|Home | About | Journals | Submit | Contact Us | Français|
Now is the time to launch the era of Alzheimer’s disease (AD) prevention research, establish the methods and infrastructure to rapidly evaluate presymptomatic AD treatments and evaluate them rigorously and rapidly in randomized clinical trials. This article is a call to arms. It contends that the evaluation of presymptomatic AD treatments must become an urgent priority, it identifies what is holding us back and proposes new public policies and scientific strategies to overcome these roadblocks. It defines the term ‘presymptomatic AD treatment,’ notes the best established biomarkers of AD progression and pathology and suggests how they could be used to rapidly evaluate presymptomatic AD treatments in the people at risk. It introduces an approach to evaluate presymptomatic AD treatments in asymptomatic people at the highest risk of imminent clinical onset and determines the extent to which the treatment’s biomarker predicts a clinical outcome. We propose an Alzheimer’s Prevention Initiative, which is now being reviewed and refined in partnership with leading academic and industry investigators. It is intended to evaluate the most promising presymptomatic AD treatments, help develop a regulatory pathway for their accelerated approval using reasonably likely surrogate end points and find demonstrably effective presymptomatic AD treatments as quickly as possible.
Alzheimer’s disease (AD) is an unacceptable problem. It takes a catastrophic toll on patients and family caregivers, and it is projected to have a financially overwhelming effect around the world in our children’s lifetime. In our opinion, the greatest roadblock in the scientific fight against AD is not necessarily the discovery of new treatments, but the means to evaluate them presymptomatically, when they may have their greatest impact, in a sufficiently rapid and rigorous way. It currently takes too many cognitively normal research subjects, too many years and too much money to evaluate more than a few presymptomatic AD treatments using clinical end points. Brain imaging and other biomarkers of AD progression and pathology have the potential to accelerate the evaluation of presymptomatic AD treatments. However, regulatory agencies are unlikely to provide accelerated approval for a presymptomatic AD treatment based solely on biomarker end points, without additional evidence from randomized clinical trials (RCTs) to conclude that a treatment’s biomarker effects are reasonably likely to predict a clinical benefit. In the meantime, sponsors are reluctant to conduct presymptomatic AD trials without a regulatory approval pathway. This dilemma may at first seem like an insurmountable ‘catch-22’, leading to a sense of nihilism and a lack of urgency, but inaction is not an option. Given both the extraordinary stakes and opportunity now at hand, we are convinced that now is the time to launch the public policy and scientific initiatives needed to accelerate the evaluation of presymptomatic AD treatments.
To begin, we define ‘presymptomatic AD treatments’ as “those interventions that are initiated before apparent cognitive decline and intended to reduce the chance of developing AD-related symptoms.” The term is introduced for two reasons:
We further divide these presymptomatic AD treatments into those that exert their beneficial effects by compensating for the underlying disease and those that exert their beneficial effects by directly or indirectly reversing or slowing down the progression of the underlying disease. We anticipate the eventual need to specify whether a presymptomatic AD treatment is initiated before or after presymptomatic biomarker evidence of disease, since this subclassification may predict a differential response to treatment and it may or may not require the treatment to be associated with an accompaning diagnostic test. Although the diagnosis of AD currently requires symptomatic evidence of the disease, the concept of a presymptomatic AD-modifying treatment anticipates the time when AD is defined on the basis of biological criteria, and related to the presymptomatic and progressively symptomatic stages of the disorder.
While we would recommend using the term ‘presymptomatic AD trial’ in place of ‘AD prevention trial’ in professional communications, the latter term may be more readily understood by the general public. It underscores the ultimate goal in the scientific fight against AD and it does not dismiss the possibility that some of the treatments now in development might have the chance to prevent symptomatic AD completely. For instance, if the amyloid-β (Aβ) peptide plays a critical role in the final common pathway associated with the predisposition to AD, and if a treatment now in development targets the right form of Aβ and is started sufficiently early, it may be possible to prevent symptomatic AD completely. For these reasons, we use the term ‘Alzheimer’s Prevention Initiative’ to describe the scientific strategy and public policies described below, and which continues to be refined in partnership with leading academic and industry investigators and other stakeholders.
There are several compelling reasons to accelerate the evaluation of presymptomatic AD treatments starting now, as described in the following sections.
