To advance excellencein the management

of clinical data

A publicAtion supported by And for the MeMbers of the society for clinicAl dAtA MAnAgeMent, inc.

• TheAnnualConferencewaswidelyap-preciated,featuringmorethan50speakersfromacrosstheglobe.Specialthankstoallofoursponsorsandthe575attendeeswhosupportedthisevent.Conferenceco-chairsPaulClarksonandCharleneDark,aswellasthesupportteamatEDI,attendedtoev-erydetailtoensureasuccessfulconference.

Thisyear,wewelcomethreenewmemberstotheBoardofTrustees:JonathanAndrus,JenniferDugganandMeredithNahm.Theybringadiversemixofexpertisefromsoftwaresolutions,thedeviceindustryandacademics.DepartingtheBoardareLindaTalley,VesnaZovkicandDebraJendrasek.Thankyouallforyourservice.Weshallmissyou!

ThisisexpectedtobeayearofsignificantchangeforSCDM,withnewglobalizationinitiatives,partnershipsandmeasurestodi-versifyandexaminethebiggerpicturesoastoencompasstheentiredatavaluechaininthedrugdiscoveryanddevelopmentprocess.ViceChairSusanHoward,SecretaryCarolGarvey,andTreasurerGreggDearhammer,allstal-wartsinthisindustry,bringimmenseleader-shiptotheExecutiveCommittee,asdoesourexecutivedirector,EloizaAltoro-Acevedo.

IamdeeplyhonoredandhumbledtoserveasthefirstinternationalchairofSCDM,andwouldliketotakethisopportunitytosincere-lythanktheentiremembershipandtheBoardfortheconfidencetheyhaveplacedinme.Fiveyearsin,myinvolvementwithSCDMhasbeenarewardingexperience.Ihavetrulyenjoyedeveryminuteofit.SCDMisaprofes-sionalandcommittedorganizationcomprisedofthoughtful,warmandcapablepeople.

This Issue1Letter from the Chair

2Letter from the Editors

3Pharmacogenomics Overview and Impact on Data Management in Clinical Research

6Pharmacogenomics and Metabolomics

9Trends in Clinical Data Management Due to Growing Dissatisfaction with EDCs

12Ethical Issues in Pharmacogenetics

Volume 16, number 4 | 2010 Winter

Letter from the ChairNimita Limaye, PhD, CCDM

DearMembers,

WelcometoabrightandsuccessfulNewYear!We

havejustclosed2010onasuccessfulnoteundertheableleadershipof2010BoardChairRalphRusso.IgreatlyenjoyedworkingwithRalphthispastyear.Hisforesight,ma-tureleadershipanddeepcommitmentaddedconsiderablevaluetoSCDM.ThankyouRalph,onbehalfoftheentireorganization!

Closeto700newmembersjoinedtheSCDMfamilyin2010.Aspartofthenewly-launchedstudentmembershipinitiative,wealsointro-ducednearly150newstudentmemberstoourorganization.Inspiteofthecurrentglob-alrecession,SCDMhasmorethan2,600memberstoday!Itisalsonotablethatnearlyone-fifthofourmembershipisinternational.

Other2010highlights:• Weendedtheyearwith580CCDMs.One-

fourthoftheseareinternational,withthemajorityfromIndia,CanadaandSouthAf-rica.ThisreflectsthegrowingrecognitionofthevalueofCCDM®certificationglobally.

• WeestablishedcollaborativeeducationalpartnershipswiththeU.S.FoodandDrugAdministration,DukeClinicalResearchIn-stitute,AHIMA,ACRP,CDISCandothers.

• WecollaboratedwiththeCenterforPhar-maceuticalPublishinginTokyotoreleaseaJapanesetranslationoftheGCDMP,ourindustrybestpractices.

• ThefocusonthechangingroleofthedatamanagerandglobaloutsourcingtrendsweredemonstratedbynewchaptersintheGCD-MPonProjectManagementfortheCDM,andonVendorSelectionandManagement.Inlinewiththis,awebinarandpre-confer-encetutorialon“OutsourcingStrategyandMethodology”wereconductedaswell. Continued on page 3

2 Data Basics • winter 2010 Return to Cover

2011 SCDM CommitteesEducation & Professional Development Michelle Meany, CCDMPat Stetser, CCDM

Marketing & Communications Nelson Lee, CCDMJim Dorsey

Products & Services Jennifer Duggan, CCDMSusan Krikorian, CCDM

Strategic Directions Jonathan Andrus, CCDM

Letter from the Editors

Directory of Advertisers11 Medidata

16 Cincinnati Childrens Hospital

DearSCDMMembers,

WelcometotheWinter2010IssueofDataBasics!

Justastechnologyhascontinuallyadvancedtogiveussmartphones,high-techgamingcon-solesandhybridcars,therehaslongbeenaneedforintroducingmedicationswithhighspecificity.Thisisevenmoreimportantdueto theideathatdrugswithhigherspecificitycanpotentiallyleadtolowerhealthcostsandfeweradverseeventsifitcanbereliablydeterminedwhowilloptimallybenefitfromacertainmedication.

Theconceptofsmartmedicinesispropelledbythedisciplineofpharmacogenomics,whichcanhelpbiotechandpharmaceuticalcompa-niesdevelopdrugsforspecificsubpopulations.Theuseofmetabolomicsfurtherservesthispurposebyhelpingresearchersfindnewmark-ersfordisease.Whilebothpharmacogenomicsandmetabolomicsmaynotseemlikecommontopicsamongclinicaldatamanagersandinthemainstreamofclinicaltrials,theyarekeydisci-

2011 Online Course OfferingsOnline Course Dates

CRFDesign Jan.10–Feb.4

DevelopingDataManagementPlans Feb.7–Mar.11

QueryProcessingandTracking/DatabaseUpdates Mar.14–Apr.8

MetricsandIdentifyingDataTrends Apr.11–May6

ProcessingLabData May9–Jun.3

SAEReconciliation,SafetyReviewandCoding Jun.6–Jul.1

DatabaseLockandRandomization Sep.19–Oct.14

ProjectManagementfortheDataManager Oct.17–Nov.11

2011 Webinar ScheduleWebinar 11:00amCentral;60-minutepresentation(30-minuteQ&A) Dates

What’sNewwithePROSystems:FDAUpdatesandRegulations Feb.17

AdaptiveTrialDesign–WhatItMeansfortheCDM Mar.17

UsingtheCDISCStandardsEnd-to-EndinClinicalTrials Apr.14

DeviceTrialStrategiesandtheCDM’sRoleinQualityAssurance May19

21CFRPart11fortheClinicalDataManager Jun.23

Biostatistics–WhatEveryClinicalDataManagerShouldKnow Sep.22&29

DataIntegrationandtheGCDMP:ANewChapterinClinicalDataManagement Oct.20

TheRoleofMetricsinClinicalDataManagement:TrackingQualityandEfficiency Nov.17

2011 SCDM E-Learning Register at http://portal.scdm.org

plineswhichprovidevalueandtimesavingsfortheoveralldrugdevelopmentprocess.

AlthoughthisissueisleanerwhencomparedtopastissuesofData Basics,wethinkyou’llfindthearticlesinterestingandtheissuewillpro-videyouwiththebackgroundforthesedisciplines.

Aswemoveinto2011,wewouldliketoen-courageourreaderstocontributenewarticles.PleasechecktheSCDMwebsiteforouredito-rialcalendar.

PleasebesuretoalsocheckouttheSCDMwebinarsandonlinecoursesscheduledthroughoutthisyear.Afullschedulehasbeenprintedbelow.Thereareawidevarietyoftopicswithsomeofthelatesttrendsinourindustry.

HappyNewYear!

Sincerely,RehanaBluntandLyndaHunterData BasicsCo-Editors

3 Data Basics • winter 2010 Return to Cover

2010 SCDMBoard of TrusteesRalph J. Russo, CCDMChairMerck

Nimita Limaye, CCDMVice ChairSIROClinpharmPvt.Ltd.

Susan HowardSecretaryGlaxoSmithKline Pharmaceuticals

Gregg DearhammerTreasureri3Statprobe

John Estrada, CCDMPGEVentures,Inc.

Carol D. Garvey, CCDMGenentech,Inc.

Debra JendrasekChilternInternational

Linda S. KingEliLillyandCompany

Steven PowellPRAInternational

Vesna E. Zovkic, CCDM

Linda Talley, CCDMPast ChairEliLillyandCompany

Continued on page 4

Pharmacogenomics Overview and Impact on Data Management in Clinical Researchby Margarita Strand, Lead Clinical Data Manager, Oncology at Novartis.

AbstractThisarticleprovidesabriefoverviewofselectedlitera-tureonPharmacogenomics(PGx)asabranchofphar-macology,anddescribesmethodsofstudyingthe

geneticbasisofpatients’responsevariabilitytoinvestigationalcompoundsinclinicalresearchwithafocusonthedatamanagementaspectsofclinicaltrialsinvolvingpharmacogenomicsdata.

