cancer world 43

�Peter Naredi: a ‘can do’ leader for Europe’s cancer surgeons�Unshackling progress inchildhood cancers�Prof Elisabeth de Vries revels in a career unburdened by expectations� Beyond survival: are the old standards still the best for evaluating new cancer drugs?

Peter Naredi

Education & knowledge through people & facts

Number 43, July-August 2011

CancerWorld

43JU

LY-AUGUST

2011

CANCER WORLD � JULY/AUGUST 2011 � 1

Contents

3 EditorialPatients need information. Is that clear?

4 Cover StoryPeter Naredi: a ‘can do’ leader for Europe’s cancer surgeons

13 e-Grand RoundDecision making in the treatment of gliomas

22 Cutting EdgeBeyond survival: what should new cancer drugs have to prove and how?

32 MasterpieceFor the love of the job: Holland’s first woman professor of medical oncologyrevels in a career unburdened by expectations and driving ambition

40 Impact FactorRituximab for follicular lymphoma: maintaining an open mindAster – another flower in the diagnostic field of lung cancer?

48 NewsroundSelected news reports

54 Spotlight on...Unshackling progress in the care of childhood cancers

60 FocusThink yourself better

EditorKathy [email protected]

Assistant EditorAnna Wagstaff

Editorial AssistantCorinne Hall

Editorial AdvisorsJacques BernierFatima CardosoFranco CavalliAlberto CostaVincent T. DeVita

Contributing WritersPaul Baas, Marc Beishon,Bruce Cheson, Simon CromptonJanet Fricker, Anna Wagstaff

Publishing AdvisorsGillian Griffith, Fedele Gubitosi

Website LiaisonCorinne Hall

Art EditorJason Harris

ProductionHarrisDPIwww.harrisdpi.co.uk

Printed byGrafiche Porpora

Cover photographJohan Gunséus/Synk

Published byEuropean School of Oncology

Direttore responsabileAlberto Costa

Registrazione Tribunale di RomaDecreto n. 436 del 8.11.2004

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

Copyright ©2011 European School of Oncology.All rights reserved

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


Editorial

Theurgent quest for novel cancertreatments engages much of ourattention, but could we be over-

looking other opportunities for increasingpatients’ survival and quality of life?

One such opportunity that is increas-ingly gaining the attention of policy makersis improvinghealth through improving healthliteracy. Defined as “the skills which deter-mine themotivation and ability of individu-als to gain access to, understand, and useinformation in ways which promote andmaintain good health,” good health literacycan play a crucial role in improving out-comes, whether it be in prevention, treat-ment, palliative care or survivorship.

Poor health literacy is associated withpoorer general health status, increased risk ofhospitalisation and a lower capacity to carefor oneself and share in treatment decisionmaking. The problem is greater among lowersocio-economic groups, ethnic minoritiesand the elderly, and consequently thesegroups have much to gain by health literacyinitiatives tailored to their needs.

A recentEuropean surveyhas shown thatone in tenpatients finds the informationpro-vided by physicians difficult to understandandmanymore struggle todecipher the infor-mationprovided on amedicine leaflet.How-ever, the problem is under-recognised andpoorly addressed by health professionals,many of whom overestimate patients’ abilityto understand and use health information.

As more and more cancers evolve intochronic conditions, we need to focus on

� Kathy Redmond � EDITOR

how to meet the needs of cancer patientswith literacy problems. As a first step weshould audit patient informationmaterials incurrent use to find out how far they werewrittenwith theneeds of less literate patientsinmind.Do they complywith the principlesof clear health communication?Would theypass the clarity test if subjected to a read-ability assessment?Similar assessments donefor other groups of patients suggestmuch ofit would not.

There is help at hand.Anumberof groupshavedevelopeduseful guides onhow towritematerials for patientswith low literacy levels.Literacy experts recommendplain language,shorter sentences and larger type sizes, witha sharp contrast between the text and back-ground. Testing draft materials on the targetaudience is also important. These are com-mon-sense recommendations which shouldbecome the gold standard for the develop-ment of all patient education resources.

Health professionals can help promotehealth literacy by using jargon-free plain lan-guage in all their interactions with patients.They can also assess patients’ literacy levelsusing one of the readily available and easy toadminister health literacy assessment tools.This could help in tailoring information to thepatient’s level ofunderstanding, in linewith thecurrent personalised approach tomedicine.

These are stepswe can all take right nowto improve cancer outcomes, which are cur-rently compromisedby adisconnect betweenwhat professionals think patients need andwhat patients actually need.

Patients need information.Is that clear?

CoverStory

4 � CANCER WORLD � JULY/AUGUST 2011

Peter Naredi: a ‘can do’ leaderfor Europe’s cancer surgeons

� Marc Beishon

There are somanyways cancer surgeons can help improve outcomes, andPeterNaredi embraces

themall.Adapting surgical approaches to thebiologyof acancer, spreading bestpractice,usingaudit

and transparency to raise theworst to the level of thebest are thingshe’s triedand tested inhisnative

Sweden.AsESSOpresident, henowhopes to enthuseEurope’s cancer surgeons to followhis lead.

As surgery is the pivotal treatment formany typesof cancer–andwill remainso for the foreseeable future – onemight expect that the discipline ofsurgical oncology would be well

entrenched in national practice around Europe bynow, especially as somuch surgery concerns cancer.But that is far fromthecase, reportsPeterNaredi, thecurrent president of the European Society of Surgi-calOncology (ESSO).Ashenotes, it is only a recog-nised speciality in a fewcountries, and there ismuchmore to the cancer surgeon’s role than just carryingout operations.

“Inmanyhospitals – suchas innorthernSweden,where I am based – there may be few medicaloncologists, and surgeons are most likely to be theones leading patients through their cancer journey,”he says. “What we are emphasising at ESSO is theneed for surgeons to participate in quality and edu-cational programmes to raise standards in oncologysurgery, and the establishment ofmultiprofessionalcentres and regional working so that patients havethe best outcomes, not just from surgery but in

other areas such as diagnosis and end-of-life care.”As he adds, there is now an unstoppable move-

ment towards auditing andpublishingoutcomedatafor hospitals andeven for individual surgeons in cer-tain countries, driven by politicians and patientgroups. As a result, the variability of cancer out-comes will become more apparent. Data from reg-istries and results from multicentre trials alreadyshow “remarkable” differencesbetween institutionsand between treatment of different tumour typesaroundEurope, and surgeons aremost often takingthe lead in diagnosis and care.

Itmightbeexpected that,with surgerybecomingmore specialised and with many surgeons focusingonly on specific areas such as urology or head andneck, the quality of cancer treatment would be anintegral part of this trend. But organ-specialist sur-geons do not necessarily have a detailed and up-to-dateknowledgeof cancer, for instance itsbiologyandmultiprofessional care, which means patients mayreceive suboptimal treatment, says Naredi. “WhatEurope lacks is the implementation of a corecurriculum in surgical oncology, which we have

CoverStory


developed at ESSO, and also the integration oforgan-specialist societies and national bodies intoESSO and other cancer societies.”

Promoting theESSOcorecurriculumisacurrentpriority forNaredi andcolleagues, as iswidening thesociety’s membership to embrace national surgicalbodies and powerful groups such as the EuropeanAssociationofUrology,whichalonehasabout12,000members.Audit andquality assurance is anotherpri-ority, for example through theEuropeanRegistrationofCancerCare (EURECCA)project,whichwas setupbyESSOandadoptedbyECCOinitially to auditcolorectal cancer surgery aroundEurope, andwhichcould be a framework for other tumours (seewww.canceraudit.eu).

Naredi, whose day job is professor of surgery atUmeå University, not far from the Arctic Circle inSweden, is himself an ideal case studyof developingsurgical and multiprofessional excellence in one ofEurope’s outposts. Since Sweden decided to estab-

lish six regional cancer centres, each focusedonkeyteachinghospitals, thenorthern regionbasedaroundUmeå has become recognised as one of the moreinnovative, despite two other regions startingmuchearlier. “This isnotprimarily aboutmoremoney,” saysNaredi. “Yes, we canmake say a 10% improvementwithmore funds,butwecanachieve30%by improv-ing what we already have in terms of the process ofgettingpeoplewithcancer symptomsdiagnosedandtreated faster and better in the right places.”

As a general surgeonwho specialises in the ‘mid-GI’ area – especially the liver and pancreas – he hashelped introducenew techniques to Swedish surgi-cal oncology.Naredi alsohas a researchbackgroundin basic science and continues to carry out work inareas such as immunotherapy and chemotherapyresistance.And as part of a general surgical team inGothenburg, hewas routinely involved in caring forpeoplewithdiseases suchas stage IVmelanoma.Thishas given him good grounding in the challenges of

JOHANGUNSÉUS/SYNK

improvingmultidisciplinary careandattracting talentto a regional university hospital.

Above all, he adds, healthcare bureaucrats needto allow clinicians such as surgeons the freedom tointroduceevidence-basedstructures thatwill improvecancer outcomes, andnot force throughchange thatdisempowers people. Naredi speaks from experi-encehere: heenjoyedagooddeal of autonomywhileat Sahlgrenska hospital inGothenburg, one of Swe-den’s leading institutions, until amerger of threehos-pitals created too much middle management,prompting him to leave alongwith other colleagues.

“But for ourpartweneed to show leadership,” hesays. “We are in fact running leadership courses foryoung surgeons inSweden, under theSwedishSur-gicalSociety, becauseourprofessionhas tobeable totell thepoliticians andadministratorswhat is best inhealthcare.” It’s also about painting a vision of whatsurgeons of the future should be doing, he adds, asthere is a degree of insecurity about their roles.

Naredi acknowledges that the European cancerworldhasnot lackedstrongcharacters,particularly from

thesurgery side,where therehavebeenquitea fewout-spoken and sometimes controversial senior figures.But like many of the younger generation who havestepped up to senior level now in oncology, he favoursa non-hierarchical, consensus-building approach thatmotivates rather than forcesotherpeople toparticipate.In his world, there is no room for the all-powerfulchief surgeonwho dominates decisionmaking.

He was the first in his family to become a doctor,influencedbyhismother,whoworked as aRedCrossnurse. “Iwanted to be an architect at first, butwhen Isaw thekindofwork Imight bedoing, suchas interiordesign inbanks, I knewIwanted todosomethingmoremeaningful and I’venever regretteddoingmedicine. Ichose surgery because I’m a practical person.”

After a residency inHalmstad hemoved to be award physician in the department of surgery atSahlgrenska hospital in Gothenburg, where he wasable to carve out a dual surgical and research career,focusing on cancer. “Iwas doing general surgery butfoundIwas learningmuchmore fromcancerpatientsthan say those who were having gall bladder or hip

CoverStory


A specialist at work.A minimalist

approach to treatingliver metastases,

which Naredi helpedto develop, has

resulted insignificantly more

patients becomingeligible for treatment JO

HANGUNSÉUS/SYNK

recognitionof surgical oncologyaroundEurope,butwedoneed to getmore surgeons interested in thebiologyof cancer and all the other aspects of treatment andcare. We need more good surgeons who understandoncology – not just dedicated surgical oncologists.”

A surgeon can be specialised in one organ, saybreast, but still learn about techniques developed inother areas such as the pelvis, says Naredi. All sur-geons need to keep up to date now with new drugssuchas targeted therapies, and the core curriculum,he emphasises, is as much about giving hospitaldepartments a framework to be a surgical oncologyteachingunit as it is about individual learning.Ashepoints out, there is no validation andaccreditationofsuch teaching capability as yet.

“Wehave tried to give theESSOcurriculum thesame format as the ones from ESMO [for medicaloncologists] andESTRO[for radiationoncologists],so thatECCO’smember societieshave standardcur-ricula,” addsNaredi. “But aswith recognition of ourspeciality, I’m not a big believer in thinking that youcan just impose it at national level – we have towork with people who join and interact with ESSOto take ithomeandadapt it for theirownsurgical soci-eties and institutions.

“We are not specifying detailed surgical proce-dures in the curriculum, just guidance on the numberof procedures. The latest hands-on learning does notbelong in the curriculum. For example, in SwedenweinvitedBillHeald fromtheUKto leadsessionsonTME[total mesorectal excision] for rectal surgery, whichthen made its way into national guidelines from ourcolorectal surgical society.Our aim at ESSO is to pro-mote the tools for implementing such best practice.”

Naredi himself benefited fromexcellent surgicalmentorship at Gothenburg, but also has a strongresearchbackground,having takenupa fellowshipattheUniversity of California in SanDiego, where hestudiedchemotherapy resistance (mainly cisplatin),andhe also has aPhD in tumour blood flow.Hehasa long collaboration with Swedish tumour immu-nologistKristofferHellstrandontheuseofhistamines

replacement operations.Everyonewith cancerhas adifferent life story and there are so many feelingsinvolved.Youhave to listen carefully to improve out-comes for survivors – nine out of ten women withbreast cancer in Sweden now survive.We also needto listen to those with poor chances, such as thosewithpancreatic cancer,whereweneed toknowhowto provide good palliative care. Treating peoplewithstage IVmelanoma, who havemiserable outcomes,has taughtmemore than any course.” Patients needonedoctorwhocanput together amultiprofessionalpackage, he says. “We shouldn’t keep sending themto see different people to take control of their care –they need confidence in one person.”

In the11hospitals thatcomprise the regional can-cer centre innorthernSweden,Naredi says only twohave a department of medical oncology. “Medicaloncologists cometo theotherhospitals asconsultantsandmay seeup to20patients in aday, butwho takescareof themafterwards? In the vastmajority of casesit is the surgeon who will be seeing patients over aperiodof severalmonths,which iswhy it is so impor-tant they have knowledge of surgical oncology.”

AsNaredi explains, surgical oncology is of courseabout excellence in treating solid tumours (althoughnot in the brain, which is the domain of the neuro-surgeon), but it also includes prevention, geneticcounselling,diagnosticandstagingprocedures, rehab-ilitationandfollow-upcare.Andtreatment for thesur-geon does not just mean resection, but also gaininga thorough understanding of the biology of the dis-eases and the use of chemo- and radiotherapy.

“AtESSOwehaveboth a core curriculumandaEuropean examination from the surgical section ofUEMS [European Union of Medical Specialists],which takesplace at either our ownconferenceor atthe European Multidisciplinary Cancer Congresseveryother year. Iwason thecommittee thatupdatedthe core curriculum, which can practically be doneover six years, althoughIcouldwriteacurriculumthatwould last a lifetime.

“I do not think it is important to push for more

CoverStory


“We also need to listen to those with poor chances,

to know how to provide good palliative care”

and interleukin in inhibiting tumour growth, whichled toNaredi being the principal investigator in sev-eral global phase III studies, although a lack of con-sistent interest fromdrug companies hasmeant thiswork has been very drawn out.

“As a young surgeon I was carrying outimmunotherapy as well as surgery on patients withmelanoma and renal cell carcinoma, and I kept upresearch in this area and incisplatin resistancewhenImoved toUmeå,” saysNaredi. “But someother sci-enceour surgical department is involved incan seemodd – for example with colleagues in themolecularpathogenesis centre we had a paper inCell in 2007

on the regulationof insulin inC. eleganswormcells.”Such work is way beyond the surgical oncology

curriculum, although it does specify that a traineeshould prepare at least one scientific paper, eitheroriginal research or a review ormeta-analysis.

WhenNarediwas in SanDiego he suddenly gotthreegreat joboffers: to takea seniorcolleague’splaceinGothenburg,movewith thecolleague toUmeå, orstay inSanDiego. “I chose to goback toGothenburgas an assistant professor, where I could continue tobenefit from great surgical leadership and also con-tinuemy research, and had some great years beforethemerger changespromptedmetomove toUmeå.”

Naredi began to specialise in liver surgery, a disci-plineonwhichhe isnowa leadingauthority, encouragedby Tore Schersten, a leading surgeon at Sahlgrenska.After focusing on conventional surgery for removingmetastases,where entire lobes areusually resected, hehas taken on amethod pioneered in France, in partic-ular by BernardNordlinger, in which smaller sectionsaround tumours are taken rather than whole lobes,whichNaredi calls the ‘Swiss cheese’method.

“As longas youkeep30%of the liver youcan takemanydifferent partswith thismethod, andweknownowwedonothave to leave largemargins around thetumours, only up to two millimeters, not the cen-timeter or sowe thoughtbefore. It’s not that patientsnecessarily dobetter thanwithwhole lobe resection,butwedon’t have to excludeasmanypeople, andwecan operate again and again on recurrences.”

Evenso, onlyone in fivepatients is currently suit-able for resection, oftenafter chemotherapy to shrinkmetastases commonly spread fromcolorectal cancer,butNaredi reports that, in recent years, five-year sur-vival rates for this grouphave advanced from40% to50% in centres such as Umeå, and even to 60% insomepatients, and such improvement is significantbecause late-stage colorectal cancer is common sothere is still a large population to target. “The ‘Swisscheese’method is the result ofunderstandingbiology,and is thewaywe should bedoing things in the21stcentury.Mysecond liver surgeonhere ishardlydoingany lobe resectionsnow, andwecanaimformoreeli-

“The ‘Swiss cheese’method is the result of understanding

biology, and is the way we should be doing things”

CoverStory


A CURRICULUM FOR SURGICAL ONCOLOGY

ESSO has put forward its core curriculum to try to tackle the ad hocway in whichsurgeons usually receive oncology training – few countries have formal nationaltraining programmes. Naredi and colleagues note that the European Board ofSurgical Qualification in surgical oncology, from the European Union of MedicalSpecialists, is probably the only formal qualification in Europe, but only five toten surgeons take this exam each year.The ESSO curriculum aims for an evidence-based approach rather than theexisting ‘common sense medicine’ now in place, and should join successfulcurricula from ESTRO (the radiation oncologists) and ESMO/ASCO, they say.It includes:� Recommendations that institutions should combine if they cannot offeraccess to facilities such as basic cancer biology facilities

� A minimum of three surgical oncologists who teach� A basic scientific curriculum that includes cancer biology, immunology andprinciples of treatment

� Evaluating and conducting clinical studies and understanding the ‘principlesand pitfalls of evidence-based medicine’

� Basic clinical requirements such as diagnosis and prognosis, implementa-tion of national guidelines, palliative surgery andmanagement of end-of-lifefeelings

� Cancer surgery itself – at least 120 cancer operations is recommended, atlast half done by the trainee

� Rotations in medical oncology and radiotherapy.The full curriculum, which Naredi says will be revisited soon to see if it needsupdating, was published in 2008 in Surgical Oncology (vol 17, pp 271–275).

we operate on, only 20% are alive after five years,which is only four or five out of 100overall.Wemustfind it earlier and we need biomarkers and bettertreatments–but inSwedenas elsewherepancreaticcancer gets very little funding and advocacy. I don’tcaremuchabout a2% increase in survivalwithanewdrug–ahugearea for research inmyview lies inearlydetection andbetter use of imaging technologies, aswell as effective treatments.”

