Reducing Barriers to Clinical Trials Data Sharing: Cooperative Science to Improve Public Health Valuing patient contributions, respecting existing research communities, and increasing discovery Rebecca Li, PhD, Executive Director
ReducingBarrierstoClinicalTrialsDataSharing:CooperativeSciencetoImprovePublicHealth
Valuingpatientcontributions,respectingexistingresearchcommunities,andincreasingdiscovery
RebeccaLi,PhD,ExecutiveDirector
MRCTCenterMission
1/31/17 2
Engagediversestakeholderstodefineemergingissuesinglobalclinicaltrialsandtocreateandimplementethical,actionable,andpracticalsolutions.
ClinicalTrialsDataSharing:Audiences
Thesharingofdatafromclinicaltrialswithotherresearchersorthegeneralpublic,includingindividualparticipantdata(rawdata),metadata,andsummary-leveldata
ClinicalTrialData
StudyParticipants
ResearchersPublic
ClinicalTrialsDataSharing:Audiences
Thesharingofdatafromclinicaltrialswithotherresearchersorthegeneralpublic,includingindividualparticipantdata(rawdata),metadata,andsummary-leveldata
ClinicalTrialData
StudyParticipants
ResearchersPublic
EuropeanRegulation
EUParliament:Regulation(EU)No536/2014(2014):
Sponsorofaclinicaltrialmustsubmit“asummaryoftheresultsoftheclinicaltrialtogetherwithasummarythatisunderstandabletoalayperson,andtheclinicalstudyreport,whereapplicable,withinthedefinedtimelines.”
Article37:Irrespectiveoftheoutcomeofaclinicaltrial,withinoneyearfromtheendofaclinicaltrialinallMemberStatesconcerned,thesponsorshallsubmittotheEUdatabaseasummaryoftheresultsoftheclinicaltrial.
EURequirespostinglaypersonssummarytoEUPortal– by2017/2018
ReturningResults- MRCTMission
Ø Goals:ReturningClinicalTrialResultstostudyparticipantsØ Developstandardsandbestpractices.
Ø Createaguidancedocument,includingtemplates.
Ø Addressperceivedbarrierstowidespreadimplementation.
Ø Returningresultsallowssponsorsandinvestigatorstorecognizeandhonortheessentialcontributionsandvolunteerismofclinicaltrialparticipants.
Ø Expectationsofacademic,industry,not-for-profitsponsorssimilar
Ø ReturningresultsisakeyaspectofImprovingTransparencyofclinicaltrialsandIncreasingPublicTrust.
ProjectScope:Communicationanddisseminationofsummaryresearchresultstoindividualparticipants
TheMRCTCenterTools&Deliverables
AnRORGuidanceDocumentforgroupswishingtoreturnresultsincluding:• Logisticsanddetailedprocessesforresultssharing• Contentofresearchresultsummaries• Culturalandhealthliteracyconsiderations• Timing
AnRORToolkitincluding:• TemplatesforPhase1,2&3,studiesendingearly• Neutrallanguageguide• Endpointslanguageguide
Goto:mrctcenter.org-- Resources– Returnofaggregateresults
http://mrctcenter.org/news/updated-versions-of-return-of-results-guidance-document-and-toolkit-released/
AlignedwithHRAguidelineswhichreferstothesedocuments
ParticipantClinicalTrialResultsSummaries-Process• Writeinunbiasedandnotpromotionallanguage• Obtainreviewbyindependentandobjectiveeditor(s)andpatient
representative(s)whenpossible• Translateintoadditionallanguagesconsistentwithtranslationsof
informedconsent• Makeavailableanindividualfromthestudysiteorneutralinformed
thirdpartytoanswerquestionsforparticipants• Makeprovisionsforvulnerablepopulationsandotherinstances• Considerastowhethertoinform,andwhomtoinform,intheevent
ofaparticipant’sdeath• Useplainlanguage(sixth-eightgradereadinglevel)• Applyhealthandnumeracyprinciples
Returnofresultstemplates
• Located in MRCT Return of Results Toolkit
• Templates for Phase 1, Phases 2 and 3, and Trials ending early
• Includes examples
• Incorporates principles of Health Literacy and Numeracy
ParticipantClinicalTrialResultsSummaries- Example
Content ExampleWhy thestudywasdone(cont.)
