TRACK 4: Bridging Silos in Biomarker Development
Day 1 | Day 2 | Download PDF
TUESDAY, SEPTEMBER 22
2:00-3:00 pm Conference Registration
3:00-3:15 Welcoming Remarks
Phillips Kuhl, Cambridge Healthtech Institute
3:15-3:45 The FDA’s Role in Improving Drug Development
Douglas C. Throckmorton, M.D., Deputy Director, Center for Drug Evaluation and Research, U.S. Food and Drug Administration
There are many pressures on the medical products endeavor, including the need for more timely and efficient development to support their marketing, and then assessment after marketing to support their best uses. There are many stakeholders with important responsibilities in the healthcare system. As regulators, the FDA has a clear role in responding to these pressures, one that includes both providing a clear path to efficient development as well as a critical role in supporting innovation and collaboration.
3:45-4:15 Oncology Clinical Trial Design: Opportunities for Rational (and Irrational) Incorporation of Biomarkers to Achieve the Goal of Individualized Therapy
Mark J. Ratain, M.D., Leon O. Jacobson Professor of Medicine; Chairman, Committee on Clinical Pharmacology and Pharmacogenomics; Associate Director, Clinical Sciences, Cancer Research Center, The University of Chicago
Biomarkers are often hailed as a panacea to reduce attrition rates of oncology drug development, thereby decreasing drug development costs. To date, the incorporation of predictive biomarkers has had mixed results, including important successes (e.g., HER2 testing for trastuzumab) and failures (e.g., EGFR testing for cetuximab). Pharmacodynamic biomarkers are similarly of theoretical value to accelerate decision-making, but in practice have had limited utility due to lack of technical validation and misconceptions of the value of a biomarker of unknown clinical importance. Novel trial designs incorporating predictive and pharmacodynamic biomarkers will be discussed.
4:15-4:45 Personalized Medicine Depends on Drug Pipeline-Efficacy Pharmacogenetics to Create New Targeted Therapies
Allen D. Roses, M.D., Jefferson-Pilot Professor of Neurobiology and Neurology; Director, Deane Drug Discovery Institute; Senior Scholar, Fuqua School of Business; Member, Duke Institute for Genome Sciences & Policy, Duke University School of Medicine
“Personalized medicine” has become a “hot topic,” discussed by many but practiced by few who are accountable for discovering and developing new medicines. I will present the viewpoint that targeted medicines that are reimbursable will drive the incentives of personalized medicine commercially. Currently academic and external investigators have the opportunity to test medicines independent of the sponsors only after they appear on the market. This creates a negative influence on drug developers in that new adverse events and focused efficacy occurs post-marketing, after a price has been set. Each reduces the potential market. Payers understand this and are willing to reimburse safe and effective medicines of value. The time to create that scenario is during drug development, especially with respect to pipeline efficacy. A medicine increases value when the “right” patients can be identified – except, when these data appear in the post-marketing period, the price never goes up with the value. Great strides in pharmaceutical development will be fueled by the prospective and integrated use of pipeline pharmacogenetics and encouragement of informational conversations with regulatory authorities, including but not limited to Voluntary eXploratory Data Submissions [VXDS].
4:45-5:15 Disease Biology as a Precompetitive Space: Emerging Opportunities for Distributed Contributors to Jointly Evolve Disease Models
Stephen H. Friend, M.D., Ph.D., President, Sage Bionetworks
Significant advances in generating probabilistic casual models as pioneered by Eric Schadt and colleagues at Rosetta Inpharmatics over the last five years have afforded an opportunity to share data not as linear files but as they reflect onto predictive models of disease. Examples will be shown that highlight the power of such models in metabolic and oncologic diseases. Emphasis will be placed on how classical target and pathway representations of disease miss capturing the essential biology needed to generate biomarkers and to develop drugs. Also to be discussed are the new organizational structures being built at the private/public interface that will allow investigators to modify and evolve each other’s models of biology. We posit that when biological data can be refl ected onto evolving disease models that clinicians and scientists in academia and industry will be able to enter into connected projects that can evolve the resolving power of disease models in ways natural to chemists but alien to most biologists.
.jpg?n=3649 "Caprion")
5:15-6:15 Grand Opening Reception in the Exhibit Hall Sponsored by
WEDNESDAY, SEPTEMBER 23
7:00 am Conference Registration Open
7:30 am Morning Coffee
8:25-8:30 Chairperson’s Opening Remarks
Stephen Naylor, Ph.D., Chief Executive Officer and Chairman, Predictive Physiology & Medicine, Inc.
