Interrogating Metastatic Prostate Cancer Treatment Switch Decisions: A Multi-institutional Survey

(1)

Prostate

Cancer

Interrogating

Metastatic

Prostate

Cancer

Treatment

Switch

Decisions:

A

Multi-institutional

Survey

David

Lorente

a,b,y

_,

_Praful

_Ravi

a,y

_,

_Niven

_Mehra

a

_,

_Carmel

_Pezaro

c

_,

_Aurelius

_Omlin

d

_,

Alexa

Gilman

e

,

Miguel

Miranda

e

,

Pasquale

Rescigno

a

,

Michael

Kolinsky

a

,

Nuria

Porta

e

,

Diletta

Bianchini

a

,

Nina

Tunariu

a

,

Raquel

Perez

a

,

Joaquin

Mateo

a

,

Heather

Payne

f

,

Leon

Terstappen

g

,

Maarten

IJzerman

g

,

Emma

Hall

e

,

Johann

de

Bono

a,

*

a_Prostate_Cancer_Targeted_Therapy_Group_and_Drug_Development_Unit,_The_Royal_Marsden_NHS_Foundation_Trust_and_The_Institute_of_Cancer_Research,_Sutton, UK;b_Medical_Oncology_Service,_Hospital_{Universitario}_La_Fe,_Valencia,_Spain;c_Monash_University _Eastern_Health_Clinical_School,_Melbourne,_Australia; d_Department_of_Oncology_and_Haematology,_{Kantonsspital}_St.Gallen,_St.Gallen,_Switzerland;e_Clinical_Trials_and_Statistics_Unit,_The_Institute_of_Cancer_Research, London,UK;f_Department_of_Clinical_Oncology,_University_College_London_Hospital,_London,_UK;g_University_of_Twente,_Twente,_The_Netherlands

a v ai l a b l e a t w w w . s c i e n c e d i r e c t . c o m j o u r n al h o m e p a g e : w w w . e u r o p e an u r o l o g y . c o m / e u f o c u s Articleinfo Articlehistory: AcceptedSeptember17,2016 AssociateEditor: JamesCatto Keywords: Castration-resistantprostate cancer Treatmentswitch Response Progression

Circulatingtumourcells Abiraterone

Abstract

Background: Evaluationofresponsestotreatment formetastaticcastration-resistant

prostatecancer (mCRPC)remains challenging.Consensuscriteriabasedon

prostate-specificantigen(PSA)andclinicalandradiologicbiomarkersareinconsistentlyutilized.

Circulatingtumorcell(CTC) countscaninformprognosisand response,but arenot

routinelyused.

Objective: Toevaluatetheuseofbiomarkersandtrendsinclinicaldecision-makingin

currentmCRPCtreatment.

Design, setting,andparticipants: A 23-part onlinequestionnaire wascompletedby

physicianstreatingmCRPC.

Outcomemeasuresandstatisticalanalysis: Resultsarepresentedastheproportion(%)

ofphysiciansrespondingtoeachoftheoptions.Weusedx2andFisher’steststocompare

differences.

Resultsandlimitations: Atotalof118physicians(22.1%)responded.Ofthese,69.4%

treated50mCRPCpatients/year.Morephysiciansadministeredfourorfewercourses

ofcabazitaxel(27.9%)thanfordocetaxel(10.4%),withnosigniﬁcantdifferenceinthe

numberofcoursesbetweenbone-onlydiseaseandResponseEvaluationCriteriainSolid

Tumours(RECIST)–evaluabledisease.Some74.5% ofrespondentsconsidered current

biomarkers useful for monitoring disease, but only 39.6% used theProstate Cancer

WorkingGroup(PCWG2)criteriainclinicalpractice.PSAwasconsideredanimportant

biomarkerby55.7%,butonly41.4%discardedchangesinPSAbefore12wk,andonly

39.4% were able to identify bone-scan progression according to PCWG2. The vast

majority of physicians (90.5%) considered clinical progression to be important for

switching treatment. The proportion considering biomarkers important was 71.6%

forRECIST,47.4% forbonescans,23.2%forCTCs, and21.1%forPSA.Although53.1%

acknowledged that baseline CTC counts are prognostic, only 33.7% would use CTC

changes alone to switch treatment in patients with bone-only disease. The main

challengesin usingCTCcountswereaccesstoCTCtechnology (84.7%),cost (74.5%),

anduncertaintyoverutilityasaresponseindicator(58.2%).

y_These_authors_contributed_equally_to_this_work.

* Correspondingauthor.ProstateCancerTargetedTherapyGroup,RoyalMarsdenNHSFoundation Trust,SectionofMedicine,TheInstituteofCancerResearch,DownsRoad,Sutton,SurreySM25PT,UK. Tel.+442087224029.

E-mailaddress:johann.de-bono@icr.ac.uk(J.deBono). http://dx.doi.org/10.1016/j.euf.2016.09.005

2405-4569/#2016EuropeanAssociationofUrology.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http:// creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

1. Introduction

The past decadehas seenan increase in the therapeutic armamentarium againstmetastatic prostate cancer,with agents proving survival benefit both in the castrate-resistant(mCRPC)[1–7]andcastration-naı¨vestages[8,9]

of the disease. This increased availability of treatment options necessitates improved biomarkers to determine treatmentresponsesmorerapidlyandfacilitateoptimised decisionsontherapeuticsequencing[10].

