New treatments are urgently needed for Alzheimer’s disease (AD). Drug discovery and development begins with target identification and treatment candidate characterization. After optimizing a molecule for efficacy and minimizing risk of toxicity, it is assessed for efficacy and toxicity in animal models of AD. If promising in animal studies, a potential treatment is advanced through clinical trials with knowledge of benefits and harms associated with human treatment accruing. Biomarkers play increasingly important roles in the drug development process. The Food and Drug Administration (FDA) is involved throughout the process and advises sponsors at key points in the evolution of trials. The only channel for developing urgently needed new therapies for AD is through clinical trials. All stakeholders including concerned patients and families, those at risk for AD, primary care physicians and specialists, neuroscientists, philanthropists, the National Institutes of Health, advocacy groups, and industry sponsors must all work together to accelerate clinical trials and bring meaningful new therapies faster to those who have or are at risk of having AD.

The Problem

A neurodegenerative and age-related disorder that affects cognition, function, and behavior,1 AD is growing to staggering proportions as the global population ages. There are 5.1 million people with AD in the US and 13 million worldwide; AD prevalence is expected to increase to 13 million in the US and more than 100 million worldwide by 2050.2 The cost of care for AD in the US is estimated at $267 billion and this will burgeon to more than $1 trillion per year by 2050 if new interventions are not developed. The need is urgent to find new drugs that will delay onset, slow progression, or improve symptoms of AD. It is calculated that a delay in AD onset by 5 years would decrease the population of people with AD by 5.7 million and decrease the expense of caring for those with AD by $367 million by 2050.3

It has been extraordinarily difficult to develop new therapies for AD. with a failure rate iof more than 99% and no new treatment approved since 2003.4 Attempts to develop disease-modifying therapies (DMTs) for AD have had a 100% failure rate.5 Every trial, however, presents opportunities to learn and improve the drug development process.6,7 In this review, we highlight drugs in the AD therapeutic pipeline and call attention to advances in trial methods, pharmaceutical agent characterization, trial population selection, and use of biomarkers to inform trial implementation, conduct, and interpretation.

The Alzheimer’s Disease Clinical Trials Pipeline

Figure 1 shows principal classes of agents in clinical trials for AD, including neuropsychiatric agents, cognition-enhancing drugs to improve cognition above baseline, and DMTs that target underlying disease biology with the intent of delaying onset of cognitive symptoms or slowing cognitive and functional decline. Figure 2 shows the number of agents in trials for each major class of drugs in each trial phase. The average time to develop a drug for AD is 13 years and the cost (including failures and capital) is $5.6 billion.8,9

Figure 1. Principal classes of agents in trials for Alzheimer’s disease.

Click to view larger

Figure 1. Principal classes of agents in trials for Alzheimer’s disease.

Figure 2. The phases of drug development and the number of agents for the major classes in each phase. Abbreviation: PK, pharmacokinetics.

Click to view larger

Figure 2. The phases of drug development and the number of agents for the major classes in each phase. Abbreviation: PK, pharmacokinetics.

The Drug Development Process

Drug Discovery

Drug development advances from target identification and candidate nomination to nonclinical efficacy and toxicity testing in animals and then to phase 1, 2, and 3 clinical trials in humans. The process begins with decisions about what biologic process to target. As pathophysiologic observations are made, hypotheses regarding pathogenesis and treatment of AD are generated, biologic targets identified, and candidate agents that may ameliorate these processes created and tested. Candidates may be natural substances, compounds derived from large libraries of molecules, repurposed agents developed for other indications and found to have activities relevant to AD, or agents derived from computer-based in silico design approaches.6 These candidates are tested in specific assays (eg, potential ß-site amyloid precursor protein cleavage enzyme [BACE] inhibitors would be tested in preparations where the activity of this enzyme can be monitored), “hits” are identified, and active agents advanced through a process of lead optimization that best balances potency, potential side effects, blood–brain barrier penetration, and pharmacokinetics.6

