruses,”saysMike Stratton, director of theWell-come Trust Sanger Institute, a genomeresearch centre near Cambridge, andleader of its cancer genomeproject. “Weshould be able to control evenmetasta-tic cancer if we can find the right com-binations of drugs, just aswe can controlHIV.” Cancer cells, like viruses, wouldfind it far harder to develop resistance todrug combinations than to single drugs,because this would require multiplemutations of the right type to take placeat the same time. So academic andindustrial research teams are steppinguptheir development of targeted anti-cancer cocktails.

For instance, GlaxoSmithKline iscarrying out clinical studies of two exper-imental drugs that are active againstadvanced melanoma. One, calleddabrafenib, targets the B-Raf gene likeRoche’s vemurafenib; the other, trame-tinib, aims at a different cell signallinggene calledMEK. The company is see-inghowwell the twodrugswork togetherand in combinationwith other promisinganti-cancer agents including one devel-oped by a Swiss competitor, Novartis.“The science is evolving at the speed oflight, which is fantastic for the patient,”says Paolo Paoletti, president of GSKOncology, a unit of theUK-based drug-maker. “But collaboration betweenpharmaceutical and biotechnologycompanies will be essential.” Such col-laboration will extend beyond drugdevelopment to diagnostic tests that telldoctorswhich treatments aremost likelyto benefit individual patients, based on

the genetic profile of their disease.Though the cost of full genome sequenc-ing is falling fast, it will be several yearsbefore it is cheap enough to apply rou-tinely to all cancer cases.

Meanwhile, a plethora of biotechcompanies are developingmore limitedtests based on ‘biomarkers’ – genes orproteins that predict a patient’s responseto particular drugs. One such companyisOxfordCancer Biomarkers, just spunout ofOxfordUniversity. “Our strategy isto take the 30 most important cancerdrugs and develop biomarkers for eachone,” saysNickLaThangue, professor ofcancer biology atOxford. “The clinicianwould take a biopsy every six months orso – and use the biomarkers to capturegenetic changes before the tumourbecomes resistant to the therapy.”

So how excited should we be aboutthe prospects of using the latest geneticbreakthroughs to cure cancer? The sci-entist who has beenworking longest onDNA is bursting with enthusiasm.James Watson, who discovered DNAwith Francis Crick in 1953 and is stillactive in genetics research at ColdSpring Harbor Laboratory in the US,says: “We havemade immense progressrecently but no one in the cancer com-munity wants to be seen jumping upand down with excitement, becauseresearchers have been over-optimisticin the past.”

But Dr Watson adds emphatically:“New science of the past yearmakesmeoptimistic that the back of most incur-able human cancersmay be broken overthe next five to 10 years.”

This article was first published in the FinancialTimes on 20 March 2012, and is reprinted herewith permission. © Clive Cookson


“The science is evolving at the speed of light,

which is fantastic for the patient”

he fragmentation of academicresearch is a major organisa-tional problem in Europe.

There is also a clear need to improvetranslational research and to overcomedifferences experienced by patientsacross Europe in access to diagnosticsand treatment, which feed into dispar-ities in therapeutic outcomes.To address these issues, the Organ-

isation of European Cancer Institutes(OECI) – a body dedicated to improvingresearch collaboration among Europe’scancer centres – identified the need todevelop a system for monitoring theresearch and care offered to patients,and also to find ways to harmonisepatient care and share the knowledgethat we are developing to improve thestandard of care. To this end, theOECIdeveloped a tool for assessing andbenchmarking the care and researchbeing carried out in cancer centres.The challenge for cancer centres

is to meet all the requirements of the

e - G R A N D R O U N D


T

OECI accreditation: is yoursa top-class cancer centre?

The European School of Oncology pres-ents weekly e-grandrounds which offerparticipants thechance todiscussa rangeof cutting-edge issues with leading Euro-pean experts. One of these is selected forpublication in each issue of Cancer World.In this issue, Mahasti Saghatchian, ofthe Institut Gustave Roussy, in Villejuif,France, reviews the OECI programme foraccrediting cancer centres, which shehelped develop and now chairs. GordonMcVie, from the European Institute ofOncology in Milan, Italy, poses questionsarising during the live presentation.It is summarised by Susan Mayor.

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

European School of Oncologye-grandround

Whether you work at a cancer unit, a cancer centre or a comprehensive cancer centre,

applying for OECI accreditation is a good way to find out how your institute compares

with the best, and get advice on how to address any shortcomings.


different stakeholders (see figure). Thekey stakeholder is the patient, and acentral aim for any cancer centre is toprovide patients with the best care.The challenge for the OECI was, firstand foremost, to develop a tool thatwould allow a cancer centre to ensureit was achieving this aim. Cancer cen-tres are also answerable to healthauthorities, and we wanted to developa tool that would help centres to collectan evidence base to prove they are pro-viding good-quality care.Wealsowanted the tool to help build

the trust of cancer professionals andother organisations in the care andresearchbeing performedby cancer cen-tres. Building the evidence base on theperformanceof each cancer centre couldstrengthen this trust. Finally, funders,including industry and thehealth author-ities, want data on activity, outcomesand production in terms of activity. Wewanted our accreditation system to beable to help cancer centres provide this.

A supportive approachMost quality assessment programmesare regulatory measures imposed byan external authority, and are usuallycompulsory. We developed the OECIquality assessment programme as asupportive measure for cancer cen-tres, and it is voluntary. It has beendesigned and developed internally bypeople from cancer centres. Peerreview is performed by people fromcancer centres and visits are chaired bythe director of another cancer centre.The standards that we have set havebeen developed and are revised everyfour years by people from the centres.

Gordon McVie [GM]: Youwork at the Gustave Roussy,one of the most prestigiouscancer centres inEurope.Whydoes your centre want to getOECI accredition?Mahasti Saghatchian [MS]:We like to think we are one ofthe best, but whenwe enteredthe programme we still feltanxious about people visitingus.Weprepared for the accred-itation and visit very seriouslyand found it a very rewardingexperience.Having a group ofauditors visit us and assesswhat we do from an externalpoint of viewwas reassuring. Itis not enough to claim that you are a com-prehensive cancer centre, you need toprove it to your peers and collect the nec-essary evidence.Gordon McVie [GM]: It’s a little bitlike submitting research to a journal.As anauthor you think your research paper isgreat, but we need other people to lookindependently and assess its value.MS: Different countries have different

assessment and accreditation systems inplace. In France we have mandatoryaccreditation for hospitals in general, butnot specifically for cancer centres.Here atthe Institut Gustave Roussy, we are goingto be the only cancer centre accredited bythe OECI.GM:The process is similar to that whichtheUS cancer centres have gone through,and provides a great deal of credibility.


THE CANCER CENTRE CHALLENGE

Cancer centres have to satisfy the requirements of manystakeholders, primary among them being patients

Trust us: we’re accredited. Wim van Harten,director of the NKI cancer centre in Amsterdam,

shows his OECI certificate, flanked by formerOECI president Marco Pierotti (left) and MahastiSaghatchian, chair of the OECI Accreditation and

Designation committee (right)

Cancer Professionalsand Organisations

Funders(Industry)

Patients Health AuthoritiesBest Care

Trust

Liabilit

y

Data

CancerCenter

The OECI toolsThe main tools that theOECI has developed are theaccreditation and designa-tion programme and theEurocanPlatform Networkof excellence, both of whichare supported by the Euro-peanCommission and focusmore on centres dedicatedto translational research.In the accreditation and

designation programme, wehave developed a tool similarto other accreditation pro-grammes in its design, witha set of quantitative andqualitative standards inte-grated in a database, a report and apeer review system. The programmeoperates at three levels: the compre-hensive cancer centre, the cancer cen-tre, and the cancer unit.A pilot amongOECI centres showed these are thethree most frequent types of centres.We developed specific criteria for eachof these. The programme includes stan-dards and a database with a set of qual-itative and quantitative data. We writea report and an improvement plan foreach centre, which are reviewed by theaccreditation board and then discussedwith the centre.

Developing the programmeThe accreditation programme has beenrunning now for 10 years. The ideawas first discussed in 2002. Then in2005 the project was launched as anaccreditation programme setting stan-dards for cancer centres. Subsequentlywe realised that this was not sufficientand we would need different assess-ments for different types of centres, dif-ferentiating between larger or smallercentres for care or research and com-prehensive cancer centres. We thendeveloped the criteria that allowedus to have a designation system for


the cancer centres, and launched themerged accreditation and designa-tion programme in 2008.We are at the end of the first version

of the accreditation and designationprogramme and will be launching thesecond version of the programme in2013. The standards are currentlybeing revised and we are reviewing thequantitative data that we ask centres toprovide.We are also working on a spe-cific programme for prostate cancercentres in collaboration with the Euro-pean School of Oncology, which hasdeveloped guidelines for prostate units.TheOECI is now turning these guide-lines into an accreditation system forprostate cancer units.The OECI accreditation and desig-

nation programme is run by a board ofpeople from cancer centres and has amanagement unit of people paid by theOECI to coordinate and run the pro-gramme (see figure above). There is alsoanaccreditationcommittee,with10peo-ple fromvarious cancer centres,which ischaired by Chris Harrison from theChristie Cancer Centre inManchester,UK. This committee reviews all thereports provided by the auditors andmakes recommendations to the board.

