REVIE W Open Access
Drug development in Alzheimer’s disease:
the path to 2025
Jeffrey Cummings
1*
, Paul S. Aisen
2
, Bruno DuBois
3
, Lutz Frölich
4
, Clifford R. Jack Jr
5
, Roy W. Jones
6
, John C. Morris
7
,
Joel Raskin
9
, Sherie A. Dowsett
8
and Philip Scheltens
10
Abstract
The global impact of Alzheimer’s disease (AD) continues to increase, and focused efforts are neede d to address this
immense public health challenge. National leaders have set a goal to prevent or effectively treat AD by 2025. In this
paper, we discuss the path to 2025, and what is feasible in this time frame given the realities and challenges of AD
drug development, with a focus on disease -modifying therapies (DMTs). Under the current conditions, only drugs
currently in late Phase 1 or later will have a chance of being approved by 2025. If pipeline attrition rates remain
high, only a few compounds at bes t will meet this time frame. There is an opportunity to reduce the time and
risk of AD drug development through an improvement in trial design; better trial infrastructure; disease registries
of well-characterized participant cohorts to help with more rapid enrollment of appropriate study populations;
validated biomarkers to better detect disease, d etermine risk and monitor d isease progr ession as well as predict
disease r esponse; more sensitive clinical assessm ent tools; and faster reg ulatory review. To implement change
requires efforts to build awareness, educate and foster en gagement; increase funding for both b asic and clinical
research; reduce fragmented environments and s ystems ; increase learning from succe sses and failures; p romote
data standardization and increase wider data sharing; understand AD at th e basic biolog y leve l; and rapid ly
translate new knowle dge into clinical development. Impro ved mech anistic understanding of dise ase onse t and
progression is centr al to mor e efficient AD d rug deve lopmen t and will l ead to impr oved the rapeutic appr oaches
and targets. The opportunity for more than a few new therapies by 2025 is small. Accelerating research and
clinical development efforts and bringing DMTs to market sooner would have a significant impact on the future
societal burden of AD. As these steps are put in place and plans come to fruition, e.g., approval of a DMT, it can
be predicted that mom entum will build, the process will b e self-su staining , and t he path to 2025, an d beyon d,
becomes clearer.
Keywords: Alz heime r’s diseas e, Disease-m odifying ther apy, 2025
Background
Increasing life expectancy has produced a dramatic rise
in the prevalence, and thus impact, of aging-associated
diseases including dementia. Alzheimer’s disease (AD) is
by far the most common dementia in late life. It is cur-
rently estimated that 46.8 million people worldwide have
dementia with an estimated global cost of dementia care
at US$818 billion in 2010 [1]. By 2030 it is estimated
that there will be 74.7 million people with dementia, and
the cost of caring for these individuals could rise to
some US$2 trillion. In the absence of effective therapies,
the estimated number of people with dementia will
reach 131.5 million by 2050.
Global leaders have set a deadline of 2025 for find-
ing an effective way to treat or prevent AD [2]. In the
United States in late 2010/early 2011, the National
Alzheimer’s Project Act (NAPA) was passed and signed
into law [3]. It required the creation of a national strategic
plan to address the rapidly escalating AD crisis and the co-
ordination of AD efforts across the federal government.
The overarching research goal of the project is to “prevent
or effectively treat Alzheimer’sdiseaseby2025”.In
December 2014, the G8 stated that dementia should be
made a global priority with the aim of a cure or approved
disease-modifying therapy (DMT) available by 2025.
* Correspondence: cumminj@ccf.org
1
Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, NV, USA
Full list of author information is available at the end of the article
© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Cummings et al. Alzheimer's Research & Therapy (2016) 8:39
DOI 10.1186/s13195-016-0207-9
Although the politically initiated 2025 deadline may not
have been based on scientific principles of disease research
or the realities of drug development, it has become a rally-
ing cry for researchers and advocates as they endeavor to
find innovative ways to develop drugs to successfully
achieve the 2025 goal.
Despite the evaluation of numerous potential treat-
ments in clinical trials [4, 5], only four cholinesterase in-
hibitors and memantine have shown sufficient safety and
efficacy to allow marketing approval at an international
level. These five agents are symptomatic treatments,
temporarily ameliorating memory and thinking prob-
lems, and their clinical effect is modest; they do not treat
the underlying cause of AD and do not slow the rate of
decline [6].
