OutcomesAssessmentAcrossMultipleSitesintheeMERGE-PGx Project

Laura J. Rasmussen-Torvik, PhD, MPH, FAHAAssistant ProfessorDepartment of Preventive MedicineCenter for Genetic MedicineNorthwestern University Feinberg School of Medicineljrtorvik@northwestern.edu

Laura J. Rasmussen-Torvik

on behalf of the eMERGE-PGxworkgroup

Outcomes

Agenda

• DescribeeMERGE-PGx

• DetailoutcomespublishedtodateineMERGE PGx• HighlightareaswhereeMERGE-PGx hasencounteredchallenges

• FuturedirectionsforeMERGE-PGx outcomesandotheroutcomesprojectstopromoteimplementation

Recruit/CollectSamples

PGRN-Seq Sequencing

ClinicalVariantValidation

ReturnResults:EHRIntegrationandCDS

Patient&ClinicianEducation

PopulateVariantandPhenotypeDataRepository(SPHINX)

SpecificAim1

SpecificAim2

SpecificAim3

eMERGEPGx - OverviewbyAim

PGRNseq:atargetedcapturesequencingpanelforpharmacogenetic researchandimplementation.Gordon AS,FultonRS,QinX,Mardis ER, Nickerson DA,SchererS.Pharmacogenet Genomics.2016Jan5.[Epub aheadofprint]

Clin Pharmacol Ther. 2014 Oct;96(4):482-9. doi: 10.1038/clpt.2014.137.

GHC/UW

NU

MSSM

VU

M/E/PSU

BCH

CCHMCCHOP

GEISINGER

MAYO

GHC/UW

NU

MSSM

VU

M/E/PSU

BCH

CCHMCCHOP

GEISINGER

MAYO

Drug-Genome pairs studyCYP2C19-ClopidogrelVKORC1/CYP2C9-Warfarin*SLCO1B1-Simvastatin* BCH DGI only VKORC1/CYP2C9-Warfarin* Geisinger and M/E/PSU also have CYP4F2-Warfarin

Pediatric Sites

Comparing site implementation details

010002000300040005000600070008000

Caucasian AfricanAmerican Hispanic Asian NativeAmerican Unknown

eMERGE-PGx samplesizebyrace/ethnicity

020040060080010001200140016001800

BCH CCHMC CHOP Geisinger GHC/UW Marshfield Mayo Mt.Sinai NU VU

eMERGE-PGx samplesizebysite

ProposedOutcomes

• Sequence9000+onpharmacogenes anddocumentvariation

• Inprogress:2nd variationpaper,HLA,CYP2D6

• Createasearchablevariantrepository

GeneticVariationamong82Pharmacogenes:thePGRN-Seq datafromtheeMERGE Network.BushWS,Crosslin DR,Obeng AO,WallaceJ,Almoguera B,etal.Clin Pharmacol Ther.2016Jul;100(1):31-3.PubMedPMID:27037844.

SPHINXisaweb-basedtoolforexploringdrugresponseimplicationsofgeneticvariationacrosstheeMERGEPGx projectcohort.

SPHINXwillcontainsdataonnearly9000subjectsfromparticipatingelectronicmedicalrecord(EMR)systems.PGRN-Seqsequencingidentifiescommonvariants,somewithknownclinicalimplications,andalsovariantsofunknownsignificance.

SPHINXhasapublic-facinggenevariantrepositoryandaprivatesearchtoolthatprovidesexploratorydatafiguresfromqueriesofvariantsummarydataplussomephenotypedata.

SPHINX:Sequence,Phenotype,andpHarmacogenomics Integration

https://www.emergesphinx.org/

ProposedOutcomes—ProcessandClinical• eMERGE-PGx sitesarecollaboratingtoreportdescriptivemetadataanddefinequantitativeandqualitativeoutcomesacrosssevendomains:recruitment,sequencing,

• Genotypevalidation– inpressatJMP

• Providereducation—inpressatPharmacogenomics

• Patienteducation– beingrevisedafterrejection

• EHRintegration-- Practicalconsiderationsingenomicdecisionsupport:TheeMERGE experience.HerrTM,Bielinski SJ,Bottinger E,Brautbar A,BrilliantM.JPathol Inform. 2015Sep28;6:50.doi:10.4103/2153-3539.165999.eCollection 2015.PMID:26605115PMCID:PMC4629307

• Actionablerarevariation:1inpress(6ACMGgenes)andAssociationofArrhythmia-RelatedGeneticVariantsWithPhenotypesDocumentedinElectronicMedicalRecords.VanDriestSL,WellsQS,StallingsS,BushWS,GordonA,etal.JAMA. 2016Jan5;315(1):47-57.doi:10.1001/jama.2015.17701.PMID:26746457PMCID:PMC4758131

Whatismissing?

