The Translational Research Working Group Developmental Pathways: Introduction and Overview

Background

Advances in cancer biology offer enormous potential to improve clinical practice and public health. More selective targeting of molecular phenomena implicated in tumorigenesis and metastasis may enable more effective and less toxic interventions. Enhanced assessment modalities promise earlier and more definitive identification and quantification of risk, as well as enhanced means of assessing and monitoring the effectiveness of interventions across the spectrum of neoplastic development and progression. Ultimately, new knowledge from molecular, cellular, and systems biology may lead to more effective preventive interventions targeted at individuals with defined cancer risk.

Rapid progress in cancer-related fundamental research has also highlighted the critical role of translational research within the nation's cancer research portfolio. Translational research may be defined as applied research that transforms discoveries arising from fundamental laboratory, clinical, or population-based research into new drugs, devices, or population interventions that are ready for definitive testing in clinical trials. In addition to pursuing development of new concepts “from benchtop to bedside,” translational research also incorporates feedback from the clinic (“bedside to benchtop”) and seeks to extend innovations from one clinical development path, such as new laboratory techniques, interventions, and assessment methodologies, to new clinical applications.⁸ Although goal-oriented, translational research is neither simple nor routinized. Breakthroughs in fundamental knowledge often pose daunting challenges of implementation, and their transformation into assessment tools and interventions that can be applied in clinical or public health practice requires high levels of conceptual creativity, technical innovation, and managerial skill.

As the distinctive character and role of translational research have become more widely understood in recent years, the National Cancer Institute (NCI) has responded with programs that target various goals in translational research. Initiatives such as the Early Detection Research Network,⁹ the Rapid Access to Intervention Development¹⁰ program, and the Specialized Programs of Research Excellence¹¹ have focused attention and resources on translation, while testing programmatic approaches to addressing the specific needs of translational research.

Unfortunately, the rate at which new drugs, devices, and population interventions reach late-stage testing has not matched the pace of the advance in fundamental knowledge. Even as its critical role is increasingly appreciated, translational research is under great stress, caught between expanding opportunities and high public expectations on the one hand, and limited financial and human resources on the other. The Food and Drug Administration's Critical Path Initiative, launched in 2004, drew broad attention to the “pipeline problem”; stimulated analysis of scientific, technological, and managerial bottlenecks along the path from discovery to product launch; and highlighted the creation of an enhanced scientific “toolkit” for product development as an important objective of public policy (1).

The Translational Research Working Group

Objectives. The Translational Research Working Group (TRWG)¹² was created in 2005 under the auspices of the National Cancer Advisory Board as a national initiative to evaluate the current status of NCI's investment in translational research and envision its future in an inclusive, representative, and transparent manner. The initiative was commissioned to build upon and extend the work of the prior Clinical Trials Working Group,¹³ which focused on recommendations to improve NCI's policies, processes, and procedures in clinical trials, with an emphasis on phase III trials. The goal of the TRWG was to enhance the productivity of the nation's translational research enterprise by identifying and addressing unmet needs, redundancies in existing programs, and bottlenecks in communication and coordination.

Membership. The TRWG was led by Ernest T. Hawk, M.D., M.P.H., who at the time was director of NCI's Office of Centers, Training and Resources; Lynn M. Matrisian, PhD., of the Vanderbilt-Ingram Comprehensive Cancer Center; and William G. Nelson, M.D., PhD, of the Sidney Kimmel Comprehensive Cancer Center at the Johns Hopkins University. Fifty-nine additional members were recruited to provide broad representation across relevant institutional, programmatic, and substantive domains, which included

• Key institutions including NCI, the Food and Drug Administration, academia, industry, and patient advocacy groups.

• Experience as principal investigators and directors in key programs across NCI divisions, including the Cancer Centers program, the Early Detection Research Network, the Specialized Programs of Research Excellence program, clinical study consortia, and investigator-initiated award programs.

• Basic, translational and clinical research and training, across the range of organ systems affected by cancer.

• Special research areas relevant to translation, including prevention, pediatrics, survivorship, genetics, imaging, drugs/immunologics, emerging technologies, behavioral sciences, and preclinical models.

Approach. To inform its deliberations, the TRWG collected information from several different sources. It reviewed key prior reports on translational research; gathered public input through open, Web-based comment forums, as well as a series of special roundtable meetings; and analyzed NCI's current portfolio of investments in translational research in both quantitative and qualitative terms. The TRWG organized itself into six subcommittees that were responsible for drafting specific recommendations on the future of the translational research enterprise. Each recommendation went through several rounds of discussion in plenary session and revision by the cognizant subcommittee until the final set of recommendations was approved by the TRWG as a whole for presentation to the NCI (2).

