Expert perspectives on the design and communication of endpoints in Alzheimer's disease clinical trials

Abstract

Disease‐targeted therapies (DTTs) for Alzheimer's disease (AD) represent a major advance, yet challenges remain on how best to interpret and communicate their benefits. As clinical development increasingly targets the earliest stages of AD (before symptoms emerge and early symptomatic AD), there is a need for novel, complementary endpoints and improved communication of treatment benefits. The Global CEO Initiative on Alzheimer's Disease convened a multi‐stakeholder working group to develop recommendations for improving the design and communication of endpoints in AD clinical trials. In the near term, communication can be improved by contextualizing treatment effects at the individual level, including framing benefits in terms of responder analyses and “time saved” over defined timeframes. In parallel, continued development of complementary endpoints is needed to enable more sensitive, scalable, and patient‐centered measurements aligned with early intervention. These recommendations aim to support stakeholders in evaluating and communicating the meaningful benefits of current and emerging therapies.

Highlights

• Global CEO Initiative on Alzheimer's Disease (CEOi) convened a multi‐stakeholder working group to develop recommendations for Alzheimer's disease (AD) endpoint design and communication

• Earlier intervention and disease‐targeting therapies are reshaping AD outcome measurement needs

• The working group identified two priorities: improving communication of existing measures and advancing novel endpoints

• Patient‐centered approaches, including responder analyses and time‐saved estimates, may improve interpretation of treatment benefits

• Novel endpoints should be sensitive to early disease changes, scalable, accessible, and aligned with patient and care partner priorities

1. BACKGROUND

Advances in treatments for Alzheimer's disease (AD) reinforce the need to convey the meaning of clinical trial results to multiple stakeholders, including patients, care partners, clinicians, and payers. Clinicians treating Alzheimer's patients are faced with the difficulty of translating between‐group differences on scales that are not used in clinical practice into concepts that will resonate with the individual patient. Widely used endpoints that span the full clinical spectrum of symptomatic AD, such as the Alzheimer's Disease Assessment Scale‐Cognitive Subscale (ADAS‐Cog) and the Clinical Dementia Rating‐Sum of Boxes (CDR‐SB), were developed for clinical research studies, including clinical trials designed to seek regulatory approval for marketing. ¹ , ² These instruments are validated and have been effective for detecting treatment effects in symptomatic AD populations within the context of clinical trials. However, they are time‐consuming and complex to administer, and their scores are not easily translated into concepts that are intuitive for patients and families.

The emergence of disease‐targeted therapies (DTTs) (also referred to as disease‐modifying therapies or, at present, amyloid‐targeting therapies) for early symptomatic AD has added renewed urgency to the need to improve how treatment benefits and risks are interpreted and communicated. ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ Unlike prior symptomatic therapies, these agents are intended only for individuals with mild symptoms and they do not produce a short‐term improvement in symptoms, requiring clinicians to explain what moderate slowing of disease progression means at an earlier stage. Commonly reported trial metrics, such as a “30% slowing of clinical decline,” reflect mean differences between treatment and placebo groups. They are not primarily intended to inform patients and care partners making individual treatment decisions. For individuals with early symptomatic AD, which includes mild cognitive impairment (MCI) or mild dementia due to AD, a clear understanding of trial results is essential for informed, values‐based decision‐making that balances the potential benefits and risks of treatment. This underscores the need for approaches that translate clinical trial outcomes into terms that clearly and accurately communicate what treatment effects may mean for day‐to‐day function, future planning, and quality of life.

Moreover, there is a growing focus on the development of interventions for the “preclinical” or “presymptomatic” stages of AD, defined as cognitively unimpaired individuals with biomarker evidence of AD pathology, corresponding to the United States Food and Drug Administration (FDA) Stages 1 and 2. ¹¹ , ¹² , ¹³ At this stage of the disease continuum, any clinical change is subtle and difficult to detect, posing challenges for outcome measurement. Several cognitive composite measures have been developed specifically to assess preclinical AD. ¹⁴ , ¹⁵ , ¹⁶ These composites represent an important advance and provide valuable tools for studying this stage of disease. Building on this foundation, there is an opportunity to further evolve endpoint development by introducing complementary, next‐generation measures that leverage novel modalities, assess additional cognitive and functional domains, and address some of the practical limitations of existing measures, such as their complex administration and need for specialized rater training, which can place strain on clinical trial sites. This is particularly critical as the number of preclinical studies requiring large populations continues to grow.