Forclinicaltrialsthatstopearly:Thisstudywasstoppedearlierthanplanned.Thiscanhappenformanyreasons.
Thisstudystoppedearlybecause[addoneofthepossiblestatementsbelow,oryourownsimpleexplanation,tothissentence.Ifthereismorethanonereason,listallthatapply.]…toomanyparticipantshadsideeffects(seebelow).…[druggenericname]didnotimprovepatientresults.…[druggenericname]wasnotaseffectiveasexpected[comparator].…[druggenericname]wasmuchmoreeffectivethanexpected.[ifapplicable,add]Thestudywasstoppedsoallparticipantshadachancetotake[druggenericname].…notenoughpeoplejoinedthestudy.[Includeastatementaboutwhatwillhappennext.…• Forsideeffects..• Forefficacy…• Forfutility…• Lowaccrual:….]1/31/17 10
NeutralLanguageGuideLanguagetoavoid Languagetoconsider
Thisstudyproved… Thisstudyfoundthat...Thisdoesnotmeaneveryoneinthatgrouphadtheseresults.
Thisstudyprovedthatusing<drugA>toprevent<disease/condition>iseffective.
Thisstudyfoundthatpeoplewith<disease/condition>whogot<drugA>had<primaryendpoint>.
Thismeansthat<DrugA>isbetterthan<DrugB>.
Inthisstudy,peoplewhogot<drugA>hadmore<studyendpoint>thansomepeoplewhogot<DrugB>withthesamehealthconditions.
<DrugA>isbettertoleratedthan<DrugB>.
Inthisstudy,fewerpatientswhotook<DrugA>had<listspecificadverseevents>thanpatientswhotook<DrugB>.
SimilarprincipleshavebeensuggestedbyTransCelerate BioPharma:RecommendationsforDraftingNon-PromotionalLaySummariesofClinicalTrialResults
EndpointDescriptionsandExamples
Endpoint Descriptionofthetypeofendpoint Exampleinsimple,plainlanguage
Composite Acompositeendpoint,astheprimaryendpoint,combinesmultipleoutcomes(e.g.death,gettingsickagain(relapse),seriousevent)andtestresultsintoonemeasureofhowwellthedrug/therapy/deviceworks.Thisisusefulwhentherearemanydifferentoutcomesthatcanhappenduringatrial.Thiscanalsobecalledacombinedormulti-partendpoint.
“TheXXXstudymeasured[patients/people]toseeifthoseinGroupA(ABCtreatment)orGroupB(XYZtreatment)livedlonger,hadfewerheartattacks,orfewerhospitalvisitsforheartfailure.Theseeventsweremeasuredtogether(combined)becauseeachoneisquiterare.Researchersalsowantedtoseeifthedrugworkedinpatientswhohadall3conditions.Thestudyfoundthattherewasnochangeinthenumberofeventsfor[patients/people]inGroupAorGroupB.”
1/31/17 12
Numeracy:Overview
• Theabilitytousebasicprobabilityandmathematicalconceptstoexplainmathematicalandstatisticalterms.
• Numeracyprinciplesinhealthliteracyfocusonsimpleexplanations,insteadofusingcomplexfractions,percentagesorstatisticalterms.