Pharmacogenomics DefinitionPharmacogenomics(PGx)isthebranchofpharmacologywhichdealswiththeinfluenceofgeneticvariationondrugresponseinpatientsbycorrelatinggeneexpressionorsingle-nucleotidepolymorphisms(SNPs)withadrug’sefficacyortoxicity.1Singlenucleotidepolymorphisms,orSNPs(pronounced“snips”),areDeoxyribonu-cleicAcid(DNA)sequencevariationsthatoccurwhenasinglenucleotide(A,T,C,orG)inthegenomesequenceisaltered.5

Althoughpharmacogenomicsisbroaderinscopeandreferstothecomplexinteractionsofgenesacrossgenome,theterms‘pharmacoge-nomics’and‘pharmacogenetics’areoftenusedinterchangeablyintheliterature.2

History of Pharmacogenomics and Current StatusRapididentificationoftensofthousandsofhumangenesandhundredsofthousandsofDNAvariationsthatmightinfluencediseasesusceptibilityhasspawnedanewfield—phar-macogenomics.3Pharmacogenomicsisanoffspringofpharmacogenetics.Thehistoryofpharmacogeneticsdatesbacktothe1900s.4

Overthepastfewdecades,duetosuchmajoraccomplishmentsasthecompletionoftheHumanGenomeProject6in2003,whichhasprovidedablueprintoftheDNApresentineachhumancell,rapidprogresshasbeenmadebyusinggeneticstoidentifythemolecularcauseofhumandisease.GenomicsresearchisnowfocusingonthestudyofDNAvariationsthatoccurbetweenindividuals,seekingtounder-standhowthesevariationsinfersusceptibilitytocommondiseasessuchasdiabetesorcancer.7

Application of PGx in Clinical ResearchThetwomajorgoalsforclinicalapplicationofPGxare:• Abilitytopredictpatientswhoareathigh

riskoftoxicityinanefforttopreventseriousadverseevents(SAEs)andensuresafety

• Abilitytopredictpatientswhoaremostlikelytoobtainthedesiredtherapeuticeffectfromthedruginanefforttoincreaseefficacy8

Bothofthesedefinepossibilitiesforincorporat-inggenotypingintoeachphaseofaclinicaltrial.

WithrecentadvancementsinthefieldofPGx,geneticfactorcannowbeconsideredanintrin-sicorpredictablefactoraffectingdrugresponseandcanbeaddedtotheotherintrinsicfactorsincludingage,gender,race/ethnicity,diseasestate,organdysfunction,etc.8

230clinicaltrialsinvolvingpharmacogenomictestinghavebeenregisteredwithClinicalTrials.govwebsite9asofOctober1,2010demon-stratingwidespreaduseofpharmacogenomicsinclinicalresearch.

PGx in Oncology ResearchTumorbiopsyallowsfordirectanalysisofabnormaltissue,whichexplainswhyPGx

Iknowthatthisisjustthebeginning,andIshallseekyourcontinuedsupporttohelpdrivethestrategicvisionthatwehavesettomakethisatrulyglobalorganizationofconsiderablerepute.

Onceagain,wishingyouallasparklingandsuccessfulNewYear,

NimitaLimaye,PhD,CCDM2011ChairSCDMBoardofTrustees

Letter from the Chaircontinued from cover

4 Data Basics • winter 2010 Return to Cover

Pharmacogenomics Overview and Impact on Data Management in Clinical Researchcontinued from page 3

analysisofefficacyincancerresearchiswellaheadofothercomplexdiseasesthatmustsolelyrelyongeneticassociationstudies.11Forexample,patientswithmetastaticcolorectalcancer(mCRC)whosetumorscarrywild-typeversionofKRASgenerespondtoPanitumumab(Vectibix)betterthanpatientswhosetumorscarryamutatedformofKRAS.

Acommercialtest–TheraScreen’sK-RASMutationtest–wasapprovedinEuropetodetectsevenmutationsfoundinmanycancertypesandisnowusedtopredictVectibixefficacy.11Patientswithnon-small-celllungcancer(NSCLC)respondfavor-ablytoGefitinib(Iressa®)—iftheyhaveparticularmutationsinthetyrosinekinasedomainoftheepidermalgrowthfactorreceptorgene(EGFR)oftheirtumor.Thefindingssuggestthereisnoneedtotrythisanti-cancerdruginpatientsnothavingthistumorEGFRgenotype.14

Stratification of Research SubjectsFigure1illustratesthecurrentstateofthedrugdevelopmentprocesswhereonlyalimitednumberofpatientsaretreatedwithaspecificdrugforanygivendiseaseduetoadverseevents.Ofthosepatientswhoarereceivingthedrug,notallrespond.10

Thefutureofdrugdevelopmentisbasedongenotype-inducedsubjectstratifica-tionandispresentedinFigure2whereeachsub-groupisrepresentedashavingadrugavailablethatistailoredtotheirgenotypewiththebenefitofreducedadverseevents.10

Regulatory PerspectiveWithactivesupportfromregulatoryagencies,PGxisalsobecominganimpor-tantaspectofdruglabelingsothatgroupsofpatientswhoaremostlikelytore-spond,andlesslikelytosufferadverseevents,canbeidentifiedinpractice.11

Increasinginterestfromregulatoryau-thoritiesonpotentialapplicationsofPGxhasraisedtheawarenessoftheuseofPGxdatainthedrugdiscoveryanddevelop-

ment.TheFoodandDrugAdministra-tion(FDA)haspublished‘GuidanceforIndustryPharmacogenomicDataSub-missions’.Itwasgeneratedspecifically‘tofacilitatescientificprogressinthefieldofPGxandtheuseofPGxdataindrugdevelopment’.12NotonlydoesitprovideguidanceonPGxdatasubmissionre-quirements,butalsoencouragesvolun-tarysubmissionofdatafordiscussionwiththeagency’snewlyformedInterdis-ciplinaryPharmacogenomicsReviewGroup(IPRG).12

ICHhasalsoissuedGuidanceforIndus-tryE15:‘DefinitionsforGenomicBio-markers,Pharmacogenomics,Pharmaco-genetics,GenomicDataandSampleCodingCategories’.

Figure 1. Current state of drug development process10

Figure 2. Future state of drug development based on genotype-induced subject stratification10

ThewillingnessofregulatoryagenciesandpharmaceuticalcompaniestoadoptacollaborativeapproachtoPGxwillbeakeyfactorinmovingPGxforward,aswellastheagency’sapprovalofvalidatedpharmacogenomictests.12

Data Management Aspects of Clinical Trials with PGx DataSeveralaspectsofaclinicaltrialinvolvingPGxdatashouldbeconsideredfromtheClinicalDataManagementperspectiveduringstudystart-upphase(Table1)andevaluatedthroughoutthecourseofthestudywiththeimplementationofqualityassuranceandqualitycontrolmeasurestoensurehighqualitydata.

Continued on page 5

5 Data Basics • winter 2010 Return to Cover

# Aspect Clinical Data Consideration from Data Manager’s (DM) Perspective1 StudyProtocol DataManagershouldthoroughlyreviewsectionsintheProtocolrelatedtoPGx,ensureunderstandingofthesample

collectionmethodologyforPGxstudiesandanalysis,evaluateimpactontheCaseReportFormandclinicaldatabasedesign(seealsoCRFsectionbelow).

2 DataManagementPlan(DMP)

DataManagershouldincorporategenetictestingcomponentandgeneticdatahandlingintotheDMPaswellasdefinetheflowofthegeneticdataanditsintegrationintotheoveralldatamanagementprocess.

3 CaseReportForm(CRF) StudyspecificCRFpagescollectinggenetictestingdataneedtobedesigned.Theuseofthedataforstatisticalanalysispurposesneedstobeevaluated,discussedwiththetrialstatisticianandalignedwiththeStatisticalAnalysisPlan(SAP).

4 Subject’sEligibilityCriteria Inclusionandexclusioncriterianeedtobecarefullyevaluatedtoaccountforthesubject’sgenotype.Receiptofresultsofmutational/geneticanalysisfromtheanalyticallaboratorypriortoBaselinevisitbecomesessentialforevaluationofeligibilitycriteria.Potentialimpactonthespeedofsubject’srecruitmentneedstobeevaluatedanddiscussedwiththeClinicalteam.

5 ClinicalDatabaseDesign Variablenames,formats,codelistsanddatasetstructureassociatedwithgeneticdataneedtobedefined.AppropriateannotationsneedtobecreatedbytheprogrammerintheAnnotatedCRFanddatabasedesignspecifications.

6 DataValidationChecks Editchecksintendedtodetectpotentialerrorsanddiscrepanciesingeneticdataneedtobedeveloped,programmed,testedandimplemented.

7 RandomizationMechanism

Dependinguponthesystemusedtorandomizethesubjectsintothestudy(e.g.InteractiveVoiceResponseSystem(IVRS),InteractiveWebResponseSystem(IWRS),etc.)randomizationscriptsneedtobewritten,programmed,thoroughlyreviewedandtestedtoensurethat,forexample,subjectwithgenotypeAisassignedtotreatmentarmA,andsubjectwithgenotypeBisassignedtotreatmentarmB(asdefinedintheStudyProtocol).

Availabilityofmutational/geneticanalysisresultsfromtheanalyticallabpriortotheBaselinevisitbecomesessentialforsubjects’stratificationandassignmenttotheappropriatetreatmentarm.

8 ExternalDataHandling(non-CRF)

Datatransferspecifications(DTS),frequency,timingandmethodofgenotypedatatransferfromtheexternalvendor(analyticallab)needtobedefinedandagreedupon.TestdatatransfershouldbecheckedforcompliancewithdatabasestructuredefinedintheDTS.

9 DataReviewandCleaning MethodandextentofPGxdatareview,cleaningandreconciliationaswellasquerymanagementshouldbedefinedintheDMP.Datalistings/reportsnecessarytoreviewthePGxdatafromtheDMperspectiveaswellaslayoutofthePatientProfiles/reportsusedbyClinicalandSafetyteamstoreviewthesafetyandefficacydataincorrelationwiththesubjectgenotypevariationsneedtobedefinedandprogrammed.