8:30-9:00 Systems Approach to Biomarker Identification
Gordon B. Mills, Ph.D., Chair & Professor of Systems Biology, University of Texas, M.D. Anderson Cancer Center
Many different approaches are available to characterize DNA and RNA. However, there is a major need for a high-throughput functional proteomics approach. We have analyzed over 300 cell lines and 3000 patient samples with reverse phase protein arrays (RPPA) encompassing over 150 antibodies of relevance to cancer. These data have identified a number of important signaling processes in different cancer lineages and their relationships to underlying genomic aberrations. Further, we have identified a series of protein markers able to classify tumors and predict outcomes. Validation of these markers could result in marked improvements in patient outcomes.
9:00-9:30 Systems Level Derived Biomarkers in Personalized Medicine
Stephen Naylor, Ph.D., Chief Executive Officer and Chairman, Predictive Physiology & Medicine, Inc.
The complexity of individual human physiology is often overlooked in the search for specific and sensitive biomarkers of health, wellness and disease. The use of a systems-level approach to determine the molecular bioprofile of an individual’s cardiovascular health will be described and compared and contrasted with individual and panels of similar biomarkers currently used in cardiovascular disease.
9:30-10:00 Practical Applications of Mechanistic Modeling
Keith Elliston, Ph.D., President and Chief Executive Officer, Genstruct, Inc.
The field of systems biology has grown to encompass both large-scale, multi-Omics mechanistic modeling approaches, as well as dynamic and quantitative modeling of biological systems. Large-scale mechanistic modeling has been applied to the discovery of new biological mechanisms, and has lead to the identification of mechanistic biomarkers. The identification of a molecular mechanism can directly lead to the discovery of relevant biomarkers, and can define activity states of systems which effectively obviate biomarkers. An approach to mechanistic modeling powered by multi-Omic datasets will be presented, along with two relevant case studies in cancer and metabolic disease.
10:00-10:30 Modeling and Simulation for Biomarker Identification and Personalized Medicine
Alex L. Bangs, Chief Technology Officer & Co-Founder, Entelos, Inc.
Disease physiology models offer a novel approach to biomarker identification. “Virtual patients” representing real-world patient diversity can be simulated under many experimental protocols, providing a rich data set for identifying biomarker candidates. Specific case studies will be presented highlighting this approach for biomarkers of efficacy, safety, and prognosis. An extension of this approach, using biomarkers and simulation for personalized medicine, will also be highlighted.
10:30-11:30 Networking Coffee Break with Poster and Exhibit Viewing
Sponsored by
11:30-12:00 Current Information Trends in Biomarker Research
Jorge Manrique, Senior BioDiscovery Consultant, Thomson Reuters
Biomarkers are becoming a key tool in enhancing the productivity of pharmaceutical research & development, both in discovery and the clinic and an essential element for regulatory purposes. It will become increasingly difficult to analyze and manage the rapidly increasing information about biomarkers across all of their utilities. A new fully indexed biomarker database, BIOMARKERcenter, will help to address this problem. Using BIOMARKERcenter we will show how biomarkers mimic the lifecycle of a drug, from discovery to approval, and show the diversity of roles and techniques currently being employed in biomarker research.
Sponsored by
12:00-12:15 Analysis of Phosphorylation Sites for Biomarker Identification and the Mechanism of Kinase Inhibitors
Wei Liu, Ph.D., Wyeth Research
A custom database of more than 8,000 phosphoylation sites for 2,793 proteins from published data was constructed. Using phosphorylation patterns for PI3K and MEK inhibitors experimentally observed in EGF-stimulated A431 cells, the custom database was used to explain the activity of kinase inhibitors including identifying off-target kinases, reconstructing the phosphorylation cascade and identifying activity biomarkers. Application workflows will be described for just phosphorylation data and how integrative “omics” analysis could be structured.
Sponsored by
12:15-12:30 Integration of Proteomics, Metabolomics, and Genomics Data in GeneSpring GX Software
Pam Tangvoranuntakul, Ph.D., GeneSpring Product Manager, Agilent Technologies
Multi-assay studies are central to systems biology research. Thus, the need for software applications that support integrative data analysis becomes critical to the discovery of linkages and concordance between different data types such as genomic copy number, gene expression, proteomics, and metabolomics. This session will focus on tools in GeneSpring GX that will allow researchers to make these discoveries. Methods of integrative data analysis in GeneSpring GX will be demonstrated by comparing genomic copy number, gene expression, and proteomics data.