Henotes though that therecouldbegenuineprac-tice-changing progress in one of his long-standinginterests, melanoma, where two targeted drugs haverecently beenapproved in theUS.Andon the surgicalside, he mentions strong results for sentinel nodetrials in bothbreast cancer (ArmandoGiuliano’sworkin theUS) andmelanoma (theMSLT-I/II trials).

After escaping fromGothenburg, where he wasfacedwith toomuchaimless administration,Narediwent toUmeå as an assistant professor, and then in2003 he became a full professor and chair of thedepartmentof surgery. “Itwasmore likeacountyhos-pitalwhen I arrived–now it is amuch larger univer-sity institutionandUmeå isa fast-growingcollegecity.I don’t regret themoveuphere for oneminute– andwehavehadnoproblemattractingyoungdoctors andresearchers here.”

Apart from helping to develop the academic hos-pital, a key advantage, he adds, has been the ability toshapeSweden’snorthern regional cancercentrearound

its major hospital, at Umeå. “Although our gov-ernmenthasmademistakes in forcinghospitalmergers – small places still need hospitals in

giblepatients in the future,maybe asmanyas one inthree. It’s like the way surgery for breast cancer hasmoved topartial removal –more is not always betterfor primary tumours and for metastases too. Ratherthan taking away more to feel safe, we must learnmoreabout thebiology.But likeany recent technique,we need to market it to get it into widespread prac-tice, just as the pharmaceutical companies markettheir drugs. It is more usual now, but I’m still givingtalks andwriting articles about it.”

Naredi adds that the liver is a challenging organwith a great deal of three-dimensional complexity,which iswhyhewasattracted to its surgery, and thereare other treatments such as perfusion and ablationto consider. There is also a lot to do in trial work onliver metastases and colorectal cancer. Testing theimpact of certain neoadjuvant (pre-surgery) treat-ments is one important area–hementions theEuro-peanEPOCstudy as one such trial.Other strategiesthat merit being tested in trials include removingmetastases before the primary tumour and afterchemotherapy, in the expectation that there is a bet-ter chance of eliminating cancer spread.

“Myothermain surgicalwork is in pancreatic can-cer, where we have improved greatly the number ofpatients we can operate on. Earlier, we were doing aWhippleprocedure ononly a fewpeople–nowwearedoingasmanyas40operationsayearatUmeå.Wehavebetter work-up with MRI and CT, and surgically wehave quality and skills we didn’t have 15 years ago.

“But the problem of course is that we aredetecting only one in five in time and, of those

CoverStory

9

On the helipad.Helicopter accessis essential for thisregional specialistcentre, which servesa huge territory,much of which iscovered in snowfor five monthsof the year

JOHANGUNSÉUS/SYNK

excels at. Cancer and death registry data are amongthe most complete anywhere, and hospitals can beinterrogated on say tumour-specific data that aremissing from thecancer registry butwhere cancer ispresent in thedeathdatabase.Butcancer registrydataare strong and are largely driven by surgeons, notesNaredi, who mentions the colorectal surgeon LarsPåhlman in Uppsala, featured in the December2004 issue of Cancer World. Påhlman has beenoutspoken about using an initiative he helped todevelop– theSwedish rectal cancer registry in1995(which now also covers colon) – to cut underper-formingcentres and surgeons for oneof the tumoursforwhich treatment quality remains highly variable.

Thishad remarkable results,with surgery feedbackalonepromotingbetter outcomes for rectal thancoloncancer, despite the latter benefiting from newchemotherapies. “I’mnot as forceful asLars – I thinksurgeons are competitive anyway and will take it onthemselves to either raise their game or stop if theyconsistently figureat thebottomofoutcomes,”hesays.

Palliative care is also getting its ownnational reg-istry, according toNaredi. “Therearemore funds fromgovernment going into this now than toother cancerregistries and it will help us measure where we canimprove factors involved in quality of life.”

Headds that the structure of the regional cancercentre allowspatients tobegenuinely representedatboard level. “Theycansay if theyareunhappywith thewaycare ismanaged. It’s not like sendingacomplaintletter– theyarepart of theprocess.”Thegeneralpop-ulation innorthernSweden is also involved inoneofthe country’s strongest biobanking projects. “Wehave100,000peoplewhogivebloodat theagesof40,50 and60, andamong themwehave identifiedpeo-plewho laterwerediagnosedwithpancreatic cancer.We have some unpublished work on possible pan-creatic cancerbiomarkers fromthisuniquebiobank.”

And in any case Sweden does best overall forcancer according to theEUROCARE-4 dataset, soit is no surprise that other countries are lookingto emulate best practice, say in tumour-specific

“A huge area for research in my view lies in early

detection and better use of imaging technologies”

CoverStory


my view – the criteria for the six regional centres,such as on education, structure, research, the cancerjourney andpatient participation, has promotedcom-petition on quality.We know that if we don’t improvequality we could lose patients to other regions.”

Naredi says there is nowmucheffort spent on try-ing to iron out the weak points in multiprofessionalworking. Rather than just a narrowmultidisciplinarytumour board, he says, there is wider participation atmeetings. “I may be the one who understands livermetastases and the best way to do surgery, but wejointly make the decision as to whether the patientshould have the treatment or not. We involve thepatient’s personal doctor, who often knows thembest, and we need people such as communitynurses to tell us if apatient isdepressedor inpain– the rest of our approach could be great, but ifwemiss factors like this thepersonhas a terriblequality of life.”

After some strugglewith the IT people,Naredi and colleagues also now haveaccess to high-quality videoconfer-encing facilities, vital tobringingmorepeopleat various locations intomeet-ings where they are all expected toplay a role in decisionmaking.

A big and stubborn challenge,as in most countries, is how toreduce the numbers of patientswhoarenot referredordiagnosedfast enough.ButNaredi feels thevariouselementsnowcombiningwill make an impact on theroughly 1 in 10 patients forwhom the process is currentlynot working, for instancebecause referrals arenotmadefor an expert appraisal for sur-gery where the patient wouldhave been eligible.

One element is use ofdata, which he says Sweden JO

HANGUNSÉUS/SYNK

superiorityofmedical oncologists, but Ihavenot seenany studies saying it is right.We are farmore experi-enced in intraperitoneal treatments, for example.But of course in large centres it will bemostlymed-ical oncologists administering systemic therapies,although inSwedentheyarealso radiotherapists.LiketheUKwehave theclinical oncologist speciality, andpractice does vary aroundEurope.”

Thatmay not endear him to ESMOcolleagues,andhe is concerned tooby the lackof trials currentlyrunbyanotherECCOmember, theEORTC. “ESSOdoesnotdo its ownstudies andweshouldbe runningthemthrough theEORTC,but at present it has veryfewof the trialsweare runningand itneedsa refresh,otherwise we may think about doing our own pan-European research.”

WithESSOpast-presidentCornelis vandeVeldenow president elect of ECCO, and Naredi also anECCOboardmember, surgical oncologydoes seemto be in the ascendancy in Europe. “I am also presi-dent of both the Swedish andNordic surgical soci-eties –but it ismost important now tobe involved atEuropean andglobal level ifwe are going to improveoncology, and ECCO is the right organisation forunity and strength.”

Naredi ismarried toSilvana, a fellowprofessor atthe university, and a neuro intensive care specialist,and they have two children, one in medical school.Like many Swedes, he’s big on outdoor pursuitssuch as cross-country skiing – just aswell as even inApril the river inUmeå is still frozen solid.

Hisplan is tocontinue todevelop the regional cen-tre atUmeå and especially the education and leader-ship side. “It takes years to get thequality youwant inoncology andwemust have continuity from the edu-cational systemand train youngdoctors tobe leaders,”he says. InEurope, he also has nodoubt therewill bea big expansion ofESSOas the organ-specialist soci-eties comeonboard, andhewill continue towork oneducational courses and quality, such as with theEuropean audit project. And quality also applies totime: anyone who books a meeting with Naredi thatdoesn’t have a key objective had better watch out.

networks such as breast, sarcoma and colorectal.Swedish surgeons, saysNaredi, havemorepolit-

ical clout via the Swedish Surgical Society thancounterparts in thecountry’smedical association, andhe says this more heavyweight presence applies atEuropean level too, andwill helpbringmoresurgeonsinto ESSO. “We have a stronger voice at Europeanlevel than the organ-specialist societies, which isone reason why I feel they will want to come underour umbrella.We are talking at present to leaders ofthe EuropeanAssociation of Urology and others inheadandneck, hepato-biliary andgynaecology soci-eties and soonabout theirmembers joiningESSO–most arenot currently individualmembers – so theycanhaveopenaccess toourconferences andcoursesas well as adding to our political voice.”

ESSO currently has about 2600 members andhas grown recently thanks to a policy of invitingnational society members to join. The last ESSOconference in2010 inBordeaux attractednearly 900people.Naredi adds that theEuropeanAssociationof Urology is likely to be the first European organ-specialist group to align itself with ESSO, whichcould give oncology ahugeboost on thecontinent byaddressing the poor treatments that occur in somecountries thanks to the ‘suboptimal’ oncologyapproach seen in certain specialities.

Like presidents of other European oncologyorganisations, Naredi is keen to get more youngpeople involved, and mentions women surgeonsespecially. “We spend a lot of our resources now oneducational events and conferences, also in col-laboration with other societies. Good examplesare the Flims fellowship courses and the ESO–ESSO masterclasses, which are not necessarilypitched at elementary levels but at experienced sur-geons too.”A young surgeons and alumni clubwaslaunched at the 2010 ESSO conference.

And like theheadsof other societies, he is robustin promoting the all-round qualities of members. “Ihavenoproblemwith surgeonsdoing systemic ther-apy – as many countries, like Sweden, do not haveregularmedicaloncologists.ESMOlikes to talkof the

CoverStory


“I think surgeons are competitive anyway and will take

it on themselves to either raise their game or stop”

e-GrandRound


Decision making inthe treatment of gliomas

Treatmentmodalities formalignant gliomas have not changed greatly in recent years, but we are

learning much more about how to tailor treatments to patients. This overview looks at the role

of age, tumour size, performance status and various predictive and prognostic biomarkers in

guiding treatment in newly diagnosed and recurrent disease.

Thetreatmentmodalities formalig-nant glioma have not changed agreat deal over the past few years,

and remain: surgery, radiotherapy andchemotherapy. Surgery is the first stepandbackbone in the treatment of glioma.Complete resection, debulking or biopsyallows for precise histopathological andmolecularcharacterisation,which isessen-tial if we are to tailor and personalise thetherapy. Radiotherapy has been used forthirty years, andwe know that it prolongssurvivalwhencomparedwithnitrosourea-based chemotherapy or best supportivecare. Chemotherapy used to be the ‘newkid on the block’, but is now the standardof care in newly diagnosed glioblastoma,concomitantly with radiation. Its value intheupfront treatment of other subtypes ismorecontroversial and thedata arenot yetconclusive.Wecommonlyusechemother-apy (nitrosoureas and temozolomide) totreat recurrent gliomaandas second- andthird-line treatment.

DECISION MAKING INFIRST-LINE TREATMENTDecisions in first-line treatment are notonly about how to treat but also who to

The European School of Oncology pres-

ents weekly e-grandrounds which offer

participants the opportunity to discuss a

range of cutting-edge issues, from con-

troversial areas and the latest scientific

developments to challenging clinical

cases, with leading European experts in

the field. One of these is selected for pub-

lication in each issue of Cancer World.

In this issue, Roger Stupp, from the

University Hospital of Lausanne, Switzer-

land, provides an update on factors that

can be used in decision-making, focus-

ing on practical aspects and everyday

questions in treating patients with malig-

nant glioma.

Olavo Feher, of the Instituto do Cancer

do Estado de São Paolo, São Paolo, in

Brazil, poses questions sent in by par-

ticipants during the e-grandround live

presentation.

It is summarised here by Susan Mayor.

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

pagina_13-19_grandround.OK.qxp:CancerWorld Template 14/6/11 15:39 Page 13

treat and when to treat. Prognostic andpredictivemarkers areused to guide treat-ment to ensure we get the most out of it.These factors includeperformance status,age, tumour size and location, andresectability. There arenot a lot of data onresectability, but we know resectedpatients dobetter. There are alsomolecu-larmarkers suchasMGMT,LOH1p/19q(t 1:19), and IDH1mutation.But towhatextent do these parameters help us ineveryday decisions inmanaging glioma ?

Performance statusBoth WHO and Karnofsky’s perform-ance status scales are commonly used.The WHO scale has largely replacedKarnofsky in oncologybecause it ismorereproducible; in neuro-oncology bothscales remain inuse. Inpractical terms itdoes not matter which one uses. Mostbenefit fromtreatmentcanbeachieved inpatients in reasonably good shape, whoare alert and largely independent, andare able to come to the outpatient clinic.

AgeWhen we started the pivotal trial withtemozolomideandradiationmore than10years ago, patients over the age of 70years were not considered for combinedmodality therapy on the grounds thattheir poorer prognosis and short survivalwould not justify a lengthy course oftreatment. But a recent trial conductedby the French neuro-oncology groupANOCEF(NEJM15:1527–1535) lookedat the value of radiation versus supportivecare in elderly patients agedover70years.The trialwasclosedearlybecause radiationtherapy improved survival over supportivecare in patients even though they wereconsidered to have poor prognosis (seefigure above). A second, Canadian, ran-domised trial showed thathypo-fraction-ated radiation gives equivalent results tostandard fractionated radiation in theeld-erly (see figure). The findings mean wecan reduceexposure to radiation and the

number of hospital visits for therapy inelderly patients.

An analysis of subgroups from theEORTC/NCIC pivotal trial comparingtemozolomide and radiation with radia-tion alone in patients aged 60–65 yearsand those aged 65–70 years shows ben-efits in both age groups in favour of com-bined treatment. The hazard ratio in the65- to 70-years age group was lessfavourable than in the younger group(0.78 vs 0.64, compared to 0.63 in thewhole trial population). These results donot suggest there is no value in com-bined modality treatment in the moreelderly group, butmay indicate the needto select patientswhowill benefit from amore aggressive approach.

The interest in elderly patients is illus-trated by two randomised trials, NOA-08and theNordic trial, reportedatASCOlastyear.TheNOA-08trial comparedan inten-sive temozolomide regimen(weekon/weekoff)with radiation inpatients agedover 65(median age 72 years). The objective,which was to show that temozolomide isnot inferior to radiation, was not attained,and toxicity with the dose-dense temo-zolomide regimenwas higher than antici-pated.With initial radiationaloneamediansurvival of 10monthswasachieved,which


e-GrandRound

was reasonably good in an elderly popula-tion compared to other trials (Wick et al.ASCO2010, abstract 2001).

The Nordic trial compared two radio-therapy regimenswith temozolomide (5/28days) in patients aged over 60 (median 70years). Hypofractionated radiation andtemozolomide seemed to be somewhatequivalent (Malmstromet al.ASCO2010abstract 2002). The verdict is still out, butthese studies show that if you select theright patients, radiation should be given.However, they also show that chemother-apy alone may be an alternative for somepatients, suchas those living far away fromthehospital and thosewhoarenot inacon-dition to travel. The ongoing NCIC/EORTC intergroup randomised trial islooking at combined modality treatment,and as this approach worked in youngerpatients, I think it should alsowork in eld-erly patients, if selected correctly.

THE ROLE OF SURGERYSeveral trials have shown that patientswho have complete tumour resection dobetter than patients who only have abiopsy. For example, the EORTC trialdemonstrated thatpatientswhohadcom-plete resection had longer survival thanthoseundergoing only partial resectionor

RADIOTHERAPY: ELDERLY PATIENTS DO BENEFIT

Trials looking at more elderly patients have shown that this group (>70 years) does benefit fromradiotherapy and that elderly patients (>60 years) can gain equivalent benefit from a lower overalldose given in fewer sessionsSource: Keime-Guibert for ANOCEF: NEJM (2007) 15:1527–1535

Roa et al. JCO (2004) 22:1583-1588. Reprinted with permission. © ASCO. All rights reserved


tions a contemporary standard does notexist as it has never been investigated.

What ofmolecularmarkers?MGMT(O-6-methylguanine-DNAmethyltrans-ferase) predicts outcome– at least that’sthehypothesis. It is aDNArepair proteinthat removes themethyl group that hadbeen transferred from temozolomideonto guanine If the gene promoter ismethylated,which is an epigenetic phe-

biopsy (seebelow,upper figure).AGermantrial aimed at increasing the completeresection ratebyusing fluorescent lights inthe operating theatre. Results showedimproved progression-free survival aftercomplete resection, and higher rates ofcomplete resection using this approach(see lower figure). This trial did not showlonger overall survival, but at least it pro-vided further evidence for the role of sur-gery. However, the extent of surgeryneeds tobebalancedagainst the risks.

Question: Considering the data inthe elderly – the results of theEORTC/NCICand theGermanandNordic trials – what is your currentapproach in elderly fit patients withgoodperformance status today,withoutresults from the randomised trials?Answer: If I have a fit elderly patient,I would give them combinedmodalitytreatment, possibly temozolomidechemotherapy combined withhypofractionated radiation. I wouldconsider exclusive temozolomidechemotherapy in a patient with amethylated tumour requiring a largeradiation field, particulary inanelderlyand cognitively frail patient. In short, Iwould go with combined modalitytreatment outside a clinical trial if I donot have a clinical trial available.Question: Would you be afraid tocombine temozolomidewithhypofrac-tionated radiation?Answer:No.

MOLECULAR MARKERSSo far,wehave seen that clinical fac-tors can give a gut feeling about howto treat a patient, but we have fewobjective factors to use in decidingwhoweshould treat andhow. I thinkexperience has a role here, and myanswer to the last question illustratesthat we sometimes deviate from theestablished standard of care for spe-cific reasons, while in other situa-

nomenon affecting gene regulation, thegene is silenced. In other words, thegene is not expressed and the cell doesnot have the toolbox to repair the DNAdamage. If this hypothesis were true,patientswith amethylatedMGMT pro-moter would benefit most from temo-zolomide chemotherapy.

Studies show thatMGMT status pre-dicts benefit from combined treatment.

Patients with an MGMT methy-lated promoter, who are missingthe tools to repair DNA damage,show most of the benefit of theaddition of temozolomide, whilein patients with non-methylatedMGMT, temozolomide seems tohaveno, ormarginal, effect on out-come (see figure overleaf).