The programmeuses a set ofquantitative and qualitativequestionnaires, which arecurrently being revised.These are integrated into anelectronic tool available witha username and passwordand there is also amanual inpaper format.When a centreparticipates in the pro-gramme, all answers to thequestionnaires are enteredinto the database. A groupof auditors, all from OECIcancer centres, carry outpeer review visits.

Participating centresCentres from all over Europe are takingpart in theOECI programme. The firstcentres to enter the programmewere inPortugal. Their healthministry asked allcancer centres to participate in an inter-national accreditation programme and,after discussing whether to follow theUS or European accreditation pro-grammes, three centres opted for theOECI programme. Comprehensivecancer centres, smaller cancer unitsand university centres have all takenpart in the OECI programme, andexperience has shown it is applicable inall of these types of centres.Of note, the Italian Ministry of

Health has decided to fund all com-prehensive cancer centres in Italy to gothrough the OECI accreditation pro-gramme. The nine Italian centres willgo through the programme over twoyears. We think this is a good exampleof a country deciding to go for a Euro-pean accreditation programme, andItaly is pioneering this.GM: How are different types of cancercentres defined? I know about compre-hensive cancer centres because I ambased at one in Milan. The EuropeanInstitute of Oncology has a hospital anda large laboratory and patients are seen by


OECI ACCREDITATION AND DESIGNATION 2012

� Board - Management Unit - Accreditation Committee

� Qualitative and quantitative questionnaires -Designation criteria

� Electronic platform (e-tool) and website - Newsletter

� OECI Auditors

� For each cancer centre in the programme

1. Quantitative and qualitative data

2. Report

3. Improvement plan


multidisciplinary teams, sowe are a comprehensive can-cer centre. The Christie can-cer centre, in Manchester,and the NKI-AVL, in Ams-terdam, have similar facili-ties. Is the IPO [InstitutoPortugués de Oncologia] inLisbon a comprehensive can-cer centre or a cancer centre?MS: It is a cancer centre, butnot a comprehensive centre,because it does not haveenough research activities.GM: Then what is a cancerunit? Have you looked at thecancer unit at the UniversityHospital of Helsinki?MS:None of the centres currently in theprogramme is a cancer unit. We arecurrently in a one-year pending periodfor Helsinki. They applied as a com-prehensive cancer centre, but it appearsthat they are a cancer department in alarge hospital. Although they have alarge amount of research and theirresearch is of high quality, we consideredthat their research and care was not suf-ficiently integrated, and that their man-agement and strategy in oncology werenot clearly separated from the rest ofthe work of the university hospital. Weconsidered that Helsinki could not bedesignated as a comprehensive cancercentre because they did not have theorganisation, strategy or logistics of acomprehensive cancer centre.We offeredthem a one-year pending period to reor-ganise and develop the processes neededto achieve these.GM: This strikes me as a very importantexample of the last part of thewhole accred-itation programme – the improvementplan.Cancer centres really appreciate thatthese accreditation visits are not just aboutchecking that everything is OK or beingcritical, but are about saying, ‘This is howyou are now, what are you going to do tomake things even better?’

Focus of the programmeThe accreditation programme focuseson what is critical to cancer care. Thisincludes:� planning and organisation of inte-grated care

� multidisciplinary teams and care� integration and translation of clin-ical and basic research into care

� education for professionals� the experience and involvement ofpatients

� monitoring and organisation ofquality improvement.

The accreditation processApplication anddesignation screeningWhen a centre applies to the pro-gramme, the first step is to definewhether you are applying as a cancerunit, a centre or a comprehensive can-cer centre. Comprehensive centreshave to provide data related to research,while cancer units and centres do not.Differences are otherwise based onvolume of infrastructure, budget andactivities. All centres then complete aself-evaluation period providing therequired information, and the OECImakes a ‘go or no-go’ decision, advising

the board on whether a cen-tre is ready for peer reviewbased on the informationprovided. This processavoids organising a peerreview in a centre that isnot ready.

Explanatory visitAn explanatory visit takesplace immediately after theapplication and ensures thateveryone in the centre isaware of the process, whatwill happen, and the infor-mation required. TheOECIhelps centres develop anaction plan. We have found

that centres have to dedicate one full-timeperson for a six-monthperiod to col-lect the required documents and data.

Self-assessment periodIn the self-assessment period, centresare required to collect documents,annual reports and data on quality,education, comprehensiveness andpatients. It usually takes at least sixmonths.We have found that gatheringdata on research activities is particu-larly difficult.

Electronic self-assessment toolThe e-tool includes a quantitative andthe qualitative questionnaire. Quali-tative questions are scored to assess thecentre’s degree of compliance to thestandards, which are as detailed aspossible in each question. Centres areasked to send reports or documents tosupport their answers. These can beuploaded onto the system. The toolautomatically shows the centres anycriteria where they are not fully com-pliant with the required standards, andhelps to build improvement plans.Quantitative questions ask for numbersrelating to infrastructure, resourcesetc. Once all the figures have been


CENTRES IN THE OECI PROGRAMME

submitted, the auditors review all thedata, and then prepare for their visit.

Peer review visitTrained auditors spend two full days onsite, during which time the centre’sactivities should continue as normal.They can visit anywhere in the centreand carry out interviews with staff oreven patients. Based on the self-evaluation reports and what they haveseen, the audit group gives a first sum-mary report about the accreditation,the designation, and points that can beimproved. The chairman of the auditgroup is always a director of an OECIcentre; two auditors are specialised ina relevant field of oncology, one is aquality manager and other personnelare from the OECI.

Peer review reportThe peer review report is developed fol-lowing some exchanges with the cen-tre. The final report highlights thecentre’s strengths and the opportunitiesfor improvement.When the report hasbeen agreed on, a decision is made onaccreditation and designation. Ifaccreditation is given, it is valid forfour years.Thecost for the full process of assess-

ment and designation is €30,000 for acomprehensive cancer centre, which ispaid to the OECI. The cost is lower forcancer centres and cancer units, at€25,000 and €20,000, respectively.

EurocanPlatformresearch programmeThe EurocanPlatform programme is aresearch programme aimed at devel-oping a designation system for excel-lent comprehensive cancer centres andcomprehensive research centres, andoutcome indicators for translationalresearch. Proof of excellence includesrecent research reports, grantsawarded, and ongoing or completed


research programmes and their out-puts. The programme is also develop-ing excellence indicators that will helpus measure impact on patients.

Summing upWe hope that the OECI accredita-tion programme and EurocanPlatformwill help us to reduce fragmentationand increase harmonisation acrosscancer centres. The aim is for centresto work more closely together, withpeople visiting each other’s centresand learning from each other in a net-work that supports competition andcollaboration.GM: How do you accredit a paediatriconcology centre? Would a paediatriconcology centre have to be accredited aspart of an adult oncology network or areyou anticipating that there could bestand-alone paediatric oncology centresaccredited in the future by OECI?

MS: Paediatric oncology is facing thesame issues as adult oncology.We are col-laborating closely on this with a net-work of European paediatric oncologypersonnel. I think we could apply theaccreditation model to paediatric units,but we might have to adapt some ques-tions or add further standards related topaediatric care.GM:What does a comprehensive cancercentre do for a patient in a district gen-eral hospital perhaps 100 kilometresfrom the centre?MS: Comprehensive cancer centres arevery important for other hospitals. Notall patients can be managed within thewalls of the cancer centre. One of thevery important standards is that centresmust have networks, where they workwith other professionals taking care ofpatients. It is the duty of the cancercentre to organise this network and theeducation of professionals within it.


The auditors. An OECI team visits the Instituto Portgués de Oncologia in Lisbon, as part of theaccreditation procedure, March 2011

network of cancer centres isplanning to transform the waythat the findings of molecular

research into non-small-cell lung cancer(NSCLC) are brought into clinical use.The16centresarepoolingclinicaldataon2400 lung cancer patients and sharinginformation on the genetic structure oftumour samples obtainedduring surgery.

The Lungscape project, part of theEuropean Thoracic Oncology Platform,aims toharmonise standardsand improvethe quality of genetic testing in cancercentres, and increase understandingabout which patients may benefit fromthe latest targeted therapies. Of the 16centres, 14 are based in nine Europeancountries, with one each in China andtheUSA.

Lungscape will eventually helponcologists select patients who stand to

benefit from innovative or experimentaltreatments. The aim is to produce effec-tive trials on subsets of maybe 50patients, instead of having to recruithundreds of patients into large ran-domised controlled trials from whichmost derive no benefit.