Prostate-specific antigen (PSA), bone scans, and Re-sponse Evaluation Criteria in Solid Tumours (RECIST) criteriaare commonlyutilized toevaluateresponses and are recommended as outcome measures by the Prostate Cancer Working Group (PCWG2) for clinical trials

[11]. However, these biomarkers have significant limita-tions.Inparticular,PSAandbonescansdonotallowearly responseassessment,andnoneofthebiomarkersprovide patient-level surrogates of clinical benefit [12,13]. This challengeiscompoundedbythelackofRECIST-evaluable diseaseinasubstantialproportionofpatients[14].Fordaily clinical practice, existing guidelines do not recommend specifictreatment monitoring,anissueaddressed by the AdvancedProstateCancerConsensusconference[15].

Thelack ofadequatebiomarkersmayimpact thedose intensityofchemotherapyandotheranticancer(hormonal, radiopharmaceutical)agents administered indaily clini-cal practice. Thefact that determining disease progres-sion in the absence of clear clinical deterioration is impossible before 12wk(owingtothe possibility of an earlyPSAorbone scan‘‘flarereaction’’)inpatients with noRECIST-evaluablediseasemay contributetoboththe administration ofmorechemotherapycycles topatients with bone-only disease (overtreatment) and a higher relianceonPSAchangesforearlytreatment discontinua-tion(undertreatment).

Circulatingtumourcell(CTC)countsareprognosticand areassociatedwithtreatmentresponseinmCRPCpatients, with recent studies indicating value as a patient-level surrogateofsurvival[16,17].Increasingevidencesuggests thatCTCscouldbeutilisedtomonitordiseaseprogressionin mCRPC [18]. However, CTC use is largely limited to academiccentresinthesettingofclinicaltrials.

Weconductedan onlinesurveyof physicianstreating mCRPC. The survey focused on how physicians make treatmentswitchdecisions,opiniononresponseindicators, utilisationof PCWG2 criteriainroutinepractice,and the valueofCTC countsto guidetreatmentswitch decisions. Theresultswillhelptoinformthedesignofaninternational

trialandhealtheconomicevaluationtoimprovetreatment switchdecisionsformCRPCpatientstoimproveoutcomes, decreaseovertreatment,andmaximiseresourceutilisation. 2. Materialsandmethods

A 23-partonline questionnaire,divided in foursectionsasoutlined below,wascompiledbytheauthors(SupplementaryFig.1):

1. Generalquestionsonclinicalpractice.

2. Familiarity with progression criteria for currently established biomarkers.

3. CTCsandtheirassessmentinpatientswithadvancedprostatecancer. 4. Clinicaldecision-makingusingresponseindicators.

E-mailsinvitingparticipationinthe surveyweresentto485UK investigatorsparticipatinginurologiccancerclinicaltrials,29physician members of the GU Group of the Swiss Group for Clinical Cancer Research,and20practisingprostatecancerphysiciansinAustraliaand NewZealand.Alink totheweb-basedsurvey(createdwith Survey-Monkey)wasincluded.

2.1. Statisticalanalyses

Descriptive statistics were used; the proportion (%) of physicians responding to each option is presented. Physicians were classified accordingtothenumberofpatientstheytreated(50vs<50patients/ year)orrecruitedtoclinicaltrials(25%vs<25%),andthenumberof cyclesofdocetaxel/cabazitaxelprescribed(4,5–6,7cycles).No pre-existing evidence wasused in choosingclassification cutoff values. Proportionswerecomparedusingax2testorFisher’sexacttest(forcell frequencies5).Apvalueof0.05wassetasthelimitforstatistical significance.Noadjustmentformultipletestingwasperformed.SPSS version21(IBMIBM,Armonk,NY,USA)wasused.

3. Results

3.1. Participantcharacteristicsandtheirclinicalpractice

Between November 21, 2014 and December 18, 2014, 118 practisingprostate cancerphysicians (22.1%)replied. Sections1,2,3,and4werecompletedby,111,106,98,and 89 physicians, respectively. Most respondents (77.1%) practised in the UK. Nearly 70% treated 50 mCRPC patients/year (Table 1). Most reported prescribing 7–10 coursesofdocetaxeland5–6cyclesofcabazitaxel(Fig.1); therewasnodifferenceinthenumberofcoursesofeither docetaxel ðpðx2

2Þ¼0:519Þ or cabazitaxel ðpðx22Þ¼0:814Þ

administered to patients with RECIST-evaluable disease compared to patients with bone-only disease. Physicians

Conclusions: A signiﬁcant proportion of physicians discontinue treatment for mCRPC

before12wk,raisingconcernsaboutinadequateresponseassessment.Manyphysicians

ﬁndcurrent biomarkersuseful,but mostrely onsymptoms todrivetreatment switch

decisions,suggestingthereisaneedformoreprecisebiomarkers.

Patientsummary: Inthisreportweanalysetheresultsofaquestionnaireevaluatingtools

forclinicaldecision-makingcompletedby118prostatecancerspecialists.Wefoundthat

mostphysiciansfavourclinicalprogressionoverprostate-speciﬁcantigenorimaging,and

thatcriteriaestablishedbytheProstateCancerWorkingGrouparenotwidelyused.