This initial retrospective obser-vation has recently been prospec-tively validated (RTOG0525;Gilbert et al.ASCO2011, abstract2006).Wecanconclude that thereare two populations of tumours:those with a methylated promoterand others with a non-methylatedpromoter, and theymaymerit a dif-ferent treatment strategy. IntumourswithmethylatedMGMT,I think that temozolomideplus radi-ation should be the backbone ofanyproposed treatment, andshouldalsobe thebackboneof anyclinicaltrial investigating the addition ofnewdrugs.For tumourswithanon-methylated MGMT, we shouldthink of options other than temo-zolomide, becausedrugswith adif-ferent mechanism of action areneeded to treat thesepatients opti-mally. The difficulty is that we donot yet have a better alternative forthese patients; and even the besttest is never 100%predictive.Untilbetter treatments are established,even patients with an unmethy-latedMGMTpromoterwill receivetemozolomide and radiotherapy.

e-GrandRound


SURVIVAL AND THE EXTENT OF RESECTION

Complete resection was associated with better survival in theEORTC trial. In the German trial, fluorescence-guided surgeryled to more complete resections, but complete resection wasassociated only with delaying disease progression and notwith improved survivalSource: (top) Adapted from R Stupp et al. Lancet Oncol (2009)

10:459–466 (figure unpublished)

(bottom) Reprinted from Stummer et al. Lancet Oncol (2006)

7:392–401 © 2006, with permission from Elsevier


THE CURRENT STANDARD OF CARETemozolomide is given sevendays aweek,including weekends (tumours do notobserveSundays),while radiation is givenfive days a week (see figure below). Withconcurrent chemoradiation therapy, dailyantiemeticprophylaxis is oftennotneeded.We use a 5-HT3 antagonist only for thefirst fewdaysof treatment (toavoidconstipation associated with pro-longed administration), beforemoving to a simple antiemetic likemetoclopramide or domperidon.

What about anti-epileptics?These are only indicated inpatientswith ahistory of seizuresand not as standard prophylaxis.It is also important to tapersteroids. All too often we seepatientswhobecomeweaker, notdue to tumour progression butbecause of steroid myopathy.

THE PROGNOSTICVALUE OF MRIIn clinical trials, MRI is usuallyperformed four weeks afterchemoradiation.However, resultsat this early timepoint aredifficult

to interpret and so thisMRImaynothavemuch value outside trials. The difficultissue ispseudoprogressionafter combinedtemozolomideand radiation therapy.Afterchemoradiation, andafter radiationalone,images with increased contrast enhance-mentmay falsely suggest tumourprogres-sion, while these changes of the

blood-brain barrier reflect inflammationdue to tumourbreakdownand repair, andwill normalise over the followingmonths.

The figure opposite showsMRI scansfor a patient with glioblastoma treatedwith temozolomide/radiotherapy in May2008. TheMRI forAugust 2008 shows aclear increase, with contrast enhance-ment, and some oedema, butwe thoughtthat it could be pseudoprogression. Wecontinued, but anMRI inOctober 2008,after a longer period when we should beable todistinguishpseudoprogression fromtrueprogression, showeda further increasein tumour size, with more oedema. Thepatient was taken into surgery but therewere no tumour cells to be seen, onlynecrosis. It is important to keep the phe-nomenon of pseudoprogression in mind,and not to take patients off treatment tooearly, particularly if they are clinicallywell.

MGMT may help in this situation.Brandes and colleagues (JCO 26:2192–2197) looked at patients who progressedafter chemoradiation therapy but con-tinued temozolomide further. Resultsshowed that some patients continued toprogresswhile others improved onMRI.

Two-thirds of patients who sub-sequently improvedhad tumourswith a methylated MGMT pro-moter, suggesting that pseudo-progression is more frequent inMGMTmethylated tumours. Inother words, pseudoprogressionmay be an expression ofincreased tumour breakdownrather than progression.

TREATMENT OF GRADE III(ANAPLASTIC) GLIOMAHistorically, the standardofcare isradiation therapy and I think it isimportant to recognise thatcertaintreatments used in the past maynot havebeen evaluatedwith thesame rigour as today. Data nowshow that we could start withchemotherapy first and then use


e-GrandRound

STANDARD OF CARE IN NEWLY DIAGNOSED GBM

Six weeks of concomitant temozolomide (seven days a week,max 49 days continuously) and radiotherapy (five days a week)followed by intermittent adjuvant temozolomide is the currentstandard of care for all patients with glioblastoma multiforme,with supportive care to combat symptoms and side-effects andcheck-ups every two to three months (X = optional)

If these results are confirmed, an alternative to temozolomide should be used in patients with non-methylated MGMTSource: Reprinted from R Stupp et al. Lancet Oncol (2009) 10:459–466 © 2009, with permission from Elsevier

MGMT PREDICTS BENEFIT FROM COMBINED TREATMENT


glioma, including oligos, as aneoadjuvant(RTOG trial, JCO 24:2707–2714) or anadjuvant (EORTC trial, JCO 24:2715–2722). No benefit from the addition ofchemotherapy could be demonstrated,even for the subset of themost chemosen-sitive oligos.

Individual treatment strategies shouldbe based on tumour size, patient age andhow aggressive a treatment one considersto be indicated. Primary chemotherapymay be an option for some patients withlarge tumours,but radiation therapymaybethebestchoice for small tumours;however,datadonot support theunconditionalpref-erence for chemotherapy. More will beknown when the ongoing internationaltrials (CATNON coordinated by theEORTC, and CODEL, coordinated byNCCTG, and inEurope by theEORTC)have completed accrual andmatured.

IDH MUTATIONSIDH (isocitratedehydrogenase)mutationswere recognisedacoupleof years agoasanimportant prognostic factor for outcome.Patients with an IDH mutation, whichusually occurs early in gliomagenesis, haveamore favourable outcome than patientswithout this mutation. IDH mutationoccurs in70%ormorepatientswith gradeII and III glioma (NEJM 360:765–773;JCO 27:4150–4154). It gives us a way toidentifywhether apatienthas a secondaryglioblastoma. I would guess that manylong-term survivors of recurrent glioblas-toma, who do well with several lines oftreatment, have IDHmutations.

TREATMENT OPTIONS FORRECURRENT GLIOMAWhile we have good data and prospectivetrials for the management of malignantglioma in theupfront setting,we lack largeand solid trials in the recurrent setting.Decisions are individual, and depend onpatients’ and physicians’ preferences, andavailability of modalities and healthcareresources.Repeat surgerymaybeanoption

radiation at progression. This could beconsidered for large tumours requiringextensive radiation therapy fields, or foroli-gos,whichhaveamore favourablenaturalhistory and where you may want to delayradiation therapy.There arenodata yet forcombinedmodality treatment, but I knowthat this approach is used frequently.

A carefully conducted German triallooked at the sequenceof treatment (JCO27:5874–5880). It randomised patientsbetween radiation first and chemotherapyat progression, or chemotherapy first andradiation at progression. The primaryendpoint was progression the secondtime. Results showed no difference inoverall outcome whether patients wereinitially treated with radiotherapy, fol-lowed by chemotherapy at first progres-sion, or the inverse sequence. However,the use of concomitant chemoradiother-apy was not investigated (this is the sub-ject of the ongoing EORTC-IntergroupCATNONtrial). Based on these results,wemay individually adapt the treatmentstrategy for each patient. Patients with asmall tumourmaybest be treatedwith sixweeks of radiation rather than a year-long chemotherapy regimen, while inlarger tumours a primary treatment withchemotherapy may be considered.

MGMT in this trial was again a strongprognostic marker; however, while one

might expect that tumours with a methy-latedMGMTpromoterwouldbenefitmostfromanapproachstartingwithchemother-apy, time to first tumour progression wassimilar in these patients regardless ofwhether theywere treatedwith radiother-apy first or chemotherapy first. The valueof MGMT in grade III tumours is prog-nostic rather than predictive and does notreadily help us chose whether to givechemotherapy or radiation therapy.

CHEMOTHERAPY FOR NEWLYDIAGNOSED ANAPLASTIC OLIGOSI deliberately use a term here that groupsoligodendroglioma and mixed oligoastro-cytoma together as ‘oligos’, because defi-nition, reproducibility and trial results arenot entirely consistent.As a general rule,pure oligodendroglioma,with a transloca-tion of the gene1;19 (LOH1p/19q) havea distinct and prolonged natural history,and better responsiveness to bothchemotherapy and radiotherapy. Theabove-mentioned German trial showedthatpatientswhohaveanoligocomponentclearly do better in terms of time to firstprogression than patients who haveanaplastic astrocytoma (JCO 27:5874–5880). Two randomised internationaltrials evaluated the addition of PCVchemotherapy (procarbazine, lomustine(CCNU) and vincristine) in anaplastic

e-GrandRound


PSEUDOPROGRESSION ON MRI AFTER COMBINED TREATMENT

Follow-up MRI scans inpatients treated withconcomitant temozolomideand radiotherapy can bedeceptive, and care must betaken not to assume patientsare progressing when in factthey are respondingSource: MRI scans courtesy of

Roger Stupp, University Hospital

of Lausanne, Switzerland


a practice treatment trial, but it tells usthat patients who have been on temo-zolomide for a long while and progressmay not benefit from re-treatment withtemozolomide.

A British randomised trial looked atPCV versus temozolomide in recurrentchemonaïve (!) glioma patients not givenchemotherapy during first-line radiationtherapy. Results suggested that temo-zolomidewas equivalent to PCVbut notnecessarily superior, although toxicitywas lower (JCO 28:4601–4608). A sec-ond randomisation between two sched-ules of temozolomide – five days of 28(standard administration schedule) and adose-dense schedule for 21 (of 28) daysshowed slightly better outcomeswith thefive-day schedule. Similarly, the recentlyreportedRTOG05025/EORTC/NCCTG-Intergroup trial failed to demonstratesuperiority of a dose-dense temozolo-mide schedule in newly diagnosedglioblastoma (Gilbert et al.,ASCO2011,abstract 2006).Wemayhave been overlyoptimistic about alternative temozolo-mide schedules.

What other alternatives do wehave?A trial comparing enzastau-rinwith lomustine (JCO28:1168–1174) provides data on lomus-tine in patients with recurrentdiseasewhohave failed on temo-zolomide and radiation. Resultsshow overall survival of sevenmonths, andprogression-free sur-vival of 19% at six months withlomustine – close to the 20%benchmark we set at the timewith temozolomide. So, lomus-tinemaybe abetter drug thanwethought, often well-tolerated butwith a substantial incidence ofprofound myelosuppression inpreviously treated patients.

VEGF inhibition forrecurrent gliomaUse of agents targeting VEGF or

VEGFR is the most recent strategy to belooked at. Data with two drugs – beva-cizumab (Avastin) and cediranib(AZD2171) – have initially been particu-larly encouraging, giving us the kind ofMRI images that getusexcited!The figureopposite shows scans for a patient beforeand after treatment with bevacizumaband irinotecan (left-hand scans), and thetumour has almost vanished.

The right-handscans in thesamefigureinclude similar findings from a trial byBatchelor and colleagues (Cancer Cell11:83–95) for a patient treated with theVEGFR inhibitor cediranib. The scansshow that the contrast enhancement dis-appears very rapidly. MRI scans the daybefore cediranib administration, the dayafter treatment, andafter fourweeks, showthat the tumour had disappeared, or hadstarted disappearing, 24 hours after givingcediranib. This is almost too good a result.It suggests that what we see with thisVEGFRinhibitor is thenormalisationof thevascular permeability and of the vascula-ture, butnotnecessarily a trueanti-tumoureffect, so some of this is a radiological


e-GrandRound

in large tumours exerting a masseffect. A randomised trial wouldbe needed to assess its true value,but I do not think that this is prac-tically feasible, as it is hard to ran-domise patients between invasivesurgery and a chemotherapy. Car-mustine wafers (Gliadel) wereapproved for recurrent gliomaundergoing repeat surgery, but theimpact and use in daily practiceremains limited. Approvedchemotherapies include temo-zolomide, carmustine, lomustineand other nitrosoureas. Irinotecan(CPT11), cisplatin, carboplatinandetoposide are occasionally used,but not formally registered. Beva-cizumab was recently approved inthe US, but it was rejected by theEuropean Medicines Agency(EMA). Re-irradiation is gainingin popularity, although it is not yet vali-dated in prospective trials.

Re-introductionof temozolomide, andalternative anddose-dense temozolomideschedules, aregaining inpopularity.Whentemozolomide was approved, mostpatients were chemonaïve. They now allhave temozolomide up front, so does itmake sense to re-expose them?

A Canadian study re-challengedpatients with progressive disease withtemozolomide. It included patients whoprogressed in the early phase of adjuvanttreatment, and thencontinued temozolo-mideonadifferentmetronomic schedule(chronic non-interrupted temozolomideadministration at 50 mg/m2). Resultsshowedalmost 30%progression-free sur-vival at six months. In patients who hadbeen on temozolomide formore than thestandard six cycles, only 10% seemed togain benefit from staying on temozolo-mide. Patients who had been off treat-ment and then started again showed a30%progression-free survival at sixmonths(see figure above).

Thiswas not a randomised trial, it was

TMZ RECHALLENGE IN RECURRENT GLIOMA

Studies exploring response to changing the dose/schedule,extending adjuvant treatment beyond the standard six cycles, orrestarting temozolomide have found varying degrees of benefitSource: Adapted from J Perry et al (for the Canadian Brain Tumor

Consortium). JCO (2010) 28:2051-2057


phenomenon of pseudoresponse ratherthana true response.Nevertheless, regres-sion of peritumoural oedema is real andoftenassociatedwitha temporary improve-ment in patients well-being.

We only have limited data with theseagents in brain tumours. Although beva-cizumab has been approved by the FDA,this is aconditional approval on thebasis ofphase II data.A randomised phase II trialin which patients were randomised tobevacizumab (with irinotecan added onprogression)or tobevacizumabplus irinote-can showed that the majority of patientscould be spared from using steroids bytreatmentwithbevacizumab,which is lesstoxic than steroids.Tumour size–asmeas-uredbycontrast enhancement–decreasedin themajority of patients.Results showedanoverall survival of aroundninemonths,similar in both arms. Although survivalappears slightly better thanwithhistoricalcontrols, trialswith cytotoxic agents alonehave shown median survival durations ofseven or eight months. So beva-cizumabmayhave somevalue,butlargelybasedona steroid-like anti-inflammatory effect, while a clearantitumour effect remains to bedemonstrated. It may improvequality of life in selected patients,without necessarily prolongingsurvival.

For cediranib, a pan-VEGFRtyrosine kinase inhibitor, a properrandomised phase III trial wasconducted. The results of theREGAL study were presentedrecently at the ESMO meeting.This trial randomised patients tocediranib alone, cediranib andlomustine, or placeboplus lomus-tine. Results showed an overallsurvival of aroundninemonths inthe two lomustine-containingarms, and eight months in thecediranibalonearm(ESMO2010abstract LBA7). Disappointingly,no benefit was seen for the com-

binationof cediranib and lomustine. Sim-ilarly to thebevacizumab, imaging showedimprovement and there was less steroiduse inpatientsoncediranib;however, itdidnot translate into improved survival.Over-all, a VEGF-inhibiting strategymay be ofsomevalue;however, the targetpopulation(e.g. large tumourswith important peritu-moural oedemaandmasseffect), theopti-mal dose and frequency of dosing, andcombinationwithcytotoxicchemotherapyremain to be determined.

SUMMARYIn conclusion, we have a few clinicalparametersonwhich tomakedecisionsonwhen to treat andwhen towithhold treat-ment in patients withmalignant gliomas.The nihilism we have had until recently,especially inelderlypatients,maybeques-tioned, and some elderly patients maybenefit from active treatment. Completetumour resection, if feasible, is associatedwith improved outcome.

In terms of molecular markers, MGMTmethylation status predicts benefit fromalkylating agent chemotherapy inglioblas-toma and is prognostic in anaplasticglioma.LOH1p/19qcharacterises a sub-groupof patients and tumourswith apro-tracted natural history. IDH mutationsoccur early in gliomagenesis andare char-acteristic for transformed lower-gradeglioma, allowing us to identify secondarygliomas that have a different geneticmakeup. They may indicate a morefavourable prognosis, and tumours thataremore likely to respond to treatment.

Question:Wehave seen overall survival ofglioblastomas convergingat around21or22months inacoupleof latephase II trials– theNABTT trials and the UCLA trial withbevacizumab and irinotecan, and temo-zolomide first-line trialswith glioblastomas.Do you think the survival in glioblastomas isshifting to the 20months hallmark?Answer: I think it is shifting, because

patients get better care.A lot of thebenefits are due to better support-ive care and the fact thatwedonotgive up, and we do repeat surgeryand multiple lines of chemother-apy. It is a conglomerate of manyinterventions rather than just oneintervention.There is always someselection bias in clinical trials.Wetend to include the better patients,because the ones with the worstprognosismay not evenmake it toa trial. A number of trials haveshown a good number of patientsprogress even after chemoradia-tion, andnevermake it to any fur-ther lines of treatment. Thisunderlines the need for ran-domised trials, because we can-not draw conclusions based onhistorical controls. This shift toimproved survivalmeanswe needcontemporary controls to helpguide decisions.The answer is ran-domised clinical trials.

e-GrandRound


Images showing recurrent gliomas before and after treatment withbevacizumab (left) and cedirinab (right) show dramatic tumourshrinkage, but this may not be true responseSource: (left) JJ Vredenburgh et al. Clin Cancer Res (2007) 13:1253–59,

adapted and reprinted by permission from theAACR (right) Batchelor et

al. Cancer Cell (2007) 11:83–95, reprinted with permission from Elsevier

VEGF INHIBITION


patient population is small andnetwork-ing is poor, recruitment to the trial canbea very slow process.A number bad things flow from this:� potentially beneficial drugs are slow

to reachpatientswho are running outof options

� the cost of the process pushes up theprice of thedrug,which could restrictpatient access

� the time and money used in gettingstatistical answers could be used forother research

� the longer it takes to answer theques-tions posed by the RCT, the greaterthe risk that the question loses rele-vance, as the standardof carechangesor greater insights are gained into theway the disease works.

There are also ethical issues about trials

Beyond survival – what should newcancer drugs have to prove and how?

� Anna Wagstaff

Demonstrating in a randomisedcontrolled trial (RCT) that anew treatment keeps patients

alive for longer has long been seen as thegold standard evidence for new cancerdrugs. That doesn’t mean it has beenuncontroversial – far from it.

On the plus side, this gold standardanswers the key question for patientsand doctors: “What is likely to keep mealive for longest?” It also gives payers astrong evidence base to assess the valueof the drug. Using ‘surrogates’ for sur-vival, such as response rate – significanttumour shrinkage – or progression-freesurvival (PFS) – how long the therapyholds the cancer at bay – are seen as farweaker measures. The notorious abilityof cancer cells to find alternative path-ways means that early indications of

response are often not sustained. Surro-gates are alsoharder tomeasure than sur-vival, where the endpoint is the finalityof death. It can be difficult to interpretwhat is happening to a tumour even onMRI, giving rise to the phenomenon ofpseudoprogressions and pseudore-sponses (see e-grandround in this issuefor a spectacular example in gliomas).Measurement of progression is also opento variations and depends heavily onhow often the patient is evaluated.