• Cost

• Assessmentspost-implementation• Baselineassumption:outcomeswouldbelargelyindividuallevelandassessedthroughtheEHR

• PlannedPGx phenotypes• MACEafterclopiodogrel• MalignantHyperthermia• Methlphenidate Response• MACEonstatin

wasimplementedinahighlyheterogeneousway.Howdoweaccountforthisandcapitalizeonthisforfutureresearch?

ResearchQuestions

ForthosethatimplementedPGx CDS…

1. Howdidyoudoit?2. Howwelldiditwork?

Credit:TimHerr,NUPhDStudent,AMIApresentation

Howdidyoudoit?w didyoudoit?Assessmentprocess:

• RepresentativesfromeMERGE-PGx workgroup• Identifieddimensionsandfactsofinterest

InformalInterviews

• Multiple-choicequestionnaire• Onerepresentativepersite,viae-mail• DGIDetails,AlertCharacteristics,andOrganizationalCharacteristics

FormalQuestionnaire

• Aggregateandidentifytrends

Analysis

Credit:TimHerr,NUPhDStudent,AMIApresentation

Howwelldiditwork? work?

Results:

AlertResponse ClinicalResponse

DrugTotalAlerts Accept Override Ignore Unknown Followed NotFollowed NoAction

Codeine 114 102 0 10 2 69 18 27Clopidogrel 65 14 46 5 0 10 40 15Simvastatin 24 22 1 1 0 11 13 0Warfarin 91 45 28 18 0 34 55 2Total 294 183

(62%)75

(26%)34

(12%)2

(0.8%)124

(42%)126

(43%)44

(17%)

Credit:TimHerr,NUPhDStudent,AMIApresentation

ConclusionsofCDSanalysis

• Despitetheseareasofagreement,thereissignificantvariationinhowPGx CDSalertsaredesignedthroughouttheeMERGE Network.

• Combined,thesedifferencescreateasignificantbarriertoanalyzingaggregatephysicianresponseviaalertlogdataalone.

• Instead,wefoundthattheeMERGE Network hascreatedaseriesofnaturalexperimentswithavarietyofalertdesignandDGIchoices.

• Single-sitestudiescancomparephysicianresponseacrossDGIsonsimilartechnicalinfrastructure.Multi-sitestudiescouldfocusoncloselytargetedanalysesofspecificDGIswheredesignchoicesallowmeaningfulcomparisons.

Credit:TimHerr,NUPhDStudent

Implementationresearchoutcomes• Adoption

• Adherence

• Fidelity

• Levelofimplementation

• Sustainability

CONCLUSIONS*:• Implementationoutcomesinstrumentationisunderdevelopedwithrespecttoboththesheernumberofavailableinstrumentsandthepsychometricqualityofexistinginstruments.Untilpsychometricstrengthisestablished,thefieldwillstruggletoidentifywhichimplementationstrategiesworkbest,forwhichorganizations,andunderwhatconditions.

*LewisCC, FischerS, WeinerBJ, Stanick C, KimM, MartinezRG.Outcomesforimplementationscience:anenhancedsystematicreviewofinstrumentsusingevidence-basedratingcriteria.ImplementSci. 2015Nov4;10:155.doi:10.1186/s13012-015-0342-x.

Site Gene/DrugPairs

PGxServiceType

(#years) CDSSupportAdopters/Regular

UsersChildren'sHospitalofPhiladelphia(CHOP)

GenotypingonlyCYP2C9CYP2c19VKORC1-1639G>A

GCmodel

(1year)

ResultssummarylettersenttoPCP;PGxspecificlettergiventopatientstosharewithprovidersoutsideEHR

Genetics,Neurology-offeredaspartofWES

CincinnatiChildren'sHospitalMedicalCenter

CYP2D6/codeineCYP2D6/SSRIs,tricyclicantidepressantsCYP2C19/SSRIs,tricyclicantidepressantsCYP2C9/warfarinVKORC1/warfarinTPMT/thiopurines

Prescriberpointofcare

(8years)

Passivealertsinelectronicorderingsystem;Resultreportswiththerapeuticrecommendationsincludingdoseadjustmentordrugselectionalternatives

Psychiatryintegratedintostandinginpatientordersby2005;infrequentusebyotherprescribers

GroupHealth GenotypeonlyHLA-B*15:02

Prescriberpointofcare

(8years)

