BUILDING AN INFORMATION PLATFORM FOR CANCER RESEARCH & EVIDENCE-BASED HEALTHCARE William S. Dalton, PhD, MD CEO, M2Gen & Director, Personalized Medicine Institute, Moffitt Cancer Center JULY 17, 2013
MOFFITT CANCER CENTER To contribute to the prevention and cure of cancer. Third largest cancer center by patient volume in the United States-located in Tampa, Florida NCI-designated Comprehensive Cancer Centers Consistently ranked among the nation s best hospitals by U.S. News & World Report 4,500 Staff - 299 Faculty 17,000 New Patients per Year 206 beds More than $70 million in active grant support $1 billion+ Economic Impact in local community 2 2
3
4
Full Disclosure: Why we lock our doors in Florida
Science and Medicine-Moving Forward 1900s 2000s Germ Theory Chemistry Physiology Pathology Physics Find It Fix It Genomics Proteomics Metabolomics Systems Biology Informatics Micro/Nano Processing Predict It Personalize It Understanding Disease Understanding Health, Disease and Complexity 2009 RALPH SNYDERMAN 6 6
The Future: Personalized Cancer Care Science, May 26, 2006
THE FUTURE OF MEDICINE Where we are today: Sequencing of the human genome has dramatically improved molecular technology and our understanding of disease The potential is HUGE! New technology will ultimately improve molecular diagnosis, identify new targets for therapy, create personalized therapies, and identify populations at risk of disease, especially cancer. The challenge: How do we translate new molecular technologies into clinical benefits for our patients? 8
THE SCENARIO The Answer: Develop a new approach to deliver cancer care that integrates research at every step of a patient s journey dealing with cancer, and create alliances with academia, industry, the cancer care community, and patients. The Opportunity: Develop a novel delivery system to incorporate new technologies and define their value in the care of cancer patients. 9
Partners contributing to the continuous cycle of discovery, translation, and delivery of health care 10
PERSONALIZED MEDICINE: A NEW PARADIGM FOR CANCER CARE Traditional Therapeutic Approaches to Cancer Care High toxicity 50 Lung Cancer Patients Surgery, Chemo and Radiation Therapy Poor prognosis Personalized Medicine: Personalized The New Way Medicine Approach RAS-Targeted Therapies ALK-Targeted Therapies EGFR-Targeted Therapies Lower toxicities Better patient outcomes 50 Lung Cancer Patients with Molecular Data 11 11
The Challenge How do we develop an integrated network information system that is obviously necessary to deal with the overwhelming amount of information being generated in basic, translational and clinical, research needed to support personalized medicine (aka, molecular medicine, precision medicine)? 12
VISION OF TOTAL CANCER CARE Identify the needs of the patient & their families Develop an evidence-based approach to meet those needs Develop markers to predict need so they can be prevented 13
THE TCC PROTOCOL The Total Cancer Care TM Protocol May we follow you throughout your lifetime? May we study your tumor using molecular technology? May we re-contact you? 14
TCC PROTOCOL Personalized Medicine August 2012, Vol. 9, No. 6, Pages 621-632, DOI 10.2217/pme.12.76 (doi:10.2217/pme.12.76) Robert M Wenham *1, Daniel M Sullivan 1, Mark Hulse 1, Paul B Jacobsen 1 & William S Dalton 1,2 Robert M. Wenham, MD 15
KEY REQUIREMENTS Early on, Moffitt leadership identified five key elements for the development and implementation of TCC: Engagement of the Community Patient Tissue and Molecular Data Advanced Information Systems Commercial Partnerships Real time delivery of information as decision tools 16
MOFFITT, M2GEN AND TOTAL CANCER CARE Moffitt + Consortium Sites Moffitt, headquartered in Tampa, Florida, is a not-forprofit institution founded in 1986 Awarded Comprehensive Cancer Center status by the NCI Moffitt has partnered with 17 additional consortium sites nationwide to enroll patients in Total Cancer Care Key Aspects of TCC Patient Consent Can we follow you throughout your lifetime? Can we study your tumor with molecular technology? Can we recontact you? ~200 patients each week are consented to TCC for life More than 90,000 patients consented, 33,000 tumors collected, 16,000 tumors profiled and 4,000 tumors sequenced M2Gen Started in 2006 to develop commercial opportunities related to Total Cancer Care Manages consortium network operations related to TCC, including consent and tissue collection Receives clinical data on patients Obtains molecular data from patient tumors Matches treatments to patients using molecular and clinical data 17