However, there is adownside tousingsurvival as thekey indicator. Itmeans thatresearchers must continue a trial untilenoughpatients have died to showa sta-tistical difference in survival. This canbea long and expensive process, especiallywhere thebenefit is incremental –which,sadly, is often the case. If the relevant


CuttingEdge

Randomised controlled trials have been the gold standard for testing new drugs, and survival is

the standard bywhich they succeed or fail.As our understanding of cancer increases, therapies

becomemore numerous and more complex, and patients live longer, is this still the way to get

the best drugs into use most quickly? If not, is there a credible alternative?

CuttingEdge


Better care.Formany cancers, survivaltimes are creeping up as the result ofimprovement in care, including support-ive care, and greater specialisation andmultidisciplinary working. This is goodnews forpatients, butmeans that survivalendpoints take ever longer to reach.

More therapies.By theendof their livesmany cancer patients will have beentreatedwith four, five or six different ‘lines’of therapy, moving on to something neweach time the previous one ceased to beeffective or the side-effects proved tootroublesome or a new more promisingdrug made it to market. For each drug

you canmeasure response rate and PFS,butdeathhappensonlyonce:howcanyoutell what contribution each drugmade tooverall survival?

Better organised patients. As patientgroups have becomemore organised andvocal, it is becoming increasingly difficultto justifyor recruit to trials thatdonotallowcontrol patients to cross to the experi-mental arm once that arm has shown itdoes better on the PFS measure. Thewhole purpose of allowing crossover is tominimise the survival gap between thetwoarms,making it hard if not impossibleto demonstrate superior overall survival.

that require patients in the control armto die early to prove the superiority of theexperimental arm,when their livesmighthave been extended had they beenallowed to cross over to the experimen-tal therapy as soon as it becameclear thatthey were showing a poorer responserate or were progressing earlier.

TIMES ARE CHANGINGThere are genuine dilemmas here, withno easy answers. But a number of trendsin cancer research and cancer care arenow changing the terms of this debate: IL

LUSTRATION:FREDVANDEELEN,W

WW.ORGANISART.CO.UK

surrogate in this setting.A consequenceof this, he argues, is that while womenare typically treated with four, five oreven more lines of treatment, after thefirst two lines, doctors and patients havelittle evidence for survival on which tobase further choices.

“I think there is a correlationbetweenthe promotion and acceptance of PFSdata, because they are the primary datathat are being generated. But from theresearch we did in preparation for ourlaunch, themessage that this is the rightendpoint to be looking at, rather thanoverall survival, seems to have beenaccepted by oncologists as well.”

When Eisai presented physicianswith a single page showing the profile ofHalaven, says Bose, their attention wasimmediately drawn to thePFSdata – thesecondary endpoint of the study. Theywere very interested in the overall sur-vival data when they saw it, he says, butthey didn’t actually look at it until it wasexplicitly pointed out. “It’s stark how theenvironment has now evolved in pre-scriber land that PFS is a valid surrogate,and they are quite convinced that it is afair and a strong endpoint, even whenthey are presented with overall survivalas primary endpoint.”

Bose hopes that Eisai’s success inshowing overall survival benefitwill chal-lengewhat he sees as a defeatist accept-ance that meeting an overall survivalendpoint is an unrealistic expectation inlate stages of disease.

He does accept, however, that thereare many situations where proving sur-vival benefit may not be possible, and itwas a delicate balancing act even in theEMBRACE trial, which demonstratedan extra two months of life for womenwithmetastatic breast cancer whowereput on Halaven as a third-line or latertreatment. Tomake the trialmore palat-able to potential participants, Eisaiallowedalmost unconstrained ‘treatmentof physician’s choice’ (TPC) as the con-

natives to lengthy RCTs be acceptable?As cancer care and drug develop-

ment move forward, should proof ofoverall survival benefit as shown in arandomised controlled trial still be thegold standard for new therapies? Howcan drug developers provide patients,doctors, regulators, trial participants andpayers with the evidence they are look-ing for?Cancer World put the questionto some of them.

COMMITTED TO SHOWINGSURVIVAL BENEFITUdayBose is EuropeanHead ofOncol-ogy at Eisai, a Japanese pharmaceuticalcompany that recently entered the can-cer fieldwithHalaven [eribulin], a cyto-toxic, thatwas approvedby theEMAthisMarch for use in advancedbreast cancer.

Bose questions how far overall sur-vival is really seen as the gold standard,citing a study that showednomore than15out of 76phase III studies inmetasta-tic breast cancer published between1998 and 2007 had overall survival astheir primary endpoint, and met thatendpoint (JCO 28:1958–1962). PFShas become an increasingly common


CuttingEdge

Shrinking patient populations.Ran-domised controlled trials are all aboutnumbers and statistical proof. Asresearchers succeed in differentiatingthe disease into hundreds of biologicallydistinct entities, the number of patientsappropriate for each trial shrinks, mak-ing RCTs less of a practical option.

Decreasing toxicity. While cardiactoxicity, for instance, can still be a seriousissue with some biologics, in general,newer therapies, including vaccines,have side-effects that are less threaten-ing and less debilitating than traditionalcytotoxics. The need to prove beyonddoubt the survival benefit of a newdrugbecomes less pressingwhere thepatientshave less to lose.

Intelligent design. RCTs can giveanswers to specific questions even if youhavenounderstanding ofwhat is drivingthe disease, or how the drug works or inwhom. Now that we understand moreabout the disease and drug developersinvest heavily in extensive preclinicaland early clinical studies to build up aclearer picture of their drug,might alter-

RCTs: GOLD STANDARD OR BLUNT INSTRUMENT?

Randomised clinical trials are used to subject hypotheses such as “patientswill live longer ondrug A than drug B” to a statistical test. They need to recruit enough patients to show, with ahigh degree of certainty, that the survival difference between the trial arms really does reflectsuperiority of the treatment rather thanhaving comeabout by chance. Thismeasureof certaintyis represented in the all-important P-value; P<0.001 being a way of saying that there is a onein a thousand chance that the survival difference shown in the trial, or an even higher differ-ence, does not represent a real and replicable difference. The smaller the difference betweenthe two arms, themore patients must be recruited to reach statistical significance.Bayesianmethodologies, in contrast, make use of all relevant knowledge gained through themultiple studies done during the process of drug development – on the role of the target,the ability of the drug to hit the target, the effect of hitting the target, perhaps the effect ofadding a second drug, the dose levels, predictive biomarkers and so on – and can incorpo-rate themmathematically as ‘priors’ into amodel that presents the strength of evidence interms of ‘credible intervals’, which are equivalent to themore familiar ‘confidence intervals’.Cancer World will look at Bayesian trial methodologies in greater depth in a future issue.

trol arm– includingbest supportive care.As it happens, saysBose, no patients

chose best supportive care – somethingthat pleased the patient representatives,he says, “because it challenges the per-ception that once a woman has gonethrough first-/second-line treatment theygive up and they don’twant anything else.”

Halaven is currentlyin a head-to-head studyagainst capecitabine,in an earlier line of dis-ease, after an anthracyclineand a taxane. This timeEisai has chosento use progression-free survival as a co-primary endpoint with overall survival.The company is clear, however, thatwhatever happens, the trial will con-tinue until there are sufficient overallsurvival ‘events’ (i.e. enough deaths) todemonstrate a significant difference insurvival. They will not do what somanyphase III trials do, andpublish an interimreport when the number of PFS ‘events’(i.e. progressions) has reached a pointwhere they are likely to show a signifi-cant difference between the two arms,and then either stop recruiting or allowpatients on the control arm to switchover to the experimental treatment.

“If we were to go out with our PFSendpoint, then in our conversationswithpayers, they may say, ‘But you compro-mised the study.You had a survival end-point, why didn’t you stick with it?’”

Bose has seen exactly this happenwith some other cancer drugs, and hedoesn’t want to repeat the mistake.There’s no great ethical achievement,he points out, in stopping a trial early orallowing patients to cross over to theexperimental arm on the basis of prom-isingPFSdata, if the consequence is thatpayers then refuse to reimburse the drugon the grounds of insufficient evidenceon survival.

There are three things hewould like

colorectal cancer, he asks, without dic-tating in the protocol what patientsshould get not only in the first line, butalso in second, third and fourth lines –which is not something physicians orpatientswould be likely to accept.And ifyou don’t, then how can you stopdifferences in overall survival beingconfounded by differences in the sub-sequent therapies?

The answer, he suggests, is to ensurethe survival benefit is sufficiently strongto show through despite the confound-ing impact of therapies taken after thetrial.And theway to do that is to identifythe patient group that derives a real ben-efit from the new treatment.

This is how Merck showed the sur-vival benefit for adding Erbitux to theFOLFIRI regimen for first-line treat-mentofpatientswithmetastatic colo-rec-tal cancer. In anundifferentiatedpatientpopulation, the combined treatmentshowed a significant improvement inPFS, with a hazard ratio of 0.85, but thedifference in overall survival failed toreach significance.However,Merckhadtaken tissue from themajority of its trialpatients, and was able to reanalyse thedata after stripping out the results frompatients with a mutated KRAS gene.This effectively doubled the responserate figures for the wild-type (normal)KRAS patients; it strengthened the dif-ference inPFSdata fromahazard ratio of0.85 to 0.696; and, importantly, the dif-ference in overall survival became statis-tically significant, showing an additional3.5months over the control arm.

“This demonstrates you can do it,”says Kisker. “It’s not just a question ofoverall survival as a primary endpoint; it’sa question of how to develop our prod-ucts. We have to address, even in pre-clinical models, how drugs might work,what is the mode of action and whatmight be potential biomarkers.You thengo for a phase I study, which should beused to identify patients who might

CuttingEdge


to see happen. One is that drug devel-opers and oncologists raise their ambi-tions and go the extra mile to try to getoverall survival data wherever they can.The second is for a regulatory and payerenvironment that encourages the pur-suit of this data, so that companieswon’t

feel they could endup penalised ifthey have strong

PFS data but fail toreach significance ontheir survival data, dueto explicable confound-ing factors.

The third, which he believes is cru-cial to being able to give payerswhat theywant, is the introduction of a value-based system of pricing that recognisesthat any given cancer drug can give dif-ferent levels of benefit depending onwhat cancer,what stage, andwhat line oftreatment it is used in.

IDENTIFYING YOURTARGET GROUP IS THE KEYOliver Kisker is vice president of globalclinical development for the oncologyunit at Merck-Serono, where he workswith a wide variety of cancer therapiesincluding Erbitux [cetuximab], theEGFR inhibitor approved foruse in somecolorectal cancers and squamous cellhead and neck cancers, and for whichMerck is nowseeking approval for use incertain non-small-cell lung cancers.

Kisker shares the view that beingable to show your drug improves survivalis always desirable, but in some indica-tions it is difficult to achieve: “In someareas, where few treatment options areavailable, it is important to demonstratethat overall survival is really better com-pared to the competitor. But if you havean indication where treatments aremuchmore available, like for colorectalcancer, it will be muchmore difficult.”

How can you prove the overall sur-vival advantage of a drugused first line in

benefit, by including even at that stage amarker that could identify the rightpatients.You thenuse expansion cohorts[add in patientswith themarker of inter-est] where you can see if these patientsreally do benefit.

“Then we have a combination ofexpansion cohort and biomarker, andthen you go to phase II, which gives amuch clearer picture of how patientsmight benefit basedonmolecular profile.Then you do additional analysis herewith furthermarkers, identify them, theright profile, the right patients, and thenyou go to phase III.”

This is the strategyMerck is follow-ing now with all its cancer drug devel-opments, says Kisker. In lung cancerthey are looking at high levels ofEGFRexpression as a marker for response toErbitux. And they are investigating theMGMT biomarker, associated withDNA repair function, as a possible pre-dictor of response to temozolomide,which is combinedwith cilengitide, theirexperimental integrin inhibitor for first-line use in glioblastoma.

Trialling the drug only in the popu-lation that responds well not onlyincreases the benefit you can show, butasKisker points out, it also decreases thenumber of ‘events’needed to prove thisbigger benefit, which means trial sizesare smaller.

But no matter how well you do thiswork, he adds, there will always be situa-tionswhere proving overall survival bene-fit simply isn’t possible.A classic exampleis where you are trialling a drug for use infirst line, when it has already beenapproved in a later setting. It is not onlyunethical but also impossible to deny a

patient in the control arm access to thetreatmentonce theyhaveprogressed to thepoint where that drug has been approvedand is freely available for use. “If youhavecrossover you need to think about it anddiscuss with the regulators about thecrossover effect of adrugalreadyapprovedfor a later stage indication.”

Despite the extensive early trial workinvolved, however, Kisker feels that newdrugs still need to prove themselves instandalone phase III RCTs. “You try toanswer questions you have raised inphase I and early phase II, but in the endyou need to show it in a phase III,because this is a requirement by regula-tory agencies.”

He doesn’t rule out the possibility ofextending phase IIs into the phase IIIs inthe future, though it’s not a designMerckcurrently uses. “I think it is an interest-ing approach, that youcould carry on andreduce time to approval. But I think wewould need to have further discussionwith agencies, because designing stud-ies in this way is not always accepted byagencies. This is one of the thingswherewe need to interact with agencies tospeed approval of drugs by using thesekinds of designs.”

Better interaction is also his solu-tion to the question of how to satisfy thedemands of payers. Kiskermakes time totalk to the people who have a say overreimbursement, to discuss the issueshe faces in developing a particular drug,and to find out from them what sort ofdata they need. “Payers are becomingmore and more important, and bothsides need to understand one another.They need to understandwherewe are,because theywant to see patients bene-

fit from therapy. We on the other handneed to see what are the points that wehave to address.”

He emphasises again the importanceof finding the right patient group. “Youneed to include as early in the trial aspossible personalisedmedicine throughstratifying patients using biomarkers. Ifyou see benefit for these patients, payerswill have nothing against it.”

ONE SIZE DOES NOT FIT ALLRafael Amado is Head of OncologyDevelopment atGlaxoSmithKline,wherecurrent strategies include starting devel-opment with a tightly defined patientpopulation–patientswith knowncancerpromoting molecular alterations, suchas BRAF mutations, for example – anddeveloping drugs, or combinations ofdrugs, that will be effective in the smallpopulation of patients whose cancersare driven by those alterations.

“It’s fair to insist on survival datawhenyouareusingbroad-spectrum, toxic treat-ments which afford only small incre-mental benefit, as in the case ofcytotoxics in most advanced cancers,”says Amado, “but when you are usingdrugs targeted to molecularly charac-terisedpopulations,whichcandrive largeeffects in surrogate endpoints, and areless toxic, I do not think that overall sur-vival needs to remain the gold standardagainst which we measure new drugs. Iunderstand thatwehave to showat leastreasonable likelihood of clinical benefit.But ifwecontinue to think of overall sur-vival as the gold standard, the develop-mentprocesswill continue tobe longandcumbersome, and it will become moreand more difficult to obtain it as an


CuttingEdge

“There will always be situations where proving

overall survival benefit simply isn’t possible”

notype. So I think it is incumbent on theinvestigators and sponsors to show thata phenomenon like this doesn’t occur.”

Oneway to do this, saysAmado, is tolook at whether the early differencebetween theoverall survival (OS)data forthe twoarmswanesover time. “If you seeOS hazard ratios that are trending pro-gressively in the wrong direction afterdrug discontinuation, that should raiseconcerns. There are also analyses onecan do looking at time to death fromdis-easeprogressionbetween test andcontrolarms,whichcanhelp rule out apotentialrebound effect,” he says.

The question thatshould be asked, hesuggests, iswhetherfailure to meet asurvival endpointwas due lack of statis-tical power, lack of atreatment effect, excesstoxicity of the treat-ment arm, compro-mise of delivery ofstandard therapy, or tumour promotionssuch as directly or via a rebound effect.“For instance,werecently learned that theuse of EGFR inhibiting antibodies inpatients with colorectal cancer harbour-ing KRAS mutations seem to indeedshorten survival.”

Amado questions the need for ran-domisedcontrolled trials as thegold stan-dard for all drug approvals, andpoints outthat Bayesian designs are often used bysponsors and the US National CancerInstitute to do proof-of-concept trials,and are endorsed by the FDA for use indevice approvals. “Traditional statisticaldesigns have the potential to slow downdrugdevelopmentparticularly indisease

endpoint. Indeed, as diseases becomemore chronic, waiting for survival willcontinue to tie up investment andresources and delay innovation.”

He mentions the controversy overthe use of overall survival as an end-point inmelanomawith innovative drugssuch as BRAF and MEK inhibitors.These are drugs with understoodmechanisms of action that have shownimpressive response rates and pro-gression-free survival in advancedmelanoma. Trials are being conductedagainst dacarbazine – an ancient andlargely ineffective cytotoxic. Carryingon the trial until enough patients die toreach statistically significant data onoverall survival would rule out the pos-sibility that PFS gainsmight fail to trans-late into longer survival, as the cancerfinds alternatives to theblockedpathway– which turns out to be a real concern.ButAmado saysGSKwould not be pre-pared to go down that road.

“We are developing a MEK and aBRAF inhibitor. Our randomised trialsuse crossover from the control to eitherMEK or BRAF after disease progres-sion, or a control arm that includes anactive targeted therapy, as we feel allpatients in these trials should have theopportunity to access these drugs.”

Amado concedes that progression-free survival is not a foolproof surrogatefor overall survival. “In the field of angio-genesis for example, there have beenpreclinical studies showing that areboundpro-angiogenic effect can occurafter withdrawal of antiangiogenic ther-apy, suggesting that while a patient canrespond during treatment, when onewithdraws the drug the tumour maycome back with a more aggressive phe-

settings with small patient populations,such as for instance non-small-cell lungcancerwithALK translocationsorBRAFmutations. To compound the problem,effective inhibitionof somegenetic aber-rations may depend on blocking morethan one target to ensure efficacy or pre-ventionof resistance.TheuseofBayesianstatistics canalsomodelnot just theover-all treatment of targeteddrugs aloneor incombination, but how their effects varydepending ona variety of factors, includ-ing biological heterogeneity”.

The RCT approach can only reallyansweroneor twoquestions ata time, says Amado, and issimply too blunt an instru-ment for these sorts of devel-opments, because there aretoo many variables: in whattype of tumour does a givendrug work best, in whichmolecular alteration, whichpathwaysdoyou target in thesetting of combinations(and which plays the pri-

mary role) and what doses do you use.Many companies are therefore alreadyrelying heavily on Bayesian approaches(usingmodellingandprobabilitymethod-ology– seeboxp24) to guide their proof-of-concept development, says Amado,and he expects that trend to continue.