PopupalertwhenevercarbamazepineprescribedregardlessofHLAstatus

Neurology,psychiatryandsometimesprimarycare

MayoClinic CYP2D6/selectSSRIs Prescriberpointofcare

(8years)

None Psychiatryphysiciansandnurses

VanderbiltUniversity CYP2C19/clopidogrelSLCO1B1/simvastatinCYP2C9/warfarinVKORC1/warfarinTPMT/thiopurinesCYP3A5/tacrolimus

Preemptiveandreactivegenotypingorderedbyprovider

Activeandpassivealertsinelectronicordering;pharmacysupportforspecificusecases

Physiciansandotherproviderswhoprescribe

Table1.SiteswithPGxImplementationandEducationPriorto2012eMERGEPGxInitiative

Lessonslearned

• Plantocaptureoutcomes(includinginstitutional-leveloutcomes)inadvance

• Decide:Arewefocusedonclinicaleffectiveness?Implementation?Both?

• Validatedimplementationoutcomesareneeded

• PGx researchersneedtolearnhowtoshareimplementationchallenges

Acknowledgements

• MembersofthePGx workinggroup(andothereMERGE workinggroups)

• Currentco-chair:CindyProws

• Formerco-chairs:DanRoden,JoshDenny

Questions?

•eMERGE-PGx continues………

PGxGene

NumberofvariantsthathaveCPICguideline(knownfunctionanddosingrecommendation)

HowmanyvariantsincolumnBareontheeMERGE"SNPlist"

TotalNumberofSNPsAssociatedwithGeneonSNP_List

NumberofadditionalvariantsonCPICwithpartialknowledge

CYP3A5 3 1 1 0

CYP2C9 2 2 215

(someknownfunction,nodosing)

CYP2C19 8 8 8

6('decreased'function&weaker

dosing+2withnofunctionbutveryrareaddedin2015)

TPMT 4 4 4 0

SLCO1B1 3 1 19

("possible"function&weakerdosing)IFNL3/IFNL4 1 1 1 0VKORC1 1 1 1 0

DPYD 3 3 6

7(withsomeknownfunctionbutno

dosingchange)WillnotreportforeMERGE

CYP2D6 18 16 45HLA-B ?? IMPUTE? 275

RYR1-isingenelistCPICguidelineinprogress ? ?

CFTR-isingenelist 13 ? ?G6PD-notinpanel >100 0 0

UGT1A1-notinpanel 4 0 0 0

ConsiderNOTreportingas2majorvariantsarenotonpanel(limitationforAfricanAmericansandLatinos)-AAcouldcomeacrossas*1/*1(*3isWT)butbe*6or*7ReportonlyforWarfarin(asforphenytoinwouldneedHLAB)

Willnotbereportedclinicallyaswecan'tgetCNVcalls

reportwithCYP2C9forwarfarin

Willnotbereported-wouldneeddiffguidelineinprogress-couldadd?;notinPharmCATyet

PharmCATdoesnotreportG6PD

PharmCATreportsCFTR;limitedutilityineMERGEpatients,willnotreport

2015guidelineforatazanavir;PharmCATdoesreportit

Comments

ThevariantpresentoneMERGEpnl(rs4149056)tagsall*alleles(*5,15,17).Allareassociatedwithdecreasedfunction,so

samerecommendation

SuggesttonotreportCYP3A5

Laboratory for Molecular MedicinePartners Healthcare Personalized Medicine65 Landsdowne Street, Cambridge, MA 02139Phone: 617.768.8500 Fax: 617.768.8513Director:CLIA#: 22D1005307

Referring Physician:

Test performed eMERGE III Sequencing panel

Important disclaimers:

Additionalnotesareincluded(columnB).ForakeypleaseseeAppendixB

Results:

Dip

loty

pe N

otes

Dip

loty

pe

Phe

noty

pe

Dip

loty

pe

Phe

noty

pe

Dip

loty

pe

Phe

noty

pe

Dip

loty

pe

Phe

noty

pe

Dip

loty

pe

Phe

noty

pe

Dip

loty

pe

Phe

noty

pe

1 1, 6, 8, 9 PM09-00655-A 38432489 16-90081 12/31/16 03/01/17 *1,*1 nl metCYP2C9*1/*1; VKORC1 Hap A/Hap A int met *1,*1 int met *1,*1 nl met *1,*1 nl met *1,*1 nl met

