Alzheimer's Disease Summit: The Path to 2025

Resources

Worldwide impact of Alzheimer's disease and government responses

Alzheimer's Disease International. World Alzheimer Report 2013: Journey of Caring — An Analysis of Long-term Care for Dementia.
A report researched and authored by Martin Prince, Matthew Prina, and Maëlenn Guerchet on behalf of the Global Observatory for Ageing and Dementia Care, hosted at the Health Service and Population Research Department, King's College London.

Prince M, Bryce R, Albanese E, et al. The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement. 2013;9(1):63-75.

Rosow K, Holzapfel A, Karlawish JH, et al. Countrywide strategic plans on Alzheimer's disease: developing a framework for the international battle against Alzheimer's disease. Alzheimers Dement. 2011;7(6):615-21.

Wimo A, Jonsson L, Bond J, et al. The worldwide economic impact of dementia 2010. Alzheimers Dement. 2013;9(1):1-11.

Wortmann M. Importance of national plans for Alzheimer's disease and dementia. Alzheimers Res Ther. 2013;5(5):40. [Epub ahead of print]

Diversifying therapeutic approaches for Alzheimer's disease

Brundin P, Melki R, Kopito R. Prion-like transmission of protein aggregates in neurodegenerative diseases. Nat Rev Mol Cell Biol. 2010;11(4):301-7.

Herrup K, Carrillo MC, Schenk D, et al. Beyond amyloid: getting real about nonamyloid targets in Alzheimer's disease. Alzheimers Dement. 2013;9(4):452-58.

Zhang B, Gaiteri C, Bodea LG, et al. Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer's disease. Cell. 2013;153(3):707-20.

Peer review and open access

Berman F, Cerf V. Who will pay for public access to research data? Science. 2013;341(6146):616-7.

Bohannon J. Who's afraid of peer review? Science. 2013;342(6154):60-5.

Franzoni C, Scellato G, Stephan P. Changing incentives to publish. Science. 2011;333(6043):702-3.

Joseph H. The open access movement grows up: taking stock of a revolution. PLoS Biol. 2013;11(10):e1001686.

Biomarkers

Blennow K, Hampel H, Zetterberg H. Biomarkers in amyloid-β immunotherapy trials in Alzheimer's disease. Neuropsychopharmacology. 2014;39(1):190-201.

Jack CR Jr, Knopman DS, Jagust WJ, et al. Hypothetical model of dynamic biomarkers of the Alzheimer's pathological cascade. Lancet Neurol. 2010;9(1):119-28.

Jack CR Jr, Knopman DS, Jagust WJ, et al. Tracking pathophysiological process in Alzheimer's disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12(2):207-16.

Public–private partnerships

Ekins S, Waller CL, Bradley MP, et al. Four disruptive strategies for removing drug discovery bottlenecks. Drug Discov Today. 2012;18(5-6):265-71.

Goldman M, Compton C, Mittleman BB. Public-private partnerships as driving forces in the quest for innovative medicines. Clin Transl Med. 2013;2(1):2.

Snyder HM, Bain LJ, Egge R, Carrillo MC. Alzheimer's disease public-private partnerships: a landscape of the global nonprofit community. Alzheimers Dement. 2013;9(4):466-71.

IT and big data

Erten-Lyons D, Sherbakov LO, Piccinin AM, et al. Review of selected databases of longitudinal aging studies. Alzheimers Dement. 2012;8(6):584-9.

Khachaturian AS, Meranus DH, Kukull WA, et al. Big data, aging and dementia: pathways for international harmonization on data sharing. Alzheimers Dement. 2013;9(5 Suppl):S61-2.

Swarup V, Geschwind DH. Alzheimer's disease: From big data to mechanism. Nature. 2013;500(7460):34-5.

Globalization and data sharing

Cummings J, Reynders R, Zhong K. Globalization of Alzheimer's disease clinical trials. Alzheimers Res Ther. 2011;3(4):24.

Kaye J, Hawkins N. Data sharing policy design for consortia: challenge for sustainability. Genome Med. 2014;6(1):4. [Epub ahead of print]

Khachaturian ZS, Petersen RC, Snyder PJ, et al. Developing a global strategy to prevent Alzheimer's disease: Leon Thal symposium 2010. Alzheimers Dement. 2011;7(2):127-32.

Sponsors

Gold Sponsors

Silver Sponsors

Bronze Sponsors

Presented by

Introduction

The U.S. National Alzheimer's Plan established goals across five domains, focused on improving prevention and treatment, enhancing care, expanding support systems, raising awareness, and tracking progress. One goal in particular—to determine how to prevent and treat Alzheimer's disease (AD) by 2025—transcends geographical boundaries. Increasing numbers of people are affected by AD worldwide, especially in low- and middle-income countries, as Bengt Winblad of the Karolinska Institute described. Governments are beginning to respond by developing national plans outlining how to care for patients and advance research and drug development to stem the rising tide.

According to Richard J. Hodes of the National Institute on Aging, NIH, the U.S. approach is to systematically solve problems by developing multi-stakeholder partnerships. In Japan, which has one of the world's oldest populations, the government focuses on promoting healthy and active aging, as described by Yuko Harayama of Japan's Council for Science and Technology Policy. Both plans reflect the increasingly prominent idea in the international Alzheimer's research community that business as usual will not adequately address the size, scope, and complexity of the AD challenge. Instead, a "big science" approach, akin to other massive multinational collaborative projects such as the Manhattan Project, will be needed, with multiple stakeholders coalescing around clear goals with shared risks and benefits.

The number of people with dementia is projected to increase dramatically by 2050, particularly in low- and middle-income countries. (Image courtesy of Bengt Winblad)

It will, however, be difficult to reach the 2025 goal. Despite massive investment by governments and industry, no new drugs have been approved in the last decade, as Jan M. Lundberg of Eli Lilly noted. Moreover, the pathophysiology of AD remains controversial and no biomarker has been proven to represent a valid surrogate endpoint. Biomarker research has shown that the disease process begins decades before symptoms appear, suggesting that treatment must be given in the early, pre-symptomatic stages to slow disease progression. Yet conducting clinical trials in the at-risk population presents additional challenges, requiring long, expensive, and risky trials.