PATIENT TISSUE AND MOLECULAR DATA: BIOREPOSITORY Patient tissues are stored in biorepository. and tissue is analyzed via genome sequencing, gene expression profiling and other analyses. Molecular analysis is used to better understand a patient s disease, and what treatment may be right for them. 18
TCC TISSUE SUMMARY 19
TOTAL CANCER CARE TO DATE 20
THE HEALTH & RESEARCH INFORMATICS PLATFORM
TOTAL CANCER CARE DATABASE: THE NEXT GENERATION HEALTH & RESEARCH INFORMATICS PLATFORM (HRI) Researcher View Cohort Identification Molecular Profiling Biomarker Discovery Comparative Effectiveness Patient View Personal Health Record Longitudinal Follow-up Personalized Search Administrators View Next Generation Health and Research Informatics Platform Operational Dashboards (i.e. new pt report) Quality & Safety Reporting Health outcomes & cost of care analysis Clinician View Point-of-care decision support (personalized medicine) Clinical Pathways Clinical Trial Matching Access for Affiliate Network 22
THE HEALTH & RESEARCH INFORMATICS PLATFORM Dana Rollison, PhD An integrated information platform that creates real-time relationships and associations from disparate data sources needed to create new knowledge for improved patient treatments, outcomes and prevention 23 23
HRI SOLUTION: ARCHITECTURE Core Front End Source Systems Cancer Registry LabVantage Data Aggregation and Storage Some representative examples of business level data domains Demographics Cancer Stage Diagnosis Treatment Drugs Labs Information Delivery Patient Cohort Examples Newly Diagnosed, Primary Pancreatic, having CEL File Capstone PathNet PharmNet Orders CEL Files Galvenon 3M VisitManager Data Sourcing Integrated Data Warehouse Data Factory Implementation Data Linkage Data Profiling Data Modeling Data Mapping Primary Breast Cancer, Survival Time >30 months, Disease Stage 1-4, Diagnosed with Type 2 Diabetes, currently on Metaformin Female with myelodysplastic syndrome, currently taking vidaza as Ist course chemotherapy, initially diagnosed in 2007-2008 24
HEALTH AND RESEARCH INFORMATICS PLATFORM The HRI Platform allows for data queries across multiple source systems, enabling realtime identification of patient cohorts for clinical trials Demographics/other non-clinical info: Gender Race Ethnicity Age at dx Vital Status Primary Site of Diagnosis, Histology, Stage at Diagnosis Basic traits for each treatment type: Chemo, radiation, surgery, hormonal therapy, immunotherapy, transplant, other Biobanking: Tumor tissue available for research CEL file available 25
TCC Data Warehouse: Phenotype Genotype; Genotype Phenotype Database is a robust, scalable dataset of oncology patients across multiple indications Database links TCC patient, clinical and molecular data and enables granular, customized and efficient searches Incorporates secondary data derived from the database clinical trial results, third-party research, etc. Standardized data format across the platform facilitates integration of data from outside sources Hub and spoke model provides standardized data quality and easy access to information 26
Four Portals to Total Cancer Care -Value Added Researcher View Cohort Identification Molecular Profiling Biomarker Discovery Comparative Effectiveness Patient View Personal Health Record Longitudinal Follow-up Personalized Search Next Generation Health and Research Informatics Platform Administrators View Operational Dashboards Quality & Safety Reporting Meaningful Use Clinician View Decision Support Clinical Pathways Clinical Trial Matching Access for Affiliate Network
TCC AS A VALUABLE RESOURCE FOR RESEARCH- SCIENTISTS The TCC Database represents an invaluable resource for research scientists, and has been utilized in several high-impact publications to-date. 28 28
SCIENTIFIC REPORTS NATURE PUBLISHING GROUP Nature Publishing Group October 2012 PMID: 2309768 29