NSCLC accounts for about 85% oflung cancers and includes adenocarci-nomas and squamous cell carcinomas.Survival rates are lowunless thedisease iscaught early –which it isn’t in themajor-ity of cases. Patients are currently gradedaccording to tumour size (T), thenumberof nodes affected (N) and the degree ofmetastasis (M). However, research isincreasingly suggesting that this TNMcategorisation is insufficient, and thefocusnow is ondefiningnumerous smallsubgroups of lung cancer patients basedon the increasingnumberof genetic alter-

ations reported to be driving the disease.EGFR mutations are present in thetumour cells of about15%of lungcancerpatients with adenocarcinoma, and theymaybenefit fromEGFRinhibitors.Trialshave shownpositive responses fromtreat-mentwithXalkori (crizotinib) for the24%ofNSCLCpatientswhose tumours showa gene fusion between EML4 andALK.Up to 90% of these tumours showed aresponse in clinical trials, and in some

The potential for personalising lung cancer therapies is expanding

rapidly. The challenge now is how to turn that potential into

reality as fast as possible so patients can reap the benefits.

S P O T L I G H T O N

PETE R Mc I NT YR E

A


Lungscape: a livinglung laboratory

cases there has still been no disease pro-gression after 15months.

In addition to these, genetic researchis identifying many new mutations thatneed to be investigated in customised

clinical trials that can be rapidly set upwith small numbers of well-targetedpatients. It is this that Rolf Stahel, pres-ident of theETOPFoundationCouncil,hopes thatLungscapewill help todeliver.

Lungscape centres keep control of thebiopsy specimens taken from patients,which remain locally stored andanalysed. Each patient’s biological infor-mation, together with their clinical data,will be anonymised and shared in whatthey are calling the Lungscape iBiobank– a virtual biobank. During September2012, Lungscape hit its target ofenrolling 2400 patients.

ETOP’s Rolf Stahel, who is a pro-fessor and medical oncologist at theUniversity Hospital, Zurich, says

Lungscape will describethe molecular land-scape of non-small-cell lung cancer inEurope by testingthe tissue samples

with molecular mark-ers and defining their

characteristics. “We will getthere by coordinating theworkof 16 different sites, which

will also allowus to establish a net-work for clinical trials where alterationsthat have been identified can be putinto early-phase trials. We will knowwhat proportion of patients will benefitandwewill have established high-qual-ity molecular testing. In addition, wewill be able to determinewhether certainmolecular changes are associated withdifferent prognostic outcomes inde-pendent of the anatomical staging.”

As a first step, they will focus onpatients found to have the ALK fusiongene and carry out a retrospective analy-sis to seewhether the course of their dis-ease and general outcome differed fromthosewithout the fusion.

Solange Peters, medical oncologistat the Lausanne Cancer Centre in M

AP:JASO

NCOOK



“Lungscape will describe the molecular landscape

of non-small-cell lung cancer in Europe”

Switzerland, is responsible for the con-tent of and access to the virtualbiobank. “These are all surgical patientswho have undergone radical surgeryforNSCLC stages I–III andwe have allthe basic demographics, tumour patho-logic characteristics and patient clinicalfollow-up annotated for these patients.We will be able to identify in this sub-set of patients several parameters: out-come in terms of recurrence of thedisease, outcomes in terms of generalprognosis as well as the contributionrelated to other parameters. In abouthalf of them we will also have dataabout recurrence.”

It is unlikely that the patients willbenefit directly from this first phase ofLungscape – aggregating and analysinganonymised data – although they maybenefit from improved genetic testingbytheir own teams.

Peters says, “We have a philosophyof wanting to leave the tissue in everycentre as, on the one hand people donot like to send their tissue materialout, and on the other hand we want toempower all the centres with thecapability to do all the testing in-house, with full quality assurance.We want all the tests to be done in-house and no tissue to travel. That iswhat makes it a virtual biobank.”

More than two years have passedsince Stahel and colleagues devel-oped the ibiobank idea at a transla-tional research meeting in Lugano inMay 2010. Data collection began inApril 2011, but it then took a further18 months to complete the registra-tion of all 16 centres and recruit the2400 patients.

Difficult terrainMoneyTherehavebeen threemainhurdles, thefirst being money. So far Lungscape hasbeen supported entirely by the industry,with an unrestricted grant from Rocheand financial sponsorship from Pfizer,which makes Xalkori, to test the sam-ples forprevalenceofALKgene fusion, aswell as support fromAbbott Molecular.Xalkori has been approved for use by theUSdrug regulators, theFDA, forpatientswithadvancednon-small-cell cancerwhoareALK positive. In Europe, it has beenrecommended for conditional approval,witha finaldecision fromtheEMApend-ing.Stahel says theyarenowpreparinganapplication for grantmoney fromtheEU.“Wefirstneeded to showwhatwecando,and we are proving that now, so I amquite optimistic that in the futurewewillbe able to get other financial resources.”

Local lawsThe second task was to ensure that cen-tres compliedwith their national legisla-tion, hadethical approval andmet all thelogistical hurdles. ETOP is run from thecoordinatingcentre inBern,Switzerland,that coordinates the InternationalBreastCancerStudyGroup(IBCSG), andsomekey staff play the same role for bothgroups.AnitaHiltbrunner, director of thecoordinating centre, explained how thishad helpedETOP andLungscape.

“We have a regulatory office who arevery experienced from the IBCSG trialswith all the needs and regulations in allthe countries.Weact as thepoint of con-tact for everything.Wework very closelytogetherbye-mail or teleconferences andwe have regularmeetings. Communica-

tionwith all the centres and investigatorsis the key. It is something you have to beon top of every day.”

The average time for a centre to beapprovedwasabout sixmonths,but sometookmore thanayear to reachagreementwith their local or national ethics com-mittees. The main issue was informedconsent frompatients.Althoughpatientsthemselves donotneed toundergo tests,the tumourmaterial removedduring theiroperationmay in future undergo genetictests that have not even been heard oftoday. Someof these testsmaybecarriedout after a patient has died.While a sur-geon inZurichwas able to get patients tosign a single form consenting to futureresearchon thismaterial, centres inothercountries had to bemore specific.

TheUKturnedout tobe themostdif-ficult country toconvince; centreshave toget both local and national approval.While two UK centres (the Royal Infir-mary inAberdeen and the Lung CancerGroup in Manchester) have joined thenetwork, oneothercentrecouldnot reachagreement.TheShanghaiChestClinic inChina is one of only two centres outsideEurope to have been included; theirnational regulations make it clear thattherecanbenoexceptionsever to the ruleabout tumourmaterial not travelling.

QualityThethirdhurdlewas thequalityof testing.Rosita Kammler, who coordinates trans-lational research at IBCSG/ETOP, saysthat centres have to achieve the samehigh standards for immunohistochem-istry and for theALK-FISH(fluorescencein situ hybridisation) test, so that resultsare comparable.



“Communication with all the centres and investigators

is the key. You have to be on top of it every day”

“Thereare several steps involved toensurethat theendresult is comparable fromonesite to the next. The first part is an inter-nal validation.A pathologist constructedtissuemicroarrays (TMAs) fromspecificidentified samples. These were sent toeach centre with the TMAmap to workup the staining.Thenext step is ablindedexternal quality assurance round. Theyreceive blinded samples and send inresults that are then reviewed.”

Following the review, some centreswere asked to undergo further practiceand testingbefore approval.Asnew testsare developed they need to keep up tospeed and Stahel says that they may getpathologists fromeachcentre together tocarry out training with a super specialistand then do another external evaluation.“One of the clear things in our mindswas to raise quality standards and toempower all the participants to be verystrong in this field.”

The future shape of researchLungscapehas started to raise its profile,making presentations at the EuropeanLung Cancer Conference in Geneva inApril 2012 and oral and poster presenta-tions at theESMOmeeting inVienna inSeptember/October 2012. Stahel saysthat the project is generating a real senseof excitement amongst themultidiscipli-nary teams that now lead theway in lungcancer diagnosis and treatment.

Genetic testing is becoming increas-ingly important to determine best treat-ment for non-small-cell lung cancer.The French National Cancer Institute(INCa) and French Government havebeen funding molecular testing for alllung cancer patients since 2011 andfind that genetic testing can savemoney.

Jean Charles Soria, profes-sor of oncology at SouthParisUniversity, told the 3rdEuro-peanLungCancerCongress,inApril 2012, that theFrenchgovernment had invested€1.7million in testing for theEGFR mutation and saved€65 million in treatmentcosts by identifying 15,000patients whowould not ben-efit from gefitinib (Iressa).

Other centres will joinclinical trials that spinoff fromLungscape. “Theway to go tothe future is anetworkofnet-works,” saysStahel. “Aswell asour big European network,youwouldhavesub-networks.In Switzerland, for example,Zurich and Basel are part ofLungscape, but the clinicaltrial could also include Lau-sanne,Geneva andBern. It isnetworking of networks thatwill allow future research.”

Peters says, “We have akind of regrettable traditionfor unselective patient trialswhich endupbeingnot beneficial for thepatient or are negative. Building thisnetwork is a way to develop proceduresto build new trials designed for subsetsof non-small-cell lung cancer patientsand not a whole crowd. That way, newtrials emerge fromtheknowledge, andnotonly the other way around.”