With this as scientific and drug development context, the Global CEO Initiative on Alzheimer's Disease (CEOi) convened a multistakeholder working group at the 2025 Lausanne XII Workshop to consider how outcome measurement and communication should co‐evolve with the field. The working group brought together patient advocates, clinicians, and academic experts with experience in AD clinical care and outcome measurement, trial sponsors, and experts in clinical trial design and measurement science. The initial recommendations developed at the Lausanne XII Workshop were revised and refined by the working group, and the recommendations presented here reflect the shared perspectives of its members.

2. THE EVOLVING MEASUREMENT LANDSCAPE

Scientific and technological advances are reshaping how AD is assessed in both clinical trials and clinical care. Blood‐based biomarkers (BBMs) for AD now enable scalable detection of underlying pathology through a simple blood draw. These tools are already being incorporated into clinical practice and clinical trial designs. ¹¹ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ As the evidence base continues to grow, BBMs may also play an important role in monitoring treatment response.

In parallel, digital cognitive assessments (DCAs) have emerged to address some of the practical limitations of conventional clinician‐administered instruments. Traditional assessments often require substantial training, administration time, calibration, and scoring oversight, creating barriers to cognitive evaluation in clinical care and/or operational complexity in clinical trials. These challenges can be particularly burdensome for smaller or resource‐constrained trial sites, which may struggle to absorb the costs of extensive rater training and maintain consistent measurement quality over time. DCAs offer the advantage of immediate automated scoring, the ability to capture more nuanced cognitive performance data, and greater standardization. ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ Moreover, they can be deployed remotely and repeatedly to detect subtle yet clinically meaningful cognitive changes in more naturalistic, real‐world settings—an advantage that is particularly relevant in the earliest stages of disease. ²⁵

Together, BBMs and DCAs represent a shift toward a more biologically informed and scalable model of AD assessment. This shift is particularly critical given the expanding global clinical trial landscape. In 2025, approximately 185 AD trials were expected to be underway in some form, largely focused on early symptomatic populations. ¹² Conducting these studies worldwide would require randomizing more than 50,000 participants and screening an estimated 450,000–550,000 individuals. Advances such as BBMs and DCAs are essential to making research at this scale feasible. These tools also have the potential to support more decentralized and flexible trial designs by enabling remote and lower‐burden data collection that allows assessments to be conducted outside traditional trial settings, thereby easing participation burdens such as travel, expanding geographic reach, and improving representation in clinical research. ¹⁹

Furthermore, with DTTs now on market for individuals with early symptomatic AD and therapeutic development advances toward preclinical intervention, far larger numbers of patients will need to be evaluated earlier in the disease course to determine treatment eligibility. This evolution places increasing importance on primary care, as this setting is often the first point of contact for individuals with emerging cognitive concerns. The scalability and accessibility of BBMs and DCAs make this role for primary care more feasible. Collectively, these tools hold promise for bringing clinical research and clinical practice into closer alignment by supporting a shared language for assessing disease progression and therapeutic benefit.

However, both classes of tools remain in relatively early stages of validation for clinical trials and real‐world clinical use, and the evidence needed to fully understand their utility continues to evolve. For example, recent findings from the EVOKE and EVOKE+ trials of semaglutide in symptomatic AD demonstrated that, while BBM data tracked with clinical assessment outcomes and showed no treatment benefit relative to placebo, some cerebrospinal fluid (CSF) biomarkers revealed modest differences between groups. ²⁸ These findings and others underscore the need for further research to clarify the relationships among biomarkers, clinical measures, and underlying disease processes, as well as their respective sensitivities to treatment effect, including how the mechanism of action of a given intervention determines the appropriateness and interpretability of these measures in the intended study population. ²⁹

While there is substantial opportunity to generate new insights through novel measures, digital or otherwise, developing the necessary evidence base to support their regulatory acceptance as key endpoints will take time. This includes ensuring that they capture outcomes meaningful to patients and care partners in an accessible way and likely evaluating how these tools compare to existing cognitive assessments. In some cases, particularly when novel tools are designed to improve sensitivity or detect changes not captured by traditional measures, evaluation frameworks may need to emphasize comparability over equivalence when benchmarking against legacy measures.

