2012 Cohort Consortium Symposium - When Bigger is Better: Combining and Designing Cohorts for Power

October 25, 2012 - Natcher Conference Center, Bethesda, Maryland


Overview

The Cohort Consortium is an extramural-intramural partnership formed by the National Cancer Institute (NCI) to address the need for large-scale collaborations to pool the large quantity of data and biospecimens necessary to conduct a wide range of cancer studies. The Consortium includes investigators responsible for more than 40 high-quality cohorts involving more than 4 million people. Through its collaborative network of investigators, the Consortium provides a coordinated, interdisciplinary approach to tackling important scientific questions, economies of scale, and opportunities to quicken the pace of research. The Symposium will focus on accomplishments of the Cohort Consortium and presentations of experiences across different NIH Institutes in combining and designing cohorts for power.

Return to Top

Purpose

The Cohort Consortium Symposium was jointly sponsored by the Epidemiology and Genomics Research Program (EGRP) of the NCI's Division of Cancer Control and Population Sciences (DCCPS), and the NCI's Division of Cancer Epidemiology and Genetics (DCEG). The objectives of the Symposium were to:

  • Identify opportunities and obstacles in combining multiple cross-disciplinary (and cross-institute) cohorts for power
  • Describe advantages and constraints of launching new "mega" cohorts
  • Identify issues in building and managing a portfolio of new cohorts across institutes

Return to Top

Agenda

View NCI Cohort Consortium Symposium Agenda
Time Topic
7:30 a.m. - 8:30 a.m. Registration
Session I: Opening Session
8:30 a.m. - 8:40 a.m. Welcome Remarks (DCCPS)

Muin Khoury, M.D., Ph.D.
Associate Director (Acting), Epidemiology and Genomics Research Program
Division of Cancer Control and Population Sciences
National Cancer Institute

8:40 a.m. - 8:50 a.m. Overview and goals of the Symposium

James R. Cerhan, M.D., Ph.D.
Chair, Cohort Consortium Secretariat
Professor and Chair, Division of Epidemiology
Mayo Clinic College of Medicine

8:50 a.m. - 9:05 a.m. Recent accomplishments of Cohort Consortium

Patricia Hartge, Sc.D.
Deputy Director, Epidemiology and Biostatistics Program
Division of Cancer Epidemiology and Genetics
National Cancer Institute

Session II: Spectrum of Experiences in Combining and Designing Cohorts for Power
Moderator: Deborah M. Winn, Ph.D.
Deputy Director, Division of Cancer Control and Population Sciences National Cancer Institute
9:05 a.m. - 9:15 a.m. National Cancer Institute (NCI) - New Initiatives
(Funding of Atherosclerosis Risk in Communities Study for cancer, survivorship, and sequencing)

Deborah M. Winn, Ph.D.
Deputy Director, Division of Cancer Control and Population Sciences National Cancer Institute

9:15 a.m. - 9:45 a.m. National Human Genome Research Institute Perspective and Examples

Lucia Hindorff, Ph.D., M.P.H.
Epidemiologist, Office of Population Genomics
National Human Genome Research Institute

Adam Felsenfeld, Ph.D.
Program Director, Division of Extramural Research
National Human Genome Research Institute

9:45 a.m. - 10:15 a.m. National Heart, Lung, and Blood Institute (NHLBI) - Perspective and examples

Diane Bild, M.D., M.P.H.
Associate Director, Prevention and Population Sciences Program
Division of Cardiovascular Sciences
National Heart, Lung, and Blood Institute

10:15 a.m. - 10:45 a.m. National Institute on Aging (NIA) - Perspective and examples

Richard M. Suzman, Ph.D.
Director, Behavioral and Social Research Program
National Institute on Aging

10:45 a.m. - 11:00 a.m. Break
Session III: Reflection and Discussion
Moderator: James R. Cerhan, M.D., Ph.D.
Chair, Cohort Consortium Secretariat
Professor and Chair, Division of Epidemiology
Mayo Clinic College of Medicine
11:00 a.m. - 12:15 p.m. Robert T. Croyle, Ph.D.
Director, Division of Cancer Control and Population Sciences
National Cancer Institute

