Workshop Report: Harnessing Neuroplasticity for Human Applications

Update: In April 2011, workshop participants published an article in Brain highlighting a number of broad themes and potential future directions that may produce therapeutic interventions to reduce disability across a range of conditions.

On April 21-22, 2009, the NIH Blueprint for Neuroscience Research hosted a workshop to promote opportunities for translating basic neuroplasticity and circuit retraining research into strategies for improving clinical outcomes across a variety of disorders and functional domains. Drs. Steven Cramer and Mriganka Sur co-chaired the meeting of approximately 30 leading scientists.  The group was charged with responding to questions focused on identifying (1) cardinal examples of plasticity with the strongest adaptive and maladaptive behavioral correlates, (2) promising interventions for manipulating central nervous system (CNS--brain or spinal cord) plasticity  toward improving human behavioral outcomes, (3) the highest-impact findings related to neuroplasticity, including how animal models have influenced the field, (4) the best measures for evaluating the anatomic basis of circuit reorganization and for monitoring functional progress in vivo, and (5) knowledge gaps and needs for translating neuroplasticity research into human interventions.  Prior to the meeting, the scientists met in four working groups spanning a wide range of conditions (adult trauma and stroke, mental and addictive disorders, neurodegeneration and aging, pediatric and developmental disorders) to address the scientific questions. Discussions at the workshop highlighted themes that were common across working groups,  as well as some that were specific to certain groups, and identified promising new directions, impediments to progress, and proposed solutions.  Several needs and areas for future emphasis were identified:

  1. A better vertical integration of basic and clinical research is needed.  Closer collaborations among basic, clinical, and translational researchers that begin in the study design stage and are interactive throughout will optimize the relevance and benefits of animal models for clinical translation.  Models are needed that better replicate the complexity of human disease mechanisms over time, that emulate human risk factors, including age, patterns of injury and post-injury experiences (i.e., models of disease and recovery, not plasticity) and that employ ecologically valid activities.  Endpoints in animal studies should be validated against endpoints in human studies; and points of overlap with respect to the effects of drugs and nonpharmacological therapies, included.  Rodent studies are well suited to investigations of cellular-molecular mechanisms, while human and non-human primate studies may be more appropriate for understanding circuit-level changes.

  2. Normative data on neuroplasticity and the postnatal development of various neural circuits, including complex and interconnected neural networks that link functional domains are needed. A deeper understanding of age effects on neuroplasticity, including critical or sensitive periods for various systems (sensory, motor, cognitive) might provide a basis for devising strategies to re-open or extend them for intervention purposes.

  3. Prevention and resilience research aimed at the prevention of disease onset, rather than post-onset treatment, is a priority.  There is a tremendous potential for cognitive neuroscience to inform and advance behavioral interventions and to promote future resilience.

  4. Methods for assessing and monitoring plasticity and circuit retraining in individual patients are needed.  Baseline assessments and validated, short-term probes (biomarkers) are needed to predict, at the individual level, who will or will not benefit by treatment, as well as by which treatment. Promising methods include transcranial magnetic stimulation (TMS) combined with functional imaging and real-time functional magnetic resonance imaging (rtfMRI), as well as other imaging and electrophysiological methods. Behavioral markers may be developed by "testing the limits" to evaluate baseline reserve (ability to improve performance given optimized conditions) and developmental reserve (ability to benefit by training).

    For monitoring circuit retraining, both neural measures and behavioral measures reflecting the focus of the circuit to be studied are critical.  Novel imaging methods, contrast probes and molecular probes are needed to extend our ability to better understand circuit-level human imaging measures at the cellular-molecular level.  Methods for standardizing how data are collected and published are needed.  Studies will need to distinguish changes resulting from intervention, development or aging, and recovery from injury.

    Retention/durability of the effects of interventions and their generalization are critical.  Thus, functional outcomes and quality of life assessments constitute an important third category of outcome measures. 

  5. Databases are needed for collecting, sharing and disseminating data among researchers.  Readily available patient registries would greatly facilitate recruitment.  As interventions are tested, databases might make available the details of randomized clinical trials.  For deep brain stimulation, a database of phenotypic data, precise brain stimulation sites and outcomes across disorders would move the field forward.  Since underlying biological vulnerabilities are critical for many diseases, databases determining associated genetic polymorphisms might facilitate the development of therapies aimed at altering protein functions to regulate enduring neuroplasticity, including drug-induced plasticity. 

