I. What we sought to accomplish
II. How the Think Tanks were organized and run
III. Overall recommendations
IV. Concluding remarks
Appendix A

I. What we sought to accomplish

A year ago (in 2003), Dr. von Eschenbach charged the Division of Cancer Biology (DCB) with conducting a series of Think Tanks to assess the state of cancer biology research and to recommend to the NCI a research agenda that would accelerate progress in cancer research. The process of fulfilling this charge began with an internal identification of scientific areas of unusual promise for rapid progress. Eight areas were chosen as the topics of Think Tanks. The areas were quite different in scope and in the maturity of the scientific disciplines involved. As a result, each Think Tank had to be structured differently, to deal with the unique questions and opportunities in each area. The eight Think Tanks are listed below, each with a brief description of its goal:

1) Tumor Immunology - This area of basic cancer biology has gone from relative obscurity to intellectual vitality over the past several years, as a result of critical basic science and translational advances, coupled with new concepts and methods in basic immunology. Immunotherapy of cancer had disputed translational potential for many years, but is now an established and expanding therapeutic modality. The field has untapped promise, but basic science advances have made clear how much still needs to be known about how the immune system and tumors affect one another. In addition, clinical advances depend on improving the rate at which immunological and other biological therapies are brought to the point where they can be used in clinical studies, and on devising clinical trials structures that accurately and efficiently allow evaluation of efficacy. These basic and translational issues were the focus of discussion.

2) Tumor Microenvironment - The microenvironment in which a tumor arises has a profound influence on its progression. Tumor cells and stromal elements such as endothelial cells, fibroblasts, lymphoid and myeloid cells, extracellular matrix and cytokines interact dynamically and depend on one another for growth and survival. The tumor microenvironment has been the focus of intense interest within NCI and the scientific community for some time because it is the key to understanding tumor biology and strongly affects the likelihood of successful cancer therapy. While a consensus has been reached that this area deserves substantial additional support, this Think Tank had the goal of specifying the type of initiative(s) that would best facilitate rapid advances in knowledge about the microenvironment.

3) Tumor Stem Cell & Self-Renewal Genes - Recent results suggesting that at least some tumors contain only a small number of cells with tumor-initiating potential (loosely termed tumor stem cells), have profound implications for carcinogenesis, cancer biology and cancer therapy. Think Tank participants discussed current evidence for the existence of tumor stem cells in different tumor types, how they might arise, and the genetic (and possibly epigenetic) pathways important in maintaining the tumor stem cell state.

4) Cell Decisions in Response to DNA Damage: Survival vs. Programmed Cell Death - When cells encounter DNA damage, they can repair the damage or decide to try to live with it, either of which leads to survival, or they can choose to undergo programmed cell death. This life or death decision involves a complex interplay between DNA damage sensing and repair pathways, the pathways controlling the cell cycle, and cell death mechanisms. With an understanding of how these decisions are made, the response to radiation and/or chemotherapy could be improved by increasing the likelihood that cancer cells would die in response to therapy-induced damage while decreasing the chance that normal cells would die. Experts in DNA damage, cell cycle control, and apoptosis, who do not meet regularly, met at this Think Tank to consider prospects for fruitful interaction.

5) Cancer Etiology: Role of Exogenous and Endogenous Chemicals - Classical studies in chemical carcinogenesis have definitively established an important role for exogenous and endogenous chemicals in cancer development and defined some of the mechanisms involved, particularly those related to DNA adduct formation. Other areas, such as the role of damage to proteins and lipids, and identifying markers of exposure, have received less emphasis. The challenge in the field is to build on what has been done by integrating with other fields of biology (particularly integrative cancer biology and cancer susceptibility) to build a system-wide understanding of the complex process of carcinogenesis. The participants discussed the connections that exist between chemical carcinogenesis and areas such as inflammation, biological carcinogenesis, reactive oxygen species and systems biology, and how to strengthen them.

6) Epigenetic Mechanisms in Cancer - Epigenetic mechanisms are those that lead to heritable changes in cell behavior without changes in DNA sequence. The dramatic changes in gene expression that characterize tumor cells come about through both genetic and epigenetic mechanisms. Dramatic advances in genetics and genomics, and smaller advances in epigenetics, have sharpened our appreciation of the critical role that epigenetics plays in development and cancer, through DNA methylation, histone modification and changes in higher order chromatin structure. The Think Tank was organized to discuss evidence for epigenetic phenomena important in cancer and to prioritize NCI activities to advance the field.