Most drug development in AD has focused on β-amyloid, the protein that aggregates and forms plaques in the AD brain. β-amyloid is thought to trigger the disease, but none of the treatments that target amyloid have yet proven effective. A second protein, tau, aggregates as neurofibrillary tangles in the AD brain and is thought to contribute to neurodegeneration. Recent evidence suggests that tau may spread from cell to cell throughout the brain. Amyloid and tau do not, however, fully explain the pathogenesis of the disease. Inflammation also appears to be necessary for the disease to develop, and genetic studies have implicated other physiological pathways. Moreover, AD is only one of several related types of neurodegenerative disorders that cause dementia. The heterogeneity of neurodegeneration is further complicated by comorbidities: most elderly people with AD or ADRD have other conditions that contribute to neurological disease. These uncertainties about the biological basis of AD create difficulties in the drug development process.

Multiple mechanisms contribute to Alzheimer's disease pathogenesis, complicating the search for treatments. (Image courtesy of Angel Cedazo-Minguez and Bengt Winblad)

The Path to 2025 Summit was organized as a series of panels on diverse but interrelated questions that must be resolved so that the field can move forward quickly:

• How can we diversify the range of therapeutic or mechanistic approaches being explored, with the ultimate goal of identifying new drug targets and more effective drugs?
• What do we need to do to develop more innovative platforms and tools, including better biomarkers, to accelerate drug development?
• How can we shrink the timeline and de-risk the drug development process?
• How can we foster innovation and outside-the-box thinking across all phases of drug development? How can we improve funding mechanisms?
• What are the most effective ways to promote global collaboration across multiple sectors? Do we need additional public–private partnerships?
• How can we interrogate the massive amount of data being generated to learn more about how AD progresses and to identify opportunities for interventions?

Highlights

A new paradigm for clinical trials is needed, focused on earlier stages of disease and incorporating novel trial designs, such as adaptive trials, with embedded biomarker studies.

Large global registries are needed to run longitudinal natural history studies, develop and validate biomarkers and outcome measures, and identify candidates for clinical trials.

Data sharing and analyses will allow the results from clinical, epidemiological, and biomarker studies to advance our understanding of the disease process and of the effectiveness of risk-reduction measures and treatments.

Innovative trial designs: learning quickly and adaptively

The opening panel discussion, moderated by Simon Lovestone from King's College London, focused on lessons learned in industry from previous clinical trials. Clinical trials are slow and inefficient. Howard Feldman from the University of British Columbia and other panelists proposed methods to accelerate trials, increase efficiency, and ensure that mistakes are not repeated. Trial designs with adaptive features enable investigators to learn from experience and modify the trial as it progresses, matching treatments to diagnoses and target populations. Adaptive trials would create a model Michael Krams of Janssen Pharmaceuticals called "solution-centric." This model would move the field away from a "compound-centric" approach, focused on a single compound or a single drug target, to instead look at multiple drugs, mechanisms, and stages of disease. Adaptive trials would also allow multiple drugs to be tested in combination, and Reisa A. Sperling of Harvard Medical School predicted that such combination testing will be needed to achieve better treatment of AD.

Regardless of trial design, data from different trials must be coordinated and integrated so that researchers can learn more about treatment effectiveness and the meaningfulness of outcome measures. However, several roadblocks have prevented the sharing and analysis of data across different studies, including a general reluctance by companies to release data and a lack of standardization in data collection. "We are wasting huge amounts of money because we can't use data to learn anything," said Richard A. Rudick of the Cleveland Clinic Lerner College of Medicine. Sharing data also facilitates efforts to optimize biomarkers. Indeed, the panelists called for extensive study of biomarkers—including markers of target engagement, downstream effects of drugs, and theragnostic (biochemical) effects of drugs—to be embedded in all trials.

Shifting to earlier stages of disease

Most trials have enrolled people with mild to moderate AD; however, this stage of disease is thought by many to be too late because substantial neurodegeneration has already occurred. Some trials included individuals with mild cognitive impairment (MCI), but treatment may be ineffective even at this stage. The AD field is moving toward secondary prevention, which aims to prevent the manifestation of symptoms in people who are at high risk of developing AD or in whom the disease process has already started. However, the early stages of AD are difficult to diagnose without significant cognitive impairment, and it is therefore difficult to demonstrate a clinically relevant treatment effect. Surrogate biomarkers that correlate with a clinical effect could provide a solution, as Eric Siemers of Eli Lilly mentioned, but none have yet been identified or validated.

Nicholas A. Kozauer of the U.S. Food and Drug Administration (FDA) described draft guidance from the agency outlining steps for conducting trials in early-stage populations. In the majority of AD drug trials, treatments must demonstrate both an effect on cognition and a global measure of function. The new guidance raises the possibility of using a different, accelerated approval process for early-stage AD drugs that would grant preliminary approval to a drug based only on the demonstration of a clinical (cognitive) effect in combination with a positive biomarker result. Post-marketing studies would be required to demonstrate the clinical meaningfulness of the effects of the drug.

Engaging people in the fight

Demonstrating treatment effectiveness in early disease stages will require large and lengthy trials with people enrolled without cognitive and functional impairments. Who will step forward to participate in such trials? The panelists suggested we need large-scale global registries of hundreds of thousands of individuals. For such registries to be created, AD will need to become prominent on the public agenda, based on increased awareness of the urgency of the problem. Ronald C. Petersen of the Mayo Clinic reported that efforts to create such registries are underway. He and others have suggested that each country should not only have a national plan but also direct sustainable funding (perhaps 1% of their national dementia costs) to Alzheimer's research and care.

Cohorts that emerge from the registries could be established as natural history studies with embedded biomarker and cognitive assessments. Individuals would be asked to agree to share their medical data in future research studies (with appropriate privacy protections) and to be willing to participate in future clinical trials. Subsets of people from these cohorts could then be enrolled in pharmacological trials. This model has already been implemented in some parts of the UK and is envisioned on a grander scale across Europe. This approach would establish trial-ready cohorts across the spectrum of AD and shift the focus of research from patients to the general population.

