At NINDS
NINDS Staff
Advisory Council
Labs at NINDS
Groups at NINDS |
Government
Other NIH Institutes
Government Resources
Other related groups
Organizations
Professional Societies
NINDS is part of the
National Institutes of
Health
Additional information about Stroke PRG
Download the Full Report in PDF Format
Download the Brief Summary in PDF Format
To request a hard copy of the report, E-mail your name and complete mailing address to: StrokePRGcopy@ninds.nih.gov
Information Update:
|
It is my pleasure to share with you this Report of the Stroke Progress Review Group (SPRG).
Over the last few years, the National Institute of Neurological Disorders and Stroke (NINDS) has embarked on a strategic planning process to identify scientific opportunities that have the potential to significantly advance the fields of neuroscience and neurology and to address unmet scientific needs that may limit that potential. Our initial efforts resulted in the publication of our first strategic plan, Neuroscience at the New Millennium in 1999. In addition to this general plan, which continues to provide a framework for the Institute's initiatives, NINDS has also recently undertaken planning efforts in specific areas, such as Parkinson's disease, brain tumor, epilepsy, and health disparities research. These efforts have come about as a result of scientific opportunity and need, as well as Congressional and public interest.
The SPRG had its origins in Fiscal Year 2001 report language from the House and Senate Appropriations Committees to the NINDS urging us to develop a national research plan for stroke. Following on the success of the Brain Tumor Progress Review Group, a joint collaboration between NINDS and the National Cancer Institute to identify priorities for research on brain tumors, NINDS decided to use a Progress Review Group to develop a plan for stroke research.
The SPRG consists of prominent scientists, clinicians, patient advocates, and industry representatives who were chosen for both their expertise and their ability to think broadly. The SPRG was charged by NINDS with identifying and prioritizing the scientific needs and opportunities required to advance the stroke research field, and developing a research plan that addresses these opportunities and needs. This report is the culmination of their efforts, and outlines both research and resource priorities in fifteen areas of basic, translational and clinical stroke research.
The report has been presented to and approved by the National Advisory Neurological Disorders and Stroke Council, and will serve as a framework for the Institute's activities in stroke research over the next five to ten years. We look forward to working closely with the SPRG, together with the larger stroke research and patient communities, to advance the research field towards a better understanding of the etiology and pathophysiology of stroke and the development of more effective methods of both preventing strokes before they occur and treating them when they do.
Sincerely,
Audrey S. Penn, M.D.
Acting Director, NINDS
This report represents the collaborative effort of scientists, clinicians, industry representatives, and patient advocates who were charged by the National Institute of Neurological Disorders and Stroke with the task of setting overall priorities for stroke research.
The executive summary of the report outlines those priorities in light of the biological and clinical complexity of stroke and the formidable challenges that have slowed progress toward effective treatments. Many priorities and directions need to be pursued in stroke research, and they are discussed in detail in the breakout session reports. Common themes emerged from those reports, however, and the Stroke Progress Review Group considers the priorities delineated in the executive summary to be the best guide to the future direction of stroke research.
This report and additional related information are available at the Stroke Progress Review Group web page on the National Institute of Neurological Disorders and Stroke web site (www.ninds.nih.gov).
TopThe Report of the Stroke Progress Review Group (SPRG) is the product of work carried out over the past several months by the SPRG, the participants at the SPRG Roundtable Meeting, and staff of the National Institute of Neurological Disorders and Stroke (NINDS). The report is based on meetings of the SPRG leadership in Bethesda, Maryland, in October 2000; of the SPRG members in Crystal City, Virginia, in March 2001; of the Roundtable Meeting participants in Denver, Colorado, in July 2001; and of weekly conference calls of the SPRG leadership and NINDS staff throughout 2001.
Special thanks are extended to the NINDS Office of Science Policy and Planning for their extraordinary organization in all phases of the SPRG process. In particular, the guidance of Dr. Paul A. Scott, Dr. Melinda Kelley, and Ms. Patricia Turner has been invaluable. The dedication of Ms. Susie Nelson, of MasiMax, Inc., has been inspiring. The completion of the report was also greatly facilitated by Ms. Catherine Dold, who served as lead science writer, and by the breakout session reports prepared by Ms. Dold and the other expert science writers (Ms. Christie Aschwanden, Ms. Martha Engstrom, Ms. Vonne Sieve, Ms. Elizabeth Staton, and Dr. Linda White) at the Denver Roundtable Meeting.
Particular thanks are also due to the co-chairs of the Roundtable Meeting breakout sessions, who worked diligently with the SPRG members and participants to plan the breakout sessions and prepare the individual breakout session reports.
Finally, the SPRG recognizes the tremendous efforts of the stroke patient advocacy groups in supporting the NINDS in developing the SPRG process, and their invaluable participation in many aspects of the work.
TopWe are pleased to submit this Report of the Stroke Progress Review Group (SPRG) to the Director and National Advisory Neurological Disorders and Stroke Council of the National Institute of Neurological Disorders and Stroke (NINDS). In their FY2001 Appropriations Committee Report language, both the House of Representatives and the Senate directed NINDS to develop a national plan for both basic and clinical stroke research. At the beginning of 2001, the SPRG accepted this charge from Dr. Gerald Fischbach, Director of the NINDS, and moved quickly to develop an appropriate plan. The expertise and efficiency of the SPRG members and of the participants of the Roundtable Meeting have produced this exciting report in less than a year, reflecting the energy and enthusiasm of the clinical, research, industrial, and advocacy communities for identifying effective treatments for stroke.
The Report of the Stroke Progress Review Group highlights the scientific research priorities that represent the next steps toward understanding the biological basis of stroke and developing effective therapies for stroke. We look forward to discussing these priorities with the leadership of the NINDS.
James C. Grotta, M.D.Members of the Stroke Progress Review Group
SPRG Co-Chairs
James C. Grotta, M.D.
University of Texas
Health Science Center at Houston
Michael A. Moskowitz, M.D.
Harvard Medical School
The National Institute of Neurological Disorders and Stroke (NINDS) is the nation's leading supporter of biomedical research on disorders of the brain and nervous system, and supports basic, clinical, and population-based research to identify and study the causes, biology, prevention, early detection, and treatment of stroke. Through years of dedicated study, researchers supported by the NINDS have amassed a significant knowledge base about stroke, and this knowledge, coupled with new technologies, is providing a wealth of new scientific opportunities. At the same time, increasing research needs and scientific opportunities require that the NINDS determine the best uses for its resources. It is necessary to identify clear scientific priorities, both to provide guidance for the scientific community and to create a benchmark against which progress can be measured.
The Stroke Progress Review Group (SPRG) was convened to identify those priorities. It is modeled after the National Cancer Institute's (NCI) planning process, which was originally established to assist the NCI in assessing the state of knowledge and identifying scientific opportunities and needs within its large, site-specific research programs. The SPRG follows on the success of the Brain Tumor Progress Review Group (BT-PRG), which was jointly established in 2000 by NINDS and NCI in recognition of the importance of brain tumor research to both institutes.
The Stroke Progress Review Group was charged with assisting the NINDS in addressing the needs of the institute's stroke research program. SPRG members were asked to take a broad view in identifying and prioritizing unmet scientific needs and opportunities that are critical to the advancement of the research field. The SPRG was specifically charged with the following goals:
The SPRG members include prominent scientists, clinicians, consumer advocates, and industry representatives from the United States and Canada who together represent the full spectrum of scientific expertise needed to make comprehensive recommendations for the NINDS stroke research agenda. Members were selected for their expertise as well as their ability to take a broad view in identifying and prioritizing the scientific needs and opportunities that are critical to advancing the field of stroke research.
In February 2001, the SPRG leadership finalized an agenda and process for the SPRG Planning Meeting. At the Planning Meeting, held in March 2001, additional members of the stroke community were identified and invited to participate in a later Roundtable Meeting. Topics were identified for Roundtable Meeting breakout sessions, and all participants were assigned to attend particular sessions. The SPRG members were assigned to co-chair the breakout sessions.
The SPRG Roundtable Meeting, held in July 2001, brought together approximately 140 leading members of the stroke research and advocacy communities, representing diverse institutions and scientific disciplines. These experts met in an open forum (both as a large group and in smaller breakout sessions) to formulate the key scientific questions and priorities for the next five to ten years of stroke research. The NINDS provided the Roundtable Meeting participants with extensive information about their research programs for use in their review. The research priorities and resource needs that the Roundtable Meeting participants identified in the course of their deliberations are outlined in this report.
After the Roundtable Meeting, an intermediate draft of this report was prepared, and multiple iterations were reviewed by the SPRG leadership and SPRG members. Upon completion of the final draft, the report was submitted for deliberation and acceptance by the NINDS Advisory Council. The report will be widely disseminated and integrated into the institute's planning activities. The SPRG will meet with the NINDS director to discuss the institute's response to the report.
TopThe Stroke Progress Review Group (SPRG) occurs at a critical juncture for the field of stroke research. We have enjoyed wonderful progress during the past "Decade of the Brain." Highlights have included a number of successful large-scale clinical trials that have enabled physicians to make evidence-based decisions regarding stroke prevention and treatment. Even better, these trials have provided us with effective therapies to reduce the risk of stroke, prevent recurrent stroke, and reduce damage in the first minutes after a stroke has occurred. We have developed sophisticated imaging techniques that enable us to diagnose stroke and its pathogenic mechanisms rapidly and precisely. We have made substantial progress in unraveling the complex cascade of hemodynamic, biochemical, and molecular changes that occur in response to ischemic injury. And we have learned about differences in stroke incidence and outcome in various populations, stimulating current research to understand these epidemiologic trends on genetic, behavioral, and lifestyle levels.
Despite this progress, the challenge to make new strides in stroke research is more urgent than ever. From a public health perspective, stroke is the third leading cause of death in the United States, and a leading cause of long-term disability. With the aging of the population, the absolute number of stroke patients in the U.S. is likely to grow substantially. Stroke is also a significant burden on public health worldwide. Our understanding of the inherited basis of human disease is increasing dramatically, but the energy and focus of such genetic research has yet to be applied fully to stroke. Despite years of laboratory and clinical research, assessing the risk of stroke for individuals remains imprecise. Moreover, stroke is still difficult for non-specialists to diagnose, it is complicated to treat, and there are few effective therapeutic alternatives. Furthermore, too few medical graduates are choosing careers in laboratory or clinical stroke research to carry out the work that is needed to change this situation.
The purpose of the SPRG was to assemble the leaders in various areas of stroke research, as well as representatives of the stroke community, who could identify the current challenges and opportunities in the field. In the end, our goal was to lay out a broad menu of research priorities that might serve to both stimulate and guide stroke research over the next decade.
The SPRG identified 15 key areas of research activity in the field of stroke and brought together experts from the basic and clinical sciences, along with representatives from industry and the advocacy community, to discuss future goals for research in each area.
Because stroke research encompasses multiple and wide-ranging disciplines, experts from diverse backgrounds were invited to take part in the process; the participants included hematologists, vascular biologists, radiologists, clinical trialists, molecular biologists, geneticists, statisticians, vascular physiologists, adult and pediatric neurologists, neurosurgeons, neuroscientists, anesthesiologists, psychiatrists, behavioral scientists, and neuroepidemiologists. Two co-chairs were identified for each of the 15 research areas and asked to lead breakout discussion sessions. The participating experts were invited to take part in as many as three of the sessions.
In July 2001, all the SPRG participants met in Denver. During that meeting, each breakout session group met to identify their top three priorities for research focus and to highlight existing problems in their respective areas, barriers to this research, scientific goals, and the resources necessary to achieve these goals. The co-chairs then prepared documents summarizing each group's findings, along with their priorities. Those documents are found under Scientific Session Reports: Full Reports of the Stroke Progress Review Group Roundtable Meeting Breakout Sessions.
In reviewing the conclusions of the 15 breakout sessions, the members of the SPRG identified five broad Research and Scientific Priorities, as well as seven Resource Priorities needed to implement such research. These priorities have broad implications -- they apply to adult and pediatric patients, to individuals with ischemic or hemorrhagic stroke, and to underserved patient groups within the population. The sections that follow summarize the highlights of the Denver Roundtable Meeting and formulate the common themes heard in the sessions into a vision for the future of basic and clinical stroke research.
The research priorities listed below represent a consensus of scientific goals expressed in many of the 15 breakout sessions. In order to satisfy these priorities and to successfully prevent and treat stroke in the future, each priority must be implemented by strong bi-directional interactions between basic and clinical stroke researchers. All of these priorities were considered equally important in accomplishing the overall goals of the PRG.
SCIENTIFIC PRIORITIESThere was universal agreement that the field of stroke is ripe for the genomic revolution that is now creating new and previously unimagined opportunities for diagnosis and treatment of neurological and other diseases.
Advances in mapping the human genome have recently made it feasible to identify and isolate genes that predispose individuals to ischemia and hemorrhage, to understand ways in which the encoded proteins modulate vascular physiology, and to recognize the cellular mechanisms of injury and death that can be initiated by stroke. For example, by identifying stroke-related genes, we can identify populations at risk with greater precision and develop more specific and effective measures for first and recurrent stroke prevention. Knowledge about the genes and proteins expressed during acute injury can help us to better classify and understand the natural history of human stroke subtypes, and to understand the biological basis for these subtypes. Furthermore, the pattern of genes expressed during stroke can be useful to detect the presence, time of onset, and extent of stroke in the emergency setting, thereby aiding physicians in diagnosing ischemia and hemorrhage more quickly and with greater accuracy. In addition, by knowing the patient's genetic profile, therapies can be individualized and tailored to variations in the genome that dictate the optimal response to specific drugs. Finally, once we identify and isolate the genes and proteins modulating responses to chronic injury and repair and determine how they work together, we may better understand the biological basis for recovery and rehabilitation and expand the limits of brain function after stroke.
Extensive studies of neurons and glia have resulted in a detailed but still incomplete understanding of ischemic injury. If we are to understand, prevent, and treat stroke effectively, we need to investigate local hemostasis and its relationship to local tissue factors, microglia, endothelium, and the cells of the blood-brain barrier, including astrocytes. At a fundamental level, there is an important need to better define the molecular influences and cell-signaling mechanisms that characterize the interactions between circulating blood elements and the blood vessel wall, extracellular matrix, glia, and neurons (together, the neurovascular unit) during ischemic and hemorrhagic stroke. These interactions critically define events that initiate ischemia, hemorrhage, brain inflammation, blood-brain barrier dysfunction, and white matter changes after stroke. Progress in the prevention, diagnosis, and treatment of stroke will depend upon a critical understanding of these interactions.
To achieve this goal, the SPRG members recommended greater focus on the process of hemostasis, platelet and leukocyte function, and particularly, aspects of their interactions that are unique to the brain. This knowledge may be useful to identify potential therapeutic targets even more specific for stroke than for thrombotic events in other organs. The SPRG members also emphasized the importance of studying the extracellular matrix proteins that play a role in the development of hemorrhage, inflammation, and blood-brain barrier dysfunction after stroke. In addition to these proteins, the SPRG members highlighted the need to study glial cells and their role in blood-brain barrier integrity, synaptic and trophic functions, inflammation, and angiogenesis. Glia and matrix proteins are fundamental to white matter structure and function, and the white matter lesions that commonly develop after stroke can cause or contribute to vascular dementia. Finally, a study of the blood-vessel wall/matrix/glial interaction would be incomplete without an emphasis on stroke risk factors such as diabetes, hypertension, atherosclerosis, and obesity, and their impact on interactions within component cellular and acellular elements of the neuro-vascular unit. We do not have a full understanding of how these very common diseases modulate hemostasis and vessel wall structure and function to specifically place the brain at high risk for stroke.
The fundamental pathophysiology of stroke is caused by an interruption of blood flow. Building on our existing knowledge of cerebral blood flow and metabolism, we should explore emerging imaging technologies that will enable us to understand the regulation and restoration of blood flow after both ischemic and hemorrhagic stroke.
We need to better understand how to optimally reestablish flow in the macro- and microcirculation. One important approach will involve the amplification of existing acute ischemic stroke therapy by accelerating the testing of devices and new pharmacologic approaches to achieve reperfusion more quickly, more completely, and safely.
Reperfusing brain quickly can improve recovery, but reperfusion can also promote mal-adaptive responses. We need to understand the consequences of reperfusion at the molecular and cellular level so that tissue survival can be optimized in the reperfused brain.
Many clinical questions must be addressed. Among them:
Despite substantial investigation into the biology of ischemic and hemorrhagic injury over the past two decades, there is still no effective therapy that targets the toxic events that develop within cells and tissues as a consequence of stroke. This type of therapy would fulfill an important need since some patients cannot be treated with clot-lysing compounds, and others could benefit from a strategy that combines neuroprotectants with clot lysis and other strategies to reduce tissue injury.
Combination therapy has been successful in treating other diseases, such as hypertension and diabetes, that have been resistant to treatments that target a single cellular or molecular mechanism. Emphasis should be given to research that promotes a more complete understanding of the natural neural pathways that protect the brain and of the blockade of pathways triggered after stroke that cause cell death alone and in combination. Therapeutic strategies based on these mechanisms should be developed, particularly after validating them by in vivo or in vitro studies in human or primate tissues or cells. As an important practical issue, drug treatment could be improved greatly if we had a better understanding of the complexity of drug delivery to the ischemic brain and optimized transport of drugs into injured tissue. This information is essential to interpret complex outcomes from clinical trials and to improve the possibility of identifying more effective treatments.
To develop combination therapy with a high probability of efficacy in humans, members of the SPRG emphasized the need to develop and validate large and small animal models that reflect the complexity and diversity of the human brain and its responses during stroke. To facilitate model development and validation, the genome of large animals (e.g., pigs, sheep, and primates) should be sequenced along with the use of mathematical and statistical methods to improve the efficiency of combination drug therapy. Molecular imaging technologies should also be developed to profile gene expression after stroke, to validate stroke in animal models, and to identify therapeutic targets. Ideally, these technologies will inform us about the molecular, cellular, and synaptic events that predict stroke outcome, response to therapy, and recovery of function in humans.
The SPRG members strongly emphasized the need to develop new therapeutic approaches to restore lost motor and cognitive function after stroke. At the moment, very little is known about the mechanisms that govern stroke recovery, and the natural history of recovery in humans and in animal models is incompletely understood. Evidence from brain injury in the clinic, particularly in children, strongly suggests that the brain does exhibit self-repair mechanisms that involve complex coordination between endogenous and exogenous elements, including blood vessels of the brain, neurons, and glial cells. However, the precise molecular and cellular events are not well understood. Nevertheless, it is becoming increasingly clear that brain recovery and remodeling occurs in response to external influences such as drugs and physical rehabilitation. To understand and perhaps amplify this process, we need to characterize the molecular and cellular mechanisms by which behavioral experience and environmental enrichment modulate the recovery process in brain after stroke. In particular, we need to develop rational pharmacological strategies based on these molecular mechanisms and determine the importance of genetic factors as a predictor of stroke outcome. Finally, we need to explore the potential use of stem cell technology as a tool to augment brain recovery in adult and pediatric stroke patients.
There is general agreement that the development of new and emerging technologies, as well as the application of existing ones, will be necessary to implement the research and scientific priorities and goals discussed above.
There is also general agreement that, more specifically, the seven resource priorities listed below, identified in many of the breakout sessions, will be necessary to meet those goals. These resources will help researchers generate and test hypotheses important to understanding all basic and clinical aspects of stroke, and to advance the prevention, diagnosis, prognosis, treatment, and rehabilitation of stroke patients. All of these priorities were considered equally important in accomplishing the overall goals of the PRG.