# ₂₀₁₆_European_Association_of_Urology._Published_by_Elsevier_B.V._This_is_an_open_access

(3)

reportedgivingmorecoursesofdocetaxelthancabazitaxel in patients with both RECIST-evaluable and bone-only disease ðpðx2

2Þ<0:001Þ. Physicians with larger patient

practicesprescribedmorecoursesofchemotherapy (Sup-plementaryTable1).

3.2. Evaluationofcurrentlyavailableresponsebiomarkers

Currentguidelinesprovidelittleinstructiononthe evalua-tionofresponsetotreatmentinmCRPC;thisisparticularly challenginginpatientswithonlybonemetastasesandno other measurable disease [15,19]. PCWG2 progression criteria(Supplementary Table2) aremainly used among patients treated within clinical trials. We evaluated the opinion of physicians on currently available biomarkers (PSA,bonescan,andCTCs)formonitoringresponse.Some

79respondents(74.5%)ratedtheseasuseful(71.7%)orvery useful (2.8%). Only39.6% reported using PCWG2 criteria mostorallofthetime,and27.3%reportedrarelyornever using the criteria (Table 2). Physicians recruiting more patientstotrialsweremorelikelytousePCWG2frequently (56% vs 25%; pðx2

2Þ¼0:001Þ = 0.001; Supplementary

Table3).

3.2.1. PSA

Atotalof59respondents(55.7%)reportedthatPSAwasa useful/very useful biomarker for monitoring response to treatment(Table2).WeaskedparticipantstoidentifyPSA progressioningraphicalexamplesshowingconsecutivePSA values to evaluatetheirability to utilizePCWG2 criteria. Only41.4%ofphysicianscorrectlyrecognisedthatatleast 12wkarerequiredtodefinePSAprogression(Fig.2A).Most physicians(84.8%)correctlyidentifiedthata25%increase fromthenadirvalue(confirmedbyasecondvalueatleast 3wklater)constituted progression(Fig.2B).Some 90.9% failedto recognisethatPSAprogression holdsevenifthe confirmatorysecondvalueislowerthanthefirst,providing bothvaluesshowa25%increasefromthenadir(Fig.2C). Only twophysicians (2.0%) answered allthree questions correctly.

3.2.2. Bonescintigraphy

PCWG2criteriadefinebonescanprogressionasaminimum oftwonewlesions,withnewlesionsobservedatthefirst 12-wk reassessment requiring a confirmatoryscan (Sup-plementary Table 2). When respondents were asked to choose from a number of definitions of bone scan progression(selectingmorethanonewaspermitted),only 39.4% answered the correct option (as per PCWG2) and discardedtheincorrectoptions,indicatingdiversityinbone scaninterpretation.

3.2.3. CTCs

Some98%ofrespondentswerefamiliarwiththeconceptof CTCs, but only53.1%recognised thatbaseline CTCs have Table1–Participantcharacteristics

Question(numberofresponses) n(%)

Q1:Specialty(n=118) Oncologist 100(84.7) Urologist 17(14.4) Other 1(0.8) Q2:Practicelocation(n=118) UK 91(77.1) Europe(non-UK) 16(13.6 Australia/NewZealand 11(9.3) Q3:NumberofmCRPCpatients

treatedperyear(n=111)

<10 3(2.7)

10–49 31(27.9)

50–99 48(43.2)

100 29(26.1)

Q4:PercentageofmCRPCpatients

enteredintoclinicaltrials(n=111)

None 6(5.4)

<25% 53(47.7)

25–49% 38(34.2)

50–74% 12(10.8)

75% 2(1.8)

mCRPC=metastaticcastration-resistantprostatecancer.

Fig.1–NumberofcyclesofchemotherapyadministeredtopatientswithResponseEvaluationCriteriainSolidTumours(RECIST)–evaluablediseaseand

bone-onlymetastaticcastration-resistantprostatecancer(mCRPC).ThefiguresummarisesrepliesforQuestions5–8(‘‘Howmanycyclesofdocetaxel/

(4)

prognosticvalue.Similarly,only50.0%and54.1% respon-dentswereawarethatapost-treatmentchangeinCTCswas associatedwithoutcomeinpatientstreatedwith abirater-oneandchemotherapy,respectively(Table2).

Majorchallengesidentifiedbyrespondentsascurrently limitingtheuseofCTCsinprostatecancerwereassaycost (74.5%),pooraccesstoCTCenumerationtests(84.7%),and uncertainty overtheir clinical utility in response assess-ment(58.2%;Table2).