7) Inflammation and Cancer - Chronic inflammatory disease has been known for many years to predispose to cancer development, and most tumor sites show evidence of ongoing inflammation. Inflammation can benefit a tumor by leading to production of mutagens, such as reactive oxygen and nitrogen species, and growth-stimulatory cytokines. At the same time, the proper inflammatory signals are necessary to allow the development of an effective immune response against the tumor. The participants in the Think Tank discussed how to sort out the positive and negative influences of inflammation in cancer, including that due to infectious agents, and how this understanding is expected to contribute to more effective strategies for cancer prevention and treatment.

8) Cancer Susceptibility and Resistance - The consensus of two prior meetings of mouse geneticists, population scientists, and statisticians is that the ability to understand human cancers as complex traits will require new statistical and computational methods for modeling gene/environment interactions, and assembling data for high-level gene network analysis. This Think Tank was set up to explore novel developments in mathematics and engineering design and their application to cancer susceptibility research, both in modeling human populations and hypothesis testing/candidate gene validation in rodent models.

9) Integrative Cancer Biology - The first Think Tank area to be addressed with a major initiative was systems biology or, in this specific context, integrative cancer biology. Cancer is sufficiently complex that it will never be adequately understood based on the expression of a few genes or the regulation of a few signal transduction pathways in the transformed cell. It is an emergent property of derangements in complex intracellular systems in a context set by equally complex interactions of the tumor cell with intercellular environment. No report is included here for the Integrative Cancer Biology Think Tank because a solicitation has already been done for Integrative Cancer Biology Programs, and these have now been funded. This was the key recommendation of the Think Tank. Recommendations from several of the other Think Tanks dealt with the need to foster this area (see below for discussion), so this is likely to remain a topic of great interest for NCI for the foreseeable future.

Back to Top

II. How the Think Tanks were organized and run

The general goal of each Think Tank was to assess a specific area of cancer biology, determining where the science is and where it is going, and asking what NCI could or should do to facilitate progress. In areas that had been assessed previously in workshops and where NCI had made a commitment to do initiatives, such as the tumor microenvironment, the emphasis was entirely on what NCI should do. In other areas, where less had been done and no funding commitment existed, the emphasis was more on assessing the state of science in the field. A general charge was developed for each Think Tank, as follows:

Questions to Consider:

• Where does the field stand today?


• What would we have to know to move the field forward dramatically?


• What are the gaps and/or roadblocks to progress?


• Are there areas of expertise that need to be, but have not been, brought to bear on the problems in this field? If so, how can this be remedied?


• What cross-cutting tool, enabling technology and/or infrastructure needs to be developed?


• What can NCI do to move the field forward?

Each Think Tank was unique in scope and maturity of the field, so each had subtly different goals and had to be organized somewhat differently. Accordingly, the organizers of each Think Tank were given guidelines or "points to consider," rather than firm rules. These organizational guidelines are included as Appendix A. A critical element of every Think Tank was the selection of outside co-chairs. These individuals are distinguished scientists with unusual breadth of vision in the area to be discussed and strong organizational ability. They were recruited very early in the planning process, and played a leading role in identifying other participants, defining the critical questions to be discussed and setting the agenda. They were also involved in assembling the reports included here, and generally took the lead if an account of the Think Tank was prepared for publication in an appropriate scientific journal. In every case, the recommendations included in the Think Tank reports are those of the co-chairs and the other participants.

Each Think Tank was organized independently, but it was obvious from the outset that some of the Think Tanks overlapped. For example, tumor immune responses are influenced by the tumor microenvironment and cancer etiology involves a strong element of inflammation, at least as a cofactor. This was seen as an asset because it made it possible to build in a level of continuity in these areas by inviting key participants to attend two or more Think Tanks. In some cases, formal participation in two Think Tanks was not possible, but advice was always sought where overlaps were anticipated to ensure that similar issues were addressed.

Back to Top

III. Overall recommendations

Each Think Tank was unique in focus and participants, and each resulted in a series of recommendations that will be addressed individually. All of the reports are appended, and for easy access the recommendations have been collected at the end of each report. This section will deal with a number of scientific and support-mechanism issues that emerged independently in multiple Think Tanks, with details on each given below. The scientific issues of the tumor microenvironment and integrative cancer biology (systems biology) were mentioned in virtually every Think Tank. Both of these subjects have been identified as high-priority areas for NCI support. The Think Tank reports provide a wealth of details on what needs to be done in these areas and reinforces their importance for cancer biology in the future. It was emphasized in these areas and others that studying cancer is impossible without also studying normal biology. The derangements in cancer can only be understood in terms of their deviations from normal behavior. This increases the scope of what must be done, and challenges us to coordinate with agencies supporting studies of normal biological development and other diseases.