Highlights

Although most drugs in late-stage development target amyloid and tau, many other potential targets could offer new therapeutic approaches.

Research and development efforts aimed at prevention, symptomatic treatment, or disease modification each require different tools, strategies, and endpoints.

Platforms should be built to link genetic changes to functional changes and to take a systems biology approach to understanding AD.

Amyloid and tau

The panel moderated by Christopher P. Austin of the National Center for Advancing Translational Sciences, NIH, explored how to build strong hypotheses leading to new therapies and treatments. The majority of therapies in late-stage development target amyloid or tau. Darryle D. Schoepp of Merck reported that despite the attention paid to these mechanisms, there is still much to learn. Amyloid, for example, exists in soluble and many different fibrillar forms, including the form that aggregates into plaques in the brain; yet researchers do not know which of these forms correlates with disease severity. Amyloid may trigger AD or drive disease progression. In either case, there may be a threshold of β-amyloid accumulation below which a person will not manifest the disease. It is unknown whether removing plaque is a viable strategy for treating AD. Trials investigating this question—such as a study of Alzhemed, an agent that binds to soluble amyloid to prevent the formation of plaques—were inconclusive, presumably because the drug failed to adequately engage the target.

Claudio Soto of the University of Texas Medical School at Houston explained an emerging concept in the study of the molecular basis of neurodegenerative diseases that postulates that protein misfolding triggers disease. Misfolded proteins called "prions" have been shown to cause a group of rare neurodegenerative diseases known as transmissible spongiform encephalopathies (TSEs). In TSEs, the misfolded protein acts as a seed or nucleus that spreads the disease throughout the brain by converting other proteins into misfolded forms. Evidence suggests that this phenomenon may occur with the proteins linked to more common neurodegenerative diseases: amyloid and tau in AD, α-synuclein in Parkinson's disease (PD), and huntingtin in Huntington's disease (HD). Although AD, PD, HD, and TSEs share a similar molecular mechanism of protein misfolding and nucleation, the proteins themselves, their regional distribution in the brain, and the clinical manifestations of the diseases are different. Importantly, however, the prion hypothesis suggests that blocking the spread of protein misfolding could stop disease progression.

Moving beyond amyloid and tau

Disappointing results from dozens of drug trials focused on amyloid have convinced many researchers of the need to move beyond amyloid and tau and consider other targets, including neurotransmitters and synaptic and metabolic proteins. Moreover, according to Dale Schenk of Prothena, prevention, treatment, and disease modification are different outcomes and may require different interventions. A preventative measure may not work as a treatment, for example.

Recent genetic discoveries point to new means to treat patients with AD, reported Rudolph E. Tanzi of Harvard Medical School. Newly discovered genes indicate that the innate immune system turns against the brain, suggesting that therapies targeting the immune system could be effective as treatments. Genome-wide association studies (GWAS) have also found genes that are associated with AD, suggesting novel avenues of treatment. However, GWAS alone can only provide information about which genes are altered at different stages of the disease. To understand how this alteration translates into pathology, the genes must be linked to functional changes. Tanzi recommended increased efforts to identify functional variants and map the variants to different stages of disease via gene sequencing approaches.

Genetics cannot fully explain a complex disease such as AD, explained Lefkos T. Middleton of Imperial College London. Genes involved in AD, along with those that are implicated in normal aging, are influenced by exposure to diet and other lifestyle components. Studies on the effects of lifestyle, diet, education, and cardiovascular risk factors are underway at various stages of disease. Recent studies on reduced AD prevalence in the UK suggest that this effect is attributable to public health campaigns to promote control of blood pressure, type 2 diabetes, and other risk factors.

The panelists, as well as other conference speakers in the audience, advocated the development of new conceptual models of disease taking a systems biology or systems neuroscience approach. These models must incorporate the multigenic nature of AD as well as the importance of the synapse, because loss of synapses correlates with dementia. Only with these models will the field be in a better position to understand AD and its complexity.

Highlights

Open, transparent, and interdisciplinary approaches to science and publishing will expand the field, bring in new concepts and tools, and encourage creative thinking.

Incentives and rewards for research should be reevaluated to promote collaboration and sharing.

A new platform could be created to engage potential volunteers for clinical trials.

Open science: sharing data, tools, and insights

The audacious 2025 goals will require new approaches to research to drive innovation, harness the ingenuity of the public, and achieve more with available funds. Large-scale public–private partnerships built in response to other grand challenges such as sending a man to the moon achieved success by engaging experts across multiple disciplines and encouraging out-of-the-box thinking. Tetsuyuki Maruyama of Takeda Pharmaceutical Company said that AD represents a similar grand challenge. To solve it, he called for more transparency and collaboration, as well as disruptive approaches that challenge existing research paradigms. The panel moderated by Cristin Dorgelo of the Office of Science and Technology Policy considered how to shake up the peer-review system and develop new research methods.

Open science is gaining a foothold in neuroscience. The Allen Institute for Brain Science has demonstrated that sharing tools and data accelerates discovery. The institute's Allan Jones noted that researchers must go further by working to organize and standardize data and tools to make information more useful and accessible. Sharing insight is perhaps the most important but difficult goal to achieve, requiring changes in culture and attitude as well as a management approach that incentivizes pre-publication data sharing.

Overhauling the peer-review system

Gustavo Stolovitzky of IBM Research and Ijad Madisch of ResearchGATE echoed the call to restructure scientific publishing. The peer-review system rewards scientific orthodoxy, not novel ideas, and cannot deal with large datasets and computational technologies. A process for rapid post-publication feedback, now missing, would spur innovation. ResearchGATE is a scientific social network providing a platform for multidisciplinary commentary on research papers. Researchers can publish negative results, share raw data, and flag papers if there are questions about the reproducibility and interpretation of results. Stolovitzky suggested allowing reviewers not only to passively review a paper but also to access the data and computational tools employed by the authors to determine whether the conclusions are warranted.