Over the past decade, the focus of drug discovery and
development efforts has shifted toward DMTs for AD;
that is, treatments whose aim is to affect the underlying
disease process by impacting one or more of the many
brain changes characteristic of AD. These treatment s
could slow the progression of the disease or delay its on-
set. Less encouragin g is that, over the same time period,
numerous candidate agents have failed in clinical devel-
opment, and no DMTs have shown a drug-placebo
difference in Phase 3 studies or received marketing
approval [7].
While AD drug failures to date are likely, in part, be-
cause the drugs tested lacked sufficient target engage-
ment or had toxic effects [8], efforts to bring new AD
drugs to market have been hindered by a number of
challenges—incomplete und erstanding of AD pathogen-
esis, the multifactorial etiology and complex pathophysi-
ology of the disease, the slowly pro gressive nature of
AD, and the high level of comorbidity occurring in the
elderly population [9]. Further challenges exist in the
clinical trial environment because overt clinical symp-
toms are not evident until considerable change has oc-
curred within the brain, the most appropriate outcome
measures have not been widely agreed upon, there is dif-
ficulty establishing and coordinating global clinical trial
networks, and strategies for identifying and recruiting
trial participant s are time- and cost-intensive [10]. As we
move to treating earlier in the disea se continuum, there
is more sensitivity around risk associated with drug use;
drug failures may be the result of studying too low a
dose in an effort to decrease the occurrence of side ef-
fect s, including amyloid-related imaging abnormalities
(ARIA), which are still poorly understood.
In this paper, we discuss the path to 2025, and what is
feasible in this time frame given the realities and chal-
lenges of AD drug development. We focus on develop-
ment of DMTs for individuals with early disease, which
may be more amenable to disease modification and most
likely to fulfill the 2025 mandate of meaningful new
therapy. Early disease is defined here as mild cognitive
impairment (MCI) due to AD (National Institute on
Aging-Alzheimer’s Association (NIA-AA) criteria [11])
or prodromal AD (International Working Group (IWG)
criteria [12]). Clinical study of individuals with preclin-
ical AD (asymptomatic persons at increased risk for
symptomatic AD) are being pursued but their longer
time frames and measurement and regulatory uncer-
tainty makes them less likely to contribute to reaching
the 2025 goal; this topic is not discussed in depth here,
but it is recognized that the challenges are likely to be
similar, though on a greater scale, to those associated
with development of treatments for MCI. Identifying
DMTs for use in pre clinical populations will take longer
given uncertainties about disease progression, clinical
outcomes, biomarkers, and regulatory views.
We provide examples of current activities and prac-
tices to help address these complex challenges and
briefly discuss activi ties that need to start now but may
not have direct impact until after 2025. We intend to set
the stage for continuing progress in the AD drug devel-
opment field, and to stimulate further discussion and ac-
tion for impact in the short term to fulfill the 2025 goal
and long term for continuing development in the AD
space beyond 2025.
AD drug development—current status
For a DMT, after pre clinical de velopment a nd initial
characterization of a n AD agent, Pha s e 1 takes approxi-
mately 13 months, Phase 2 approximately 28 months,
and P hase 3 appr oximately 51 mo nth s, followed by
regulatory review of approximately 18 months [13]. In-
cluding preclinical development, the total de velopment
time reaches 160 months (more than 9 years). The cost
of de veloping a DMT for AD, including the cost of fail-
ures, is estimated at $5.7 billion in the current environ-
ment [13].
To determine if availability of DMTs by 2025 is a real-
istic goal, we can work backwards from 2025 (Fig. 1). If
the current timeline remains unchanged, approval of a
DMT by 2025 requires that the agent be under regula-
tory review by 2023/2024. Thus, Phase 3 studies will
need to start by 2019 to allow sufficient time for recruit-
ment, treatment in trial and analysis/interpretation, and
Phase 2 will need to start in 20 16/2017. Therefore, for
approval by 2025, potential AD DMTs need to be in late
Phase 1 now, and most compounds with a chance of
success by 2025 will currently be in Phase 2 or 3.
The number of DMTs in clinical trials is small consider-
ing the need. Currently in Phase 1 through Phase 3
development, there are approximately five active immuno-
therapies, 11 passive immunotherapies, and 55 small
molecules (including but not limited to DMTs) [14].
The main targets are amyloid, followed by inflammatory
Cummings et al. Alzheimer's Research & Therapy (2016) 8:39 Page 2 of 12
mediators/factors and tau, and the dominant approach is
passive immunization targeting amyloid. Reviews of select
DMTs in development are provided by Scheltens et al.