10 AdverseEvents/SAEs Correlationbetweensafetysignalsandgenotypeshouldbeevaluated(particularlyforSAEs)duringsafetymonitoringandAEclassification.

11 ClinicalDatabaseLock DataManagerneedstoincorporateitemsrelatedtoPGxdataintotheclinicaldatabaselockchecklistandensurethechecklistisfollowed.

12 Training Forin-housetrials,theinternalstudyteampersonnelaretrainedontheprocessofcollectingandhandlingPGxdata.

ForoutsourcedtrialswhereaContractResearchOrganization(CRO)isinvolved,clearunderstandingoftheroles&responsibilitiesbytheCRO’sDMteamwithrespecttoPGxdatahandlingshouldbeensured.

Table 1 – Data Management Aspects of Clinical Trials with PGx Data

Pharmacogenomics Overview and Impact on Data Management in Clinical Researchcontinued from page 4

Conclusions Pharmacogenomicscanprovidesubstan-tialefficiencyinclinicalresearchbyfacili-tatingtheconductofsmallerclinicaltrialswhiletargetinggroupsofpatientswithsimilargeneticbackground.Theapproachofdeterminingthegenotype/phenotyperelationshipinindividualdrugresponsewillprovidephysiciansandresearcherswiththekeyinformationthatallowsthemtopreciselyprescribeordesigntherightdrug,attherightdose,fortherightpatient.13

Blendingthecomponentsofgeneticsandpharmacology,PGxhascreatedanewparadigminthepharmaceuticalland-scape.15Withawidespreadadoptioninclinicalresearch,PGxhasapotentialtobecomeacorecomponentofconvention-aldrugdevelopmentinthenearfuture.ConsideringtheimportantroleofClini-calDataManagement,DataManagersshouldbepreparedtohandlePGxdatabymeansofacquiringknowledgeaboutthissubject,welcomingopportunitiestolearnandgainexperience,leveraging Continued on page 16

technology,anddevelopingandimple-mentinggoodpharmacogenomicsdatamanagementpractices.

References 1 http://en.wikipedia.org/wiki/Pharmacogenomics2 Clinical Use of Pharmacogenomic Tests in 2009,

Leslie J Sheffield, Hazel E Phillimore, Clinical Biochemistry Review, Vol. 30, May 2009: 55-65

3 The New, New Pharmacogenomics, Malorye Branca, Bio-IT World, September 2002

6 Data Basics • winter 2010 Return to Cover

Pharmacogenomics and Metabolomicsby Nimita Limaye, PhD, CCDM, VP and Global Head, Strategic Data Services and Medical Writing, SIRO Clinpharm Pvt. Ltd.

Pharmacogenomicsandmetabolomicsaretwoofthekeydriversofthemassive“omics”explosion.Systemsbiologyaimstofacilitateacomprehensiveanalysisoftheinteractionsofcomplexbiologicalsystemsusingkeytoolssuchaspharma-cogenomicsandmetabolomicstodevel-oppredictivemodelsofhumandisease.

Integratedanalysisoforganandsystem-levelresponsesusinginnovativecomputationalanalyticaltechniqueswouldhelpidentifybiomarkersanddesignclinicaltrials,factoringinbothgenotype-phenotypeaswellasgenotype-envirotyperelationships.Theapplicationofhealthinformaticstothedataoverflowresultingfromtheapplicationofsuchtech-nologieswoulddrivefaster,morefocusedandtargeteddrugdiscovery,reducehealthcarecostandpavethewayforper-sonalizedmedicine.

Pharmacogenomicsandmetabolomics,alongwiththeother‘omics,’suchasepigenomicsandneurogenomics,aresomeofthetoolswhicharespearheadingthenewageofpredic-tive,preventiveandpersonalized(PPP)medicine.

Pharmacogenomicsisabranchofpharmacologywhichdealswiththeinfluenceofgeneticvariationsondrugre-sponseinpatientsbycorrelatinggeneexpressionorsingle-nucleotidepolymorphismswithadrug’sefficacyortoxicity.1Thus,itaimstooptimizedrugtherapywithrespecttothepatient’sgenotypetoensuremaximumefficacyandmini-mizeadverseeffects.Itfollowsapolygenicorgenome-wideapproachtoidentifyinggeneticdeterminantsofdrugre-sponse,asopposedtopharmacogeneticswhichexaminessinglegeneinteractionswithdrugs.Itisimportanttore-memberthatwhilethenumberofpolymorphismsingenesencodingdrug-metabolizingenzymes,drugtransportersanddrugtargets,aswellasdisease-modifyinggenesthathavebeenidentifiedcontinuestoincrease,mostdrugeffectsandtreatmentoutcomesaredeterminedbytheinterplayofmultiplegenes,andthisiswherepharmacogenomicscomesintothepicture.

Thesetechnologiesarenotonlybeingusedtotreatcriticalillnesseslikecancer,cardiovasculardisordersanddiseasessuchasHIV,tuberculosis,asthmaanddiabetes.AgoodexampleistheresearchonWarfarin(ananticoagulant)whichhasshownthattwogenesimpacttheoptimaldosingofWarfarin.Thesegeneticfactorsaccountfor30percentto35percentofthevariabilityinthedosing,whileclinicalfactorsareresponsibleforonly17percentto21percent.2

Pharmacogenomicsisalsobeingappliedto‘theranostics’–themergerbetweentherapeuticsanddiagnosticsintheformofaDNAtesttopredictapatient’sresponsetothedrug–andhasbeeneffectivelyusedtoselectpatientsub-populationsforclinicaltrials,forexampleintheK-rastestwithCituximabandEGFRtestwithGefitinib.3Similarly,genesignaturemicroarraystorelatepatternsofgeneexpres-sionwithspecificclinicaloutcomes,suchasthe70-genesignature“Mammaprint”,aU.S.FoodandDrugAdminis-tration-approvedtoolcommercializedbyAgendia,arebeingusedforbreastcancerprognosisandhavebeenreportedtobe77percentto81percentaccurate.4Thisnewsetofmo-leculardiagnostictoolscanbeusedtoindividualizeandoptimizedrugtherapy.

Pharmacogenomicsnotonlyservestoidentifynewtargetsforthedevelopmentanduseofdrugsinspecific,identifi-ablesubpopulations,butalsoplaysasignificantroleinreducingtheincidenceofadverseeventsbyexcludingpa-tientswhoarelikelytosuffersuchevents.Thiswillprob-ablyservetolowerthecostofhealthcareandmakephar-macogenomicsaninvaluabletoolforpredictive,aswellaspreventive,medicine.5

Thiscouldbedonebyidentifyingpolymorphismsthatpredisposepatientstoadversedrugeffects,possiblybyobtaininggenomicDNAfrompatientsenteredonlargePhaseIIIclinicaltrialsofanewagent,andthenretrospec-tivelysearchingforpolymorphismsthatpredisposeasmallsubsettotoxicities.6Thus,predictivemedicinewouldin-volvemappingthefuturehealthhistoryofanindividualandwillmandatethecost-effectiveandswiftsequencingofthehumangenome.Thisisnotmerelylefttoone’simagi-nationasIBMhasannouncedthatitisintheprocessofdevelopingaDNAchipthatcouldbeusedinahandheldmedicaldeviceonwhichpatientscandepositasampletogettheirDNAreadinamatterofsecondsorminutes.7Thiswouldcostbetween$100to$1,000.Thetimeframeofthisprojectisthreeyearsanditintendstomakepersonalizedmedicineaffordableandquick.

Metabolomicsisthe“systematicstudyoftheuniquechemi-calfingerprintsthatspecificcellularprocessesleavebe-hind.”8Itisthestudyofthemetabolome,theentiremeta-boliccontentofacellororganismatagivenmoment.Themetabolomerepresentstheendproductsofgeneexpression.Whiletranscriptomics(mRNAgeneexpressiondata)andproteomics(theuseofaproteomicstoolkittofacilitatebiomarkerprofiling)analysesdonottellusthewholestory

Continued on page 7

7 Data Basics • winter 2010 Return to Cover

ofwhatmightbehappeninginacell,metabolicprofilingcangiveaninstantaneoussnapshotofthephysiologyofthatcell.Metabolomicsattemptstotabulateandquantifyallthesmallmoleculeswithinasampleandtofindnewmarkersfordiseaseormetabolitepatternsasindicatorsofnutritionalstatus.Ametabolicsignatureobtainedfrombodyfluidsortissueswouldserveasametabolicfootprinttoassessindi-vidualhealthanddiseaserisk,inamannersimilartoDNAfingerprinting.

Metabolomicsisanoffshootofgenomicsandproteomics,asitallowsfortheevaluationofbothgenotype-phenotypeaswellasgenotype-envirotyperelationshipsandisbeingusedinpharmacology,pre-clinicaldrugtrials,toxicology,transplantmonitoring,newbornscreeningandclinicalchemistry.