Alzheimer’s disease is the most common form of disabling cognitive impairment in older people, afflicting approximately one tenth of people older than 65 years of age and almost half of those older than 85 . With the skyrocketing number of people living to older ages, the financial toll of AD is projected to become overwhelming by the time today’s young adults become senior citizens [2,3]. According to a recent report , more than 35 million people are afflicted by AD and other dementias; this number is estimated to nearly double every 20 years, to more than 65 million in 2030 and more than 115 million in 2050, and there are likely to be an even greater number of people in the earlier symptomatic stages of AD.
With the quickening pace of scientific discoveries, researchers continue to add to the list of currently available, sufficiently safe and well-tolerated interventions, suggested but not yet proven to postpone the onset and reduce the risk of symptomatic AD. Interventions proposed to reduce the risk of symptomatic AD include a variety of currently available healthy lifestyle [4,5] and dietary interventions , medications approved for other indications [7-9] and dietary supplements . A treatment that delayed the onset of symptomatic AD by only 5 years without increasing longevity would potentially reduce the number of patients by half , illustrating the impact that a symptomatic AD-postponing treatment could have on this rapidly growing public health problem.
The list includes numerous Aβ-modifying medication and immunization treatments, a much smaller number of tau-modifying and microtubule-stabilizing treatments, and treatments suggested to target other elements of the postulated pathogenic cascade . If any one of these treatments is targeting critical pathogenic events, is sufficiently safe and well tolerated, and is introduced early enough, it could have an enormous public health benefit. There will soon be enough safety and tolerability data to permit the evaluation of some of these treatments in asymptomatic people at increased risk for symptomatic AD.
It has been suggested that certain Aβ-modifying treatments may be most effective if started before the appearance of both fibrillar Aβ and neurofibrillary neuropathology. There is a compelling reason to evaluate promising treatments presymptomatically, whether or not they significantly slow down the clinical course in symptomatic patients, as long as the treatment is considered sufficiently safe and well tolerated. While there is a growing concern that investigational AD-modifying treatments now in RCTs may offer too little help too late to exert a profound effect even in patients in the mildest symptomatic stages of AD, we and others hold onto the hope that symptomatic AD patients may still have at least some therapeutic response to these treatments, we postulate that symptomatic AD patients may benefit from a combination of treatments that target early and late stages of the disease (e.g., a combination of Aβ- and tau-modifying therapies), and we wish to emphasize the importance of doing everything in our power to address the needs of our symptomatic AD patients and families. To be clear, the accelerated evaluation of presymptomatic AD patients should not come at the expense of our patients.
Cognitively normal people at unusually high imminent risk of symptomatic AD include carriers of relatively rare early-onset AD-causing PS1 , PS2  and APP  mutations close to their family’s estimated median age at clinical onset, people with two copies of the APOE ε4 allele, a late-onset AD susceptibility gene , close to their estimated median age at clinical onset and, according to initial studies, older people with a cerebrospinal fluid (CSF) biomarker profile characterized by high total (t)- or phospho (p)-tau and low Aβ42 levels [16,17].
While researchers continue to investigate a number of different brain imaging, biological fluid and other biomarker measurements for use in the early detection and tracking of AD (see references [18-22], for example), to date the best established biomarkers of AD progression and pathology include reduced fluorodeoxyglucose (FDG) PET measurements of the cerebral metabolic rate for glucose (CMRgl) in brain regions preferentially affected by AD (Figure 1A) [23-25]; structural MRI measurements of brain shrinkage (i.e., hippocampal, entorhinal cortex and whole brain volume loss, ventricular enlargement and gray matter loss and cortical thinning in regions preferentially affected by AD (Figure 1B)) [26-28]; PET measurements of fibrillar Aβ burden (using 11C-labeled Pittsburgh Compound-B (Figure 1C) or 11C- or 18F-labeled ligands now in development) [29-37] and certain CSF analytes (i.e., low CSF Aβ42 levels, high t- or p-tau levels, or a combination of low Aβ42 and high t-tau or p-tau levels) [16,17,38,39].