The Developmental Pathways Concept

Need for the pathways. Early in the deliberations of the TRWG, it became apparent that one of the characteristic attributes of the translational research enterprise is its sheer complexity: the diversity of specialized expertise and the extent of crossdisciplinary collaboration required for successful translation, the variety of different activities needed to advance a concept to its realization in tangible form, and the many and varied relationships among the researchers and research institutions, sponsors, and resources involved. In seeking to develop wise recommendations for advancing translational research, TRWG members first faced the challenge of grasping the essential features of the enterprise.

A further challenge was the development of a shared vocabulary to enable productive deliberation. For example, at first, the TRWG found it difficult to agree upon a definition of translational research. Published literature on the topic as well as definitions used by different NCI programs varied. The President's Cancer Panel¹⁴ had recently considered translational research from a very broad and inclusive perspective, considering the full spectrum of development from basic discovery through clinical and regulatory development, adoption, and dissemination. The TRWG decided to focus on that portion of the translational continuum designated as “early translation” by the President's Cancer Panel—the translational process that follows fundamental discovery and precedes definitive, late-stage trials.

The differing viewpoints on the definition of translational research highlighted the importance of opening the “black box” of translation and clearly identifying its functional elements and their relationships. Without such clarity, discussion of the strengths and weaknesses of translational research would be impressionistic and imprecise, and it would be difficult to formulate convincing and well-motivated recommendations.

Origin and function of the pathway concept. As the TRWG began to examine in detail the functional elements and their relationships that comprise the domain of early translation, it soon became apparent that the most powerful insights emerged from considering translational research as a set of developmental processes—or pathways—focused on various clinical goals. A decision was made to try to capture the working group's collective knowledge about these developmental processes in the form of graphical representations that could provide a shared basis for discussion.

The approach to graphical representation implemented by the TRWG borrows heavily from the flowchart, a schematic process representation that is widely used in engineering. Familiar examples include the flowcharts used in software engineering to represent the algorithm underlying a computer program, and the process flow diagrams commonly used in chemical and process engineering to describe production sequences and infrastructure.

The software engineer's algorithmic flowchart model includes a number of elements that make it especially appropriate as a tool for elucidating the workings of translational research. Most important among these are the explicit representation of start- and end-points, the flow of control, decision points and criteria, as well as the possibility of activity loops that iterate until a condition is met.

It was anticipated that, applied to translational research, the algorithmic flowchart could be used to

• specify activities and decision points along the path by which fundamental scientific discoveries are transformed into clinical modalities;

• clarify dependencies among different steps, as well as events that occur in parallel; and

• show feedback loops and iterative processes that are embedded within the development process.

Development of the pathways. The starting point for development of the pathways was to identify the clinical goals that drive translational research. Drawing upon the Food and Drug Administration's regulatory approach, a distinction was drawn between pathways for the development of drugs versus biologicals versus devices. It was soon recognized that important developmental distinctions could be identified within these broad categories, and further refinements were proposed. As proof of concept, four rough drafts of flowcharts in the algorithmic style were prepared to characterize the developmental process for new clinical modalities including (a) drugs or biologics, (b) immunologics, (c) interventive devices, and (d) assay/biomarker devices. TRWG members found these sufficiently promising to warrant a focused effort to refine them for use.

Informal subcommittees of interested TRWG members were convened to review the initial drafts and provide expert input about the pathway elements and relationships. Each draft was developed through several rounds of review and revision, with input provided initially by the respective subcommittees, and then by the TRWG as a whole, and by attendees to the first public roundtable meeting. The key objective of the iterative process was to assure the accuracy and realism of the pathway details by testing them against the collective knowledge and experience of experts in each domain.

Eventually, it became apparent that the pathways fell into two broad domains characterized by their clinical goals: assessment and intervention. Within each of these domains, the translational products could be applied to various populations or used to characterize the process of neoplastic development and progression at various times. Thus, assessment devices can be used for screening, early detection, diagnosis, prediction, prognosis, or response assessment. Similarly, interventions may variously be used for primary, secondary, or tertiary prevention, or for treatment.

Further discussion clarified the distinctive developmental characteristics of life-style interventions compared with other interventions such as drugs, immunologics, and devices, and of imaging techniques compared with other assessment modalities such as in vitro assay/biomarker technologies. Thus, the TRWG finally agreed on six distinct developmental pathways to clinical goals: two assessment modalities and four interventive modalities.