This test does not report all pharmacogenomic variants that might alter protein function. Therefore, a normal result does not exclude the possibility that a patient has an increased, intermediate or poor metabolizer phenotype. This risk may vary among ethnic groups. This assay cannot determine if multiple variants are present on the same copy or different copies of the gene, leading to an inability to definitively assign a diplotype and phenotype. This test does not detect copy number variants (CNV). Extrinsic factors (e.g. diet, smoking status, co-administered medications) and intrinsic factors (e.g. gender, age, weight, renal or hepatic function) may affect drug response. In addition, certain ethnic populations may have an increased prevalence of rare genetic variants not reported by this assay that could affect drug response. These factors need to be taken into consideration when interpreting genetic test results. The CPIC Guidelines and PharmGKB website should be consulted for interpretation of all results presented here. These guidelines and frequent updates are found at https://cpicpgx.org/ and https://www.pharmgkb.org/. Please consult a clinical pharmacologist for additional information.

Geneticist Approval

PHARMACOGENOMICS REPORT

SLCO1B1NM_006446

.4CYP2C19

NM_000769.1

TPMTNM_000367

.2

CYP2C9NM_000771.3

VKORC1NM_024006.5

IFNL3NM_172139.2

DPYDNM_000110

.3

IND

EX

AC

CES

SIO

N

LAST

NA

ME

(PA

TIEN

T ID

)

FIR

ST N

AM

E (S

TUD

Y ID

)

DA

TE R

ECEI

VED

DA

TE R

EPO

RTE

D

SAMPLEPGx REPORT

TPMTc.238 TPMTc.460 TPMTc.719 TPMTc.626-1 DiplotypeWT(G/G) WT(G/G) WT(A/A) WT(G/G) TPMT*1/*1WT(G/G) WT(G/G) WT(A/A) MUT(A/A) TPMT*4/*4WT(G/G) WT(G/G) WT(A/A) HET(G/A) TPMT*1/*4WT(G/G) WT(G/G) MUT(G/G) WT(G/G) TPMT*3C/*3CWT(G/G) WT(G/G) HET(A/G) WT(G/G) TPMT*1/*3CWT(G/G) MUT(A/A) WT(A/A) WT(G/G) TPMT*3B/*3BWT(G/G) MUT(A/A) MUT(G/G) WT(G/G) TPMT*3A/*3AWT(G/G) HET(G/A) WT(A/A) WT(G/G) TPMT*1/*3BWT(G/G) HET(G/A) HET(A/G) WT(G/G) TPMT*1/*3AWT(G/G) HET(G/A) HET(A/G) HET(G/A) TPMT*3A/*4HET(G/C) WT(G/G) WT(A/A) WT(G/G) TPMT*1/*2HET(G/C) WT(G/G) WT(A/A) HET(G/A) TPMT*2/*4HET(G/C) WT(G/G) HET(A/G) WT(G/G) TPMT*2/*3CHET(G/C) HET(G/A) WT(A/A) WT(G/G) TPMT*2/*3BHET(G/C) HET(G/A) HET(A/G) WT(G/G) TPMT*2/*3AMUT(C/C) WT(G/G) WT(A/A) WT(G/G) TPMT*2/*2

DiplotypeTestResultforTPMT

CodedGenotype/Phen

otypeSummaryb

EHRPriorityResult

NotationcDrug LMMInterpretationandBackground CPICConsultation(Interpretation)TextProvidedwithTestResultd

Azathioprine(immunosuppressant)

LowerconcentrationsofTGN(thioguaninenucleotide)metabolites,highermethylTIMP(secondarymetaboliteofmercaptopurine),thisisthe"normal"pattern.Startwithnormalstartingdose(e.g.,2–3mg/kg/d)andadjustdosesofazathioprinebasedondisease-specificguidelines.Allow2weekstoreachsteadystateaftereachdoseadjustment

Mercaptopurine(immunosuppressant)

LowerconcentrationsofTGN(thioguaninenucleotide)metabolites,highermethylTIMP(secondarymetaboliteofmercaptopurine),thisisthe"normal"pattern.Startwithnormalstartingdose(e.g.,75mg/m2/dor1.5mg/kg/d)andadjustdosesofmercaptopurine(andofanyothermyelosuppressivetherapy)withoutanyspecialemphasisonmercaptopurinecomparedtootheragents.Allow2weekstoreachsteadystateaftereachdoseadjustment

Thioguanine(immunosuppressant)

LowerconcentrationsofTGN(thioguaninenucleotide)metabolites,butnotethatTGNafterthioguanineare5-10xhigherthanTGNaftermercaptopurineorazathioprine.Startwithnormalstartingdose.AdjustdosesofthioguanineandofothermyelosuppressivetherapywithoutanyspecialemphasisonTG.Allow2weekstoreachsteadystateaftereachdoseadjustment

TPMT,*1/*1,Extensivemetabolizer.Thiopurines:Standarddosage.Thiopurinemethyltransferase(TPMT)metabolizesthiopurineprodrugs(azathioprine,6-

mercaptopurine,andthioguanine)intotheiractivethioguaninenucleotide(TGN)

metabolites.TPMTalleles*3A,*3B,and*3C,leadtoreducedproteinexpressionand/orfunctionandcarriersmaybeatriskoflife-threateningmyelosuppressionwhentreatedwithstandarddosesofthiopurines.Adapted

fromRelling2011.