RESOURCE PRIORITIESBreakthroughs in science and technology have altered immeasurably the practice of neurology and have improved our understanding of basic disease mechanisms. Gene microarrays, in particular, are a recent breakthrough developed from advances in miniaturization, microfabrication, and high-density chip technologies that provide state-of-the-art platforms for genomics, proteomics, and pharmacogenetics. Microsystems such as these may become useful to generate data reflecting changes in enzyme activity, protein-protein interactions, and receptor-ligand binding plus gene expression. In addition, chip technologies may one day provide an individualized molecular portrait of stroke and its recovery course as well as a blueprint for therapy. Nominating and then testing candidate genes or mechanisms of interest individually will no longer be required, as thousands of different gene candidates can be assessed simultaneously within a single drop of fluid. Used in conjunction with molecular imaging techniques, markers in blood or other body fluids may then be used to profile stroke as it evolves. By applying these techniques, we may learn how molecules compromise cells, as well as parse their individual contributions to stroke pathogenesis.
Whereas gene chips and arrays use micron-based technologies, nanotechnologies focus on even more miniaturized systems and the manipulation, assembly, and targeted delivery of molecules into nanoparticles for applications such as biosensing, drug delivery, and cell repair. At such dimensions, nanoparticles may be particularly useful because the blood-brain barrier becomes less of an obstacle to drug delivery during stroke.
High-throughput initiatives, albeit exciting, are expensive and require centralized resources and high-throughput data analysis (bio-informatics). The sheer weight of the information generated, which is often non-intuitive and cryptic, can be daunting, and will require advanced data processing capabilities. The SPRG embraces the notion that emerging micro- and nanotechnologies will make it possible to investigate stroke in ways not previously possible.
One theme raised in many breakout sessions was the need to reconcile clinical and laboratory disciplines in all areas of stroke research. This includes better models of stroke disease, especially in primates.
Improved animal models are needed to accomplish all five of the SPRG research and scientific priorities; their availability would help to advance drug development as well as our understanding of basic stroke biology. In this task, special emphasis needs to be given to species differences in hemostasis, inflammation, white matter content, and brain size, as well as vascular considerations such as ana-tomical distribution and regulation. Model validation is deemed essential and will require, at a minimum, the use of microarray and imaging tools and the development of physiologically based behavioral and pharmacological assays that accurately reflect the human condition in both short- and long-term studies.
Potential applications include:
Models can also be useful in the context of developing a more comprehensive understanding of stroke recovery and developing treatments that enhance stroke rehabilitation and maximize the potential for restoration of function. Accordingly, these models can be used to:
Because animal models for stroke are technically difficult to develop and often require special facilities for surgery, imaging, and housing (e.g., primates), we need to encourage collaborations between groups dedicated to perfecting these models and laboratories applying these models to complementary research interests. Development of models of both brain ischemia and hemorrhage remains a high priority.
Brain imaging already has revolutionized the diagnosis and management of stroke. We need to develop new imaging techniques. We also need to better understand the existing modalities, to improve our understanding of stroke pathophysiology and recovery, and to provide a trans-lational link between experimental advances and clinical applications.
Imaging techniques could be used to:
In addition to the above, there are important unmet needs that require further technology development and validation. The needed technologies include:
Thus, the SPRG places a high priority on developing and validating new imaging markers and techniques to facilitate the spatio-temporal assessment of stroke at both the tissue and the molecular level.
Clinical trials are necessary to define how to best apply basic research advances to the treatment of patients. Building on the conspicuous successes of NINDS-sponsored clinical trials in stroke prevention and treatment during the past decade, new, better designed clinical trials will use innovative approaches to get needed answers efficiently and expeditiously.
Clinical trials in stroke are time consuming and expensive. These constraints may serve to discourage innovative or start-up strategies as well as drain the good will and resources of funding agencies, investigators, clinical resources, and patients.
We need to develop more streamlined clinical trials of prevention strategies and acute stroke therapy by improving trial design, developing and testing new outcome measures, and forming clinical trial consortia/networks. Depending on the questions asked and the population studied, both large simple trials and smaller focused trials with surrogate endpoints should be explored. Proper guidelines for such studies should be developed. We need to develop outcome measures with more relevance to the patient as well as measures that might be more sensitive to therapeutic effect than those currently used. Treatment trials in certain clinical areas have been relatively ignored and need more attention; these include intracerebral hemorrhage, pediatric stroke, and rehabilitation.
There also needs to be greater collaboration between industry and academia in designing, prioritizing, and funding clinical trials. Networks of collaborating centers and individuals interested in conducting clinical trials should be established to help prioritize resources and expedite trial execution. The establishment of specialized centers pioneering translational research in acute stroke will expand treatment options for acute stroke when prevention fails. Finally, we need to develop better methods for encouraging more physicians, patients, and advocacy organizations to participate in clinical trials. Clinical trials are the front lines in the fight against stroke and will define the next generation of treatments used in clinics throughout the world.
The SPRG recognizes that existing and future preventive and acute interventional therapies need to be more widely and rapidly adopted by health care workers and patients. This could be accomplished by a better understanding of the existing barriers to health services implementation pertaining to stroke, including the lack of incentives, information, and essential personnel and technologies. We need more accurate and universal information regarding existing practice patterns and we need to understand the administrative barriers to obtaining medical resources and care.
To promote better implementation we need to develop and test interventions aimed at improving community practice, and partner with payors and other groups in stimulating good practice. We need greater regional collaboration to develop multidisciplinary teams (i.e., stroke center networks) that can better overcome the local barriers that exist to implementing stroke prevention and therapy, including health disparities.
A national stroke surveillance system would establish a database of stroke that would help characterize the public health burden of stroke and identify those populations that need special emphasis. An effective database would include substantial socioeconomic and ethnicity detail that is often unavailable when doing epidemiologic analyses. A data-base would also facilitate the study of the many stroke-related conditions that occur too sporadically for randomized comparison studies.
Such a database will also facilitate the application of targeted genetic analyses. The complex variability of the stroke phenotype requires such a database in order to carry out research on stroke genetics. Genetic databases are also needed. These would be particularly important in sharing and sifting through the exploding information in this area. A centralized genomic/proteomic/ bioinformatic facility would support the establishment of such a database.
Stroke information that is widely available to clinicians in an electronic format would help to foster the development of collaborative consortia and standardized methodology for conducting research and for patient management, and might help increase implementation of therapies.
It is clear that prevention, diagnosis, and treatment of stroke is a public health problem that is too large to be managed only by stroke specialists. Yet training of other medical personnel in modern stroke management is currently inadequate. Non-neurologists and neurologists alike need more exposure to the advances made in the field of stroke. Even more important, the next generation of medical personnel should receive an educational curriculum, starting early in professional school, that ensures they will be more knowledgeable about stroke.
We also need to improve the training of neurologists in the emerging disciplines that will be critical for researching and applying new stroke therapies, including genomics, endovascular therapy, imaging, and rehabilitation. Existing barriers to such cross-training should be identified and eliminated.
As we focus our research on the interface between circulation and the brain, the lack of neuropathological information about and expertise needed to effectively study stroke is recognized as a major deficiency.
Stroke is the third leading cause of death and a major disabler of the American people. Although many challenges lie ahead, we are currently experiencing an extraordinary and unprecedented time of scientific growth and technological breakthroughs. Our greatest advances in stroke research have been made in the prevention of stroke through surgical and drug therapies. Early stroke treatment with t-PA (tissue plasminogen activator) has reinforced the belief that stroke is a treatable disease. However, now we are in great need of new treatments that reduce damage and promote recovery once a stroke has occurred. To attain these reachable goals, we will require new initiatives and new applications of technologies that can advance the field of stroke in the laboratory and at the bedside.
The research and scientific priorities and resource priorities identified by the SPRG provide an outline for academia, industry, government, and patient advocates to guide progress in stroke research. Commitment and joint sponsorship among these vested communities to address these priorities will facilitate the development of creative solutions to prevent, diagnose, and treat stroke in the current decade and beyond.
TopCo-Chairs: Bruce M. Coull, M.D., and Donald Heistad, M.D.
Participants:
Richard A. Cohen
Frank M. Faraci
Mary Gerritsen
Gary Gibbons
Willa A. Hsueh
J. Paul Muizelaar
Stephen M. Schwartz
Katherine Woodbury-Harris
Research on basic vascular biology has now provided us with the underpinnings needed to understand vascular diseases in specific organs. This critical information includes an understanding of how vessels develop, as well as the molecular and functional differences between the endothelial and smooth muscle cells making up arteries, veins, and capillaries.
It is likely that extension of this basic biology to the neurovasculature will lead to fundamental new directions in cerebral vascular biology (as it already has in the biology and pathology of other organ systems). Major advances in the understanding of causes and treatment of stroke (especially cerebral hemorrhage) are likely to be delayed until this improved understanding of neurovascular biology is achieved. As a precedent, lessons learned from vascular biology were critical for recent advances in treatment of myocardial infarction.
Obvious areas of interest include characterization of physiological responses to acute stimuli and to major risk factors, especially hypertension and diabetes, as well as inflammatory diseases involving the nervous system. The availability of comprehensive human and mouse genetic data, as well as newly developed technologies including arrays, genetically altered mice, and proteomics, should move the field of cerebrovascular biology very rapidly.
CHALLENGES AND QUESTIONSThe neurovasculature has many properties that are not found in other organ systems. These properties likely account for a great deal of neurovascular pathology, including not only obvious targets such as neurovascular spasm and stroke, but also less obvious targets as diverse as brain metastases, develop-mental anomalies, and inflammatory diseases.
We already know that brain endothelium is distinctive, as manifested by the blood-brain barrier, for example. The role of smooth muscle cells in the distinctive characteristics of cerebral vessels is not well understood, and the likely role of smooth muscle in determining the unique phenotype of cerebral endothelium has not been explored. Adventitia, which has emerged as an important tissue in regulation of blood vessels, is less prominent in cerebral vessels than in extracranial vessels. The functional implications of this structural difference are not clear.
A number of opportunities for study already exist based on current knowledge of vascular biology. For example, although we have some knowledge of the effects of risk factors on cerebral blood vessels, other important risk factors (including atherosclerosis, diabetes, smoking, and aging) have received little attention. Newly recognized risk factors, including hyperhomocystinemia and chronic inflammation, may be fertile areas for study.
Although inflammation and infection may play an important role in cerebral vascular disease, new classes of anti-inflammatories directed at vascular adhesion molecules, chemotactic factors, and the death receptor family have received little attention in the neurovascular system.
The role of oxidative and antioxidant mechanisms in cerebral vessels is an especially promising area of research. Recent studies suggest that oxidant injury may be an underlying mechanism in vascular injury in response to a variety of stimuli.
BARRIERS
Current and future research should focus on building the basic knowledge of vascular biology needed to proceed with more broadly based efforts with a disease focus. The obvious challenge is to leverage current knowledge of vascular biology with the opportunities offered by new methods to accelerate research. Applications of arrays, genetically altered mice, and proteomics, combined with our existing, finite knowledge of the entire set of transcribed genes, should greatly accelerate research in this area.
Priority 1:
Understand developmental and basic aspects of cerebral vascular biology.
The basic discoveries of developmental vascular biology have identified mechanisms underlying not only the formation of blood vessels, but also the mechanisms of vascular response to injury in general. Brain-specific vascular biology is needed to identify the precise mechanisms underlying neurovascular disease. Specific questions to address include:
Priority 2:
Understand mechanisms of response to injury.
Reactive oxygen species are products of metabolism in ischemia, and are produced by specific enzymes. Oxidative mechanisms may regulate vasomotor responses of cerebral vessels to ischemia, the inflammation accompanying brain ischemia, remodeling associated with cerebral vasospasm, and chronic effects of risk factors on cerebral vascular structure and function. Specific areas to be explored include:
Priority 3:
The application of developmental and basic aspects of cerebral vascular biology and mechanisms of response to injury can provide a deeper understanding of vascular patho-physiology of great importance to stroke. The approaches outlined above can address the consequent effects of recognized risk factors for stroke and may help to elucidate new stroke risk factors. Research priorities include:
large number of resources now exist that may enhance the study of the neuro-vasculature, but they have not been fully evaluated. Resources that should be evaluated include:
Training and research needs include:
Co-Chairs: Justin A. Zivin, M.D., Ph.D., and Joseph S. Beckman, Ph.D.
Participants:
Michael Chopp
David J. Fink
Gary Fiskum
Frank W. Marcoux
Michael A. Moskowitz
Lennart Mucke
Robert Rothlein
David S. Warner
STATEMENT OF THE PROBLEM
The cause of neuronal death in stroke is deceptively simple. When blood flow to any part of the brain is stopped for only a minute, neuronal function is impaired. A complex cascade of events is set in motion that irreversibly damages the brain over the next few hours. When blood flow can be restored within three hours of the onset of stroke, a substantial fraction of brain function can be rescued. However, much of the injury is irreversible after six hours. Any combination of therapeutic approaches that could somehow expand this narrow window would have significant health benefits.
Over the past few decades, remarkable progress has been made in unraveling the mechanisms that cause neurons and glial cells to die after stroke. By the early 1980s, a variety of clever methods had been developed to induce focal strokes in small-animal models, which enabled large numbers of therapeutic and transgenic approaches to be tested rapidly and inexpensively. Hundreds of studies have demonstrated that cooling the brain, blocking the actions of excitatory neurotransmitters and inhibiting free radicals, nitric oxide, proteases, and caspases, can reduce infarction by as much as 70 percent in these animals. This rich literature has demonstrated that the complex cascade leading to permanent brain injury can be stopped at many points if initiated in the proper time window.
The problem remains that no individual approach, with the exception of thrombolytic agents given within three hours, has translated into a clinically useful treatment for stroke. We need to understand why animal stroke models fail to predict clinical trial results. We need to better characterize the complex interactions between different components in the cascade initiated by ischemia in animal models, and we need to understand how they relate to human disease. Recent progress in other degenerative diseases, as well as advances in genomics and proteomics, can be better applied to provide insight into cellular mechanisms of injury and death activated in stroke.
CHALLENGES AND QUESTIONS
The brain is more susceptible to ischemic injury than is any other organ. Failure of energy production causes a flood of neurotransmitters to be released from neurons, which further amplifies the damage. A major contributor in this regard is the release of the excitatory amino acid glutamate, which activates several different types of channels to allow toxic concentrations of calcium and zinc to enter neurons. A large research effort has involved the development of antagonists of these glutamate channels, but clinical trials in acute stroke have been disappointing. The toxic effects of glutamate may occur too rapidly to be prevented in treating stroke patients. However, prophylaxis may be possible, and protection from brain ischemia resulting from surgery can potentially aid 400,000 Americans a year.
In the past decade, significant progress has been made in understanding the intracellular signaling cascades involved in the cell death process. Many genes have been identified that are involved in Alzheimer's and Parkinson's disease, amyotrophic lateral sclerosis, and other forms of neurodegeneration, and the functional consequences of these mutations are gradually becoming known. A variety of approaches may help rescue these damaged neurons, to restore their function. Understanding these molecular mechanisms is helping to broaden our understanding of neuronal injury and death in stroke.
When stroke occurs, many neurons die by a process called apoptosis, where a series of "suicide" enzymes become activated and internally digest neurons. Drawing from progress in cancer and other fields, many strategies have been developed to block apoptotic death and these same treatments have been shown to protect the ischemic brain. This type of cell death takes time and so might be successfully treated hours after a stroke has occurred.
Mitochondria, small organelles in cells that produce energy in the form of adenosine triphosphate, have a central role in initiating the apoptotic cascade. Additionally, for many years, mitochondria have been viewed as no more than small engines that stall in response to a lack of oxygen and fuel during ischemia, but are ready to restart immediately if blood flow is restored. However, prolonged ischemia can cause mitochondrial damage, which leads to further injury. A variety of metabolic interventions may help preserve mitochondrial function and improve stroke outcome.
Many cells in the damaged brain die by less well-understood necrotic mechanisms. For example, at the edge of an ischemic region there is marginally perfused tissue that can potentially be salvaged; this tissue is known as the ischemic penumbra. Furthermore, it is not known which factors determine whether a brain region suffers a diffuse loss of neurons or frank infarction. Still, even in forms of cell death that are less well understood, certain pathways can be pharmacologically inhibited. For example, DNA repair by PARS may be such a mechanism that can be potentially reversible with appropriate therapy.
In the days following a stroke, damaged regions of brain undergo a broad-scale necrosis, which causes the death of all types of cells, including astrocytes, Schwann cells, oligodendrocytes, supporting microvessels, and neurons. Identifying the interactions between the different cell types in brain that lead to the development of this necrosis is a crucial need. For example, the responses of supporting cells, like astrocytes and microglia, in the brain during a stroke are critical determinants of injury and an area that needs further investigation. Microglia are important for defending the brain from infection, and activation of these cells can produce a wide range of proinflammatory and toxic molecules that will further damage the brain. Advances in understanding neuroinflammation from diseases like multiple sclerosis and infections may be useful in understanding reactions of the brain to ischemia. Additionally, swelling of the brain, or edema, is a major complication of stroke that occurs days after the initial injury. The basis for swelling in the brain parenchyma needs to be better understood.
In some instances, the brain tissue does not die shortly after the onset of ischemia. Brief periods of ischemia can produce long-lasting changes in the brain that substantially increase its resistance to subsequent longer ischemic challenges. This suggests that the brain initially responds to stroke by inducing protective mechanisms, which can be overwhelmed with sustained ischemia. Some changes, such as acidification of the brain, were first viewed as exclusively damaging, but later found to also be protective. For example, it is possible that acidification can protect neurons by diminishing the activation of glutamate channels to prevent subsequent damage.
There is also growing recognition that mechanisms that protect the brain from some types of insults can have negative consequences in other circumstances. Pathways that are protective early in stroke may amplify injury as the stroke evolves.
A crucial next step is to examine how different brain components interact during a stroke to produce injury. How does the cell death cascade initiated by stroke evolve over time and how do the various biochemical steps interact? If one part of the cascade is prevented, what factors decide whether the remaining tissue will die? It will be necessary to revisit many previous results using recent technological advances, to take into account the multiple actions contributing to brain injury.
Barriers
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:Improve animal models and endpoints to more closely reproduce the complexity and diversity of human disease.
Priority 2:
Better define the interactions between components of the ischemic cascade in multiple animal species and in relation to human cerebral dysfunction and recovery.
An enormous effort has focused on investigating individual mechanisms that lead to infarction. These studies are appropriately funded as hypothesis-driven projects. However, there is a great need to understand how these different processes interact. For example, energy failure induces the necrotic death of neurons. If necrotic death is prevented, will the same neurons die a short time later of apoptosis?
Areas that should be investigated include:
Priority 3:
Develop methods to investigate how simultaneously altering several components of ischemic injury modulates the evolution and final outcome of stroke.
Combination therapies will almost certainly be used to treat stroke patients, but few combinations have been adequately tested in preclinical models. Thrombolysis is currently the only approved method for treatment of acute stroke. All patients who meet the treatment criteria receive this therapy, but it is ineffective in the majority of these patients. For the large majority of patients who do not meet the rather stringent guidelines for thrombolytic therapy, this intervention can be ineffective and even harmful. Combinations of thrombolytics plus neuroprotectives or various classes of neuroprotectives may be synergistic.
RESOURCES NEEDED
Priority 1:
Priority 2:
Priority 3:
Funding is needed to support preclinical investigations. Resources should also be devoted to developing mathematical models and statistical methods that can improve the efficiency of combination studies. Other resources needed are:
Co-Chairs: Randolph J. Nudo, Ph.D., and Frank R. Sharp, M.D.
Participants:
Michael Chopp
Steven C. Cramer
Seth P. Finklestein
Wolf-Dieter Heiss
Alan J. Jacobs
Barbro B. Johansson
Thomas A. Kent
Michael A. Moskowitz
Evan Y. Snyder
STATEMENT OF THE PROBLEM
It is known that limited functional recovery can occur during the weeks and months after stroke. The current challenge to the scientific and clinical community is to develop new therapeutic approaches to restore lost function.