3.3. Clinicaldecision-makinginCRPC

According to PCWG2, clinical progression is defined as worseningpainandanalgesicuse,deterioratingqualityof life, urinary or bowel compromise, or a need for new anticancer therapy. Of these, only worsening pain is associated with outcome in prospective clinical trials

[20]. Almost all physicians (90.5%) considered clinical progressiontobeimportantfordrivingtreatmentswitches. Table2–Evaluationofcurrentlyavailablebiomarkers,CTCsanduseofProstateCancerWorkingGroup(PCWG2)criteriainmCRPC

Question(numberofresponses) n(%)

Q9:Suitabilityofcurrentlyavailablebiomarkers(PSA,bonescans,CTCs)inmonitoringdiseaseinmCRPC(n=106)

Veryuseful 3(2.8)

Useful 76(71.7)

Notveryuseful 25(23.6)

Poor 2(1.9)

Q11:SuitabilityofPSAasachemotherapyresponsemarkerinmCRPC(n=106)

Veryuseful 3(2.8)

Useful 56(52.8)

Notveryuseful 44(41.5)

Poor 3(2.8

Q10:UseofPCWG2criteriafordecision-makingwhentreatingpatientswithmCRPC(n=106)

Always 3(2.8)

Mostly 39(36.8)

Sometimes 35(33)

Rarely 12(11.3)

Never 17(16)

Q14:FamiliarwiththeconceptofCTCs(n=98)

Yes 96(98)

No 2(2)

Q15:BaselinenumberofCTCsatstartofchemotherapyisprognosticforoverallsurvivalinmCRPC(n=98)

Yes 52(53.1)

No 0(0)

Unsure 46(46.9)

Q16–17:ChangeinnumberofCTCsisassociatedwithresponseinmCRPCduring(n=98):Chemotherapy

Yes 53(54.1) No 0(0) Unsure 45(45.9) Abiraterone Yes 49(50) No 0(0) Unsure 49(50)

Q18:ChallengesassociatedwithuseofCTCsinprostatecancer(n=98)

Cost 73(74.5)

Lackof/uncertaintyaboutprognosticsigniﬁcance 43(43.9)

Lackof/uncertaintyaboutpredictiveinformationontreatmentresponse 57(58.2)

DifﬁcultyininterpretingchangesinCTCnumber 41(41.8)

PooraccesstoCTCenumerationtechnology 83(84.7)

Other 4(4.1)

Q20:LikelihoodofswitchingorstoppingchemotherapyinanasymptomaticmCRPCpatientwithPSAincreaseat12wkandnoradiologicprogression(n=95)

Deﬁnitely 0(0)

Likely 16(16.8)

Unlikely 70(73.7)

Deﬁnitelynot 9(9.5)

Q21:LikelihoodofswitchingorstoppingabirateroneorenzalutamideinanasymptomaticmCRPCpatientwithPSAincreaseat12wkandnoradiologic

progression(n=95)

Deﬁnitely 0(0)

Likely 9(9.5)

Unlikely 68(71.6)

Q23:LikelihoodofusingCTCchangesalone,independentlyofPSAorbonescanﬁndings,inguidingdecision-makingtoswitchorstoptherapyinanmCRPC

patientwithbone-onlydisease(n=89)

Deﬁnitely 1(1.1)

Likely 29(32.6)

Unlikely 55(61.8)

(5)

Some71.6%and47.7%feltRECISTandbonescan progres-siontobeimportant,andonly23.2%and21.1%feltCTCand PSAprogressiontobeimportant,respectively.

Overall, 55.7% considered PSA useful/very useful in guidingtherapy,butonly21.1%considereditimportantfor decision-making (Fig. 3). Physicianswho considered PSA and bone scans important/very important for decision-makingdid not havea betterunderstanding ofresponse criteria(Supplementary Table 4).Only 30%ofphysicians whoconsideredPSAimportant/veryimportantinguiding

treatment switches acknowledged that at least12 wkis neededtodefinePSAprogression(SupplementaryTable4). In thecaseof anasymptomaticmCRPC patient witha rising PSA at 12 wk but no evidence of radiologic progression, mostphysicianswere unlikelyto switch/stop chemotherapy(83.2%)orabiraterone/enzalutamide(90.5%). Only33.7%ofrespondentswerereadytouseCTCchanges alone, independently of PSA or bone scans, to guide switching/stopping therapy in patients with bone-only disease; among those who acknowledged the value of

Fig.2–Evaluationofprostate-specificantigen(PSA)progressioncriteria.ThefiguresummarisesrepliesforQuestion12.Participantswereshownthree

differentPSAbiomarkerscenariosforpatientswithbone-onlydisease.ThepercentageofparticipantswhobelievedthescenariocorrespondedtoPSA

(6)

CTCsasaresponse-biomarker,theproportionwas43.5%. Of those who were likely/very likely to switch on CTC changes alone independently of PSA or bone scans, a largerproportionwerephysicianswhofeltthatcurrently available biomarkers are not very useful/poor in moni-toringdisease(p=0.03;SupplementaryTable5).Among those who were unlikely/unwilling to switch on CTC changes alone, 57.6% cited uncertainty over predictive informationontreatment responseasachallengein use of CTCs, with 52.5% and 42.4% citing uncertainty over prognosticsignificanceanddifficultyininterpretingCTC changes,respectively.