Among the issues dealing with support mechanisms, it was not a surprise that the participants uniformly identified a need for additional and more flexible mechanisms to support multi-disciplinary, generally multi-institutional efforts to address complex issues like integrative cancer biology and the tumor microenvironment. A related concern was the difficulty of establishing training programs that prepare students and investigators for research in a multi-disciplinary environment. To balance this concern, each Think Tank also emphasized the indispensable contribution by, and the continuing importance of, research supported by investigator-initiated R01 grants.

Some of the issues were unexpected. Inflammation was recognized as important enough to deserve its own Think Tank, but it was surprising how prominent a role it occupied in other Think Tanks. This suggests that it needs an even more prominent role than had been envisioned in initiatives to study the tumor microenvironment, where inflammation is a nearly constant finding. A related common theme was the role of microbial flora as a cofactor in tumor development. While biological carcinogenesis, with an emphasis on cancer caused by viruses, has always been supported within NCI, examination of a cofactor role for microbes that are not directly transforming has lagged behind. This is one of several areas that bridge cancer biology and etiology. Support for technology development, in general, appears to remain a challenge despite the addition of many new programs in recent years. Think Tank participants consistently reported limitations in funds for reagent preparation (e.g., monoclonal antibodies), model development (both genetically engineered animals and complex, three-dimensional tissue cultures), and state-of-the-art imaging. Funding for critical resources needs to be factored into plans in many areas, but it was also surprising that in some instances Think Tank participants recommended that NCI make available resources, including reagents, databases, animals and facilities, that already exist. This indicates that more effort needs to be put into ensuring that all members of the cancer research community are aware of the resources NCI currently provides. If there are problems of quality or access with existing resources, these need to be evaluated as well.

The Tumor Microenvironment

Evidence continues to accumulate that growth and migration of normal epithelial cells are subject to many levels of regulation by neighboring cells, extracellular matrix, and local levels of soluble signaling molecules. Cancer cells lose critical aspects of these controls, but they lose them gradually and rarely lose them all. Thus, one way of looking at cancer initiation and progression is as an iterative and progressive renegotiation of constraints carried out between a developing clone of epithelial cancer cells and its stromal microenvironment. This perspective suggests two principal lessons. First, attempts to understand tumor behavior or to treat cancers must take into account far more than the intrinsic properties of the malignant cells to be successful. And second, attempts to model tumor behavior must go beyond using tumor cell lines cultured on plastic surfaces, to three-dimensional culture systems and in vivo studies. The Tumor Microenvironment Think Tank provided a detailed blueprint for integrated studies, and several of the other Think Tanks emphasized specific aspects of the microenvironment that are often overlooked in overviews of the subject.

1) Characterize all of the cellular and non-cellular components of the normal, wounded, and tumor tissue microenvironments. The microenvironment that sustains and shapes tumor stem cells would also be of interest. Characterization would involve probing the genomics and proteomics of stromal and tumor cells, and developing antibodies and other reagents useful for visualizing, quantitating and comparing different microenvironments, and storing the data in public databases. Static characterization would rapidly be extended to studies of dynamic interactions, using real-time imaging methods.

2) Facilitate communication and data exchange through caBIG.

3) Create and make available to the research community three-dimensional tissue culture and animal models in which microenvironmental influences can be studied.

4) Delineate the role of the microenvironment in tumor progression and metastasis, and in response to radiation and/or chemotherapy, including characterization of the metastatic tumor microenvironment.

5) Determine the role of inflammation in shaping the tumor microenvironment in the earliest phases of tumor initiation and during progression. Delineate how inflammatory processes facilitate or inhibit the development of an effective antitumor immune response. It will be critical to determine the role of the microbial flora in establishing the nature and degree of inflammation.

6) Understand the effects of androgens and/or estrogens on inflammation and other aspects of the microenvironment.

7) Explore the role of host genetics in influencing stromal elements, and determine whether mutations or epigenetic changes in the stroma influence tumor growth and capacity to metastasize.

8) Encourage the use of state-of-the-art imaging technologies in microenvironment studies and the development of techniques to bypass current limits on imaging. This is critical because only imaging has the capacity to capture the dynamic and contact-mediated aspects of this complex system.

9) Provide an environment conducive to interdisciplinary training programs.