Over the past decade, potential mechanisms that underlie recovery of motor and cognitive function after stroke have begun to emerge. In addition to the resolution of acute pathophysiologic events associated with ischemia, several long-lasting processes have been identified that may play a role in recovery. Animal models recently have provided detailed information regarding neuroanatomical and neurophysiological plasticity in the undamaged cortical tissue during recovery. However, the degree to which each of the long-term alterations in neural, glial, and vascular systems contribute to behavioral recovery is still not known. Also, while modern neuroimaging techniques have advanced our understanding of the long-term changes in brain function after stroke, these techniques have yet to address changes at the cellular and molecular level during the recovery process.
These new insights into the mechanisms underlying brain remodeling and the role of motor experience after stroke in modulating those mechanisms have already resulted in promising novel therapeutic approaches in chronic stroke. Once these processes are better understood, it should be possible to identify patients who could benefit from a particular intervention, and to devise therapeutic interventions to maximize functional recovery. The goal of restorative neurology and neuro-rehabilitation over the next decade should be to design successful clinical trials based on the underlying mechanisms of recovery. Such information is potentially of critical value in defining molecular targets for restorative therapies.
CHALLENGES AND QUESTIONS
Mechanisms
Self-Repair Mechanisms
"Self-repair" mechanisms are constitutively triggered during the acute and subacute phases following stroke. The cellular basis of some of this plasticity is programmed into neural networks and progenitor/stem cells. What are the signals that trigger this response? What are the signals that terminate this response once the acute/subacute phase has passed? Can the expression of repair signals be prolonged such that the window for repair is left open longer? Can the window be re-opened in the chronic phase? What other signals might trigger or enhance repair, such as cytokines, chemokines, transcription factors, or signaling molecules?
The roles of neurogenesis, angiogenesis, and other proliferative responses (e.g., glial proliferation) in the recovery process following ischemia are still unclear. The molecular mechanisms that lead to these processes, as well as the role played by the new cells in behavioral recovery, need to be uncovered. The precursor cells that proliferate following ischemic injury need to be characterized and the molecules that control their growth and differentiation delineated. Is there a coupling between angiogenesis and neurogenesis? Further insight into the role of these processes in recovery will likely lead to novel targets for therapy.
Endogenous and Exogenous Factors
While it may be possible to enhance the effects of endogenous substances (growth factors, neurotransmitters, receptors, and others) as therapeutic approaches, there is still a need to identify which substances improve recovery and which impede recovery in experimental models. In addition, we need to characterize the effects of endogenous inflammatory cells (i.e., within to the central nervous system) and exogenous inflammatory cells on the recovery following stroke and how the role of each can be manipulated. Though the inflammatory response to ischemia has been studied, many issues related to the role of nonneuronal elements (microglia, macrophages, leukocytes, lymphocytes, etc.) after ischemia are still unresolved.
Increasingly, restorative approaches utilizing exogenous substances (e.g., growth factors, other small molecules, d-amphetamine) are being investigated. However, the mechanisms of action of these substances are only partially understood. In addition, behavioral experience after stroke (physiotherapy, environmental enrichment, etc.) is now known to play a substantial role in recovery. Again, the underlying mechanisms are still unclear. Finally, increasing evidence points to an interaction of behavioral experience and pharmacotherapy. What are the mechanisms by which experience can modulate the effects of drugs such as amphetamine?
Pre-Existing Factors Related to Recovery
Cellular, molecular, and network changes during recovery should be characterized as functions of degree and location of injury, degree of recovery, age, gender, race, stress, and environmental enrichment. The molecular processes that underlie such interactions consider other disciplines also exploring those issues.
Developmental Models
It is known that the plasticity of the newborn and neonatal brain is much greater than the adult brain. What is the molecular basis for limitations for recovery in adults? Understanding the factors that support plasticity in the developing brain may lead to the potential to activate or augment this process for therapeutic purposes. Is the process of development literally recapitulated at the molecular level, much as early 20th Century neurologists believed that recovery of the organism's behavior mimicked the development and loss of reflexive behavior? In other words, what lessons can developmental biology lend to recovery?
Network Processes
The likelihood that focal lesions of brain produced by focal ischemia, or isolated loss of neurons following global ischemia, lead to network disturbances and compensatory changes in excitatory and inhibitory connections needs to be explored at the molecular, cellular, electrophysiological, systems, and behavioral levels. Though individual cells are often studied in isolation, the changes in networks have only recently been approached. The roles of dividing cells, synaptic pruning and synapto-genesis, and changes in excitatory and inhibitory circuits need to be better defined in order to better delineate points at which therapies might be useful.
Definition of Functional Recovery
The differences between brain recovery and behavioral compensation have not been adequately defined or studied. There is a need to define these differences using clinical, imaging, and other parameters, as the two may arise on the basis of different brain events. This may aid in designing clinical trials, by more precisely defining the behavioral outcomes measures of interest.
Pre-Clinical and Clinical Models
Pre-Clinical Models
Animal models can provide information regarding the complex interactions of large numbers of neurons in central nervous system circuits. New ways to record and process the large sets of multidimensional data that can result from such models are needed; that might best be accomplished by incorporating scientists from fields such as mathematical modeling and statistics into the field of stroke recovery. There is also a need to develop animal models of sensory and cognitive deficits that can be used for performing better preclinical studies of functional recovery after stroke.
Clinical Models
Validated methods are needed to define and characterize stroke impairment and disability. Models need to consider clinically significant covariates such as the nature of disability, the location of injury, and the type of injury. Defining sources of population heterogeneity will be critical to designing and inter-preting studies of new therapeutics. Criteria must be developed for adequate testing of drugs, cells, and other therapies in animal models before proceeding to human trials (e.g., time window, dose-response, lesion type/ size/location, etc.).
Integration of Basic and Clinical Research
How can appropriate patient candidates be identified for eventual treatment based on animal studies? Markers are needed for in vivo observation and monitoring of key processes related to stroke recovery and the effects of drugs or endogenous/exogenous neural progenitors in animal and human studies.
Monitoring and Measuring Recovery
The natural history of recovery in humans and in animal models is incompletely understood; paradigms often cannot be compared between laboratories, and there is considerable evidence that the process of recovery differs depending on the location of injury and the functions that are affected. Advances in imaging technology will play a large role in better characterizing these events.
Molecular Neuroimaging
What can be learned from molecular neuroimaging studies of the pharmaco-kinetics of drugs and bioactive molecules (e.g., growth factors) used to enhance the restorative capacity of brain tissue? Can treatment effects on molecular mechanisms and synaptic/ enzymatic activity be demonstrated by functional imaging methods? Also, new imaging technology is needed to monitor neuronal plasticity and treatment effects at cellular and network levels of analysis. There is a need for high-throughput screening tools for evaluating therapies.
Analogous neurophysiological tools are needed in animal models. More animal models and studies, especially primate models and studies, are needed that can correlate functional MRI and other imaging parameters with electrophysio-logical and cellular and molecular outcomes following stroke.
Monitoring Network Processes
It will be important to identify the essential neuronal networks responsible for good recovery following stroke and to define, image, and monitor these networks using established and new imaging techniques (e.g., PET, SPECT, fMRI, MRS, TMS, MEG, EEG, etc.). The ability to monitor these networks and other image parameters with adequate spatial and temporal resolution will provide independent outcome measures of stroke recovery in clinical trials. How does the process of recovery differ in various regions and lesions? Why is there adequate redundancy for some functions and not others? What role do inter- and intrahemispheric interactions play in promoting, inhibiting, or supplementing specific functions and how can positive effects be encouraged? Such studies will require new approaches that include functional and molecular imaging and interventions that directly address causality. Such information may be critical to predicting and understanding treatment responses in this setting.
Treatment and Enhancing Recovery
There is a current lack of knowledge on patient selection for specific treatment protocols. Treatment approaches using growth factors, as well as progenitor/ stem cells and engineered cells hold promise, but require further study. There is increasing support for the independent and interactive effects of physiotherapy/ environmental enrichment and pharmacotherapy, but important gaps must be filled by further animal studies and controlled clinical trials. A viable mechanism for delivery of pharmaco-therapeutic treatments is still a problem. We need the ability to manipulate molecular events underlying the potential mechanisms. Appropriate preclinical models that systematically address lesion size, location, and recovery period are needed to help reduce the possibility of poor design in future clinical trials.
Interactive Effects of Physical Rehabilitation and Other Therapies
How does physical therapy (and other behavioral experiences, such as environmental enrichment) interact with new therapies such as trophic factors, promoting neurogenesis, and cellular transplantation? What effect, if any, does rehabilitation have on molecular events (e.g., distribution kinetics of labeled cells and molecules) that potentially influence behavioral outcome?
Clinical Trial Design
Treatment paradigms need to be evidence-based and hypothesis-driven and supplemented as much as possible by mechanistic experimentation, so that even failures will yield important new knowledge. A variety of potentially therapeutic treatment options to stimulate recovery already exist, such as enriched environment or performance of specific physical tasks, adrenergic stimulation, growth factors, stem cells, and engineered transplants. However, in the absence of a better understanding of the process of recovery, it is possible that trials addressing these approaches may fail as a result of improper timing, inadequate models, or inadequate outcome measurements in human trials.
Clinical trial design for stroke recovery treatments remains a conundrum that needs to resolved. How can we predict outcomes in order to properly identify patients appropriate for a new therapeutic intervention? How do we identify patients with comparable prognosis? How do we separately evaluate sensorimotor and cognitive recovery after stroke? What can be learned from controlled clinical trials of the efficacy of various physiotherapeutic and drug-supported treatments in the rehabilitation of stroke?
How can we manipulate molecular events underlying angiogenesis, neurogenesis, and neuronal remodeling? Are there pharmacological methods (e.g., ephrin and notch proteins) to induce neurogenesis, angiogenesis, or neuronal remodeling? Can cellular therapies such as endogenous inflammatory cells and exogenous cells (e.g., marrow stromal cell, stem cell, cord blood) act as growth factor "factories" that respond to the neurotrophic needs of the tissue?
Timing of recovery interventions may be critical. Many models equate acceleration of short-term recovery (the first few days after infarct) with long-term processes that are likely to be fundamentally different events. This may be a critical error if applied to patients for whom the experimental paradigms are not appropriate. Lesion size and location may be important factors that will restrict the range of applications. An approach that may be beneficial in one model (e.g., cortical infarct) may fail or even be detrimental in another model (e.g., white matter lacunar infarct).
New approaches for transport of large molecules across the blood-brain barrier are needed to deliver therapeutic agents (viral vectors, opening barrier, etc.). Approaches that are most promising for the development of pharmacotherapeutic tools for stroke recovery are still unclear. Modulators of neurotransmission and growth factors appear to be the leading candidates at present.
There is evidence that physiotherapy modulates the effects of pharmacological treatment for stroke recovery (e.g., amphetamine). Are there other methods that can enhance or target the effects of pharmacotherapy, such as electrical stimulation, psychological state, or environment?
Can cellular and gene therapies be combined with tissue engineering, including the use of biomaterials (e.g., biodegradable synthetic scaffolds that provide templates for exogenous or endogenous cell growth and/or secrete various molecules that might promote repair, neuroprotection, angiogenesis, or neurogenesis)? Can cellular therapies be used to create natural pumps or factories of therapeutic proteins beyond or in addition to replacement of degenerated cells? Cellular replacement approaches should recognize the importance of nonneuronal elements (e.g., astrocytes). Recovery of function may require reconstitution of the entire milieu.
Efforts to integrate and orchestrate multifaceted, multidisciplinary approaches over time should be encouraged (e.g., neuroprotection, cell replacement, molecular therapies, tissue engineering, neurite-outgrowth promotion, and remyelination).
BARRIERS
RESEARCH AND SCIENTIFIC PRIORITIES
Although the following three priorities will likely be advanced in concert, they represent a logical progression from (1) basic science insights into the underlying mechanisms of recovery, to (2) new neuroimaging techniques for monitoring these mechanisms in humans, to (3) new clinical interventions based on mechanistic targets. This strategy should yield improved therapies quickly utilizing approaches already identified as promising. More importantly, basing the development of both monitoring and treatment strategies on a solid foundation of basic science should result in new interventions that have yet to be proposed.
Priority 1:
Understand the molecular, cellular, and network changes in the brain that lead to good versus poor behavioral recovery following stroke.
Priority 2:
Develop neuroimaging and other methods for detecting molecular, cellular, synaptic, and circuit mechanisms of recovery following stroke that can predict outcome.
Priority 3:
Develop new clinical interventions based on mechanistic models of recovery.
RESOURCES NEEDED
We propose the establishment of multicenter networks for collaborative studies in animals and humans focused on single issues, including imaging. Centers require not only imaging and other tools, but also teams of scientists to address mechanisms of recovery and methods of measuring recovery, and to develop treatments to enhance recovery. The focus of these networks must be on translational approaches to stroke recovery, because of the strategy outlined in the priorities. Novel interventions will be based on underlying mechanisms of recovery. Specific therapeutic agents will likely have effects on specific targets that are involved in brain plasticity mechanisms. These interventions will then be tested in animal models, including primate models, to verify their specific effects and efficacy. Recovery progression in human stroke survivors will be closely monitored at molecular, cellular, and network levels using new neuroimaging techniques.
In addition, significant support is needed for the training of basic and clinical scientists in several areas: studying the mechanisms of recovery from stroke and brain injury, developing outcome measures, developing neuroimaging tools, and assessing therapies for improving stroke recovery. It has been difficult to recruit basic neuroscientists into the field of stroke, especially in the development of preclinical models. Other fields (engineering, computational modeling, etc.) have not been drawn to this area. A major effort must be made to lure bright, young scientists from these areas into an increasingly interdisciplinary field.
The scientific stroke community also recognizes the importance of maintaining a variety of animal models for preclinical assessment, including animals with complex brains. If we are to continue supporting non-human primate studies for stroke recovery, primate centers need to be responsive to the needs of researchers to obtain animals with required characteristics (e.g., B virus-free, particular age, sex) at a reasonable cost. Regional primate centers are no longer reliable resources, except for a host institution's investigators.
TopCo-Chairs: Pak H. Chan, Ph.D., and Giora Z. Feuerstein, M.D., M.Sc.
Participants:
Mark P. Goldberg
Neil Granger
John M. Hallenbeck
Chung Y. Hsu
Patricia D. Hurn
Michael A. Moskowitz
Gary A. Rosenberg
David M. Stern
Raymond A. Swanson
STATEMENT OF THE PROBLEM
Acute and chronic dysfunction of the neurovascular (NV) unit leads to cerebrovascular disease such as stroke and other cerebrovascular disorders.
CHALLENGES AND QUESTIONS
Brain Microvascular Endothelium
Glial Cells
Matrix and Matrix-Modulating Proteases
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:
Cerebrovascular endothelium: in order to understand the role of endothelial cells in the NV unit under both normal and disease conditions, elucidate the mechanism of endothelium signaling pathways that regulate survival or death as well as factors that support an antiinflammatory/antithrombotic phenotype.
Priority 2:
The role of glial cells in neurovascular function: in order to understand the role of glial cells in the operation of the NV unit in both normal and abnormal conditions, elucidate the mechanism of astrocyte and microglial growth and differentiation as well as interaction with the endothelium and matrix milieu.
Priority 3:
Matrix proteins and matrix-regulating proteases: in order to understand the role of the matrix in normal and disease conditions of the NV unit, elucidate the proteins and enzymes (MMPs, ADAMs) that compose and regulate the structure and function of the matrix.
RESOURCES NEEDED
Co-Chairs: Gregory J. del Zoppo, M.D., and Bryce Weir, M.D.
Participants:
Alastair M. Buchan
Bruce M. Coull
Jawed Fareed
James C. Grotta
John M. Hallenbeck
Julian T. Hoff
Eng H. Lo
James H. Morrissey
Bruce R. Ransom
Zaverio M. Ruggeri
Bradford S. Schwartz
STATEMENT OF THE PROBLEM
Stroke is a vascular disorder with neurological consequences. Vascular thrombosis is responsible for the majority of ischemic strokes. But loss of vascular integrity during focal cerebral ischemia is responsible for ischemia-related hemorrhagic transformation and contributes to spontaneous hemorrhage in the non-ischemic brain.
Successful treatments for ischemic stroke are limited to a few antithrombotic approaches, but unfortunately, these treatments carry a significant risk of hemorrhage. Such hemorrhage imparts significant injury, but there is no effective treatment or management strategy. Hemorrhage-related injury erodes benefit and has a negative impact on the outcome of stroke trials.
Downstream effects of the reduced blood flow that occurs during ischemic stroke target the microvasculature. The integrity of cerebral microvessels (which contain the endothelium, matrix, and astrocyte end-feet, as well as smooth muscle cells and pericytes) is important for normal vascular hemostasis and for preventing hemorrhage. Hemostatic factors and other enzymes are compartmentalized in the central nervous system (CNS). The increased incidence of intracerebral hemorrhage during thrombocytopenia suggests that normal platelet function is necessary for cerebrovascular integrity.
Although vascular hemostasis primarily involves the blood and the luminal aspect of the endothelium, it also involves the vascular compartment and the neuronal tissues. The cerebral blood vessel is the interface between the blood and neuronal cells. Very little is known about the interactions of the microvascular compartment and the neuronal/glial compartment under normal conditions and during ischemia; it is known, though, that vascular interventions very early in thrombotic stroke can limit the extent of ischemic injury. It is expected that increased understanding of these interactions can be applied to new therapies for ischemic stroke and hemorrhage.
Furthermore, little is known about the particular contributions of the brain microvasculature to normal hemostasis and their relation to abnormal neuron function. This includes a significant lack of information about how the endothelium and the glial compartments interact with each other and regulate vascular responses to ischemia. Families of proteases are generated from both the vascular and the nonvascular compartments during ischemia, which participate in tissue injury. These include hemostatic factors and thrombin, plasminogen activators, matrix proteases, and other enzymes. Selective protease activation in the microvasculature is essential for maintaining hemostasis, but also accompanies brain ischemia. The balance of these processes in the CNS is unknown. Antithrombotic agents (antiplatelet agents, anticoagulants, and plasminogen activators) significantly increase the risk of intracerebral hemorrhage, while limiting vascular thrombosis. Developing and properly testing protease inhibitors and antithrombotics that reduce the potential for hemorrhage are likely to improve patient outcome for acute interventions.
CHALLENGES AND QUESTIONS
A major challenge that needs to be addressed is our poor understanding of several factors: (1) the interaction between cerebral microvessels and the neuronal tissues they serve, (2) the role that vascular integrity plays in spontaneous hemorrhage and in hemorrhagic conversion of the cerebral infarct, and (3) the impact that inhibitors of thrombosis and of protease generation have on these processes. This requires an understanding of the relationships between normal cerebral endothelial cells and astrocytes in capillaries, their contributions to normal hemostasis, and their responses to ischemia and parenchymal hemorrhage.
These issues can be addressed by increasing our knowledge about hemostasis in the normal brain and under pathophysiologic conditions, with a focus on intravascular, vascular, and perivascular targets. Here, the relationship of hemostasis within the microvessel (which necessarily involves the vessel wall) to astrocytes and to brain cell functions must be considered. (Although not a focus of these considerations, it is understood that activation of inflammation also involves activation of coagulation and of platelets. Platelets and other cellular elements provide surfaces upon which proteases are generated.)
Current and future research priorities should take advantage of advances in parallel areas of cardiovascular research, particularly with regard to mechanisms of injury, tools for understanding those mechanisms, and therapeutic approaches. Smooth translation from in vitro work to animal models and human stroke is required. New specific targeted inhibitors of coagulation factors, their receptors, and platelets can be used as probes to examine their roles in microvascular integrity and hemorrhage. Overall, we need a significantly better understanding of the roles that the microvasculature plays in cerebral hemostasis and its responses to ischemia and hemorrhage.
BARRIERS
Interdisciplinary Barriers
Poor communication exists between researchers in vascular biology, hemostasis, and neurobiology. For instance, while more sophisticated antithrombotic agents are being used in clinical trials of ischemic stroke, there is little understanding of their effects on neuronal tissues and how to limit their contributions to hemorrhage. A multidisciplinary approach to understanding hemostasis in the CNS is likely to significantly increase the possibility of positive outcomes in clinical trials of stroke.