Thehumangenomehasnowbeenfullysequencedandisfreelyaccessible.Whileapproximately2,900endogenousorcommonmetabolitesaredetectableinthehumanbody,metabolomicsisyettomaturetothatlevel.TheHumanMetabolomeProjectisa$7.5millionGenomeCanadafundedprojectlaunchedinJanuary2005.Itaimstopro-videinsightintodrugmetabolismandtoxicology,providingalinkagebetweenthehumanmetabolomeandthehumangenome.9

Geneticheterogeneitydoesnotconsiderenvironmentalandotherexternalcontributionstoindividual’sbiologicalcondi-tion;however,themetabolicphenotypeofanindividualcanbestatisticallymodeledpriortodrugadministrationandcanbeusedtopredictpost-dosepharmacokineticresponse.Systemsbiologybuildsuponthereconstructionofbiologicalsystemsusingmetabolomicdataandusesthesetoderive testablehypothesesandthepredictivemodels.Changesintheconcentrationandfluxofthemetabolitesallowforabetterunderstandingofregulatorymechanismsandmetabolicnetworks.Thishelpsfurtherinterpretwholecellbehavior.

Detailedstudiesonmetabolomicshavebeenconductedprimarilyonmicroorganismsandplants.Therehasbeenlimitedworkdoneonanimalsorhumans,withthemainfocusbeingonbiofluids,ratherthanoncells.Evaluatingchangesinmetabolicprofileswillplayakeyroleinper-forminganintegratedanalysisofgenefunctionanditsrelationshipstophenotypes,andwouldhaveasignificantimpactonbiomedicine.

Drugscreeninglibrariestypicallyscreenfordesirablephar-macokineticscharacteristicsandbiophysicaldescriptorsof

druglikenessorleadlikeness.However,inordertoentercells,drugsrequiresolutecarriersakintonaturallyoccur-ringintermediarymetabolites(endogenites)andarelikelytointeractinasimilarmannertothem.Cheminformaticsmoleculardescriptorsarenowbeingusedtoassessmetabo-lite-likenesswhichisbeingusedasacriterioninthedesignandselectionofpharmaceuticaldruglibraries.9

Whilemassspectroscopyandnuclearmagneticresonancearewell-establishedtoolsformetoboliteanalysis,therate-limitingstepcouldbereadyaccessibilitytometabolomicsdatabases.

Attheendoftheday,accessibilitytobiobanks(repositoriesofbiologicalsamplessuchasDNAortissues),withdiffer-entkindsofinformationattachedtosamplesallowingformappingtodistinctpopulationsubtypes,wouldallowonetocorrelatedata.DatasuchasfrequenciesofmarkerslikeSNPs(singlenucleotidepolymorphisms),usingautomatedmoleculartechniques,aswellasthenecessarybioinformat-icstools,genefunction,andidentifyingmarkersthatmightplayaroleintheaetiologyofcommondiseases.

ThethreemajorpubliclyavailabledatabasesthatserveascentralrepositoriesforDNAandproteindataareGenBank,maintainedbytheNationalCenterforBiotechnologyInformation(NCBI),andDNADatabankofJapanandEuropeanMolecularBiologyLaboratory(EMBL),main-tainedbytheEuropeanBioinformaticsInstitute.Similarly,datafromtheHumanMetabolomeProjectwillbefreelyaccessibleinanelectronicformattoallresearchersthroughtheHumanMetabolomeDatabaseandallcompoundswillbepubliclyavailablethroughtheHumanMetabolomeLibrary.Othercomprehensivemetabolomicdatabasesin-cludeBiGG,SetupX,BinBaseandSYSTOMONAS;whereassomedrugdatabasesincludeDrugBank,Therapeu-ticTargetDatabase(TTD),PharmGKB,STITCHandSuperTarget.Forexample,PharmGKBcontainsdataonmorethan20,000genes,morethan3,000diseases,morethan2,500drugsandmorethan53pathways.Italsohasdetailedinformationon470geneticvariants(SNPdata)affectingdrugmetabolism,andhelpsrelategeneticvaria-tionsinindividualstodifferencesinreactionstodrugs.11

Thus,enterprisedatamanagementsolutionsandinnovativecomputationalanalyticaltechniquesarerequiredtointe-gratealllevelsofbiologicaldataorganizationfromsource-to-outcomecontinuumtoallowforthecost-effectiveanaly-sisofdatausingasystemsbiologyapproach.

Pharmacogenomics and Metabolomicscontinued from page 6

Continued on page 8

8 Data Basics • winter 2010 Return to Cover

Pharmacogenomics and Metabolomicscontinued from page 7

Personalizedmedicineisaboutgettingtherighttreatmentfortherightpatientattherighttime.Itssuccesslieswiththeeffectiveintegrationofpharmacogenomics,molecularmedicine(includinggeneticandgenomicsequencing,metabolomics,andproteomics)andhealthinformationtechnologies.Healthinformaticsisdefinedastheintersec-tionofinformationscience,computerscienceandhealthcare,whichwillpotentiallyallowformoleculardatatobelinkedtootherhealthinformation,primarilybywayofelectronicmedicalrecordsthatincludebothphysicianandpatientinput.

In2003,Hewlett-PackardformedapartnershipwithPart-nersHealthCareinanefforttoaccelerateclinicalgenomicsandadvancetheconceptofindividualizedmedicinebyintegratinggeneticknowledgeintothehealthcaresystem.Microsoftpartneredwithmorethan35otherorganizationsandinstitutionstoformtheBioITAlliance,analliancethatisworkingtoadvancetranslationalandpersonalizedmedi-cinebybetterintegratingscienceandtechnologyintohealthcare.Microsofthasalsodevelopedasoftwareplat-form,calledAmalga,thatwillhelpadvancepersonalizedmedicine.1Thiscanbeusedto“assimilatelargequantitiesofdiversedata,includingelectrocardiograms,magneticresonanceimagingscans,dynamicangiograms,ultrasoundimagesandultimately,genomicinformationprovidingavisualgatewayforinstantaccesstotheinformation,andallowingresearcherstomakeandprovetheirhypotheseswithinminutesinsteadofmonths.”

PhaseIvolunteersareroutinelyscreenedfordrugmetabo-lizingenzyme(DME)polymorphisms.12ThisscreeningisincreasinglyperformedroutinelyatmanypharmaceuticalcompaniesandpatientsenrolledinPhaseIIandPhaseIIIclinicaltrialsarebeinggenotypedtocorrelateefficacywithgeneticmarkers.Finally,theeffectiveapplicationofphar-macogenomicsandmetabolomicsintheclinicwillbebasedonthereproduciblecorrelationofthesedatawithdataobtainedfromotherestablishedclinicalmeasurements.

ToquoteHenig,froma1989New York Timesarticle,“Inthenot-so-distantfuture,wecanexpecttowalkintoaphysician’sofficeforanannualphysicalandwalkoutwithablueprintofourgeneticinheritance–andwiththeknowl-edgeofthemostlikelycauseofourowndeath.”13

References:1. M. M. Zdanowicz Concepts in Pharmacogenomics, 6.2. J. L. Fackler. and A. L. McGuire. Paving the Way to Personalized Ge-

nomic Medicine: Steps to Successful Implementation. Curr Pharmacoge-nomics Person Med. 2009 June 1; 7(2): 125.

3. Theranostics: Guiding Therapy. R and D directions staff, 1 – 14.4. Medscape Today. MammaPrint™: Another Milestone in Personalized

Medical Care for Breast Cancer: MammaPrint™. Expert Rev Mol Di-agn. 2009;9(5):417-422. © 2009 Expert Reviews Ltd.

5. http://www.ornl.gov/sci/techresources/Human_Genome/medicine/phar-ma.shtml

6. Pharmacogenomics. Nature Biotechnology 18, IT40 - IT42 (2000) doi:10.1038/80079.

7. N, Limaye. Personalising bioinformatics. Pharma Asia. September 01st, 2010.

8. Life in Metabolomics and Clinical Research. P. Phaphale, Abstracts of Phamarmacometabolomics and Pharmacogenomic Research work which Presented in Metabolomics Society Conf. in Boston, USA in Aug.2008.

9. K, Davies. The Human Metabolome Project. BioIT World.com April 12th, 2007

10. S. Carney. ‘Metabolite-likeness’ as a criterion in the design and selection of pharmaceutical drug libraries. Drug Discovery Today: highlighted review 11 August, 2009.

11. http://www.metabolomicssociety.org/database.html12. D.W. Nebert. Polymorphisms in Drug-Metabolizing Enzymes: What is

their clinical relevance and why do they exist? Am, J. Hum. Genet. 60: 265-271,10, 1997.

13. R. M. Hening BODY & MIND; High-Tech Fortunetelling. The New York Times, Published: December 24, 1989.

Nimita Limaye, PhD, CCDM, holds a doctorate in biotech-nology from Pune University and is currently working as VP and Global Head (Strategic Data Services & Medical Writing) at SIRO Clinpharm Pvt. Ltd. In her current role, she contrib-utes to winning and establishing key engagements, ensuring sustainability and profitability and driving strategy for engage-ments across clinical data management, biostatistics and pro-gramming and medical writing. She has also been trained as a black belt in Lean Six Sigma and plays a key role in driving operational excellence. She is 2011 Chair of the SCDM Board of Trustees.

Please nOte:

SCDM does not sell its membership list and does not condone the use of the online membership database for electronic broadcast marketing activities.

9 Data Basics • winter 2010 Return to Cover

Trends in Clinical Data Management Due to Growing Dissatisfaction with EDCsby Mitch Scurtu, Owner, Gobal Technologies

Agrowingdissatisfactioninthepharmaceuticalandbiotechin-dustrieswithcurrentelectronicdatacapturesytems(EDCs)onthemarketcanbesummedupalongthreedirections:• Qualityofclinicaldatabases• Costandtimetodeliveraqualitydatabase• Inabilitytoaccommodatelarge-scaledatamanagementop-

erations.

