2021 NIH Alzheimer's Research Summit: Gaps and Opportunities

The NIH Alzheimer's Disease (AD) Research Summits are key strategic planning meetings tied to the implementation of the first goal of the National Plan to Address Alzheimer's Disease: to treat and prevent AD by 2025. They bring together a multi-stakeholder community, including government, industry, academia, private foundations, and patient advocacy groups to further integrated, translational Alzheimer’s research. The goal is to accelerate the development of effective, disease-modifying, and palliative therapies for cognitive as well as neuropsychiatric symptoms of Alzheimer’s disease.

The 2012, 2015 and 2018 Alzheimer’s Research Summits delivered gaps and opportunities that served, following review by relevant Institute and Department advisory councils, as the basis for developing research milestones. These milestones detail specific steps and success criteria for the NIH and other stakeholders toward the development of effective treatment and preventions for Alzheimer’s. The milestones span the entire Alzheimer’s research landscape, including basic, translational, clinical, and health services research, and serve as the basis for the development of the NIH Alzheimer's Disease Bypass Budget.

The goal of the 2021 Summit was to feature progress toward achieving the AD research implementation milestones and to continue the development of an integrated multidisciplinary research agenda necessary to enable precision medicine research and accelerate the development of successful therapies for AD.

More than 90 leading experts on Alzheimer's disease, other neurodegenerative disorders and other complex diseases joined forces as speakers and co-chairs. Through a multi-step, iterative process, these experts identified a series of gaps and opportunities that addressed the topics of each Summit Session.

The gaps and opportunities from the 2021 NIH Alzheimer's Research Summit refine and expand the AD/ADRD milestones research framework and will help guide all AD/ADRD research stakeholders across the public and private sectors, towards meeting the research goals set forth in the National Plan to Address Alzheimer's Disease.

The gaps and opportunities are organized around the Summit's seven programmatic sessions:

SESSION ONE: Deconstructing Disease Complexity - from Populations to Single Cells, from Genes to Multiscale Modes

1A. Expand GWAS studies in large non-European cohorts.

1B. Expand the generation of multi-omic data from globally representative populations, by supporting larger diverse cohort studies with systematic molecular characterization of the target tissue, cerebrospinal fluid (CSF) and blood so that brain-derived omic data can be linked to CSF/blood data to enable multi-scale modeling that is well-powered for discovery.

1C. Improve human tissue banking and distribution practices to improve reliability and reproducibility of human brain studies on brain aging and AD/ADRD; improve and standardize baseline tissue quality preparation and documentation standards.

1D. Continue and expand support for systems-based, integrative approaches that combine clinical endo-phenotyping, multi-omic and mechanistic research (using multiple experimental models) aimed at:

understanding the dynamic interaction between the brain and peripheral systems (metabolic, vascular, immune, etc.) during "normal" aging and AD/ADRD neurodegeneration.
the impact of sex differences on the molecular trajectories of brain aging and AD/ADRD to identify sex-specific factors of risk and resilience.
the molecular determinants of resilience in aging and in response to various types of AD/ADRD risk.

1E. Support comprehensive in vivo studies of healthy versus pathological aging in cellular, rodent and non-human primate models systems that reflect various subtypes of AD/ADRDs, guided by multiple clinical and molecular biomarkers to identify and characterize cell-autonomous and non-cell-autonomous mechanisms by which the brain interacts with peripheral organs with aging.

1F. Expand support for functional characterization of susceptibility variants to better understand the role of common/rare/private/somatic variation of AD/ADRD pathology, and single-cell activity.

1G. Develop research tools to better define the function of microglia, astrocytes and oligodendrocytes including their functional states and subtypes.

1H. Support the development of community-wide, open-source computational tools and pipelines for standardized data processing and analysis for diverse data types generated in AD/ADRD research, as well as their downstream integration in multi-scale models. These should come with clear documentation and best practices for application.

1I. Use Artificial Intelligence (AI) and deep learning for discovery and to improve subtyping of dementia, diagnosis, prognosis and personalized interventions. Include diverse global cohorts to ensure that AI predictive models work well and fairly for people of all ancestries and demographics.

1J. Develop links between the AD/ADRD research community and the research community supported by the BRAIN Initiative's Cell Census efforts to accelerate the development of a new armamentarium of cell type-selective molecular genetic tools, new model systems and gene therapy-based therapeutics.