Several of these imaging and CSF biomarkers have shown value in the presymptomatic detection and tracking of AD. For instance, we and our colleagues have used imaging techniques in cognitively normal initially late-middle-aged APOE ε4 homozygotes, heterozygotes and noncarriers to detect or track CMRgl reductions (Figure 2), whole brain shrinkage (Figure 3), and higher fibrillar Aβ burden (Figure 4) associated with an increased genetic risk for AD (Figures (Figures22--4).4). These studies suggest how the imaging techniques could be used to evaluate presymptomatic AD treatments in a few hundred cognitively normal late-middle-aged APOE ε4 homozygotes or heterozygotes in 2-year proof-of-concept RCTs [26,27,31,40-46] and complement the work of other researchers [30,47-58]. Researchers have also used imaging techniques to show CMRgl declines, whole brain shrinkage and increased fibrillar Aβ burden in cognitively normal carriers of early-onset AD-causing mutations (Figure 5) [35,59-71], and to demonstrate that fibrillar Aβ burden in cognitively normal older adults is associated with subsequent rates of progression to symptomatic AD . Moreover, they have reported low Aβ42 levels and high t-tau and p-tau levels in cognitively normal APOE ε4 carriers [72,73], and that the combination of high t-tau or p-tau levels and low Aβ42 levels predicts subsequent rates of progression from cognitive normality to the early symptomatic stages of AD [16,17,74]. These presymptomatic AD biomarkers could be used to rapidly evaluate presymptomatic AD treatments in APOE ε4 carriers, early-onset AD-causing mutation carriers or research participants with presymptomatic biomarker evidence of AD in proof-of-concept RCTs.
Despite the urgent need to evaluate promising presymptomatic treatments for AD and the scientific opportunities now at hand, several roadblocks continue to hold us back. These are discussed in the following sections.
As noted previously, it takes too many cognitively normal people and too much time to find demonstrably effective presymptomatic AD treatments using clinical end points, such that enough people develop the clinical end points needed to distinguish between active treatment and placebo groups with adequate statistical power. Although a small number of presymptomatic AD RCTs have been performed [75-78], they have required thousands of subjects, many years of treatment and restriction of the sample to older adults in order to achieve sufficient statistical power. If we continue to rely on the use of clinical end points in presymptomatic AD RCTs, we will not be able to evaluate the range of promising presymptomatic treatments or those treatments started in middle age, when several treatments (e.g., hormone-replacement, cholesterol-lowering and blood pressure-lowering therapies) have been postulated to have their most significant therapeutic effects. Pharmaceutical and biotechnological companies are reluctant to sponsor prevention trials that take longer than their duration of marketing exclusivity, and the government may be reluctant to support many presymptomatic AD trials at the same time, given their extraordinary expense. Therefore, given the stakes, “who in his right mind would consider studying just one of these [presymptomatic AD] treatments at a time?” [GINGRICH N, ALZHEIMER’S STUDY GROUP, PERS. COMM.].
Brain imaging and CSF biomarker methods could be used in APOE ε4 carriers and other at-risk subjects to evaluate presymptomatic AD treatments in RCTs. Still, these studies will not provide a sufficient path for the accelerated regulatory agency approval of presymptomatic treatments until RCTs provide sufficient evidence to conclude that a treatment’s effect on these biomarker end points is reasonably likely to predict a clinical benefit [26,79,80]. In the next section, we propose a strategy to address this critical need.
For now, some humility is needed when it comes to a reliance on these biomarkers in RCTs. It is not yet known how much a given biomarker will budge in response to different treatments. Similarly, it is not known whether the treatment might have a confounding effect on the biomarker unrelated to AD progression (e.g., a change in brain swelling, synaptic activity, glucose metabolism or arborization or analyte turnover), as illustrated in the RCT of the first active Aβ immunization therapy . Finally, researchers still do not know the extent to which the biomarker effects of different AD-modifying treatments, alone or in combination, will predict a clinical benefit. While the only way to answer these questions is to characterize and compare each of the promising biomarkers in clinical trials, this surrogate marker development effort introduces significant cost and complexity, and it is typically not the primary focus of sponsors seeking to evaluate a particular drug product in the most cost-effective way.
As previously noted, there has been little financial incentive for companies to conduct time-consuming and expensive presymptomatic AD RCTs; there has been little financial incentive to investigate off-patent treatments, including health-promoting lifestyle interventions, and there has been little short-term financial incentive to include multiple expensive brain imaging and other biomarker measurements in RCTs to help determine which biomarker is mostly likely to predict a treatment’s therapeutic effect. It is time to make this effort an urgent priority. Given the stakes, it is time for federal governments to play a leadership role in this critically important endeavor.