1/31/17 13
NumeracyPrinciples:Implementation
• Lessismore– howcriticalarethenumbers?• Providefewerchoices– choosestrategicallywhichoptionstoshow• Dothemath– calculateorconvertnumbers,readersareunlikely
toconductevenbasicmath• Givenumbersmeaningandcontext– explainwhatnumbersmean• Usecommontermsandimaginableformats• Usevisuals• Usewholenumbers• Useconsistentdenominatorsandtimeframe• Naturalfrequenciesvspercentages– “1outof10”maybemore
usefulthanpercentagesbecauseitgivescontextandimagery
From:“HealthLiteracyMissouriBestPracticesforNumeracy,”citedinMRCTReturnofResultsToolkit.(2016),Appendix4 14
14%Or
About 1 in 7
Example
Example
1/31/17 16
In 20% (or 1 in 5) of patients,tumors got at least 30% smaller
In 80% (or 4 in 5) of patients,tumors did not get at least 30% smaller
Returnofresults:MRCTCenterworkgroup
Academic/MedicalCenter:CarmenAldinger– MRCTCenterMarkBarnes- Ropes&Gray,LLP/MRCTCenterBarbaraBierer- Brigham&Women'sHospital/MRCTAssuntaDeRienzo - Brigham&Women'sHospitalAlla Digilova – MRCTCenterRebeccaHLi– MRCTCenterHollyFernandezLynch- HarvardLawSchoolPearlO'Rourke- PartnersHealthCareNesri Padayatchi - Univ.ofKwaZulu-NatalAmishShah- MRCT/HarvardLawSchoolZacharyShapiro– MRCT/HarvardLawSchoolPatrickTaylor- Children'sHospital,BostonSarahWhite- PartnersHealthCareElizabethWitte– HarvardMedicalSchoolSabune Winkler– HarvardMedicalSchoolIndustry/TradeAssociations:SalvatoreAlesci – PhRMARichardBergstroem – EFPIAElizabethGarofalo - NovartisPharmaAGLauraHagan- MerckSeranoSandraHayes-Licitra – Johnson&JohnsonAngelikaJoos – MerckSharp&DohmeBarbaraKress– MerckSarahLarson– BiogenIdecDavidLeventhal– PfizerCraigLipset – PfizerLaurieMyers– Merck(CO-CHAIR)
AlexNasr– AbbVieMaryAnnPlummer– J&J(priorCO-CHAIR)SandyPrucka – LillyBenRotz – LillyBethRoxland– Johnson&JohnsonJessicaScott– GSKInstitutionalReview Boards:DavidForster- WIRBCopernicusGroupMaryOster– NEIRBJimSaunders- NEIRBNonprofit:Behtash Bahador – CISCRPPhyllisFrosst - PersonalizedMedicineCoalitionZachHallinan– CISCRPMarcWilenzick– InternationalAIDSVaccineInitiativePatientAdvocates:NicolaBedlington– EuropeanPatientsForumDeborahCollyar– PAIR(COCHAIR)DavidHaerry – EuropeanAIDSTreatmentGroupCherylJernigan- SusanG.KomenYannLeCam – EURODISMarcelloLosso - HIVRAMOSJanePerlmutter – GeminiGroupResearch/ConsultingFirms:BarbaraGodlew - TheFAIRECompany,LLCPierreGervais- QTResearchPauloLacativa - CCBRClinicalResearchDavidWalling– CollaborativeNeuroScience
QuestionsonPlainLanguageSummaries?
1/31/17 18
ClinicalTrialsDataSharing:Audiences
Thesharingofdatafromclinicaltrialswithotherresearchersorthegeneralpublic,includingindividualparticipantdata(rawdata),metadata,andsummary-leveldata
ClinicalTrialData
StudyParticipants
ResearchersPublic
OurWorkinDataSharing
1/31/17 20
••Harmonizinglanguage&agreementsamongsponsorsPolicy
••Loweringbarrierstodatasharing••AcademiccreditfordatasharingAdvocacy
••Creatingaplatformfordatasharing••Makingdataonotherplatformsmorediscoverable
Implementation
Policy
1/31/17 21
••Harmonizinglanguage&agreementsamongsponsors
Policy
• DataContributoragreement
• DataUseagreement
• Informedconsentagreementlanguage• IRPPrinciplesdocumentdefinedforparticipatingIRPs
Policy
1/31/17 22
••LoweringbarrierstodatasharingAdvocacy
• Conveningstakeholderswhoarecurrentlysharingandthosewhodonotcurrentlyshare
• Lowerthebarriersforthosewhodonothavetheresourcesorknowledgetosharedata
• Advancepolicychangesthatenableacademiccreditfordatathatismadewidelyavailablefortheadvancementofscienceandmedicine
Policy
1/31/17 23
Createasustainableglobaldata-sharingplatform
Implementation
• LaunchVivli,anon-profitcenterforglobaldatasharing
• Makingdataonotherplatformsmorediscoverablethrougharobustsearchengine
MajorClinicalTrialDataSharing:Platforms
• CSDRPLATFORM- leadingindustrymulti-sponsorrequestsite
• Clinicaltrials.gov – searchabledatabaseincludingsummaryresults
• J&J- YODA project- YalepartnerswithJ&J/Medtronic
• Duke ClinicalResearchInstitute– BristolMyersSquibbStrategicInitiative(SOAR),whichsupportsopenaccesstoclinicaltrialsdata
• ProjectDatasphere – cancercomparatordata,andmore
• NIHdatarepositoriesand(BIOLINCCand60+others)
• FDA Oncology’sdataaggregationeffort- InformationExchangeandDataTransformation(INFORMED).