SESSION TWO: Enabling Infrastructure and Incentives to Improve Research Rigor, Reproducibility and Translatability

2A. Provide support for on-demand resources for pre-aged mouse models of late-onset Alzheimer's disease (LOAD) being developed by the NIA-supported MODEL-AD (Model Organism Development & Evaluation for Late-Onset Alzheimer's Disease) Centers.

2B. Develop enabling infrastructure for non-human primate models (such as marmosets genetically engineered with Alzheimer's disease risk variants) similar to the resources for the development of mouse models of LOAD by the MODEL-AD Centers to bridge the translational gap between rodents and humans.

2C. Enable back translation of failed therapeutics in order to understand the drivers of failure in the clinic by providing access to clinical trials data and biosamples, and by evaluating failed therapeutics in a modern preclinical efficacy testing pipeline that applies new and emerging translationally-relevant measures including post-drug treatment transcriptomics, proteomics, and metabolomics.

2D. Rigor and reproducibility in clinical research requires understanding of which factors contribute to different results. To this end:

support engagement and participation of diverse participants in clinical research to fully understand the impact of social determinants of health (SDOH) as well as ancestry.
support uniform DNA assessment (minimum GWAS, maximum sequencing).
support standard assessment of plasma and CSF biomarker assays.
support uniform stool sample collection.

2E. Establish interoperability across disparate NIH-supported AD/ADRD data repositories by improving the data infrastructure and enabling data access through "passports" or similar authentication mechanisms to make it easier for investigators to:

identify where data is located
request data from multiple locations
make it easier to combine data from multiple sources.

2F. A global initiative is needed to overcome barriers to the sharing of human data (genetic and other types) to enable large international collaborations, due to different data sharing policies and restrictions [such as the General Data Protection Regulation (GDPR)].

2G. Funding agencies should require researchers to document the provenance of reagents and protocols to ensure that their claims are based on the use of well-validated research tools.

2H. Leveraging the success of Alzheimer's Preclinical Efficacy Database (AlzPED), build systems and processes for transparent sharing of experimental results for an expanded set of early-stage target validation and perturbation data.

2I. Establish training programs to increase research rigor in the practices of early-stage experimental scientists and develop all publicly-funded resources in a manner that promotes and supports researchers that are diverse in expertise, training, and resources.

SESSION THREE: Accelerating Therapy Development - Open Science from Targets to Trials

3A. Improve access to post-mortem neuropathology data and associated longitudinal antemortem data and biosamples and standardization across assays to enable more effective use of datasets generated by different research groups.

3B. Continue to build new, cross-disciplinary translational research teams/networks that use open science practices to accelerate the validation of emerging targets/therapeutic hypothesis and the preclinical to clinical translation of novel candidate therapeutics.

3C. To increase research rigor and accelerate translation, funding agencies should require researchers to provide evidence that they are working with validated reagents and protocols including validated research tools (antibodies, cell lines, measurement kits, internal standards), common criteria for evaluation of experiments and outcomes and timelines for data deposition/publication.

3D. Expand chemical biology approaches aimed at developing high quality, open-source probes necessary to interrogate cell-specific molecular mechanisms of AD/ADRD and to characterize cellular and molecular mechanisms of action of candidate treatments in preclinical animal models of AD/ADRD.

3E. Establish a publicly-funded AD drug development program that enables drug discovery and development work build on federally-funded innovation to be carried out entirely in the public domain with a goal to generate innovative new medicines that are priced fairly for all.

3F. Promote open drug development through regulatory change to allow for longer regulatory data protection for approved drugs (4 extra years) that would require:

sharing of all data including negative ones, as well as materials and tools
no patents that would prevent anyone else manufacturing and selling the product
reasonable price cap.

SESSION FOUR: Diversifying the Therapeutic Pipeline to Develop Precision Medicines

4A. Expand the infrastructure for collecting, aggregating and analyzing multi-omic and deep endophenotyping data derived from diverse (ethnicity, sex, age) participants.

4B. Enhance the ability to collect, store, analyze and share specimens, images and as much biomarker data as possible from clinical trials.

4C. Continue to support the discovery of therapeutic targets for diverse biological domains and expand support for rapid and rigorous preclinical validation of novel targets emerging from large-scale systems biology efforts.