The Alzheimer’s Study Group , Leaders Engaged on AD (LEAD), the Alzheimer’s Association, Act-AD, Prevent Alzheimer’s Disease 2020 and other stakeholders [82-85] have proposed new public policies to help transform AD research and care. Here, we propose a strategically focused Alzheimer’s Prevention Initiative, which consists of new scientific strategies, public policies and the sense of urgency needed to launch an era of AD prevention research and find demonstrably effective presymptomatic AD treatments as soon as possible . In our opinion, it is not only important to launch an initiative like this, but to do so in a way that maximizes the chance for success. This will entail extensive collaboration with scientific, industry and public partners, capitalizing on the successes and lessons of other collaborative efforts such as the Alzheimer’s Disease Cooperative Study (ADCS), Alzheimer’s Disease Neuroimaging Initiative (ADNI) and the Dominantly Inherited Alzheimer’s Network (DIAN). Success will be measured, to some degree, by the extent to which a new initiative such as this can leverage and complement the goals and aims of other related programs, as well as the strengths of individual investigators and advisors. Transparency and accountability will be critical, along with the imperative to assure that the many ethical issues that will arise are handled as thoughtfully as possible.
Despite the roadblocks discussed previously, we remain optimistic regarding the ability to evaluate presymptomatic AD treatments. These suggestions are discussed in the following sections.
We propose to evaluate promising Aβ-modifying treatments using clinical, cognitive and multiple biomarker end points in asymptomatic early-onset AD-causing mutation carriers, APOE ε4 homozygotes close to their estimated median age at clinical onset and older adults with a high-risk CSF biomarker signature [16,17,74] in 60-month RCTs using fibrillar Aβ PET, FDG PET, structural MRI, CSF and cognitive measurements, as well as clinical end points. If the Data Safety Monitoring Board (DSMB) found no promising biomarker or clinical effects during the course of the study (i.e., no evidence that the active treatment was associated with the predicted effects on measurements of Aβ or tau pathology, regional CMRgl decline, brain shrinkage, long-term memory or clinical ratings), it could declare futility, thereby giving those people at the highest imminent risk for symptomatic AD access to other investigational treatments in presymptomatic therapeutic trials while conserving precious resources. If instead, the DSMB found promising biomarker or clinical effects, the study would continue using biomarker, cognitive and clinical end points.
This presymptomatic AD treatment trial/surrogate marker development paradigm would make it possible to evaluate investigational presymptomatic treatments trials sooner than otherwise possible (i.e., with safety and tolerability data from fewer patients), since it would provide the data needed to show the extent to which an established treatment’s effects on different biomarkers predicts a clinical benefit. This paradigm would provide some of the evidence needed to approve presymptomatic treatments solely on the basis of reasonably likely surrogate end points in future prevention trials, and would help trial sponsors select the most suitable biomarker end points for these trials. We postulate that the combined effects of an amyloid-modifying treatment on measures of Aβ pathology and one or more theoretically downstream measurements (i.e., regional CMRgl decline, whole brain shrinkage, t-tau or p-tau increases and/or long-term memory decline) predict subsequent rates of conversion to symptomatic AD onset (e.g., a Clinical Dementia Rating  global score >0). If the treatment was not associated with a clinical benefit, this paradigm would raise new questions about the amyloid hypothesis and Aβ-modifying treatments, accelerating the investigation of AD mechanisms and treatments unrelated to Aβ.
Longitudinal studies such as the ADNI , the recently initiated DIAN and our study of initially late-middle-aged cognitively normal APOE ε4 homozygotes, heterozygotes and noncarriers [27,31,40,41,43,46] include a range of promising brain imaging biomarker measurements of AD progression and pathology. Clinical trials of promising AD-slowing treatments should also include multiple biomarkers, not just to secure regulatory approval for the intervention under study, but to help establish surrogate end points for future prevention trials. As suggested below, federal governments could provide the tax incentives or public–private partnership funding needed to validate surrogate end points as quickly as possible. The use of multiple biomarkers would help to overcome the possibility of imaging modality-specific confounding treatment effects, for example, the modality-specific effects of an Aβ-modifying treatment on MRI measurements of whole-brain volume unrelated to the treatment’s effects on neuronal loss or the modality-specific effects of treatment on FDG PET measurements of CMRgl unrelated to synaptic loss. In hindsight, for instance, it would have been beneficial to have included FDG PET measurements to evaluate an effect of the first active Aβ immunization therapy on disease progression in light of the treatment’s unanticipated effects on MRI measurements of brain volume .