3.4. Treatment switches in mCRPC

The final part of thequestionnaire askedrespondents to consider scenarios involving clinically stable mCRPC patientswith bone-onlydisease.Fora>25%PSArisebut aCTCdeclineto<5cells/7.5ml(‘‘favourable’’CTC conver-sion)andastablebonescanat12wk,92.1%ofrespondents wouldnotswitch/stoptherapy(Fig.4A).Theproportionfell to68.5%if thebonescanshowedincreasedtraceruptake butnonewlesions(Fig.4B).Fora50%fallinPSAbutaCTC riseto5cells/7.5ml(‘‘unfavourable’’CTCconversion)at 12 wkandstable disease according to a bonescan, only 11.2%wouldswitch/stoptherapy(Fig.4C).Fora50%PSA decline and CTC conversion from ‘‘unfavourable’’ to ‘‘favourable’’ count at 12 wk, but two new lesions on a bonescan,mostrespondents(70.8%)reportedtheywould notswitch/stoptherapy(Fig.4D).

Respondents who believed that post-treatment CTC changes were associated with treatment response were morelikelytoswitch/stoptherapyonCTCprogressionasin

Figure 4C (p=0.023), and were more likely to continue

treatmentwithCTCresponseasinFigure4B(p=0.003)and

Figure4D(p=0.005;SupplementaryTable3).

4. Discussion

Itisimperativethatmorepreciseresponsebiomarkersthat canguidemorerapididentificationofdrugresistanceand treatment termination are developed to minimise the overtreatment of patients with ineffective therapies, decreasethetoxicityofineffectivetreatment,andmaximize the utilisationof resources.Weconductedthis surveyto evaluate current practice in clinical decision-making by physiciansspecialisedinthetreatmentofCRPC.Ourresults highlightdifficultiesintheapplicationofcurrent biomark-ers inthetreatmentofadvancedprostatecancerindaily clinicalpractice.

Are physicians giving too muchchemotherapy, ortoo little? The optimumnumber of chemotherapy coursesis unclear.IntheTROPICtrial,althoughamaximumnumber oftencyclesofchemotherapywasallowed,amedianofsix courseswas reported,and28%ofpatients completedten courses [7]. This is similar to numbers reported in expanded-access programmes [21,22]. In TAX-327, in whichthenumberofcyclesofdocetaxel wasnotlimited to ten,the mediannumberof cyclesinthethree-weekly docetaxelarmwas9.5[23].Oursurvey,however,indicates that a significant number of physicians discontinue treatment beforefour courses(12 wk) oftreatment;this isespeciallytrueforcabazitaxel.Accordingtooursurvey, early discontinuation does not appear to be related to radiologic disease progression, since no difference in the number of chemotherapy courses between RECIST-evaluableandbone-onlydiseasewasreported.

Fig.3–Importanceofdifferentbiomarkersinclinicaldecision-making(stoppingtherapy)inmetastaticcastration-resistantprostatecancer.Thefigure

summarisesrepliesforQuestion19.Participantswereaskedtorankeachofthedifferenttypesofdiseaseprogressionlistedfrom1(extremely

important)to6(notatallimportant)intheirclinicaldecisionstoswitchorstoptherapy.RECIST=ResponseEvaluationCriteriainSolidTumours;

(7)

Fig.4–Decision-makingfordifferentbiomarkerscenarios.ThefiguresummarisesrepliesforQuestion22.Participantswereshownfourdifferent

biomarkerscenarioscombiningprostate-specificantigen(PSA),circulatingtumorcells(CTCs),andbonescanfindingsforclinicallystablepatients.The

(8)

How familiarare physicians with consensus response criteria?Inoursurvey,althoughmostphysicians consider-ing currently available biomarkers (74.5%), and PSA in particular (55.6%), to be useful for monitoring disease, knowledge of the specificPCWG2 criteria is suboptimal. PCWG2requiresaconfirmatoryvalueatleast3wkaftera firstprogressingPSA,andrecommendsdiscardinganyearly (before12wk)PSAincreaseowingtothepossibilityofPSA ‘‘flare’’,reportedin16.7%ofpatientsinTAX-327[24].Inour survey,many physiciansfailedto acknowledgethe possi-bilityofaPSAflareinevaluatingPSAprogression.

Concerns regarding the interpretation of bone-scan imaging were also identified. Only around 40–60% of mCRPC patients are evaluable according to RECIST, with many patients having bone-only disease [14]. PCWG2 criteriaindicatethatbonescanscan onlybeusedforthe assessmentofprogressionandnotresponse.Newlesionsat the first 12-wk assessment require a confirmatory scan, sinceearlybone-scan‘‘flare’’isnotuncommon[25].Only 39.4% of respondents followed the PCWG2 definition of bonescanprogression,despiterecentstudiesindicatingan associationbetweenradiographicprogression-freesurvival (combininga bone scan andRECIST) and survival in the COU-302phase3trial[26].