3. RECOMMENDATIONS FROM WORKGROUP PARTICIPANTS

The following recommendations were developed to support trial sponsors, clinicians, researchers, patient advocates, and regulators in assessing and communicating the meaningful benefits of current and emerging AD therapies and provide a forward‐looking framework to guide the continued evolution of endpoint design (Table 1).

TABLE 1.

Working group recommendations.

Recommendation	Specific action
Improve communication of trial endpoints	Use responder analyses to convey treatment efficacy as the likelihood that an individual patient remains progression‐free over defined timeframes Leverage time savings analysis to communicate treatment effects in terms patients and care partners intuitively understand Frame treatment benefit at the individual level to better reflect the variability of AD and make treatment benefits more patient‐centered and actionable Optimize the use of minimal clinically important differences (MCIDs) and other defined thresholds to ensure that interpretations of treatment effect align with outcomes that patients and care partners perceive as most meaningful Develop communication strategies that consider diverse stakeholder needs and perspectives
Build toward novel, complementary endpoints	Develop new clinical endpoints that reflect the shift toward earlier intervention Incorporate patient and care partner input Prioritize accessibility to achieve broad adoption Capture individual variability to provide a more granular view of trajectories and distinguish clinically meaningful change Establish clear standards for validating and demonstrating comparability of next‐generation endpoints Initiate proactive dialogue with regulators and other stakeholders

3.1. Improve communication of trial endpoints

3.1.1. Use patient‐friendly analyses to translate treatment benefit into patient‐centered outcomes

To more accurately convey the real‐world impact of AD therapies, group‐level trial results should be complemented by analyses that translate findings into more interpretable, patient‐centered outcomes for clinicians, patients, and care partners. Approaches such as responder analyses and time‐savings estimates can help contextualize treatment effects in ways that are more tangible and meaningful for decision‐making. Rather than focusing solely on average changes in cognitive or functional scale scores, these strategies provide insight into how treatment may influence outcomes that matter most to patients.

Responder analyses identify the proportion of individuals who achieve a clinically meaningful outcome over a defined period. In the context of AD, responder analyses can be used to quantify the percentage of treated individuals who remain below a prespecified threshold of disease progression, offering a more intuitive understanding of therapeutic benefit than mean between‐group differences alone. Importantly, the thresholds used in these analyses should reflect outcomes that are meaningful to patients and care partners, such as maintaining independence in activities of daily living or delaying progression to the next stage of disease. Table 2 provides an example of a patient‐centered explanation of a responder analysis focused on preserving independence in daily activities.

TABLE 2.

Examples of patient‐friendly explanations of analyses used to communicate treatment benefit.

Parameter	Example result	Patient‐centered explanation
Responder analysis	25% of patients receiving treatment versus 33% on placebo experienced a clinically meaningful increase in dependence over 18 months (defined as ≥X‐point worsening on a functional scale)	“Over 18 months, about 3 out of 4 people receiving the treatment maintained their current level of independence in daily activities, compared with about 2 out of 3 people who did not receive the treatment.”
Time saved estimate	∼6‐month delay in clinical decline (defined as ≥X‐point worsening on a cognitive or functional scale) over 18 months versus placebo	“With treatment, people may experience the level of decline at 18 months that might otherwise have occurred about 6 months earlier without treatment.”
Goal attainment scaling	Participants receiving treatment showed significantly less decline on the ADCS‐MCI‐ADL * compared with placebo at 18 months.	“You mentioned that independently managing your finances is an important goal for you. In the 18‐month trial, people receiving treatment showed less decline in daily activities compared with those not receiving treatment. This may mean that treatment could help you maintain your ability to manage important daily activities for longer, although individual experiences may vary. We can monitor these activities over time to assess how the treatment is affecting you.”
HR	HR = 0.75 for progression to AD dementia over 18 months with treatment vs placebo	“Over the next year and a half, your rate of progression to the next stage of disease may be about 25% lower with treatment than without treatment.”

Abbreviations: AD, Alzheimer's disease; ADCS‐MCI‐ADL, the Alzheimer's Disease Cooperative Study–Mild Cognitive Impairment Activities of Daily Living; HR, hazard ratio.

^*ADCS‐MCI‐ADL is a commonly used caregiver‐reported measure designed to assess an individual's ability to perform everyday activities and maintain functional independence.