ETOP is already planning phase 2of Lungscape when it will switch fromretrospective analysis to generatingprospective trials of about 1000patients who will be followed from thepoint of diagnosis.Although the central

biobank will remain anonymised, eachcentre will use the genetic testing todetermine treatment.

That is when it becomes really excit-ing, as Lungscape will follow long-termoutcomes of patientswho have receivedpersonalised treatment. Stahel said,“Two years fromnowwewill have a lot ofbiomarker datawith clinical correlationsand we will have begun a prospectiveLungscape. Five years fromnowwewillhave demonstrated which of the bio-markers adds to the diagnosis in additionto the anatomical staging.”



“Five years from now we will have demonstrated

which of the biomarkers adds to the diagnosis”

CENTRES IN THE LUNGSCAPE NETWORK

There are currently 16 centres in the Lungscape net-work. The majority are in Europe, but the networkalso includes one specialist centre in the US andone in China, to widen the research base and toallow data to be compared inside and outsideEurope.University Hospital Leuven, BelgiumUniversity Hospital Aarhus, DenmarkSt James Hospital Dublin, IrelandOspedale Clinicizzato Chieti, ItalyNetherlands Cancer Institute Amsterdam,

the NetherlandsFree University Medical Centre Amsterdam,

the NetherlandsUniversity Medical Centre Maastricht,

the NetherlandsMedical University Gdansk, PolandVall d’Hebron University Hospital Barcelona, SpainUniversity Hospital Valencia, SpainUniversity Hospital Basel, SwitzerlandUniversity Hospital Zürich, SwitzerlandRoyal Infirmary Aberdeen, UKLung Cancer Group Manchester, UKShanghai Lung Cancer Centre, ChinaRoswell Park Cancer Institute, Buffalo, USA

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


Gastro-oesophageal cancer –is CROSSing over so hard to do?

Suboptimal studies had established preoperative chemoradiation as the pre-ferred strategy in themanagement of localised oesophageal cancer (LEC)and gastro-oesophageal cancer. The recent CROSS trial has now demon-strated considerable benefit frompreoperative chemoradiation over surgeryalone in patients with LEC.But are these results only reinforcing advocatesof the preoperative chemoradiation strategy?

he incidence of gastro-oeso-phageal adenocarcinoma hasbeen rising for the past three

decades, possibly owing to the dramaticincrease in the BMI of adults in manysocieties, which has led to chronicgastro-oesophageal reflux disease andBarrett’s oesophagus.1–3 Squamous-cell carcinoma remains the most fre-quent histology in the endemic areas ofthe world, whereas adenocarcinoma isnow the most common form of gastro-oesophageal cancer in the USA and

parts of the Western World.4 Histori-cally, the management of localisedoesophageal cancer (LEC) has been asource of intense debate. Complexitiesin the clinical decision-making forpatients with LEC include the locationof the primary tumour, histological sub-type and tumour grade, clinical T andN stages, length of the tumour, abilityof the patient to withstand surgery, andprevailing practice patterns. Theremaybe unity in how to manage early-stagedisease endoscopically (for example,

stage Tis [carcinoma in situ] and T1atumours), but opinion is divided in termsof how to manage thoracic T2–T3tumours with any N stage or T1N+tumours. In patientswho canwithstandsurgery, thoracic LEC is best managedby multimodal therapy – preoperativechemotherapy or preoperative chemora-diation – because the five-year survivalrates from primary surgery are dismal,5

although high-volume centres havereduced surgicalmortality considerably.6

Multidisciplinary evaluationbefore start-ing any therapy is encouraged; however,it is not the norm in many countries(such as China and India, where mostpatients undergo surgery directly). Infact, preoperative chemotherapy for tho-racicLEC is largely abandoned inNorthAmerica but remains popular in manyEuropean countries.Regarding the preoperative chemo-

therapy strategy, a NorthAmerican ran-domised trial of this approach reportednobenefit,7 anda largerBritish trial from theMedicalResearchCouncil demonstratedonly marginal benefit.8 Preoperativechemoradiation, however,maybe estab-lishing itself as the strongest contenderamong all strategies.Results from theCROSS trial9 con-

This article was first published inNature Reviews Clinical Oncology vol. 9 no. 9, and is published with permission.© 2012 Nature Publishing Group. doi:10.1038/nrclinonc.2012.122

CLINICALONCOLOGY

impactfactor

M A R I E L A B L U M A N D J A F F E R A J A N I

T



ducted in Europe, have demonstratedconsiderable benefit from preoperativechemoradiation over surgery alone inselected patients with LEC. Yet theCROSS study may be reinforcing onlythe subscribers of the preoperativechemoradiation strategy and may notconvert the proponents of preoperativechemotherapy or primary surgery. Wefeel that the report published by vanHagen et al.,9 which represents thelargest trial of its kind,may be transfor-mative, although, we have our doubtsabout its impact on global approaches toLEC. Van Hagen et al.9 randomlyassigned 368 LEC patients with histo-logically confirmed squamous-cell car-cinoma or adenocarcinoma (tumourstage T1N1 or T2–T3 with any N) toreceive preoperative chemoradiationwith paclitaxel and carboplatin in com-binationwith 41.4Gy of 3-D conformalradiation technique in 23 fractions givenfive days per week (n=180) or surgeryalone (n=188).Althoughthe tumours were wellstaged at baseline (evenif no PET was carriedout), patients were stillselected according toage (18–75 years),weight loss (≤10%), andtumour not exceeding8 cm in length or 5 cmin width. With a median follow up of45.4months, themedian overall survivalfor the group receiving preoperativechemoradiation was 49.4 months ver-sus 24 months for the surgery-onlygroup (HR 0.67, 95%CI 0.495–0.871;P=0.003). The five-year overall survivalrate was 47% versus 34%, favouring thechemoradiation group. This benefit wasobserved in both histological subgroupsstudied; however, the effect in the ade-nocarcinoma group (the largest cohort ofthe two histological subtypes: 275patients vs 46)wasmarginal (P=0.049).Chemoradiation did not lead to exces-

sive toxicity. Other benefits from pre-operative chemoradiation included ahigher rate of R0 resection and, asexpected, a higher rate of pathologicalcomplete response in the surgical spec-imen. Data also supported the use of amoderate radiation dose of 41.4 Gy.The CROSS trial is a well con-

ceived and well-executed study thatestablishes level 1 evidence for preop-erative chemoradiation for thoracicLEC and gastro-oesophageal cancersstage T1N1 or T2–T3 with any Nstage. However, we are doubtful thatthese results will help establish a uni-form global strategy for this group ofLEC patients. This is in part becausevan Hagen et al.9 are not optimisticabout the preoperative chemoradia-tion strategy for patients with thoracicLEC and leave the door open to otheroptions (even though the evidence ofthe benefit of other options is dubious).What will it take for us to have one

global strategy forthis group of patientswithLEC?Theanswerto this quandary isunclear.However, thoseoncologists who preferpreoperative chemo-radiation are on firmground because of theCROSS trial results,

and thosewho believe in surgery first orpreoperative chemotherapy have little tobase this on.We recommend preopera-tive chemoradiation as the preferredapproach for the group of patients stud-ied in the CROSS trial. We do notendorse surgery first or preoperativechemotherapy under normal clinicalconditions, in which there is no con-traindication to radiation. It is time toCROSS over.Significant challenges remain when

dealing with a difficult disease such asLEC.Wemustdevelopstrategies that areappropriate for eachhistological subtype,

”“

for each anatomic location of LEC, andfor each stage group of LEC. We mustexploit the informationprovidedbybetterimaging, such as PET.Wemust developimagingmethods thatprovidehighly spe-cific information – those that can imageproliferation, apoptosis, hypoxia, receptorproteins, etc. before and after chemora-diation.Circulating tumourcells,mRNA,DNA, andmiRNAcan also increase ourunderstandingof theaggressivenessof thecancer.Wemustcarryout in-depthanaly-ses of the molecular biology underlyingoesophageal andgastro-oesophageal junc-tion cancer and the genetic profile of thepatients.Wemust focus more onmeth-ods to enable the immune system torecognise oesophageal and gastro-oesophageal cancer. Finally, we mustidentify patients who do not requireoesophagectomy because their cancer ishighly sensitive to chemoradiation.We feel this can be achieved through

establishing validated biomarkersignatures and/or sophisticated imagingstudies. We must, therefore, strive foroesophageal and gastro-oesophagealjunction preservation and customisa-tion of therapy. These are real challengesto deal with while we are debating ourtherapeutic preferences. These prefer-ences are usually empiric in nature andhelp only a few patients, while wesubject each one of our patients to thetoxicity of preoperative therapy and sig-nificant life-altering consequences ofoesophagectomy.10 Wewould be remissnot to mention that we must also

Yet, the CROSS studymay be reinforcingonly the subscribersof preoperativechemoradiation

Key pointTheCROSSstudy, which provided excel-lent evidence in support of preoperativechemoradiation therapy of patients withlocalised gastro-oesophageal junctioncancer, establishes a platform we canbuild on.