Stolovitzky also suggested crowdsourcing as a new means of scientific research. The DREAM (Dialogue for Reverse Engineering Assessments and Methods) project employs this approach, posing a series of questions with relevant data and inviting people to propose solutions. The Alzheimer's Disease Big Data DREAM Challenge #1, to be launched in the first half of 2014, was announced at this Summit. Established in partnership with the Global CEO Initiative on Alzheimer's Disease and Sage Bionetworks, the challenge focuses on finding better predictors of AD risk and protective factors in Alzheimer's Disease Neuroimaging Initiative (ADNI) populations. Participants in the challenge are asked to build predictive models using cognitive, imaging, genomic, and other data from a large cohort of volunteers. The top-performing teams will be invited to present their data and co-author a paper. The idea behind the DREAM challenges, according to Stolovitzky, is to promote collaboration, democratize data, and accelerate knowledge extraction.

A new platform for clinical trial recruitment

The current system for AD clinical trials is broken: trials are too big, take too long, cost too much, and have so far failed to identify an effective treatment. Michael T. Ropacki of Janssen Alzheimer Immunotherapy proposed a new registry model he calls the Registry Recycling Model, which would create a registry of minimally characterized potential trial volunteers in the pre-competitive space.

Beginning with this broadly heterogeneous group, researchers collect longitudinal data (including neuropsychological testing and biomarkers) and identify subjects showing early signs of cognitive impairment to assign to appropriate trials. If subjects assigned to a trial are deemed ineligible, they are "recycled" to the larger registry, where they remain until eligible for recruitment to a future trial. This model thus retains volunteers with baseline and longitudinal data so that subsequent trials can avoid these early recruitment steps and launch more quickly and efficiently. In addition, the baseline data allow investigators to observe changes in performance before a trial begins, making it possible to select a more homogeneous group with early signs of decline. The traditional approach, which selects subjects based on memory test scores without regard for baseline performance, results in greater heterogeneity within each trial and highly variable responses to treatment.

The proposed new approach accomplishes more than funneling volunteers to appropriate clinical trials. Data collected on subjects before enrollment in a trial can be used as run-in data to show disease progression before treatment is initiated. In addition, these data can be used to test and validate neuropsychological measures, composites, screening measures, and biomarkers in the pre-competitive space, lowering the cost and improving the efficiency of future trials.

Highlights

Novel functional biomarkers should be embedded into prevention trials, with the goal of identifying a biomarker that can provide a readout of functional changes over a short period of time.

Imaging and cerebrospinal fluid (CSF) biomarker data from all trials should be collected into a common, open database.

Bioinformatics and biostatistical platforms should be built to integrate data from biomarker studies to better understand how biomarkers can stratify patients, predict progression, and assess risk.

Biomarkers in clinical trials

The search for effective treatments for AD and related disorders requires biomarkers at multiple stages and for multiple purposes, including to identify appropriate subjects for a trial; to indicate the stage of disease, degree of neurodegeneration, and presence and extent of comorbid pathologies; to show whether a drug has engaged the target for which it was intended; and to indicate a treatment effect. The panel moderated by Jeffrey S. Nye of Janssen Research & Development aimed to identify the characteristics of a better biomarker.

According to David M. Holtzman of Washington University, existing biomarkers can determine reasonably well whether pathology is present and who is likely to convert from cognitively normal to impaired. The challenge, however, is to develop functional biomarkers to embed in clinical trials that could show whether a treatment improves brain function or reduces degeneration.

Clinical trials require standardized procedures for data collection and analyses, and regulations governing these standards need to be aligned between different countries, explained Diane Stephenson from the Critical Path Institute (C-Path). Without standardization, there is no way to combine or compare data from studies of disease progression. Typical biomarker results also need to be defined for different populations (groups defined by age, gender, race, disease stage, and other factors) so that outliers can be identified, and data should be collected to demonstrate how changes in biomarker values can be used diagnostically or prognostically. Kaj Blennow of the University of Gothenburg pointed to substantial progress toward global standardization. These efforts have also stimulated biotech companies to develop fully automated analysis tools to perform biomarker assessments.

Novel biomarkers

The development of agents for imaging tau pathology is a major advance in the quest for AD biomarkers. Tau protein accumulates in the brain in AD and several related disorders, and the process is considered a sign of neurodegeneration. Early studies suggest tau levels in the brain correlate with cognitive decline; thus, tau imaging may provide a surrogate marker to track neurodegeneration and cognitive impairment.

Markers of other pathological proteins are also needed, according to John Q. Trojanowski of the University of Pennsylvania. While β-amyloid and tau are the best known proteins, others such as α-synuclein and TDP-43 are also important. Moreover, all of these proteins can exist in multiple forms: cleaved into different sized fragments, chemically modified, aggregated, or soluble. We do not yet know the optimal variants of various proteins or how well they correlate with disease outcomes.

Several physiological pathways, such as the inflammatory pathway, are important in neurodegenerative disease, and markers of the inflammatory response could be very informative. Because the culminating event linking pathology to cognitive impairment appears to be loss of synapses, synaptic markers rank high on the list of critical biomarkers.

Clifford R. Jack Jr. from the Mayo Clinic cautioned against thinking too simplistically about biomarkers. As AD progresses, multiple pathogenic mechanisms are engaged at different stages and with different degrees of severity, and biomarkers are needed for each mechanism and stage. Most elderly people have concurrent conditions, so biomarkers for heart disease, diabetes, and other diseases should also be assessed to understand all the factors contributing to an individual's symptoms. He recommended beginning with easily accessible and economical tests, followed by more expensive and invasive tests for people in high-risk pools based on clinical and genetic factors.

The regulatory point of view

In the short term, researchers aim to identify a biomarker or biomarkers that regulatory agencies would accept as a means to track disease progression, and to identify an enrichment biomarker or biomarkers that could be used to ensure that studies enroll appropriate candidates. For example, the European Medicines Agency (EMA) has already approved several enrichment biomarkers that suggest a person may be in the early stages of AD. However, the agency's Maria Isaac noted that a disease progression biomarker must meet much higher standards than an enrichment biomarker. From the regulatory point of view, a treatment effect must be shown to be clinically meaningful.

Highlights

Biomarkers that predict clinical efficacy must be identified more quickly. To accelerate the process, companies should share data on active treatment and placebo arms from interventional studies.

Alternative therapeutic approaches such as regeneration, neurostimulation, and medical foods should be explored.

