HL7 v3 Clinical Genomics – Overview The HL7 Clinical Genomics Work Group Prepared by Amnon Shabo (Shvo), PhD HL7 Clinical Genomics WG Co-chair and Modeling Facilitator HL7 Structured Documents WG CDA Co-editor CCD Implementation Guide Co-editor GTR Implementation Guide prime editor HL7 RIMBAA WG, Co-chair

The Mission of HL7 Clinical Genomics Work Group The HL7 Clinical Genomics Work Group (CGWG) supports the HL7 mission to create and promote its standards by enabling the communication between interested parties of clinical and genomic data related to an individual. The focus of the CGWG efforts is the personalization of the genomic data – the so-call ’omics differences in an individual’s genomic – and its association with relevant phenotypic and clinical information. Associations to interpretive/expected phenotypes will be modeled as knowledge that can be utilized to transform an individual's data into meaningful information. CGWG will facilitate the development of common standards for clinical research information management across a variety of organizations -- including national and international government agencies and regulatory bodies, private research efforts, and sponsored research -- and thus the availability of safe and effective therapies by improving the processes and efficiencies associated with regulated clinical research. CGWG will strive to achieve common semantics across the clinical and research environments. Consequently, the group will start each standardization effort in Universal specifications that later on can be refined to specific realms.

Overview of Activities Three Tracks: v3: Family History (Pedigree) Topic Genetic Variations Topic Gene Expression Topic CMETs defined by the Domain v2: v2 Implementation Guides * The IG “Genetic Test Result Reporting to EHR” is modeled after the HL7 Version 2.5.1 Implementation Guide: Orders And Observations; Interoperable Laboratory Result Reporting To EHR (US Realm), Release 1 CDA: A CDA Implementation Guide for Genetic Testing Reports Common: Domain Analysis Models for the various topics A Domain Information Model (v3) describing the common semantics Semantic alignment among the various specs       Normative       DSTU       Informative

HL7 Clinical Genomics v3 Static Models Family History CDA IG Normative RCRIM LAB Reference DSTU Utilize Utilize Utilize Utilize Comments Reference Other domains Genetic Loci Constrained GeneticVariation Utilize Implementation Topic Genetic Locus Constrained Gene Expression Implementation Topic Domain Information Model: Genome Utilize Phenotype (utilizing the HL7 Clinical Statement)

To achieve semantic interoperability… …we need standard specs derived from a Central Health RIM: Clinical Guidelines EHR Pharmacy Health RIM Clinical Trials Clinical Documents encapsulate Clinical Genomics Imaging Bioinformatics Data Models Orders & Observations Central Health RIM (e.g., an extended HL7 V3 Reference Information Model): Bio & medical-informatics standard specs are derived from the same RIM

The DSTU GeneticLocus Model (deprecated) Focal Areas: Expression Data The Locus and its Alleles Sequence and Proteomics Sequence Variations (Clinical) Phenotypes

The Underlying Paradigm: Encapsulate & Bubble-up Genomic Data Sources Bridging is the challenge… Clinical Practices HL7 CG Messages with encapsulated data associated with HL7 clinical objects (phenotypes) Knowledge (KBs, Ontologies, registries, reference DBs, Papers, etc.) HL7 CG Messages with mainly Encapsulating HL7 Objects Encapsulation by predefined & constrained bioinformatics schemas EHR System Bubbling-up is done continuously by specialized DS applications Decision Support Applications Bubble up the most clinically-significant raw genomic data into specialized HL7 objects and link them with clinical data from the patient EHR

Encapsulate & Bubble-up Example Genetic Counseling DNA Lab Sequencing Example… HL7 CG Messages with encapsulated sequencing data associated with clinical phenotypes HL7 CG Messages with a Sequence HL7 Object encapsulating the raw sequencing results Knowledge Sources on genetic variants (e.g., OMIM) Encapsulation by a constrained BSML schema EHR System Bubbling-up is done dynamically by specialized applications, e.g., sequence analyzing programs Decision Support Applications Bubble up the most clinically-significant SNP data into HL7 SNP and Mutation objects and link them with clinical data from the patient EHR

Omics in the LS DAM - Molecular Biology

Omics in the LS DAM - Experiment

Omics in the LS DAM - Specimen

Gene Expression Topic (outdated) Domain Analysis Model (DAM) Passed informative ballot Based on several models for gene expression data along with extensions

The Domain Information Model - Genome Encapsulating Obj. Entry Point: Geneome Bubbled-up Obj. Individual Allele Non-locus specific data Genetic Locus Genetic Loci Expression Attributes Bio Sequence Polypeptide Expression Data genotypephenotype Sequence Variation (SNP, Mutation, Polymorphism, etc.) Variation Attributes Phenotype

Example: Family History XML Encoding Taken from a patient pedigree, the portion related to patient’s daughter (in collaboration with Partners HealthCare & other HL7 CG SIG members) Bubble up… Point back… Point back to the raw data of this relative providing “personal evidence” To phenotype and beyond….