Quality of clinical databaseCurrentEDCsonthemarketaremostly,ifnotexclusively,query-baseddatamanagementsystems.Asmallfractionofallerrorsinadatabasearecleanedor“prevented”up-front.Thebulkoferrorsarecleanedwitheditchecks/queriesintheback-end,afterthedatahasbeenenteredinthedatabase.Thissitua-tionismostlikelytobeoverwhelmingtodatamanagers.Youhavemanythousandsorevenmillionsofdatapointsinthedatabase.Editchecksarelookingforerrorsinthedatabasewithahitormissrateof,atbest,50percent.Outofthetypicalninetypesoferrorsinaclinicaldatabase,editcheckscancatch,atbest,threetofourtypesoferrors.Itisliketheproverbiallook-ingforaneedleinahaystack.Ifithappensthattheerrorsaretobecleanedattheendofthestudy,thisdatacleaningjobcanbecomeunreasonable,giventimelines.

Inthegoodolddays,errorsinaclinicaldatabaseweremeasuredbyfindingdiscrepanciesbetweenfinalSASdatasetsandsourcedocumentsfromtheclinic.Errorrateswereassessedconse-quently.Currently,EDCscannotcopewiththemanyerrorsinthedatabaseandfallbackonmeasuringthequalityofclinicaldatabaseswithqueryrates(i.e.,queries/page).Iftherearefewqueriesperpage,thequalityoftheclinicaldatabaseisconclud-edtobeOK.Someflowsinthiswayofthinkingare:

• WhatisanOKdatabase?Whatisthequalityassurance?Qualityis,andmustbe,ameasurableentity.Generalstate-mentsaboutthequalityofaclinicaldatabasearelesscon-vincing.

• Editchecksareprogrammedupfrontandtestedon“clean”and“dirty”mockdatatomakesuretheeditcheckisfir-ingonerrors,butnotoncleandata.Inthebackendwearedealingwithreallifeclinicaldata,whichisadynamicentityandmayormaynotconformtowhatevermocktestdatawasgeneratedupfront.ContinuousreviewingandeditingandchangingofeditchecksasrealdataisaccumulatingisatbestatimeconsuminganddifficultjobtocoordinateformostoftheEDCs.

• Whatabouttheerrorstheeditchecksdidnotcatch?Howmuchoftheerrorsdidtheeditchecknotcatchandhowaretheseerrorsplayingintothequalityoftheclinicaldatabase?

AsfarasIamaware,EDCvendorsdonotprovidetheclientwithpositivenumericalassessmentofclinicaldatabasestheway

weweredoingitwithqualitycertificatesbearingamongothers,theerrorrates.Youwerelookingatoverallerrorratesperdata-base,errorratesperSASdatasets,errorrateforcriticaldata,whichofcoursehastobezero.ValidatingtheEDCsystem,trainingallpersonnel,andvalidatingEDCprocessesareallpartsofmonsterprocessesthataresupposedtobeconducivetothequalityofclinicaldatabasesandsupposedtoreplaceaplainanddirectmeasureofthequalityofaclinicaldatabase.Thosehavemanypartsandareapoorpredictoroftheresultingqualityofthedatabase.

Cost and time to deliver a quality databaseInarecentPhaseIIB/PhaseIIIstudyweworkedinoneofthetopEDCsatatopcontractresearchorganization.Halfwaythroughthestudy,wehad52,000queries.Astheindustrycostforonequeryresolutionisaround$100,thecostofcleaningthe52,000querieswas$5.2million.Thewholedatamanage-mentbudgetforthestudywas$320,000.FourmanagersinchargeofthisstudylefttheCROsuccessively,becausetheyfeltatalossinattemptingtofitthe$5.2millioncostintoa$320,000budget.Uppermanagementpracticeinthiscaseistoshiftthecostofcleaningclinicaldatatootherfunctionslikeclinical,medical,orclinicalsitecostsandclaimdatamanage-mentasuccessfuloperation,stayingwithinitsbudget.Obvi-ously,uppermanagementisdoingthenextbestthingtheyknowof.

Needlesstosay,timetocleanthedatabase,performingallQC,QAoperations,lockingdatabaseanddeliveringclinicaldatatotheclientwasdelayedbymonthsandtheblameandresponsi-bilitywasevenlyspreadtoallparticipatingfunctionsinthestudy.Theclienthadtoquietlybeartheconsequences,sinceallclinical,sitemanagement,datamanagementandstatisticsfunc-tionswereoutsourcedtous.

Inability to accommodate large scale data management operationsMorethan50,000querieshalfwaythroughthestudywitheachqueryhavingitsownresolutioncycleoftheoreticallyfiveto10days,andpracticallytwotosixweeks,madeforcomplicated,attimesimpossible,coordinationwork.

Addtothisoneormoreadditionalstudiesandthejobofsimul-taneouslycleaningupdatabasesbecomesanightmare,ifnotimpossible.PharmaceuticalcompanieswithhundredsofactivestudiesrunningsimultaneouslycannotexpectthatasingleCROcantakecareofalltheirstudies.Therefore,theyaredistributingtheirclinicaltrialsamongmultipleCROoperations.Morerecent,pharmaceuticalcompaniesapprehensiveaboutCROperformanceareoutsourcinglessthanfulldatamanagement

Continued on page 10

10 Data Basics • winter 2010

Trends in Clinical Data Management Due to Growing Dissatisfaction with EDCscontinued from page 9

Return to Cover

servicestoCROs.Theyareoutsourcingfunctionslikeprogram-mingeditchecks,testingeditchecks,buildingdataentryscreens,developingCaseReportForms(CRFs),andenteringdataintoadatabase.ThatiswhatNelsonLee,attheSocietyforClinicalDataManagement,noticedinhisreviewofsomelessonslearnedfrombuildingafunctionalserviceproviderpartnership,publishedinInternational Clinical TrialsinMay2010.Growingdissatisfactionofpharmaceutical/biotechcom-paniesandalackoftrustintheCROs’capabilitiesofhandlingfulldatamanagementservicesinthecurrentquery-basedEDCs,forcedthesponsorcompaniestoscalebacktheirexpectationsofwhataCROcandeliver.

CurrentEDCsystemsonthemarketarequery-basedsystemsandarethereforenotcapableofaccommodatinglarge-scaledatamanagementoperations.

Trends in Clinical Data ManagementGrowingdissatisfactionwithcurrentquery-basedEDCsonthemarketistransformingtraditionalCROsintofunctionalserviceproviders.Everythingthatdoesnotworkwithcomplexandcomplicatedquery-basedEDCsispushedofftothecoordina-tionnightmareoffunctionalservicesprovision.Thetrendofintroducingquery-basedEDCsseemstobeabust,andanewcycleofthinkingandactinginthedirectionofnewsolutionstotheclinicaldatamanagementpuzzleisunderway.

BothpharmaandtheEDCvendorsarerecognizingthistrendandareattemptingto“retool”.Majorpharmacompaniesarediscretelyinquiringintotheintelligentdatamanagementmod-el,whichsomeofthemhadintheiroperationsmorethan10yearsagointheformofhomegrowndatamanagementsystems.CROs,unhappytobegivenlessofafulldatamanagementservicecontract,arerethinkingtheiroptions.ThetopCROcompanyiseagertore-introducetheintelligentdatamanage-mentmodel.Moreandmore,EDCvendorsareinvolvingintheiroperationsandespeciallyintheirnamesfeatures.OneEDCvendorevencallsitsEDCIDAM,whichstandsforintel-ligentdataacquisitionandmanagement.AnotherEDCvendorandCROserviceprovideriscallingitsEDCIDC–forintelli-gentdatacapture.

Whatisthesignificanceof“intelligence”andhowisthismak-inglifeofclinicaldatamanagersmoretolerable?

Intelligence-baseddatamanagementsystemsareintentonpreventingerrorsenteringclinicaldatabasesratherthancatch-ingthematthebackendafterdataentersdatabases.Thiscon-ceptgoesbacktotheintelligentdatamanagementmodel.The-oryofthismodeldrawsheavilyonthescienceofqualityknowninengineering,bankingandretail.Thistheorydealswithiden-

tifyingtheerrortypesinaclinicaldatabase,identifyingthesourcesoftheseerrors,andteachinghowtodevelopintelligencetocontainerrorsourcesandspecificwaysofpreventingallmechanicaltypesoferrors.Extensiveresearchintheintelligentdatamanagementmodelidentifiedninetypicalerrorsinaclini-caldatabaseandthreemajorerrorsources,whichareallstudyindependent.Eightofthesetypicalerrorsaremechanicalerrors(somedatamanagersliketocalltheseerrorsstupiderrors)andonetypewasalogicalerror.Logicalerrorsaredefinedaserrorsthatrequiresomeclinicaland/ormedicalreasoningtobere-solved.Progresshasbeenslowindevelopingacompleteintelli-gencestructuretopreventtheeighttypesofmechanicalerrorsandcontainthethreemajorerrorsources.ThisisthereasonwhypharmacompaniesandmajorCROsinvolvedintheearlypromotionoftheintelligentdatamanagementmodelyieldedtotheintroductionofquery-basedEDCmorethan10yearsago.