[15] and Hendrix et al. [16]. Attrition is high, and few
compounds, at best, will be approved in the 2025 time
frame.
Accelerating AD drug development to 2025:
overcoming key challenges
Hastening new drug development with a view to imple-
menting drug treatment and prevention strategies by
2025 will require advances within the clinical trial set-
ting as well a s progress in the regulatory environment.
We discuss specific measures that are needed in these
areas, and provide examples of where they are already
being implemented.
Clinical trial environment
Speeding up clinical development, particularly in Phase
2 and 3, requires innovations and improvements in study
design and trial execution, as well a s more effective re-
cruitment measures and disease detection/monitoring.
Trial design and operation
Traditionally, individual study phases are completed be-
fore moving to the next phase of the study. However, as
has been the case in immunotherapy development, com-
bined Phase 1/2 clinical trials may speed development;
that is, instead of conducting a Pha se 1 trial for toxicity
and a separate Phase 2 trial for efficacy, it may be appro-
priate to integrate these two phases into one study of in-
dividuals with AD. Study sponsors can consider an
adaptive Phase 2/3 study design, whereby accumulating
trial data are used to guide modification of one or more
specified aspects of the study design, for example redu-
cing the number of dose arms, or extending or shorten-
ing the length of the trial without undermining its
validity and integrity. Use of such an adaptive trial de-
sign places greater emphasis on Phase 2 learnings as
guides to pharmaceutical decision-making (for example,
whether to continue development of an investigational
drug). While AD drug development could be reduced by
months or even years using an adaptive design, there is
some skepticism about its value with concern of errone-
ous trial modifications as a result of the “noise” with our
current cognitive measures as well as with non-validated
biomarkers. Intensive study of novel study designs will
be required to understand their appropriate role within
the AD trial setting and potential for drug development
acceleration.
While the advance of a drug from Pha se 2 t o Phase 3
is a step closer to the goal, there is a risk of wasting
both time and money if this decision is based on se c-
ondary analysis and subgroup findings when the pri-
mary endpoint is not met in Pha se 2. Rigorous
adherence to pre-specified outcomes a nd avoidance of
over-interpreting subgroup data, a s well as greater un-
derstanding of the test agent in Phase 2 and appropri-
ate primary endpoint selection, are crucial and will
help preserve resources for agents with a higher likeli-
hood of success.
Patient re cruitment and enrollment are the most time-
consuming aspects of the clinical trial process and this is
particularly so for Phase 3 studies where sample sizes for
AD trials of DMTs are often more than 1000 partici-
pants. Recruitment to AD trials is notoriously slow, and
can take years for AD Phase 3 trials. Speeding up re-
cruitment is a key area where clinical development time-
lines could be impacted, reducing Phase 2 and Phase 3
study times by many months.
Efforts are needed to increase recruitment include
expanding public and healthcare professional (HCP) under-
standing of AD as well as increasing their awareness of op-
portunities for AD trial participation, whether in currently
recruiting trials or in future trials through voluntary inclu-
sion on AD registries. HCPs also need to be more informed
and prepared for screening of patients and referral to mem-
ory centers. Patient networks and advocacy groups will play
an increasingly important role in engaging the general
population and raising their awareness of AD and demen-
tia, as well as AD trial participation opportunities, for ex-
ample through the creation of more Dementia Friendly
Communities (DFCs) (e.g., The Heart Ring Movement
campaign in Japan, and The Dementia Friends program
Fig. 1 Current timeline for clinical development of disease-modifying drugs for approval by 2025 [8]
Cummings et al. Alzheimer's Research & Therapy (2016) 8:39 Page 3 of 12
initiative launched by the Alzhei mer’s Society in the UK).
This will be particularly important in countries where there
is currently a less enthusiastic attitude towards clinical trial
participation. Voices of politicians, entertainment celebri-
ties, sports figures, religious leaders, and other stakeholders
may ultimately be necessary to fully galvanize populations
into action.
To streamline enrollment in AD clinical trials and
speed up recruitment, registries of healthy aged and
symptomatic individuals are vital. As well as providing a
repository for information about people with a specific
condition, registries provide a connection between
people who are willing to participate in research and
those recruiting for studi es. More advanced registries
with standardized demographic, genetic, biologic, cogni-
tive, and environmental information on potential partici-
pants could reduce the length of clinical trials further
after the initial eligibility testing ha s been established.