Translation of Animal Model Studies to Clinical Trials
Significant differences exist between rodents and primates/humans in terms of hemostasis, and there are also significant differences in the results of interventions in rodent models of focal ischemia and human ischemic stroke. Furthermore, there are major differences between rodents and humans in important receptor-ligand interactions in hemostasis. These factors also raise the concern that vascular-tissue interactions in the rodent brain may not translate to human brain. Although rodents account for the bulk of preclinical testing, beneficial outcomes observed in other animal models of human stroke also have failed to translate into clinical success thus far. Attention to the role of white matter injury is necessary. Furthermore, modeling of spontaneous intracerebral hemorrhage and of hemorrhagic transformation has been curtailed by a lack of understanding of the causes. This has limited approaches to the clinical treatment of these conditions. Finally, there are significant limitations in the availability of primates and novel rodent strains for experimental studies.
Technological Barriers
There is a discordance between experimental work on cerebral vessels at the molecular level and the ability to image only large cerebral vessels. The rational development of biochemical and molecular imaging tools for translation from model systems to human application has been largely ignored. Clinical imaging has focused on changes in anatomy and physiology. Significant advances in functional imaging (including fMRI, PET, and SPECT) now allow collection of data about damaged brain at the physiologic/organ level. However, fundamental knowledge of ischemic and hemorrhagic cerebral pathology is now being collected at the cellular and molecular level. Molecular imaging tools that span the range from model system to human brain are needed.
Clinical Trial Design
Despite the apparent success of antithrombotics in the treatment of ischemic stroke, the majority of clinical trials are negative or too small. Furthermore, preclinical testing and exploration of mechanisms in animal models is often inadequate. Proper use of antithrombotics and appropriate clinical trial designs are required to efficiently examine more novel agents in ischemic stroke. Specific antithrombotic agents can be used as probes. Strikingly, there have been similar limitations placed on our understanding of the mechanisms of antithrombotic-related hemorrhage, spontaneous hemorrhage, and hemorrhagic transformation. As most trials involving antithrombotics are sponsored by industry, early communication between scientists and clinical trial designers in the development of the projects is required. The absence of information regarding brain penetration of antithrombotics, related side effects, matching drug action to stroke subtype, and efforts to design preclinical and clinical trials with similar principles are of concern.
RESEARCH AND SCIENTIFIC PRIORITIES
These three research priorities are closely related in that each seeks to extend molecular studies to clinical application.
Priority 1:
Understand the normal biology of cerebral microvessels, as well as the interactions of cerebral microvascular endothelial cells, matrix, astrocytes, and other related cells in response to focal ischemia and parenchymal hemorrhage.
The temporal responses of endothelial cells and astrocytes to ischemia occur together, and in relation to neuron injury. Several factors imply a high level of interaction between these cells that may be extended to the neurons they serve. These factors include the interaction of endothelial cells and astrocytes to generate matrix during development, their use of the same signaling molecules (e.g., Ca++), and pores that allow entry of small molecules into the astrocyte compartment. These factors imply that the endothelial cell-astrocyte relationships may act as a unit. Subclinical ischemia may involve perturbations of these relationships. To better understand these interactions, we need to:
Priority 2:
Understand the requirements for hemostasis and platelet function within the normal CNS, and how they are perturbed by ischemia, parenchymal hemorrhage, mechanical interventional approaches, and antithrombotic agents.
Activated platelets and products of coagulation accumulate in ischemic microvessels, implying changes in normal endothelial cell function and loss of vascular integrity. Little is known about the processes that underlie normal hemostasis in the brain, their responses to ischemia, or their effects on astrocyte and neuron viability. Ischemia initiates the appearance of serine proteases, metalloproteinases, and other proteases within the vasculature and the brain tissue.
To understand this area better, we need to:
Priority 3:
Develop fundamentally new strategies to limit the impact of hemorrhage on brain function in the setting of focal cerebral ischemia and spontaneous hemorrhage.
Present treatment strategies for spontaneous hemorrhage and hemorrhagic conversion have had limited utility because of a poor understanding of (1) the pathobiology of the events, (2) the effects of hemorrhage on the brain parenchyma, and (3) the potential targets. Inadequate modeling has also contributed to this issue. Furthermore, while outcomes for modeling of ischemic stroke are reasonably well-defined, outcomes for the injury caused by hemorrhage are uncertain or diffuse. We need to:
RESOURCES NEEDED
Priority 1:
Understanding the complex biology of cerebral microvessels will require multidisciplinary efforts and the availability of high-quality reagents and models. Resources needed include:
Priority 2:
Appropriate in vitro and in vivo studies are required to evaluate the roles that the hemostatic (and vascular) systems play in the normal CNS and during ischemic and hemorrhagic injury. Resources needed include:
Priority 3:
The causes of brain hemorrhage are inadequately understood and there are no standard therapeutic approaches for spontaneous hemorrhage. Resources needed to address this issue include:
The co-chairs of this session thank Drs. Robert Rosenberg, Maiken Nedergaard Sidney Strickland, and Charles Esmon for additional contributions to this session's report.
TopCo-Chairs: Philip B. Gorelick, M.D., M.P.H., and Costantino Iadecola, M.D.
Participants:
Helen C. Chui
Frank M. Faraci
Steven M. Greenberg
Lennart Mucke
William J. Powers
Barbara Radziszewska
Kenneth J. Rockwood
Ingmar Skoog
Vascular causes of cognitive impairment are common, especially in the elderly. It is estimated that as many as one-third of those who suffer a stroke have post-stroke dementia. Vascular disease is considered to be a major contributor to slowly progressive dementia. Among persons at high risk of stroke, such as Asians, vascular dementia may be more prevalent than Alzheimer's Disease (AD).
As stroke and AD are both common in the elderly, many stroke patients may have concomitant clinical and neuropathologic changes of AD. In these cases it may be difficult to decide which changes caused or contributed to the dementia. Vascular brain disease, often clinically silent, may also be important for the clinical manifestations of dementia in individuals with mild Alzheimer encephalopathy. Differentiation of pure vascular cognitive impairment (VCI) from AD, or from mixed VCI/AD, is therefore problematic. In contrast to AD, the definition, clinical picture, course, risk factors, markers, preventatives, treatments, and vascular biology of VCI and mixed VCI/AD are not well established.
In this text we denote VCI as a heterogenous group of syndromes in which there is cognitive impairment with cerebrovascular disease (CVD). Some forms of CVD lead to cerebrovascular brain injury (CVBI) sufficient to be associated with cognitive impairment, whereas other forms of CVD and CVBI do not lead to cognitive impairment. Descriptive information is needed about the pathophysiology of VCI to establish a causal link between CVD and cognitive impairment.
VCI is a heterogeneous condition that can result in a wide range of clinical deficits and manifestations. VCI may manifest structurally as large- or small-vessel territory infarction, rarefaction of the white matter, Aß peptide deposition, brain hemorrhage, or a combination of these states. The spectrum of cognitive impairment may vary from isolated dysfunction in one or two cognitive domains to dysfunction in many domains. When research diagnostic criteria for vascular dementia are compared, there is substantial variation in prevalence, sensitivity, specificity, and inter-examiner reliability across the criteria. Isolation of a homogeneous, definable vascular dementia syndrome has proven difficult. Definition of VCI subtypes may be a more promising approach. Similarly, the heterogeneity of VCI provides conceptual and operational challenges in measuring the effectiveness of treatment.
Risk factors for vascular forms of cognitive impairment seem to be similar to those for cerebrovascular diseases (e.g., hypertension, diabetes mellitus, lipid disorders, atrial fibrillation), and now have also been linked to AD. Furthermore, the AD susceptibility gene, Apolipoprotein E (ApoE), may also be important in VCI, as this gene has also been linked to an increased risk for atherosclerosis and myocardial infarction. However, it is not known whether such similarity in risk factors and susceptibility genes reflects the difficulty in differentiating AD from mixed VCI/AD syndromes. Cerebral amyloid angiopathy (CAA) may lead to brain dysfunction and injury in both VCI and AD, by shared or different mechanisms. The apparent overlap of the risk factors for vascular dementia with those for stroke and AD has important implications for preventive strategies, but has not been studied in sufficient depth.
The cellular and molecular pathology of VCI has not been defined. An understanding of these factors is needed to provide insight into the mechanisms of these conditions. Such knowledge will heighten our ability to identify risk factors and biological markers, and to develop rationally based prevention and treatment strategies, which are non-extant to date. The study of the molecular pathogenesis of VCI is complicated by the heterogeneity of these conditions, their multifactorial pathogenesis, the complexity of the reaction of the brain tissue to vascular and neuronal injury, and the interaction between genetic and epigenetic factors in the resulting tissue damage. While considerable emphasis has been placed on the mechanisms of the damage produced by acute severe ischemia and static cognitive impairment, very little is known about the effects of chronic moderate ischemia on neurons and white matter, from the standpoint of molecular, cellular, network, and cognitive changes.
CHALLENGES AND QUESTIONSBARRIERS
Interdisciplinary Barriers
There is poor communication among clinical researchers and between clinical researchers and basic scientists. Although substantial expertise exists in these scientific realms, there is a paucity of cross-communication. Furthermore, most clinical and epidemiologic studies are being carried out in relative isolation, with disparate study methodologies and study populations. A large-scale, systematic and integrated approach including both clinical and basic arms for the study of vascular dementia is lacking.
Nosology Barriers
There are several major limitations in current nosology. First, current definitions and classifications of vascular dementia direct attention to the relatively late stages of cerebrovascular disease and vascular-related brain injury, and may not address the early or "brain-at-risk" stages of the disorder. Second, there is no fundamental agreement on the neuropsychologic parameters that best characterize VCI. Current diagnostic criteria, when compared, differ substantially in parameters such as sensitivity, specificity, and inter-examiner reliability. Furthermore, the criteria may be difficult for a general practitioner to apply in the community at-large. Third, when there is substantial overlap between two conditions, such as between AD and CVD, categorical approaches to nosology have limited utility. Alternative approaches to classification should be sought.
Technology Barriers
Suitable quantitative structural and functional neuroimaging tools are undergoing technology assessment to determine their role in the diagnosis and staging of VCI. However, their value in the study of vascular dementia has not been established, and they are not yet ready for use in the community at-large. In the basic science community, techniques for studying cerebrovascular function in transgenic mouse models are not as widely available as those for investigations in larger mammals. This is an obstacle to the study of the alterations in cerebrovascular function in mouse models relevant to VCI.
Disease Model Barriers
There is a lack of appropriate in vivo and in vitro models with which to study chronic ischemia and VCI. Rodents have relatively little white matter and may not be suitable models for the white matter rarefaction frequently seen in humans. Therefore, primate models may be needed for investigations involving white matter. Transgenic mouse models of AD and CADASIL are promising, but they do not address the heterogeneity and complexity of vascular dementias. Interaction with genetic and epigenetic factors, species differences, and difficulties with modeling human cognitive deficits in lower animals are also barriers.
Tissue Resources Barriers
There is a general lack of brain tissue from affected individuals for rigorous clinical-pathological correlation, for correlation with ante-mortem imaging, and for genetic and molecular investigations. There are no central resources for tissue collection, storage, and distribution to the scientific community. Modalities for tissue processing, preservation, and analysis most appropriate for vascular dementia syndromes have not been defined, and are likely to differ from those used for AD. Animal models relevant to vascular dementia and related risk factors are not always readily available, presenting another obstacle to extensive study by the scientific community.
RESEARCH AND SCIENTIFIC PRIORITIESPriority 1:
Develop diagnostic criteria or alternative classification schemes for CVBI and VCI subtypes, define stroke and cardiovascular disease risk factors and markers, and identify novel risk factors and markers for these conditions.
Priority 2:
Investigate the molecular pathology of VCI subtypes and mixed dementias, and develop preclinical treatment strategies.
Priority 3:
Develop and test prevention and treatment modalities in patients with VCI subtypes.
RESOURCES NEEDED
The co-chairs of this session thank Dr. Gustavo Roman for additional contributions to this session's report. Top
Co-Chairs: Eric Boerwinkle, Ph.D., and Joseph P. Broderick, M.D.
Participants:
Mark J. Alberts
Larry Atwood
Dorit Carmelli
Tatiana Foroud
John A. Hardy
Thomas P. Jacobs
Steven Kittner
James F. Meschia
STATEMENT OF THE PROBLEM
The identification of the genetic underpinnings of stroke has lagged substantially behind progress made in other common neurologic disorders such as Alzheimer's and Parkinson's diseases. Only a few genes have been linked to stroke, and these are found only in rare families. These include stroke phenotypes that involve primarily the cerebrovascular system and brain, such as CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; Notch 3 gene), familial cavernous hemangiomas, and hereditary intracerebral hemorrhage (ICH) secondary to amyloid angiopathy (e.g., Icelandic variant). In addition, other systemic diseases due to genetic mutations -- such as sickle-cell disease, polycystic kidney disease, MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, stroke-like episodes), Fabry's disease, Factor V Leiden deficiency, and other inherited coagulation disorders -- have been associated with ischemic and hemorrhagic stroke in some patients. For the overwhelming majority of patients with stroke, however, the genetic risk factors are unknown.
A few small linkage studies in Finnish and Japanese populations have identified chromosomal regions associated with intracranial aneurysms. One specific gene in the Icelandic population has been identified as a potential stroke candidate gene (unpublished data). Several candidate gene polymorphisms have also been investigated in case-control type studies. The most convincing evidence thus far for any candidate gene is the association between Apo E4 and E2 genotypes and lobar intracerebral hemorrhage.
Despite limited current knowledge regarding the genetics of stroke, identification of stroke genes represents the clearest path to a better understanding of the mechanisms underlying stroke. Once the genes have been identified, their effects can be examined in animal models. Identification of the relevant genes, the proteins that they code, and the function of these proteins will lead to new innovative strategies for primary and secondary prevention of stroke. Eventually, preventive or acute therapy may be chosen based upon genetic makeup.
CHALLENGES AND QUESTIONS/ BARRIERS
The identification of stroke genes is more complicated than is the identification of genes for other neurologic diseases. Stroke has a very high short-term mortality rate that makes recruitment of living relative pairs much more difficult than it is in the case of diseases with long survival times, such as Parkinson's and Alzheimer's. Because most strokes occur in the elderly, affected relatives are also more likely to have died from other co-morbid conditions.
Stroke and stroke subtypes have many known environmental risk factors, whereas diseases such as Parkinson's and Alzheimer's have only a few, inconsistently identified environmental risk factors at present. The many environmental factors associated with stroke may confound linkage analyses if they are not accounted for, and may complicate the analysis in general.
There are many mechanisms by which stroke occurs and definition of phenotypes is crucial to identification of susceptibility genes. Phenotypes include:
This phenotypic variability likely will be reflected in genetic heterogeneity as well; there is probably more than one gene underlying stroke. Thus, at the very least, genetic studies of stroke should be initially designed to address each of the major stroke subtypes separately: subarachnoid hemorrhage due to ruptured intracranial aneurysms (with inclusions of unruptured intracranial aneurysm), intracerebral hemorrhage, and ischemic stroke. A recent report from the company deCODE genetics on the Icelandic population, concerning the identification of a stroke gene related primarily to ischemic stroke and transient ischemic attacks, provides optimism about identifying susceptibility genes, even without further subtyping of ischemic stroke.
Because of the high mortality, late age of onset, and likely genetic heterogeneity of stroke patients, recruitment of sufficient numbers of affected relative pairs is critical to successful identification of susceptibility genes. Genetic studies of complex diseases such as stroke require particularly large numbers of affected individuals and need to be performed in genetically isolated populations as well as in more diverse populations. Other common diseases, such as hypertension and diabetes, have required sampling and genotyping of thousands of affected relatives, and the requirements for stroke are likely to be similar.
Once genes are identified in families where there is aggregation of stroke, the attributable risk (impact of the genetic mutation as a cause of stroke) within the population as a whole and within the population of stroke patients will need to be determined. This determination will require population-based samples of stroke patients and matched controls.
How specific genotypes may modify acute and preventative pharmacologic therapies for stroke will be an important future issue. The only reported relationship thus far is the association between an Apo E2 genotype and response to t-PA in acute ischemic stroke. While other gene-therapy interactions have yet to be identified, it will be critical to collect genetic samples in clinical trials to begin to address this issue.
Genotype is likely to be related not only to the mechanism of stroke but also to recovery following stroke. The collection of genetic samples in ongoing and future clinical trials will provide a central resource that will help address this issue in the future, even though current hypotheses about genotype and recovery are lacking.
RESEARCH AND SCIENTIFIC PRIORITIES
Our overall long-term goal is to identify genetic polymorphisms that are related to mechanisms of stroke occurrence, response to therapy, and recovery following stroke. To accomplish this goal, we recommend the following research priorities.
Priority 1:
Perform genetic studies of families in which stroke aggregates.
Individual studies are needed for each of the primary phenotypes of intracranial aneurysms (ruptured and unruptured), intracerebral hemorrhage, and ischemic stroke, since these major stroke subtypes are likely to have different susceptibility genes. Large sample sizes will be needed to adequately power these studies, regardless of study design. Ongoing and future case-control studies, and other study designs, will be critical complementary approaches to identification of susceptibility genes and their impact in the population at large.
Priority 2:
Standardize the methodology for genetic studies of stroke.
Clear and consistently used definitions of stroke phenotypes, standardized collection of environmental risk factor data for inclusion in data analysis, standardized methods of consent for participation in genetic studies of stroke, and standardized methods of obtaining and storing genetic samples from ongoing and future stroke studies, are needed. Such standardization will provide the essential data to allow future combinations of data from several studies for meta-analysis of the effect of possible stroke genes.
Priority 3:
Use information from existing clinical/genetic databases for related diseases (e.g., studies of hypertension and diabetes) and from ongoing clinical stroke trials for genetic studies of stroke, to increase the power to identify, test, and characterize susceptibility genes contributing to the risk for stroke.
There are very large ongoing cohort and epidemiologic/genetic studies of conditions (such as hypertension and diabetes) that are known risk factors for stroke. Stroke as an endpoint has been poorly explored in these databases and new studies and study designs would need to be designed to identify stroke within these populations. An easily accessible listing of such databases and information on how to gain access to the principal investigators of these studies will be critical for clinical/genetic researchers who wish to explore genes relevant to stroke occurrence in these populations.
Clinical trials of stroke should be encouraged to include collection of genetic samples. These samples can be used to explore the pharmacogenetics of response to therapy and adverse events.
RESOURCES NEEDED
Co-Chairs: David Greenberg, M.D., Ph.D., and Roger P. Simon, M.D.
Participants:
Stephen Barnes
Steven M. Greenberg
Barbara Handelin
Mary P. Stenzel-Poore
Katherine Woodbury-Harris
Justin A. Zivin
STATEMENT OF THE PROBLEM
Technological advances in genomics, proteomics, pharmacogenomics, and bioinformatics have the potential to advance research on the biology of stroke and to contribute to diagnosis and treatment. The best ways to apply these emerging fields to stroke are still uncertain, however.
Because the techniques involved are new, the cross-disciplinary expertise, collaboration, and training programs that will be required for them to benefit patients with stroke do not yet exist. To derive maximum clinical benefit from these new advances, it will be essential that:
The development of inflammatory changes that occur both in the central nervous system and in the periphery modify stroke outcome. At present there is little known about factors or predisposing physiology that influence the response to ischemia. There is also little known about the initial response to ischemia that influences the predilection for subsequent stroke.
CHALLENGES AND QUESTIONS
BARRIERS
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:
Facilitate basic science access to genomic, proteomic, and bioinformatics technology and develop methods of data analysis, interpretation, and standardization as they relate to stroke.
Priority 2:
Create a structure for defining stroke at a molecular and mechanistic level. This should include the creation of methods and protocols for the collection of biological samples for genomic, genetic, and proteomic studies linked to current and future National Institutes of Health (NIH) trials. With basic research as a template, identify molecular markers or profiles of stroke and stroke risk in patients to include gene profiling (functional genomics) and inflammatory/cellular mediators.