However,theintelligenceinfrastructureisnowacompletelymatureandknownentity,capableofdeliveringhighqualitydatabaseswithinitialerrorratesof0.1percentto0.4percent.AninitialerrorrateisdesignatedastheerrorratemeasuredasdiscrepanciesbetweenSASdatasetsandsourcedocumentsfromtheclinicatthetimethedatahaspastdataentry.Fornon-criti-calclinicaldata,thisisalreadyacceptablequalityfortheU.S.FoodandDrugAdministration.Simpledataentryintoadata-baseviaanintelligentdatamanagementmodelEDCisdeliver-ingFDA-acceptablequalitydatabases.Logicalerrorsaremadebyhumanintelligenceand,therefore,machineintelligencecannothandlethem.Logicalerrorswillhavetobecleanedoutofthedatabasetheold-fashionedwaybymanualreview.Thegoodnewsisthatstatistically,logicalerrorsareonlyasmallfractionofalltheerrorsinadatabase(0.4percentorless).Badnewsisthatlogicalerrorsarethemostimportanterrors,whichcanmakeorbreakastudy.ClinicaldatabasesworkedinanintelligentdatamanagementmodelEDCaretypicallyhighqualitydatabasesdeliveredatafractionofthecosttypicalforquery-basedEDCs.

Example:ForaPhaseIstudy,datamanagementserviceswasquotedat$80,000inatypicalEDC;samestudywasquotedinanintelligentdatamanagementEDCat$5,000.Needlesstosaywhattheclient’spreferencewas.

AtthebackendofadatabaseworkedinanIntelligentDataManagementEDC,thereareonlylogicalerrorstobecleaned,thereforetimetolockisgreatlyreduced.AndbecausetherearenoeditcheckstobeprogrammedandnoqueriestocontendwithinanintelligentdatamanagementEDC,itisideallysuitedforlarge-scaledatamanagementoperations.

12 Data Basics • winter 2010 Return to Cover

Continued on page 12

Ethical Issues in PharmacogeneticsCarol Isaacson Barash, PhD, Principal, Genetics, Ethics & Policy Consulting

Pharmacogeneticsisthestudyofhowgenesinfluenceanindividual’sre-sponsetodrugs.Thoughthefieldwouldseemtobebrandnew,itisreallyhalfacenturyold.Inthe1950s,scientistsfirstidentifieddeficienciesinenzymesthatexplainedadversereac-tionstodrugsandthatthesedeficien-ciescouldbeinherited.

Forexample,earlyresearchshowedthat10percentofAfricanAmericanmenservingintheKoreanWarbe-cameanemicafteringestingananti-malarialdrug,whichrarelycausedproblemsforCaucasiansoldiers.Pin-pointingthecausetookyearsofstudy:• Theanemicreactionwasdeter-

minedtobecausedbyavariationoftheG6PDgene,andthisvariationwasfoundtobecommonamongpeopleofAfricandescentbutnotsoamongCaucasians.

• Itwaslaterdiscoveredthatthenormalformofthegenemakesanenzymethathelpsprotectredbloodcellsagainstcertainchemicals.Lackingthatprotectiveeffect,thosewiththevariantformarevulnerabletodeleteriouseffects.

• Sincethattime,numerousotherenzymevariantshavebeenidenti-fiedandfoundtocauseadversereactions.Suchadverseeffectswereidentified,untilrecently,bytrialanderrormethods.Specifically,drugswereadministered,andanindi-vidual’smetabolismofthatdrugwastrackedbyrecordingtheamountofby-productintheirurine.

TheHumanGenomeProjecthasen-abledustoidentifythemolecularcom-positionoftheenzymesinquestionsothatwecanstudycorrelationsbetweengenotypic(genetrait)andphenotypic(physicaltrait)variability.Thesead-vanceswillincreasinglyenableustodetectindividualswhoarelikelyto

experienceadversereactionstomedi-cineswithouthavingtousepotentiallydangerousmethodsoftrialanderror.

Incomingyears,wearelikelytolearnthatparticularsinglenucleotidepoly-morphisms(SNPs)areassociatedwithsensitivitiesorresistancestochemicalcompoundsintheenvironment.Sci-entistsarenowrushingnotonlytoidentifycommonSNPs,buttodeter-minewhatdrugeffectscanbecorre-latedtothem.Thisknowledgeisen-abling“personalizedmedicine,”ortheabilitytotailortherapyforpositiveefficacyandsafety,thusavoidingtrialanderrorprescriptionsandimprovingpatientcare.

Pharmacogenomicsisarecentoffshootofpharmacogeneticswithabroaderscope.Forexample,itattemptstounderstandnotonlythemolecularcompositionofgeneticvariantsassoci-atedwithdrugresponse,butalsothebehaviorofthosevariants,includinghowthosegenesaffectdrugreceptorsites.Pharmacogenomicsisenablingdrugmanufacturerstodevelopthera-peuticagentsthataretargetedtore-ceptorsanddesignedtoreverse,ordramaticallymitigate,thesourceofthehealthproblem.Gleevac,onesuchsmartdrug,hasdemonstratedstun-ningsuccessinfightingchronicmy-eloidleukemia.Thus,theabilitytoprofileapatient’sgenevariationscanguideboththedevelopmentofnewdrugs,aswellastheselectionoftreat-mentprotocolsthatwillmorelikelyminimizeharmfulsideeffectsandensuremoresuccessfuloutcomes.

Ethical Issue #1: “Good” or “Bad” Allocation of Scarce Resources?Manybelievethatpharmacogenomics,likeothernewfieldsspawnedbytheHumanGenomeProject,representa

misallocationofresources.Ratherthanembarkonlearninghowgenesindicateapredispositiontodiseaseanddevelop-ingcuresandenhancements,orexperi-mentingwithwaystochangethehu-mangermcell,globaleffortsshouldbespentonsolvingmoreurgentproblemsfacinghumanity,suchasglobalfamineoraccessibilitytopotablewater.

Otherscontendthatpharmacogenom-ics,inparticular,offersenormouspo-tentialforclinicalbenefitstopatients,aswellaseconomicbenefitsforhealthcaredelivery.Theargumentsinfavorofpharmacogenomicsinclude:• IntheUnitedStatesalone,adverse

drugreactionsarethoughttokillabout100,000hospitalizedpatientsannually.Itisbelievedthatmanyofthesereactionsareduetogeneticvariantsandthusmanyofthesedeathscanbeavoidedbytestingpeopleforadversedrugresponsebe-foreadministeringdrugs.However,thescienceandtechnologyforsuchtestsareintheirinfancy.

• Another2.2millionpatientsincurserious,non-fatalreactions.Physi-cians,asstatedintheirHippocraticOath,areobligatedtodonoharm.Canthisobligationbefulfilledwhentheinformationavailabletophysiciansabouthowparticularmedicineswillfareintheirpatientsissomeager?Atpresent,physiciansgenerallyhavenowayofknowinginadvancewhetherthedrugtheyprescribewillorwillnotcauseanadverseeffectintheirpatients.

• Thissituationisfurthercompound-edbythefactthatmostadversedrugreactionsresultfromthefactthatmedicinesare“aone-size-fits-all.”Inotherwords,althoughmedi-cinesaretakenindifferentdosagesdependingonsymptoms,patient

Continued on page 13

13 Data Basics • winter 2010 Return to Cover

Ethical Issues in Pharmacogeneticscontinued from page 12

age,weightandotherclinicalfac-tors,thesecriteriamaynotbead-equatetoensurethataparticularmedicinewillbesafeandeffectiveforaparticularindividual.Untilrecently,therehasbeennoalterna-tivetoeitherdevelopingorprescrib-ingmedicines.Pharmacogenomicspromisestotaketheguessworkoutofdevelopingandprescribingsafeandeffectivedrugs.

Ethical Issue #2: What is a fair distribution of burdens and benefits in developing the field of pharmacogenomics?Moniesandpeople(asresearchsubjectsandasresearchers)willdevelopthefieldtothepointthatcustomizedmedicinewillbepossible.Whowillbenefit?• Theavailabilityofthisnewtechnol-

ogymaybeinitiallycostly,andthusaccessibleonlytothosewealthyenoughtopayforboththetestandthedesignerdrugbestsuitedtothem.Yet,thecostwilllikelydiminishsoastobecomeafford-abletomost.However,willlowercostsinfluenceapersontosubmittotherequiredgenetictesting,thuscreatingthreats,ifnotviolations,toone’sautonomy(thebasictenetofbioethics)?

• Researcherswhohaveinvestmentsincompaniescompetingintheirfieldmaybeinaconflictofinterestiftheyareconductingresearchforsuchacompany.Substantialconcernsaboutconflictsofinterestasbothathreattoqualityresearch,aswellastothewellbeingofresearchsubjects,haveaboundedfordecades.1

• Arecentstudyfoundthatpoliciesgoverningconflictsofinterestsatmajormedicalinstitutionsvariedconsiderablyinbothdisclosurere-quirementsandthenatureofper-mittedacademic-industryrelation-

ship.Thisopensthedoortothepos-sibilitythataninterestinfinancialgaincouldoverpoweraninterestineitherachievingvalidresearchorprotectingthewell-beingofsubjects.2

• Thereareseveralexamplesinthehis-toryofmedicalresearchwherethepatientpopulationstandingtobenefitfromadvances(i.e.,peoplewhohavedonatedtheirtime,bodies,andheartstoresearch)thoughcompensatedperstandardNationalInstitutesofHealth(NIH)terms,didnotreceivethean-ticipatedmedicalbenefitsbecausenewtherapieswereunaffordablewhentheybecamecommerciallyavailable,ornotcoveredbyinsurers.Thefollowingex-amplesillustratethis:• NumeroussufferersofGauchierDis-

easewhohelpedcompaniesdevelopsafeandeffectivetreatment(clini-calresearch),weredeniedaccesstotreatmentsbyinsurancecompaniesthatrefusedtocoverthehigh-costtherapies.Thesepatientscouldnotaffordtopaycostsoutoftheirownpocket.