  6. The development of novel interventions focused on inducing and harnessing neuroplasticity is needed.  This might involve a wide range of approaches, including a neuropharmacology for plasticity (ampakines, trophic factors, etc.), brain stimulation, behavioral training, cognitive training, physical exercise, and use of virtual-reality manipulations with simultaneous measurement of brain activation using rtfMRI to establish correlations between behavioral and brain physiological changes in parallel and in real time.  Multidimensional training strategies, e.g., that combine motor, perceptual and cognitive elements, the use of assistive devices (e.g., robotics) for motivating and maximizing training, and approaches that increase reserve and/or promote compensation in addition to blocking molecular events for pathogenesis are of interest.  Interventions that modulate the timing of critical periods, e.g., by promoting the maturation of inhibition (neurotrophins, GABAergic modulators), are also of interest.

    Interventions are needed across the spectrum of injury severities, including for severely impaired patients Currently, most trials focus on subjects with moderate impairments.  As the availability of residual or altered networks declines, the need for intrinsic biological manipulations of plasticity (cellular and neural repair strategies), in combination with training protocols, increases. 

    Once developed, scalable methods for delivering training to subjects, e.g., through the use of software and the Internet, would dramatically enhance dissemination at a low cost and offer potential for high impact, possibly worldwide, following initial development costs.

  7. More agile, mechanistic clinical trials are needed to advance neuroplasticity-based translational research.  Trials should incorporate both clinical and mechanistic endpoints, mediators, and moderators, so that we learn from both successful and unsuccessful trials. 

    Biomarkers are needed across the spectrum of biomarker science, e.g., clinical, behavioral, imaging, chemical, to monitor development and aging, as well as neuroplastic responses to conditions.  Basic, clinical, and translational research collaborations that begin in the design phase toward this end could be transformative. 

    Interventions might focus on circuit-level processes underlying domains of relevance to many disorders and types of rehabilitation, e.g., motor and sensory functional recovery, cognitive control, mood, motivation, declarative and procedural learning, using both behavioral and neural measures (structural and functional).  A circuit-based approach may enable the use of human diseases as models with which to compare and contrast the results of experimental manipulations to gain greater insights.  The disease may be less relevant than the neural pathways affected and spared. 

    There is a need for more strategic staging of pilot studies to resolve issues necessary for better and adequately powered randomized clinical trials.  Well-defined training protocols are needed to assess their potential utility.  Dose-response curves for experimental interventions are needed prior to larger trials.  The optimal timing and duration of treatment may be found by assessing gains during pilot studies at regular intervals of training to establish plateaus.  A pipeline for the initial screening of interventions might be geared toward moderate effect sizes and limited numbers of subjects (~ 20-25 per preliminary trial).  Well organized preliminary trials might be organized to make or break new interventions within a three-year period. 

    An important lesson of animal and human studies is that biological interventions, e.g., stimulation or pharmacology, are unlikely to lead to improvements unless conjoined with a practice paradigm for functional goals.   Such combined interventions require sophisticated pilot trial designs, however. 

    New statistical methods are needed to analyze variation in relationship to course/trajectories and treatment responses, particularly in developing populations and in those with converging deficits and comorbidities.

  8. A consortium of researchers (basic, clinical, and translational) might be formed to interact with one another iteratively to devise training interventions focused on neural circuits and related behavioral targets that share common denominators across diseases that serve as human platforms for circuit plasticity.  Results of experimental manipulations might be compared and contrasted across diseases to gain greater insights.  For example, manipulations of declarative and procedural learning are critical requirements for rehabilitative cueing, practice paradigms and retraining strategies such as for stroke and trauma, but could be further developed by techniques and insights developed by experts in maps of mood, motivation, aging and circuit decompensation.  Hypotheses and conceptual and practical issues could be addressed to induce, monitor, and measure plasticity-based therapies within the context of flexible pilot studies of interventions.  A shared or federated database might be developed to maximize the dissemination and utility of the results. 

    Such an approach might go beyond incremental studies and, instead, develop and optimally devise and test potentially robust interventions in sufficient numbers of subjects.  These collaborators might seek interactions with Clinical and Translational Science Award (CTSA), Research Centers in Minority Institutions (RCMI), and other research centers, disease-related foundations, and early career investigators to foster further interactions and progress. Partnerships of NIH, industry, community-based research entities, foundations, and other research funders might share target discovery and support clinical trials.