Patient Centered Outcomes Research NCI Grand Opportunity Grant (PI: Fenstermacher, ~$4M total costs) Specific Aims: 1. Expand data management resources, integrating automated data extraction methodologies & creating user interfaces to data for researchers & clinicians 2. Unite functionalities in biomedical informatics, biostatistics, clinical trials & information technology to: Develop data quality standards to support CER Develop new methodologies to integrate data from multiple sources by generating data metrics & data representation strategies 3. Assess CER infrastructure using an iterative pilot project based on ongoing & future clinical trials to: Determine impact of CER infrastructure to reduce time and cost of performing clinical trials at TCC Consortium sites Determine information & technology gaps
Survivorship Health Outcomes Drs. Jacobsen & Malafa Florida Initiative for Quality Cancer Care Procedures for measuring quality of care in treatment of colorectal, breast, and non-small cell lung cancer at Moffitt and 10 selected oncology practices (388 cases) Provides feedback to each participating site Feedback can be used to evaluate need for and implement quality improvement efforts Data Warehouse utilized to constantly improve evidenced based guidelines
POWER OF THE TCC DATA WAREHOUSE The TCC Database represents a truly unique and powerful resource that allows for the creation of associations and rapid learning. Clinical Data Integration of patient s clinical data, tissue and molecular data provides: Tissue TCC Data Warehouse Molecular Data Ability to discover new biomarkers and drug targets Enhanced ability to match patient to right treatment or clinical trial Superior data set for clinical decision-making Improved ability to track and understand patient treatments, outcomes and costs 32 32
CLINICAL TRIAL MATCHING
CHALLENGES FACING TRIAL SPONSORS Current Challenges with Clinical Trials Trial Activation too slow Trial accrual too slow Patients do not want to leave home 80% of cancer care delivered locally Trials are searching for patients Total Cancer Care Approach Develop a multi-state Consortium for conducting trials Maintain longitudinal clinical data on patients Obtain molecular data from patients tumors Use this database to match patients to trials Shorter Timelines Fewer Patients Biomarker- Driven 34
TRIAL MATCHING CASE STUDY, HRI DEMONSTRATION AND Q/A DAN SULLIVAN, MD
Case Study: Breast Cancer Phase 2 Clinical Trial Background Drug X is a potent small molecule inhibitor of Target A, Target B and Target C These targets are over-expressed in ER+ breast cancer Hypothesis: Blocking these targets with drug X can delay drug resistance and improve clinical benefit. Tumor amplification of these targets is a predictive biomarker for drug X benefit. Amplification is over-expression of target A, B or C 36
COHORT ID TRANSMED SCREENSHOT Bread Crumbs and Linkages Multiple Views Possible Filter/view Tracking Proprietary and Confidential 37 37
PATIENT SELECTION All Patients Female Only Alive >50 y/o Primary Site = Breast Infiltrating Ductal or Mixed Histology Stage 3 or 4 ER Positive Residual Tumor Available Has CEL File 418,117 252,243 41,825 28,724 13,267 8,674 1,033 561 330 193 38
Search of TCC database for Eligible patients Clinical Eligibility Results MCC 36 clinically eligible with CEL file 65 clinically eligible with tissue that can be assayed 17 Consortium sites 157 clinically eligible with CEL file 265 clinically eligible with tissue that can be assayed 39
MOLECULAR ELIGIBILITY SEARCH
TARGET A EXPRESSION BREAST CANCER 41 41
TARGET B EXPRESSION BREAST CANCER 42 42
TARGET C EXPRESSION BREAST CANCER 43 43
Gene Expression Results Target A Target B Target C 44
Results from Clinical and Molecular Searches Clinical Search 193 clinically eligible with CEL file (go to molecular search) 330 clinically eligible with tissue available Molecular Search Target amplified include 15 MCC subjects and 82 from consortium sites Over-express 1 target only (70%), 2 targets (28%) and 3 targets (2%) Consortium sites 4 sites with 10-17 subjects (including MCC) 4 sites with 5-9 subjects 6 sites with 4 subjects 45
Next Steps to Enroll Subjects Re-contact form to sites with sufficient subjects to get clinical status update For those subjects eligible, send FFPE for definitive CLIA test for target amplification; also send tissue from those clinically eligible but no CEL file Determine sites to activate trial; use commercial IRB Contact treating MD and patient to enroll 46