Specifically, the following parameters should be included in a re-evaluation of the definition of stroke:
In addition, researchers should:
Priority 3:
Recruit geneticists and molecular biologists into collaborations with stroke researchers (both clinical and basic science), with the objective of achieving near-term progress in capitalizing on current technologies. In parallel, enhance the recruitment of young and mid-career scientists into existing training programs.
Very few people have expertise in these new fields and also in stroke. Moreover, investigators in these fields and stroke researchers typically have limited contact with each other in their institutions, at meetings, and through the literature. However, this sort of combined expertise and interdisciplinary interaction will be critical if genomic, proteomic, and bioinformatic approaches are to be applied to stroke in a useful way.
High priority should be given to the establishment of:
RESOURCES NEEDED
Co-Chairs: John Detre, M.D. and William J. Powers, M.D.
Participants:
Joseph P. Broderick
Alastair M. Buchan
Michael Chopp
Andrew M. Demchuck
Konstantin-Alexander Hossman
Weili Lin
Michael E. Moseley
Steven Warach
Katherine Woodbury-Harris
Justin A. Zivin
STATEMENT OF THE PROBLEM/PROGRESS REVIEW
The field of neuroimaging has produced major advances in the diagnostic evaluation of cerebrovascular disease during the past 25 years.
Computerized Axial Tomography
The development of computerized axial tomography (CT) in the early 1970s revolutionized the diagnosis of acute stroke, primarily because of its ability to detect acute hemorrhage. Today, CT remains the standard diagnostic imaging procedure in this setting. Intracerebral hemorrhage size on CT has proven to be accurate for prognosis of ICH. Serial CT scanning showing progressive enlargement of intraparenchymal hemorrhage early after onset has identified an important cause of post-hospitalization deterioration. CT has also proven useful in the management of patients following ruptured aneurysms to detect hydrocephalus, re-bleeding, and cerebral infarction.
The accuracy of CT for detecting hemorrhage led to its use as a requirement for determining eligibility for treatment with thrombolytic agents, currently the only approach with demonstrated efficacy in the therapy of acute stroke. Delayed CT also permitted the visualization of the precise area of the brain that was damaged by cerebral infarction. This has, in many cases, improved differentiation of anterior circulation infarcts from posterior circulation infarcts, permitting more appropriate use of carotid endarterectomy for secondary prevention. Recent attention to CT in the hyperacute period has demonstrated its sensitivity for detecting early signs of cerebral infarction within six hours of onset. However, even with the accepted widespread use of CT for stroke, solid data demonstrating that the use of CT improves patient outcome is lacking in most instances.
Magnetic Resonance Imaging
Proton magnetic resonance imaging (MRI) is noninvasive and provides improved resolution and tissue contrast for detecting ischemic changes, vascular anomalies, and evidence of prior hemorrhage. MRI is widely used in the clinical evaluation of transient ischemic attacks and stroke, and has shown increased sensitivity over CT for detecting small and posterior fossa infarcts. Its sensitivity in detecting acute intracranial blood remains unknown, primarily due to a lack of studies in this area. The development of diffusion-weighted MR imaging and its quantitative correlate of regional measurement of the apparent diffusion coefficient (ADC) has made it possible to visualize cerebral ischemia very early after onset, before CT- and other MRI-detectable changes occur. While much still needs to be learned about the biological basis of this signal, clinical practice is defining its sensitivity, specificity, and temporal evolution for diagnostic purposes. Yet, studies prospectively demonstrating that the increased sensitivity of MRI for early and small lesions actually improves patient outcome or reduces costs remain to be carried out.
Arteriography
Catheter arteriography remains the gold standard for defining the anatomy of intracranial and extracranial blood vessels. The pathophysiological relevance of arteriography has been demonstrated empirically by the close correlation between the degree of arteriographic carotid stenosis and the subsequent risk of stroke in medically treated patients. Arteriography has also been extensively used for the diagnosis of and surgical treatment planning for aneurysms and arteriovenous malformations and for the diagnosis of vasospasm after subarachnoid hemorrhage.
Noninvasive Visualization of Vessels
Due to the invasive nature of arteriography and the small, but real complication rate, a great deal of effort has been devoted to the development of alternative techniques to visualize the vasculature. These alternative methods also have the potential to provide more rapid images, with the possibility of use at the bedside. Carotid duplex Doppler ultrasound has been developed into a useful screening tool of the extracranial carotid arteries for severe stenosis and can be used to study the characteristics of the carotid plaque. However, variations in machine and operator characteristics make it necessary to perform validation studies versus arteriography in each individual laboratory. Transcranial Doppler ultrasound of the intracranial vessels is also widely used to detect intracranial stenosis and occlusion, though its accuracy and clinical utility remain to be fully established. Flow-sensitive MR techniques also provide information about the cerebral and cervical blood vessels and have found widespread application, including screening for aneurysms. Use of rapid-sequence CT following intravenous bolus injection of X-ray contrast agents also has recently been used to study the cerebral and cervical vessels. The accuracy and clinical utility of MR and CT angiography versus catheter angiography remains to be formally established by appropriate studies.
Physiological Imaging
The development of positron emission tomography (PET) in the mid-1970s launched the modern era of physiological imaging in cerebrovascular disease. Physiological neuroimaging of cerebrovascular pathophysiology in humans provides a critical translational link to laboratory research. The relevance of information derived from cellular and animal models to human disease can be determined and, vice versa, human pathophysiological data can be used to design more appropriate experimental systems. PET remains the standard for quantitative, accurate, regional measurements of cerebral blood flow and metabolism. Its complexity and expense have thus far precluded widespread general use in cerebrovascular disease, but this may change with the growth of PET facilities for oncology imaging. PET has provided important new insights into the pathophysiology of both acute and chronic cerebrovascular disease. MRI and CT approaches have also been developed for examining cerebral blood flow and related hemodynamic parameters, and are much more widely available than PET, though their accuracy has not been conclusively established.
Imaging of Acute Stroke
The development of clinically practical physiological imaging in acute ischemic stroke has led to a great deal of interest in trying to identify brain tissue that is viable early on but will go on to die if left alone. Accurate identification of this physiological "ischemic penumbra" would be extremely valuable because it would facilitate clinical trials and allow clinical care of acute stroke to be targeted to those patients who can benefit from treatments designed to salvage tissue. A variety of measurements, including cerebral blood flow and nonquantitative perfusion alone or in combination with other tissue signatures, have been proposed for this purpose. However, few validation studies have been conducted to demonstrate that these techniques can accurately identify viable but doomed tissue.
Methods for portable imaging of physiological change, primarily using optical imaging methods, are also under current development. While these approaches afford the promise of providing diagnostic information at the bedside or in the ambulance, significant technical and methodological challenges remain to be met before they can be considered for clinical application.
Recent advances in understanding the cellular and molecular changes that accompany ischemic events in animal models have provided many new insights into the mechanisms of ischemic injury. Concomitant with this is a need to develop imaging methods that will allow these processes to be visualized in human patients. Because of its high sensitivity, PET imaging with radionuclide tracers is most likely to meet this need, though efforts to link paramagnetic contrast agents to molecular markers for use with MRI are also underway.
Imaging of Chronic Cerebrovascular Disease
While much of the interest in brain imaging in cerebrovascular disease has focused on its use in the evaluation and management of patients presenting with acute stroke, there are numerous opportunities for imaging methods to contribute to the evaluation and management of chronic cerebrovascular disease. Applications include optimization of pharmacological therapy, stratification of patients for prophylactic procedures such as extracranial-intracranial bypass surgery, and the use of sequential imaging to better define the natural history of cerebrovascular disorders.
Recovery of function from stroke is of paramount importance to patients and families, yet the mechanisms of recovery remain poorly characterized, and it is unclear which interventions influence outcome. Functional imaging methods could potentially assist in this issue by providing direct visualization of brain function, even in the absence of overt behavioral manifestations, though the validity of imaging markers for neural activity in the setting of cerebrovascular disease also needs to be validated.
CHALLENGES AND QUESTIONS
In the next decade, neuroimaging will advance the treatment of cerebrovascular disease in two important ways.
First, neuroimaging of cerebrovascular pathophysiology in humans will provide a critical translational link to laboratory research. Direct translation of the results of mechanistic and preclinical pharmacological studies in experimental systems to therapeutic interventions in humans have not met with much success. While there are many potential reasons for this, the differences in pathophysiology between human and animal model systems undoubtedly contribute. Neuroimaging provides the methodology to directly test whether specific mechanisms identified in animal systems are important in humans. Similarly, a better understanding of human pathophysiology can lead to the development of improved animal models. Furthermore, studies of drug delivery and of the biological effects of different treatment interventions will be of value to screen potential therapies for efficacy and to determine the proper parameters for dose, duration, and time window. This approach will require not only the use of currently available technology but also the development of new methods to image molecular, cellular, and functional processes.
Second, neuroimaging will provide predictive information to stratify clinically similar patients into different outcome groups. Since neuroimaging provides both physiological and structural information about cerebrovascular disease over and above that available from conventional risk factors and the clinical examination,it can be used to further improve the prediction of outcome for both acute and chronic cerebrovascular disease. Such outcome measures may initially include anatomic endpoints such as reduction of infarct size, but must eventually rely on clinically relevant outcomes such as functional status or subsequent stroke. This predictive information can be used in two important ways to advance clinical treatment trials. First, selection of patients based on physiological or structural criteria can identify groups of patients most likely to benefit from a given treatment. Second, if a predictive factor is demonstrated to be reliably and precisely related to an important clinical endpoint, it can be used to stratify enrollment or, in the best of circumstances, as a surrogate marker in Phase 2 or Phase 3 trials.
Data obtained in individual cases or small series demonstrate the feasibility of all of these approaches. For example: (1) neuroimaging of ischemic stroke has demonstrated that the evolution of damage takes place over hours, not minutes, (2) neuroimaging of intracerebral hemorrhage greater than six hours old has failed to document any evidence of ischemia, (3) imaging of the carotid arteries is necessary for selection of patients for carotid endarterectomy, and (4) PET measurement of oxygen extraction fraction determines stroke risk in patients with symptomatic carotid occlusion and has provided a basis for an intervention trial.
The challenge is to further implement these approaches more widely.
More extensive studies are needed to define the pathophysiology of human cerebrovascular disease in all ages, encompassing acute and chronic ischemia, atherosclerosis, hemorrhage, and recovery in populations. The value of neuroimaging to predict outcome under a variety of different conditions must be validated by rigorous studies, most likely requiring prospective multicenter trials. This validation approach, while accepted for therapeutics, represents a newdirection in the evaluation of imaging.
BARRIERS
The following barriers to achieving these advances were identified:
Priority 1:
Define the relevant pathophysiologic mechanisms in human cerebrovascular disease as determined by molecular and functional neuroimaging, and integrate this knowledge with appropriate data from experimental systems to develop new therapeutic approaches.
This priority employs neuroimaging as the critical translational link between laboratory experiments and human disease. It employs existing methodology and supports the development of new methods for elucidating the pathophysiology of human disease and testing specific mechanistic hypotheses derived from animal model systems for their relevance in human disease. This approach uses neuroimaging as a scientific measurement to study pathophysiology and thus requires careful validation studies for accuracy and a clear understanding of the biological basis of each neuroimaging modality and measurement. The goal is to integrate this information as a means to develop new treatment strategies.
Priority 2:
Identify and prospectively validate neuroimaging markers of tissue injury for prediction of clinical outcome in large patient samples.
This priority refers to existing markers of vascular, physiological, and functional derangement as well as the development of new markers of function and injury. The application is to all aspects of acute and chronic cerebrovascular disease, including pre-symptomatic vascular disease, ischemic stroke, intracerebral hemorrhage, and aneurysmal subarachnoid hemorrhage. This approach uses neuroimaging for clinical prediction and thus requires well-designed studies free of bias that provide empiric proof of accuracy in the clinical setting in which it will be applied.
Priority 3:
Identify neuroimaging markers of potentially salvageable brain tissue in acute stroke.
This priority refers to the application of existing or novel neuroimaging measures of vascular and/or parenchymal function to determine the potential for response to therapy. Pursuit of this priority will further refine the concept of the "ischemic penumbra" and will seek to expand the treatment window for patients presenting with symptoms of acute stroke.
RESOURCES NEEDED
EXHIBIT I
Specific Areas for Future Research
Co-Chairs: George Howard, Dr.P.H., and Philip A. Wolf, M.D.
Participants:
Larry Atwood
Larry B. Goldstein
Edgar J. Kenton III
Steven Kittner
Nancy E. Mayo
James F. Meschia
Lewis B. Morgenstern
Ralph L. Sacco
Katherine Woodbury-Harris
STATEMENT OF THE PROBLEM
Stroke is the third leading cause of death in the United States today but current understanding of its etiology and variations (among groups and over time) is insufficient to provide the foundation needed for effective strategies to reduce stroke mortality and morbidity in the foreseeable future. Stroke remains a challenging disease to address with epidemiological methods for several reasons, including:
Lack of Data
While there are selected exceptions (the Framingham Heart Study, the Rochester, Minnesota Epidemiology Project, the Northern
Manhattan Stroke Study, the Atherosclerosis Risk in Communities study, the Cardiovascular Health Study, and others), further
advances in stroke epidemiology will require additional detailed population-based data on stroke risk and outcome.
Particular issues include:
Lack of Accurate Subtype-Specific Data
As noted above, the processes underlying stroke differ by disease subtype, and specific studies may be needed within broad subtypes such as infarction or hemorrhage. This implies a need for sufficient resources to draw meaningful inferences within subtype, and the need to develop classification procedures that are not defined by the risk factors for the disease under study.
Poorly Defined Racial/Ethnic and Socioeconomic Factors
Race/ethnicity is substantially confounded with socioeconomic status and cultural factors, and attempts to provide adjustments require the quantification of socioeconomic status beyond the education/income/occupation measures currently employed. Cultural contexts needed to understand differences by race/ethnicity and socioeconomic status include measures of discrimination, stress, acculturation, language barriers, differences in access to healthcare, and others.
Poor Understanding of the Epidemiology of Outcome Following Stroke
Improving our understanding of the prevention of secondary or subsequent stroke events among those with prevalent disease is necessary. The risk factors for subsequent stroke may differ substantially from the risk factors for the first stroke, and a deeper understanding of these factors is critical to preventing such subsequent events.
In addition to subsequent stroke events, epidemiology of outcome as characterized by cognitive and functional disability following stroke is lacking. The identification of factors associated with positive functioning outcomes is the first step of interventions that may maintain or improve the life of the stroke patient, as well as caregivers, following stroke events.
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:
Characterize the public health burden of stroke and establish subpopulations for special emphasis.
The foundation approach to characterizing the public health burden of stroke is to develop a national, population-based surveillance system to establish incidence rates for stroke, both overall and by stroke subtype. This surveillance system needs to be prospectively designed to provide detailed information on incidence rates, with strata defined by age, geographic region, and race/ethnicity. These strata-specific incidence estimates will serve as the foundation to estimate the proportion of the well-known differences in stroke mortality rates between these strata that are attributable to differences in incidence (rather than case fatality, which will be addressed below), as well as the case-mix of stroke subtypes between these strata. Of equal importance, this resource will provide the mechanism to prospectively track temporal changes in incidence, allowing for appropriate shifts in resources and research efforts in response to changes in the incidence rates (both overall and by stroke subtype), as well as shifts between the strata defined by demographic factors.
A well-designed surveillance system naturally results in a sizable cohort of stroke patients. The follow-up of this cohort will serve the secondary aim of establishing the magnitude and determinants of the public health burden associated with the post-stroke period. Specifically, the cohort can be used to estimate the costs associated with long-term treatment, and the mortality, recurrence, and morbidity (including both recovery and cognitive decline) associated with the stroke event. For each of these domains, the cohort can also be used to establish the determinants that place stroke patients at risk for differential outcomes.
Finally, links between the national surveillance cohort and administrative databases can be investigated, to provide determinants of cost-effectiveness and patterns of resource utilization on the national level.
Priority 2:
Establish the new determinants of stroke and its consequences and identify subgroups with varying risk.
The current understanding of the determinants of stroke and its consequences is only partially successful in providing prognostic information on stroke incidence, mortality, morbidity (including recovery and cognitive decline), recurrence, and subclinical disease (including silent cerebral infarction). For each of these domains, additional case/control and cohort studies are needed to refine the information and understand the prognostic factors. Issues that need to be addressed include:f
Each issue needs to be addressed overall as well as within strata defined by ethnicity, gender, stroke subtype, age, geographic
region, and socioeconomic status.
Priority 3:
Integrate epidemiology into clinical management and prevention.
The hypotheses addressed in epidemiological studies need to be designed so that the results collected are useful to the anticipated users of this information. Specifically, studies should be designed to meet the information needs of investigators from the fields of policy development, randomized clinical trials and primary prevention, systems development, and the lay public and healthcare workers at all levels. In addition, hypotheses should be developed to address issues raised in related fields, including cardiovascular disease, peripheral vascular disease, subclinical disease, and brain injury.
Finally, there is a need to establish systems for data sharing by other investigators and clinicians.
RESOURCES NEEDED
Co-Chairs: Karen L. Furie, M.D., M.P.H., and Ralph L. Sacco, M.D., M.S.
Participants:
Joseph P. Broderick
Larry B. Goldstein
Philip B. Gorelick
Jonathan L. Halperin
Robert G. Hart
George Howard
S. Claiborne Johnston
Walter N. Kernan
Barbara Radziszewska
Rose Marie Robertson
Don B. Smith
Philip A. Wolf
STATEMENT OF THE PROBLEM
Epidemiological studies carried out over the last decade have clearly identified factors associated with increased risk of ischemic stroke and intracerebral hemorrhage. In addition, clinical trials have identified interventions that can reduce initial and recurrent stroke risk. Both vascular risk factors and stroke-specific factors, such as atrial fibrillation and carotid stenosis, have been demonstrated to be prevalent and modifiable. However, recent work has shown that a significant proportion of stroke-prone individuals do not receive appropriate therapy, despite dissemination of evidence-based recommendations. This is likely due to a combination of factors, including failure in patient compliance, lack of adherence to evidence-based guidelines by healthcare providers, and the inability of the healthcare system to provide adequate resources.
Additionally, while the prevalence of stroke risk factors varies across populations in the United States, the greatest burden of risk factors is borne by an underserved segment of the population. Thus, those at greatest risk are the least likely to benefit from recent advances in initial and recurrent stroke prevention.
The current issues in initial and recurrent stroke prevention focus on the following points:
Effective risk factor modification offers great potential to significantly reduce stroke incidence and recurrence. An understanding of why proven therapies are not being used, especially in vulnerable populations, is essential for designing and testing implementation strategies. The current challenge is to take what has been learned through cardiovascular prevention research and incorporate it into stroke prevention interventions. The efficacy of these strategies to modify a risk factor, affect an intermediate endpoint, or reduce stroke incidence can then be tested.
CHALLENGES AND QUESTIONS/ BARRIERS
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:
Improve implementation of existing, proven stroke prevention guidelines by (1) identifying barriers to such implementation by assessing the individual, healthcare providers, and the healthcare system, (2) studying methods of overcoming these barriers, and (3) supporting the development of research evaluating the effectiveness of innovative initial and recurrent stroke prevention interventions, especially in underserved populations and minority racial/ethnic groups.
Appropriate management (behavior modification, lifestyle changes, management of vascular risk factors, and medical/surgical therapy) may reduce initial and recurrent stroke risk and improve both the length and quality of life. Despite proven benefit, however, patients commonly fail to achieve accepted goals for risk factor modification before and after stroke. The reasons for failure to implement guidelines need to be explored in a variety of healthcare settings. One issue may be the dissociation between population and individual benefit. Special considerations may need to be given to younger populations and to individuals without traditional risk factors.