10) Translate the basic knowledge obtained to improve diagnosis and early detection of cancer, and to discover and validate therapeutic targets derived from the tumor microenvironment.

Think Tanks other than the one focused specifically on the tumor microenvironment have influenced our sense of priorities in this area. Tumor Immunology emphasized the critical roles of the tumor and lymph node microenvironments in determining the effectiveness of antitumor immune responses. The Tumor Stem Cell Think Tank noted that the long-term proliferation potential of a tumor can controlled by a small number of tumor stem cells. Hence, the niche that supports those stem cells must be characterized and stem cell markers developed so that the most important subset of interactions can be studied. The Inflammation Think Tank highlighted evidence that inflammation is strongly associated with tumor initiation and progression, suggesting that the common inflammatory aspects of the microenvironment facilitate tumor development. At the same time, there is the paradox that inflammation, at least in infectious diseases, is associated with brisk immune responses, but tumors generally exhibit evidence of ongoing inflammation while suppressing immune responses. It is clearly insufficient to say that inflammation is present or absent; the phenomenon must be dissected to its molecular roots if interventions are to be developed. In Cancer Etiology, the macroenvironmental influences of both chemicals and microbes were discussed. Chemical exposures have impact far beyond mutagenesis, extending into a range of effects on the microenvironment. Striking evidence also exists in some tumor types that the normal microbial flora strongly influences the microenvironment. Viruses and bacteria must be assessed as important components of, or cofactors for a permissive tumor microenvironment. Finally, it was emphasized in Cancer Susceptibility that some of the complexity that has made it difficult to identify the genes responsible for individual variation in susceptibility comes from the context dependence of gene expression. Individuals may have a series of genetic polymorphisms that would predispose to cancer development, but it may not make a difference unless the microenvironment in which the tumor must develop supports the expression of the susceptibility gene(s).

Integrative Cancer Biology

As Fiscal Year 2004 is ending, NCI is funding a series of Integrative Cancer Biology Programs, the first organized foray into systems biology in the context of cancer. The Think Tanks provided strong evidence that this is an important direction for the Institute to pursue and that its influence will be felt throughout cancer research. Integrative Cancer Biology is complementary to the reductionist studies that comprise the majority of the grant portfolio. It attempts to address the complexity of many interactive and interdependent biological processes, starting by making high-throughput measurements of critical parameters. Biological lessons will be extracted from the masses of data through the use of advanced bioinformatics tools and the construction of predictive computational models of the cancer process. Although integrative biology is most often identified with the analysis of signal transduction pathways and other regulatory circuits within a single cell, it is equally applicable to complex processes involving multiple cells and extracellular molecules. Thus, the tumor microenvironment can be fully characterized only through the use of high-throughput analytical methods, and participants in the Think Tank acknowledged that a predictive model of the interactions that drive the microenvironment can be obtained only through the methods of integrative biology. The same is true of the related fields of tumor immunology and inflammation.

Other areas explored in the Think Tank process are similarly dependent on the emerging field of integrative cancer biology. Epigenetic influences on gene expression are mediated through DNA methylation, covalent modifications of histones, and higher order chromatin structure effects. Progress in the field is absolutely dependent on being able to measure these molecular changes on a genome-wide scale and knowing how to extract meaningful patterns from this mass of data. The DNA damage response and the cell cycle and cell death machinery are complex, interacting systems each made up of many quasi-stable molecular complexes. The dynamics and interactions of these systems must be understood in detail if they are to be manipulated for patient benefit in cancer. Determining how tumor stem cells function and how they come about involves achieving a molecular understanding of the cell regulatory pathways that underlie "stemness," or the ability of some cells to maintain unlimited replication potential and to divide asymmetrically into another stem cell plus a daughter cell committed to differentiation. The stem-cell genetic program is in turn heavily influenced by the cell and molecular milieu of the stem-cell niche. Etiology and susceptibility are two different views of the earliest stages of tumor development. They have been recognized for some time to involve numerous genetic and environmental influences impinging on complex cellular homeostatic mechanisms. These disciplines all need high-throughput data, bioinformatics support, and computational modeling to address critical issues.