XML Fusion: Encapsulation of Raw Genomic Data HL7 v3 XML Raw genomic data represented in Bioinformatics markup

Interpretive Phenotype The Phenotype Model Observed Phenotype Interpretive Phenotype

The Genetic Variation CMET (passed normative in Jan. 2010) Genetic Testing Order Timing issues: collecting specimen, extracting genetic material, identifying genomic observations, interpretation Genetic Loci Genetic Report (CDA) Participants (including specimen) Genetic Locus Individual Allele Associated data (vocab. Controlled) Sequence Variation Observed or Interpretive phenotypes Sequence (observed or reference)

The HL7 RCRIM CT Laboratory Model- The Pharmacogenomics Extension Clinical Trial Utilizes the Clinical Genomics CMET Specimen Enrolled Subject Pharmacogenomics Test Consent to Genotype Genetic Lab

The Gene Expression CMET Draft Participants Genetic Loci GTR Report Genetic Locus Associated observations Gene Expression

The CG V3 Query Model: Query by Parameter Phenotype parameters Starting point with query identifiers and attributes participants parameters GeneticLoci parameters GeneticLocus parameters Miscellaneous parameters

V2 Implementation Guides The IG “Genetic Test Result Reporting to EHR” passed informative ballot It is modeled after the HL7 Version 2.5.1 Implementation Guide: Orders And Observations; Interoperable Laboratory Result Reporting To EHR (US Realm), Release 1 Is used in a pilot of information exchange between Partners Healthcare and Intermountain Health Care

V2 update for January 2012 Genetic variations Cytogenetics

The v2 Message Structure

V2 Sample Message OBR|1||PM-08-J00094^HPCGG-LMM^2.16.840.1.113883.3.167.1^ISO|lm_DCM-pnlB_L^Dilated Cardiomyopathy Panel B (5 genes)^99LMM-ORDER-TEST-ID||20080702000000|20080702100909|||||||||234567891^Pump^Patrick^^^^^^NPI^L||||||20080703000000|||F||||||00000009^Cardiovascular^99HPCGG-GVIE-INDICATION^^^^^^Clinical Diagnosis and Family History of DCM|&Geneticist&Gene&&&&&NPI^^^^^^^HPCGG-LMM&2.16.840.1.113883.3.167.1&ISO|||||||||||||||55233-1^Genetic analysis master panel ^LN SPM|1|||119273009&Peripheral blood&SNM3&&&&0707Intl&&Blood, Peripheral|||||||||||||20080702000000 OBR|2||PM-08-J00094-1^HPCGG-LMM^2.16.840.1.113883.3.167.1^ISO|55232-3^Genetic analysis summary panel^LN|||20080702000000|||||||||||||||20080703000000|||F||||^PM-08-J00094&HPCGG-LMM&2.16.840.1.113883.3.167.1&ISO OBX|1|CWE|51967-8^Genetic disease assessed^LN||399020009^DCM-Dilated Cardiomyopathy^SNM3^^^0707Intl||||||F|20080702100909|||||||||||Laboratory for Molecular Medicine^L^22D1005307^^^CLIA&2.16.840.1.113883.4.7&ISO|1000 Laboratory Lane^Ste. 123^Cambridge^MA^99999^USA^B

CDA IG: Genetic Testing Report (GTR) Define an implementation guide for a genetic testing report that is both human readable and machine-processable Target at all types of GTR producers, e.g., genetic labs, clin. geneticists Readable content is larger in scope E.g., detailed description of the tests performed along with references Machine-processable should be limited, e.g., exclude raw data Ballot a Universal IG; then derive  specific types of GTR: Healthcare & Research Realm-specific guides Omic-specific guides Developed using the MDHT open source tool (OHT)

GTR - Design Principles Follow existing report formats commonly used in healthcare & research Emphasize interpretations & recommendations Provide general background information on tests performed Reference HL7 Clinical Genomics instances (e.g., v3 or v2 GeneticVariation and Pedigree) as the place holders of full-blown raw genomic data and fully-structured family history data Utilize patterns of ‘genotype-phenotype’ associations in the HL7 v3 Clinical Genomics Domain Implement them as ‘clinical genomic statement’ entry-level templates (see next slide), enabling meaningful use of the data

The Clinical Genomic Statement An abstract Clinical Genomic Statement (CGS) template that Has at its core a genomic observation (e.g., a DNA sequence variation) If it’s a reportable finding, then it should be associated with indications and interpretations, specimen and genomic source class The major finding can be associated with associated observation (e.g., amino acid change) Optionally, performers may be specified (overriding header performers) The CGS abstract template is instantiated by specialized CGS’s, e.g., for genetic variations or cytogenetics Associated Observations Indications Genomic Observation Interpretations Performers Clinical Genomic Statement Specimen Genomic Source

Narrative and Structured Data All CGS structured data items shall be part of clinical genomic statement (CGS) instances so that parsing applications can find the full semantics explicitly represented in one coherent structure In the case of the overall interpretation, it is part of CGS that has references to the various testing interpretations Sub-sections such as Indications, Interpretations and Specimen are mainly for presenting narrative, but they may also contain structured data In this way, it is possible to have less redundant documents, e.g., in the case where all tests reported in a GTR document have the same indication, an Indications section in the Summary section consists of a full-blown indication observation which all CGS indication observations reference CGS structured data may point to the respective narrative in sub-sections (by means of XML ID)