Once barriers have been identified at the three points in the prevention paradigm (individual, healthcare provider, and healthcare system), studies should evaluate novel interventions to improve initial and recurrent stroke prevention. Interventions aimed at the individual may include education, behavior modification, attention to health state preferences, use of support groups, empowerment, and incentives. Healthcare providers may benefit from education, clinical informatics, and incentive programs. Modification of healthcare systems may include changing access patterns, implementing quality assurance programs, and restructuring patient care reimbursement. In the design of these studies, it is essential to measure the clinical efficacy and cost-effectiveness in a variety of healthcare settings, especially in underserved and minority racial/ethnic groups. The use of intermediate endpoints in the design of pilot studies may provide justification to proceed with larger confirmatory clinical trials.
Priority 2:
Develop and examine the effectiveness of quantitative risk factor assessment tools that can identify stroke-prone individuals who need aggressive risk factor management and initial and recurrent stroke prevention intervention, with particular emphasis on underserved populations and minority racial/ethnic groups.
Identifying and treating those at highest risk of developing cerebrovascular disease remains a major challenge and a public health imperative. The use of risk assessment tools by healthcare providers and individuals can contribute greatlyby identifying even low-risk persons likely to benefit from intervention. Risk assessment tools that include innovative markers of risk can improve upon conventional models. These assessment tools must be flexible, in order to incorporate new risk factors and to allow for calculation of risk in the case of missing data (incremental models). The development of these tools must take into consideration issues of risk-factor measurement (such as discrete vs. continuous variables), as well as validation in multiple racial/ethnic groups.
Priority 3:
Support research designed to identify and evaluate innovative stroke prevention treatments and strategies.
There is great promise that future epidemiological studies will identify new stroke risk factors. Pharmacological research could help identify new agents capable of modifying these factors, and subsequent randomized clinical trials could then evaluate the efficacy of these new prevention treatments or strategies. For example, recent evidence that HMG-CoA reductase inhibitors and angiotensin-converting enzyme-inhibitors play multifunctional roles in preventing stroke has increased the number of potential medical therapies. Next-generation antiplatelet agents and thrombin inhibitors may similarly expand the role of antithrombotic therapy in first and recurrent stroke prevention. New antiinflammatory agents may help reduce the risk of stroke and retard atherosclerosis progression. In addition, there may be specific dietary and lifestyle practices that will prove to help reduce the risk of stroke. These and other as-yet-unproven therapies need further investigation through randomized clinical trials. Further successful results could lead to important modifications of future comprehensive stroke prevention recommendations.
RESOURCES NEEDED
Co-Chairs: Thomas G. Brott, M.D., and Patrick D. Lyden, M.D.
Participants:
Jeffrey I. Frank
Anthony J. Furlan
James C. Grotta
James D. Gustafson
Randall T. Higashida
Chelsea Kidwell
Thomas G. Kwiatkowski
Rose Marie Robertson
Judith A. Spilker
Marc K. Walton
The problems associated with acute stroke care can be divided into three related and sometimes overlapping areas:
Acute Care Infrastructure
Problems with acute care infrastructure have been identified at a number of different levels, from the number of individuals who are treated, the time it takes for these individuals to be treated, and the experience of the teams that respond to their clinical situation.
One option for improving the overall care of individuals with stroke is to center this care around community hospitals, with support and coordination for tertiary centers designated as comprehensive stroke centers. People with symptoms of stroke typically present at their community hospital first; it is unlikely that acute stroke patients would all present to a single, tertiary stroke center. For this reason, health care systems should focus on community-based infrastructure. In terms of the time it takes for people to respond to the symptoms of stroke, continued public education efforts will be critical. NINDS and voluntary health agencies have already provided support for these programs, but it is important that future educational efforts urge the public to treat the symptoms of an acute stroke as an emergency situation, and educate emergency personnel to deliver these individuals to the best, closest facility. The quality of care currently varies among community hospitals. A primary stroke center should include a round-the-clock stroke team, the ability to triage patients appropriately, and the availability of immediate CT scanning -- services not offered at many community facilities. Too many hospitals are currently accepting stroke patients without adequate preparation or resources. For these reasons, the designation of stroke centers by the Brain Attack Coalition could be very helpful in this regard, however further translational research will be needed to ensure that the most appropriate methods are used for designating stroke centers, monitoring their performance, and improving their patient care
Stroke Center Designation
Many groups consider the designation of stroke centers as an appropriate solution to the problem of hospitals that are not sufficiently prepared to manage stroke patients. A system for this designation has been considered to be a high priority, and as a result, the Brain Attack Coalition published guidelines for primary stroke centers in June 2000.
However, the official authorization of stroke centers is still a subject of ongoing discussion. Who should be in charge of this authorization? NINDS and/or a professional society are two options, though it will likely take a multidisciplinary effort to ensure that appropriate guidelines are implemented.
It is believed that the designation of stroke centers would be valuable to the field, since it is already known that the prompt use of thrombolytics improves outcomes, and stroke centers that are well-equipped, well-staffed, and prepared are able to deliver these drugs more quickly. However, considerable translational research will still be needed in order to demonstrate the long-term benefits of stroke centers on patient outcomes. As part of this effort, centers will need support for the collection of appropriate outcome and process data.
Research Infrastructure
At present, stroke research is also hindered by a lack of research infrastructure that will enable the translation of basic science findings to clinical phases of development. The development of centers or programs that could conduct studies and provide a framework within which collaborations could form has very broad support. Research centers could also contribute to the development and testing of therapies.
The identification of an entity/entities that would be responsible for the administrative functions of a stroke research center is another unresolved issue. Commercially-sponsored translational research typically focuses on a single, potentially profitable agent. By contrast, the type of translational research infrastructure that is needed would enable hypothesis-driven research to be conducted. In an appropriate framework, the findings from these studies could be translated into projects with significant clinical relevance.
In terms of funding for translational research, it is unlikely that this type of infrastructure would be funded privately. Research progress in acute stroke is likely to be incremental, and pharmaceutical companies are often skeptical that treatments for acute stroke will be profitable. For these reasons, they may hesitate to devote resources to these specific research needs.
Cooperative Clinical Trials
The concept of cooperative clinical trials has emerged from research centers in the field of cancer research, in which the collaboration of several centers has been necessary to investigate rare forms of cancer. Individual institutions that treat stroke patients are in a similar situation - they may not see sufficient numbers of patients to support a clinical trial, in some cases because the affected individuals are not reaching the hospital in the necessary time frame to enroll in a study. Cooperative clinical trials in stroke could help centers to overcome this obstacle, and they would offer the additional benefit of bringing good researchers, and their ideas, together.
Device Trials
In terms of the testing of devices, we believe that a paradigm shift in how stroke treatment and prevention research is conducted is needed. As an example, high-risk patients may be referred for several different procedures that involve devices: carotid stenting, intracranial aneurysm therapy, and endovascular coiling. However, comparison studies involving these approaches (e.g. surgery vs. endovascular coiling) have not yet been conducted.
In addition, clinical researchers have proposed a number of additional medical devices for use in ischemic stroke therapy. These devices typically remove clots mechanically or deliver therapeutic hypothermia. The device industry traditionally has brought such products to market under regulations that are significantly different, and less stringent, than those that govern pharmaceutical products. However, with the increasing need for additional Phase 1 and 2 studies, small companies with limited resources are beginning to look for academic collaborators who can organize and manage such trials. These collaborations can complement the needs of academic researchers, who often find it difficult to organize such trials alone, because they are not sufficiently familiar with the device field, device regulations, and with the related branch at the Food and Drug Administration (FDA).
Diversity Issues
The study of diversity issues as they relate to acute stroke treatment is also an important priority, since the Centers for Disease Control and Prevention have demonstrated that healthcare outcomes are disparate across minority groups and geographic regions. Since we do not have a complete understanding of why these disparities occur, more research in this area will be needed.
Therapeutic Innovations
Diagnostic Issues
Another significant issue related to stroke therapy is the amount of time and extent of effort required to make a reasonable and confident diagnosis of stroke. Issues of timing are very critical in caring for stroke patients, yet clinicians do not have a diagnostic tool for stroke that works well and is rapidly available. This problem is compounded by the fact that the presentation of stroke is often clinically complex. Better diagnostic tools, and designated stroke centers could both help to facilitate difficult diagnoses.
Combination Therapy
Another issue facing clinicians and researchers is the number of single treatment agents that have not performed well in clinical trials. The results to date indicate that the complex cascade of events following brain ischemia may not be easily interrupted by a single agent. For this reason, the concept of combination therapy, which has improved cancer treatment, should also be explored for stroke.
Although the testing of treatment agents in combination may be a reasonable approach, these combinations have not been studied in humans and have been seldom studied in animals. Many factors have influenced the evolution of these studies: traditional factorial designs require prohibitively large numbers of patients, involvement of pharmaceutical companies in these trials may be limited because of competition issues, and studies based on empirical evidence have encountered difficulties in the review process. Successes in cancer chemotherapy suggest, however, that combination therapy will almost certainly be necessary to develop effective treatments for stroke. Enhanced funding for trials of combination therapy, and improvements in the capability of centers to conduct trials of hyperacute therapy are both critical needs. Incorporation of thrombolytics and potentially synergistic therapies should be considered a high priority as these studies move forward. For these reasons, NINDS should make the evaluation of combination chemotherapy for hyperacute stroke a key priority over the next five years.
Endpoints
As treatments are considered and improved, it will be important to address quality of life endpoints. Often, researchers and clinicians consider the treatment of stroke as an acute issue, but endpoints are often measured several months after initial treatment has taken place. A 90-day endpoint, for example, would be affected by interventions used acutely, but it will also be affected by follow-up treatment that was performed in the weeks and months following the stroke. Thus, reconsideration of time points may be needed such that researchers can maximize the number of endpoints that can be analyzed, with the least impact from co-morbid conditions.
BARRIERSAcute Care Barriers
Reimbursement
Reimbursement for acute stroke care is a critical issue, and it will likely impact the implementation of new research findings in the future. Although thrombolytic drugs may be covered by insurance, insurers do not always appreciate the complex medical management that is required to administer these drugs. However, more recently they have begun to recognize the considerable cost savings that can be achieved with the application of thrombolytics.
Infrastructure
As described previously, an infrastructure for acute stroke treatment is needed -- in particular, one that is based on community hospitals and community response teams. More facilities are needed that can respond quickly and effectively when individuals present with symptoms of an acute stroke, and these approaches should also be cost-effective for the centers. Caring for most stroke patients in an intensive care unit, for example, is not likely to be productive for most patients, or cost-effective for the hospital.
Professional Staff
Another critical need in the field of acute stroke research, and in stroke research in general, is the expansion of the scientific and clinical workforce. Both fundamental pre-clinical and translational research will be required to address the problems outlined in this report. Traditional training mechanisms are available to the young investigator, but no programs specifically develop stroke researchers. In an effort to encourage professionals to enter the field of stroke research, NINDS has created loan repayment programs. However, these programs may not influence an adequate number of people to select a career in stroke. NINDS could further enhance these efforts by fostering a program that recruits and trains clinical investigators specifically focused on stroke. Once an interest in stroke research has been stimulated, such young investigators will transition to more traditional training programs and grants. As an addition option, stroke centers could also contain a professional development component to expand the workforce.
In addition to the problem of recruitment and training of new investigators, additional training of neurologists is also needed. Exposure to developments in stroke therapeutics, especially in emergency care, endovascular methods, and critical care, would enhance the ability of neurologists to treat individuals who present with a possible stroke. Similarly, clinicians who are not neurologists may also be caring for stroke patients. These emergency medicine and primary care physicians would also benefit from training in the proper recognition and management of stroke. This information should be incorporated as early as possible into the medical education system.
Research Barriers
An efficient system for the introduction of safe and effective devices for treating stroke is a high priority. In the past, companies that develop these devices have hoped for a rapid approval of their products, in order to recoup their research investments as quickly as possible. Traditionally, safety standards for these devices were considered to be higher than were the required standards of efficacy. However, efficacy is still a critical concern, and there is agreement in the field of stroke therapeutics that new devices should not be used prior to a demonstration of efficacy. However, efficacy trials can be both large and time-consuming, and device companies tend to be small, start-up businesses that have difficulty supporting such studies. In order to resolve this problem, several steps will likely be required, including: 1) enhanced interactions with the device companies, to familiarize them with the complex requirements of device approval, 2) improved awareness at FDA of the requirements and design of good efficacy trials, and 3) new methods of conducting efficacy trials to enhance their efficiency.
Another barrier to research is the lack of good intermediate markers that can be correlated with ultimate clinical outcome following stroke. We understand how stroke affects an individual's function, and leads to lesions that are apparent on brain imaging, but we know very little about other changes. Such changes, if identified, might be used in place of the typical stroke outcome measures (patient function and lesion size).
Along these same lines, there is a critical need to develop markers of stroke that complement traditional stroke scales. Brain imaging is still in early development, and will require considerable further investigation. Novel markers of ischemia, possibly including serum or cerebrospinal fluid tests, will be required to facilitate rapid bedside diagnosis and staging. Exciting areas of research are underway, and include the characterization of serum markers that may differentiate cerebral hemorrhage from ischemia. Such a test could stimulate the development of ways to begin stroke therapy in the field, at an earlier time point following stroke onset.
Lastly, the views of the individuals with stroke are an important concern for the field of acute stroke treatment. A meaningful result to a clinician or a healthcare system may be very different than a meaningful result to a patient. Clinicians may focus on basic functions, like movement, speech, and vision, or activities of daily living such as cooking, bathing, and dressing. By contrast, a patient may be much more concerned about regaining speech and the ability to comprehend language, read, or watch television. Additional research is needed on patient responses to stroke-induced deficits and patient perceptions of disability.
Therapeutic Innovation Barriers
Diagnostics -- the "EKG"
When individuals present in the clinic with conditions that mimic stroke, considerable effort is spent on separating the cases of true stroke from these "mimics." A simple, reliable, and accurate serum marker might allow faster diagnosis and facilitate the triage of all patients. Similarly, improvements in brain and vessel imaging might also aid in the accurate diagnosis of stroke, and importantly, identify patients most at risk for hemorrhage.
Combination Therapy
As described in a previous section, a renewed emphasis on combination therapy is needed in the field of stroke. Although pharmaceutical companies may encounter proprietary roadblocks to cooperation, methods must be identified which will allow these studies to move forward. NINDS could aid in this effort by investing in two types of research:
1) methodological investigations into how to design efficient trials of combinations, and 2) combination trials themselves.
Time Constraints
Brain tissue dies rapidly following stroke, thus the speed with which therapies need to be introduced presents a major barrier to effective treatment. More research is needed on the mechanisms of cell death after stroke, such that therapies that delay this degeneration can be developed. With such therapies in hand, effective treatments could be offered to a larger number of patients.
RESEARCH AND SCIENTIFIC PRIORITIESPriority 1:
Reperfusion: Therapeutic agents that open blood vessels in more patients and that do so better, faster, and more safely, are greatly needed.
It is quite possible that devices available in the future will aid in the removal or dissolution of clots. However, such devices must be rapidly deployed by healthcare practitioners, and at present, the time required for this deployment limits their potential utility. Trials must be designed to test the efficacy of these devices, and teams must be created that are ready to mobilize and use these interventions in an appropriate but realistic time frame (under one hour), in order to achieve patient benefit.
Lytic drugs that are more powerful and versatile, and open more arteries with fewer hemorrhages, are also greatly needed.
In addition to a need for new drug therapies, clinicians still do not know the best method for delivering these therapies. The best delivery model should be defined, and it may involve a network system, stroke teams, or single centers. The Brain Attack Coalition has already proposed one model (comprehensive and basic stroke centers linked in a network) that could be tested in a rigorous design.
Studies are also needed to identify ways to reduce hemorrhage. Current options include using safer lytics with or without concurrent cytoprotection, and improved patient selection. Devices to induce hypothermia may also reduce the rateof hemorrhage and augment the beneficial effects of thrombolysis.
Priority 2:
Biology and pathobiology: A paradigm shift to a focus on brain blood vessels is needed.
The primary events that cause ischemic and hemorrhage stroke -- the intraluminal molecular and cellular pathological processes within the brain blood vessels, particularly those of the aging brain -- are largely uncharacterized. The neuroscience of stroke -- ischemia, cellular injury, inflammation, necrosis, apoptosis, neuronal reorganization and repair -- are usually secondary events. Since they are of fundamental importance in the development of stroke, the intraluminal generation of thrombus, the lodging of that thrombus, lysis of the thrombus (or not), and the interplay of endothelial cells, platelets, and other cellular constituents of blood should be thoroughly explored.
In addition, pathology of the blood vessel walls also underlies the problem of brain hemorrhage. The basic vascular pathology of this form of stroke is even less well understood than is the pathology of ischemic stroke, and should be more fully investigated.
Priority 3:
Clinical trials and establishing the utility of cytoprotection: a shift is needed from single agent trials to combination trials.
A number of clinical trials have suggested that a single cytoprotectant is unlikely to work as an effective treatment for stroke. The design of these trials has, and may continue to be debated, however the pathobiology of ischemia suggests that a multi-modal approach may be more successful. Combination therapies have already proven to be highly successful in other fields of treatment, including cancer chemotherapy and in acute myocardial infarction. This area of research may require NINDS leadership, as proprietary issues related to the development of therapies by pharmaceutical companies may slow the formation of research collaborations.
As combination trials get underway, two types of research will be needed. Initially, multi-modal therapies should be tested in Phase 2 and Phase 3 trials. A second goal will involve improvements in primary trial design such that efficiency is enhanced.
As mentioned previously, the combinations that are most likely to be effective may involve the addition of a putative cytoprotectant to thrombolytic therapy. Multiple types of thrombolytics, such as those applied intravenously and intra-arterially, should be studied as part of this effort. In addition, multiple cytoprotective agents, with or without thrombolysis, could also be evaluated. Another consideration in the development of combination therapies is a reperfusion strategy; those approaches that include such a plan may be more likely to succeed.
RESOURCES NEEDED
Continued Investments in Stroke Research
Stroke is the second leading cause of mortality in the world, and a greater investment of the NINDS/NIH in stroke grants and contracts would enhance the progress in this field.
The funding mechanisms used by NINDS in past have worked well; rapid turn-around contract mechanisms, along with creative requests for applications have led to the rapid evaluation of multiple interventions, including hemodilution, naloxone, and t-PA. By contrast, current grant mechanisms may not facilitate the rapid application of new ideas, due to the time required to develop applications, and move through the review process. The new Specialized Program of Translational Research in Acute Stroke (SPOTRIAS), supported by NINDS, may help to resolve some of these issues, but this type of program should be strengthened and enhanced. NINDS is also encouraged to streamline the review process and time-to-funding, and is also encouraged to commit a sufficient level of support for the development of stroke centers.
Stroke Center Certification
The certification of stroke centers is another critical need in the field. As plans to develop centers move forward, it will be important to consider that the ability to bring stroke centers closer to patients may be a more productive goal than bringing patients to stroke centers more rapidly. Certification of these centers, also described above, should recognize competence for both basic stroke care (intravenous t-PA) and advanced care (endovascular treatment).
Further, designated stroke treatment centers would also have a critical role to play in the development of a consortium to conduct clinical trials - both to test new therapies and to help identify more meaningful outcome measures.
As they are developed, the certified stroke centers would need to be subject to a system of checks and audits similar to the trauma system. Voluntary compliance with guidelines has not worked well, and not all hospitals that have agreed to meet the guidelines developed by NINDS have followed them in practice. As a result, patients are taken to hospitals that claim to be ready to provide stroke care, but are not. This issue is a sensitive one in the field of acute stroke care, however there are potential solutions to this problem. Collegial pressure can be effective in encouraging centers to follow guidelines, as can linking identification of a center to reimbursement or participation in clinical trials.