“When you have so many variablesand are trying to test a combinationagainst a given standard, youendupwithmultiple-arm studies.And if you are notincorporating the knowledge that youget fromevery patient you endupwith alarge proof-of-concept trial that is oftenvery difficult to interpret beyond the pri-mary endpoint and safety. So eventuallyproof of concept is going to bemore and

CuttingEdge


“The RCT is too blunt an instrument for these sorts

of developments, because there are too many variables”

more iterations of trials inwhicharmsgetadded on and arms get graduated orremoved.We are now often using theseadaptive and Bayesian designs.”

Neither the EMA nor the FDA hasruled out approving a drug on the basisof evidence generated using Bayesianmethodologies, and Amado hopes thatthe field will evolve in that direction,particularly in situationswith highly bio-logically segmented populations wherethe use of novel agents result in largetreatment effects. “In oncology we stillhave towait for the first example of a fullapproval to come out of a Bayesiandesign. But I think that if regulators arewilling to accept proof of concept basedon Bayesian design as end of phase IIdata, it is only one step removed fromaccepting these designs for approval.Consider that large effects inOSmedi-ated by a novel agent in a phase III RCT

require amuch higher and faster rate ofdeath events occurring in the controlarm than in the novel therapeutic arm.The question iswhether allowing excesspatients to die at a higher and faster ratein the control arm is appropriate whenwe know that the novel drug is highlyactive from phase II studies. Forinstance, when phase II studies alreadysuggested that PFSwith a novel agent issubstantially longer thanPFSor evenOSobserved with traditional chemother-apy, one could argue there is a loss ofequipoise in randomising patients to thecontrol arm”.

GSK has been in discussions withregulators inUS andEU to reach agree-ment on the design of RCTs for use inregistering new combination therapies.“Whenusing combination therapies youhave to supply proof of the contributionof each compound to the benefit of the

combination, and to do that one needsrelatively large trials involving at leastthree arms. One way to decrease thesizeof the trial is byusing a surrogateend-point (e.g. PFS instead of OS) in one ofthecomparisons.Another step to simplifythe development of two unapproveddrugs is to use one of them as a com-parator, rather than including a fourtharm for an approved standard. This canbe done if the drug has significant activ-ity as a single agent in phase II; althoughsuch a trial, if successful,would result inapproval of the combination alone, itwould likely not support approval of eachof the agents asmonotherapy.”

For the payers, says Amado, the bigissue may become whether paying forbothdrugs up front offers better value formoney than using the two drugs in suc-cession. “It is incumbent on us todemonstrate that the value of the com-bination goes beyond an endpoint suchas progression-free survival or even over-all survival, because these comparisonsare done to single agents and not tosequencingmultiagents.”Wewill have todemonstrate that using combinationsupfront is superior to the sequential useof each drug in terms of clinical out-comes and cost-effectiveness. In thecase of BRAF and MEK it is possiblethat sequential usemaybe of no value asdrugs may be cross resistant; in thatcase the only possible use of themwouldindeed be in combination.

TAILOR-MADE TRIALSHilary Calvert is director ofAnticancerDrug Discovery at the University Col-lege London Partners, where he isinvolved inmany phase I trials. He con-fesses to an ambivalent attitude on theneed always to demonstrate overall sur-vival benefit. Whatever else cancerpatients may want from a drug, saysCalvert, we can be pretty confident thattheywant it tomake them live longer, sowe do need to show that can happen,

WHAT DO PATIENTS WANT?

Cancer patients do want to live. But at what cost? As survival times increase, issues of qual-ity of life become increasingly important, and drug developers are now encouraged, by reg-ulators and payers, to incorporate quality of life measures into their trial designs.How best to do this remains a problem.� Studies show that doctors consistently rate side-effects as less significant than they are

rated by patients – and it tends to be doctors rather than patients who fill in the trial forms.� Even where patients are asked to rate side-effects on a scale, the frequency or sever-

ity may say little about howmuch it matters to the patient – diarrhoeamay be less debil-itating if you don’t have to be out and about a lot; loss of feeling in the fingers, or disfiguringrashes affect people different ways. Even indicators such aswhether the patient can con-tinue working depend to some extent on what options they have.

� Patientsmayalsohave reasons to hide from their doctors the severity of side-effects if theythink that telling the truthmay lead to thembeing taken off a treatment theywant to keep.

� Evidence on how patients see the trade-off between extramonths of life and quality of lifeis scant and somewhat contradictory. A study done in 1990showed that patients are pre-pared to take a greater hit on their quality of life for some extra time than their doctors (orthe general public) would consider acceptable (Slevin et al.,BMJ300:1458–1460). Amorerecent study presented at the7th EuropeanBreast Cancer Conference (Sheik-Youssouf etal., EJC Suppl 8:77), which looked exclusively at patients with metastatic breast cancer,suggests doctors require the offer of an additional two to six months of life as enough toconsider trying a new therapy rather than the standard options, while almost two-thirds ofpatients want the promise of at least 10months’ additional survival.


CuttingEdge

beenpossible to develop anumber of dif-ferent drugs. I’m not saying it wouldwork for all forms of cancer, but this iswhy I am a bit ambivalent about thevery rigorous approach.”

Calvert doesn’t deny that approvinga drug on progression-free survival canlead to wrong decisions. Iressa (gefi-tinib) for non-small-cell lung cancerwasa case in point – it was approvedby the FDA on the basis ofits PFS figures, but it failedto show significantimprovement insurvival. Thismightnever have becomeclear if there hadbeenno requirementto show overall sur-vival benefit. Avastin(bevacizumab) wasanother similar case.Calvert suggeststhat, given what weknowabout the fiendish ability of cancercells to mutate in order to keep multi-plying, the possibility that their responseto being deprived of vascularisationwould be to becomemore invasive in thesearch for alternative sources of bloodmight have been anticipated by bothdevelopers and regulators.

Which is all very well to say in hind-sight, but is there any way to say inadvance when disease-free or progres-sion-free survival may be an acceptablesurrogate and when it is not? “I can’tthink of a rule that would tell you that.Looking at a particular drug and itsmechanismof action I could give you anopinion – it might well be wrong. It’s agood question but a tough one.”

especially where the drug is fairly toxic.But then he cites the history of devel-opment ofAIDS therapies, which tookplacewithout any of the stringent regu-latory controls imposed in cancer.

“We’ve seen an absolute revolution insurvival in HIV and they’re now sayingthat it maybe takes five years off yourexpected lifespan rather than killing youwithin a few years.With all the enthusi-asmandemphasis put onAIDSresearch,there’s nowabout five different targets intheHIV system and about five differentdrugs available for each one. Physiciansdon’t have any restrictions onprescribingthemandnorhave they everhad toprovethey are value for money.”

Progress in AIDS therapies was allabout finding the combinations thatwork best, whichmakes it an interestinganalogy to current approaches in cancer.“What they do is they look at the viralload, and theymeasure the changes veryquickly until they find the combinationthat works.” IfAIDS research had beenforced to jump through the sorts of hoopsstill required of cancer therapies, whereyou have to prove each individual drugwith a set of trials for overall survival, saysCalvert, they would never have pro-gressed as fast – if at all.

He concedes, however, that AIDS,like heart disease and many other con-ditions, has something that cancer lacks:a good surrogate endpoint. Cancer hasno equivalent of viral load or cholesterollevel, which have been shown to corre-late closely with survival. “Maybe theclosest analogy in cancer is chronicmyeloid leukaemia, where you can lookfor the BCR-ABL fusion, so you have aquick marker, and consequently it has

What Calvert is saying, in effect, is thatcancer is just too complex and varied forgolden rules or one-size-fits-all gold stan-dards. By the same logic, he agrees thatBayesian trial designs should replacethe gold standardRCT in certain settingsin the future, despite its quite formida-ble complexity. “I think traditionalhypothesis testingmethodologymaywellget very clumsy for things where wehave a rationale for selecting quite

small subsets of patients and giv-ing them different things.We do need amathemati-cal logical approach thatwill take our subjectivityout of whether we thinksomething is working ornot. But the classical RCTwith a 0.05 P-value and onehypothesis that you acceptor reject on the basis of itmay be too blunt an instru-ment for that.”

That then leaves thequestion of howdrugdevelopers are going toconvince reg-ulators and payers of the risk–benefitand value of their products, without anygold standards for approval? Just like thedevelopment process, says Calvert, youhave to do it on a drug-by-drug (or com-bination-by-combination) basis. “Peopleneed to take on board getting the rightexpertise onto the committee thatmakesthe judgement [on approval or reim-bursement], and really engage in a lot ofdetail about each drug, its mechanism,and the best way to evaluate it. I don’tthink there is a global solution, but ifyouknow inenoughdetail how things arehappening, you cancomeupwith a goodplan for each individual drug.”

“Cancer is just too complex and varied for

golden rules or one-size-fits-all gold standards”

CuttingEdge


For the love of the jobHolland’s first woman professor of medical oncology revels in a career

unburdened by expectations and driving ambition

� Simon Crompton

Elisabeth de Vries is enjoying investigating whether we can push the potential of imaging

techniques to the point where a patient’s response to a drug can routinely be measured in an

outpatient clinic. But she worries that the energy and creativity of her young students will be

stifled by the pressures of preordained career paths.

Traditionally, women have thought differ-ently about careers thanmen, saysElisabethde Vries, the first female professor of med-

ical oncology in theNetherlands.Women play lifeby ear, in the knowledge that children, family andunforeseen circumstancesmay get in theway of thebest laid plans. Men, historically, have followedtheir ambitions.So de Vries is apologetic that she can’t tell me

about grand plans fulfilled over her 40 year careerin theNetherlands and on the international stage.But she needn’t be.As awomanwho edged herselfto the fore of the emerging discipline of medicaloncology in the ’80s and ’90s, and now stands at thevery top of her profession, her career has real sig-nificance. DeVries, Head ofMedical Oncology atthe University Medical Centre in Groningen, is aKnight of the Order of the Netherlands, a visitingprofessor at theDanaFarberCancer Institute in theUnited States, andwon the ESMOaward in 2009

for “an outstanding contribution to the develop-ment of oncology in Europe”.And if (as she admits) she has always had a ten-

dency to overcommit herself, it is a mark not somuch of personal ambition, as of a keenly feltresponsibility on behalf of her sex.“I’ve been endlessly on boards as the only female

representative, and unfortunately, somanywomenhave been needed that I simply couldn’t do it all.”She remembers how her work on national andinternational committees revealed to her just howeasily (and subconsciously) gender could influ-ence decisions; and the disbelief of male doctorswhen female doctors became pregnant shortlyafter being awarded fellowships – as if they shouldchoose a better time.She also points out that it wasn’t so long ago that

she used her initials on research papers – neverher first name. In the past, she was all too awareof research in the 1990s indicating that papers


Masterpiece

studies have shown that women doctors havesmaller offices than male ones. Her own, ratherspacious one, was offered soon after she hadpointed this research out to colleagues. She has agood-humoured, often amazingly detached, realismabout the medical world and her place in it.

INTO A ‘MALE’ PROFESSIONThe daughter of a paediatrician and a nurse, deVrieswas exposed tohospitals fromanearly age andsoon decided she wanted to help people and be anurse. It was only when she went to secondaryschool that she realised shecouldbeadoctor likeherfather. So she trained inmedicine inherhome town,Groningen, and having at first thought of herself asa paediatrician, shebegan todevelop a special inter-est in internalmedicine. She spent a fewmonths inLondon, studying endocrinology at theNorthMid-dlesex Hospital, and remembers one person therewho had a deep influence on her outlook.

submitted by an author who was obviously awoman were less likely to reach publication.“I have the feeling that it doesn’t happen any

more,” she says. Times are changing, and 70%–80%of medical students in the Netherlands are nowfemale. Yet when it comes to the high-flying med-ical oncologists who make a name on the interna-tional stage, she suspects that men may have thehighest profile for a while yet. She’s found thatwomen doctors are unwilling to blow their owntrumpets even on curriculum vitae. “Men aregood at this whole thing of status, whether it be thecar, the house or the career. It works, andwe’re lack-ing that gene.”But deVries is no club-wielding feminist.When

Imeet her, in the cavernous atrium of theUniver-sity Medical Centre in Groningen – a new hospi-tal so well organised that patients are buzzed alongbroad corridors in golf buggies and every patientward has a view – she laughs about the fact that

Masterpiece


ANTOINETTEBORCHERT

“Therewas a female regis-trar from India, who wasverybeautiful, alwaysworebeautiful silk gowns, andwas very hard working. Ifound her a brilliant doc-tor. When she had toresuscitate a patient whohad just come inbyambu-lance, she simply took upher long gown and sat onthe trolley, and I realisedthen that, okay, maybewomenshouldn’t do inter-nal medicine, but if shecoulddo it, fromIndiaandwearing silks, thenmaybeI could too. She certainlyinfluencedme.”De Vries completed

aPhDinacute leukaemiain1982, and thenspent ayear as a research fellowlearningmoreaboutmed-ical oncology at the City of Hope Medical Centre,California–onthebasis that, though itwouldbechal-lenging, “if it didn’tworkout, it didn’tmatterbecausewomen didn’t have to work anyway!”But it did work out fine, and she returned to the

Department ofMedicalOncology atUMCGronin-gen as a senior staffmember in1983,where shehasbeen based ever since. Her career has straddledpatient care, education and influential translationalresearch. She worked on several types of cancer,with a focus on breast cancer and neuroendocrinetumours, and isparticularly interested inpersonalisedtreatments, using interdisciplinary research toimprove diagnosis and treatment of a range of can-cers.Her current research lines are aimedat increas-ing thesensitivityof tumours toanti-cancerdrugs, andmolecular imaging to support this.It’s molecular imaging that she wants to talk to

me about, “because that’s what’s bothering me

most”.Only in the past year has she decided to con-centrate on it for research because its potential isbecoming clear.

THE EXCITING POTENTIAL OFMOLECULAR IMAGINGMolecular imaging techniques allow biologicalprocesses at cellular andmolecular levels to be visu-alised andmeasured in living patients.With knowl-edge about the heterogeneity of cancers and theneed for targeted therapies increasing, imagingofferstheprospect ofmonitoringhowtreatments affect thebiological processes that influence cancer growth.Afluorescent or radioactive label, for example, can beaddedtoaproteinorantibody that is attractedbywhatis believed to be an important tumour characteristic–HER2 expression in breast cancer, for example.What’s exciting about the techniques, says de

Vries, is that the scansmayoffer vital informationon


Masterpiece

Meet the extended family. Though she may have felt obliged to stay tied to Groningen morethan she might have liked, de Vries happily combined raising two children with working her wayup to professor of oncology, and even welcomed in an additional child along the way!

“I realised then that, if she could do it,

from India and wearing silks, then maybe I could too”

how a patient is responding at a very early stageof treatment – within days even. This has clearimplications for tailoring treatments to patients –and controlling drug budgets. She points to somespectacular longitudinal and cross-sectional scansof a patient who was injected with a radioactivetracer linked to an antibody against HER2. Yellowpatches show the areas where the tracer wasabsorbed – patches in the liver and the boneswhere there are clearly metastases.Then the patient was treated with a drug that

reducesHER2expression.Asecondsetof scans, twoweeks after treatment started, reveals that the yellowpatches have contracted, and become riven withholes. The drug is affecting its target.“It shows that thecharacteristics in the lesionsare

changingwhenyou treat thepatient.But if the treat-menthadn’t affectedHER2expression as theoncol-ogist hoped, the findings might be used to inform achange of treatment.”There are implications for screening and drug

development, as well as patient welfare. What’sparticularly interesting for de Vries at the momentis that her research, in collaboration with researchcentres in other countries, is indicating that, usingthe latest technology, fluorescence can be detectedin tissue far better than originally believed. Tumour-targeted fluorescent tracers can be detected duringendoscopy or surgery, or evenusinghandheldprobesthat can pick up light several centimeters under tis-sue. This is a better option than using radioactivity,which has obvious risks for patients.“Molecular imaging allows you to see thebehav-

iour of thedrug in thebody as awhole. Is it reachingits target in the tumour?How long is it staying there?Areyoudosingproperly?Thisgivescluesalso to speedup drug development and decisionmaking becauseyou really knowwhat’s going on. You can fuse theseimageswithCTorMRI,providing informationaboutthecharacteristics of the lesionand its exact location,which is very useful for surgeons too.”As the years of researchhaveprogressed, deVries

hasbecomepersuaded thatmolecular imagingmaybeof the greatest use in developing novel therapies.“I think it is within reach that we can label

novel drugs not onlywith radionucleotides but alsowith fluorescent tracers, and then routinely checkcertain lesions over time in the outpatient clinic,without theneed for smart people around you all thetime. That would be really nice for drug develop-ment. I still have to prove it, but I have the feelingwe’ll make progress in the future.”

FOCUSED ON THE POSITIVESDeVries likes to talk about the present rather thandig into the past or peer too far into the future. It’sher current research that interests her. Equally,she’s a great believer in a positive attitude, of livingin the present – as an oncologist, she’s all too awareof people who have put too much store on waitinguntil retirement to enjoy life, only to find it accom-panied by illness. “Life is too short not to appreci-ate those important little moments that make youhappy as an oncologist, like making the right deci-sion, or a patient getting better than you expected,or your PhD students doing well.”She remains deeply influenced by NannoMul-

der, ahaematologist and thenoncologist,who super-vised her PhD thesis in Groningen. Whatever theproblem,hemade it a discipline to thinkof ten solu-tions. “He is a brilliant thinker. The ideas weren’talways feasible, but from ten you usually had some-thing tochoose. I think it is ahugeadvantage tomeetpeopleearly in yourcareerwhoseeopportunities, nothurdles, everywhere. He certainly influenced mydecision togo intooncology, andsee itmoreasachal-lenge tobemet, at a timewhen itwas seenas secondrate by others in internalmedicine.”Somedoctors in theearly ’80s, she recalls, thought

that young internistswhowanted to go intooncologywere strange: why would you want to go into a spe-cialty where there was so little to do for the patient?How things have changed, de Vries reflects. She isdeeply proud of being in if not the first then the

Fluorescent tracers can be detected during

endoscopy or surgery, or even with handheld probes

Masterpiece


secondcohort of this newprofession, andoneof thevery first women. She repeats again and again whata good decision shemade to go into oncology.“It’s so exciting to be in a profession where

every year you see change. That’s nowmore true inoncology than other fields. It’s always nice to seeprogress in patients, and we are now better athelping them live longer than ever before. But theother exciting thing is that biology is helping us findnewmechanisms to treat cancer – though never asfast as you want to. It means you have to keep onstudying and learning to acquire new insights andunderstanding of pathways and mechanisms. Ifthat’s what you like doing, it’s wonderful that some-where like here you can translate it into somethingthat works in the clinic.”Despite her international outlook – deVries has

been a member of numerous EORTC and ESMOcommitteesand is involved in theEuropeanAcademyofCancerSciences– shehas spenther entire careerin Groningen, working her way up to assistant pro-fessor at the Department of Medical Oncology in1983, then associate professor in 1989 and thenfull professor in 1997.She admits to seriously considering working

elsewheremany times. “Thingsmight havebeendif-ferent,” she says, with a touch of regret. “But every-thing has to fit.” The needs of the family haveobviously played a part in decisions to stay put. Sheis married to a gastroenterologist, and has twodaughters now in their mid 20s – one a physicalchemist specialising in nanotechnology, the othernearly qualified as a medical doctor. But childcareissues never got in the way of her career (de Vrieshas always worked full-time) and despite the guiltthat she and other parents suffer as a result of not

staying at home, she observes wryly that it has hadno negative effect on her children whatsoever.Staying inGroningenhas allowedher todowhat

shewants to do, on the clinical, research and teach-ing fronts. She is a significant figure in the nationalcancerworld, a vice chair of theDutchCancerSoci-ety, and a member of the Health Council of theNetherlands since2008. It is important, she says, topresent the medical perspective when high-levelhealth policy decisions are beingmade. “I think thatdoctors have to speak up, for example, on smoking.I’mnot sure thatmost of us like to do it, but someofus need to.”