The Challenge of Comparing the Normal and the Tumor State

The National Cancer Institute has finite resources and cancer biology presents an enormous array of promising areas of investigation directly relevant to the NCI mission. To maximize the impact of our efforts, it is tempting to focus exclusively on the cancer state, leaving studies of the normal state to other funding agencies. In six of the Think Tanks, participants explicitly recommended against this course of action, pointing out a need to understand cancer in the context of normal biology. In the Tumor Microenvironment, normal constraints on cell growth and mobility are gradually loosened as the tumor develops. We need to know much more about the normal constraints individually, and about how they are coordinated at a systems level, before the tumor microenvironment can be fully characterized. The Cell Decisions in Response to DNA Damage are similarly complex and also must be better described in the normal case before they can be manipulated for therapeutic benefit in cancer. In Tumor Immunology, the major advances in understanding that have occurred in the last ten years have come from conceptual advances in immunology as a whole. The critical questions that remain are the same for basic immunology, autoimmunity, chronic infectious diseases and cancer, although the perspectives on the questions differ slightly among these fields. Inflammation in cancer has a marked stimulatory effect on cancer growth not because of its intensity, but because it fails to resolve the way acute, physiological inflammation does. It shares this characteristic with autoimmunity and certain chronic infections, and it is from a comparison of all these states that further understanding of the process, and the ability to modulate it selectively, will come. Cancer appears to use the stem-cell program of several tissues to further its own causes, but so little is known about the regulatory program within the normal tissue stem cell and the cell-cell interactions of the stem-cell niche that it is difficult to characterize cancer stem cells or to determine the path by which they became transformed. Epigenetics is similarly a young field, in which a great deal of basic knowledge must be accumulated before its role in cancer can be clarified.

The challenge is to identify those elements of these fields that the NCI should attack with its own resources and those where it should work in coordination with other NIH Institutes and other funding agencies. Leveraging of resources is difficult, but necessary. Some of the NIH Roadmap areas are relevant to scientific issues listed above, but a great deal more remains undone. There is no large-scale project on epigenetics on the horizon, despite its documented importance in many human diseases. Similarly, while the NIH has some coordinated activities related to human embryonic stem cells, tissue-specific stem cells (with the exception perhaps of hematopoietic stem cells) have received scant attention. The Think Tank recommendations make it clear that catalyzing larger-scale studies of some critical cross-cutting biological issues must be a high priority for NCI to provide the necessary context for progress against cancer.

Mechanisms to Foster Collaborative, Interdisciplinary Research

NIH grants, built around the R01 traditional research grant, have been the engine of creativity that has brought us to the current exciting point in cancer research. The Think Tank participants uniformly acknowledged the past and continuing importance of these individual grants. During the Think Tanks, however, they focused on needs that are difficult or impossible to meet through this mechanism. These were generally large-scale efforts, especially those that required input from scientists in diverse disciplines. The Tumor Microenvironment Network, described above, is an example of the recommendations, but similar networks were suggested in immunotherapy, stem-cell research, epigenetics, etiology and susceptibility. In other cases, less formal (and smaller scale) resources for collaboration were recommended. The Inflammation Think Tank recommendations included one to "sponsor interactive fora that interface experts drawn from different disciplines to address the multi-faceted topic of inflammation and cancer at a deeper level." The Cancer Etiology Think Tank made several recommendations for collaborative undertakings, including one for instrument development and use. The Epigenetics Think Tank recommended formation of a working group to discuss and begin outlining a Human Epigenome Project. In this case, the smaller initiative could lead to the development of a large-scale effort.

For very large projects, such as a Tumor Microenvironment Network, the size and expense of the proposal makes it appropriate for funding through the RFA process. The NCI has available a wide variety of funding mechanisms and governance models for large projects of this type. It is noteworthy that in recent years RFAs at NCI and elsewhere at NIH have increasingly emphasized collaboration and interdisciplinary teamwork, part of a well publicized trend toward "team science." The large projects included in the Think Tank recommendations, including the Tumor Microenvironment Network, exemplify that trend. These recommendations came from the outside Think Tank participants, rather than NCI staff, indicating that the trends in RFAs reflect the current thinking of the research community.

Many of the recommendations were modest in cost and involved more coordination than direct research support. These recommendations were made because there are very few investigator-initiated NIH funding mechanisms that can support any of these varied activities. Critical problems in cancer research and other areas of biomedicine increasingly require a variety of expertise and/or the sharing of data or reagents in a manner that is not facilitated or sometimes even possible when support comes exclusively from grants to individual principal investigators. Constraints on collaborative and interdisciplinary research also exist at research institutions. Rigid departmental structure, intellectual property policies and concerns about indirect costs can make some types of research more difficult.