12:30-2:00 Lunch on your own
2:00-2:30 IT Frameworks for Supporting Analysis of Complex Data Sets in Translational Medicine Research
Gary K. Mallow, Ph.D., Director, Merck Global Services: MRL IT
Being able to coherently assemble complex data sets and then provide those to analysts is a necessary component of modern translational medicine research. For example, mutations in signaling pathways are emerging as good predictors of response efficacy for small molecule therapies. Developing those signatures requires integrating data from preclinical studies with clinical patient data so that preclinical signatures can be related to patient information and clinical subtypes, preclinical predictive hypotheses validated, clinical datasets identified to predict response to therapy, and trials designed to test those hypotheses. This talk will describe a high-level IT framework designed to support robust analysis, including data Standards, IT infrastructure, QA and QC processes, warehousing and data query and dataset assembly. It will also suggest areas where maturation of IT frameworks is needed.
2:30-3:00 New Biomarkers: Opportunities and Challenges
Anastasia M. Khoury Christianson, Ph.D., Senior Director and Global Discipline Leader, Biomedical Informatics, AstraZeneca R&D
There has been an increasing need for new or better Biomarkers to evaluate safety and efficacy in the clinic. Identifying such markers is part of the challenge, while evaluating them and deciding on the best marker(s) to use in a particular study might be an even greater challenge. This presentation will describe an integrative approach for Biomarker identification and evaluation which takes advantage of existing knowledge. It will also consider ways of capturing Biomarker knowledge for reuse.
3:00-3:30 A Comprehensive Combinatorial Biomarker Discovery Strategy
Raymond Ng, Ph.D., Professor, Department of Computer Science, The University of British Columbia
Monitoring cardiac transplant subjects for acute rejection requires the use of endomyocardial biopsy, an invasive and expensive procedure. Over the past four years, the Biomarkers in Transplantation team developed single platform Omic biomarkers for the diagnosis of cardiac allograft rejection. The team establishes a comprehensive analysis pipeline for combining genomic, proteomic, and clinical variables. The combinatorial biomarker panel identified using this analysis pipeline has a superior performance to that of the single platform Omic biomarkers.
3:30-4:30 Networking Refreshment Break with Poster and Exhibit Viewing
4:30-5:00 Molecular Synergy of Driver Genes in Colon Cancer
Mark Chance, Ph.D., Professor and Director, Center for Mass Spectrometry & Proteomics, Case Western Reserve University
Candidate driver genes that are frequently mutated in colorectal cancer (CRC) and other cancers have been discovered by sequencing and genotyping of human colorectal tumor biopsies. These frequently mutated genes lie on well-connected sub-networks of proteins in the human interactome. We show that when considered together, the proteins (crosstalkers) in these sub-networks are more robust discriminators of disease than single genes or gene signatures. Furthermore, studies of established mouse models illustrate driver gene synergy in intestinal epithelial cancers (e.g., p21-/- and APC1638N+/- single mutant mouse models have modest phenotypes while the double mutant mouse exhibits an aggressive tumorigenic phenotype). With a view to identifying molecular signatures that mediate the cross-talk between these drivers and their interactions, we have developed a bioinformatics pipeline that will integrate various levels of molecular interaction data to identify CRC network signatures.
5:00-5:30 Leveraging Biomarkers: Integrating Genomic and Proteomic Data for Biomarker Discovery
Karin Rodland, Ph.D., Science Lead, Biological Sciences, Pacific Northwest National Laboratory (PNNL)
Efforts to identify biomarkers for early diagnosis or prognosis of cancer and other disease have often focused on a singular molecular species, with preference given to mRNA, micro RNA, proteins, auto-antibodies or metabolites based on available technologies and model systems. Each one of these measurements provides a snapshot of cell function, but a dynamic understanding of disease processes really requires the integration of all these modalities to the extent possible. Particularly in the context of developing non-invasive assays, it is important to leverage gene xpression data with protein abundance data. Doing so not only provides more confidence in the biomarker, but can also provide insight into disease mechanisms.
5:30-6:00 Integration of Omics Data Reveals Multi-Level Regulation of Biological Systems
Jun Zhu, Ph.D., Associate Scientific Director, Genetics, Rosetta Inpharmatics, a wholly owned subsidiary of Merck and Co., Inc.
There are many large Omics data sets, such as genomics, transcriptomics, proteomics, metabolomics and more. Each type of data covers different aspect of biological systems. Integration of different Omics data can generate a more comprehensive view and help to better understand biological systems.
6:00 Close of Day