Development of Emergency Department Investigators and Protocols
In the field of cardiology, most patients enrolled in trials of acute myocardial infarction treatments are identified first by investigators in emergency medicine departments. In this field, advances in therapy have been transferred smoothly to clinical practice because emergency medicine physicians provide the initial clinical care.
For stroke, initial clinical care is also provided by emergency medicine physicians. However, this group has not moved forward as rapidly with stroke research protocols as they have with cardiology protocols. Given that the success of emergency room interventions (t-PA) has been established, more urgent interventions are needed, and emergency personnel may be the best equipped to administer these interventions, it will be important for research to be fostered among these investigators.
Education and Telemedicine
The training of established clinicians in stroke care is an important goal, and financial incentives for physicians who are already in practice may help to facilitate this training. However, a more achievable approach may be the education of residents and medical school students. For practitioners at some community hospitals, telemedicine may help to facilitate the application of urgent interventions, by enabling them to achieve an increased understanding of stroke treatments. Along these lines, public education is also an important goal, and can aid in the recognition and treatment of stroke.
Endpoints; Surrogate Markers of Outcomes
Individuals with stroke need to be identified more quickly than they are at present, so that treatments can be provided faster and more safely. To achieve this goal, a rapid, reliable, sensitive, and specific marker of stroke is needed. Ideally, a fingerstick test would be developed that would allow paramedics to diagnose strokes in the field. Rapid identification of individuals with stroke would accelerate the overall medical response: medics could transport more quickly, the stroke team could assemble more rapidly, and needless additional testing could be avoided. Several candidate markers are already available and many appear promising enough to proceed with clinical validation. NINDS could provide needed support for the development of such markers.
New Ideas from Basic Scientists
Researchers must be encouraged to explore new areas of biology and pathobiology, such as:
Links Between Basic and Clinical Scientists
Advances in the treatment of myocardial infarction typically have involved the use of a device in combination with appropriate medical management. This is the result of effective translation of basic science findings on compounds such as aspirin, into informative clinical trials.
Small, project-type mechanisms that link clinical trials investigators with basic scientists are needed. This type of program could lead to productive collaborative relationships and information exchange between these very different groups of investigators. If designed well, the process can speed the development of animal models and the testing of therapies, by facilitating basic science studies that closely replicate the clinical situation.
The movement of an idea from basic proof-of-concept work through to clinical deployment requires an array of resources not found in most stroke centers. In other diseases, however, the "center" model has worked, notably in cancer therapeutics.
In order for an effective center for translation to be developed, the resources needed would include a stroke response team trained to give urgent stroke therapy and a biostatistical unit available to guide the development of new therapies and diagnostics. Such centers are also ideal places to train new clinician-investigators, so a fellowship core could be included as well. Other services that might benefit basic investigators include the collection and storage of blood and other human tissues for future use. NINDS could foster the development of such centers, using the SPOTRIAS or other mechanisms. As in basic investigation, the potential future benefits of such research activity cannot be predicted; it is certain, however, that such centers could provide the critical elements needed to efficiently translate basic discoveries into public health benefits.
Improved Research Infrastructure
In some ways, a renewed focused on community-based healthcare is new, however prior NINDS trials have used centers that led community-based networks. For example, in the rt-PA for Acute Stroke Trial, most of the patients were entered at community hospital emergency departments by stroke teams from the academic medical center. Due to time constraints, acute stroke trials require immediate treatment in the emergency department where the patient presents.
The Brain Attack Coalition proposed designated stroke centers, a system in which a comprehensive stroke center would lead
a network of primary stroke centers. The SPOTRIAS center grants could be fashioned to support such an infrastructure. These
community networks could, then, provide the support for a variety of trials. Also, such centers could also participate in
the collection and storage of blood and other specimens for use in genetic studies. An inevitable benefit of these centers
would be an increase in the number of treated patients, both in and outside of clinical trials.
Top
Participants:
Joseph P. Broderick
Stephen L. George
Philip B. Gorelick
S. Claiborne Johnston
Michael Krams
David E. Levy
John R. Marler
David S. Nilasena
Miriam T. Rodriguez
Judith A. Spilker
Marc K. Walton
Janet Wilterdink
Philip A. Wolf
STATEMENT OF THE PROBLEM
There is no substitute for randomized clinical trials to define the benefits and risks of interventions for stroke prevention and treatment. Given the burgeoning understanding of the diseases causing stroke, the number of potential interventions, and the magnitude of stroke's burden on society, there are too few stroke trials. The overall goal of NINDS-sponsored clinical trials in stroke is to decrease the burden of neurological disease due to stroke and vascular diseases of the brain, with the corollary objective of gaining new knowledge about biology and pathophysiology. It is no exaggeration that NINDS-sponsored trials in stroke over the past two decades have been landmark trials that influence daily the management of stroke patients throughout the world. But we must move in new directions and accelerate the advances in cerebrovascular disease.
Research Areas
Stroke is caused by several different disease processes, each with distinct interventional strategies for prevention, acute care, and recovery. Primary prevention refers to interventions undertaken prior to the clinical occurrence of a stroke or transient ischemic attack. In contrast, secondary prevention refers to treatment to prevent stroke in those who have survived an initial stroke or transient ischemic attack (about 20 percent of incident strokes occur in those with prior stroke or transient ischemic attack). The distinction between primary and secondary prevention is blurred, since many strokes are unrecognized by patients and their physicians; apparently asymptomatic, subclinical or "silent" strokes detected by MR imaging are common in the elderly. Potential interventions for prevention include treatment of traditional and novel risk factors (e.g., hypertension management, lipid lowering using the statin class of drugs, folate supplementation for elevated homocysteine, and others) and antithrombotic agents.
In addition to clinical and subclinical stroke (i.e., temporally discrete, focal brain lesions), vascular disease of the brain is increasingly appreciated to play a role in progressive cognitive decline in the elderly, albeit through mechanisms incompletely defined at present. Future clinical trials testing interventions aimed at vascular diseases causing stroke should include cognitive assessment and quality of life measures in addition to counting stroke events. Clinically recognized stroke events represent only the tip of the iceberg of vascular injury to the brain.
Treatments and diagnostic procedures of uncertain value are widely used for treatment or prevention of stroke. The need for testing of such widely practiced procedures and treatments depends on their intrinsic risk, as well as their total cost to society. Clinical trials of certain widely used procedures offer a potential win-win situation: whatever the outcome of the trial, stroke patients benefit. "Positive" results justify the wider use of such procedures, while "negative" results discourage their use, saving money and avoiding needless risks. While scientifically less attractive (in a perfect world, treatments would not be widely used without adequate validation of benefit), "negative" trials are consistent with and mandated by the primary mission of NINDS clinical trials.
Scientifically, there is more appeal in clinical trials that evaluate new interventions that emerge from laboratory research and/or from epidemiological studies. Use of t-PA and the neuroprotectants for acute stroke are examples of such interventions that have been evaluated in recent clinical trials. When testing these types of novel interventions, a "positive" outcome is required to impact patient care, although "negative" results can significantly shape future research directions.
Diagnostic procedures are increasingly expensive and are sometimes risky. It is particularly challenging to assess their overall value in clinical trials because their potential benefit to patients is often linked to the availability of established treatments, whose use, in turn, depends on the results of the diagnostic test. Trials of diagnostic procedures should be reserved for situations where the costs or risks are high and the management implications are important (based on availability of validated interventions). Ongoing studies of MR imaging to predict potential late-responders (i.e., more than three hours from stroke onset) to t-PA are an example. Not only must the accuracy of the MR technique be determined, but also the incremental benefit afforded to all the patients who undergo the procedure (whether treated with t-PA or not) must be characterized. An inherent difficulty in studying diagnostic tests is the standardization of procedures at different clinical sites; this difficulty is also encountered in their eventual application to clinical use, but is particularly challenging during early phases of development and application.
Collaboration Issues
Barriers to collaboration between pharmaceutical and medical device companies and NINDS exist concerning validation of new agents and diagnostic procedures. Proprietary interests often necessitate added complexity in a clinical trial, primarily because of regulations that must be followed to gain FDA approval. Government participation in the clinical testing of a drug for stroke is not entirely welcomed by pharmaceutical companies for many reasons, including uncertainty about whether collaboration in a government-sponsored trial threatens future profits, and concerns about the control and publication of trial data. In addition, when similar agents are available for prevention or treatment of stroke, pharmaceutical companies are often reluctant to undertake the large trials necessary to compare two different, potentially active agents. For example, the control of blood pressure has well-established benefits for stroke prevention; does the specific type of antihypertensive agent make a difference?
Social and Economic Issues
Primary prevention of stroke is a complex issue that must consider patient preferences, societal values, and medical economics, in addition to the biology of the disease. There is often a disparity between what is good for public health and what is accepted by people, patients, and physicians.
For example, treatment of hypertension remains underutilized in our relatively affluent society, even after more than two decades of high-quality trials have shown profound reduction in stroke and despite the availability of generally well-tolerated antihypertensive drugs.
In contrast, an area of notable success in primary prevention of stroke has been the use of warfarin in patients with atrial fibrillation. Of more than two million Americans with this cardiac rhythm disturbance, which predisposes to stroke, almost half are now treated with warfarin following the NINDS-sponsored Stroke Prevention in Atrial Fibrillation (SPAF) trials and other clinical trials. This has prevented tens of thousands of strokes yearly. For these patients, the number-needed-to-treat for one year with warfarin to prevent a stroke is typically about 40; in other words, they are at relatively high risk. For the average middle-aged person with mild hypertension, the risk of stroke is lower; the number-needed-to-treat for one year to prevent one stroke is several hundred.
One key to more widely applied interventions for primary stroke prevention will be reliable identification of relatively high-risk populations, which are more likely to accept the need for preventive therapy.
In addition, primary prevention trials concerning vascular disease of the brain should not focus exclusively on stroke outcomes, but rather on more global measures of the healthy brain potentially impacted by interventions. Vascular factors may have a major role in age-related cognitive impairment and are more common and perhaps even more important than clinical stroke events. In short, large primary prevention trials involving people at particular risk, to test interventions aimed at vascular disease of the brain in all of its manifestations, are warranted.
Clinical trials also need to be more efficient and economical. Novel design strategies that test multiple interventions in factorial designs, incorporate futility monitoring and innovative statistical design when appropriate, and use more sensitive global outcomes as well as potentially more sensitive outcomes such as cognition and quality of life, must be further developed. Large "simple" trials are appropriate to address certain issues; however, they need to be balanced with more complex trials that also seek to incorporate study of the disease process.
Progress has been slowed unnecessarily by the slow recruitment of patients in stroke trials. Many trials have taken several years to recruit participants; other than a lack of organization, there is little excuse for this, given the huge number of Americans with cerebrovascular diseases. Currently, answers that should take a few years to obtain can take a decade or more. In addition, the relatively high cost of clinical trials has limited the number and types of trials sponsored.
To address these issues, the development of a clinical trial network, including both academic and practicing physicians, and an infrastructure along the lines successfully used in oncology and cardiology, should be explored.
Increased efficiency will also require the development and inclusion of important new trial methodologies. For example, there is a need for trial designs that can accommodate the complexity of acute stroke. Questions to consider in such trials might include the best ways to develop combination therapy for acute stroke while determining the correct dose and duration of treatment and also accounting for drug interactions. In addition, it might be helpful to develop adaptive designs with real-time data retrieval and adaptive treatment allocation to optimize learning about the research question. Also, it would be important to be able to maximize the information that can be obtained from each patient and the benefits to patients that can be obtained from each trial. Another consideration could be a decision-theoretic termination rule, for stopping a trial at the earliest point at which sufficient information is available to conclusively answer the research question or, when appropriate, to determine futility.
Inadequately Explored Areas
Primary reliance on investigator-initiated clinical trials has led to uneven clinical study of the different aspects of stroke. For example, several clinical trials have evaluated stroke prevention in patients with atrial fibrillation and in patients with cervical carotid atherosclerosis, whereas no trials have been conducted in intracerebral hemorrhage. Less common causes of stroke relevant to young adult and pediatric stroke patients (e.g., arterial dissections) are unlikely to be brought to trial given current mechanisms. Recent availability of funding for pilot clinical trials has allowed testing of methodological innovations important for study of less common stroke subtypes.
Future Needs
In short, too few clinical trials have been conducted in stroke prevention and treatment, given the burden of the disease and the availability of potentially efficacious interventions. Some of the specific areas that need more research include:
In order to accomplish all that needs to be done in stroke trials, the clinical trial process needs to be more efficient, economical, and expedient, without inhibiting individual investigator enthusiasm and innovation.
CHALLENGES AND QUESTIONS
Some of our major challenges are as follows:
NINDS could take a major step toward reducing the cost of clinical trials and increasing recruitment by developing a clinical trial investigator network to recruit participants for multiple trials. Such a network could involve both academic and practice-based physicians and could obviate the need to rebuild the clinical site recruitment infrastructure from scratch for each new trial. More efficient use of trial machinery should result in less expense. Because of the relative maturity of development of stroke clinical trials and the large number of patients, stroke research may be more readily adaptable to the clinical trials consortium concept than are some other areas of neurologic research. In the future, expansion of the consortium or the parallel development of consortia for other areas of neurology may be useful.
Other issues that need to be considered include:
BARRIERS
Barriers to improved stroke clinical trials include:
RESEARCH AND SCIENTIFIC PRIORITIES
Priority 1:
Identify needed clinical trials in stroke prevention, acute treatment, and recovery.
The NINDS should undertake leadership in organizing large, simple clinical trials for primary prevention of vascular injury to the brain (including hemorrhagic and ischemic stroke), and for evaluation of rehabilitation therapies for patients with vascular brain injury. The NINDS should encourage and support investigator-initiated clinical trials addressing areas such as the following:
This list should be reviewed regularly by a working group and updated as understanding of stroke advances. The goal would be to more rationally prioritize trials in different areas of stroke as well as to take advantage of specific research opportunities.
Priority 2:
Organize an NINDS-based clinical trial investigator network to execute clinical trials expeditiously and efficiently.
Such a network could involve scores, or even hundreds, of clinical investigators (and sites) who would enroll patients in several ongoing trials, enhancing rapid and less expensive recruitment due to economies of scale. Inclusion of community-based (i.e., non-academic) investigators would permit access to a broader range of participants in a "real-world" setting. The network would minimize the need to re-create the various components of trials every time a trial is proposed (whether institute- or investigator-initiated). The organization and management of this network would require considerable effort and expense; its performance and value would require critical evaluationperiodically to assure that its objectives are being met.
Priority 3:
Encourage novel clinical trial methodology.
The "pre-review" process thatis now being developed by the NINDS Clinical Trials Group should be expanded in order to assist investigators in the initial submission of the highest-quality grant applications and encourage the use of efficient designs and novel outcomes. For planning grants, the review process should be accelerated.
Additional Priorities
The NINDS and the stroke research community should:
RESOURCES NEEDED
Priority 1:Priority 2:
Priority 3:
The co-chairs of this session thank Dr. Barbara Tilley for additional contributions to this session's report. Top
Co-Chairs: Pamela W. Duncan, Ph.D., P.T., and Timothy J. Schallert, Ph.D.
Participants:
Leora R. Cherney
Steven C. Cramer
Larry B. Goldstein
Leslie J. Gonzalez-Rothi
James C. Grotta
Judi Johnson
Theresa A. Jones
Richard F. Macko
Nancy E. Mayo
Katherine J. Sullivan
John Whyte
Steven L. Wolf
Priority 1:
Investigate the neurobiology of recovery.
RationaleObjectives
Priority 2:
Promote evidence-based investigations of innovative therapies compatible with principles of neural plasticity and learning.
Rationale
Priority 3:
Evaluate the organization of rehabilitation services.
Rationale
RESOURCES NEEDED
Participants:
Joe Acker
Susan C. Fagan
Walter N. Kernan
William T. Longstreth, Jr.
John R. Marler
George A. Mensah
Andrew Nelson
David S. Nilasena
Linda Williams
Rhys Williams
A major goal of NINDS-supported research is practical: to reduce the burden of neurological disease for the U.S. population. This practical goal can only be realized when the research results are implemented appropriately and well in real-world practice. The broad aims of health services implementation research are to (1) help patients and healthcare workers interpret clinical research so they may understand which interventions lead to valued outcomes, (2) identify when valued interventions are not being used optimally, and (3) develop, test, and promote practical ways to implement those interventions.
Failing to address these issues leads to a variety of specific problems:
There are general challenges and barriers to the three research issues outlined above.
Challenges to Measuring Outcomes
Challenges to Evaluating Current Practice Patterns
Challenges to Developing, Testing, and Promoting Health Services Interventions
Priority 1:
Measure outcomes that people care about.
Priority 2:
Identify when care is inconsistent with best evidence.
Priority 3:
Develop, test, and promote ways to optimize practices based on best evidence.
APPROACHES TO ADVANCING THESE PRIORITIES
Advances in diminishing the burden of stroke in the U.S. require that we overcome the classes of barriers mentioned above. The three research priorities speak to overcoming the barriers. Studies performed within the framework of these priorities will serve to advance the uptake of clinical practice based on best evidence.
The three research priorities are derived from experience in health services implementation research that suggests this effort is most effective when it proceeds in a logical sequence. Such a sequence is listed in Statement of the Problem: (1) identify interventions that lead to valued outcomes, (2) identify "targets of opportunity" (i.e., community practices that are not optimal), and (3) develop, test, and promote practical ways to implement those practices.
Each step in this sequence can be addressed with a variety of methodological tools. Identifying interventions that lead to valued outcomes may be as simple as evaluating a clinical trial with crucial outcome measures. When a single study is not sufficient, other approaches can be considered, such as evidence synthesis based on meta-analysis and/or decision modeling.
Identifying "targets of opportunity" can be accomplished by performing population-based surveillance studies to examine whether there are significant discordances between practices supported by best evidence and actual practice. These studies can also be used to identify barriers that may impede optimal practice. In specific circumstances, other issues and research methods may be relevant. For example, where cost is an important consideration, cost of illness and cost-effectiveness studies can be employed.
The quintessential approach to evaluating clinical interventions, including health services interventions, is the randomized controlled trial. Substantial progress has been made in advancing the science of health services implementation trials. When a randomized controlled trial is not feasible, other designs, such as before and after studies with concurrent controls can be reasonable alternatives.
TopCo-Chairs: John M. Hallenbeck, M.D., and Richard J. Traystman, Ph.D.
INTRODUCTION
The brain does best if ischemia can be prevented and it does second best if ischemia can be reversed within the first few hours. Ischemia and hemorrhage set into motion many injury mechanisms and injury-modifying mechanisms. We need to understand these mechanisms in exquisite molecular detail. We must find better ways to subdue vessel activation as it begins to threaten thrombosis or hemorrhage. We must find ways to block or attenuate the systems of injury mechanisms that extend brain damage in acute stroke. We need to understand recovery and repair mechanisms and learn how to promote them. We can assemble the expertise and apply the technology to do these things.
OVERARCHING THEMES
Several overarching themes emerged from the 15 Roundtable Meeting breakout sessions. In particular, the stroke community needs to focus on:
Several of these themes are discussed below.
GENOMICS AND PROTEOMICS TECHNOLOGIES
Many of the overarching themes are best discussed in the context of the impending genomic and proteomic revolution.
New genomics and proteomics technologies, such as DNA and protein chips, together with miniaturization down to nanostructures, hold tremendous potential for providing personalized medical care to the individual patient that is based on predicted molecular mechanisms. If we can successfully filter and digest cryptic, non-intuitive data and connect it to clinically relevant information, the resultant molecular profiling should suggest many hypotheses that can be tested at both preclinical and clinical levels. For optimal translation, every effort should be made to increase the relevance of preclinical models to clinical disease. Some crucial points are addressed below.
Genomics and proteomics technologies will operate in a major new research paradigm.
Data analysis will require multidisciplinary teams, including computer scientists skilled at mining databases, biostatisticians, bioengineers, molecular and cellular biologists, pharmacologists, and basic and clinical stroke researchers.