Cetuximab dosing by rash – is thescaling of EVEREST meaningful?

The small EVEREST trial has shown that the concept of guiding cetuximabdose escalation using the clinical parameter of acneiform skin rash is safe.However, as no significant increase of cetuximab efficacy could be observed,data from the ongoingEVEREST II trialmust be awaited before dose esca-lation can be considered for clinical use.

kin toxicity in patients receivingcetuximab is positively associ-atedwith clinical outcome.The

EVEREST study was conducted toexamine the pharmacodynamics, phar-macokinetics, pharmacogenetics andsafety of cetuximab dose escalation in aphase I/II setting in patients withmetastatic colorectal cancer.1 The studyspecifically investigatedwhether highercetuximab doses would lead to a higheroccurrence of grade 2 or 3 acneiformrash and superior treatment efficacy. Inthis study, 89 patients with or withoutminor (Common Terminology CriteriaforAdverseEvents [CTCAE] grade 0 or

1) acneiform skin rash after 21 days ofreceiving the standard dose of cetux-imab (250 mg/m² per week after initial400 mg/m²) were randomly assigned toreceive either escalated cetuximabdosesof up to 500mg/m² per week or to con-tinue with standard dosing. Patientsshowing grade ≥2 skin toxicity (n=77)continued standard cetuximab dosinguntil progression and served as a controlgroup. As with the BOND study,2

patients were eligible to enrol afteririnotecan failure andwere administeredirinotecan (180mg/m² every otherweek)as the chemotherapeutic backbone. Thepublished data of EVEREST1 focus on

the pharmacokinetic parameters, toxic-ity analyses and efficacy of the differenttreatment arms.As expected, cetuximab serum con-

centrations rose under the influence ofincreased cetuximab administration.In terms of toxicity, there was noobvious difference in haematologicaladverse events between patients receiv-ing the standard dose and those receiv-ing the elevated regimen. Consideringthe non-haematological events associ-ated with therapy, the proportion ofpatients developing hypomagnesaemiarose, and grade ≥2 cetuximab-relatedskin toxicities occurred at higher fre-quencies in the dose-escalated cohort,as predicted. In addition, the authorsreport efficacy data (objective responserate [ORR], progression-free survivaland overall survival) with no signifi-cant differences between the differ-ent treatment groups. However, a trendtowards higher ORR (30% vs 43%) anddisease control rate (70% vs 83%) wasseen in the dose-escalated group.Acneiform skin toxicity is a class

effect of all EGFR-targeting drugs cur-rently in clinical use, including erlotinibfor pancreatic and lung cancers andcetuximab and panitumumab for thetreatment of metastatic colorectal andhead-and-neck cancers.2–4 Retrospective

This article was first published online inNature Reviews Clinical Oncology vol. 9 no. 10, and is publishedwith permission.© 2012 Nature Publishing Group. doi:10.1038/nrclinonc.2012.142

S E B A S T I A N S T I N T Z I N G A N D H E I N Z - J O S E F L E N Z

S

identify adults who are at high risk fordeveloping oesophageal cancer, anddetect and treat their cancers early. Wemay have the roadmap for makingprogress against LEC, butwe seem tobewalking inmultiple directions. It is timeto collaborate.

References1. H Hampel, NS Abraham, HB El-Serag. (2005)Meta-analysis: obesity and the risk forgastroesophageal reflux disease and its complications.Ann Intern Med 143:199–211

2. LM Brown, SS Devesa, WH Chow. (2008)Incidence of adenocarcinoma of the esophagus amongwhite Americans by sex, stage, and age. JNCI100:1184–873. LM Brown et al. (1995) Adenocarcinoma of theesophagus: role of obesity and diet. JNCI 87:104–1094. Answers™. (2012) Esophageal cancer [online]http://www.answers.com/topic/esophageal-cancer5. TW Rice et al. (2009) Worldwide esophageal cancercollaboration. Dis Esophagus 22:1–86. JD Birkmeyer, Y Sun, SL Wong et al. (2007)Hospital volume and late survival after cancer surgery.Ann Surg 245:777–7837. DP Kelsen et al. (1998) Chemotherapy followed bysurgery compared with surgery alone for localizedesophageal cancer. NEJM. 339:1979–84

8. WH Allum, SP Stenning, J Bancewicz et al. (2009)Long-term results of a randomized trial of surgery withor without preoperative chemotherapy in esophagealcancer. JCO 27:5062–679. P van Hagen et al. (2012) Preoperativechemoradiotherapy for esophageal or junctional cancer.NEJM 366:2074–8410. P Viklund et al. (2006) Risk factors forcomplications after esophageal cancer resection: aprospective population-based study in Sweden. AnnSurg 243:204–211

Author affiliationsMariela Blum and Jaffer Ajani, Department ofGastrointestinal Medical Oncology, University of TexasMD Anderson Cancer Center, Houston, Texas



analyses have shown a correlationbetween the grade of acneiform rashand the efficacy of anti-EGFR therapy –irrespective of the drug, underlying dis-ease or whether the drug is given incombination with radiotherapy. Forexample, in colorectal cancer, the gradeof acneiform rash is directly associatedwith the length of the observed survival.5

In trials that investigated the efficacy ofEGFR-targeting drugs in combinationwith chemotherapy, patients with colo-rectal cancer who did not experienceany skin toxicity from theEGFR therapyhad shorter survival periods thanpatientstreated with chemotherapy alone. Thiscorrelation between rash and survivalhas been shown for erlotinib,3 cetux-imab andpanitumumab.4,5 However, thetrials each showed an overall survivalbenefit for EGFR-targeting agents withchemotherapy over the chemotherapyalone arms, irrespective of skin rash.These benefits were significant forerlotinib in pancreatic cancer (P=0.03;HR0.81)3 andhada trend to significancein cetuximab (P=0.48; HR 0.91)4 andpanitumumab(P=0.072;HR0.83).5 Sev-eral host-related factors have been pro-posed tobepredictiveof thedevelopmentof acneiformrash.Aside fromthe fact thatyounger male patients (<65 years) aremore likely to develop acneiform rash,3

molecular factors such as the single-sequence CA repeat intron-1 polymor-phism of the EGFR also predict thelikelihood of acneiform rash.Acneiform rash is attributed to the

direct inhibition of EGFR expressed inundifferentiated proliferating ker-atinocytes in the basal and suprabasallayers of the epidermis and outer layersof the hair follicle. This inhibition leadsto premature keratinocyte differentia-tion and increased cell–cell attachmentaswell as reduced growth andmigration,which collectively disrupt the formationof a normal, protective epidermal barrier.Indeed,Fracassoandcolleagues revealed

that elevated serum concentrations ofcetuximab in patients receiving esca-lated cetuximabdoses (with amaximumserum level reached at doses of400mg/m²perweek)were accompaniedby reduced EGFR protein expressionin the skin compared with patientsreceiving the standard dose.6 Thisdecreasewas already evident at a cetux-imab dose of 250 mg/m² per week andtherewas a trend to even lower levels ofEGFR expression in the 400 mg/m²cohort.6A separate study that focused onsaturation of theEGFRwith cetuximabin patientswith head-and-neck tumourswho were treated with cetuximabshowed that EGFR saturation withinthe tumour was dose dependent; aninitial cetuximab loading dose of400 mg/m² followed by 250 mg/m²weekly achieved almost complete sat-uration of EGFR in the tumour tissue.7

These data suggest that the acneiform

rash reflects the saturation of EGFR inthe tumour. Furthermore, the studydemonstrated that a small number ofpatients might benefit from 400 mg/m²per week cetuximab in terms of down-regulatingEGFRin theskin– resulting inacneiform rash– andmaximisingEGFRsaturation by cetuximab in the tumour.An inflammatory response– including

increased production of cytokines suchgranulocyte-macrophage colony-stimu-lating factor and recruitment of inflam-matory cells – also contributes to thedevelopment of the characteristic rash.8

This inflammatory response suggests thatEGFR inhibition increases the recruit-

”“

mentof inflammatorycells involved in theimmune response against the tumour.Such an immune reactionmight furtherexplain thecorrelationbetweentheoccur-rence of acneiform rash and a superioroutcome, which is observed irrespectiveof KRAS mutational status.4 Although acorrelationbetweenacneiformgradeandthe inflammatory response inside thetumour in the presence of cetuximabtherapy has not been shown, Taberneroand colleagues have demonstrated thatdownregulation of proteins downstreamof EGFR – such as phosphorylatedEGFRandphosphorylatedMAPK–andupregulationofSTAT3(signal transducerandactivator of transcription3protein) inthe skindooccur in tumours treatedwithcetuximab in adose-dependentmanner.9

The efficacy of anti-EGFR drugs isalso influenced by the presence ofgenetic alterations, which inhibit down-streampathway signalling. In colorectalcancer, in addition to KRAS codon 12and 13 mutations, NRAS, BRAF(V600E), PI3KCA and AKT1 (E17K)mutations aswell as expression ofPTENare associated with resistance to cetux-imab. Furthermore, EGFR mutationsandKRASamplificationwere also notedto be associated with cetuximab resist-ance.10 If the mechanism by whichtumours are resistant to cetuximab ther-apywas inherited, dose escalationwouldnot likely show benefit. However,patients without the genetic mutationsthat confer inherited resistance, and

...there was no obvious differencein haematological adverse eventsbetween patients receiving thestandard dose and thosereceiving the elevated regimen

Key pointHigher doses of cetuximab are well tol-erated and lead to increased grade ≥2acneiform rashes; we await the resultsfrom the prospective EVEREST II trial,which will test whether patients withincreased acneiform rash have longeroverall survival.



whose tumours do not show completeEGFR saturation when receiving stan-dard doses of cetuximab, might benefitfrom escalated cetuximab doses. Thosehigher dosesmight further downregulateEGFRexpression in the skin and, there-fore, the patientsmight develop a highergrade of acneiform rash.As almost 90% of patients treated

with cetuximab display some grade ofacneiform rash, the therapeuticwindowof cetuximab dose escalation might benarrow. Although the small size of theEVEREST trial limits the conclusionswe candraw, these preliminary datawar-rant further studies.Weawait the data ofthe prospectiveEVEREST II trial beforewe can consider the clinical implica-tions of the concept of escalating cetux-imab guided by the clinical parameter ofacneiform rash.