Caveats. In creating these pathways, the TRWG was aware that such idealized representations cannot capture the full complexity of the real world. The pathways were intentionally represented at an intermediate level of detail—approximating 15 to 20 summarized key steps and decisions—rather than the hundreds that might be represented, for example, in a process engineer's comprehensive rendering. For each activity, decision point, parallel path, or feedback loop, it was understood that many more variations can occur—and indeed have occurred—in practice, and that not all steps may occur in each instance. In addition, the diagrams did not capture the full range of possible interactions between the pathways, nor did they address the ways in which insights gained from late-stage clinical trials can influence the development process. As the TRWG was focused on public investments in translational research, there was no attempt to address the influence that market conditions, projected financial return, or reimbursement considerations might have on development pathway decisions made in the commercial sector, however important these decisions are. Finally, although the pathways were intended to be globally relevant in their decriptions of the scientific process, differences in regulatory policies may limit their generalizability across the world.

Details and Commonalities of the Pathways

Basic elements of the pathways

Each pathway is built using a limited set of graphical elements. A rounded rectangle at the top of each diagram indicates the origin of the process. Square-cornered rectangles are used to indicate activity steps. Conditional tests, or decision steps, are represented as diamonds. Unidirectional arrows indicate the direction of the activity sequence, and the direction of transfer of supporting tools from their parallel development paths to the main path of modality development. Bidirectional arrows are used to indicate codevelopment or concurrent, interactive refinement. A rectangle with inset oval is used to conserve space in representing an iterative refinement process.

The generic pathway

The generic pathway in Fig. 1 captures the common underlying structure of the pathways in simplified form. Five phases of the development process are represented.

Credentialing. The initial phase of the developmental pathways draws attention to the process by which an advance in fundamental science is recognized as having potential clinical application, and the criteria that are used to determine whether investment in its development is warranted. Three key decision criteria are highlighted that are applicable to all of the pathways: strength of the empirical basis for anticipating clinical utility of modalities based on a discovery (scientific validation); clinical need, broadly construed (for example, cancer types of relatively low prevalence or that disproportionately affect underserved populations may be given extra weight); and anticipated feasibility of the development process. These three criteria were collectively called the “Decision to proceed” in the final report of the TRWG.

Supporting tools. The second phase acknowledges the role of enabling tools in the developmental process, and in particular, the fact that advancing a concept through the development process often poses technical challenges that require collateral development of new enabling technologies.

Creation of modality. This phase represents the actual reduction of the new concept to a tangible modality.

Preclinical development. The penultimate phase of early translation encompasses refinement of the new modality for real-world practicability, including, as appropriate, compliance with regulatory standards for safety and for manufacturing quality and consistency. This section of the generic pathway includes a representative feedback loop to capture the iterative character of much translational work (note that although this phase is designated “preclinical,” it may include clinical studies required to develop certain components of the modality). See the discussion of image-based assessment modalities elsewhere in this Focus (3).

Clinical trials. The final phase of early translation—initial validation of the new modality in humans—requires early trials to determine whether there is sufficient evidence of safety and efficacy to warrant the resource expense, effort, and burden on patients of a definitive trial of the new modality.

Two different classes of modality

The six types of clinical modality addressed by the pathways fall into two distinct classes: assessment modalities and interventive modalities. The concept of an assessment modality is a generalization of the familiar notion of a diagnostic device or technique and includes any tools intended to characterize the cancer-related health status of an individual. The information generated by such tools may be used for screening, diagnosis, staging, response assessment, prognosis or prediction, and is intended to inform decisions about preventive or therapeutic interventions. Any specific assessment modality might be used for more than one of these purposes. Interventive modalities are intended to change the cancer-related health status of an individual, via either prevention or treatment. The generic pathway applies equally to both classes of modality.

Assessment modalities.

• Biospecimen-based assessment modalities (e.g., biomarker assays).

• Image-based assessment techniques (computed tomography, magnetic resonance imaging, nuclear imaging, etc.).

Interventive modalities.

• Agents (drugs or biologics).

• Immune response modifiers (vaccines, cytokines, etc.).

• Interventive devices (e.g., radiotherapy devices, image-guided interventions).

• Life-style alterations (e.g., changes in risk-related behaviors or environmental exposures).

Details of the scope and distinctive features of each pathway are presented in the articles in this Focus (3–8).

Parallel paths of development

A key feature of the developmental pathways for the interventive modalities is the inclusion of three parallel paths, representing development of the modality itself as well as development of two different classes of supporting tools: tools for characterizing and evaluating the effects of the modality, and tools for defining the cohort for which the modality is appropriate. These parallel paths correspond to the three essential elements that converge in clinical trials and may be referred to by the mnemonic “ICE”: Intervention, Cohort, and Evaluation or Endpoint.