THE EXCITING POTENTIAL OF YOUNGONCOLOGISTSIn Groningen, a university town jam-packed withstudents on bicycles, it has been the medical andPhD students she teaches and supervises thathave kept her feet grounded and her brain buzzingwith new ideas. De Vries says that much of herresearch, including her work on imaging, has beenfired by their creativity and knowledge of newtechnology. So as she looks ahead to the next cou-ple of years, one of her main aims is to train moreyoung people into independent doctors and scien-tists. But sheworries about the increasing burdensbeing put on them.“Wehave all these rules, requirements, forms to

fill out, all the administrative burden associatedwith trials. Life for young doctors and scientists hasbecome much more demanding from that per-spective, and if they want to do research becauseit’s inspiring and gives them the chance to seepatients regularly, it’s difficult to give them thesame opportunities as ten years ago. I worry that we


Masterpiece

“It is a huge advantage to meet people early in your

career who see opportunities, not hurdles, everywhere”

“Doctors have to speak up, for example on smoking.

I’m not sure most of us like to, but some of us need to”

are going to lose these people who are trying totranslate findings from lab to clinic.” She believesit is her role to create a supportive setting whichmakes students independent but helps them overthe time-consuming hurdles.De Vries has enormous faith in the abilities of

young people, if not tied down by the bonds ofexpectation imposed by parents and society. Thisfaith, it turns out, is rooted in part in anunusual per-sonal experience that also helps explain her reluc-tance to make plans for the future.Around a decade ago, she explains, the family

got what de Vries calls ‘a borrowed daughter’ – athird child who came from another home. Heryoungest true daughter brought her home.“She came from a background where people

didn’t go to high school, a disrupted family, and inthe end my daughter thought it would be a goodidea if she became part of our family, and she did.”

The girl was never formally adopted, but all the par-ties involved were happy with the arrangement.“She brought in a different background, and shemade us realise that education and child-rearingseem important, but actually a lot of things chil-dren do is through their own inspiration, theirown drive.” The young woman has now finished aMasters degree in education and is about tobecome a teacher.“So this is a gift,” says de Vries. Sometimes, she

points out, the best things in life are unexpected.She reflects that for young women now, starting acareer inmedicine or another profession, things aremuch harder than they were for her 30 years agobecause nowadays a course is charted out andthey are expected to do well. “I didn’t have goals –I didn’t have to reach any particular goal, so Ididn’t disappointmyself!” Thankfully, she didn’t endup disappointing anybody.

“I worry that we are going to lose these people who

are trying to translate findings from lab to clinic”

Masterpiece


Don’t tie them down. De Vries tries to protect her students (pictured here alongside staff members) from all the expectations,rules and bureaucracy that could end up sapping their enthusiasm and creativity

ImpactFactor

Rituximab for follicular lymphoma:maintaining an open mind

� Bruce Cheson

Newdata from thePrimaryRituximabAndMaintenance study provide the strongest support for

the use of rituximab maintenance in patients with follicular lymphoma. However, further

considerations of cost, inconvenience, toxic effects, efficacy of retreatment and lack of survival

benefit should focus future clinical research on more-effective induction strategies.

Few drugs have made as great animpact on how patients withlymphoma are treated as the

chimeric anti-CD20 monoclonal anti-body rituximab. Chemoimmunother-apy regimens incorporating rituximabwere the first strategies in decades toprolong the survival of patients withdiffuse large B-cell non-Hodgkin’slymphoma (DLBCL), follicularlymphoma and chronic lymphocyticleukaemia. Moreover, rituximab is

often the platform on which newertherapies are developed. However,until all patients are cured, thereremains room for improvement in ourtreatments.

Given the already high completeresponse and overall response rateswith chemoimmunotherapy, oneattractive strategy has been to increasethe duration of response with post-induction treatment, such as mainte-nance rituximab. Unfortunately,

maintenance rituximab has notdemonstrated a benefit in patientswith DLBCL or chronic lymphocyticleukaemia. Nevertheless, the use ofmaintenance rituximab for the treat-ment of follicular lymphoma is wide-spread. Results from the Lymphocarestudy suggest that 45% of patients inthe USA are being treated with thisstrategy following chemoimmuno-therapy induction.1 Indeed, new datafrom the Primary Rituximab And


This article was first published in Nature Reviews Clinical Oncology 2011 vol.8 no.5, and is published withpermission. © 2011 Nature Publishing Group. doi:10.1038/nrclinonc.2011.35, www.nature.com/nrclinonc

ImpactFactor

Maintenance (PRIMA) study providesupport for the use of maintenance rit-uximab,2and will be discussed below.

What other available data arethere to support the maintenancerituximab approach? Martinelli etal.3 randomly assigned 202 patientsto four weekly rituximab infusionsfollowed by observation or mainte-nance using one dose every twomonths for eight months. At amedian follow-up of 9.5 years, 45%of previously untreated patientsremained event-free in the pro-longed therapy arm compared with22% in the control arm, but withno significant survival advantage(P=0.0813).

Ardeshna and co-workers con-ducted a three-arm randomised trialin which patients with stages II–IV,asymptomatic, non-bulky follicularlymphoma were randomly assignedto either watch-and-wait, weekly rit-uximab for four doses alone, or fourdoses of weekly rituximab followedby two years of maintenance.4 Timeto next therapy and progression-freesurvival (PFS) favoured the mainte-nance group. The controversial con-clusion was that this approachshould become the standardapproach. However, the follow-upwas only 32 months, there was nosurvival advantage, and manypatients might still prefer to waituntil progression before initiationof therapy, when more-effectivetreatments might become available.Moreover, there were no data onresponsiveness to second-line ther-apy (first systemic treatment forwatch-and-wait and second systemictreatment for rituximab-treatedpatients). Hochster et al.5 treatedpatients with cyclophosphamide, vin-cristine and prednisone (CVP) fol-lowed by rituximab maintenancewith four weekly doses every six

months for two years. Maintenanceprolonged PFS, but not overall sur-vival. Other groups evaluated main-tenance rituximab in the relapsedsetting, with prolongation of responseduration or PFS6 but without a sig-nificant overall survival benefit.

However, until now, no study hadaddressed the most important ques-tion: does maintenance rituximabimprove the outcome of patients ini-tially treated with the standard ofcare – chemotherapy plus ritux-imab? An important publication bySalles et al.2 of the PRIMA studyaddresses this issue. Indeed, theFDA approved rituximab for main-tenance therapy of follicular lym-phoma in January 2011, a decisionbased largely on results from thistrial. Previously untreated patientswith follicular lymphoma weretreated with R-CHOP (rituximab,cyclophosphamide, adriamycin,vincristine, prednisone), R-CVP (rit-uximab and CVP) or R-FCM (ritux-imab, fludarabine, cyclophosphamide,mitoxantrone) at the choice of thetreating physician. Patients who

achieved either a complete remis-sion or partial remission were ran-domly allocated to either no furthertherapy or to rituximab maintenanceevery two months for two years (oruntil disease progression occurred).Of the 1217 patients who enteredthe induction phase, 503 underwentmaintenance and 513 underwentobservation alone. At a median fol-low-up of 36 months, the PFS was74.9% in the maintenance arm com-pared with 57.6% in the observationarm (HR 0.55, 95%CI 0.44–0.68).The benefit was observed regardlessof designated pretreatment charac-teristics including treatment regi-men, age, sex, Follicular LymphomaInternational Prognostic Index(FLIPI) or response to induction.In addition, event-free survival wasimproved in the group receiving rit-uximab maintenance. More patientswere in complete remission at theend of maintenance than in theobservation group. At the time ofpublication, there was no differencein overall survival because of thelow number of events.

As maintenance rituximab con-sistently prolongs PFS, why notdeliver it to all patients? Recognisingthat I am in the minority by avoidingmaintenance rituximab, I will sharemy rationalisations. First, thistherapy is expensive and timeconsuming. Moreover, the optimalmaintenance schedule and durationare not known. In addition, therewere more adverse events in themaintenance arm of the PRIMAstudy (56% vs 37%), notably grade2–4 infections, with one death fromfulminant hepatitis B.2

What is also not clear is whethermaintenance provides an advantageover retreatment upon relapse.Many patients who had previouslyresponded to rituximab respond again


ImpactFactor

to the same therapy, often with alonger response duration. Hainsworthet al.7 randomly assigned relapsed andrefractory patients to maintenance orretreatment upon relapse. Mainte-nance prolonged PFS, with no survivaladvantage.

The results of the RESORT trial(E4402; NCT00075846, which wascompleted in October 2008), in whichpatients received one weekly dose ofrituximab for four weeks followed byeithermaintenance every threemonthsuntil progression or observation withretreatment upon relapse, will be ofinterest to address this issue.

Second, there are toxic effectsassociated with maintenance ritux-imab including neutropenia, grade 3and 4 infections,8,9 and a small risk ofpotentially fatal, progressive multifocalleukoencephalopathy. Follicular lym-phoma is a disease characterised byrepeated relapses. Therefore, anotherconcern is whether prolonged ritux-imab may compromise responsivenessto subsequent therapy.

In the CORAL study, Gisselbrechtand colleagues compared R-ICE (rit-uximab, iphosphamide, carboplatin,etoposide) versus R-DHAP (rituximab,dexamethasone, cytarabine, cisplatin),both with stem-cell transplantationin relapsed DLBCL.10 One of thestrongest predictors of inferior out-come was receiving rituximab in a pre-vious regimen. Data from the PRIMAstudy on responsiveness to salvagetherapies are not yet available.

One alternative might be to useanother agent after induction therapy.Potential candidates include radioim-munotherapy, newer antibodies such ashumanised anti-CD20s, galiximab(anti-CD80), and epratuzumab (anti-CD22), lenalidomide, small-moleculeproapoptotic agents, and signaling

pathway inhibitors, such as thosedirected against Bruton tyrosine kinaseor PI3K.

Although the PRIMA study pro-vides the strongest support yet forthe use of maintenance rituximabin patients with follicular lymphoma,some of us will continue to waituntil studies demonstrate an impacton survival and any further compli-cations that may appear over time asa result of maintenance rituximab.Most importantly, the availability ofnewer, more-effective targeted ther-apies may provide a solution – if youhave better induction, then main-tenance becomes irrelevant.

References

1. C Flowers et al. (2010) Use of

maintenance rituximab (R) in the United

States following R-based induction for

follicular lymphoma (FL) [abstract]. JCO 28

(Suppl. 15):a8100

2. G Salles et al. (2011) Rituximab

maintenance for 2 years in patients with high

tumour burden follicular lymphoma

responding to rituximab plus chemotherapy

(PRIMA): a phase 3, randomised controlled

trial. Lancet 377:42–51

3. G Martinelli et al. (2010) Long-term

follow-up of patients with follicular

lymphoma receiving single-agent rituximab at

two different schedules in trial SAKK 35/98.

JCO 28:4480–4484

4. KM Ardeshna et al. (2010) An intergroup

randomised trial of rituximab versus a watch

and wait strategy in patients with stage II,

III, IV asymptomatic, non-bulky follicular

lymphoma (grades 1, 2 and 3a). A

preliminary analysis [abstract]. Blood 116:a6

5. H Hochster et al. (2009) Maintenance

rituximab after cyclophosphamide,

vincristine, and prednisone prolongs

progression-free survival in advanced

indolent lymphoma: results of the

randomized phase III ECOG1496 study.

JCO 27:1607–1614

6. MH van Oers et al. (2010) Rituximab

maintenance treatment of relapsed/resistant

follicular non-Hodgkin’s lymphoma: long-

term outcome of the EORTC 20981 phase

III randomized intergroup study. JCO

28:2853–2858

7. JD Hainsworth et al. (2005) Maximizing

therapeutic benefit of rituximab:

maintenance therapy versus retreatment at

progression in patients with indolent non-

Hodgkin’s lymphoma – a randomized phase

II trial of the Minnie Pearl Cancer Research

Network. JCO 23:1088–1095

8. MH van Oers et al. (2006) Rituximab

maintenance improves clinical outcome of

relapsed/resistant follicular non-Hodgkin’s

lymphoma in patients both with and without

rituximab during induction: results of a

prospective randomized phase 3 intergroup

trial. Blood 108:3295–3301

9. R Forstpointner et al. (2006) Maintenance

therapy with rituximab leads to a significant

prolongation of response duration after

salvage therapy with a combination of

rituximab, fludarabine, cyclophosphamide,

and mitoxantrone (R-FCM) in patients with

recurring and refractory follicular and mantle

cell lymphomas: results of a prospective

randomized study of the German Low Grade

Lymphoma Study Group (GLSG). Blood

108:4003–4008

10. C Gisselbrecht et al. (2009) R-ICE

versus R-DHAP in relapsed patients with

CD20 diffuse large B-cell lymphoma

(DLBCL) followed by autologous stem cell

transplantation: CORAL study [abstract

8509]. JCO 27:436s


Author affiliations: Georgetown University Hospital, Lombardi Comprehensive Cancer Center, Washington, DC, USA

Practice pointMaintenance rituximab prolongs pro-gression-free survival of patients withfollicular lymphoma,butmaynot yetbestandard therapy for all patients.

ImpactFactor

ASTER – another flower inthe diagnostic field of lung cancer?

� Paul Baas

Mediastinal staging of patientswith lung cancer is used to avoid futile thoracotomies. Endoscopic,

oesophageal and bronchial ultrasound procedures aremethods to identify involved lymphnodes.

TheASTER study indicates that the sensitivity of these new techniques is high, reducing the

number of futile thoracotomies and improving outcomeswhen combinedwithmediastinoscopy.

Lung cancer has a very high inci-dence and is themost lethal can-cer type worldwide, accounting

for 12.7% of the total cancers diag-nosed.1 Patients presenting withlocalised tumours that can be resectedcompletely tend to achieve the best out-comes after treatment; therefore, stag-ing of the mediastinum is of greatimportance.Cervicalmediastinoscopy isthe standard procedure to investigatethemediastinum.Under general anaes-thesiology a small incision ismade in thecollar just above the manubrium. Avideoscope is introduced and proceedsalong the trachea. Lymphnodes in frontor on both sides of the trachea can bevisualised and sampled (paratracheal,ventral and subcarinal lymph nodes)

for histological examination.Annema et al.2 compare the use of

standard mediastinoscopy with a com-bination of endobronchial ultrasound(EBUS) and endoesophageal ultrasound(EUS) for mediastinal nodal staging oflung cancer; if no cancerwas detected inthe experimental arm,mediastinoscopywas performed. EUS and EBUS have aclear advantage over mediastinoscopyin that theyprovide improved coverage ofthe mediastinal lymph node stations(see figure).3 Theoretically, the combi-nation ofEUSandEBUS should lead tobetter staging and reduction of the num-ber of futile thoracotomies; theASTERstudy examined this theory. A directcomparison ofEUSand/orEBUSversusmediastinoscopy was not performed

because the current guidelines indicatethat mediastinoscopy is the standard ofcare.4 Therefore,Annemaet al.2 includedthis analysis in the experimental arm todetermine the additional value of thecombination.The results were as expected; the

experimental armproducedbetter resultsthan mediastinoscopy alone and out-comeswere substantially improvedwhenendosonography (EBUS and EUS) andmediastinoscopy were combined. Withtheexceptionof ipsilateral or contralateraldisease detection, direct invasion of thetumour can be visualised more easilywith endosonography than with CT orPET imagingormediastinoscopy.Positivemediastinal lymph node samples wererecorded in 41 of 118 patients in the


This article was first published inNature Reviews Clinical Oncology 2011 vol.8 no.4, and is published withpermission. © 2011Nature Publishing Group. doi:10.1038/nrclinonc.2011.22, www.nature.com/nrclinonc

ImpactFactor

control arm and in 56 of 123 patientswho underwent endosonography alone.The addition of mediastinoscopy to theexperimental arm revealed an additionalsix lymphnodemetastases that had pre-viously beenmissedby endosonography.The number of unnecessary thoraco-scopies prevented was 21 (18%) in theexperimental group comparedwith nine(7%) in the control arm.The ASTER study is the first ran-

domised trial that presents data on thesensitivity and negative predictive valueof both arms. The study hasmany strongfeatures; it is an investigator-initiated,multicentre, prospective study con-ducted in experienced centres and dataare presented on an intent-to-treat basis.Features of this study have similarities tothe implementation of the PET scan inthe staging of lung cancer. The use ofPET scanning has reduced the numberof futile thoracotomies by 50%5 and hasbecome part of the accepted guidelinesin theWestern world.Nonetheless, a number of criticisms

canbemade about the study byAnnemaet al.2 Learning to use EBUS requiresextensive training and keeping expertiseat a high level requires aminimumnum-ber of cases per operator per year. Thus,identification of specific referral cen-tres will be of importance. Coughing,patient distress andhypoxia can hamperthe endobronchial procedure, leadingto incomplete endobronchial oroesophageal examination. Proper patientselection for the use of midazolam orpropofol anaesthesia can reduce theseproblems. One of the advantages of theendosonography procedure is thatmanyinstitutes will use rapid on-site exami-nation (ROSE), informing the broncho-scopist during the procedurewhether ornotmore punctures are required. In theASTER study, ROSE was used only inthe experimental arm.1 The limited

amount of cytologicalmaterial obtainedby punctures is of concern whenendosonography alone is performed inpatientswhopresentwith positivemedi-astinal lymph nodes.Histology samplesare preferred to test for molecular bio-markers such asEGFR-activatingmuta-tions, EML4-ALK translocations orKRAS mutations. This list of potentialbiomarkers is growing and requires aminimum amount of histological mate-rial. The cytologicalmaterial obtainedbypunctures is, at the moment, insuffi-cient andmight increase the number offalse-negative results.The conclusion of this newapproach

is simple: endosonography using a com-bination of EBUS and EUS is here tostay and will allow a quick selection ofpatients suitable formajor surgical inter-ventions in lung cancer.Mortality is nearzero, as ismorbidity.Mediastinal bleed-ings are extremely infrequent and per-sistent hoarseness due to lesions of therecurrent laryngeal nerve has not beenreported. Does this information meanthat mediastinoscopy is now in its pre-

terminal stage? Not yet. There willalways be an indication for this proce-dure, such as the need for histologicalmaterial, or in case of restaging afterinduction therapywhenendosonographyhas failed to identify previously involvedlymphnodes.6,7As applicable to surgicalprocedures, quality assurance, trainingandmaintenance of expertise remain ofgreat importance.