• ACanavan’sDiseasesupportgrouphasbeeninstrumentalinhelpingacompanydeveloptreatmentbyrais-ingresearchfundsaswellassupply-ingresearcherswithwillingresearchparticipants.Thegroupissuingresearchfacilities,notforfinancialreturnoninvestment,butfortheopportunitytoplayanactiveroleinfurtheringresearch/treatmentgoals.3

Ethical Issue #3: Will individualized medicine be used ethically?Knowingifapersonwillrespondtoadruginwaysthataresafeandeffectiveforthatindividualwillenablepatientstoavoidmedicationsthataredangerousorineffectiveforthem.

Thisisnottosaythatgenesaretheonlykeytocures.Environmentplaysarole,

too.Dietaryandlifestylebehaviorsarestilllikelytoaffectthesafetyandefficacyofmedicinesforpar-ticularindividuals.Inaddition,variationindrugresponseisnotlimitedtomicropolymorphisms.

Environmentalfactorsalsoplayarole(suchassunexposure,drug/druginteraction,drug/foodinterac-tion).However,scientistsarepoisedtouncoverwhythemetabolismofparticularindividualsabsorbsanddispelspharmaceuticalsinaparticu-larmanner.

Considerthefollowinghypotheticalclinicalscenarioasillustratingsomeoftheethicalissuesthatcanariseinclinic:• A42-year-oldmanofScandi-

naviandescentpresentstohisphysicianwithageneralfeelingofmalaise.

• Fiveyearspreviouslyhewasdiagnosedwithhighserumcho-lesterol,whichheattemptedtocontrolwitharegimenofexer-ciseanddietaryregulation,withnosuccess.Hisphysicianthenprescribedforhimdrugtherapy.

• Beforeagreeingtotakethepre-scribedmedication,thepatientretrievedvolumesofinformationfromtheweb,including,butnotlimitedto,peer-reviewedjournalarticlesabouthisconditionandhisphysician’sfirstchoicedrug.

• Aftersixmonthsoftherapytherewasonlyamodestloweringofcholesterollevels,sothemedi-cationwaschanged.Afterninemonthsonthesecondmedica-tion,therewasstillnomarkedeffect.

• Bythetimethepatientwasabletoseehisphysicianagain,anewertherapyhadbecomeavail-

Continued on page 14

14 Data Basics • winter 2010 Return to Cover

Continued on page 15

able.Thisnewdrughadbecomethephysician’sfavorite.Thephysi-cianadvisedthepatienttoswitchtothisnewdrug,andthepatientwaseagertotryit.Threeweekslater,thepatientcametoseethephysiciantocomplainofcontinuedmalaise.

Thepatientmayhavebeenbetterservedifhehadundergonethefollow-inggenetictests,theresultsofwhichcouldhaveprovidedvaluablemanage-mentinformation:• Test1:Apre-dispositionaltestto

determinewhetherthepatienthasapolymorphismassociatedwithplaquedevelopmentleadingtocoronaryheartdisease.

• Test2:Toseewhetherthepatienthasapolymorphismassociatedwithanon-responsetothemedi-cation(thenewestmedicine).ApositiveTest2indicatesthatthepatientlacksanenzymeneededtometabolizethedrug.Theabsenceoftheenzymemeansthatthedrugisdispelledfromthebodywithoutabsorption.

• Test3:Toseewhetherthepatienthasapolymorphismwhichindi-catesthepresenceofanenzymeresponsibleformetabolizingthedosagetooslowly,makingthedruginthatdosagetoxictothepatient.

• Therationalsequenceoftestingis1-3.

Ifthepatienttestsnegative,meaningthathedoesnothavethepolymor-phismassociatedwithplaquedevelop-ment,thenhishighcholesterolposesnohealthriskandmedicationtolowercholesterollevelsarenotindicated.Ifthepatienttestspositive,meaningthathedoeshavethepolymorphism,thenheispredisposedtocoronaryheartdisease(CAD)byvirtueofbeingaplaquemaker.Inthiscase,cholesterol-loweringmedicationisindicated.

Ethical Issues in Pharmacogeneticscontinued from page 13

Ethical Issue #4: Whose right predominates?Thefatherofaresearchsubjectopenedaletteraddressedtohischildandlearnedthathischildhadenrolledinageneticresearchstudy.• Theletterindicatedthatforthe

purposeofresearch,theresearchfacilityhadobtainedsomeofthefather’smedicalrecords.Thefatherobjectedtowhatapparentlywasnon-consensualdisclosureofhismedicalinformation,evenforthepurposeofobtaininganinformativefamilyhistorytobeusedtoprovideoptimalcarefortheson/daughter.

• Outraged,thefatherphonedtheOfficeforHumanResearchPro-tections(OHRP)5oftheU.S.De-partmentofHealthandHumanServices,andprotestedthattheresearchersobtaininghisfamilyhistorywithouthisexplicitcon-sentconstitutedaviolationofhisprivacyrights.OHRP,sidingwiththefather,blockedtheoffspringfromusingthefather’sinformationandforbadeanyfurtherattemptstoobtainmoreinformationonthegroundsthatanindividual’srighttoprivacyandautonomyispara-mount.

Amongtheinterestinganddifficultissuesinthiscaseisthefactthatitchallengesustothinkdeeplyabouttheweightedvaluesweassigntofirstprin-ciples,namelytherighttoprivacy.Whoserightpredominatesinthiscase-thefather’sorthechild’s?

FederalmedicalprivacyrulesundertheHealthInsurancePortabilityandAc-countabilityAct(HIPAA)spellouttherequirementstoensureprivacyofallindividualsinthecontextofelectronichealthinformationtransmission.Al-thoughseveralsectorsofthehealthcareindustryopposedtheadoptionof

theserulesbasedoncostandimpracti-calities,HIPAAremainsabarriertounauthorizedaccessofprivatehealthinformation.TheGeneticInforma-tionNondiscriminationAct(GINA),whichpassedin2008,protectsindi-viduals,exceptthosewhoseinsuranceisobtainedthroughtheirsmallbusi-nessemployerandtheuninsured,fromsubstantialcostburdensbasedontheirgeneticprofile.However,fornumer-ousreasons,itisfarfromclearifandhowtheseruleswouldsupportthefather’sclaim.4

Ethical Issue #5: Could individuals be arguably coerced to undergo genetic testing in the context of the State’s right to protect the public?Inknowingthatthepresenceofanindividual’sgenevariantscanpredictthatcertaindrugswillnotachievethedesiredtherapeuticbenefit,itwouldnotseemfar-fetchedtoimaginethatMedicaidandMedicarewouldnotwanttopayfordrugsthatdonotwork.Furthermore,thosethatresultingreatercostburdens,suchashospital-izationtotreatadversesideeffects.Particularlygivenstatebudgetdeficits,thecostincentivetotestbeneficiariesfordrugresponseinadvanceofpre-scribing,ifnotalsothedesiretoensurepositivepatientoutcomes, wouldseemcompelling.

Forexample,roughly5percentofthepopulationdoesnotcatalyzethecon-versionofcodeinetomorphineandthusachievenopainrelief.Knowingthatnon-codeinepainmedicationisindicatedinadvanceofprescribingandrealizingthefailedefficacyisbutoneexampleofhowconsiderablecostcouldbesaved(perfectlygoodmedi-

15 Data Basics • winter 2010

Ethical Issues in Pharmacogeneticscontinued from page 14

Return to Cover

cineswouldnotbethrownawayandcostlyadversereactionswouldnotoccur).Whetherthestatewouldim-posegenetestingasaconditionofenrollmentorcontinuedbenefitsisunclear,giventhestate’slegalpreroga-tivetoenactlegislationforthepurposeofprotectingitscitizens.

ConclusionInspiteofourbesteffortstoanticipateandresolveethicalquandariesarisingfromtheapplicationofnewgenetictechnologies,itislikelythatunexpect-edconflictswillarise.Thosediscussedinthisarticlearenotintendedasanexhaustivelist.

Theethicalissueshereareremarkablysimilartothosestandardlyinvokedinpre-dispositionaltestingdiscussions.Yet,arguablythestakesarelowerhere.Theriskofpsychologicalharmis,forthemostpart,farlesssubstantialthantestingforalateonsetdisorderlikeHuntington’sdiseaseforwhicheffec-tivetreatmentdoesnotexist.Still,intheabsenceofguidanceaboutwhatconstituteshighandlowstakes,ethi-callydefensibledecision-makingre-quiresacknowledgementofthecom-petinginterestsandabroadenoughscopeofconcerntoanalyzehowanapparentlowriskcanbecomeahighriskandviceversa.

Pharmacogeneticswillpermitgeneprofilingtoanswerquestionsaboutmedicineresponses,aswellasenableresearcherstodesignbetterandsafermedicines.Thescienceanditsapplica-tionsarerealtodayandwillbeincreas-inglycommonincomingyears.Whilethelikelihoodthatindividualswillbeshutoutfromhealthinsurancebecausetheydonotrespondtoasingledrugorbecauseaparticulardrugformulaistoxictothemisextremelylow,aswouldbeemploymentexclusions(inhiring,promotingorjobresponsibili-ties),theissuesunderscoretheimpor-tanceofdebatingtheethicaluseofpharmacogeneticsmorewidely.