• OPENTRIALS – indexesallfreelyavailableinformation,noIPD
• EMADatabase– CSRssubmittedtotheagencyaspartofaMAA
CurrentPlatformDifferentiatingFactors
FACTOR VARIABLES
Data types Summary leveldata IPDdata*
Reviewprocess No review Discretionary Review
DataStandards Specificformatrequired Multipleformatsaccepted
Contributor Members only Anyresearcher
Requestor Researcher/Expert Public/no qualifications
Hosts data No Yes
Downloadabledata No Yes
Review ofrequest None Discretionary
Analysis space Noneprovided Spaceandlicensed toolsprovided
Specificityofplatform Specificto sponsortype,geographyortherapeuticarea
Multi-stakeholder, global,alltherapeuticareas
Connect/search otherplatforms
No Yes
1/31/17 25*IPD– IndividualParticipantleveldata
CurrentPlatformDifferentiatingFactors
FACTOR VARIABLES Vivli
Data types Summary leveldata IPDdata*
Reviewprocess No review Discretionary Review
DataStandards Specificformatrequired Multipleformatsaccepted
Contributor Members only Anyresearcher
Requestor Researcher/Expert Public/no qualifications
Hostsdata No Yes
Downloadabledata No Yes
Review ofrequest None Discretionary
Analysis space Noneprovided Spaceandlicensed toolsprovided
Specificityofplatform Specificto sponsortype,geographyortherapeuticarea
Multi-stakeholder, global,alltherapeuticareas
Connect/search otherplatforms
No Yes
1/31/17 26*IPD– IndividualParticipantleveldata
CurrentGap
Weandothershaveidentifiedsignificantcurrentchallengestoutilizingexistingdataonclinicaltrialsforfurtherresearch:• Manyacademiciansandothersdonothaveameanstomakedata
availableinaturn-keyfashion.• Althoughtechnologyhasmadeiteasiertomakedataavailable,
dataarestilldifficulttodiscover.• Arobustcentralizedsearchenginedoesnotexisttolocatedata
acrossthedifferentdatageneratorsanddataplatforms.• Combingdatasetsfromdifferentgeneratorsisresource- andtime-
intensiveduetoinconsistentadoptionofdatastandards,datarequirements,securitystandardsandpolicies.
1/31/17 27
WhyVivliisNeeded
• Datahostingcapacity– VIVLIisageneralaccessdataplatformthatflexiblydesignedtomeetglobal
capacityneeds
• Analyticfunctionality/value– currentfederatedarchitectures(e.g.,SENTINEL,PCORnet)offeronly
limitedanalyticcapabilities(e.g.,counts)– Toenablemeta-analysisandotheraggregatedanalyses,datasetsneedto
beheldinasinglehostenvironment– ThegreatertheproportionofIPDdatasetsheldbyonehost,thegreater
theabilitytodoaggregatedanalyses– NootherplatformaimsforVivli’sscaleandscopeofIPDhosting
1/31/17 28
CenterandBackground
TheMulti-RegionalClinicalTrialsCenterofBrighamandWomen’sHospitalandHarvardUniversity(MRCTCenter)haschampionedeffortstoenhancesharingof,
andaccessto,clinicaltrialsdataaroundtheworld.
