New Frontiers Program (NFP) Consensus Conference on Innovations in Imaging - Final Report
Prepared by:Graham Wright, PhD
Sunnybrook Health Sciences Centre
April 13, 2007
Summary
In an extensive consultation effort, a broad spectrum of imaging scientists, clinical imagers, clinicians, epidemiologists, policy makers, and industry representatives exchanged perspectives and worked toward a common vision of gaps in the process of moving innovations in imaging toward broad clinical application in stroke, lung, vascular, and cardiac disease, as well as strategies to address these gaps. To be useful, imaging information must influence patient management in a way that ultimately improves clinical outcomes and population health. Establishing such connections will require a coordinated effort relating imaging measures to disease pathophysiology, exploiting this information to direct therapy, and tracking differential health outcomes based on the characterization of the disease process and the associated patient management strategy. This will be facilitated by the creation of clusters of expertise to establish core resources and a common lexicon for image acquisition, analysis and interpretation linked to standardized characterization of disease progression at the molecular, cellular, and organ level. These cores should be linked in a broader network to facilitate ongoing development and evaluation of new biomarkers and associated imaging technologies and the utilization of associated measures at all stages of clinical trials of new therapies. This includes establishing tools for incorporating this information into extended databases while addressing security and privacy concerns. An imaging network should interact closely with other networks of expertise focused on various diseases and technologies to integrate their various databases into a rich resource where the entire research community can probe the full spectrum from genomic patient characterization through measures of pathophysiology progression and associated treatment decisions to health outcomes.
1. Objectives
Responding to a request for proposals from the Institute of Circulatory and Respiratory Health (ICRH), our objective was to hold a consensus conference involving imaging scientists, industry, clinical imagers, clinicians, clinical epidemiologists, payers, and policy makers to: 1) define gaps between basic imaging research and clinical imaging applications in areas of circulatory and respiratory sciences, including imaging technologies associated with the heart, lung, brain (stroke), and blood vessels, and, 2) determine how to measure the impact of existing and novel imaging technologies and imaging biomarkers in disease diagnosis, earlier risk detection, and better patient management including the development of novel treatments.
2. Process / Background Material:
Broad consultation has been a central principle in this effort. Following the plan outlined in the initial proposal for the conference (App. A), the steering committee interacted with numerous communities to develop the background information and to identify conference participants (including an internationally recognized expert in each area). In addition to these communities, organizers worked with the staff at ICRH to identify representatives from imaging, tracers, and pharmaceutical industries to attend and provide perspective on potential areas for funding partnerships.
The communities engaged in pre-conference discussions included those associated with the 4 major disease foci of the ICRH (Stroke, Respiratory, Vascular, and Cardiac) as well as those involved in evaluation, prioritization and policy making (Clinical Trials and Outcomes / Policy). Experts across many disciplines (imaging scientists, clinical imagers, clinicians, clinical epidemiologists and policy makers) participated. The communities with a disease focus identified clinical needs and imaging opportunities that would serve as examples to guide identification of gaps in the process of moving ideas from basic research toward broad clinical application. Similarly, the clinical trials and outcomes / policy communities reviewed recent efforts to evaluate emerging imaging technologies and identified examples to illustrate the challenges associated with such evaluations and potential strategies to address these challenges. Preliminary background material and a literature review in each of these areas are included in App. A. The results of subsequent community discussions are summarized in the reports from the various communities in App. B.
The conference structure was developed by the organizing committee in cooperation with professional facilitators. The 2-day conference was held March 30-31, 2007. The list of participants is provided in App C. The agenda and associated presentations are provided in App. D. The format involved plenary sessions and discussions among break-out groups, generally moving from a summary of perspectives from experts in each area toward consensus on the gaps in the "translation" process aimed at moving ideas from basic research toward broad clinical application and strategies to address those gaps. The first day emphasized the perspective of imaging scientists and clinicians in the various disease communities. Following summaries of examples of clinical needs and imaging opportunities, different categories of opportunities were discussed in terms of the magnitude of the effect, ease of evaluation, and ease of implementation. Finally, ways to move these opportunities toward broad clinical use were discussed and common themes were identified. Following a summary of the first day, the second day began with the perspective of experts in clinical trials, health outcomes, and health policy. Break-out sessions followed aimed at developing consensus among all parties on the infrastructure and programs needed to address the major challenges (identified as relating imaging measures to molecular mechanisms of disease and to disease pathophysiology at the organ level, as well as incorporating imaging into trials evaluating novel therapies and developing strategies to directly evaluate the impact of novel imaging tools in broad clinical application). Each break-out group was lead by both an expert in imaging and an expert in evaluative methods (clinical trials / health outcomes) and involved the full spectrum of expertise at the conference. In the final session, summaries of the break-out sessions were presented, followed by comments from representatives of industry, provincial health ministries, and the ICRH indicating where each saw the focus of their interests and hence their potential investment in mechanisms to address the identified challenges. The session concluded with a brief discussion of the conference itself and ways to improve such conferences in the future (summarized in App E).