Time‐savings analyses provide another patient‐friendly way to analyze treatment benefit by estimating how much additional time treatment may provide before a person reaches a meaningful milestone in cognitive or functional decline. ⁶ , ⁷ , ³⁰ For example, rather than describing treatment benefit solely as slowing decline on a cognitive or functional scale, clinicians may explain that, on average, individuals receiving therapy might experience a level of decline at 18 months that might otherwise have occurred approximately 6 months earlier without treatment (Table 2). The relevance of a given time‐savings estimate will vary depending on an individual's priorities, goals, and life circumstances. Thus, a 6‐month delay in progression may be contextualized in terms of outcomes that are personally meaningful to the patient (e.g., additional time participating in valued activities or delaying the need for increased caregiving support) to help patients and care partners better understand the practical implications of treatment effects in everyday life. Notably, qualitative analyses assessing patient and caregiver impact have not historically been designed to incorporate time‐savings frameworks, representing an important area for future work.

Earlier treatment initiation has been shown to increase the amount of “time saved,” as slowing decline during earlier stages of disease preserves disproportionately more functional time than treatment initiated at later stages. ⁸ However, to ensure accurate interpretation, time‐savings explanations must be explicit about what outcomes are being delayed and transparent about the limitations of the trial window. Estimates of time saved are often bounded by the study period, and shorter durations may therefore underestimate long‐term benefits. This is illustrated by analyses of the anti‐amyloid DTTs donanemab and lecanemab, for which initial analyses estimated approximately 5 to 6 months of time saved over an 18‐month study period. ³¹ , ³² Subsequent longitudinal modeling of continued lecanemab treatment estimated substantially greater time savings over a 10‐year period. ³³ These models estimated time savings of 2.5 years for progression from MCI due to AD to mild AD and 3.5 years for progression to moderate AD. Among individuals in the low‐amyloid group (< 60 centiloids), the estimated time savings increased to 6.0 and 8.3 years, respectively. Longer‐term results from the TRAILBLAZER‐ALZ 2 long‐term extension similarly showed greater time savings with donanemab beyond the initial 18‐month trial period. ³²

Clinical staging frameworks may further support interpretation of responder and time‐savings analyses by anchoring treatment effects to recognizable phases of disease progression. In oncology, treatment benefit is often communicated in terms of delaying progression to a more advanced disease stage or increasing the probability that an individual will remain in their current disease stage over a defined period. A similar conceptual framework may help patients and clinicians interpret therapeutic benefit in AD by framing treatment effects as delaying progression between clinically meaningful stages of disease. However, unlike oncology, AD progression is more heterogeneous and transitions between stages are often less discrete. Consequently, staging frameworks in AD should be applied cautiously and supplemented with individualized assessments of cognitive and functional change to ensure that projected benefits remain clinically meaningful and reflective of real‐world patient experience.

3.1.2. Align communication of treatment effects in clinical trials with patient goals and priorities

Aligning communication of treatment effects with patient goals and priorities requires translating clinical trial outcomes into concepts that are relevant to an individual's everyday life. Group‐level averages, although essential for evaluating therapeutic efficacy in clinical trials, often fail to reflect the highly individualized experience of AD or the variability in outcomes that patients prioritize. ⁹ , ³⁴ , ³⁵ Several communication approaches may support this effort, including risk‐reduction framing, goal‐attainment scaling, and simplified explanations of hazard ratios. Together, these strategies help contextualize treatment effects in ways that better support patient understanding, individualized care discussions, and shared decision‐making.

Framing outcomes in terms of risk reduction or goal attainment improves understanding of therapeutic benefit by shifting the focus from numerical changes on cognitive or functional scales to outcomes that patients perceive as personally meaningful. ³⁶ , ³⁷ , ³⁸ Risk‐reduction framing can help patients contextualize the likelihood of avoiding or delaying outcomes they want to avoid most, such as losing the ability to live independently. Similarly, goal‐attainment scaling provides a structured approach for identifying and monitoring patient‐specific priorities over time. This approach allows patients and care partners to identify goals they value the most, such as maintaining the ability to drive or manage finances, and assess whether a therapy supports those goals over time (see Table 2 for an example).