*1/*1TPMTExtensivemetabolizer

Normal/Routine/LowRisk

PLANNEDSUPPLEMENTALINFORMATION

Phenotype EHRPriorityResultNotation

DrugswithCPICGuidelines

DosageSummary* References-seeAppendixD

GeneBackground

CYP2C9Normalmetabolizer&NormalVKORC1expression

CYP2C9Normalmetabolizer&IntermediateVKORC1expression

CYP2C9Normalmetabolizer&LowVKORC1expression

CYP2C9Intermediatemetabolizer&NormalVKORC1expression

CYP2C9Intermediatemetabolizer&IntermediateVKORC1expression

CYP2C9Intermediatemetabolizer&LowVKORC1expression

CYP2C9Poormetabolizer&NormalVKORC1expression

CYP2C9Poormetabolizer&IntermediateVKORC1expression

CYP2C9Poormetabolizer&LowVKORC1expression

CYP2C9Normalmetabolizer Normal/Routine/LowRisk Standarddosage

CYP2C9Intermediatemetabolizer Abnormal/Priority/HighRisk

CYP2C9Poormetabolizer Abnormal/Priority/HighRisk

Allshouldconsultsourceslistedatright

(colD).Warfarin(anticoagulant)

Usethealgorithmsavailableon

http://www.warfarindosing.organdFigures2&3ofJohnson_2017

Johnson_2011,Johnson_2017

CommongeneticvariantsinthecytochromeP450-2C9(CYP2C9)andvitaminK-epoxidereductasecomplex1(VKORC1)enzymeshaveaneffectonwarfarinmetabolism.CarriersoftheCYP2C9*2,CYP2C9*3,andVKORC1-1639Aallelesareatincreasedriskofbleedingandmayrequirelowerdosesofwarfarin.(Johnson_2011)Thealgorithmsavailableonhttp://www.warfarindosing.orgshouldbeusedtodeterminewarfarindosing.The2016guidelineupdate(Johnson_2017)includesadditionalallelesinCYP2C9,CYP4F2,andtheCYP2CclusterthatareNOTincludedinthistestpanel.Theupdatealsohasimportantancestry-specificdosingrecommendations,andmustbecarefullyreviewed.

Phenytoin/fosphenytoin(anticonvulsant) Caudle_2014

TheCPICguidelinesforphenytoinincludethegenesCYP2C9andHLA-B.(ThistestdoesnotincludeHLA-B;theHLA-B*15:02allelehasbeenassociatedwithanincreasedriskofSJS/TEN.)However,SupplementalTable13bofCaudle_2014providessomerecommendationsbasedonCYP2C9diplotypeintheabsenceofHLA-Btesting.Decreasedosage

CYP2C9&VKORC1-theseareinterpretedtogetherforwarfarindosing.

TPMT

TPMTNormalmetabolizer Normal/Routine/LowRisk

Azathioprine,Mercaptopurine,

Thioguanine(thiopurines,immunosuppressant)

Standarddosage

Relling_2011,Relling_2013

Thiopurinemethyltransferase(TPMT)metabolizesthiopurineprodrugs(azathioprine,6-mercaptopurine,andthioguanine)intotheiractivethioguaninenucleotide(TGN)metabolites."IndividualswhoinherittwononfunctionalTPMTallelesareat100%riskforlife-threateningmyelosuppression,duetohighTGNs,iftheyreceivechronictherapywithconventionaldosesofMP(orazathioprine).DespitehavinghigherTGNsthanwild-typehomozygotes,only~30–60%ofpatientswhoareheterozygousforTPMTareunabletotoleratefulldosesofMPorazathioprine"(Relling_2011).

TPMTIntermediatemetabolizer Abnormal/Priority/HighRisk Decreasedosage

TPMTPoormetabolizer Abnormal/Priority/HighRisk

Greatlydecreasedosage

Themesinimplementationscience

• Understandingthe“voltage”gap

• Successfulapproachestoprogramdissemination

• Tensionbetweenlocalcontextandfidelity