This infrastructure would include national registries of potentially interested, cognitively normal people for both proof-of-concept and pivotal presymptomatic AD trials, along with the means to ascertain genetic, brain imaging and other biomarker indicators of AD risk in these RCTs.
Apart from new scientific strategies, new public policies are needed to accelerate the evaluation of presymptomatic AD treatments. We will discuss our recommendations in the following sections.
The extension of marketing exclusivity has promoted the development of orphan drugs for rare disorders. Offering extended marketing exclusivity for demonstrably effective presymptomatic AD treatments, perhaps starting the time-clock with completion of the expensive, large and time-consuming pivotal prevention trials, would galvanize the evaluation of promising presymptomatic treatments, and it could help avert a financially overwhelming crisis.
Treatments without a sufficient marketing incentive, such as currently available medications with limited or no patent coverage, dietary supplements and health-promoting lifestyle interventions that have been suggested to reduce the risk of AD, could be funded with federal support.
This policy would accelerate the development of reasonably likely surrogate end points for the evaluation of promising AD-slowing treatments, and it would provide a path for the accelerated regulatory agency approval of presymptomatic treatments.
It is important that public–private partnerships that are needed to evaluate promising presymptomatic treatments are created, and surrogate markers are developed in the most rapid and rigorous way. This partnership would include academia, industry, federal and regulatory agencies, and other important stakeholders.
Now is the time to make the scientific understanding and prevention of AD a national priority. The Alzheimer’s Prevention Initiative needs to be established with a sense of urgency, public policies and research infrastructure put in place, and presymptomatic AD treatment/surrogate marker development RCTs planned in order to launch an era of AD prevention research to find demonstrably effective risk-reducing and prevention therapies, and avert an overwhelming crisis. Leading academic and industry researchers, policymakers and other stakeholders need to be brought together in order to accomplish the goals set forth in the most appropriate, inclusive and effective way. Let us find demonstrably effective presymptomatic AD treatments without losing a generation.
The authors thank R Caselli, K Chen, A Fleisher, D Bandy and all of their other research colleagues and their valued study participants for their contributions to the development of presymptomatic brain imaging end points. The authors would also like to thank their academic, industry, foundation and National Institute on Aging colleagues for their feedback and support; S Madsen and P Thompson for generating Figure 4 using our data and their beautiful computer graphics software.
Financial & competing interests disclosure This article was supported by grants from the National Institute on Aging (R01AG031581 and P30 AG19610 to EMR), the Banner Alzheimer’s Foundation and the state of Arizona. EM Reiman has received consulting fees from Amnestix/Sygnis, Élan, Eli Lilly and GlaxoSmithKline for his advice on the design of research studies. His group has had study contracts with AstraZeneca, Avid, GlaxoSmithKline, and Kronos Life Sciences. PN Tariot has received consulting fees from Acadia, AC Immune, Allergan, Eisai, Inc., Epix Pharmaceuticals, Forest Laboratories, Genentech, Medavante, Memory Pharmaceuticals, Inc., Myriad Pharmaceuticals, Sanofi-Aventis, Schering-Plough and Worldwide Clinical Trials. He has also received consulting fees and research support from Abbott Laboratories, AstraZeneca, Avid, Baxter Healthcare Corp., Bristol Myers Squibb, Élan, GlaxoSmithKline, Eli Lilly, Medivation, Merck and Company, Pfizer Inc., Toyama, and Wyeth Laboratories. He has received educational fees from the Alzheimer’s Foundation of America and is listed on a patent ‘Biomarkers of Alzheimer’s Disease’. Banner Health, the investigators’ parent organization, has a patent pending on the use of biomarkers in the evaluation of presymptomatic AD treatments. It was filed in 2004, and EM Reiman is listed as the inventor.
The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Papers of special note have been highlighted as:
of considerable interest