A draft of this document, together with all the supporting material, was then posted on a website accessible to the conference participants for comment and modification prior to final editing and submission by the organizing committee. All the information on the site has also been made available to the broader communities involved in pre-conference discussions.
3. Results and Conclusions
As noted above, the observations and conclusions at each step of the process are presented in Appendices A, B, and D. Here, we present a synthesis of these observations following the two main themes: (1) gaps / challenges in moving innovations in imaging toward broad clinical application and (2) strategies to address these challenges.
3.1 Gaps / Challenges
The central challenge in moving innovations in imaging to broad clinical application is linking imaging measures to health outcomes. Understanding this link is critical to all facets of the "translation" process from development of new imaging technologies through appropriate clinical implementation to policy decision making regarding differential allocation of resources to improve population health. That said, imaging measures can only affect outcomes by influencing patient management and, more specifically, when imaging information is integrated with the application of effective therapies. This suggests the need for coordinated efforts to guide and evaluate imaging developments in the context of broader efforts to understand disease progression and to develop new therapies altering disease progression.
This challenge must be addressed in an environment where there is: a limited understanding of disease progression; heterogeneity across individuals in manifestation of disease and associated clinical management; and rapid evolution of imaging tools.
These issues indicate that the interplay between development and evaluation of imaging methods is inherently an iterative process, integrating mechanistic studies with prospective clinical trials and retrospective analyses exploiting rich databases.
Linking imaging measures to outcomes can be broken down into a cascade of related challenges:
- Relating imaging measures to molecular and cellular processes;
- Relating imaging measures to disease pathophysiology at the organ level;
- Use of imaging measures in prospective clinical trials of therapies;
- Developing practical tools for direct evaluation of imaging technology.
Relating imaging measures to molecular and cellular processes has rapidly evolved into the field of molecular imaging with associated foci in the development and study of imaging markers conjugated to targeting reagents. Experimental studies must be designed and interpreted in the context of a rapidly evolving literature in systems biology. These, together with studies relating imaging measures to disease pathophysiology, constitute basic mechanistic work needed as a foundation for the exploitation of imaging measures in broader clinical trials.
Prospective clinical trials of therapeutics with hard outcomes as endpoints should integrate imaging at all stages where possible. In such trials, imaging can play a role in patient selection, guidance of therapy selection and delivery, measurement of specific therapeutic action and biological effect, and prediction of longer term outcomes. While imaging measures may have some intermediate role as surrogate outcomes in such trials, they cannot be used as a substitute for hard outcomes.
The utility of all the aforementioned work depends on the quality of the imaging measures. The group from the Duke Clinical Research Institute advocates 4 stages of evaluation in human studies:
- Phase I - Proof of Concept: Establish how accurately an imaging measure reflects the pathophysiology in studies of highly selected populations;
- Phase II -Test Characteristics: Determine Sensitivity and Specificity in cohort studies involving multiple tertiary centres;
- Phase III - Efficacy / Implementation Trial: Pragmatic randomized study comparing a new test with current strategies in terms of effects on decision making, cost/benefit, and, ultimately effect on outcomes;
- Phase IV - Appropriateness and Imaging Registry - monitoring of image quality and ongoing use (clinical indications for which imaging is performed and effect on decision-making affecting patient management).
Clearly, imaging must be integrated into a much broader effort to develop a rich characterization of the disease process which is ultimately linked to outcomes. Many of the linkages can only be achieved retrospectively so it is important to work toward comprehensive image registries (including raw data together with image acquisition parameters) linked to registries of interventions and outcomes as well as other measures, such as registries of genomics, so that information can be tracked longitudinally in individual patients and across large populations. These databases will serve as the substrate for hypothesis-driven development of models of disease progression and interactions between patient genomics, environmental influences, and medical and surgical interventions. Such information must be broadly accessible to facilitate investigator-initiated studies probing linkages and the incorporation of the results of such studies to enhance the resource. It is only within a richer description of disease progression that the most cost-effective diagnosis and therapy strategies for individual patients can be determined.