References1. E Van Cutsem et al. (2012) Intrapatient cetuximabdose escalation in metastatic colorectal canceraccording to the grade of early skin reactions: therandomized EVEREST study. JCO 30:2861–682. D Cunningham et al. (2004) Cetuximabmonotherapy and cetuximab plus irinotecan inirinotecan-refractory metastatic colorectal cancer.NEJM 351:337–3453. MJ Moore et al. (2007) Erlotinib plus gemcitabinecompared with gemcitabine alone in patients withadvanced pancreatic cancer: a phase III trial of theNational Cancer Institute of Canada Clinical TrialsGroup. JCO 25:1960–664. J Douillard et al. (2010) Randomized, open-label,phase III study of panitumumab (pmab) withFOLFOX4 versus FOLFOX4 alone as first-linetreatment (tx) for metastatic colorectal cancer(mCRC): efficacy by skin toxicity (ST) [abstract3528]). JCO 28 (Suppl.):15s5. HJ Lenz et al. (2006) Multicenter phase II andtranslational study of cetuximab in metastaticcolorectal carcinoma refractory to irinotecan,oxaliplatin, and fluoropyrimidines. JCO 24:4914–216. PM Fracasso et al. (2007) A phase I escalatingsingle-dose and weekly fixed-dose study of cetuximab:pharmacokinetic and pharmacodynamic rationale for

dosing. Clin Cancer Res 13:986–9937. DM Shin et al. (2001) Epidermal growth factorreceptor-targeted therapy with C225 and cisplatin inpatients with head and neck cancer. Clin Cancer Res7:1204–138. F Mascia et al. (2010) EGFR regulates theexpression of keratinocyte-derivedgranulocyte/macrophage colony-stimulating factor invitro and in vivo. J Invest Dermatol 130:682–6939. J Tabernero et al. (2010) Pharmacogenomic andpharmacoproteomic studies of cetuximab in metastaticcolorectal cancer: biomarker analysis of a phase I dose-escalation study. JCO 28:1181–8910. LA Diaz Jr et al. (2012) The molecular evolution ofacquired resistance to targeted EGFR blockade incolorectal cancers. Nature 486:537–540

Author affiliationsSebastian Stintzing, Department of Oncology andComprehensive Cancer Centre, Klinikum Groβhadern,University of Munich, Munich, Germany. Heinz-JosefLenz, USC/Norris Comprehensive Cancer Center,Keck School of Medicine, Los Angeles, California

Competing interestsSebastian Stintzing declares associations with Amgen,Merck KGaA and Roche. Heinz-Josef Lenz declaresan association with Merck KGaA. See the originalarticle online for full details of the relationships.

� 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

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

www.cancerworld.org

cancerworld

N E W S R O U N D


newsround

Glioblastoma: temozolomideoffers alternative to radio-therapy in elderly patients� The Lancet

Both temozolomide and hypofractionatedradiotherapy should be considered as stan-

dard treatment options for elderly patientswithglioblastoma, a phase III study from theNordicClinical Brain Tumour Study Group (NCBTSG)has found.

In 2004 chemoradiotherapy with temo-zolomide became the standard of care forpatientswith glioblastoma, but its introductionwas based on a pivotal study inwhich patientswere aged 70 years or younger. In other studies,increasing agehas been shown to be a negativeprognostic factor, leading to the suggestionthat elderly and frail patients might not beviewed as candidates for combined therapy.

To define an evidence-based treatment forpatients aged 60 years or older with glioblas-toma, NCBTSG investigators undertook a ran-domised trial to compare health-related qualityof life and safety in patients randomised tosingle-agent temozolomide chemotherapy,short-course hypofractionated radiotherapy(34.0 Gy administered in 3.4 Gy fractions overtwo weeks) or standard six-week radiotherapy(60.0Gy administered in 20.0Gy fractions oversix weeks). Both patients and staff were awareof treatment assignments.

Between February 2000 and June 2009,342 patients with newly diagnosed, histologi-

neutropenia (n=12)andthrombocytopenia (n=18).“We found that temozolomide chemother-

apy is a potential alternative to radiotherapy inelderly and frail patients,”write the authors. Theresults, they add, support the predictive value ofMGMT promoter methylation as a useful bio-marker in guiding treatment decisions aroundtemozolomide.

In an accompanying commentary, PhioanhLeia Nghiemphu and Timothy Cloughesy, fromtheUniversity of California at LosAngeles,write,“TheNordic study is awell balanced randomisedtrial that provides provocative results andgreatly contributes to the understanding ofgeriatric neuro-oncology. For patients aged 70years and younger, radiation at least did notnegatively affect survival and provides insightinto a molecular subgroup that might be wellsuited to treatment with temozolomide.”

A collective effort towards systematic pri-oritisation of the effects of all factors on prog-nosis, they add, will enable classification ofprognostic subgroups for prospective investi-gation and eventually lead to definition of rel-evant optimum treatments.

� A Malmström, B Henning Gronberg, C Marosi

et al. Temozolomide versus standard 6-week

radiotherapy versus hypofractionated radiotherapy

in patients older than 60 years with glioblastoma:

the Nordic randomised, phase 3 trial. Lancet

Oncology, September 2012, 13:916–926

� P Nghiemphu, T Cloughesy. Glioblastoma

therapy in the elderly: one age does not fit all. ibid,

pp 857–858

cally confirmed glioblastoma (WHO grade IVastrocytoma) from 28 centres in Austria, Den-mark, France, Norway, Sweden, Switzerland andTurkey were recruited.

In the study, 291 patientswere randomisedacross three treatment groups: temozolomide(n=93), hypofractionated radiotherapy (n=98),and standard radiotherapy (n=100). An addi-tional 51 patients were randomised acrossonly two groups: temozolomide (n=26) andhypofractionated radiotherapy (n=25).

Results for the three-group randomisationshow that median overall survival was6.0 months for standard radiotherapy versus8.3 months with temozolomide (HR 0.70,95%CI 0.52–0.93,P=0.01); and7.5monthswithhypofractionated radiotherapy (HR 0.85,95%CI 0.64–1.12, P=0.24).

In the two-group randomisation, overallsurvival was 8.4 months for patients whoreceived temozolomide versus 7.4 months forpatientswho received hypofractionated radio-therapy (HR 0.82, 95%CI 0.63–1.06; P=0.12).

For patients older than 70 years, survivalwas better with temozolomide (HR 0.35;P<0.001) and with hypofractionated radio-therapy (HR 0.59; P=0.02), in comparison withstandard radiotherapy.

An additional finding was that patientstreated with temozolomide who had tumourMGMT promoter methylation showed signifi-cantly longer survival than those without (HR0.56; P=0.02).

As expected, the most common grade 3–4adverse events in the temozolomide group were

Selected reports edited by Janet Fricker


N E W S R O U N D

R-CHOP best forolder patients withmantle-cell lymphoma� New England Journal of Medicine

For older patients with mantle-cell lym-phoma, a rituximab-based chemotherapy

regimen followed by rituximab maintenancetherapy improves survival, according to theresults of a study by the European Mantle CellLymphoma Network.

Treatment options for older patients withmantle-cell lymphoma are limited, as the stan-dard first-line therapy approach of high-dosecytarabine, followed by autologous stem-celltransplantation, is usually not feasible. Themedian age at diagnosis for mantle-celllymphoma is about 65 years.

In the current study, Habbeje Kluin-Nelemans and colleagues, from theGroningenUniversity Medical Centre, in the Netherlands,compared two induction regimens, followedby two different maintenance therapies forthose showing a response.

Between January 2004 and October 2010investigators randomly assigned 560 patientsaged 60 years or older with mantle-cell lym-phoma, stage II to IV, whowere not eligible forhigh doses to one of two alternative treatmentarms: six cycles of rituximab, fludarabine andcyclophosphamide (R-FC) every 28days or eightcycles of rituximab, cyclophosphamide, dox-orubicin, vincristine, and prednisone (R-CHOP)every 21 days. Altogether 532 of these patientswere included in the intention-to-treat analy-sis and 485 in the primary analysis.