Four Portals to Total Cancer Care Researcher View Cohort Identification Molecular Profiling Biomarker Discovery Comparative Effectiveness Patient View Personal Health Record Longitudinal Follow-up Personalized Search Next Generation Health and Research Informatics Platform Administrators View Operational Dashboards Quality & Safety Reporting Meaningful Use Clinician View Decision Support Clinical Pathways Clinical Trial Matching Access for Affiliate Network
MyMoffitt Patient Portal A secure gateway into a patient s personal health information. Desire to create tangible benefit to patients for participating in TCC Enhance communication with their care providers Electronic Patient Questionnaire Documented treatment plan Avenue for research Mark Hulse, RN Jen Camps 48
Background Information Patient Name: RETREAT DEMO Patient MRN: 598432 Age at Diagnosis: 63 Patient DOB: 1/5/1947 Clinical Information Site: OVARIAN CANCER SEROUS ADENOCARCINOMA TNM Pathological Pathological stage group: 3C T Status: 3C N Status: 1 Number of Lymph Nodes Examined: 10 Surgery date 11/15/2010 Surgery Date: 11/15/2010 Operation: Hysterectomy Chemotherapy 01/20/2011 to 04/15/2011 Pathological M Status: 0 Number of Lymph Nodes Positive: 02 Start Date: 1/20/2011 End Date: 04/15/2011 Name of Chemo Agents: PACLItaxel, CARBOplatin 49
50
The Patient Portal Page My TCC 51
Patient Portal Statistics 92% of new Moffitt patients
PATIENT PORTAL STATISTICS 53
Total Cancer Care TM Online Web Based Consenting 54
THE CANCER CENTER ALLIANCE: NATIONAL HEALTH & RESEARCH INFORMATION EXCHANGE FOR CANCER
DESIGNING A NEW RESEARCH & HEALTHCARE NETWORK MODEL Offices & Clinics Hospitals & Healthcare Networks Insurers Research Information Exchanges Personal Health Records Researchers Centers & Networks Researchers Genomic Data & Annotation Services Patients Dalton, Fenstermacher, et al, Clin Cancer Res; 16 (24) December 15, 2010 56
GUIDING PRINCIPLES FOR DEVELOPING THE CANCER CENTER COLLABORATION Inclusiveness Participation by Cancer Centers in their respective communities Accessibility of Data Real time access to research data through RIE using a hybrid model of a federated and central data warehouse: a network model Public-Private Partnerships Establish trusted business relationships across sectors to achieve long-term program sustainability 57
MOVING FORWARD AND ADDRESSING CHALLENGES Challenge of capturing data in the source systems themselves Creation of regional consortia that collect patient-level data (clinical, environmental, molecular, outcomes data) and expand to community hospitals and practices Development of a Hub-Spoke architecture with the Hub promoting centralized data management and the Spoke providing a data mart to create customized perspective and uses of data Development of Data Provisioning and Governance Systems for all stakeholders Development of applications to enhance querying functionality Development of a business model to build and sustain the RIE that capitalizes on private-public partnerships 58
Total Cancer Care Follow Patients Throughout Their Lifetime Database Patient Consent Clinically Annotated Biorepository Molecular Analysis THE RAPID LEARNING ENVIRONMENT Pre-Clinical Assessment Biomarker Prevalence Query Biomarker QA and Diligence Tissue-Based Biomarker Analysis Development of Companion Diagnostic Primary Xenograft- Assays of Response Cohort Feasibility Eligibility Assessment In Silico Clinical Trial Design Biomarker Expression Analysis Knock-Out Questionnaires Precision Trial Patient Outreach Trial Activation Biomarker Verification Patient Verification Patient Enrollment Adaptive Learning Signature Refinement Adaptive Design Precision Historical Control Comparison Long-Term Analysis Following Patients Throughout Their Lifetime to Identify Unmet Need Phase 4 Analysis Simulate & Model Develop Strategy for Drug Combinations and New Targets The foundation of the rapid learning system Pre-clinical validation of biomarker as determinant of response Pre-trial go/no-go decision and enhanced trial design capability Previously identified, trialready populations Superior data mining and analytical capabilities Modeling and Simulation Shortened drug development timelines and enhanced probability for patient response 59
THANK YOU!