Nonclinical Assessment

Lead compounds are tested in animals for both efficacy and toxicity. Efficacy assessments are typically conducted in transgenic (tg) mice created by inserting 1 to 5 human autosomal dominant AD (ADAD) mutations into the mouse genome and testing the offspring with the mutation for expression of AD pathology. These tg species develop amyloid pathology but may not have tau changes or marked cell death; these animals allow assessment of the drug effects on amyloid biology, but they are imperfect models of AD and do not predict the effects of drugs in human AD.10 Mutations of genes involved in familial frontotemporal dementia (FTD) have been used to create tg mice with tau pathology and have been combined with ADAD mutations to develop tg species that express both amyloid and tau pathology.11 Knock-in and knock-out gene strategies have been used to interrogate the effects of increasing and decreasing aspects of cellular pathways to make the case for drugs that increase or decrease these activities as treatments for AD. Animals with tg alterations from birth more closely resemble human ADAD than the typical later-onset form of AD. Rabbits, guinea pigs, some species of dogs, and senescence accelerated mouse (SAM) models have late-life AD-like changes and may be more like late-onset AD.12 Most nonhuman species typically do not have AD-like changes with aging, and most are resistant to insertion of transgenes and creation of tg species. For example, there are few rat or nonhuman primate models of AD.

The FDA requires demonstration of safety, usually in 2 species, prior to human administration. Ascending dose toxicity studies are typically done in mice or rats; dogs are usually observed for cardiac safety. Dosing studies establish the no observed adverse event level (NOAEL), often used to guide human dosing.

Phase 1

Phase 1 clinical trials are first-in-human (FIH) exposures and necessarily involve some risk because animal experience is not always predictive of human outcomes. Phase 1 trials should establish the pharmacokinetics (PK) of the treatment agent, blood-brain penetration and entrance into the central nervous system (CNS), and maximum tolerated dose (MTD). Critical PK features to be determined in phase 1 include half-life of the agent, time to maximum concentration (Tmax), maximum concentration (Cmax), bioavailability of orally administered agents, metabolic and elimination pathways (eg, liver metabolism through the CYP450 system, fecal excretion, urinary excretion), number and biological activity of major metabolites, interaction with other drugs commonly used in the elderly population and metabolized through shared pathways, and MTD.13 The MTD is particularly important because failure to show a drug-placebo difference at later stages of development inevitably raises the question of whether an optimal dose was used. Safety and tolerability are determined in phase 1 with symptom reports, electrocardiography (ECG), and blood tests (eg, liver enzymes, endocrine measures, and more).

Phase 1 studies are typically conducted in dedicated inpatient units with participants observed continuously during FIH dosing. Cohorts (typically 4-8 individuals receiving the active agent and 2-4 individuals receiving placebo) are given increasing doses in single ascending dose (SAD) studies. Once safety and tolerability of single doses are known, cohorts are given repeated doses for 14 to 28 days in multiple ascending dose (MAD) cohorts with each receiving a higher dose than the last. Participants in phase 1 studies are typically healthy volunteers under age 55, but 1 or more cohorts may include participants over age 55. Phase 1/2 studies may include cohorts of patients with AD. In the case of immunotherapies, in which treating healthy volunteers could have long-term consequences, the entire SAD/MAD program may be conducted in participants with AD. Efficacy cannot usually be assessed in phase 1 although computerized cognitive measures or pharmacoEEG measures are sometimes included in phase 1 studies of purported cognition-enhancing agents.14