Results showed that the four-year survivalratewas 47% for R-FC versus 62% for R-CHOP(P=0.005). Furthermore, 10%of patients in theR-FC group died during the first remission ver-sus 4% in the R-CHOP group.

Complete remission rates were 40% forR-FC versus 34% for R-CHOP (P=0.10), andprogressive diseasewas14%forR-FC versus 5%for R-CHOP.

Among those re-randomised to mainte-nance therapy, 58%of those receiving rituximab

were in remission after four years versus 29%receiving interferon-alpha (HR for progressionor death 0.55, 95%CI 0.36–0.87, P=0.01). Fur-thermore, among patientswhohad a responseto R-CHOP, the four-year survival of patientsreceivingmaintenance therapywith rituximabwas 87%, versus 63% for interferon-alpha(P=0.005).

“In conclusion, older patientswithmantle-cell lymphomawhohave a response toR-CHOPand continue to receive rituximab as mainte-nance therapy have a longer life expectancythan those who receive maintenance therapywith interferon-alfa,” write the authors.

The outcomes for R-FC, they add,were dis-appointing, given the high expectations thatthis regimen had in the early 2000s. In future,they add, it might be “attractive” to combinerituximab-based maintenance regimens withother drugs shown to be active againstmantle-cell lymphoma, such as bendamustine or mol-ecularly targeted approaches.

“However, physicians need to be aware ofthe potential interactions between the initialtherapy and the maintenance regimen,” writethe authors.

� H Kluin-Nelemans, E Hoster, O Hermine et al.

Treatment of older patients with mantle cell

lymphoma. NEJM 9 August 2012, 367:520–531

Observation effectivein prostate cancerwith low PSA� New England Journal of Medicine

F or men with localised prostate cancerdetected throughprostate-specific antigen

(PSA) testing, radical prostatectomydelivers nosurvival benefits over 12 years follow-up incomparison to observation alone, the PIVOTstudyhas reported. Subgroupanalyses of theUSstudy, however, suggest surgery reduces mor-tality amongprostate cancer patientswith highPSA levels.

Although the lifetime risk of receiving a

diagnosis of prostate cancer is about 17%, therisk of dying from the disease is approximately3%, suggesting conservativemanagementmaybe appropriate for somemen. But the observa-tion option is rarely offered due to lack of evi-dence from randomised trials.

In the Prostate Cancer Intervention versusObservationTrial (PIVOT) study, TimWilt, fromtheUniversity of Minnesota School of Medicine,Minneapolis, and colleagues conducted a ran-domised trial to compare radical prostatectomywith observation in men who had received adiagnosis of clinically localisedprostate cancer inthe early eraof PSA testing. BetweenNovember1994and January 2004, 731menwith localisedprostate cancer were randomised to radicalprostatectomy (n=364) or observation (n=367).Patients, who had amean age of 67 years, wererecruited from44DepartmentofVeteransAffairssitesandeightNationalCancer Institute sitesandfollowed through until January 2010.

Results during a median follow-up of 10years show 47.0% of men assigned toradical prostatectomy died compared with49.9% assigned to observation (HR 0.88,95%CI 0.71–1.08;P=0.22). Additionally, 5.8%ofmen assigned to radical prostatectomy diedfrom prostate cancer or treatment comparedwith 8.4% assigned to observation (HR 0.63,95%CI 0.36–1.09; P=0.09). During the first 30days after surgery, perioperative complicationsoccurred in 21.4% of men undergoing radicalprostatectomy, with the most common com-plication being wound infections (4.3%).

Among men with a PSA value greater than10 ng/ml, surgery reduced all-cause mortalityby 13.2%. Among men with intermediate-risk tumours (determined by a PSA value of10.1–20.0 ng/ml, a Gleason score of 7, or a stageT2b tumour), those randomly assigned to surgeryhada31%relative reduction in all-causemortal-ity comparedwith those assigned to observation.

“Our findings support observation formenwith localized prostate cancer, especially thosewhohave a lowPSA value and thosewhohavelow-risk disease,” write the authors. The study,they add,was conducted in the early era of PSAtesting. “The current practices of performingrepeated PSA testing, using a lower PSA

N E W S R O U N D


threshold for biopsy, obtaining more tissue-biopsy cores, and performing a repeat biopsyafter initially negative findings increase thedetection of smaller volume indolent cancers.…These factors increase the likelihood of overdiagnosis and overtreatment.”

In an accompanying commentary, IanThompson from the University of Texas HealthScience Center, San Antonio, and CatherineTangen from the FredHutchinson Cancer Cen-ter, Seattle, stress that the men most likely tobenefit from therapy are thosewhose prostatecancers pose the greatest risk of death fromcancer. “The screening, detection, and treat-ment we provide must focus on cancers thatmatter, and future clinical trials must do so aswell,” they write.

� T Wilt, M Brawer, K Jones. Radical prosta-

tectomy versus observation for localised prostate

cancer. NEJM 19 July 2012, 367:203–213

� I Thompson, C Tangen. Prostate cancer –

uncertainty and a way forward. ibid pp 270–271

Pyridoxine notrecommended inhand-foot syndrome� British Journal of Cancer

While pyridoxine (vitamin B6)may reducethe incidence of severe hand-foot

syndrome (HFS) and the need for capecitabinedose modifications in patients with advancedcolorectal or breast cancers, no antitumourbenefits were detected, the CAP-IT study hasreported. Routine use of pyridoxine for HFS,conclude the UK investigators, should not berecommended.

Pyridoxine is frequently used to treatcapecitabine-inducedHFS. SinceHFS resemblesthe ratdiseaseacrodynia, knowntobecausedbypyridoxinedeficiency, treatmentwithpyridoxinehasbeenproposed. There is, however,noevidenceforbenefit. In thecurrent study, PippaCorrie andcolleagues from Addenbrooke’s Hospital, Cam-bridge, performed a randomised placebo-

Study suggests newstandard of care forplatelet transfusions� The Lancet

Therapeutic platelet transfusions couldbecome the new standard of care for

patientswithhaematologicalmalignancieswhohave received autologous stem cell transplan-tation, a German study suggests. Prophylacticplatelet transfusion, however, should remain thestandard for patients with acute myeloidleukaemia forwhomspecial attention is neededdue to increased risk of central nervous system(CNS) bleeding.

Routine prophylactic platelet transfusion isthe standard of care for patients with severethrombocytopenia with morning plateletcounts of 10x109 per litre or lower. However,whether such transfusions are necessary forclinically stable patients with no bleeding haslong been debated. Small studies performed30 years ago showed favourable results for thetherapeutic strategy (where transfusions areoffered following bleeds), but these results areno longer considered applicable to current clin-ical practice due to changes in chemotherapydose intensities and supportive care.

In 2005 and 2006 two single-centre pilotstudies showed that a new strategy of thera-peutic platelet transfusionwas feasible,with noincreased risk inmajor bleeding and a substan-tially reduced number of platelet transfusionscomparedwithhistorical controls. In the currentstudy, the same team, led by Hannes Wandtfrom the KlinikumNuremberg Nord, Germany,investigated whether these results could bereproduced prospectively in amulticentre ran-domised study.

In the study, patients aged 16–80 yearsundergoing intensive chemotherapy for acutemyeloid leukaemia or autologous haematopoi-etic stem-cell transplantation for haematologi-cal cancers were randomised to receive platelettransfusions either when bleeding occurred(therapeutic strategy, n=199) or whenmorningplatelet counts were 10x109 per litre or lower

controlled trial todeterminewhetherpyridoxineavoided the need for capecitabine dosemodifi-cations and furthermore improved outcomes.

Altogether 106 patientswith amedian ageof 73 years scheduled for palliative single-agentcapecitabine (65% of whom had colorectalcancer and 35% breast cancer) were ran-domised in a 1:1 ratio, betweenDecember 2004and June 2009, to receive either concomitantpyridoxine (50mg;n=53) ormatching placebo(n=53) three times daily, commencing on theday capecitabine chemotherapy was initiated.Treatment continued until disease progression,toxicity or patient preference. After discontin-uation, patientswere followedup for 12weeks.

Results showed that 37% of patients ran-domised to pyridoxine avoided capecitabinedose modifications versus 23% randomised toplacebo (RR 0.59, 95%CI 0.29–1.20; P=0.15).Furthermore, 9% of patients in the pyridoxinegroup experienced grade 3/4 HFS-relatedadverse events versus 17% in the placebogroup(P=0.26). Therewas a trend towards pyridoxinedecreasing progression-free survival (PFS), witha median PFS duration of 7.4 months for pyri-doxine and 9.9 months for placebo (HR 1.62,95%CI 0.91–2.88; P=0.095).