These findings suggest that decisions to switch treat-ment are challenging for physicians treating advanced prostatecancer.PCWG2guidelinesacknowledgedifficulties in assessing progression according to clinical symptoms alone because of ‘‘subjectivity’’ [11]; however, this was overwhelmingly acknowledged as the most important determinant of disease progression in routine practice. RECISTcriteriarankedsecondinimportance,despitebeing useful for only some patients. Interestingly, only 39.6% commonlyusePCWG2criteriaforclinical-decisionmaking. Whenconfrontingphysicianswithclinicalscenariosbased on CTC, PSA and bone scan information no significant predominance of one biomarker was found. Physicians generallycontinuedtreatmentinthefaceof‘‘contradictory’’ biomarker information (ie, rising CTCs with falling PSA; fallingCTCswithrisingPSA;orfallingCTCsandPSAwith new lesions on bone scan), for which current European Association ofUrology and EuropeanSocietyfor Medical Oncologyguidelinesdonotofferclearrecommendationson optimal decision-making. Importantly, we observed no significantdifferencesinthefamiliaritywithPSAorbone scan progression criteria (questions 12 and 13), the importance of each ofthe biomarkers in the decision to switchorstop therapy(question19),orthelikelihoodof switchingorstoppinginthefaceofthedifferentproposed biomarker scenarios (question 20) between physicians treating in high-volume centres (50 patients/yr) and thoseinlow-volumecentres(<50patients/yr).Thesedata suggesta needformore precise biomarkersto reporton response andprogression, sincepatients today appearto continuereceivingtreatmentdespitebiomarkersindicating alackofresponse.

CTCcountholdspromiseasaresponsebiomarker,with well-establishedprognosticutilitythathasbeenvalidated prospectivelywithchemotherapy[16,27],abiraterone[17],

andenzalutamide[28].Acombinationoflactate dehydro-genaseandCTCsisapatient-levelsurrogateofsurvival[17], and post-treatmentchangesare robustlyassociated with outcome [29,30]. Moreover, CTC counts have greater sensitivity andspecificity andinform on outcome earlier thanchangesinPSAdo[30,31].However,onlyhalfofthe responding physicians were familiar with available CTC data,withveryfewpreparedtostopabiraterone(9.5%)or chemotherapy (16.8%) on the basis of CTC progression. Nonetheless, physicians cognisant of available CTC data were more willing to guide treatment according to CTC changes.Costwasreportedasamajorcaveattotheroutine useofCTCs,althoughmostofthiscouldbeeasilyrecouped byearlierdiscontinuationofineffectivetreatment.

Weacknowledgeanumberoflimitationstoourstudy. Thereturnratewas22.1%,andnotallphysicianscompleted theentiresurvey.Reasonsfornotcompletingthesurveyare unknown, although this could be related to the lack of compensation offered. Furthermore, no distinction was madebetweenacademicandnonacademiccentres,andno comparison was made between UK-based and non–UK-basedphysicians.Tomaximisetheyieldofinformationand study participants,thesize ofthequestionnaireincluded onlythreequestionsonbiomarkercriteria,whichmay be insufficienttofullyevaluatephysicianknowledge. 5. Conclusions

Inconclusion,ourdataindicatethatmorepreciseresponse biomarkersandphysicianeducationareneededto interro-gateoutcomeindailyclinicalpracticeinmCRPC,andthatit is likely that many patients are being over- and under-treated. Many physicians rely on the highly subjective reporting of symptoms for treatment switch decisions. Physician education on thesechallenges, andestablished workinggroupcriteria,areneeded,asareprospectivetrials to clinicallyqualify biomarker utility,improve treatment switch decisions and patient outcome as well as change clinicalpractice.

Authorcontributions:DavidLorentehadfullaccesstoallthedatainthe study andtakes responsibilityfor theintegrity ofthe data andthe accuracyofthedataanalysis.

Studyconceptanddesign:Lorente,Ravi,deBono. Acquisitionofdata:Allauthors.

Analysisandinterpretationofdata:Allauthors. Draftingofthemanuscript:Allauthors.

Criticalrevisionofthemanuscriptforimportantintellectualcontent:All authors.

Statisticalanalysis:Lorente,Gilman,Miranda,Porta,Hall,deBono. Obtainingfunding:None.

Administrative,technical,ormaterialsupport:None. Supervision:None.

Other:None.

Financialdisclosures:DavidLorentecertifiesthatallconflictsofinterest, includingspecificfinancialinterestsandrelationshipsandaffiliations relevanttothesubjectmatterormaterialsdiscussedinthemanuscript (eg,employment/affiliation,grantsorfunding,consultancies,honoraria,

(9)

stockownershiporoptions,experttestimony,royalties,orpatentsfiled, received,orpending),arethefollowing:HeatherPaynehasattendedand received honoraria for advisory boards, travel expenses to medical meetings,andservedasaconsultantforAstraZeneca,Astellas,Janssen, Sanofi Aventis, Takeda, Amgen, Ipsen, Ferring, Sandoz, Roche, and Novartis.LeonTerstappenislistedasaninventoronUSpatentsrelated totheCellSearchsystem,therightsofwhichareassignedtoJohnson& Johnson,andisthechairmanoftheDepartmentofMedicalCellBioPhysics attheUniversityofTwente,whichhasreceivedresearchfundingrelated totheCellSearchsystemfromJohnson&Johnson.AureliusOmlinhas servedonadvisoryboardsforAstraZeneca,Astellas,Bayer,Janssen,Pfizer andSanofiAventis,andhasreceivedresearchfundingfromTEVAand JanssenandtravelsupportfromAstellas,Bayer,andSanofiAventis.David LorentehasservedonadvisoryboardsforBayerandhasreceivedtravel supportfromJanssen.DilettaBianchinihasreceivedtravelgrantsfrom Janssen.Theremainingauthorshavenothingtodisclose.

Funding/Supportandroleofthesponsor:None.