Phase 2

Phase 2 is a critical aspect of drug development and results are essential to planning and conducting the pivotal phase 3 registration trials in preparation for discussions of a new drug application (NDA) and marketing approval by the FDA.15 Phase 2 has sometimes been called the “learn” and phase 3 the “confirm” of a “learn and confirm” drug development paradigm.16 At the end of phase 2 and prior to progressing to phase 3, the development program should determine: brain penetration of the agent, target engagement by the drug, and 1 to 3 doses to be advanced to phase 3. Brain penetrance should be shown for all drugs with primary effect in the CNS postulated. This is done by measuring cerebrospinal fluid (CSF) levels of the agent at specific time points after oral or intravenous administration. A plasma to CSF ratio can be established, and the amount of the drug available at the target estimated. Target engagement studies are required to show that the drug is doing in humans what is posited to be required for efficacy. For example, the effect of BACE and γ-secretase inhibition on CNS amyloid synthesis can be shown with the stable isotope-labeled kinetics (SILK) technique.17 Gamma-secretase inhibitors produce a detectable change in the types of amyloid detectable in the serum.18 Imaging can also be used to show target engagement; fluorodeoxyglucose (FDG)-positron emission tomography (PET) can demonstrate drug-placebo differences in brain metabolism and amyloid PET can show drug-placebo differences in fibrillar plaque amyloid following treatment with amyloid-lowering antibodies.19 Functional MRI (fMRI) and EEG can be used to show circuit level changes following drug exposure in phase 2 studies.20 Circuit function is required for normal cognition and behavior, and circuit effects may increase the likelihood of showing clinical benefit in large phase 2 and phase 3 studies. Programs for DMT development in which target engagement is not established in phase 2 are at high risk of failure in phase 3. Dose-finding and dose-response relationships inform phase 3 dose decisions and should be explored in phase 2. Phase 2 may be split into phase 2a for proof-of-concept and target engagement and phase 2b for dose-finding; both are often conducted in a single trial. Table 1 shows the tools typically used in clinical trials in different types (eg, AD dementia, prodromal AD, preclinical AD) of AD populations.

Cognition should be measured in phase 2 studies but it may not be necessary to show a significant drug-placebo difference at this phase for all programs. Cognitive effects should be seen in phase 2 trials of cognition-enhancing agents. Behavioral benefits should be shown in phase 2 of agents targeting neuropsychiatric symptoms. Programs for DMTs aimed at slowing disease progression compared to placebo may advance to phase 3 without demonstrating a drug-placebo difference on clinical tools in phase 2. An effect on target engagement biomarkers in phase 2 combined with directional benefit on cognitive measures may be sufficient to matriculate a DMT candidate agent to phase 3. Larger, longer phase 2 trials supporting clinical and biologic effects increase confidence in potential success of phase 3.

Phase 3

Phase 3 goals are to meet standards for regulatory approval of a new drug. The FDA requires 2 well-conducted clinical trials showing benefit of the active agent compared with placebo on measures of cognition and function or cognition and a global measure in people with AD dementia. Trials of drugs treating neuropsychiatric symptoms are also typically required to show benefit on behavior targets in 2 well-conducted trials.

Trials in patients with mild cognitive impairment (MCI) due to AD or prodromal AD may be approved on the basis of benefit demonstrated on a single composite outcome such as the Clinical Dementia Rating—Sum of Boxes (CDR-sb) although it should be shown that the drug-placebo difference is not derived exclusively from an effect on cognition. The AD Composite Score (ADCOMS) is a score derived from performance on standard clinical trial tools including the CDR-sb, AD Assessment Scale—Cognitive Subscale (ADAS-cog), and Mini-Mental State Examination (MMSE) and shown to be sensitive to change in the MCI/prodromal phase of AD.21 Improved sensitivity to change facilitates capturing a drug-placebo difference in this early phase of AD.

Preclinical/secondary prevention trials of DMTs that involve participants who have no cognitive deficits on traditional measures and therapies cannot feasibly show cognitive benefit. Approval for treatment in this period might depend on showing the reduction or delay of change in a biomarker or set of biomarkers or delay in cognitive decline on sensitive measures of cognition. Conditional approval based on compelling evidence from biomarkers followed by demonstration of clinical benefit after initial approval is an option for approval in this very early stage of AD. More experience and more data are needed from prevention trials to inform regulatory decisions


Biomarkers play an increasingly important role in clinical trials.22 In phase 1 trials, biomarkers are important for adverse event detection (eg, liver function tests, blood pressure, ECG). In phase 2, biomarkers confirm the diagnosis of AD (eg, amyloid PET or CSF amyloid measures), demonstrate target engagement (eg, reduction of brain amyloid with amyloid PET), inform analyses (eg, effects in people with the apolipoprotein ε4 [ApoE-ε4] allele), and monitor adverse events (eg, amyloid-related imaging abnormalities [ARIA]). In phase 3, biomarkers support the diagnosis of AD, demonstrate disease-modification, inform analyses, and provide evidence on adverse events. Table 2 shows the role of biomarkers in the different phases of drug development.