“Pyridoxine appeared to reduce the inci-dence of grade 3/4 HFS and the need forcapecitabine dosemodifications, although thisdid not translate into an improvement in out-come from chemotherapy itself; the trendtowards poorer PFS in the pyridoxine arm wasnot statistically significant,” write the authors.“Whether pyridoxine might in fact negativelyinfluence chemotherapy efficacy is intriguing,although not conclusive,” they add.

The authors also refer to an earlier study,which reported significantly lower tumourresponses to capecitabine at the higher pyri-doxine dose level, while a second study, involv-inguseofpyridoxine inadvancedovariancancer,found it reduceddurations of response to treat-ment with hexamethylamine plus cisplatin.

� P Corrie, R Bulusu, C Wilson, et al. A

randomised study evaluating the use of pyridoxine

to avoid capecitabine dose modifications. Br J

Cancer 7 August 2012, 107:585–587


N E W S R O U N D

(prophylactic strategy,n=197).Altogether190ofthe patients had acute myeloid leukaemia and201hadundergoneautologous transplantation.

The studywas undertaken between Febru-ary 2005 and May 2010 at eight haematologycentres in Germany.

Results show that, for all patients, the pri-mary endpoint of platelet transfusions occurredin 2.44% of patients in the prophylactic groupversus 1.63% in the therapeutic group(P<0.0001). For those with acute myeloidleukaemia, transfusions occurred in 2.68% ran-domised to theprophylactic groupversus 1.83%to the therapeutic group, representing a 31.6%reduction (P<0.0001).While for thosewhohadautologous transplantation, transfusionsoccurred in 1.8% in the prophylactic groupversus 1.18% in the therapeutic group, repre-senting a 34.2% reduction (P=0.0193).

For patients undergoing autologous trans-plantation, randomisation to the therapeuticarmdid not increase the risk ofmajor haemor-rhage; but for those with acute myeloidleukaemia in the therapeutic arm, the risk ofnon-fatal grade 4 bleeding significantlyincreased (mostly CNS) compared to the pro-phylactic group (P=0.0095).

“Our findings show that the number ofplatelet transfusionswas significantly lower, byroughly a third, in the therapeutic group than inthe prophylactic group. However, this clinicallymeaningful differencemust beweighedagainstthe increased bleeding risk,” write the authors.

The new strategy of therapeutic platelettransfusions inpatientswhohave receivedautol-ogous stemcell transplantation, theyadd, shouldbeusedonly byhaematology centreswhere staffare experienced in the approach and can react ina timelyway to first signs of CNS bleeding.

In an accompanying commentary, NeilBlumbergandcolleagues fromtheUniversity ofRochesterMedical Center, NewYork,write, “Theemerginghypothesis, ... is that transfusedplateletsmight be promoters of arterial and venousthrombosis, tumour growth, and metastasis.These possibilities provide additional reasons tofavoura restrictivepolicy forplatelet transfusionin view of themoderate benefits of transfusionshown in autologous transplant patients.”

� H Wandt, K Schaefer-Eckart, K Wendelin et

al. Therapeutic platelet transfusion versus routine

prophylactic transfusion in patients with

haematological malignancies: an open-label,

multicentre, randomised study. Lancet, published

online 7 August 2012, doi:10.1016/S0140-

6736(12)60689-8

� N Blumberg, J Heal, G Phillips, et al. Platelets

– to transfuse or not to transfuse. ibid, doi:10.1016/

S0140-6736(12)60983-0

Survival advantagefor centralisationof vulvar surgery� European Journal of Cancer

Centralisation of care for womenwith vul-var squamous cell carcinoma (SCC) is

associated with improved survival, a Dutchstudy has reported.

In 2000, guidelines from the Dutch Soci-ety of Obstetrics and Gynaecology recom-mended centralisation of care for patientswith vulvar SCC. Benefits identified for thisapproach included the development ofexpertise, and the opportunity to give patientsappropriate treatment from experienced cli-nicians using new techniques that mightimprove prognosis and/or lower treatment-related morbidity. The strategy was alsothought to facilitate training and research.

The cornerstone of treatment for vulvarcarcinoma is surgery, which offers an excellentchance of cure. In other rare malignancies,such as oesophageal and pancreatic carcino-mas, associations have been found between thevolume and/or specialisation of a hospital onthe one hand and better survival on the otherhand. In recent years, treatment of patientswith early-stage vulvar SCC has shifted frominguinofemoral lymphadenectomy to the sen-tinel lymph node dissection (SLND) procedure.To meet quality standards, it has been sug-gested that surgeons should perform SLNDsurgery at least 510 times per year. For a raretumour such as vulvar SCC, with an annual

incidence of one to two cases per 100,000women, this would require centralisation.

In the current study, Loes van den Eindenand colleagues, from Radboud University,NijmegenMedical Centre, in the Netherlands,set out to determine whether guidelines hadbeen adopted and whether such adoptionhad resulted in improvements in survival. Theguidelines were introduced in 2000. For thestudy, data on all patients diagnosed withvulvar malignancies between 1989 and 2008in the eastern part of the Netherlands wereretrieved from the population-based cancerregistry held by the Comprehensive CancerCentre IKNL. Data for patients diagnosedbefore the introduction of guidelines(1989–1999) were compared with those forpatients diagnosed after (2000–2008).

A total of 382 patients with vulvar SCCwith invasion >1mm, who had an indicationfor groin surgery, were included in the analy-sis. In the first decade, 62% (123 out of 198patients) were treated in a specialised oncol-ogy centre, which increased to 93% (172 outof 184 patients) in the more recent period(P<0.0001). The five-year relative survival was69% for the first period, compared to 75% forthe second period. After adjustment for ageand stage, being treated in a specialised oncol-ogy centre was found to be an independentprognostic factor for survival. Patients treatedin a specialised oncology centre in the period2000–2008 appeared to have comparablefive-year relative survival rates compared topatients treated in specialised centres in theperiod 1989–1999.

“In conclusion, the present study showedthat centralisation of the treatment ofpatients with vulvar SCCwho need groin sur-gery has beenwell adopted in the Eastern partof the Netherlands. Being treated in a spe-cialised oncology centre is associated with abetter survival,” write the authors.

� L van den Einden, K Aben, L Massuger et al.

Successful centralisation of patients with vulvar

carcinoma: a population-based study in the

Netherlands. Eur J Cancer, September 2012,

48:1997–2003

M Y W O R L D


Maria João Cardoso is head breast surgeon at the Champalimaud Cancer

Centre in Lisbon. She founded the patient support centre Mama Help, and

leads a research group at the Institute for Systems and Computer

Engineering, in Porto, on improving outcomes in breast reconstruction.

� Why I chose to work in cancerBecause it was and still is a mysteriousdisease with so many things yet to dis-cover. Working in an institution dedi-cated to translational research, and beinga part of the process, makes your careermuch more interesting.

� What I love most about my jobPatients’gratitude! When a patient looksat you as the one who pushed deathaway, it has a profound effect on you.

� The hardest thing about my jobWhen disease progresses and we have totell patients the bad news. In thosemomentsyou feel yourpatient’sdespair asif it were your own. But you have to moveon and learn to deal with that feeling.

� What I've learnt about myselfThat patients are my best teachers. Theyare usually strong and optimistic andeven when they are facing a difficult sit-uation, they manage to see the best sideof it. I’ve learnt from their examples andfound ways to use what I’ve learnt.

� I'll never forget...The opening of our first support centrefor breast cancer patients in 2011, called

Mama Help. It was a project in my headfor almost 10 years. Being able to trans-form it into reality, and watching thebenefit it brings to patients... simplymarvellous!

� A high point in my careerThe invitation by Dra. Fátima Cardoso,director of the Breast Unit, to be thehead breast surgeon at the Champali-maud Cancer Centre. I felt very impor-tant and honoured. To be able to workwith one of the best specialists in breastcancer in the world was a dream cometrue.

� I wish I were better at...Being more tolerant with others. Whenyou work hard you tend to evaluate oth-ers by your own parameters and it canoften be difficult to accept even minorfailures. I feel that younger doctors aresometimes really afraid of me – I can seeit in their faces!

� What I value most in a colleagueHonesty. When you work in a team,your input, as well as others’, is importantto achieve the best results. Intellectualand scientific honesty are fundamentalingredients.

� The most significant advance inmy specialty in recent yearsImmediate breast reconstruction andsentinel node biopsy. Quality of life isvery important, particularly with pro-longed survival. These procedures offerpatients two major improvements intheir quality of life.

� My advice to someone enteringmy specialty today would be...Beahardworker.Youwillnotachieveany-thing without hard work. When you lovethis work you never stop. You finish yourappointments when your patients wantyou to, not when you finish your dailyagenda! And after 12 hours of surgeryand an absolutely impossible week youstill spend your weekend correcting arti-clesor rushing tomeet researchdeadlines.

� What I wish I’d learnt atmedical schoolHealth economics. When you start work-ing, you have no idea about health costs.We have learnt to act as others before usdid. Private and public medicine hasenormous costs and oncology is one of itsmajor consumers. To serve patients bet-ter we need to have at least some idea ofhow much it will cost to treat them.

My World

cancer world 51

Documents