With sufficient resources, NCI could address all of these recommendations through available mechanisms such as contracts, supplements and workshops. DCB fully intends to address as many of the high-priority recommendations as possible through such efforts. What this will not do, however, is alter the fact that few such efforts can be initiated directly by the research community through investigator-initiated mechanisms. Some collaborative research can be carried out through Program Projects, but these offer a limited range of flexibility. They typically have a small number of projects, all of which normally extend for the entire project period. There is also a perception that multi-institutional Program Projects are difficult to get funded. SPORE-like grants were recommended in a couple of cases, based on the added flexibility of a changing cast of projects and integral training, but these are possible only in response to an initiative. The DCB Activities to Promote Research Collaborations (APRC) program is relatively flexible, but it is relatively small and short-term in nature. What is needed is a highly flexible, permanent program open to investigator-initiated applications to support modest-scale, collaborative, interdisciplinary research efforts. DCB will work toward the design of such a program.

Collaborative and Interdisciplinary Training

Each recommendation for a collaborative and/or interdisciplinary research program was accompanied by a recommendation for a program that would train students, postdoctoral fellows and established investigators to take optimum advantage of the opportunities such a program would create. There are differences in opinion as to the ideal way to attract and train scientists comfortable with both wet-lab biology and computational modeling, for example, or with both chemistry and biology, but there is universal agreement that any barriers that exist between disciplines need to be removed. Some barriers exist at the level of research institutions, but NCI must strive to create incentives for flexible training programs that will meet the needs for the next generation of research scientists. Three types of suggestions about training were made during the Think Tanks. One was to incorporate training into large-scale interdisciplinary initiatives. This was done in the Integrative Cancer Biology Programs, and is a feature of some other NCI programs. The second was to place some leverage back in the hands of graduate students by inaugurating individual pre-doctoral fellowships in which the range of subdisciplines and the mentor(s) could be determined by the graduate fellow and not the institution. The third was to reserve a portion of NCI postdoctoral training grants for explicitly interdisciplinary programs. While institutions are moving to respond to the need to change training paradigms, the Think Tank process made it clear that NCI must work to facilitate and accelerate such change.

Back to Top

IV. Concluding Remarks

The Think Tank process provided an opportunity for both NCI funded investigators and DCB scientific staff to think broadly and deeply about the directions cancer biology must take in support of the goals of the NCI.

The individual Think Tank reports are attached. Each of the individual recommendations deserves, and will receive, consideration by DCB staff. The resulting initiatives, both large and small, will reflect a unified agenda for cancer biology over the next several years.

The cross-cutting themes, summarized here, all derive from the need and emerging ability to study cancer biology as a complex system. Integrative cancer biology will be the necessary new fundamental discipline driving this transformation. We believe that understanding the tumor microenvironment is an initiative that can focus the new tools and approaches and provide results with dramatic translational impact. To support this endeavor and others, the emerging trend toward team science as a component of cancer biology research needs to be reinforced with a reengineering of NCI funding mechanisms and approaches to encourage more collaborative and interdisciplinary interactions, accompanied by continued support of an investigator-initiated research portfolio.

Back to Top

Appendix A

Guidelines for Organization of Think Tanks

1. Think Tank Goals - Concrete goals should be set early in the planning process, and modified as necessary. The goals should be communicated to the participants before the Think Tank.

2. General Structure of Agenda - Think Tanks need to be distinguished from scientific workshops. Presentations, if any, should be short and not involve primary data. Slide talks should be banned or held to a strict minimum. The bulk of the time should be devoted to brainstorming and structured group discussion.

3. Number and Type of Participants - To optimize discussion, the number of outside participants should generally be limited to fifteen. The participants should cover the critical subdisciplines and points of view. It is desirable to have a mix of established authorities and rising stars. It may also be useful to have a few key individuals who have useful perspectives but are not identified with the field in question. Achieving all of this with such a small number of participants requires judgment, and the mix will be different for each Think Tank.

4. Think Tank Summary - Each Think Tank must produce a summary of discussion and recommendations for internal use. The summary should be done in a timely fashion, and should be done by, or have substantial input from, the outside Chair(s). Depending on the goals of the Think Tank, it may be useful to produce a meeting report or review article for publication in an appropriate journal. This should be decided in advance, in consultation with the outside Chair(s).

5. Schedule/Location - Barring unusual circumstances, Think Tanks should be held locally. The length is expected to be in the range of 1.5-2 days. The schedule of an evening session, followed by a full day, followed by a morning has become common, but should be modified as appropriate. Breakout sessions are fine, as needed.

Back to Top

Next Section >