Food and Drug Administration Involvement

The FDA is involved throughout the drug development process.23 The FDA advises drug development sponsors on the trial conduct and outcomes necessary to achieve approval for their candidate agent. Consultation with the FDA begins at phase 1 and continues after drug approval because monitoring for safety continues as long as an agent is on the market. The FDA issues guidances that inform sponsors about specific aspects of drug development and approval from manufacturing to product labeling. Figure 3 shows the pivotal episodes of consultation by the FDA with drug development sponsors; the consultations are not limited to those shown and advice can be sought by the sponsor whenever needed. The FDA may also create advisory panels of experts with open hearings including sponsors and the lay public to gain input on whether drugs should be approved.

Figure 3. Main episodes of Food and Drug Administration consultation in the drug development process.

Click to view larger

Figure 3. Main episodes of Food and Drug Administration consultation in the drug development process.

The Importance of Referring to Clinical Trials

The only channel for development and approval of safe and efficacious drugs for AD is through clinical trials. Without trials, no new drugs would become available and no progress in preventing, delaying, slowing, improving or reducing AD is possible. Scientists and biopharmaceutical companies cannot win the battle against AD without the help of cognitively normal at-risk individuals, patients with symptoms, caregivers, and primary care practitioners (PCPs) and specialists who refer people to trials. People with AD and those at risk and their family members and caregivers must become citizen-scientists comprising an army committed to defeating an enemy that is attacking millions of Americans and will afflict millions more if treatments are not found. Nontrial specialists and PCPs form the first line of defense in the care of individuals with AD, offering detection, diagnosis, and the treatment with existing therapies.24 Counseling of caregivers and referral to community resources can provide relief of stress for family members. Providing standard care, however, is not enough. As a team of concerned individuals, we must find new treatments that improve lives, preserve cognition, and ensure dignified and vital aging.

Table 3 shows the number of trial participants needed for all current on-going trials (40,871) and for all trials currently recruiting (27,277) participants. Recruitment to trials is currently so slow that the period of recruiting for trials exceeds the period of drug exposure in the trial, turning an 18-month treatment period into a 3- to 4-year trial. This slow recruitment increases the cost of trials, delays go/no go decisions critical to accelerating drug development, discourages companies from working in the AD area, and ultimately slows the ability to get new therapies to those in urgent need of treatment. The current costs and delays of drug development are unacceptable and all citizens concerned about their own brain health and that of their families, friends, and fellow citizens must find ways to assist in developing new treatments through personal participation in trials, referral and support of family members and friends in trials, encouraging doctors and opinion leaders to refer to trials, and participating in educational and awareness-raising events that will lead to greater trial participation. The federal government and the National Institutes of Health (NIH) can also contribute through public education and awareness campaigns and through supporting research on the optimal means of engaging participants in clinical trials. The current rate of trial participation is unacceptably low, and we must all renew our commitment to being part of the generation that will bring meaningful new therapies to people with AD and those at risk of this mind-robbing illness.

1. Masters CL, Bateman R, Blennow K, et al. Alzheimer’s disease. Nat Rev Dis Primers. 2015;1:15056.

2. 2018 Alzheimer’s disease facts and figures. Alzheimer Dement. 2018;14(3):367-429.

3. Alzheimer’s Association. Changing the trajectory of Alzheimer’s disease: how a treatment by 2025 saves lives and dollars. Chicago, IL; 2015.

4. Cummings JL, Morstorf T, Zhong K. Alzheimer’s disease drug-development pipeline: few candidates, frequent failures. Alzheimers Res Ther. 2014;6(4):37-43.

5. Cummings J, Lee G, Ritter A, et al. Alzheimer’s disease drug development pipeline: 2018. Alzheimers Dement. 2018;4:195-214.

6. Cummings J, Ritter A, Zhong K. Clinical trials for disease-modifying therapies in Alzheimer’s disease: a primer, lessons learned, and a blueprint for the future. J Alzheimers Dis. 2018;64(s1):S3-S22.