Sections order constraint GTR Overall Layout Sections order constraint

GTR Rendered – The Header Draft that has not been clinically validated

GTR Rendered – Summary Section Draft that has not been clinically validated

GTR Rendered – Genetic Variation Sections Draft that has not been clinically validated

GTR Rendered – Test Information Section Draft that has not been clinically validated

GTR Main Hierarchies Test Details Section specimen indications interpretations test performed findings test information Abstract section template w/common sub-sections: Extended by specialized sections Genetic Variations Section Genetic variations CGS Cytogenetic Section Genetic variations CGS Gene Expression Section Gene Expression CGS Abstract CGS template: Extended by specialized CGS’s Clinical Genomic Statement (CGS) Genetic Variations CGS GV Associated Observations GV Interpretive Phenotypes Cytogenetic CGS Cyt Associated Observations Cyt Interpretive Phenotypes Gene Expression CGS GE Associated Observations GE Interpretive Phenotypes

GTR UML Model - Section Outline

GTR UML Model - Summary Section

GTR Genetic Variation Section

Clinical Genomic Statement Extended by specialized Clinical Genomic Statements

Interpretive Phenotype Observation

GTR XML Snippets – Indications Section Summary Section Indication’s narrative Indication’s structured data

GTR XML Snippets – Specimen Section Specimen’s narrative Specimen’s structured data

GTR XML Snippets – Overall Interpretation Section Interpretation’s narrative Structured Interpretation

GTR XML Snippets – Genetic Variation Section Genetic Variation associated observations

GTR XML Snippets – Genetic Variation Section (cont.) Genetic Variation indication Genetic Variation interpretation

CDA GTR Ballot Status Balloted as DSTU and passed in October 2010 Still under ballot to refine & reconcile ballot comments Main issues: Vocabulary: Universal spec vs. Realm (e.g. mandate the use of LOINC code?) Binding syntax (align with new vocabulary spec and the respective SDWG guidance for CDA IGs) Layout: Semantics – compare to recommended layouts in the literature Syntactic – works closely with MDHT developers to adhere to SDWG guidelines Sections specific to every type of genetic test (derived from abstract) Section and Entry level template ids registration (when layout agreed) Suggestion to add drug safety template (considered for future use)

Alignment Among the Various Specs v3 specs and CDA are all based the RIM CDA GTR-IG will be based on CDA R3 Depending on the “right side” of R3, if it allows RIM-based domain models, then alignment is trivial v3-v2 alignment: Proposal: represent semantics with v3 and implement it in various ways, one of which is v2; develop an “v2 ITS” for the v3 models See proposal made by Amnon in a separate presentation (click here to see that presentation)

Utilizations in HL7 Clinical Trials: HL7 RCRIM Work Group (clinical trials specs) utilized the CG DSTU model (Genetic Locus) in their Pharmacogenomics message, which was an extension of the CTLab message (an approved but expired DSTU) Laboratory: The Lab Work Group might utilize a constrained version of the Genetic Variation model in their next release if the Lab Result message

Selected Implementation v2 Exchange of genetic testing results between Intermountain and Harvard v3 The Family History spec is used in Mass General Hospital Expanding to other family history applications including the US Surgeon General Family History tool The Genetic Variation model is used in Hypergenes (a European project on essential hypertension, http://www.hypergenes.eu/) The Pedigree and Genetic Variation models are used in Italy, the Rizzoli institute in Bologna CDA GTR has been used in uHealth – a PHR/EHR system in Korea

HL7 WG Health Check – Need to Improve! Active projects SWOT current 3 year plan current Mission and charter current Co-chair post-WGM survey participation Ballot presence Minutes posted since last WGM Last listserv activity Wiki presence WG conference calls schedules Steering division conference call participation Steering division co-chair (TSC representation) election participation WG rep at steering division WGM WG meetings at WGM scheduled WG has an approved DMP based on review of the updated template

Planning ahead May 2012 WGM (Vancouver) Weekly conf. calls Schedule (from Tuesday Q3 to Thursday Q2) Joint meetings AP – Wed Q4 CDS - informal OO+AP – Wed Q1 Weekly conf. calls Continue Tuesday’s 11EST Submit ‘renewed’ PSSs for GTR and Omics Prepare to ballot GTR, Omics & Sequencing storyboards in May 2012

Summary Small group coping with Various HL7 formats: v3, v2 and CDA Clinical & Research environments Developing a DAM and component models (CMETs) to be used in other HL7 domains Genetic Variation Gene Expression CDA Genetic Testing Report (GTR) Bridge from raw data to human readable reports and bubbled-up data Model-driven development of standards (use of MDHT CDA Editor)

The End Thank you for your attention…  Questions? Contact Amnon at shabo@il.ibm.com Comments of general interest should be posted to the CG mailing list at clingenomics@lists.hl7.org