Crowdsourcing can be used to gather data about off-label drugs, over-the-counter drugs, and non-pharmacologic therapies used by people with ADRD.

Using 21st century tools for clinical trials

The panel moderated by Howard Feldman looked at new approaches to reduce the clinical trial timeline. Innovation, according to Bernard H. Munos of InnoThink, requires bold thinking and coordination. To encourage both, he suggested inviting physicists, mathematicians, researchers from diverse medical disciplines, and experts from other fields to join the discussion. He also advocated using 21st century tools such as the Internet, data mining, biosensors, and smartphone applications to collect data from millions of people. These data could provide a more complete understanding of the natural history of diseases such as AD.

Michael W. Weiner from the University of California, San Francisco, suggested that the Internet is best used to build cohorts and conduct rough assessments to identify candidates for clinical trials. His team has established the Brain Health Registry in partnership with Lumos Labs (creators of the Lumosity "brain games") to gather longitudinal cognitive data on thousands of people. Data gathering alone, however, does not provide enough information to build populations for clinical trials, added Samuel E. Gandy from the Mount Sinai School of Medicine. Researchers also need novel computational approaches that can integrate diverse types of data from many sources.

Investigating novel therapeutic strategies and disease mechanisms

Innovation means more than simply using the latest technologies, as Gregory A. Petsko from Weill Cornell Medical College pointed out. Novel therapeutic strategies and alternative mechanisms of disease should be investigated, and we need a better understanding not only of neurodegenerative diseases but also of metabolic and other types of disease. For example, there is evidence that abnormal protein aggregation may underlie AD, suggesting that other diseases involving cells that are unable to eliminate unwanted proteins may provide clues about mechanisms of AD. Experts in diseases that may have relevance for AD, such as lipid diseases and lysosomal storage diseases, could provide guidance for applying alternative strategies to AD research.

Many physiological pathways may be altered in the AD brain. For example, neurogenesis has been shown to be impaired in several mouse models of AD, suggesting that pro-neurogenic compounds could be beneficial. Evidence also suggests that exercise induces the production of brain-derived neurotrophic factor (BDNF), a nerve growth factor that may be the link between increased exercise and better brain health. Other possible approaches to treatment include neurostimulation and drugs that enhance neural plasticity.

Andrea Pfeifer of AC Immune and other panelists also championed combination therapy, which could target multiple parts of a single pathway or multiple pathways. Molecules have been identified that target both β-amyloid and tau, for example, but questions about combination therapy remain. Which molecules work in combination? What type of animal studies are needed to define the best combinations? How should a combination therapy trial be conducted? How can target engagement be assessed? Which biomarkers are relevant? And how can the safety of drugs used in combination be assessed? Combination therapy could also include non-pharmacologic lifestyle approaches such as diet and exercise.

Accelerating the identification of biomarkers that predict clinical efficacy

Across several disease areas, the main barrier to drug development is our inability to predict, in phase II studies, which compounds will ultimately prove effective. Gary Romano of Johnson & Johnson suggested that the most effective innovation and the quickest way to solve this problem would be to share data from both active treatment and placebo arms of a trial. He suggested finding out why companies do not share these data and engaging legal and intellectual property experts to find solutions.

The panelists agreed that the main barrier to investigating out-of-the-box ideas is lack of funding in general, and particularly, the tendency of agencies and companies to fund more conventional approaches. However, the world has faced similar challenges with HIV/AIDS and other health problems with huge unmet needs and severe economic and public health consequences. With the support of HIV/AIDS patient and advocacy groups, research organizations and companies developed procedures that enabled data sharing with adequate privacy protections for patients and providers.

Highlights

Alzheimer's disease is a huge social problem with a drag on economic growth. The finance industry is an important partner in the effort to find creative funding solutions.

Funding strategies proposed to support research and drug development in Alzheimer's disease range from engaging private philanthropy to establishing targeted investment funds.

Diverse funding approaches

The high cost of research and drug development for AD, combined with the urgency of the challenge, requires substantially more resources than are available from traditional sources such as the federal government, research institutions, and charitable organizations. In this session, moderated by George Vradenburg of the Global CEO Initiative on Alzheimer's Disease, panelists provided perspectives and approaches to fundraising aimed at both large and small investors and supporting early- to late-stage drug development. Organizations represented by the panelists included a large medical research institution, public–private partnerships funding novel, high-risk research, financial institutions with products that address specific issues in medical care and research, and new investment initiatives aimed at democratizing the investment world.

The traditional approach to fundraising for research at large medical centers is to appeal to wealthy donors who are grateful for care they have received. At Weill Cornell Medical College, Robert J. Appel of Appel Associates took a different approach, focused on unmet medical needs such as AD. Because there are no positive outcomes for this disease, his team used scare tactics to appeal to people's concerns about their own personal risk and the threat to society if the disease is not controlled.

Gregory C. Simon suggested that crowdfunding is another option for increasing the capital available for investment in disease research. His new venture, Poliwogg Holdings, allows small accredited investors to contribute in $5000 units to a diversified fund focused on investments in dementia research. Poliwogg is based on the notion that people want to invest in their passions and will take chances for projects they find inspiring. Investments from Poliwogg will focus on the beginning of the drug development pipeline, raising money for high-risk but high-reward projects.

Social impact investing

Labeeb M. Abboud has worked for many years with the International AIDS Vaccine Initiative, a global nonprofit public–private partnership that has raised approximately $1 billion to advance the development of an AIDS vaccine. He is now working with a new social impact fund called the Global Health Investment Fund (GHIF), which is piloting an effort to bring private sector capital into late-stage product development for diseases that disproportionately affect developing countries. The partnership brings together the Bill & Melinda Gates Foundation, governments from several countries, the International Finance Corporation (IFC) of the World Bank, a multilateral group of pharmaceutical companies, financial institutions, and philanthropic donors. The idea is to invest in late-stage drug development projects that could produce a significant return on investment, with guarantees against losses provided by the Gates Foundation and the Swedish International Development Cooperation Agency (Sida).