In the most enthusiastic scenario, genomics and proteomics may turn into a major engine that leads and drives much of stroke research. The molecular mechanisms, systems of mechanisms, and targets derived from human tissue samples may activate stroke research on a pathway that begins at the bedside, goes to the bench (e.g., cell culture systems, transgenic models), and then goes back to the bedside for clinical trials. Developing hypotheses and designing research based on molecular mechanisms derived from human rather than animal tissues may help to resolve some of the difficulties that have been encountered in translating preclinical stroke research into approved clinical applications.
Genomics and proteomics technologies may require a systems biology approach to understand gene and protein interactions.
These technologies will involve a "combinatorial explosion," raising many questions about how combinations of factors interact and influence each other. Such an approach is appropriate for a multifactorial problem like the progression of brain injury during acute stroke. Data derived from the new technologies should encourage the study of the interrelationships among molecular mechanisms of injury and their effects on all brain structures, rather than a preoccupation with a circumscribed area of interest, which has tended to characterize current stroke research. This approach to research should encourage a more global view.
To realize the full predictive value of genomics and proteomics analysis, we need massive clinical and population analyses (a bioinformatics database) that will identify and validate the predicted batteries of disease-relevant molecular markers (biomarkers and surrogate measures). This must be accomplished before the markers can be used in routine clinical diagnosis, individualization of therapy, and prognosis.
We don't know yet whether genomics and proteomics technologies will follow the development pattern of personal computers, resulting in miniaturized lab-on-a-chip devices in every lab and clinic, or whether they will remain expensive "mainframe" instruments found only in central locations. The NINDS should decide (perhaps with other institutes) what role is appropriate to facilitate the availability of this technology for stroke researchers and how best to assist with their education to exploit the full potential of these technologies. For example, three years ago NCI established a joint initiative with FDA, the Tissue Proteomics project. The stated goal of the project is to "originate and complete technology for studying proteomic networks and signal pathways in small quantities of microdissected human tissue cells directly from biopsy specimens." It is oriented toward immediate, patient-based clinical applications. NCI might be a good strategic partner for NINDS in this area.
One possible effect of the genomics/proteomics paradigm is that research groups will increase their cooperation and reduce competition, since input from many laboratories and clinical centers will be necessary to build the bioinformatics database. Currently, however, data submitted to the database does not remain proprietary, so it may be difficult to ensure that individual research effort is recognized and rewarded. Our research community will need to address this problem.
A proliferation of injury mechanisms during acute brain ischemia has been revealed by conventional research approaches during the past decade. This list is likely to expand exponentially as a consequence of this new research paradigm. We may have to devise ways to counteract and control multiple injury mechanisms at the same time in order to achieve a marked attenuation of progressing brain damage in acute stroke. One approach to this that could be facilitated by the new research paradigm is the discovery and characterization of master regulatory switches and molecular mechanisms that simultaneously counteract multiple mediators of injury and confer resistance to ischemic brain damage.
STROKE MODELS
Discussion of stroke models permeated many of the Roundtable Meeting breakout sessions.
The scientific research community has developed many varieties of stroke models over the past years. These models encompass stroke, subarachnoid hemorrhage, intraventricular hemorrhage, vasospasm, and global ischemia. These models have been developed for two reasons: one, to try to recreate the human disease in an animal, and two, to create a model that allows us to study and dissect mechanisms of formation of injury and mechanisms of neuroprotection from injury.
For example, in developing animal models of focal ischemia (middle cerebral artery occlusion), there are different methods available to occlude the vessel. It can be coagulated, clipped with a neurosurgical clip, clogged with thrombo emboli, or blocked with a thread or filament. There are both similarities and differences between these models of middle cerebral artery occlusion, and there is always the question of whether any of them actually represent a model of human disease. However, each of these models can be studied to examine mechanisms of injury and mechanisms of neuroprotection. This can equally be stated for other models of injury as well.
The question is, why have these animal models not predicted stroke outcome in clinical trials? This issue arises often, given the number of negative clinical trials. For many of the animal models, specific pharmacologic agents have been protective, yet when these same agents were used in patients in clinical trials, the results were disappointing. In the past, most have placed fault on the animal models, saying they are not appropriate models of human disease and therefore results obtained in the animal may not have been predictive of a drug's efficacy in humans.
Fault may also be placed on the clinical trials, however. One problem is that trials may be organized in a way that is not completely based on preclinical data. For example, why design a clinical trial such that a drug must be given at three hours after the stroke event, when no preclinical studies have looked at the efficacy of the drug at that time? Preclinical data for the drug may have shown effectiveness when administered prior to ischemia, or at reperfusion, but the drug had never been administered at three hours following ischemia. Another possible issue is drug dosing. What may be an effective dose in animals may not be the most effective in humans.
Thus, fault can be given to both the researcher, who may not study a drug completely from a dose/response analysis or a window of opportunity analysis, and to the clinician, who may design and perform a clinical trial not based on appropriate preclinical data.
BASIC SCIENTIST AND CLINICIAN INTERACTION
The problems seen in carrying animal model data to clinical trials points out the need for strong interaction and collaboration between the basic scientist and the clinician, another overarching theme of the Roundtable Meeting breakout sessions.
Basic scientists and clinicians must work together to design proper clinical trials based on appropriate preclinical data. They need to discuss precisely what information the clinician needs in order to design an appropriate clinical trial, and which animal experiments the researcher must do to allow the clinician to base the design of the trial on those data. This collaborative relationship is critical for the design of both preclinical studies and the clinical trial itself. The proper integration and collaboration of researchers and clinicians will clearly lead to better translational research for the good of the stroke patient.
COMBINATION THERAPY
Ischemia results in brain injury, and over the years many potential mechanisms of injury and neuroprotection have been identified. But as the mechanisms of neuroprotection each have been dissected out, not one has been found to be capable of completely ameliorating the injury produced. Therefore, it seems reasonable that no single agent will be found that, when administered alone, will ameliorate all injury or completely protect the brain. It is more likely that agents will be given in combination (combination therapy) as a "cocktail," with effects on several different protection mechanisms. One might have to administer, for example, an excitotoxic amino acid antagonist and an O2 radical scavenger together, in order to protect the brain better than either agent could do individually.
The preclinical studies required to determine which agents to administer, at what dose, and the timing of administration of the agents, however, are extremely complicated. To perform dose/response curves for two or more agents in combination with dose/response curves for windows of opportunity is time consuming and complex.
There is also the question of who would fund such studies. These types of studies traditionally have not fared well at NIH Study Sections. They are perhaps not reductionistic, molecular, or mechanistic enough, not to mention innovative enough, to do well. Thus, there may have to be a change in the culture concerning how these types of grant proposals are reviewed, in order to appreciate and foster these types of studies.
MULTIDISCIPLINARY TRAINING AND RESEARCH PROGRAMS
Stroke is a disease of the vasculature, of the blood vessels. Perhaps we need to return to the study of the vessels and their elements to discover new potential cures for stroke. Areas that need to be studied more carefully include methods to unclog vessels, mediators released from the clot in the vessels, the cerebral endothelium and its responses, reperfusion, hyper- and hypoperfusion, vascular inflammation, and new, unique agents to vasodilate cerebral vessels.
In order to accomplish all this, we need to have individuals trained in these areas, and/or we need to attract individuals from other fields (e.g., neurology, neurosurgery, neurosciences, physiology, radiology, pharmacology, pathology, molecular biology, genetics, biomedical engineering, and microcirculation) into the area of stroke research. This would bring individuals with the best creative minds and with techniques from different areas to study stroke.
Postdoctoral programs and programs for mid-career individuals who want to redirect their efforts towards stroke and cerebrovasculature are needed to attract individuals into the stroke field. How do we attract and train the vascular wall biologists, for example, to direct their attention to the field of stroke? Stroke involves multidisciplinary activities, and perhaps interdepartmental or even inter-university individuals. How do we foster these multidisciplinary, intra- and inter-university approaches to the study and understanding of stroke?
This will not be easy to accomplish, but one way to be successful is through collaborative grant mechanisms. The SPRG strongly encourages the NINDS, NHLBI, and other institutes, within or outside the NIH, to facilitate the development of new, innovative programs to accomplish the goals outlined above.
Along with this comes the challenge of implementing clinical trials. Clinical trials are complicated, multidisciplinary, multi-institutional, and very expensive to perform. The SPRG recommends that the NIH and industry work together to creatively move stroke trials forward. Developing stroke databases of epidemiological, genetic, and imaging data are similarly challenging and would benefit from collaborative efforts among the NIH institutes and industry.
These are difficult challenges that need to be addressed. However, one thing is certain: there are many individuals, clinicians and researchers alike, from many different fields (as evidenced from attendees at the Stroke PRG Roundtable Meeting) who are poised to work together to advance our understanding of stroke mechanisms and translate that understanding into better prevention, diagnosis, and treatment of this major cause of death and disability in the United States. These intellectual resources, combined with the necessary fiscal resources, will provide a powerful impetus for alleviating the devastating effects of stroke on our society.
TopJoe Acker, E.M.T.-P., M.P.H.
Birmingham Regional Emergency Medical Services System
Mark J. Alberts, M.D.
Nothwestern University Medical School
Beth M. Ansel, Ph.D.
National Institute of Child Health and Human Development
David M. Armstrong, Ph.D.
Center for Scientific Review, National Institutes of Health
Larry D. Atwood, Ph.D.
Boston University School of Medicine
Stephen Barnes, Ph.D.
University of Alabama, Birmingham
Joseph S. Beckman, Ph.D.
Oregon State University
Toby N. Behar, Ph.D.
National Institute of Neurological Disorders and Stroke
Eric Boerwinkle, Ph.D.
University of Texas
Health Science Center at Houston
Joseph P. Broderick, M.D.
University of Cincinnati College of Medicine
Thomas G. Brott, M.D.
Mayo Clinic, Jacksonville
Alastair M. Buchan, M.D.
University of Calgary
Neil S. Buckholtz, Ph.D.
National Institute on Aging
Dorit Carmelli, Ph.D.
SRI International
Pak H. Chan, Ph.D.
Stanford University School of Medicine
Leora R. Cherney, Ph.D.
Rehabilitation Institute of Chicago
Arlene Chiu, Ph.D.
National Institute of Neurological Disorders and Stroke
Michael Chopp, Ph.D.
Henry Ford Hospital
Helena C. Chui, M.D.
University of Southern California
Richard A. Cohen, M.D.
Boston University School of Medicine
Steven C. Cramer, M.D.
University of Washington
Bruce M. Coull, M.D.
University of Arizona College of Medicine
Gregory J. del Zoppo, M.D.
The Scripps Research Institute
Andrew M. Demchuk, M.D.
University of Calgary
John A. Detre, M.D.
University of Pennsylvania Hospital
Gabrielle A. deVeber, M.D.
Hospital for Sick Children
Robert A. Dobie, M.D.
National Institute on Deafness and Other Communication Disorders
Pamela W. Duncan, Ph.D., P.T.
University of Kansas Medical Center
J. Donald Easton, M.D.
Brown Medical School
Susan C. Fagan, Pharm.D.
University of Georgia
Frank M. Faraci, Ph.D.
University of Iowa College of Medicine
Jawed Fareed, Ph.D.
Loyola University Medical Center
Donna M. Ferriero, M.D.
University of California, San Francisco
Giora Z. Feuerstein, M.D.
DuPont Pharmaceutical Company
David J. Fink, M.D.
University of Pittsburgh
Seth P. Finklestein, M.D.
Massachusetts General Hospital
Gary Fiskum, Ph.D.
University of Maryland, Baltimore
Tatiana Foroud, Ph.D.
Indiana University School of Medicine
Jeffrey I. Frank, M.D.
University of Chicago
Lawrence Friedman, M.D.
National Heart, Lung, and Blood Institute
Karen L. Furie, M.D., M.P.H.
Harvard Medical School
Anthony J. Furlan, M.D.
Cleveland Clinic Foundation
Stephen L. George, Ph.D.
Duke University Medical Center
Mary Gerritsen, Ph.D.
Genentech
Gary Gibbons, M.D.
Morehouse School of Medicine
Meighan Girgus
American Stroke Association
Mark P. Goldberg, M.D.
Washington University School of Medicine
Larry B. Goldstein, M.D.
Duke University Medical Center
Leslie J. Gonzalez-Rothi, Ph.D.
VA Medical Center and University of Florida Brain Rehabilitation Research Center
Philip B. Gorelick, M.D., M.P.H.
Rush Medical Center
Neil Granger, Ph.D.
Louisiana State University
Health Sciences Center
Steven A. Greenberg, M.D.
Brigham and Women's Hospital
Steven M. Greenberg, M.D., Ph.D.
Massachusetts General Hospital
David Greenberg, M.D., Ph.D.
Buck Institute for Age Research
James C. Grotta, M.D.
University of Texas
Health Science Center at Houston
James D. Gustafson
Possis Medical, Inc.
Antoine M. Hakim, M.D., Ph.D.
Canadian Stroke Network
John M. Hallenbeck, M.D.
National Institute of Neurological Disorders and Stroke
Jonathan L. Halperin, M.D.
Mount Sinai Medical Center
Barbara Handelin, Ph.D.
Kenna Technologies, Inc.
John A. Hardy, Ph.D.
Mayo Clinic, Jacksonville
Christina A. Harrington, Ph.D.
Oregon Health & Science University
Robert G. Hart, M.D.
University of Texas
Health Science Center at San Antonio
Wolf-Dieter Heiss, Prof. Dr.
Max Planck Institute for Neurological Research
Donald D. Heistad, M.D.
University of Iowa
Randall T. Higashida, M.D.
University of California, San Francisco Medical Center
Deborah G. Hirtz, M.D.
National Institute of Neurological Disorders and Stroke
Julian T. Hoff, M.D.
University of Michigan
Konstantin-Alexander Hossmann, M.D, Ph.D.
Max Planck Institute for Neurological Research
George Howard, Dr.P.H.
University of Alabama, Birmingham
Chung Y. Hsu, M.D., Ph.D.
Washington University School of Medicine
Willa A. Hsueh, M.D.
University of California, Los Angeles
Carole Hudgings, Ph.D., R.N.
National Institute of Nursing Research
Patricia D. Hurn, R.N., Ph.D.
The Johns Hopkins University School of Medicine
Costantino Iadecola, M.D.
University of Minnesota
Alan J. Jacobs, M.D., Ph.D.
Layton BioScience, Inc.
Thomas P. Jacobs, Ph.D.
National Institute of Neurological Disorders and Stroke
Barbro B. Johansson, M.D., Ph.D.
Lund University
Judi Johnson, Ph.D., R.N.
North Stroke Center and North Minneapolis Health Care
S. Claiborne Johnston, M.D., M.P.H.
University of California, San Francisco
Theresa A. Jones, Ph.D.
University of Washington
Markku Kaste, M.D., Ph.D.
University of Helsinki
Thomas A. Kent, M.D.
University of Texas Medical Branch
Edgar J. Kenton III, M.D.
Main Line Jefferson Health System
Walter N. Kernan, M.D.
Yale University School of Medicine
Chelsea Kidwell, M.D.
University of California, Los Angeles Medical Center
Kathleen King
National Stroke Association
Steven Kittner, M.D., M.P.H.
University of Maryland School of Medicine
Michael Krams, M.D.
Pfizer Global Research and Development
Thomas G. Kwiatkowski, M.D.
Long Island Jewish Medical Center
Dennis Landis, M.D.
Case Western Reserve University
David E. Levy, M.D.
DOV Pharmaceutical, Inc.
Weili Lin, Ph.D.
University of North Carolina, Chapel Hill
Eng H. Lo, Ph.D.
Harvard Medical School
William T. Longstreth, Jr., M.D.
University of Washington
Claudia J. Louis, M.B.A.
American Heart Association
Patrick D. Lyden, M.D.
University of California, San Diego School of Medicine
Richard F. Macko, M.D.
Baltimore VA and University of Maryland
Peter R. MacLeish, Ph.D.
Morehouse School of Medicine
Frank W. Marcoux, Ph.D.
Pfizer Global Research and Development
John R. Marler, M.D.
National Institute of Neurological Disorders and Stroke
David B. Matchar, M.D.
Duke University Medical Center
Nancy E. Mayo, Ph.D.
McGill University
George A. Mensah, M.D.
Centers for Disease Control and Prevention
James F. Meschia, M.D.
Mayo Clinic, Jacksonville
Mary Ellen Michel, Ph.D.
National Institute of Neurological Disorders and Stroke
Lewis B. Morgenstern, M.D.
University of Texas
Houston Medical School
James H. Morrissey, Ph.D.
University of Illinois College of Medicine
Michael E. Moseley, Ph.D.
Stanford University
Michael A. Moskowitz, M.D.
Harvard Medical School
Lennart Mucke, M.D.
University of California, San Francisco
J. Paul Muizelaar, M.D., Ph.D.
University of California, Davis
Andrew F. Nelson, M.P.H.
HealthPartners Research Foundation
David S. Nilasena, M.D., M.S.P.H.
Health Care Financing Administration
Randolph J. Nudo, Ph.D.
University of Kansas Medical Center
LaRoy P. Penix, M.D.
Morehouse School of Medicine
Audrey S. Penn, M.D.
National Institute of Neurological Disorders and Stroke
William J. Powers, M.D.
Washington University School of Medicine
Barbara Radziszewska, Ph.D.
National Institute of Neurological Disorders and Stroke
Bruce R. Ransom, M.D., Ph.D.
University of Washington
Rose Marie Robertson, M.D.
Vanderbilt University
Kenneth J. Rockwood, M.D.
Dalhousie University
Miriam T. Rodriguez, M.S.W.
Magee Rehabilitation Hospital
Gary A. Rosenberg, M.D.
University of New Mexico School of Medicine
Robert Rothlein, Ph.D.
Boehringer Ingelheim Pharmaceuticals Inc.
Zaverio M. Ruggeri, M.D.
The Scripps Research Institute
Ralph L. Sacco, M.D., M.S.
Columbia University Neurological Institute
Timothy J. Schallert, Ph.D.
University of Michigan
Bradford S. Schwartz, M.D.
University of Illinois College of Medicine at Urbana-Champaign
Stephen M. Schwartz, M.D., Ph.D.
University of Washington
Paul A. Scott, Ph.D.
National Institute of Neurological Disorders and Stroke
Frank R. Sharp, M.D.
University of Cincinnati Medical Center
Patti Shwayder
National Stroke Association
Roger P. Simon, M.D.
Robert S. Dow Neurobiology Laboratories
Ingmar G. R. Skoog, M.D.
University of Goteborg
Don B. Smith, M.D.
University of Colorado
Health Sciences Center
Evan Y. Snyder, M.D., Ph.D.
Harvard Medical School
Judith A. Spilker, B.S.N.
University of Cincinnati
Mary Stenzel-Poore, Ph.D.
Oregon Health & Science University
David M. Stern, M.D.
Columbia University College of Physicians and Surgeons
Katherine J. Sullivan, Ph.D., P.T.
University of Southern California
Raymond A. Swanson, M.D.
University of California, San Francisco and VA Medical Center
Richard J. Traystman, Ph.D.
The Johns Hopkins University School of Medicine
Marc K. Walton, M.D., Ph.D.
Food and Drug Administration
Steven Warach, M.D., Ph.D.
National Institute of Neurological Disorders and Stroke
David S. Warner, M.D.
Duke University Medical Center
Bryce Weir, M.D.
University of Chicago
John Whyte, M.D., Ph.D.
Moss Rehabilitation Research Institute
Linda Williams, M.D.
Indiana University School of Medicine
Rhys Williams, Sc.D.
Bristol-Myers Squibb
Janet Wilterdink, M.D.
Rhode Island Hospital
Philip A. Wolf, M.D.
Boston University School of Medicine
Steven L. Wolf, Ph.D.
Emory University School of Medicine
Katherine Woodbury-Harris, Ph.D.
National Institute of Neurological Disorders and Stroke
Justin A. Zivin, M.D., Ph.D.
University of California, San Diego
Last updated September 15, 2008