The availability of serial assessment information within a
database could permit selection of trial participants
based on disease trajectory. The availability of a trial-
ready cohort in which both amyloid and cognitive status
are known could potentially provide the greatest time
saving in treatment development; however, the cost of
establishing such cohorts is significant. Finding mecha-
nisms to avoid “labeling” participants, ensuring security
and privacy of data, and respecting transgenerational
inferences are all key to development of successful
registries.
AD registries currently in place to help with more
effective recruitment to clinical trials include:
– The Alzheimer’s Prevention Registry [17]
– The Dominantly Inherited Alzheimer Network
Trials Unit (DIAN-TU) Expanded Registry [18], to
aid recruitment of individuals at risk of having a
gene mutation that causes dominantly inherited AD
to trials of potential DMTs;
– The Brain Health Registry [19], a global online
registry for anyone age 18 years and older interested
in research of new treatments for AD, and other
conditions that affect brain function (includes
opportunities for clinical trial participation);
– The Global Alzheimer’s Platform (GAP) initiative
[20, 21];
– The Cleveland Clinic Healthy Brains Registry
(healthybrains.org) [22];
– The Alzheimer's Disease Cooperative Study (ADCS)
[23], a fede ral-university collaboration and also part
of the Alzheimer Prevention Initiative; and
– The Join Dementia Research initiative [24]intheUK.
Linking individual registries will further increase the po-
tential participant pool, and that is beginning to occur.
The Brain Health Registry has recently partnered with
GAP to dramatically increase the registry’s database size to
tens of thousands of new registry members; under the
joint effort, supporters hope to have 40,000 people regis-
tered before the end of 2016.
Other channels of recruitment, including community
screenings, advocacy forums, educational programs, so-
cial media campaigns, and use of conventional media
(television, radio, newspapers), can function to refer po-
tential participants to registries or directly to trial sites.
Trial execution can be far more efficient if there is an
integrated standing network of clinical trial sites. These
clinical trial platforms may include local disease regis-
tries, trial-ready cohorts, and optimized administrative
procedures, and they are increasingly being pursued as a
way to assure less redundancy (e.g., through acceptance
of a standardized budget and contract language and an-
nual rather than trial-specific rater training and certifica-
tion) and greater speed compared to the existing clinical
trial procedures. GAP, for example, hopes to reduce the
clinical testing cycle by 2 years or more through the de-
velopment of certified clini cal trial sites and, as already
discussed, registries with cohorts of trial-ready patients.
Clinical trial execution is dependent upon Institutional
Review Board (IRB)/Ethics Committee (EC) approval,
and with multi-center trials this invariably means ap-
proval by numerous individual IRBs/ECs. The burden on
IRBs/ECs and on sponsors and clinical investigators who
are seeking review for multicenter trials is considerable;
there is unnecessary expense, duplication of effort, and
delays. Greater reliance on a centralized (even national)
IRB review process could reduce this time-consuming
problem.
Fostering stronger ties between clinical prac tice and
research could also speed trial recruitment and increa se
trial efficiency. Comprehensive AD Centers, whereby
clinical activities and research efforts are brought to-
gether so that patient c are and clinical study of AD can
occur in a more integrated environment, are being
established. Examples include: The Gérontopôle in
Toulouse, France [25]; the Salpêtrière Dementia Research
Center, in Paris, France; the Amsterdam Dementia Cohort
[26]; the German Dementia Competence Network [27];
the Cleveland Clinic Lou Ruvo Center for Brain Health,
USA; and the University of Southern California Alzhei-
mer’s Therapeutic Research Institute (USC ATRI), USA.
Detection and monitoring of disease—clinical assessment
tools
As we shift clinical focus from the study of symptomatic
treatments in populations with more advanced disease to
DMTs in populations with earlier disease, several chal-
lenges arise with the use of currently available assessment
tools. While DMTs will potentially slow cognitive decline,
Cummings et al. Alzheimer's Research & Therapy (2016) 8:39 Page 4 of 12
they may not provide immediate improvement like exist-
ing symptomatic treatments. Cognitive instruments, such
as the Alzheimer’s Disease Assessment Scale-Cognitive
subscale (ADAS-Cog), Mini-Mental State Examination
(MMSE), and neuropsychological test items show rela-
tively little change over time in individuals with early
symptoms of disease, primarily due to ceiling effects in
many of the items that make up these scales (i.e., there are
ceiling effects, and parts of the test do not properly cap-
ture subtle changes over time). In addition, since func-
tional worsening occurs later in the disease process and
seems to follow cognitive decline [28, 29], individuals with
no or minimal cognitive symptoms are likely to have no
functional compromise, limiting the relevance of func-
tional and global assessments.