Liam O'Neil of Merrill Lynch, Surya Kolluri of Bank of America, and Ben Taylor of Goldman Sachs described other social impact investment products. O'Neil described a partnership model in which investors provide up-front funding for projects that could substantially benefit a third party, and then share in that party's savings. In an offering structured to raise funds for a specific intervention directed at caring for people with dementia, for example, governments could guarantee repayment based on the savings they achieve. A government-guaranteed Alzheimer's bond for research could function in a similar way, Vradenburg noted. If a million people in the U.S. invested $1000 in such a bond, with the government guaranteeing repayment of the principle, $1 billion could be available for research. Over several years, this investment could create substantial savings by reducing costs to Medicare.

Venture philanthropy and other public–private partnerships

Henry McCance described the Cure Alzheimer's Fund (CAF), a venture capital firm he founded with two other families touched by AD. He likened venture philanthropists to venture capitalists, describing both as proactive rather than reactive investors who focus on transformational rather than incremental projects and are active in the projects they fund. CAF gave a group of leading researchers funding free of red tape for high-risk projects, with the caveat that they work collaboratively with other CAF-funded researchers.

Chas Bountra of the University of Oxford described the Structural Genomics Consortium (SGC), an early-stage public–private partnership established with funds from the Wellcome Trust, the Canadian government, and nine corporate partners. The consortium funds projects to identify and generate proteins, structures, small molecule inhibitors, and antibodies as tools for drug development. Data and knowledge generated from these projects are disseminated to the scientific community before publication, and the assets developed with SGC funds are freely available. Researchers are encouraged to use these tools to find new targets for drug development in various disease areas. The open-access structure of the consortium encourages investigators to pool resources and share risks as a means to drive innovation, reduce duplication and waste, and expedite drug development.

Highlights

Global public–private partnerships will be needed to achieve the 2025 Alzheimer's disease research goals.

Building on existing partnerships, consortia, and networks may be more efficient than creating new partnerships. Partnerships should give voice to patients.

Public–private partnerships have already achieved success in drug development challenges, including solving legal and regulatory issues.

The need for partnerships

Achieving the goals of the U.S. National Alzheimer's Plan requires more than innovative thinking and increased funding. We also need better cooperation among stakeholders. Partnerships are becoming more important, according to moderator Howard Fillit of the Alzheimer's Drug Discovery Foundation, as academic institutions move toward technology transfer, pharmaceutical companies scale-down internal research, philanthropic organizations take a more active role in drug development, and both the NIH and the FDA call for more collaboration. The challenge is to organize partnerships efficiently.

Several public–private partnerships have already advanced research through collaboration among academic, industry, and regulatory scientists, as well as patient advocates, philanthropic organizations, payers, and policy makers. Working in the pre-competitive space, these projects provide solutions and tools to expedite, de-risk, and lower the cost of drug development.

AD-related partnerships include the Alzheimer's Disease Neuroimaging Initiative (ADNI) and other collaborations managed by the Foundation for the National Institutes of Health (FNIH), a non-governmental organization founded by the U.S. Congress to bring in resources from the private sector and promote interaction with the NIH. David Wholley of the FNIH described the Biomarkers Consortium, a partnership that included the FDA and 29 private sector partners, nonprofits and pharmaceutical companies among them. The consortium aims to develop and qualify biomarkers across four disease areas, including AD. Other AD-related public–private partnerships include the Alzheimer's Association's Research Roundtable, the Dominantly Inherited Alzheimer's Network, and the Coalition Against Major Diseases (CAMD).

Partnerships that are not AD-specific have also advanced drug development. In Europe, for example, the Innovative Medicines Initiative, developed by the EU and the European Federation of Pharmaceutical Industries and Associations, promotes better science and decision making in industry and regulatory agencies, according to its executive director Michel Goldman. One project involves the creation of a large registry of older adults, intended to facilitate understanding of the natural history of AD and other diseases of aging. The organization hopes to launch an adaptive trial design project in collaboration with investigators in the U.S. and Canada.

An overabundance of consortia?

Maria C. Carrillo of the Alzheimer's Association expressed concern that the growing number of consortia may result in duplication and overlap of effort. The Alzheimer's Association promotes communication among AD-related organizations. Leaders from the association, CAMD, and the FNIH meet quarterly to share progress and look for collaborative opportunities. The association also recently published a compendium of AD-related public–private partnerships, which will be updated annually. Carrillo reported that this review of consortia confirmed there is not currently substantial overlap in their activities, in part because many partnerships are formed to address specific needs.

CAMD, one of seven consortia established by the Critical Path Institute, was founded to close the gap between emerging interventions and regulatory decision making. Martha Brumfield described the successful collaboration between CAMD and the Clinical Data Interchange Standards Consortium (CDISC) to develop therapeutic data standards for AD clinical research. These standards will be required for clinical trial submissions to the FDA beginning in 2017. CAMD also built an open database of placebo-arm data from 24 clinical trials and developed the first modeling and simulation tool to be endorsed by the FDA and the European Medicines Association (EMA) for use in designing clinical trials.

The DIAN Pharma Consortium, funded by the NIH since 2008, is designed to follow the ADNI model of biomarker validation in a uniquely informative population of people with autosomal dominantly inherited genetic mutations that confer near certainty that dementia will develop at a predictable age. According to Randall Bateman, who leads the DIAN Therapeutic Trials Unit at Washington University, DIAN was conceived as a network for use in future clinical trials. To implement a trial across DIAN centers worldwide, partners would be needed to provide additional funding, expertise, and drugs. Today, the DIAN Pharma Consortium includes 11 pharmaceutical companies, and Washington University has received a $4.2 million grant from the Alzheimer's Association to support clinical trials of anti-amyloid therapies in the DIAN population. A significant accomplishment of the consortium is the implementation of a contract enabling the 11 companies to contribute money and expertise to the project collaboratively, proving, as Bateman noted, that it is possible for competitors to work together when the need is urgent.

Taking partnerships to the next level

James C. Greenwood of the Biotechnology Industry Organization argued for the creation of a public–private partnership on the order of the Manhattan Project. It would conduct a longitudinal and comprehensive study of aging to identify markers that precede dementia and to understand the mechanisms underlying those changes. Pharmaceutical companies alone cannot fund such a large project, so commitment and financial support will be needed from the federal government, patient advocacy organizations, and others.