For coordinated mechanistic studies, clinical trials, and database development, standardization of image acquisition, analysis, animal models, etc is needed. This will require close interaction among experts in each facet of the measurement process together with extensive communication across the many disciplines and communities involved in characterizing disease precursors and disease progression as well as treatment management, outcomes assessment and health policy.
3.2 Strategies to Address Gaps
Determination of all the infrastructure needed to properly address these challenges is beyond the scope of the conference. That said, an understanding of the larger context is necessary if targeted contributions are to be integrated effectively. Hence, recommendations first describe the larger context and then identify specific needs within that context.
Central to the strategy is a national imaging network to facilitate coordination of effort. Within such a network, we anticipate the development of clusters by areas of expertise, disease focus, imaging technology, and nature of question (from basic work in molecular imaging / molecular effect to clinical trials and registries). The principal functions of this network will be effective information exchange through standardization of data reporting and through websites, web meetings, teleconferences, workshops and symposia for scientific discussion and training. Network development should be guided by an organizing committee representing the broad spectrum of expertise involved in the consensus conference. The primary responsibility of the committee is coordination, providing a point of contact for potential partners and a clearing house for information.
Core activities / services within the network should include:
- organizing scientific meetings;
- creating, cataloguing and advertising national imaging research resources (centres of excellence in imaging technologies, chemistry labs for reagent / tracer / biomarker development, pathophysiology analysis labs, image analysis labs, core labs for image interpretation, training programs, databases/image registries);
- developing consensus on standards for archiving of images and associated acquisition parameters, image quality metrics, definition of image-based measures of anatomy and function and analysis methods, development and dissemination of standard disease models for use in imaging studies, diagnostic protocols;
- multi-centre coordination of imaging participation in clinical trials of therapeutics and clinical evaluation /comparison of novel diagnostic imaging technologies;
- management of regulatory issues (i.e. templates for interaction with Health Canada on device, method, and reagent approvals; assessment of safety) and privacy issues (e.g. data anonymization);
- providing a point of contact for linkage to other disease-based and technology-based networks (e.g. Genome Canada, Stem Cell Network, clinical trials networks, Centre for Stroke Recovery, Resuscitation Outcomes consortium, physiome project, DICOM standards committee) and for linkage of imaging registries to registries of genomics and health outcomes.
A major need associated with interactions across networks is the development of a common lexicon for cross-referencing information. For instance, descriptors of anatomy and physiology in models of disease progression need to be aligned with image measures aimed at describing these dimensions. Similarly, descriptors of molecular and cellular function must be aligned with associated targeting and reporting moieties. Key dimensions of clinical decision making in patient management and associated outcomes measures must also be aligned with descriptors used for image interpretation.
With such a structure, one can then put into context the role of imaging in characterizing disease, guiding treatment decisions, and affecting outcomes. Examples applied to stroke, respiratory, vascular, and cardiac diseases are found in the appendices.
3.3 Potential Roles of Funding Partners
Industry and provincial ministries of health are among the potential funding partners who could benefit from the development of various aspects of the proposed network infrastructure. Engagement with the research community would be greatly facilitated by clearly identified points of contact and standardization of tools across multiple research sites. Industries represented at the conference included those involved in the development and marketing of pharmaceuticals, biomarkers, and imaging technology.
The pharmaceutical industry is focused around understanding the disease process and associated molecular mechanisms. Imaging is seen as a tool in the development and evaluation of new pharmaceutical products. It can demonstrate effects at intermediate steps between application of therapy and clinical outcome: are characteristics of the disease being targeted by the drug evident in preclinical models and patients in a study population; does the drug get to the target; does it affect the expected mechanism; are there surrogate imaging measures showing early changes in the pathophysiological state that predict long-term outcomes. As a result, this group would be very interested in supporting and exploiting mechanistic studies linking imaging measures to disease pathophysiology at the molecular, cellular, and organ levels. The pharmaceutical industry also has a clear role in supporting the use of imaging as part of clinical trials evaluating the effectiveness of new therapies. Similar interest would be expected from industries associated with the development and marketing of interventional devices.