Robert N. Hoover, M.D., Sc.D.
Director, Epidemiology and Biostatistics Program
Division of Cancer Epidemiology and Genetics
National Cancer Institute

Julie E. Buring, Sc.D.
Professor of Medicine, Brigham and Women's Hospital / Harvard Medical School
Professor of Epidemiology, Harvard School of Public Health

William J. Blot, Ph.D.
Professor of Medicine, Vanderbilt University
Chief Executive Officer, International Epidemiology Institute, Ltd.

Stephen Chanock, M.D.
Chief, Laboratory of Translational Genomics
Director, Core Genotyping Facility
Division of Cancer Epidemiology and Genetics
National Cancer Institute

12:15 p.m. End of Symposium

Return to Top

Videocast

Those who were unable to participate in person may view an NIH videocastExternal Web Site Policy of the Symposium.

Return to Top

Workshop Summary

  • Session I: Opening Session
    View Session I Details

    Welcome Remarks - Muin Khoury, M.D., Ph.D.

    The field of epidemiology is entering a new era that will emphasize translational research. It will be necessary to move forward with research that translates scientific discoveries into major public health impact and medical interventions. In fact, the Director of the National Cancer Institute (NCI) stated that he expects the dramatic revolution in epidemiology to be the ability to define cancers by their genetic subsets. It is important to remember that:

    1. Discovery research is not enough;
    2. Outcomes/applied research is imperative;
    3. This change in research direction will be driven by collaborations, technology of many kinds, knowledge integration, and multi-level analyses; and
    4. There should be an emphasis on translational research.

    Overview and Goals of the Symposium - James R. Cerhan, M.D., Ph.D.

    The goals of the Symposium are to identify the opportunities and obstacles in combining cross-institute cohorts for greater power, describe the advantages and constraints of launching new "mega" cohorts, and identify the issues in building and managing a portfolio of new cohorts across Institutes.

    Recent Accomplishments of the Cohort Consortium - Patricia Hartge, Sc.D.

    The NCI Cohort Consortium has seen great success and many recent accomplishments. Its projects are described online at http://epi.grants.cancer.gov/Consortia/cohort_projects.html, and descriptions of the member cohorts are available at http://epi.grants.cancer.gov/Consortia/members/. Many genomic, biological marker, and behavioral and environmental studies are contained within the Consortium - genomics studies being the "string" that ties it together. New directions for Consortium research areas include second cancers, survival, and non-cancer/cancer-risk studies.

  • Session II: Spectrum of Experiences in Combining and Designing Cohorts for Power
    View Session II Details

    National Cancer Institute New Initiatives - Deborah M. Winn, Ph.D.

    For NCI, there is great importance in not only maintaining and streamlining databases, but also in creating linkages between them. The National Virtual Cancer Registry Resource is an example of a data linkage from cohorts of all the U.S. cancer registries. The Resource is not centralizing the storage of data, but rather centrally linking all cohort registries. Other databases that require streamlining and improvement are the Death Master File and the National Death Index.

    National Human Genome Research Institute (NHGRI) Perspective and Examples - Lucia Hindorff, Ph.D., M.P.H., and Adam Felsenfeld, Ph.D.

    NHGRI funds a large portion of collaborative science. Gene-Environment Association Studies (GENEVA), for example, is a collaboration of 16 studies funded by NHGRI, and creates statistical power because of its large size. Additional data power can be obtained from utilizing cancer research cohorts and studies to contribute to non-cancer studies, such as the Population Architecture using Genomics and Epidemiology (PAGE) studies.

    Many challenges exist with research collaborations (e.g., heterogeneity of strengths/weaknesses, infrastructure lacking or too much, lack of shared vision); therefore, it is imperative to create standardization and central coordination between cohorts.