Gigi Hirsch, executive director of the Center for Biomedical Innovation and program director of the New Drug Development Paradigms (NEWDIGS) program at MIT, agreed that broader multi-stakeholder collaborations will be needed. The main goal is to create an opportunity for stakeholders to deconstruct complex challenges and build models that optimize risk sharing. While public–private partnerships have been effective drivers of progress, multi-stakeholder collaborations that broaden the group of partners to include non-governmental organizations, patient groups, and others can and should go further to build on and achieve the full impact of partnerships. Hirsch proposed working backward from the desired outcome (curing and preventing AD) to create a roadmap of multi-stakeholder collaborations that captures the strategic value of each partnership. This roadmap could identify gaps to be filled by other collaborations.

Highlights

The Geoffrey Beene Foundation Alzheimer's Initiative sponsored the Global Neurodiscovery Challenge to spur research on gender-based differences in AD.

Four finalists competed for a $50 000 prize in online voting by both scientists and members of the public, with an additional $50 000 prize awarded to the second place finisher in a surprise announcement.

Searching for gender-based differences in Alzheimer's disease

According to Reisa A. Sperling, women are nearly twice as likely as men to develop AD, and do so about 12 years earlier, but the biological reasons underlying this increased risk are not well understood. Data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and the Australian Imaging, Biomarker & Lifestyle Flagship Study of Ageing (AIBL) suggest that women have faster rates of brain shrinkage, decline more rapidly on cognitive measures, and show earlier biomarker changes in the preclinical stage. While hormones may explain some of these differences, there are probably other contributory factors. Data from large biomarker studies and prevention trials now beginning may allow researchers to better understand how gender contributes to AD risk.

Crowdsourcing innovation

On April 29, 2013, the Geoffrey Beene Foundation Alzheimer's Initiative announced the Global Neurodiscovery Challenge, titled Mind the Data: Search for Gender-based Differences in Alzheimer's Disease, in partnership with the FNIH, the Institute of Medicine, the Society for Women's Health Research, Women Against Alzheimer's, the 21st Century Brain Trust, BrightFocus Foundation, Brigham and Women's Hospital, and InnoCentive. The challenge, explained the Beene Foundation's Meryl Comer, asked participants to use large, publicly available datasets and novel analytical approaches to better understand gender differences that influence the development of AD.

Finalists were selected by a panel of judges with expertise in neuroscience, neurology, genetics, and women's health. According to Marietta Anthony, one of the judges, projects were judged on three criteria: originality of the hypothesis, rationale for the hypothesis, and quality of the comprehensive research plan. Four finalists were selected in the first round. Elizabeth Mormino, a postdoctoral fellow in the laboratory of Reisa Sperling, was selected as the "best hypothesis" awardee.

Following the first round, the competition moved to a crowdsourcing phase, where scientists and members of the public had the opportunity to vote for the most promising project. Each finalist prepared a video, made available online, to describe their project.

Live vote selects the winner

Nearly 6300 votes were cast online before the Path to 2025 Summit and participants had the opportunity to cast live votes at the event. The winner, Enrico Glaab from the Luxembourg Centre for Systems Biomedicine at the University of Luxembourg, aims to examine whether certain genes are expressed differently in men and women, particularly whether the USP9Y gene is protective in men.

Close behind Glaab in the voting was a project proposed by John Quackenbush and Kimberly Glass of the Dana-Farber Cancer Institute, which aims to study gender differences in cellular networks. In a surprise announcement, Anne Whitaker of Sanofi stepped forward to announce that Sanofi US will provide an additional $50 000 to fund this project.

Highlights

A global, longitudinal, and comprehensive database is needed to improve the efficiency and speed of drug development.

Expanding the breadth of data collection beyond research and clinical studies to include data from payers, caregivers, and individuals will allow for a more accurate and holistic understanding of disease and drug effects.

Decision making in health care and drug development will be improved if diverse types of data can be made more accessible and understandable.

Building a global, integrated database

The projects proposed during the Summit—from registries that will feed volunteers into natural history studies and clinical trials, to large, adaptive global trials—have in common the challenge of managing massive amounts of data. A panel moderated by Andrew Holzapfel of the Global CEO Initiative on Alzheimer's Disease looked at the changing face of health care and how new data systems can advance research.

It is a high priority to assemble multinational data, with longitudinal results across many dimensions that can be used to develop predictive models. Many projects have been established but there remains a need for broad agreement on how to coalesce, expand, and build on these efforts. According to Zaven S. Khachaturian from PAD2020: The Campaign to Prevent Alzheimer's Disease by 2020, some 150 databases relevant to studies of brain aging and AD exist but the type, format, and rigor of data collection varies. Standardizing these data, and ensuring consistent data collection across multiple sites, is essential but challenging.

There is an engineering challenge facing the AD field. We need new technology to detect the most reliable change that predicts the effectiveness of an intervention, and we need to build infrastructure to analyze these big datasets. Data may come from biomarker studies and cognitive, behavioral, and other assessments, and not only from studies of AD but also from studies of aging and related diseases. The payoff from such systems will be some 10 to 15 years in the future—too distant to qualify for traditional funding streams from the NIH or industry. Thus, creative, large-scale (global) investment strategies will be needed.

Expanding the range of data collected

Several panelists talked about the huge gap in data collection and storage methods in health care compared to financial services and other industries. Jeffrey Hammerbacher from Mount Sinai Medical Center, and formerly of Facebook, asked why the most talented software engineers and data scientists are working on consumer web problems rather than on health care, and why health care companies have not followed companies like Google, Walmart, and Goldman Sachs in building the capability to store and analyze large amounts of data.

Kris Joshi from Oracle noted that data needs in health care are different from those in other big science projects or the financial sector. In health care, major challenges arise from the size and variability of data from clinical, molecular, and genomic platforms, as well as from variability in the quality of data. Oracle is building systems for data management that speed data capture in clinical trials.

There are also huge datasets from electronic health records and claims, added David Schuette of Knowledgent. The large volume of structured and unstructured information adds to the complexity of data management. Systems must be able to aggregate diverse data to allow investigators to build hypotheses that can be proven or disproven quickly.

