EMERGE & Beyond Workshop Summary

eMERGE & BeyondWorkshop SummarySharon E. Plon, MD, PhD, FACMGNational Advisory Council for Human Genome Research(NACHGR)February 12th, 2018

eMERGE & Beyond Webpage forWorkshop Summary & Video Cast

eMERGE &BeyondWorkshopGoals12Review thecurrent goals andaccomplishmentsof eMERGEIdentify conceptsand gaps ofinterest in thefield of GenomicMedicine that apotential newphase of eMERGEcould address

eMERGE Phases (2007-2019)Phase IPhase IIPhase III2007-20112011-20152015-2019 Can EMR andbiobank be used forgenomic research? Genome-widegenotyping GWAS Can genomicfindings be appliedin clinical care andhow? Clinicalimplementationpilots GWASELSI Research Can sequence data inclinically relevantgenes be used toassess penetrance andimprove clinical care? Electronic phenotypingSequencingClinical implementationGWAS

eMERGE Phase III GoalsSequence and assess clinically relevant genes in about 25,000individualsAssess the phenotypic implications of these variantsIntegrate genetic variants into EMRs for clinical careCreate community resources

The eMERGE Network

eMERGE Organization CharteMERGE SUPPLEMENTSeMERGE SUBGROUPSFamilial Implications of RORHLAInfobuttonROR LegalConsiderationsParticipant SurveyPhenotype VariablesGeocodingPhenotyping – OMOPModelHealth Care ProviderSurvey

Deliverable: Development of aneMERGEseq Platform241ANCESTRY PANEL56184FINGERPRINT SNPS 854SITE SPECIFIC272HLA 53109TotalGenes SITE'S TOP 6 1551TotalSNVsFinaleMERGEPanel Clinical reports are generated on the “Consensus ActionableList” and any specific genes or SNVs requested by individualsites

Process of return of results in Phase 3across the eMERGE sitesAll 9 sites return 68 common genes plus institutionalgenes. However, no two sites have the same process.MOST (7/9) sites follow this protocolResultto siteReview bycommitteeContactparticipantUploadto EMRReturn byGCUploadto EMRUploadto EMRContactparticipantMINORITY (2/9)of sites follow this protocolInformPCPReturn by GC orspecialist

EMR Phenotypes Development WorkflowTool Support Phenotype algorithm and data dictionary are indevelopmentCreatev Share algorithm with project teamv Standardize Phenotype Developmentv Standardize data collection Algorithm and Data Dictionary in review byvalidation site(s)Validatev Share algorithm with validation teamv Validate algorithmv Validate Data Dictionary Share and implement algorithm and data dictionaryfor multi-site data collectionSharePublishv Validate Dataset against Data dictionary Phenotype published and Algorithm is sharable topublic

Development of EMR Phenotypes ineMERGE

eMERGE & Beyond Workshop 10/30/2017Major topics discussed and recommendationsdeveloped1. Electronic Phenotyping for Genomic Research2. Evidence Generation for Genomic Medicine3. EMR Integration of Genomic Results andAutomated Decision Support4. Novel and Disruptive Opportunities in GenomicMedicine

Key Recommendations: ElectronicPhenotyping for Genomic Research Focus on developing better phenotyping methodsand technologies, such as: Continuum of disease severity rather thanabsent/present binary disease Longitudinal phenotyping to gain information onthe time course of conditions Experiment with alternative or more automatedphenotyping strategies to improve speed andefficiency Alternate approaches to manual phenotypevalidation Find more efficient ways to pool, normalize andanalyze data across consortium sitesEnvironmentPatient stateCare teamElectronic medical record

Key Recommendations: EvidenceGeneration for Genomic Medicine Develop and document best practices from thestudies already completed in eMERGE Seek appropriate balance between standardizationand experimentation with study designs Improve capture of standardized family history dataacross all site Consider different study designs including moreautomated ROR to clinicians Study the value and impact of reporting negativeresults Create data standards for new types of genomicmedicine “data objects” Ensure study designs provide for longer-termoutcomes assessment of testing, ROR anddownstream testing

Key Recommendations: EMR Integration ofGenomic Results and Automated Decision Support Further automate clinical pipelines for variant classification Develop tools and standardized displays to synthesize and presentinformation at the point of care with a focus on decision support Narrow the scope of eMERGE to develop CDS in high-priority areas Develop user-centered designs through eCDS Build foundations that promote shareable eCDS across centers Develop closed-loop CDS with automated outcome assessment tools

Key Recommendations: EMR Integration of GenomicResults and Automated Decision Support (cont) Explore patient-specific factors thatmight increase or decrease theimportance of genomic information inthe CDS Develop roadmap for naïve adopters ofgenomic data implementation in EMR Develop standard extraction of EMR forresearch across sites within and outsideeMERGE

Key Recommendations: Novel and DisruptiveOpportunities in Genomic Medicine Develop real time variant interpretationusing publicly available data (eg ClinGen)matched to patient’s phenotype Assess crowdsourcing of variantinterpretation Develop methods to efficiently reinterpret genomic results over time Apply deep learning techniques to thecharacterization of uncertain variants,drug targets, etc.ORSymptomsFamily historyEMR with Genomic data (VCF file)

Key Recommendations: Novel and DisruptiveOpportunities in Genomic Medicine (cont.) Link EMR-derived phenotypes with otherclasses of -omics data Develop innovative ways to presentsequence information (apps) to physicians Enhance clinical methods for assessingpathogenicity and penetrance Encourage patient-centered datagovernance Evaluate the legal and ethical implicationsof directly contacting relatives of patientspotentially harboring deleterious variants

Overall Summary of Recommendations Decide on the appropriate balance between innovation andstandardization across sites to increase power of data analysis Expedite variant classification by automation, machine learning andcrowdsourcing. Test innovative ways to present genomic results to physicians,patients and perform longitudinal follow-up of patients. Try to shiftsequencing and ROR early in the funding cycle. Increase efficiency of developing EMR-derived phenotypes Increase engagement of diverse patient populations on ELSI issuesrelated to genomic testing and use of electronic health records Facilitate usefulness of eMERGE tools and analyses across multipleresearch consortia and diverse healthcare systems

AcknowledgementsPlanningCommittee Rex Chisholm Dan Masys Howard Mcleod Sharon PlonWorkshopAttendees eMERGE attendees External attendeeseMERGE NHGRIProgram Staff Jyoti Gupta Sheethal Jose Rongling Li Teri Manolio Robb Rowley Ken Wileyhttps://www.genome.gov/27569445/

eMERGE & Beyond Workshop 10/30/2017 Major topics discussed and recommendations developed 1.Electronic Phenotyping for Genomic Research 2.Evidence Generation for Genomic Medicine 3.EMR Integration of Genomic Results and Automated Decision Support 4.Novel and Disruptive Opportunities in Genomic Medicine