Vivliisbeingformedtoprovidegovernanceandmanagementofaplatform—agnostictodisease,country,sponsorandfunder—toreducebarrierstoclinical
trialsdatasharing.
1/31/17 29
Variousstakeholderpositions
• Industry– Datasharingon“multi-sponsor”orsingleinstitutionplatforms– Attheleadingedgeofdatasharing– Allrequirereviewfordata
• Academic– Noinstitutionalsolution;creatingURLsorusingother(e.g.NIH)resources– Individuals:(1)Ifshare,prefertodelegatemanagementwithopenaccess
or(2)closelyheldunderstrictcontrolsforaccess
• Non-profitandpatientadvocacygroups– Generallyencouragingofdatasharingwithappropriateparticipant
informationandconsent– Appreciatethat“standaloneplatforms”losethevalueofinteroperability
andofcollectiveadvantageof“bigdata”
1/31/17 30
1) Enablinginteroperabilityofdatafrommultiplesources;
2) Hostingdataforstakeholdersthatdonothavetheabilitytodoso;
3) Coordinatingandpartneringwithexistingdata-sharinginitiatives,policies,andprocessesasappropriate;
4) Promotingreasonedsolutionstochallengesofdatasharing.
31
Scope
Mission
1/31/17 32
Promote,coordinate,andfacilitateclinicalresearchdatasharingthroughthecreationandimplementationofa
sustainableglobaldata-sharingenterprise
TheUniqueRemitofVivli– EnhancingDiscovery
Advancedmetadatasearchanddiscovery capability
Simplified access request systemtodataresidingonotherplatforms
1/31/17 33
Searching multipledatabasesina
fragmentedlandscape
Discovery ofdatacanbechallenging
Morecommunitiesandpartners=morediscoverabledata
Astheenhancedmetadatacatalogmatures,moredata,includingexternallyhosteddata,willbediscoverablethroughVivli
TheUniqueRemitofVivli– ProvidingVisibilityandCombiningDatainaCentralizedPlatform
CombinedIndustry,Academic,andBiotechdatadiscoverablethroughonesearchengine
Anonymized IndividualParticipant-LevelData(IPD)availableforrequest
1/31/17 34
VivliSolutionsOfferings
1/31/1735
VivliSearchandRequestTool
VivliYodaCSDR
GSK Lilly Takeda RocheEtc. JnJ Medtronic Vivli
More
Other
IRP IRP IRP IRP
OpenSearch UserRequest ApprovedRequest
Secureanalytic
Environment
Dr.XData
Vivli
Yoda
GSK
Centralizedsearchandrequestportalfordatahostedonmultipleplatforms
EnhancedMetadataformoreprecisesearchresults
SecurespacetocombineIPDdatafrom
multiplesources,includinguploadofacademicdata
Hostingforclinicaltrialdata,includingmintingDOI forpublicationpurposes
Respectingothercontributorreview
processesanddatausetermswhileprovidinguserwithcentralizedmechanismforrequest
Examplesofexistingplatforms
ExampleofNewSearchFunctionality
1/31/17 36
DevelopmentAdvisorsandSupporters
1/31/17 37
TheMRCTCenterwouldliketothankthefollowingsupportersandfunders*fortheircontributionstotheVivliproject:
• DeloitteConsulting• GlaxoSmithKline• TheInstituteofMedicine• TheLauraandJohnArnoldFoundation• Johnson&Johnson• TheLyda HillFoundation• PfizerFoundation• WellcomeTrust
*Inalphabeticalorder
ChallengeswithLaunchingaNewPlatform
1/31/17 38
• CreateincentivestructureswithinVivli forsharingwithexternalmandatestorequireclinicaltrialdatasharing
• Harmonizingrequirementsofdiversestakeholders• AcceleratingtheVivlitimelinetolaunch• Maintainlowfinancialbarrierstoentry
Joinus
• Partnerwithus- ourmissiontoistoaccelerateyourdiscoveriesthroughintegratingofdataglobally acrossplatformsanddatabases
• Pleasecollaboratewithusbymakingyourdatadiscoverable
• ConsiderlearningmoreaboutsupportingtheVivli effort
1/31/17 39
THANKYOU
1/31/17 40