There is a need for more sensitive and responsive in-
struments for use in these early stages of AD; in particu-
lar, more sensitive and spe cific cognitive assessment
tools to capture subtle clinical decline are required to
identify individuals with minimal symptoms and discern
treatment effects among participants with earlier dise ase.
As a result of the early involvement of the medial tem-
poral lobe in AD pathogenesis, a specific memory profile
has been reported in AD that is characterized by a
diminished free recall ability that is only marginally
improved by cueing (amnestic syndrome of the hippo-
campal type) [30]. The Free and Cued Sele ctive Recall
Reminding Test (FCSRT) can be used to detect impair-
ment of free and cued recall and identify patients with
MCI with high sensitivity and specificity [31, 32]. Com-
puterized neuropsychological assessment may offer a
greater degree of sensitivity, and assessments can be per-
formed more frequently so within-subject changes can
be detected more easily. Unfortunately, computer ex-
perience can influence computerized test performance
and this is likely to be more of a challenge in elderly in-
dividuals. Exploration of these emerging alternatives
may lead to new standard assessments in the AD trial
environment.
Functional assessment remains a key challenge in AD
drug developm ent, and more sensitive tools to assess
deficits in function are urgently needed. Function is
commonly assessed using Activities of Daily Living
(ADL) measurement instruments. ADL is divided into
Basic Activities of Daily Living (BADL), which includes
self-maintenance skills such as bathing, dressing or eat-
ing, and Instrumental Activities of Daily Living (IADL),
which involves more complex activities such as using
public transportation, managing finances, or shopping.
These instrumental activities generally require a greater
complexity of neuropsychological organization and are
therefore likely to be vulnera ble to the early effects of
cognitive decline. Although there is incomplete agree-
ment about which IADL domains are typically impaired
in prodromal AD and which types of instruments may
detect those best, it is clear that activities requiring
higher cognitive processes are the most consistently af-
fected items. New instruments for assessment of IADL
functioning including items measuring the domains of
financial capacities, keeping appointments, task comple-
tion time, decision making, speed of performance, and
task accuracy [33] are needed.
Development and validation of new scales de novo is a
long process. Recent efforts have focused on developing
composites, which capture only those components from
existing scales that have the ability to discern decline in
early AD populations; for example, by eliminating items
from the ADAS-Cog that appear less sensit ive to early
changes and combining the remainder with items from
other instruments of cognition and/or function, sensitivity
to change and reduced variability can be achieved. The
Food and Drug Administration (FDA) has also indicated
that a single composite outcome may be appropriate for
trials of individuals with MCI/prodromal AD [34]. Com-
posites for the study of individuals with MCI include the
integrated Alzheimer’s Disease Rating Scale (iADRS) [35],
comprised of the scores from two widely accepted mea-
sures, the ADAS-Cog and the ADCS-instrumental Activ-
ities of Daily Living (ADCS-iADL), and the AD
Composite Score (ADCOMS) [36], comprised of four
ADAS-Cog items, two MMSE items, and all six Clin ica l
Dementia Rating-Sum of Bo xes (CDR-SB) items.
ADCOMS which has been demonstrated to have im-
proved sensitivity to clinical decline over individual
scales in individuals with earlier symptoms of AD.
New tools are also needed for pre clinical AD trials. The
ADCS-Preclinical Alzheimer Cognitive Composite
(ADCS-PACC) is a cognitive composite and the pri-
mary outcome measure in the Anti-Amyloid Treatment
in Asymptomatic AD (A4) Trial of individuals with pr
eclinical AD [37]. Other instruments for preclinical tri-
als are being assesse d.
Ultimately, the goal is to ensure that any assessment
tool employed provides clinically meaningful informa-
tion. In studies of populations with early AD symptoms,
it mig ht be appropriate to consider cognition alone as a
primary endpoint, and this may require a better under-
standing of the clinical meaningfulness of cognitive
changes and their ability to predict functional decline. A
composite composed of appropriate cognitive and func-
tional components would also be of use. More sensitive
tools, whether cognitive or functional, could help speed
clinical development by shortening recruitment time
and reducing the required samp le size.
Detection and monitoring of disease—biomarkers
Biomarkers have become instrumental in efficient clin-
ical development of drug entities for many diseases,
Cummings et al. Alzheimer's Research & Therapy (2016) 8:39 Page 5 of 12