These approaches may also support more balanced discussions regarding the relationship between treatment benefit and treatment‐associated risks. Many clinicians and patients have difficulty contextualizing the safety risks associated with DTTs relative to their potential benefits. Framing benefits in terms of postponing outcomes that patients want to avoid may provide a more meaningful reference point against which risks can be weighed. When patients and care partners understand benefit as the possibility of delaying outcomes they consider most consequential, they may evaluate treatment‐related risks differently, enabling more informed, values‐aligned decisions that reflect both potential benefit and acceptable tradeoffs.

Simplified explanations of statistical measures commonly reported in clinical trials, such as hazard ratios, may further improve communication of treatment effects. Hazard ratios are often difficult for patients to interpret in practical terms. Translating these results into more accessible language, including absolute terms or natural frequencies, may therefore improve understanding of the likelihood of benefit over a defined timeframe. For example, rather than presenting a hazard ratio alone, clinicians may explain how treatment may reduce the rate of progression to a more advanced stage of disease over a certain time period (Table 2).

3.1.3. Optimize the interpretation and use of minimal clinically important differences

The concept of minimal clinically important differences (MCIDs) is frequently used to contextualize clinical change by defining the smallest within‐patient change on an outcome measure that is considered meaningful to patients, care partners, or clinicians that would justify a change in clinical management. ³⁹ , ⁴⁰ When appropriately applied, MCID thresholds provide an important framework for interpreting disease progression and treatment response at the individual level. Importantly, MCID is an individual‐level construct and cannot be used to interpret differences between treatment groups or the clinical meaningfulness of average treatment effects. ⁴¹ , ⁴²

In AD, MCID thresholds have been derived largely from clinician‐anchored assessments rather than patient‐defined priorities, which may limit alignment with outcomes that individuals and care partners perceive as most meaningful. ⁴³ Interpretation of MCID thresholds must consider the characteristics and limitations of the clinical measures on which they are based. A commonly cited MCID for mild dementia is 1.63 points on the CDR‐SB. However, because clinical progression is very slow, the CDR‐SB increases by about this amount in 18 months. Therefore, completely stopping progression over an 18‐month trial would be considered only borderline meaningful. ⁴⁴ Furthermore, the CDR‐SB does not explicitly incorporate the timing of clinical progression, which is often highly relevant to patients and care partners. These limitations underscore the importance of careful interpretation and application of MCID frameworks to ensure alignment with achievable benefits and outcomes individuals value most. Rigorous patient experience data, such as that developed by the Alzheimer's Disease Patient and Caregiver Engagement (AD‐PACE), can help ensure MCID frameworks are applied in ways that accurately reflect what people living with AD and their care partners identify as meaningful.

3.1.4. Consider diverse stakeholder needs and perspectives

Interpretation and communication of trial outcomes must be tailored to the needs of distinct audiences. Diagnosis and management of AD involves a range of clinicians, including neurologists, geriatricians, neuropsychologists, geriatric psychiatrists, and primary care providers, each with different priorities, levels of expertise, and decision‐making contexts. Primary care clinicians, for example, may favor concise, actionable interpretations that support shared decision‐making, whereas specialists may seek more detailed data to inform treatment selection and longitudinal management. Trial endpoints must therefore be meaningful and accessible to a diverse clinical audience, enabling physicians to translate trial results into patient‐centered discussions of therapeutic benefit.

Beyond clinical practice, other stakeholder groups also engage with trial outcomes for different purposes and require distinct modes of communication. Patients and care partners benefit most from clear, concrete explanations that link trial results to everyday functioning and quality of life. Payers and regulators prioritize consistent and transparent framing of outcomes that can be mapped to functional ability, health system impact, or resource utilization. Advocates and sponsors, meanwhile, play a critical role in maintaining clarity and alignment across educational materials and public‐facing messaging.

Because each group uses trial measures differently, flexible and audience‐specific communication strategies are essential. Ongoing efforts to better understand stakeholder perspectives, alongside the development of shared standards for translating treatment benefits across audiences, will be critical to advancing more meaningful interpretation and use of trial outcomes.

3.2. Build toward novel, complementary endpoints

3.2.1. Develop new clinical endpoints that reflect the shift toward earlier intervention

The modernization of AD measurement requires developing new and complementary clinical endpoints that are sensitive to subtle early changes in clinical status, while also being practical and scalable for use across both research and routine clinical care. Importantly, improving outcome measurement will require fundamentally new approaches that capture patient‐relevant constructs, leverage emerging biological and digital innovations, and align with an evolving understanding of AD pathophysiology and disease progression.