IntheUnitedStates,45millionpeoplelackanyhealthinsurance,andthusareatthemercyofhospitals’budgetsforunrecoverableexpenditures.Further,theseindividuals,andthemanymoremillionsofpeoplewithhealthinsur-ance,havenoaccesstosophisticatedmedicalcareduetolimitsimposedbyinsurers,especiallyfor-profitmanagedcareorganizationsandself-insuredemployers.Whethercustomizedmedi-cinewillbeavailabletoallremainsalargeunknown.Ifhistoryisahinttohowthisnewfieldwillbeused,weoughttoactnowtoensurethatthebenefitsareavailabletoall.

© 2001, American Institute of Biologi-cal Sciences.

References:1. Science Magazine, Vol. 290, 8 Dec. 2000,

p. 1873: “Studies trace patchwork of conflict policies,” B. Agnew.

2. The New England Journal of Medicine, Vol. 343, 30 Nov. 2000: “Conflict-of-interest policies for investigators in clinical trials,” Bernard Lo, Leslie E. Wolf, Abiona Berkeley.

3. Science, Vol. 290, 10 Nov. 2000, p. 1062: “Genetic Testing: Families Sue Hospital, Scientist for Control of Canavan Gene.” Eliot Marshall.

4. U.S. Dept. of Health and Human Services. Office for Civil Rights - HIPAA. “Medical Privacy - National Standards to Protect the Privacy of Personal Health Information.” http://www.hhs.gov/ocr/hipaa/

5. U.S. Dept. of Health and Human Services. “Code of Federal Regulations: Protection of Human Subjects.” http://www.hhs.gov/ohrp/humansubjects/guidance/45cfr46.htm. HHS (OHRP) home page: http://www.hhs.gov/.

Carol Isaacson Barash, PhD, is a recognized expert in genetic testing, ethics and the integration of genomic diagnos-tics into health care delivery. She works with leading non-profits, academic centers, and industry on commercializa-tion of new genetics-oriented products as well as marketing/communication, and patient education related to genetics and medical ethics.

16 Data Basics • winter 2010

4 Pharmacogenomics: Historical Perspective and Current Status, Kalow W., Methods in Molecular Biology, 2005; 311:3–15. 21

5 http://www.ornl.gov/sci/techresources/Human_Genome/faq/snps.shtml

6 Human Genome Project: http://www.ornl.gov/sci/techresources/Human_Genome/home.shtml

7 Highly Scalable Genotype Phasing by Entropy Minimization, Gusev, A. Mandoiu, I.I. Pasa-niuc, B., IEEE/ACM Transactions on Computa-tional Biology and Bioinformatics (TCBB), Vol. 5, Issue 2, April-June 2008: 252 – 261

8 Role of pharmacogenomics in drug discovery and development, A Surendiran, SC Pradhan, C Adithan, Indian J Pharmacol, Vol. 40, Issue 4, August 2008: 137-143.

9 http://clinicaltrials.gov/ct2/results?term=pharmacogenomics

10 Ethical perspectives on pharmacogenomic profil-ing in the drug development process, Amalia M. Issa, Nature Reviews Drug Discovery, Vol. 1, April 2002: 300-308

Pharmacogenomics Overview and Impact on Data Management in Clinical Researchcontinued from page 5

11 Pharmacogenetics in drug discovery and develop-ment: a translational perspective, Allen D. Roses, Nature Reviews Drug Discovery, Vol. 7, No. 10, October 2008: 807-817.

12 Pharmacogenomics: integration into drug discov-ery and development, Johnson K, Thompson J, Power A., Current Topics in Medicinal Chemis-try, Vol. 5, No 11, 2005: 1039-1045

13 Pharmacogenetic Application in Drug Develop-ment and Clinical Trials, Michael M. Shi,

Michael R. Bleavins and Felix A. de la Iglesia, Drug Metabolism & Disposition, Vol. 29, No. 4, April 2001: 591-595.

14 Advances in pharmacogenomics and individual-ized drug therapy: exciting challenges that lie ahead, Daniel W. Neberta, and Elliot S. Vesellb, European Journal of PharmacologyVol. 500, Issues 1-3, October 2004: 267-280.

15 Pharmacogenetics: potential for individualized drug therapy through genetics. Johnson JA. Trends in Genetics, Vol. 19, No. 11, November 2003: 660-666.

Pleasecontactinfo@scdm.orgifyouhavequestionsaboutregistrationforupcomingmeetings,advertising,renewalofmembershiporifyouneedtoupdateyourcontactinformation.

@Info

scdm.org

Visit us at www.scdm.org

SocietyforClinicalDataManagement,Inc.555E.WellsSt.,Suite1100Milwaukee,WI53202-3823

Phone:414.226.0362Fax:414.276.3349

E-mail:info@scdm.org

Return to Cover

Margarita Strand, CCDM, has more than 10 years of clinical data management expe-rience in oncology, CNS and cardiovascular therapeutic areas, and currently holds the position of Lead Clinical Data Manager, Oncology at Novartis.

She is an active contributor and a member of the SCDM GCDMP and Publications taskforces, and is a VIP Member of Cam-bridge Who’s Who Registry of Empowering Executives, Professionals and Entrepreneurs, Honors Edition 2008-2009.

17 Data Basics • winter 2010 Return to Cover

Submission RequirementsEditorial Board(alsoknownasPublicationsTaskforce)

Chandra Wooten, CCDM Taskforce LeaderGileadSciences,Inc.206.832.2016chandra.wooten@gilead.com

Margarita Strand, CCDMTaskforce Co-leaderNovartisPharmaceuticalsCorporation862.778.2923margarita.strand@novartis.com

Rehana BluntAmgen805.447.5729rehanab@amgen.com

Kit Howard, CCDMKestrelConsultants,Inc.734.576.3031kit@kestrelconsultants.com

Lynda L. Hunter, CCDMPRAInternational913.410.2065hunterlynda@praintl.com

Frank McDonaldGenentech,Inc.650.467.7091mcdonald.francis@gene.com

Michelle Meany, CCDMNovartisOncology862.778.6263michelle.nusser-meany@novartis.com

Derek C. Perrin, CCDMPfizerGlobalResearchandDevelopment860.732.2720derek.perrin@pfizer.com

Janet Welsh, CCDMBoehringer-Ingelheim+49735154-97905janet.welsh@boehringer-ingelheim.com

Kimberly JahnkeSocietyforClinicalDataManagement414.226.0362kjahnke@scdm.org

Tristan JohnsonSocietyforClinicalDataManagement414.226.0362tjohnson@scdm.org

Publication PolicyWewelcomesubmissionofmaterialsforpublicationinData Basics.Materialsshouldpreferablybesub-mittedinelectronicform(MSWord).AcceptanceofmaterialsforpublicationwillbeatthesolediscretionoftheEditorialBoard.Thedecisionwillbebasedprimarilyuponprofessionalmeritandsuitability.PublicationmaybeeditedatthediscretionoftheEditorialBoard.

NeitherSCDMnortheData BasicsEditorialBoardendorsesanycommercialvendorsorsystemsmen-tionedordiscussedinanymaterialspublishedinData Basics.

Advertising PolicyADRATES** x1 x2 x3 x4FULLPage $1,064each $1,008each($2,016) $960each($2,880) $906each($3,624)

HALFPage $740each $700each($1,400) $670each($2,010) $630each($2,520)

QTRPage $450each $425each($850) $402each($1,206) $378each($1,512)

**Adsarenet,non-commissionable.

Advertiserspurchasingmultipleadpackageswillhavetheoptionofplacingthoseadsanytimewithinthe12-monthperiodfollowingreceiptofpaymentbySCDM.

QuarterPage=(3⅝inchesx4⅞inches) HalfPage-Vertical=(3⅝inchesx10inches)

HalfPage-Horizontal=(7½inchesx4 ⅞inches) FullPage=(7½inchesx10inches)

MECHANICALREQUIREMENTS:Donotsendlogo/photos/imagesfromwordprocessingsoftware,presentationsoftwareorwebsites.Filesshouldbesavedinthenativeapplication/fileformatinwhichtheywerecreatedataresolutionof300dpiorhigher.AcceptablefileformatsincludeAI,EPSandhighresolutionPSD,JPEG,TIFandPDF.

PAYMENT:Paymentmustbereceivedwithadvertisement.Spacereservationscannotbemadebytelephone.ThereisNOagencydiscount.Alladsmustbepaidinfull.

CANCELLATIONS:Cancellationsorchangesinadvertisingrequestsbytheadvertiseroritsagencyfivedaysorlaterafterthesubmissiondeadlinewillnotbeaccepted.

GENERALINFORMATION:Alladsmustbepre-paid.Publisherisnotliableforadvertisementprintedfromfaultyadmaterials.AdvertiseragreestoholdSCDMharmlessfromanyandallclaimsorsuitsarisingoutofpublicationonanyadvertising.SCDMassumesnoliability,includingbutnotlimitedtocompensatoryorconsequentialdamages,foranyerrorsoromissionsinconnectionwithanyad.SCDMdoesnotguaranteeplacementinspecificlocationsorinagivenissue.SCDMreservestherighttorefuseorpulladsforspaceorcontent.

Pleasesubmitallforms,artwork,andpaymentsto: SocietyforClinicalDataManagement,Inc. Phone: 414.226.0362 555E.WellsSt.,Suite1100 Fax: 414.276.3349 Milwaukee,WI53202 E-mail: info@scdm.org Website: www.scdm.org