Telemetry data such as that captured on smartphones is readily available and could be very valuable for clinical trials and health care in general. Personal-use technologies can empower people to track their health, make decisions about their care, and participate in clinical trials. Peter L. Levin of Amida Technology Solutions also pointed to data provided by caregivers, such as information about the level of care needed by the patient and subjective assessments of changes in behavior, as useful in analyzing the effects of an intervention.

The challenge of big data analytics, according to Levin, is knowing what to look for. Stephen H. Friend of Sage Bionetworks pointed out that as medicine moves from an observational to a mechanistic approach, we need better systems to link different data sources. For example, genomic data by itself has little meaning; it must be linked to detailed phenotypic data to make sense of the functional consequences of changes in gene expression. It is unclear how to capture rich phenotypic data in AD. Friend mentioned that patients are often willing to provide these data, and their contributions could fill the void.

Making sense of the data — data sharing and accessibility

Acquiring, managing, and analyzing the huge amounts of data needed for AD research will require partnerships, not only among academic scientists and industry but also among payers, regulators, and importantly, patients. Academic and industry scientists must be given incentives to share their insights and break traditional boundaries.

It will also be necessary to ensure that health care and patient data are normalized and usable. Joshi pointed to the changing economics of health care as a lever for change in this area. Payers already have access to tremendous analytic capabilities based on claims data, and health systems need equally robust information about clinical care, quality, and cost.

Shari M. Ling from the Centers for Medicaid & Medicare Services (CMS) echoed Joshi's comments about the changing economics of health care and highlighted a push to improve data acquisition, management, and analyses. CMS is building quality-reporting programs and incentivizing providers across the care spectrum to focus on quality. One critical data element that has often been overlooked is how the patient feels. Ling argued for a system that supports parallel processing, so that concepts that resonate across different specialties and experiences, such as what constitutes a clinically meaningful treatment outcome, can be explored simultaneously. This approach will increase the need for stakeholders representing different interests to share data, insight, and tools.

Highlights

Alzheimer's disease and related dementias represent a global public health challenge that will require a coordinated global response.

On December 11th, 2013, the G8 met in London for a summit on dementia, providing an opportunity for the Alzheimer's community, including those represented at this meeting, to be heard on the world stage.

There is a need to strategically align dementia-specific efforts with broader work in neurodegenerative disease in different countries.

A global plan

The AD Summit brought together a multinational group of Alzheimer's disease experts, with representation from the U.S., Australia, the UK, Canada, Japan, and the Organisation for Economic Cooperation and Development (OECD), comprised of 34 member countries across North America, Europe, and Asia. In the final panel, moderated by George Vradenburg, participants summarized the work ahead to reach the goal of effective treatment and prevention of Alzheimer's disease by 2025.

Elettra Ronchi of the OECD said the organization regards AD as a grand challenge across scientific, social, fiscal, and health care domains. In pursuit of their mission to enable collective action, the OECD brought together experts from Europe, Asia, and the U.S. in 2012 to discuss how to anticipate the needs of a 21st century aging society. The joint statement produced by the conference focused on the need for innovation in biotechnology and nanotechnology to accelerate treatment and diagnosis and the need to adapt governance frameworks and modernize regulatory systems to expedite translational medicine.

Many countries have introduced a national plan for AD, and there was a call for all countries to develop one. The United States released its National Plan in May 2012. Donald B. Moulds from the U.S. Department of Health and Human Services explained that the plan, which will be updated annually, includes a mandate to work internationally and strategically to align research with other countries that are conducting similar or collaborative research projects. As part of this effort, the National Institute on Aging, NIH, worked with the Alzheimer's Association to create an International Alzheimer's Disease Research Portfolio (IADRP), a publicly available database to coordinate funding strategies and leverage resources worldwide.

Regardless of whether every country has a national plan, there is a consensus that AD must be tackled globally and collaboratively, as Yuko Harayama of the Council for Science and Technology Policy in Japan reported. The Japanese government and others have made political commitments to tackle AD. Indeed, Canada has made deep investments in a consortium that they hope will add value to international collaborations, according to Howard Feldman. Likewise, the incoming government in Australia, which has had a national plan since 1992, has committed increased funds to AD research, added David Ames of the Australian Imaging, Biomarkers & Lifestyle Flagship Study of Ageing (AIBL). Feldman said the biggest risk for the field, which could slow global progress, is that countries individually and collectively will fail to recognize the immensity of the problem and the need to join in worldwide initiatives.

Data sharing and data linkage across countries

The UK has committed substantial resources to advancing clinical trials in AD, as Simon Lovestone reported. Their resources are focused in three main areas: the establishment of a readiness cohort, capitalizing on the widespread use of electronic medical records across the country; innovation in clinical trials through the development of adaptive trials; and data sharing to build frameworks for data linkage that are compatible across countries.

The OECD has also called on member governments to align data management and data sharing. The organization believes big data will change the way science is pursued, Ronchi noted, pointing to the need to assess global capacity and create systems to harness data resources. She also mentioned the need to look at privacy and confidentiality from a global perspective.

Aligning global efforts

Feldman suggested that efforts should coalesce around an achievable goal, such as funding a collaborative platform to enable speedier trials at lower cost. Other panelists called for an extension of trials to developing countries, because many of the assessment instruments currently used are not appropriate for non-Western populations.

While recognizing that health care systems around the world are unprepared to deal with the increasing numbers of people who will be affected by dementia in the coming years, the panelists expressed optimism that the momentum of this Summit, along with efforts by OECD, the Group of 8 (G8), and other organizations, will create more inter-agency collaboration, data sharing, and open science.

What are the barriers to the development of a multinational adaptive trial, and what steps are needed to overcome these barriers?

What have the studies on non-pharmaceutical approaches to Alzheimer's disease treatment told us, and how can we advance more multinational, non-pharmaceutical studies?

What will be the cost of success? If trials of anti-amyloid treatments are successful, there will be additional costs to the health care system for treatment (infusion) and diagnostics. Do we have the capacity to treat people?

What economic impact will be felt if people live longer, but possibly with less need for institutional care?

What can we learn from economic analyses for each therapeutic approach?