The imaging and biomarkers industries have direct interest in the commercialization of new technologies and the proposed four phases of technology evaluation. These industries are natural partners with imaging and clinician scientists involved in early technology development. They also have a strong interest in early disease detection and risk stratification and hence ways of linking imaging measures to underlying molecular and cellular processes predicting disease progression.
Provincial ministries of health have a role in technology assessment, primarily at the stages of implementation and monitoring of appropriate use (Phase III and IV). Several provinces - notably, Ontario, Quebec and Alberta - have developed specific programs for health technology assessment. Provinces also have a strong interest in the development of imaging registries for Phase IV studies and have the capacity to link such registries to population databases characterizing genomic profiles, monitoring treatment strategies, and health outcomes. Cost effectiveness studies are of critical interest at this level to provide evidence supporting decisions about relative allocation of resources.
3.4 Opportunities for Research Funding Agencies
The role of the research funding agencies (where ICRH has provided leadership through this initiative) is to engage individual academic researchers in this process of moving innovations in imaging toward broad clinical application. At the level of the CIHR New Frontiers Programs, this involves establishing mechanisms for the effective exchange of data and ideas among researchers and between researchers and the communities that can best define the needs and exploit the results.
At the conference, participants were encouraged to consider a coordinated response to the current CFI/CIHR sponsored Clinical Research Initiative (CRI) aimed at developing infrastructure and programs in support of clinical research. Information management infrastructure needed to establish and effectively use comprehensive image registries is seen as a core technology cutting across and linking to various disease foci.
Given its breadth, the CRI is an excellent opportunity to develop interactions across networks. That said, there remains a need to develop the infrastructure and programs specific to imaging and covering the spectrum from mechanistic to outcomes studies. A request for proposals in this area should facilitate the establishment of a national imaging network framework and the development of teams within that network focused around various core activities as described earlier. These teams may have a primary focus by disease or technology but should facilitate interactions across these dimensions. Of greatest importance, the products of team activities should be broadly accessible, in order to enhance the opportunities for investigator-initiated basic research and associated contributions to the larger effort.
Appendices
| Appendix A: | Proposal | 9 | |
|---|---|---|---|
| Appendix B: | Summaries from pre-conference consultations | ||
| Disease Focus: Stoke | 29 | ||
| Disease Focus: Respiratory | 33 | ||
| Disease Focus: Vascular | 35 | ||
| Disease Focus: Cardiac | 36 | ||
| Clinical Trials | 39 | ||
| Outcomes / Policy | 41 | ||
| Appendix C: | List of Participants | 42 | |
| Appendix D: | Conference Agenda and Presentations | 44 | |
| Day 1: | Opening Remarks and Introduction | 46 | |
| Presentations by Disease Foci | |||
| Disease Focus: Stoke | 55 | ||
| Response by: Alastair Buchan, University of Oxford | 60 | ||
| Disease Focus: Respiratory | 74 | ||
| Response by: Peter Macklem, McGill University | 80 | ||
| Disease Focus: Vascular | 100 | ||
| Response by: Jean-Claude Tardif, Montreal Heart Institute | |||
| Disease Focus: Cardiac | 110 | ||
| Response by: Marcelo DiCarli, Brigham & Women's Hospital | |||
| Break-out Sessions by Disease Focus | |||
| Disease Focus: Stoke | 124 | ||
| Disease Focus: Respiratory | 127 | ||
| Disease Focus: Vascular | 128 | ||
| Disease Focus: Cardiac | 132 | ||
| Day 2: | Presentation by Clinical Trials | 138 | |
| Response by: Manesh Patel, Duke University Research Institute | 149 | ||
| Presentation by Outcomes / Policy | 155 | ||
| Response by: Penny Ballem, Former Deputy Minister of Health, BC | 164 | ||
| Break-out Sessions | 177 | ||
| Response from Group 1 | 178 | ||
| Response from Group 2 | 183 | ||
| Response from Group 3 | 187 | ||
| Response from Group 4 | 189 | ||
| Appendix E: | NFP Meeting Evaluation | 196 | |
Note: All appendices are currently available on the website for the consensus conference (imagingresearch.desktop.com/consensusconference/). If access to the website is required, please contact 416-480-5738.