Depending on the details of a new modality, it is possible that some or all of the required supporting tools may already exist. However, these parallel paths have been made explicit to acknowledge that, in many cases, at least some of the required tools will not exist, and their parallel or codevelopment will be prerequisite for the viability of the new modality.

It will be apparent from examination of the detailed pathways that, for interventive modalities, many of the supporting tools required in the development process are themselves assessment modalities.

A further aspect of parallel development can be seen most clearly in the immune response modifiers pathway. Unlike drugs, which in principle are pure molecular species, immune response modifiers are often composite modalities. An immune response modifier may combine an antigen, a delivery vector, and an immune modulator, with the overall therapeutic effect dependent on the presence of all three components. The immune response modifier development pathway acknowledges the parallel and interactive development of these components of the modality itself. A simpler representation of parallel development of components may be seen in the pathway for image-based assessment modalities, which realize their functionality through a combination of agent, imaging platform, and technique.

Although the immune response modifier pathway is the one most reliant on the composite approach, it became apparent during refinement of the other pathways that other modalities may also take on a composite character, and in such cases may require attention to the timing and interaction of the development paths for the component elements. For example, agents are often used in combination regimens, and the future may see more widespread development of rational combinations of targeted agents designed to block distinct steps in oncogenic pathways, to synergistic effect. Analogous combinations may be imagined for the other interventive modalities as well.

Uses of the Pathways

The developmental pathways were initially envisioned primarily as a tool to facilitate the deliberations of the TRWG by clarifying essential aspects of early translation, establishing a common language, and suggesting background investigations (including analyses of the NCI translational research portfolio) whose results would inform debate. Working group members used the pathways to help understand the challenges faced by translational researchers and to identify ways to help the translational research process function more effectively. The effort of fleshing out the different pathways clarified their commonalities and their differences, highlighted the distinct challenges associated with specific modalities, and drew attention to modalities, such as life-style alterations, which traditionally had been underappreciated in this context. The pathway diagrams stimulated fruitful discussion among TRWG members and participants in the first TRWG public roundtable about relationships among different elements of the translational research effort, resources needed, and barriers that stand in the way of more rapid progress. A mapping of NCI programs that provide funding for translational research against the steps of the pathways, together with case studies of the progress of specific innovations through the relevant pathways, provided grist for fruitful discussions about possible new models for funding, adjustments in policies of existing programs, and improvements in communication and coordination across NCI.

The pathways also play an explicit role in several of the TRWG's recommendations (2) serving two main functions: as an analytic framework for assessing the content, character, and balance of NCI's portfolio of translational research activities; and as a foundation for innovative, milestone-based approaches to assessing and managing progress in translational research, at both project and program levels.

Research program management

The pathways can facilitate the development of new research coding systems, and the enhancement of existing ones, to properly capture the distinctive features of translational research—compared with fundamental research on the one hand or late-stage clinical trials, dissemination and implementation research on the other. In turn, such enhanced coding systems can facilitate more effective monitoring, evaluation, and management of translational research programs. Tracking systems based on such codes can provide insight into the scope of the activity of a program, relative to that required to advance concepts to fruition, and also into the balance of a research portfolio across the different modalities. Milestones based on the phases and activities captured in the pathways can be used to combine project-level information to develop measures of progress at the program level. The availability of detailed, explicit development schemas enables a more penetrating analysis of gaps and bottlenecks in a goal-oriented activity such as translational research, and provides a rational basis to prioritize activities, investments, and program improvements.

The analytic framework provided by the developmental pathways is also useful at the level of the individual translational research project at all stages from planning and proposal writing, through implementation to evaluation.

Research project management

In planning a project, the pathways facilitate analysis of gaps in the existing research base and help target proposed activities to most expeditiously advance the development process. In proposal writing, the pathways can help in arriving at a concise statement of goals and in elaborating convincing rationales for a proposed scope of activity, as well as in constructing project management plans that clarify dependencies among project elements; in sum, they can tie resources, pathway-based milestones, and timelines together in a robust and realistic way. During project implementation, pathway-based management can facilitate prediction of potential impediments, measurement of progress, determination of root causes of departures (in either direction) from timelines, and midstream adjustment of management plans as required. A more transparent account of dependencies among resources and activities can enhance the effectiveness of collaboration, including management of budgets and interfacing/handoffs between activities. Finally, milestone-based planning and implementation can enable improved fiscal management, provide a basis for demonstrating return on investment, and create possibilities for innovative, milestone-based funding mechanisms.