Details of the references cited in this article can

be accessed at www.cancerworld.org


Practice pointThe combination of endoesophagealultrasound and endobronchial ultra-sound offer the physician a new andless-invasive method to stage and re-stage themediastinum. Its reach is supe-rior to that of the mediastinoscopy andfuturedevelopments inmolecular genet-ics will allow the required analysis ofspecimens formolecularmarkers.

Author affiliations: Department of Thoracic Oncology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

Transversal view of the chest at the level of the main carina. The red dotted line indicates the fieldapproachable by mediastinoscopy; the black unbroken line for endobronchial ultrasound and thedotted black line for the endoesophageal ultrasound

NewsRound


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

Cytarabine: lowdose as effectiveas high dose in AML� NEJM

Intermediate dose cytarabine was shown tobe as effective as high-dose cytarabine in

the treatment of acute myeloid leukaemia(AML) with less toxicity, an investigator-ledstudy has concluded.

Cytarabine (ara-C) is one of the corner-stones of treatment for AML. Although high-dose cytarabine is now used routinely forinduction and consolidation therapy it has notbeen compared in studies with intermediate-dose cytarabine, which could result inmaximalanti-tumour effects with less toxicity.

In the current study Bob Löwenberg andcolleagues, from the Erasmus University Med-ical Centre (Rotterdam, the Netherlands),compared outcomes for 821 patients with

(HR=1.14, 95%CI 0.81–1.60; P=0.45).In the first three months there were 72

deaths in the high-dose group versus 52 inthe intermediate-dose group (HR=1.41;P=0.057). However, at five years there wereno significant differences between the inter-mediate-dose group and high-dose group inthe rate of probability of relapse, event-freesurvival or overall survival. High-dose cytara-bine provided no clear advantage for anyprognostic subgroup.

After the first cycle, 61% of patients in thehigh-dose cytarabine group suffered grade 3to 4 adverse events versus 51% in the inter-mediate-dose group (P=0.005). Specifically,skin reactions and gastrointestinal and oculartoxic effects were noted. Additionally in cycle2, more patients in the high-dose group suf-fered prolonged hospitalisation and delayedneutrophil recovery, and in cycles 2 and 3more patients in the high-dose group suffereddelayed platelet recovery.

“The results suggest that the anti-

AML (aged 18–60 years) and 39 patients withrefractory anaemia with excess blasts (RAEB),who were randomly assigned to high-dosecytarabine (n=429) or intermediate-dosecytarabine (n=431).

The high-dose group received a dose-escalated regimen of 1000 mg/m2 of cytara-bine every 12 hours in cycle 1 and 2000mg/m2

twice daily in cycle 2. The intermediate-dosegroup, received cytarabine at a dose of200 mg/m2 given by continuous intravenousinfusion for 24 hours during cycle 1 of induc-tion therapy and 1000 mg/m2 by infusion for3 hours twice daily during cycle 2 of inductiontherapy. For the third cycle patients with acomplete response received consolidationtherapy with chemotherapy (mitoxantrone–etoposide) or underwent autologous or allo-geneic stem-cell transplantation.

Results show that, at a median follow-up of five years, complete remission rateswere 80% for the intermediate-dose groupversus 82% for the high-dose group

NewsRound


leukaemic effects of cytarabine may reach amaximum at doses well below the maximumtolerated dose,” conclude the authors, adding“… the high-dose cytarabine regimen resultedin considerable toxic effects, was significantlymore myelosuppressive, and required moreplatelet transfusions and prolonged hospi-talization. Myelosuppression of high-dosecytarabine appears cumulative and is carriedover to post remission chemotherapy.”

� B Löwenberg, T Pabst, E Vellenga et al.

Cytarabine dose for acute myeloid leukemia.

NEJM 17 March 2011, 364:1027–1036

Eribulin delivers overallsurvival benefit inmetastatic breast cancer� The Lancet

Eribulin produced a significant improve-ment in overall survival in women with

heavily pre-treated metastatic breast cancerwhen compared to treatments selected bydoctors, the EMBRACE study has reported.

It is widely recognised that a great needexists for new treatments to improve overallsurvival in women with advanced or recurrentmetastatic breast cancer, particularly thosewith heavily pre-treated disease. Eribulinmesilate is a non-taxane microtubule dynam-ics inhibitor that is a structurally modifiedsynthetic analogue of halichondrin B, a nat-ural product isolated from the marine spongeHalichondria okadai.

In the phase III EMBRACE trial, led byChris Twelves from the University of Leeds inthe UK, 762 women with metastatic breastcancer who had received a median of fourprevious chemotherapy regimens from 135centres in 19 countries were randomly allo-cated, in a 2:1 ratio, between November 2006and November 2008, to treatment with eribu-lin (n=508) or to the treatment of physician’schoice (n=254). The treatment of physician’schoice (TPC) arm represented a mix of agents

(both approved and non-approved formetastatic breast cancer) intended to mirrorclinical practice at the time of the study. In theTPC arm 96% received chemotherapy, withvinorelbine, gemcitabine and capecitabinebeing the most frequently used agents.Patients and investigators were notmasked totreatment allocation.

Results show that overall survival was13.1 months in women assigned to eribulinversus 10.6 months in women assigned toTPC (HR=0.81, 95%CI 0.66–0.99; P=0.041).Furthermore, median progression-free sur-vival was 3.7 months with eribulin versus2.2 months with TPC (HR=0.87, 95%CI 0.71–1.05; P=0.137).

Asthaenia or fatigue occurred in 54% ofpatients on eribulin versus 40% on TPC, andneutropenia occurred in 52% of patientsreceiving eribulin versus 30% receiving TPC.Peripheral neuropathy was the most com-mon adverse event, leading to discontinuationfrom eribulin in 5% of patients.

“This ... study establishes a potential newstandard treatment for women with heavilypre-treated metastatic breast cancer, forwhom there was previously no chemotherapytreatment with proven survival benefit,” writethe authors, adding that on the basis of theresults, eribulin has received approval in theUSA for patients who have received at leasttwo chemotherapeutic regimens for the treat-ment of metastatic breast cancer, with previ-ous treatments including an anthracyclineand a taxane.

Eribulin, they add, has a manageableprofile of toxic effects, short infusion times,and is easy to administer with no require-ment for premedication to prevent hyper-sensitivity. Further evaluation of eribulinearlier in the natural history of breast canceris now warranted.

� J Cortes, J O’Shaughnessy, D Loesch, et al.

Eribulin monotherapy versus treatment of

physician’s choice in patients with metastatic

breast cancer (EMBRACE): a phase 3 open-

label randomised study. Original text. Lancet

12 March 2011, 377:914–923

Small proportionof second cancersrelated to radiotherapy� Lancet Oncology

Around 8% of second cancers that developin adult cancer survivors are related to

radiotherapy, an analysis of the US Surveil-lance, Epidemiology and End Results (SEER)cancer registries has found. The majority ofsecond cancers, said the authors, are attrib-utable to lifestyle or genetics.

Radiotherapy reduces the risk of cancerrecurrence, promotes tumour control, andimproves survival. However, with improve-ments in survival, the long-term risks fromradiotherapy, including the risk of developinga second cancer, have becomemore important.

In the current study Amy Berrington deGonzalez and colleagues, from the NationalCancer Institute (Bethesda, Maryland), under-took a comprehensive and systematic analysisof data recorded in the US SEER cancer reg-istries on 15 solid cancer sites in adults whohad been routinely treated with radiotherapy.Patients were aged 20 years or older and hadbeen diagnosed with their first primary inva-sive solid cancer between January 1973 andDecember 2002. Due to the five-year lagbetween radiation exposure and solid-cancerinduction, investigators excluded patients whosurvived less than five years from treatment.

Relative risks (RRs) for a second cancer inpatients treated with radiotherapy versuspatients not treated with radiotherapy wereestimated using Poisson regression analysisadjusted for age, stage, and other potentialconfounders.

Altogether 647,672 adult cancer patientsin the cohort survived for five years or longerand were followed up for a mean of 12 years.The proportion of patients who received radio-therapy as part of their initial cancer treatmentvaried from 23% for non-small-cell lung can-cer to 79% for testicular seminomas.

Results showed that the attributable riskof a second cancer to radiotherapy was 5%

NewsRound


for cancers of the oral cavity/pharynx, 12%for the salivary glands, 7% for the rectum,10% for the anus, 5% for the larynx, 6% forthe lung, 15% for soft tissue, 5% for femalebreast, 17% for the cervix, 9% for theendometrium, 10% for the prostate, 24% forthe testes, 4% for the eye/orbit, 9% for thebrain and 7% for the thyroid. Overall theattributable risk was 8%.

In general, the investigators found thatrelative risk was highest for organs thatreceived greater than 5 Gy, decreased withincreasing age at diagnosis, and increasedwith time since diagnosis.

“These findings can be used by physiciansand patients to put the risk of radiation-relatedcancer into perspective when compared withthe probable benefits of the treatment,” writethe authors, adding that studies are nowneeded of secondary cancer risks related tonewer radiotherapy treatments.

The strengths of the study include its sys-tematic approach, large sample size and long-term follow-up, with the main limitationbeing lack of treatment randomisation, pro-viding a potential for confounding factors.

� A Berrington de Gonzalez, RE Curtis, SF Kry,

et al. Proportion of second cancers attributable

to radiotherapy treatment in adults: a cohort

study in the US SEER cancer registries. Lancet

Oncol April 2011, 12:353–360

Colonic stenting delivers noadvantages over emergencysurgery in malignantcolonic obstructions� Lancet Oncology

Colonic stenting offers no decisive advan-tage over emergency surgery in patients

with acute malignant colonic obstruction,concluded a Dutch study.

Around 7%–29% of patients with colo-rectal cancer present with bowel obstruc-tions that require emergency surgery to

the authors, adding that further studies areneeded to establish whether specific groups ofpatients might have experienced greater ben-efit in either group.

While colonic stenting can be used as analternative to emergency surgery, write theauthors, caution should be exercised due toconcerns over overt and silent perforations,which are more likely to occur with stentsand might result in distant seeding of malig-nant cells.

In an accompanying commentary, LouisWong Kee Song and Todd Baron from theMayo Clinic (Rochester, Minnesota), wrotethat until improvements in colonic stentdesign are addressed, endoscopic preoperativecolonic stenting should be undertaken inselected centres and in selected patientsdeemed most likely to benefit.

� J Evan Hooft, WA Bemelman, B Oldenburg et

al. Colonic stenting versus emergency surgery for

acute left-sided malignant colonic obstruction: a

multicentre randomised trial. Lancet Oncol April

2011, 12:344–352

� L Wong Kee Song, T Baron. Stenting for acute

malignant colonic obstruction: a bridge to

nowhere? ibid, pp 314–315

Denosumab representstreatment option forbone metastases inprostate cancer� The Lancet

Denosumab proved better than zole-dronic acid, the standard of care, in

preventing skeletal-related events in menwith bone metastases from castration-resistant prostate cancer, an internationalphase III study has found.

Bone metastases are a major burden formen with advanced prostate cancer. Histo-logical findings, together with analysis ofbone turnover markers, suggest that excessosteoclastic activity is responsible for bone

restore luminal patency (unblock the pas-sage). Emergency surgery is associated withmortality rates of 15%–34% and morbidityrates of 32%–64%. In the early 1990scolonic stenting was introduced to restoreluminal patency in patients with malignantobstruction on the left side of the colon,with uncontrolled studies suggesting thatstent placement before elective surgerydecreases mortality, morbidity and the num-ber of colostomies. Additional advantagesthat have been suggested for the temporaryprocedure are that it enabled accuratetumour staging and prevented the need forsurgery in patients found to have dissemi-nated disease.

Jeanin van Hooft and colleagues fromthe University of Amsterdam, in the Nether-lands, set out to establish whether colonicstenting delivers better health outcomesthan emergency surgery. Between March2007 and August 2009, 98 patients withacute obstructive left-sided colorectal can-cer from 25 hospitals in the Netherlandswere randomly assigned in a 1:1 ratio tocolonic stenting as a bridge to elective sur-gery (n=47) or emergency surgery (n=51).

At six months investigators found nodifference between treatment groups inglobal health status (assessed with the QL2subscale of the EORTC quality-of-life ques-tionnaire). Mean global health status was63.0 (SD 23.8) in the colonic stenting groupversus 61.4 (SD 21.9) in the emergencysurgery group (P=0.36). Furthermore, nodifference was recorded for 30-day mor-tality (P=0.89), overall mortality (P=0.84),morbidity (P=0.43) and stoma rates at latestfollow-up (P=0.35). The most common seri-ous adverse events were abscess (three in thecolonic stenting group versus four in theemergency surgery group), perforations (sixversus none), and anastomotic leakage (fiveversus one).

“In this multicentre randomised trial,colonic stenting or emergency surgery didnot have any distinct benefits for global healthstatus, mortality, morbidity, other quality-of-life dimensions, and stoma rates,” conclude

NewsRound


destruction in metastatic disease. Denosumabis the first fully human monoclonal antibodydeveloped to specifically target RANK ligand,a key mediator of osteoclast formation, func-tion and survival. In studies denosumab hasbeen shown to reduce bone resorption,tumour-induced bone destruction andskeletal-related events.

In the current study, led by Karim Fizazifrom the Institut Gustave Roussy (Villejuif,France), investigators from 342 centres in39 countries randomised 1904 men withcastration-resistant prostate cancer and noprevious exposure to intravenous bisphos-phonate in a 1:1 ratio to receive 120 mgsubcutaneous denosumab plus intravenousplacebo (n=950), or 4 mg intravenous zole-dronic acid plus subcutaneous placebo(n=951), every four weeks.

Results show that the median time tofirst on-study skeletal-related event was 20.7months with denosumab versus 17.1 monthswith zoledronic acid (HR=0.82, 95%CI 0.71–0.95; P=0.0002 for non-inferiority; P=0.008for superiority). Adverse events were recordedin 97% of patients on denosumab versus 97%on zoledronic acid, and serious adverse eventswere recorded in 63% of patients on deno-sumab versus 60% on zoledronic acid. Theonly differences were raised rates of hypocal-caemia (13%) and osteonecrosis (2%) in thedenosumab group.

“We have shown that denosumab is bet-ter than the established therapy, zoledronicacid, for the delay or prevention of skeletal-related events in patients with advancedprostate cancer,” write the authors, addingthat two limitations of the studywere that thedouble-dummy design did not allow them toobjectively measure the benefits of subcuta-neous versus intravenous administration, andthat the protocol prevented them from assess-ing treatment benefits in patients with severerenal dysfunction at baseline.

In an accompanying commentary,Jeanny Aragon-Ching from George Wash-ington University Medical School, describesthe advantages of using denosumab overzoledronic acid.

“Denosumab is easier to give (subcutaneous)than is zoledronic acid, allowing for shortervisit times and applicability in various physi-cians’ office settings by removing the need foran infusion clinic. Furthermore, denosumabreduces the need for management of acutephase reactions and renal monitoring or doseadjustments, although caution should beexercised with patients who have poor base-line kidney function.”

Further quality-of-life and pain responsedata, she adds, would have been helpful, sincefatigue, bone pain and asthenia were reportedalmost equally in both groups.

� K Fizazi, M Carducci, M Smith et al.

Denosumab versus zoledronic acid for treatment

of bone metastases in men with castration-

resistant prostate cancer: a randomised, double-

blind study. Lancet, 5 March 2011, 377:813–822

� J Aragon-Ching. Unravelling the role of

denosumab in prostate cancer. ibid pp 785–786

Success of spermretrieval depends onchemotherapy used� Journal of Clinical Oncology

Sperm retrieval using testicular spermextraction (TESE) coupled with intracyto-

plasmic sperm injection (ICSI) resulted insperm retrieval in 37% of patients who hadundergone previous chemotherapy, reporteda US study. The study – representing thelargest series of postchemotherapy microdis-section TESE–ICSI yet – found that the successof fertility techniques was related to the typeof cancer that patients had originally beendiagnosed with.

Advances in chemotherapy have led togreater longevity for young men, with thepreservation of fertility and paternity becom-ing increasingly important as a quality of lifeissue. It has been estimated that up to two-thirds of men undergoing chemotherapyremain persistently azoospermic (no measur-

able level of sperm) after treatment. Whilemen rendered persistently azoospermic havetraditionally been considered sterile andreferred for adoption or use of donor sperm,there is growing recognition that fertility canbe salvaged with TESE–ICSI.

In the current study Peter Schlegel andcolleagues, from the New York PresbyterianHospital (US), retrospectively identified 73patients with persistent postchemotherapyazoospermia from a series of testicular spermextraction procedures performed betweenJune 1995 and December 2009 by a single sur-geon in 892 patients. The results show thatspermatozoa were retrieved in 37% ofpatients (27 of 73), with an overall spermretrieval rate of 42.9% (36 of 84). This resultedin a 57.1% fertilisation rate per injected oocyteand an overall live birth rate of 42%. Alto-gether there were 15 deliveries involving atotal of 20 children.

When the sperm retrieval rate was strat-ified according to indications for chemother-apy, the highest retrieval rates were seen inpatients with testicular cancer (85.7%), fol-lowed by neuroblastoma (50%), leukaemia(50%), non-Hodgkin’s lymphoma (36.4%),Hodgkin’s lymphoma (25.9%) and sarcoma(14.3%). “Sarcoma patients tended to have thelowest sperm retrieval rate due to high ratesof exposure and higher doses of alkylatingagents,” write the authors.

With the mean time elapse since chemo-therapy of 18.6 years, this led the authors toquestion whether sperm retrieval closer tothe time of chemotherapy might have led toa higher success rate.

“Our data demonstrates that many menwith long-term azoospermia afterchemotherapy can still have their fertility sal-vaged with the use of assisted reproductivetechniques,” conclude the authors.

� W Hsiao, PJ Stahl. EC Osterberg et al.