4D. Maintain a broad perspective of candidate therapeutic strategies including reversal or regenerative therapeutics for the treatment of advanced disease.

4E. Advance the discovery of resilience-based precision therapeutics through the development of cross-species integrative computational models and the use of bio realistic 3D tissue culture models.

4F. Support the development of novel ex vivo and in vitro models (iPSC-based, organoid, etc.) derived from diverse populations for use in precision medicine research and in therapy development.

4G. Accelerate therapy development for emerging targets (including network modules) by developing translatable biomarkers evaluated in preclinical animal models of LOAD and in human peripheral fluid samples from well-phenotyped representative cohorts.

4H. Enable more facile clinical development by using target engagement endpoints versus clinical endpoints.

4I. Create opportunities and avenues for closer interaction between researchers developing new INDs and researchers supported NIA's enabling infrastructure programs involved in developing data resources, translational tools and new mechanistic insights about the complex biology of the disease.

4J. Evaluate and update current trial funding models to ensure the ability of small biopharma companies to test novel first-in-class therapeutics in the context of the full range of expenses and time frames required for robust evaluation and sharing.

4K. Expand support for mechanisms to rapidly advance clinical testing similar to concepts now routine in the oncology field ("first in patient", basket trial design concepts, test multiple agents/hypothesis in patients at once).

SESSION FIVE: Emerging Biomarkers Landscape

5A. Expand support for positron emission tomography (PET) radiotracers for AD/ADRD pathologies and emerging targets; critically needed are PET radiotracers for alpha-synuclein, subtype-selective tau-PET radiotracers and PET ligands for neuroinflammation.

5B. Support research with advanced MRI modalities to help define vascular and other structural and functional pathophysiological features.

5C. Support systematic evaluation of AT(N) and other emerging biomarkers (including omics biomarkers) within the context of environmental, socio-cultural and behavioral factors across multi-ethnic populations to enable a true precision medicine approach to treatment and prevention.

5D. Expand the AT(N) framework into a more comprehensive framework that can take into account ADRDs. Testing expanded AT(N) models within the context of multiple neurodegenerative disease and across diverse populations will be key to advancing precision medicine for AD/ADRDs.

5E. Develop a "hands free" freesurfer-like, open-source digital data processing pipeline.

5F. Develop data science/AI methods to generate dynamic digital indices as the next generation of digital biomarkers to open the path for new FDA gold standards and break free from the current digital biomarker validation.

5G. Support the use of passive monitoring devices (e.g., voice assistants) to develop clinically validated biomarkers that can be used to identify prodromal disease states.

5H. Wearable and home sensor devices are fundamental for clinical trials. Support the validation of different measures in the patient populations of interest and the development of the technology needed to register devices, safely collect, and store device data, and the analytical tools.

5I. Support validation studies of blood biomarkers as outcomes in clinical trials to predict clinical efficacy.

5J. Identify and develop cell-specific biomarkers of disease in CSF and plasma linked to brain pathology.

5K. Expand multi-omic strategies to identify and characterize fluid and imaging translatable biomarkers that can be utilized in pre-clinical testing in animal models and in human clinical trials.

5L. Develop diagnostic and prognostic algorithms based on easily accessible and cost- and time-effective tests (such as blood biomarkers and brief cognitive tests) to:

differentiate AD dementia from other dementias (for clinical practice and for clinical trials)
predict cognitive decline and AD dementia in patients with mild cognitive impairment (for clinical practice and clinical trials) and identify asymptomatic/pre-clinical AD (for clinical trials).

5M. Enable close collaboration between NIH and FDA to ensure that biomarker development translates to therapeutic development.

SESSION SIX: Advancing Drug Repurposing and Combination Therapy Development

6A. Continue and expand support for a variety of translational bioinformatics approaches (network medicine, network pharmacology, deep learning) combined with conducting deep molecular endo-phenotyping across diverse cohorts to accelerate rational drug repurposing and the development of multi-target drugs, and drug combinations.

6B. The absence of noncoding RNA in current drug repurposing databases such as iLINCS and CLUE limits the opportunities for drug repurposing for AD/ADRD; support the development of iLINC-type data resources that utilize relevant disease related cell types and full transcriptome profiling.

6C. Enable better access to electronic medical records (EMR) data, including EMR data from the Veterans Administration Health Initiative Care System and the All of Us to aid the identification and validation of FDA-approved, repurposable drugs and the identification of responder vs. non-responder phenotypes.