    Data that already have been obtained are vast and can lead to significant research advancements if appropriately aggregated and harmonized. The streamlining process does not need to be perfect; creating solutions to even a small majority of problems should be considered a success. Harmonization of phenotyping data should be encouraged, both retrospectively and prospectively.

    National Heart, Lung, and Blood Institute (NHLBI) Perspective and Examples - Diane Bild, M.D., M.P.H.

    NHLBI has been successful in combining cohorts (synthetic cohorts) and designing cohorts (e.g., National Longitudinal Mortality Study). Large sample sizes in cardiovascular research have allowed for subgroup studies that have found modest trends and the integration of epidemiology with other complex, interacting disease factors (e.g., environment, disease comorbidities). Overlapping interests between NCI and NHLBI can assist in such collaborations.

    National Institute on Aging (NIA) Perspective and Examples - Richard M. Suzman, Ph.D.

    Although NIA is one-sixth the size of NCI, the cost of age-related diseases matches the cost of cancer diseases to NCI. Synthetic cohorts would be a great boon to NIA by offering large sample sizes and statistical power.

    NIA funds many national cohorts. With these cohorts, administrative linkages are helpful (e.g., with Medicare, Social Security, National Death Index). Harmonizing global cohorts is imperative because there is much to be learned about how trends in the United States compare to other countries. Harmonization, both ex-ante and ex-post, is key for making comparative research feasible and obtaining appropriate replication of genetic associations.

  • Session III: Reflection and Discussion
    View Session III Details

    Panelists: William J. Blot, Ph.D., Julie E. Buring, Sc.D., Stephen Chanock, M.D., Robert T. Croyle, Ph.D., and Robert N. Hoover, M.D., Sc.D.

    Panelists and Symposium attendees discussed issues uncovered by the presentations. Discussion topics and points are highlighted below.

    • Linking Cancer and Non-Cancer Studies/Cohorts
      • Expanding the utility of NCI cohorts to examine other phenotypes would be optimal. Cancer studies have an advantage over other disease studies due to their large size. "Thick" studies tend to have extensive participant evaluation and include measures, outcomes, and factors from other diseases (or not limited to one specific disease). "Thin" studies tend to focus on one disease and do not measure an extensive number of ancillary or alternative outcomes/factors. Cancer studies can tend to be thin.
      • More than one-half of cohort studies were designed with cancer as the primary endpoint; having interest in and examining other outcomes and exposures has created an efficient and value-added infrastructure situation. Because new projects are difficult to launch, the aim should be to use the extensive resources that the community has already. The creation of synthetic cohorts allows for prodigious data usage and collaboration between cancer and non-cancer outcome researchers.
      • One roadblock to cohort collaboration (that leads to cohort underutilization) is that some researchers do not have the infrastructure to collaborate. Cost and administrative issues also can make collaborations difficult (some researchers are precluded from examining anything other than their specific disease focus). The cohorts that can collaborate, should collaborate.
      • NIH can strip away silos by achieving blended funding and co-funded initiatives. This is the strongest available source of help in the endeavor for the long term.
    • Linking Cohort Studies With Other Large Sources of Data
      • The ability for NCI cohort studies to link with centers for other services (e.g., Medicare and Medicaid) still is in its infancy but allows researchers to better exploit co-morbidities and data analyses. The cost of obtaining data and linking studies with the Centers for Medicare and Medicaid Services (CMS) can be reduced by using NIH as a liaison.
      • Using CMS data is productive; however, there is a need to expand to data sources beyond Medicare/Medicaid. NIH has a strong skill set, and this should be used to link to other databases. Aggregation of other data types (e.g., molecular and environmental) is critical, and the cancer community should lead in this direction. To encourage further collaborations, easily accessible and cost-effective methods of getting information should be devised.
    • Improving Cohort Studies
      • The domains that cohorts cover (e.g., disease, people, research) are not distributed evenly. Certain regions and people have been captured, whereas other groups and regions have been missed. Legacy cohort studies that examine individuals for extended periods of time, followed by examination of their offspring, are optimal. This has been a struggle due to the long-term financial commitment.
    • Government Policy and Cohort Studies
      • There is an assumption that the United States has an antiquated health care system; however, since the inception of the Affordable Care Act, there has been a rapid improvement in U.S. health care. The collaborations among doctors and practices (prompted by the economics of payment reform) have consolidated people and networks. Epidemiologists can "piggy back" onto this new system infrastructure and gain a significant research advantage.
    • The Public's New Ability
      • For the first time, commercial companies (e.g., 23andMe, Inc.) are putting genetic information in the hands of the public. Tools and kits will be available soon to the general public to construct their own genomes and metabolic profiles. They will not be necessarily getting genetic advice from classically trained academic institutions. This sector must be pursued, and a dialogue must begin. This should not be seen as a crisis for the academic community, but rather as an opportunity.
    • The Future of Cohort Studies
      • The technologies available are changing rapidly, creating many new ways to identify various outcomes. The work in genetics, metabolomics, and so forth make this an informative time in research. The capacity of these new technologies currently is stabilizing, which allows for a period of time to observe how technologies can be applied to research questions and how integrated analyses can be reached.
      • The Cancer Genome Atlas (TCGA) and other large-scale cancer genomics projects and companies are shifting to germ-line epidemiology. Following the cancer genomics work that is slated to end in 2014, NCI will be switching to a "clinical mode." The idea is that prevention is just as important as outcomes and treatment. The cohort research community should focus on what can be learned biologically, both directly and indirectly, for long-term preventive capabilities. The next set of studies that will follow up on TCGA must be conducted.
      • There is a relationship between innovation and "big science." There are efforts currently to combine somatic mutations and metabolomics research; however, the consortia should assist in accelerating these technologies. For example, cellular telephones can be used to passively collect information (e.g., sleep patterns, light levels, exposures), or photographs could be taken of cohort participants' meals, and these ideas should be adopted by researchers.