Successful treatment of post chemotherapy

azoospermia with microsurgical testicular sperm

extraction: the Weill Cornell experience. JCO

doi: 10.1200/JCO.2010.33.7808, published

online 14 March 2011

Unshackling progressin the care of childhood cancers

� Marc Beishon

Youngcancer patients face a specific set of problems that canonly be resolved throughaconcerted

and coordinated effort by national and EU policy makers, researchers, regulators, funders and

service providers.A meeting held in the run up to International Childhood Cancer Day reviewed

howwell we are doing, andwhat is urgently needed to do better.

For anyone unsure that therereally is overbearing regulationoncancer research inEurope,a visit to any gathering of pae-diatric oncologists and others

involved with child cancers would soonput themstraight. In thewords of one sen-ior clinician: “Wehaveaclinical trialsdirec-tive that allows national re-interpretation,no platform for European approval, oneset of rules that applies to all types of study,no adaptation to risk, overwhelmingbureaucratic burden and it has been con-quered by regulatory fundamentalists.”

So said Stefan Bielack, medical direc-tor ofpaediatriconcologyatStuttgart’sOlgachildren’s hospital, speaking at a stake-holder meeting held at the European Par-liament in Brussels ahead of InternationalChildhood Cancer Day, and hosted bySlovenian MEP Alojz Peterle, himself an

adult cancer survivorwhohashelped restarttheMEPsAgainst Cancer (MAC) group.

The fundamentalists, Bielackexplained later toCancer World, are thosewho strictly follow the regulatory rules tobeabovecriticism,butgrowat theexpenseof the ‘rationalists’,whoexercise judgementin thepursuitofbetterprogress.The terms,he adds, are those ofDavidStewart at theMDAnderson in theUS, and colleagues,commenting on what they see as dimin-ishing returns from the narrow and dys-functional ‘efficacy versus safety’approachin clinical cancer research in general (formore on this see Equipoise lost: ethics,costs and regulation of cancer clinicalresearch JCO 28:2925–2935).

But for paediatric oncologists likeBielack,working inanevenmorecomplexregulatory regime than in the US, thestraitjacket of clinical trial regulationshas

reachedabsurdproportions for childhoodcancers,which rely almost totally on inves-tigator-driven research, given that there isa limitedmarket to interest pharmaceuti-cal companies. “There is toomuchgarbageto too many recipients,” he said, in refer-ence to theseeminglyunendingcascadeofpaperwork to meet the varied require-ments of a wide range of organisationsthat can play by different rules.

Developing cancer drugs and refiningtheiruse inchildren is essential, saidGillesVassal, head of clinical and translationalresearch at theGustaveRoussy Institute,pointing to themajor role that chemother-apy has played in reaching the 80% curerate over the last 50 years. “We need tointroduce more safe and effective drugsinto standardcare,” he said, “and there aresuch drugs in development – about 800now for adults – but children are denied


Spotlighton...

Spotlighton...


tion for an orphan disease,” she said.The message is clear – that there are

centresandnetworksacrossEuropewhichcoulddomuchmore if theyhad access tomorenewdrugs and improvedprofiling ofthemanyunlicensedones alreadyused inpaediatriconcology.Vassal talkedabout thehopespinnedon theEuropeanpaediatricregulation of 2007, which requires phar-maceutical companies to submitnewadultoncology drugs for paediatric investiga-tion plans (PIPs) to the European Medi-cineAgency (EMA). “But four years later,where are we?Yes the process is in place,butonly23oncologydrugshaveaPIPandnot all of thesewill be completed.We arenot seeing an increase so far in the num-

access to them,which is an issue not justfor oncology but for all paediatrics.”

This is not for lack of trying on behalfof the paediatric oncology community,commentedRuthLadenstein,presidentofSIOPE, the European Society for Paedi-atricOncology.Themajority ofEuropeanchildren with cancer are treated in trials,she said, andmultidisciplinary approachesto treatment havebeen important in driv-ing the cure rate to its present high level.“Wehavemore than250 specialised cen-tres around Europe and we’ve been net-working since the late 1960s.About 50%of children are treated in phase I to IIItrials and 30% in standard treatmentapproaches with prospective studies, butless than 5% are in pharma-sponsoredtrials.”Also important, she added, are themanyhigh-level research teamsdedicatedto tumourbiology. “This is a unique situa-

berofdrugs inearly-phasepaediatric stud-ies in the European Union – there arefewer than tennow,while in theUS therearemore than 30.”

NO STRATEGYFOR DRUG DEVELOPMENTAt present, pharmaceutical companiesseepaediatricdevelopment as a regulatorycompliance issue inEurope rather than astrategic research priority, he said, andthere is no role for cooperative groupsbeyond contributions from individualexperts. “Europe lacks a strategy for drugdevelopment forchildren,”headded,com-paring the situation with the US, wheresince 1997 theNational Cancer Institute

Drug A or drug B? Europe’s paediatric oncologists are leading efforts to address the many obstacles todeveloping evidence on the best way to treat young patients like this one; most are still being treatedwith therapies that have never been approved for their particular indication

SIO

PE

has funded a programme fordrug companies to make prod-ucts available to cooperativegroups for paediatric trials. Asa result, major opportunitiesto address childhood cancerthrough thePIPprogrammearebeingmissed.

Childhood cancer resear-cherswill push formorestrategicuse of the European paediatricregulation (and PIPs), and ofcourse for the reformof theclin-ical trialsdirective,whichshouldhappen in some form next year.Bycoincidence, on the samedayof the meeting in Brussels theEuropeanCommission issueda‘conceptpaper’containinga ‘pre-liminary appraisal’ of the mostsuitableways to address someofthekeyconcerns in thedirective,such as how risk is determined.

Jan-Willem van de Loo, sci-entificofficer for cancer researchin thehealth sectionat theEuro-peanCommission,wasnot ableto comment on the directive’sreform, but he did provide anoverviewof theEU’s commitment to sup-porting research andcare through the var-ious framework programme (FP) projectsand networks.

Most notable, in the area of paediatriconcology, isENCCA(EuropeanNetworkforCancer inChildren andAdolescents),a four-year FP7 programme coordinatedby Ruth Ladenstein that aims to buildsustainable research via a ‘virtual institute’across Europe (for more on both Laden-stein and ENCCA see Cancer WorldMarch/April 2011).

Others include collaborative research

projects such as PROTHETS,which looked at prognosticmarkers and therapeutic targetsin Ewing’s sarcoma, and Pan-Care, which is building a data-base on long-term childhoodcancer survivors to lookat trendssuch as late-effects.

Van de Loo highlighted theexplicit focus in FP7 on investi-gator-drivenclinical trials, andontrials to obtainmarketing autho-risation for paediatric use of off-label drugs – a big gap in therecentEUpaediatric regulationaccording to Ladenstein. Oneexample is theworkof theEuro-pean Paediatric Oncology Off-Patent Medicines Consortium(EPOC),which isexamining thepharmacokineticsofdoxorubicin– a drug that is widely used inpaediatric oncology, despite thescarcity of data on correct dosesfor young children.

Another helpful develop-ment has been the establish-ment of aEuropeanNetwork of


Spotlighton...

Major opportunities to address childhood cancer

through the PIP programme are being missed

A success story. Diagnosed quickly,referred to the right specialist centre,treated effectively – Olivia Ferrary describedher experience of having a rare kidneycancer to show the meeting what allchild cancer services should aspire to

HIS

PAPH

OTO

GRAPH

Y

The ear of thePresident. Jerzy Buzek,

President of theEuropean Parliament,

was among thoseattending the SIOPE

conference. He ispictured here (right),

with fellow PolesSidonia Jędrzejewska

MEP (centre),and child cancer

specialist PiotrCzauderna (left)

PaediatricResearch runby theEMA, andtasked with promoting collaboration, as itis primarily a ‘network of networks’.

FUNDING REMAINS A BARRIERBut oncologists such as Vassal are scepti-cal that thecurrent frameworkprogrammewill delivermore ‘calls’ for cancer researchfunding, and Richard Sullivan, from theCentre forGlobalOncoPolicy inLondon,noted that a new report he has co-writtenon the state of child cancer research inEurope (see box) shows that fundingremains short-term and ‘fragile’, and sup-port in somemember states is poor. “Newmechanismsareneeded forcomplex trans-lational research infrastructure –weneedto innovate all the time,” he said.

The need to unshackle the researcheffort is, however, onlyhalf the story. JerzyKowalczyk, fromthechildren’s hospital inLublin, Poland, talked of the need toimprove the standards of care acrossEurope.AsymposiuminLublin twoyearsago laid the basis for SIOPE to drawup aset of minimum European standards ofcare for children with cancer, and a proj-ect to identify besthealthcarepractices inpaediatric oncologyhasnowstartedunderthe auspices of the European Partner-ship forActionAgainstCancer.Next stepsincludepreparingnational versions of thestandards, convincing national agenciesand the EU to issue regulations, andbuilding a registry of child cancer centres.

Kowalczyk expressed disappointmentthat “politicians showed little interest” inthe 2009 meeting, but there is an oppor-tunity to put that right at the EuropeanStandards ofCare forChildrenwithCan-cer conferenceon20–21October inWar-saw this year, led by thePolishMinistry ofHealth under Poland’s EU presidency.Jolanta Kwaśniewska, President of theCommunicationwithoutBarriers founda-tion, and a former ‘first lady’of Poland, is aleading supporter of the meetings and ofchild cancer clinics in her country.

Present at the Brussels meeting were

representatives of the thousands of childand teenage cancer survivors and theirparents for whom good-quality servicesand unhindered progress in developingnewtherapies are so important.OliviaFer-rary talked of her experience of being suc-cessfully treated for a rare formof renal cellcarcinomaatGreatOrmondStreet hospi-tal in London. A video was also shown ofteenagers, which came from Jimmy-teens.tv, aproject startedatSt James’s hos-

pital in Leeds, UK, where young peoplewith cancer are given cameras to recordtheir experiences. There are 600 suchvideos now from theUK and Ireland, andthe producer, Claire Pope, is looking toincludemore from other countries.

The term ‘therapeutic orphan’was firstcoinedback in1968 todescribe the lackofdrug development for children, but theredoes finally seemtobeconcertedaction toimprovematters substantially.

Spotlighton...


The state of paediatric research‘The state of research into children with cancer across Europe: new policies for a newdecade’ is a research report with input frommore than 30 leading European paediatriconcologists, ledbypast SIOPEpresidentKathyPritchard-Jones, and fundedby theEUEurocancercomsproject. It looksat the fundingandextentofpaediatriconcology inEuro-pean countries and also compares the effort with the rest of theworld.Findings include:� In Europe, Sweden and the Netherlands have done the most basic paediatric

oncology research but the differences between countries are not large� Papers from theNetherlands are themost cited, followedby those from theUS, the

UK and Sweden� There is relatively little collaborationbetweenNorthAmerica andEurope.However,

EUmember states arecollaborating increasinglywitheachother, especiallyGermanyand theNetherlands, and also Switzerlandwith France,Germany and Italy

� In most European countries except Spain, private non-profit funding sources out-number government support, but almost half thepapers boreno acknowledgement– “amarker of fragile, short-term funding”

The report includes snapshotsof countries fromexperts, finding forexample thatno inter-national trial hasopened inPolandsince2007; in Italy efforts arebeingmade tocutdownthe largenumber of centres (54) seeing child cancer, someofwhichhave fewer than10patients a year; and those countries that do have strong government funding includeFranceandGermany,whereas theUKandSweden relymoreoncharitableorganisations.A survey of opinion leaders done for the report revealed the following to be priorities:� AdequateEUfunding to support aEurope-wide clinical trials network to assistwith

testing and dissemination of novel therapies and techniques� A reduction of EU trial bureaucracy/regulations to remove barriers to investigator-

led clinical trials, which could include a European trials bureau� Better understandingby regulatorypolicymakers of the level of risk for childrenpar-

ticipating in trials (currently overestimated by insurers as well)� The creation of a European parent/survivor organisation and a common European

information portal� The creation of a European childhood cancer epidemiological registry� EU support for harmonising of treatments through pan-European guidelines.The report is at www.eurocancercoms.eu


Focus

ISSACROSE/ALAMY

Think yourself betterAlternativemedical treatments rarely work. But the placebo effect they induce sometimes does.

This article, published in The Economist twomonths ago, may have a particular relevance for

oncology, where understanding the science is so important that it can be easy to overlook the

contribution of the human touch.

On May 29th EdzardErnst, the world’s firstprofessor of comple-mentary medicine, willstep down after 18 years

in his post at the Peninsula MedicalSchool, in south-west England.Despitehis job title (and the initial hopes ofsome purveyors of non-mainstreamtreatments), Dr Ernst is no breathlesspromoter of snake oil. Instead, he andhisresearch group have pioneered the rig-orous study of everything fromacupunc-ture and crystal healing to Reikichannelling and herbal remedies.Alternativemedicine is big business.

Since it is largelyunregulated, reliable sta-tistics arehard to comeby.Themarket inBritain alone, however, is believed to beworth around£210m($340m),with onein five adults thought to be consumers,andsometreatments (particularlyhomeo-pathy) available fromtheNationalHealthService.Around the world, according toan estimatemade in 2008, the industry’svalue is about $60 billion.Over the yearsDrErnst andhis group

have run clinical trials and publishedover 160meta-analyses of other studies.(Meta-analysis is a statistical techniquefor extracting information from lots ofsmall trials that are not, by themselves,statistically reliable.) His findings are


stark. According to his “Guide to Com-plementary and Alternative Medicine”,around95%of the treatments he andhiscolleagues examined– in fields asdiverseasacupuncture,herbalmedicine,homeo-pathy and reflexology – are statisticallyindistinguishable from placebo treat-ments. In only 5% of cases was thereeither a clear benefit above andbeyond aplacebo (there is, for instance, evidencesuggesting that St John’s Wort, a herbalremedy, can helpwithmild depression),or even just a hint that something inter-estingwashappening to suggest that fur-ther researchmight be warranted.It was, at times, a lonely experience.

Moneywas hard to comeby. Practition-ers of alternative medicine becameincreasingly reluctant to co-operate asthe negative results piled up (a row in2005with an alternative-medicine lobbygroup foundedbyPrinceCharles did nothelp),while traditionalmedical-research

bodies saw investigations into thingslikeAyurvedic healing as awaste of time.Yet Dr Ernst believes his work helps

address a serious public-health problem.He points out that conventional medi-cines must be shown to be both safeand efficacious before they can belicensed for sale. That is rarely true ofalternative treatments, which rely on amixture of appeals to tradition and tothe ‘natural’wholesomeness of their prod-ucts to reassureconsumers.That explainswhy, for instance, somehomeopaths canmarket treatments formalaria, despite alack of evidence to suggest that suchtreatments work, or why some chiro-practors can claim to cure infertility.Despite this lack of evidence, and

despite the possibility that somealterna-tive practitioners may be harming theirpatients (either directly, or by convincingthem to forgo more conventional treat-ments for their ailments), Dr Ernst also

Focus

Dr Ernst believes that doctors can usefully learn from the

chiropractors, homeopaths andAscendedMasters

I will help you feel better. Whether or not theirtreatments have any merit, the time and

attention alternative therapists canspend in consultations, and their sense of

assurance and belief in the therapiesthey are proposing, can make a real

difference to the wellbeingof their patients

believes there is something that conven-tional doctors canusefully learn fromthechiropractors,homeopathsandAscendedMasters. This is the therapeutic value ofthe placebo effect, one of the strangestand slipperiest phenomena inmedicine.

MIND AND BODYA placebo is a shammedical treatment– a pharmacologically inert sugar pill,perhaps, or a piece of pretend surgery. Itsmain scientific use at the moment is inclinical trials as a baseline for compari-son with another treatment. But justbecause themedicine is not real does notmean it doesn’t work. That is preciselythe point of using it in trials: researchershave known for years that comparingtreatment against no treatment at allwill give a misleading result.Giving pretend painkillers, for

instance, can reduce the amount of paina patient experiences.A study carried outin 2002 suggested that fake surgery forarthritis in the knee provides similarbenefits to the real thing.And the effectscan be harmful as well as helpful.Patients taking fake opiates after havingbeenprescribed the real thingmay expe-rience the shallow breathing that is aside-effect of the real drugs.Besides beingbenchmarks, placebos

are a topic of research in their own right.On May 16th the Royal Society, theworld’s oldest scientific academy, pub-lished a volume of its Philosophical

Transactions devoted to the field.One conclusion emerging from the

research, says Irving Kirsch, a professorat Harvard Medical School who wrotethe preface to the volume, is that theeffect is strongest for those disordersthat are predominantlymental and sub-jective, a conclusion backed by a meta-analysis of placebo studies that wascarried out in 2010 by researchers attheCochraneCollaboration, an organi-sation that reviews evidence formedicaltreatments. In the case of depression,says Dr Kirsch, giving patients placebopills can produce very nearly the sameeffect as dosing them with the latestantidepressant medicines.Pain is another nerve-related symp-

tomsusceptible to treatment by placebo.Here, patients’ expectations influencethe potency of the effect. Telling some-one that you are giving him morphineprovides more pain relief than sayingyou are dosing him with aspirin – evenwhenboth pills actually contain nothingmore than sugar. Neuro-imaging showsthat this deception stimulates the pro-duction of naturally occurring painkillingchemicals in the brain.A paper inPhilo-sophical Transactions byKarinMeissnerof Ludwig-Maximilians University inMunich concludes that placebo treat-ments are also able to affect the auto-nomic nervous system, which controlsunconscious functions such as heart-beat, blood pressure, digestion and the

like. Drama is important, too. Placeboinjections are more effective thanplacebo pills, and neither is as potent assham surgery. And the more positive adoctor iswhen telling a patient about theplacebohe is prescribing, themore likelyit is to do that patient good.Despite the power of placebos,many

conventional doctors are leery of pre-scribing them. They worry that to do sois to deceive their patients. Yet perhapsthe most fascinating results in placeboresearch – most recently examined byTedKaptchuk andhis colleagues atHar-vard Medical School, in the context ofirritable-bowel syndrome – is that theeffectmaypersist even if patients are toldthat they are getting placebo treatments.Unlike their conventional counter-

parts, practitioners of alternative medi-cine often excel at harnessing theplacebo effect, saysDrErnst. They offerlong, relaxed consultations with theircustomers (exactly the sort of “goodbed-side manner” that harried modern doc-tors struggle to provide).And theybelievepassionately in their treatments, whichare often delivered with great and reas-suring ceremony. That alone can beenough to do good, even though themagnets, crystals and ultra-dilute solu-tions applied to the patients are, bythemselves, completely useless.

This article was first published in The Economist on19 May 2011. It is reprinted here with permission.© The Economist Newspaper Limited London (2011)


Focus

Neuro-imaging shows that this deception stimulates the

production of naturally occurring painkilling chemicals

They offer the sort of long, relaxed consultations

that harried modern doctors struggle to provide

cancer world 43

Documents

cancer worldshould

new cancer drugs

andmaintain good health

good health literacycan

poor health literacy

increasingpatients survival

edgebeyond survival

rights reservedcancer