Acknowledgments:WeacknowledgesupportfromProstateCancerUK andMovember to theLondon Movember Prostate CancerCentre of ExcellenceatTheInstituteofCancerResearchandRoyalMarsden,and throughan Experimental Cancer MedicalCentre(ECMC) grantfrom CancerResearchUKandtheDepartmentofHealth(C51/A7401).The authorsacknowledgeNHSfundingtothe NIHRBiomedicalResearch CentreattheRoyalMarsdenandTheInstituteofCancerResearch.David LorenteconductedthisworkintheMedicineDoctorateframeworkofthe UniversidaddeValencia.HeatherPayne’sworkwassupportedbythe UCLH/UCLComprehensiveBiomedicalResearchCentre.

AppendixA. Supplementarydata

Supplementary data associated with this article can be found,intheonlineversion,atdoi:10.1016/j.euf.2016.09. 005.

References

[1] deBonoJS,LogothetisCJ,MolinaA,etal.Abirateroneandincreased survivalinmetastaticprostatecancer.NEnglJMed2011;364:1995– 2005.http://dx.doi.org/10.1056/NEJMoa1014618.

[2] RyanCJ,SmithMR,deBono JS,etal.Abirateroneinmetastatic prostate cancerwithout previous chemotherapy. N EnglJ Med 2013;368:138–48.http://dx.doi.org/10.1056/NEJMoa1209096. [3] ScherHI,FizaziK,SaadF,etal.Increasedsurvivalwith

enzaluta-mide in prostate cancer after chemotherapy. N Engl J Med 2012;367:1187–97.http://dx.doi.org/10.1056/NEJMoa1207506. [4] BeerTM,ArmstrongAJ,RathkopfDE,etal.Enzalutamidein

meta-static prostate cancer before chemotherapy. N Engl J Med 2014;371:424–33.http://dx.doi.org/10.1056/NEJMoa1405095. [5] KantoffPW,HiganoCS,ShoreND,etal.Sipuleucel-T

immunother-apy for castration-resistant prostate cancer. N Engl J Med 2010;363:411–22.http://dx.doi.org/10.1056/NEJMoa1001294. [6] ParkerC,NilssonS,HeinrichD,etal.Alphaemitterradium-223and

survivalinmetastaticprostatecancer.NEnglJMed2013;369:213– 23.http://dx.doi.org/10.1056/NEJMoa1213755.

[7] deBonoJS,OudardS,OzgurogluM,etal.Prednisoneplus cabazi-taxelormitoxantroneformetastaticcastration-resistantprostate cancerprogressingafterdocetaxeltreatment:arandomised open-label trial.Lancet2010;376:1147–54. http://dx.doi.org/10.1016/ S0140-6736(10)61389-X.

[8] SweeneyCJ,ChenY-H,CarducciM,etal.Chemohormonaltherapy inmetastatic hormone-sensitiveprostate cancer. N EnglJ Med 2015;373:737–46.http://dx.doi.org/10.1056/NEJMoa1503747.

[9] JamesND,SydesMR,ClarkeNW,etal.Additionofdocetaxel, zole-dronicacid,orboth toﬁrst-line long-termhormonetherapy in prostatecancer (STAMPEDE): survival results from an adaptive, multiarm,multistage,platformrandomisedcontrolledtrial.Lancet 2015;6736:1–15. http://dx.doi.org/10.1016/S0140-6736(15) 01037-5.

[10] LorenteD,MateoJ,Perez-LopezR,deBonoJS,AttardG.Sequencing of agents in castration-resistant prostate cancer. Lancet Oncol 2015;16:e279–92. http://dx.doi.org/10.1016/S1470-2045(15) 70033-1.

[11]ScherHI,HalabiS,TannockI,etal.Designandendpointsofclinical trialsforpatientswith progressiveprostatecancerandcastrate levelsoftestosterone:recommendationsofthe ProstateCancer Clinical Trials Working Group. J Clin Oncol 2008;26:1148–59.

http://dx.doi.org/10.1200/JCO.2007.12.4487.

[12]FlemingMT,MorrisMJ,HellerG,ScherHI.Post-therapychangesin PSAasanoutcomemeasureinprostatecancerclinicaltrials.Nat Clin Pract Oncol 2006;3:658–67. http://dx.doi.org/10.1038/ ncponc0664.

[13]ScherHI,WarrenM,HellerG.Theassociationbetweenmeasuresof progression andsurvivalincastrate-metastaticprostate cancer. Clin CancerRes 2007;13:1488–92. http://dx.doi.org/10.1158/ 1078-0432.CCR-06-1885.

[14]Scher HI,MorrisMJ,KellyWK.Prostatecancerclinicaltrialend points:‘‘RECIST’’ingastepbackwards.ClinCancerRes2005;11: 5223–32.http://dx.doi.org/10.1158/1078-0432.CCR-05-0109. [15]GillessenS,OmlinA,AttardG,etal.Managementofpatientswith

advancedprostatecancer:recommendationsoftheStGallen Ad-vancedProstateCancerConsensusConference(APCCC)2015.Ann Oncol 2015;26:1589–604. http://dx.doi.org/10.1093/annonc/ mdv257.