By serving as transparent, readily understandable tools for goal-oriented management at the project level, the pathways can also help advance the efforts of research sponsors to cultivate a distinctive research culture that is optimally matched to the needs and objectives of translational research.

Coordination of research efforts

The “roadmaps” provided by the developmental pathways can facilitate the identification of complementary elements in the development process and, correspondingly, of especially promising opportunities for collaboration. As noted in the previous section, pathway-based project management plans can enhance the coordination of activities within a project; the explicitness and transparency of the pathways may be especially helpful in addressing the challenges of coordinating effort across the diverse research disciplines and cultures that contribute to progress in translational research.

By clarifying the nature, relationships, and significance of different elements of a crossdisciplinary translational research plan, the pathways may also help improve the quality of study section review of multiple-principal investigators proposals in translational science, including collaboration as appropriate across study sections to mobilize the needed expertise to evaluate such proposals effectively.

Teaching

The function of the pathways is to abstract and clarify key concepts, activities, and dependencies from the immense complexity of real-world translational research. As a set of tools/concepts/processes, they can be valuable for didactic purposes.

Effective coordination of diverse activities across different disciplines to achieve translational goals requires not only deep scientific knowledge but also substantial management skill. However, existing training programs for physicians, scientists, and other biomedical research professionals rarely address management issues in a clear or systematic way. The developmental pathways provide a platform to introduce key management concepts in a way that directly addresses the challenges that leaders and participants in cross-disciplinary, team-based translational research will face. Indeed, the very nature of the pathways, as a cross-disciplinary application of traditional engineering and project management flow chart tools, highlights the need for a cadre of individuals with formal training that encompasses management and public administration as well as the clinical and biological sciences.

It can be difficult for the research community to convey the nature and significance of translational research to other professions, to legislators and other policy makers, to the news media, and to the lay public. How does it differ, on the one hand, from fundamental research and, on the other, from late-stage clinical trials or practice-based health services research? What is its unique and essential role in the continuum of cancer research? What are the specific resources on which it depends, including funding as well as in-kind resources such as biospecimen repositories that depend critically on public understanding and cooperation? In the face of high public expectations for advances in cancer prevention, detection, and treatment, what are reasonable timelines for translation into the clinic after a promising scientific discovery has been publicized? Versions of the developmental pathways—adjusted to an appropriate level of complexity and expressed in terminology accessible to the lay public—could and should be valuable tools for getting the message across.

Future Development of the Pathways

As the members of the TRWG gained experience with the developmental pathways, many came to view them as more than an expedient device to facilitate the efforts of the working group; they were recognized as powerful tools that may have broader and longer-lasting value. In keeping with the ever-changing state of translational science, the developmental pathways are envisioned as living tools. It should be expected that they will be revised and evolve over time, to reflect the evolution of translational research itself.

Optimal application of the pathway concept will depend on careful choice of the level of abstraction appropriate to the intended use. The six TRWG developmental pathways presented in this Special Focus were designed to render the complexity of translational research tractable in the context of a working group of researchers, patient advocates and policy makers charged within envisioning the future of NCI's translational research enterprise. Tradeoffs between realism, simplicity, and transparency were guided by the need to capture those features that were essential to wise decision making about future policy.

Versions of the pathways intended for other uses may serve more effectively with a different set of tradeoffs. For example, as noted above, pathways designed to facilitate communication with lay audiences may be further simplified. On the other hand, pathways intended to support intensive management analyses may require additional detail. Professional specialty groups may wish to elaborate on certain elements or phases while simplifying others, or may split existing pathways into different versions that vary in details for subclasses within the specified modalities. Versions intended for didactic use in management training may be annotated with relevant guidance. The TRWG invites and encourages adaptation of the pathways, with a variety of specific content, and to different levels of abstraction and detail as needed, to serve different purposes.

To facilitate the broad application and continued development of the pathways, NCI plans to establish a Web-based repository for the pathways. Contributions of derivative versions, collateral materials, case studies of applications, and reviews of best practices in applying the pathway concept to cancer research will be solicited.

Finally, although the pathways were developed specifically in the context of translational cancer research, this approach to analysis and representation of developmental processes should be applicable to translational research targeted at other diseases as well. NCI welcomes the opportunity to work with other institutes of the NIH and with other sponsors of biomedical research, public and private, to promote further development and broader application of pathway-based tools to enhance the productivity of translational research.

Disclosure of Potential Conflicts of Interest

B. Reid is a consultant to the NCI-Early Detection Research Network.