7. Cummings J. Lessons learned from Alzheimer disease: clinical trials with negative outcomes. Clin Transl Sci. 2017;11(2):147-152.

8. Scott TJ, O’Connor AC, Link AN, et al. Economic analysis of opportunities to accelerate Alzheimer’s disease research and development. Ann N Y Acad Sci. 2014;1313:17-34.

9. Cummings J, Reiber C, Kumar P. The price of progress: funding and financing Alzheimer’s disease drug development. Alzheimers Dement. 2018;4:330-343.

10. Sabbagh JJ, Kinney JW, Cummings JL. Animal systems in the development of treatments for Alzheimer’s disease: challenges, methods, and implications. Neurobiol Aging. 2013;34(1):169-183.

11. Sterniczuk R, Antle MC, Laferla FM, et al. Characterization of the 3xTg-AD mouse model of Alzheimer’s disease: part 2: behavioral and cognitive changes. Brain Res. 2010;1348:149-155.

12. Bates K, Vink R, Martins R, et al. Aging, cortical injury and Alzheimer’s disease-like pathology in the guinea pig brain. Neurobiol Aging. 2014;35(6):1345-1351.

13. Friedman LM, Furberg CD, DeMets DL, et al. Fundamentals of Clinical Trials, 5th ed. New York, NY: Springer; 2015.

14. Jobert M, Wilson FJ. Advanced analysis of pharmaco-EEG data in humans. Neuropsychobiology. 2015;72(3-4):165-177.

15. Gray JA, Fleet D, Winblad B. The need for thorough phase II studies in medicines development for Alzheimer’s disease. Alzheimers Res Ther. 2015;7(1):67.

16. Sheiner LB. Learning versus confirming in clinical drug development. Clin Pharmacol Ther. 1997;61(3):275-291.

17. Potter R, Patterson BW, Elbert DL, et al. Increased in vivo amyloid-beta42 production, exchange, and loss in presenilin mutation carriers. Sci Transl Med. 2013;5(189):189ra77.

18. Portelius E, Zetterberg H, Dean RA, et al. Amyloid-beta(1-15/16) as a marker for gamma-secretase inhibition in Alzheimer’s disease. J Alzheimers Dis. 2012;31(2):335-341.

19. Sevigny J, Chiao P, Bussiere T, et al. The antibody aducanumab reduces Abeta plaques in Alzheimer’s disease. Nature. 2016;537(7618):50-56.

20. Sperling RA, Dickerson BC, Pihlajamaki M, et al. Functional alterations in memory networks in early Alzheimer’s disease. Neuromolecular Med. 2010;12(1):27-43.

21. Wang J, Logovinsky V, Hendrix SB, et al. ADCOMS: a composite clinical outcome for prodromal Alzheimer’s disease trials. J Neurol Neurosurg Psychiatry. 2016;87(9):993-999.

22. Cummings J. The role of biomarkers in Alzheimer’s disease drug development. Adv Exp Med Biol. 2019;1118:29-61.

23. Mitchel JT. FDA relations during drug development. Dialogues Clin Neurosci. 2000;2(3):213-217.

24. Galvin JE, Meuser TM, Morris JC. Improving physician awareness of Alzheimer disease and enhancing recruitment: the Clinician Partners Program. Alzheimer Dis Assoc Disord. 2012;26(1):61-67.

JC has provided consultation to Acadia, Accera, Actinogen, Alkahest, Allergan, Alzheon, Avanir, Axsome, BiOasis Technologies, Biogen, Diadem, EIP Pharma, Eisai, Genentech, Green Valley, Grifols, Hisun, Idorsia, Kyowa Kirin, Lilly, Lundbeck, Merck, Otsuka, Proclara, QR, Resverlogix, Roche, Samus, Samumed, Sunovion, Suven, Takeda, Teva, Toyama, and United Neuroscience pharmaceutical and assessment companies.  JC acknowledges funding from the National Institute of General Medical Sciences (Grant: P20GM109025) and support from Keep Memory Alive.
KZ is the Chief Executive Officer of CNS Innovations and has provided consultation to Green Valley Pharma.