[16]ScherHI,JiaX,deBonoJS,etal.Circulatingtumourcellsas prog-nosticmarkersinprogressive,castration-resistantprostatecancer:a reanalysis of IMMC38 trial data. Lancet Oncol 2009;10:233–9.

http://dx.doi.org/10.1016/S1470-2045(08)70340-1.

[17]ScherHI,HellerG,MolinaA,etal.Circulatingtumorcellbiomarker panelasanindividual-levelsurrogateforsurvivalinmetastatic castration-resistantprostatecancer.JClinOncol2015;33:1348–55.

http://dx.doi.org/10.1200/JCO.2014.55.3487

[18]YapT,a,LorenteD,OmlinA,OlmosD,deBonoJS.Circulatingtumor cells:amultifunctionalbiomarker.ClinCancerRes2014;20:2553– 68.http://dx.doi.org/10.1158/1078-0432.CCR-13-2664.

[19]Heidenreich A,BastianPJ, BellmuntJ, etal.EAUguidelineson prostate cancer. Part II: treatment of advanced, relapsing,and castration-resistant prostate cancer. Eur Urol 2014;65:467–79.

http://dx.doi.org/10.1016/j.eururo.2013.11.002.

[20] ArmstrongAJ,Garrett-MayerE,OuYangY-C,etal.Prostate-speciﬁc antigenandpainsurrogacyanalysisinmetastatic hormone-refrac-toryprostatecancer.JClinOncol2007;25:3965–70.http://dx.doi. org/10.1200/JCO.2007.11.4769.

[21]HeidenreichA,BracardaS,MasonM,etal.Safetyofcabazitaxel insenioradultswithmetastaticcastration-resistantprostate can-cer:resultsoftheEuropeancompassionate-useprogramme.EurJ Cancer 2014;50:1090–9.http://dx.doi.org/10.1016/j.ejca.2014.01. 006.

[22]Wissing MD, VanOort IM, GerritsenWR, etal. Cabazitaxelin patients with metastatic castration-resistant prostate cancer: resultsofacompassionateuseprogramintheNetherlands.Clin GenitourinCancer2013;11:.http://dx.doi.org/10.1016/j.clgc.2013. 04.004,238–50.e1.

[23]TannockIF,deWitR,BerryWR,etal.Docetaxelplusprednisone or mitoxantrone plusprednisone for advancedprostate cancer. N Engl J Med 2004;351:1502–12. http://dx.doi.org/10.1056/ NEJMoa040720.

(10)

[24]Berthold DR, Pond GR, Roessner M, de Wit R, Eisenberger M, Tannock AIF. Treatment of hormone-refractoryprostate cancer withdocetaxelormitoxantrone:relationshipsbetween prostate-speciﬁcantigen,pain,andqualityofliferesponseandsurvivalin theTAX-327study.ClinCancerRes2008;14:2763–7.

[25]RyanCJ,ShahS,EfstathiouE,etal.PhaseIIstudyofabiraterone acetate in chemotherapy-naive metastatic castration-resistant prostate cancerdisplayingboneﬂare discordant with serologic response.ClincancerRes2011;17:4854–61.http://dx.doi.org/10. 1158/1078-0432.CCR-11-0815.

[26]MorrisMJ,MolinaA,SmallEJ,etal.Radiographicprogression-free survivalasaresponsebiomarkerinmetastaticcastration-resistant prostatecancer:COU-AA-302results.JClinOncol2015;33:1356– 63.http://dx.doi.org/10.1200/JCO.2014.55.3875.

[27]GoldkornA,ElyB,QuinnDI,etal.Circulatingtumorcellcountsare prognosticofoverallsurvivalinSWOGS0421:aphaseIIItrialof docetaxelwith orwithoutatrasentan formetastatic castration-resistantprostatecancer.JClinOncol2014;32:1136–42.http://dx. doi.org/10.1200/JCO.2013.51.7417.

[28] FleisherM,DanilaDC,FizaziK,etal.Circulatingtumorcell(CTC) enumerationinmenwithmetastaticcastration-resistantprostate cancer(mCRPC) treated with enzalutamide post-chemotherapy (phase3 AFFIRM study).J ClinOncol 2015;33(15Suppl):5035.

http://meeting.ascopubs.org/cgi/content/abstract/33/15_suppl/ 5035

[29] LorenteD,OlmosD,MateoJ,etal.EarlyCTCdeclineasabiomarker ofresponsetotreatment incastration-resistantprostatecancer (CRPC):analysisoftheCOU-AA-301andIMMC38trials.JClinOncol 2015;33(15 Suppl):5014. http://meeting.ascopubs.org/cgi/ content/abstract/33/15_suppl/5014

[30] deBonoJS,ScherHI,MontgomeryRB,etal.Circulatingtumorcells predictsurvivalbeneﬁtfromtreatmentinmetastatic castration-resistantprostatecancer.ClinCancerRes2008;14:6302–9.http:// dx.doi.org/10.1158/1078-0432.CCR-08-0872.

[31] OlmosD,ArkenauH,AngJE,etal.Circulatingtumourcell(CTC) countsasintermediateendpointsincastration-resistantprostate cancer(CRPC):asingle-centreexperience.AnnOncol2009;20:27– 33.http://dx.doi.org/10.1093/annonc/mdn544.