      Regarding the adaptation of mobile technology for research purposes, it must be decided exactly who/what will create these adaptive tools. Additionally, there must be a critical evaluation and validation of these technologies being used for research purposes. (There often are unintended artifacts.)

    There has been a historical lack of appreciation for population studies, especially when most of the people who make funding decisions do not have an adequate grasp on what these studies do and their importance. This is not likely to change in the near future. Advances in science and technology provide opportunities, and NCI has been helpful in obtaining these advantages (e.g., sequencing); however, it would be cheaper if cohorts could consort with each other.

Return to Top

Return to Top

Planning Committee

  • James Cerhan, M.D., Ph.D., Mayo Clinic
  • William J. Blot, Ph.D., Vanderbilt University and International Epidemiology Institute, Ltd.
  • Julie Buring, Sc.D., Brigham and Women's Hospital and Harvard University
  • Robert Hoover, M.D., Sc.D., Epidemiology and Biostatistics Program (EBP), Division of Cancer Epidemiology and Genetics (DCEG), National Cancer Institute, NCI
  • Elio Riboli, M.D., M.P.H., M.Sc., Imperial College
  • Deborah Winn, Ph.D., Division of Cancer Control and Population Sciences (DCCPS), NCI
  • Anne Zeleniuch-Jacquotte, M.D., New York University
  • Muin Khoury, M.D., Ph.D., Epidemiology and Genomics Research Program (EGRP), DCCPS, NCI
  • Patricia Hartge, Sc.D., EBP, DCEG, NCI
  • Stephen Chanock, M.D., Laboratory of Translational Genomics and Core Genotyping Facility, DCEG, NCI
  • Daniela Seminara, Ph.D., M.P.H., EGRP, DCCPS, NCI
  • Nonye Harvey, M.P.H., EGRP, DCCPS, NCI
  • Scott Rogers, M.P.H., EGRP, DCCPS, NCI

Return to Top

Contact

For further information or questions about the Symposium, contact Nonye Harvey, M.P.H. or visit the Cohort Consortium website.

Return to Top