6D. Support deep molecular profiling of biosamples from non-pharmacologic clinical trials to identify responder phenotypes and to inform the rational development of therapeutic strategies that combine pharmacologic and non-pharmacologic interventions.

6E. Overcome remaining barriers to sharing patient-level data and biosamples from failed AD/ADRD clinical trials including trials from industry sponsors, to enable the identification of responder phenotypes and repositioning of failed drugs to the right patient population.

6F. Support the development of a set of harmonized outcomes to be used across trials to enable comparative effectiveness analyses to accelerate the identification of interventions (or combinations of interventions) that are most effective and for whom.

6G. Support research focused on identifying effective strategies to modify behavior to increase adherence and compliance in trials and best ways to support and maintain behavior modification and participant engagement across diverse populations, after the trial is complete.

6H. Support research aimed at discovering the optimal ways to integrate digital technologies into current and future trials and optimize the analysis and interpretation of the collected data.

6I. Establish a national working group to provide evidence-based guidelines for the development of protocols for delivery of non-pharmacologic interventions.

6J. Expand support for open science practices for sharing data and analytical tools.

SESSION SEVEN: Understanding the Impact of the Exposome on Brain health to Advance Disease Prevention

7A. Develop a precision environmental health approach for AD/ADRD prevention (i.e., an individualized risk assessment and interventions to prevent disease). Identify windows of susceptibility prior to disease pathology and clinical onset and characterize the molecular intermediates linking exposures and dementia to provide mechanistic evidence that can be translated into disease prevention strategies.

7B. Leverage exposome AD/ADRD studies to address health disparities research questions, including the impact of structural racism and inequity across the life course. Develop and promote practical and effective ways to improve population-representativeness of AD/ADRD research resources, ensuring wide-spread inclusivity across all groups, including NIH-designated priority populations.

7C. Enable comparative analysis of the exposome within and among populations to understand the impacts of exposures on disease outcomes and to gain deeper understanding of the functional biology linking those exposures to disease outcomes.

7D. Invest in the development of surrogate measures of the impact of the exposome on brain aging, AD/ADRD, in easily obtainable biosamples (hair, blood, and saliva).

7E. Examine the role of sex/gender on determining the exposome and modifying the response to the exposome in AD/ADRD.

7F. Expand research on the role of pollution on dementia risk. Embed measurements of pollution in cohorts with uniform and regular dementia evaluations. Understand the timing aspect of the effects of pollution on dementia risk. Utilize newly available data on weather patterns and climate change and the interplay of this on pollution exposure and ADRD.

7G. Develop methods/technologies to directly assess diet from a patient specimen (e.g., from the microbiome or metabolome) rather than indirectly and inaccurately from diet records.

7H. Leverage existing consortia and data resources to identify known diet-AD relationships (positive or negative) that likely have microbiome underpinnings, and devise strategies to test these mechanistically.

7I. Develop AD-focused multi-omics data resources that include molecular read-outs of microbial activity that can be integrated with host genetics data to enable inference of causal relationships between AD risk loci and microbial observations.

7J. Enable integration of molecular read-outs of microbial with longitudinal clinical records, to enable unbiased and precise identification of microbial perturbations that may predispose individuals to dementia.

7K. Support the development of mobile health technologies to address the lack of true measures of the exposome (i.e., all exposures across the lifecourse) and the lack of longitudinal objective data on cognition at scale.

7L. Develop and characterize polygenic animal models for LOAD and related dementias that faithfully reflect the heterogenous genetic and environmental etiology of these disorders including the impact of the exposome and diverse genetic backgrounds.

7M. Support multi-omic profiling of the pan-exposome in 3D cerebral organoids. Systematic perturbation of iPSC-derived cerebral organoids from AD cases and controls, with a range of exogenous factors (i.e., pathogens, toxicants, pollutants, etc.) to characterize the molecular, cellular and neuropathological networks that are perturbed by each factor, would enable the detection of shared, as well as distinct mechanisms that are associated with different aspects of the exposome.

7N. Develop open and democratized data resources and FAIR data practices to increase the reach of AD/ADRD-relevant exposome research across a wide-array of public, academic, industry, government and other stakeholder groups. Encourage the adoption of exposomics-related ontologies for data interoperability.