As interventions move earlier in the disease course, endpoint design must also account for the continuous and often fluctuating nature of cognitive and functional change in AD, particularly during the preclinical stage (i.e., FDA Stages 1 and 2) when progression is subtle and non‐linear. In addition, if DTTs for early AD become part of standard clinical care, future trials may increasingly rely on active‐comparator or add‐on designs rather than untreated control groups, potentially attenuating observable differences in clinical decline between study arms. This may make incremental benefits of investigational therapies more difficult to capture over typical trial durations. These considerations highlight the need for sensitive endpoints that can capture meaningful changes in disease trajectory over time and reliably detect treatment effects within practical study durations. Such endpoints may reduce reliance on increasingly lengthy trials while still providing robust evidence of therapeutic benefit.

A particular challenge in preclinical AD is that sensitivity to early change does not necessarily equate to immediate clinical meaningfulness. A measure may detect a subtle difference in cognition or function before an individual experiences a change that they or their care partner would consider meaningful. One approach to addressing this challenge is to evaluate whether sensitive measures can predict future meaningful clinical outcomes. For example, a digital cognitive measure that detects subtle changes in preclinical AD may be clinically meaningful if those changes are associated with subsequent decline in function or loss of independence. In this context, an individual may not yet experience meaningful functional impairment at the time of assessment, but a sensitive measure may identify changes that signal an increased likelihood of meaningful impairment in the future. Establishing these relationships can help bridge the gap between sensitivity to early change and clinical meaningfulness.

The potential for sensitive endpoints to support early intervention with treatments that alter disease trajectory may also necessitate a shift in how clinical meaningfulness is understood. As patients and other stakeholders recognize the importance of intervening during the window of opportunity to prevent or delay potentially irreversible disease progression, the clinical value of altering an early disease trajectory may be better captured through the concept of “time saved” (see Section 3.1.1 for further discussion). Altering disease trajectory can delay downstream functional milestones, so translating subtle point differences into estimates of time saved may provide an additional bridge between statistical sensitivity and real‐world, patient‐centered impact.

To further support this effort, novel endpoints ought to be built upon the substantial literature that exists on what matters to Alzheimer's patients and care partners at the earliest disease stages, with specific endpoints co‐designed using standard outcomes research methods (see Section 3.2.2 for further discussion on co‐design). ⁶ , ⁹ , ¹⁰ , ⁴⁵

3.2.2. Co‐design with patients and care partners

Co‐design with patients and care partners can help ensure that new endpoints capture outcomes people care about. Their perspectives are critical for identifying which changes feel meaningful, which tradeoffs are acceptable, and how burden and benefit should be balanced. Incorporating these perspectives early can also improve face validity and acceptance of new tools among both patients and clinicians. ⁴⁶ Moreover, patient and public involvement is increasingly expected by major journals and reporting frameworks, including the Consolidated Standards of Reporting Trials (CONSORT) 2025 guidelines, which recommend reporting how patients and the public were involved in trial design, conduct, and dissemination. ⁴⁷

Importantly, co‐design efforts should include diverse representation across the target patient population, as well as a range of care partner perspectives (e.g., spouses and adult children), to ensure that endpoints meet the needs of different demographic and social groups. Co‐design workshops, in‐depth interviews, and focus groups can be used to elicit lived‐experience themes, map them to potential endpoints, and assess face validity with patients and care partners. Within these settings, techniques such as concept elicitation and cognitive interviewing can help identify outcomes that are meaningful to patients and care partners and assess how they interpret proposed endpoint concepts or assessment tasks. ⁴⁸ , ⁴⁹ Patient preference methods, such as ranking exercises, can be incorporated to characterize preferences and tradeoffs among endpoint characteristics, including accuracy, burden, frequency, and mode of administration. Participatory and iterative approaches can further engage patients and care partners in designing assessment tasks, digital interfaces, and testing procedures, with usability testing in real‐world settings used to refine candidate endpoints. Together, these approaches support the development and refinement of candidate endpoints to ensure that they capture changes that are meaningful to patients and care partners while balancing measurement performance, feasibility, and burden.

3.2.3. Design for accessibility to achieve broad adoption

For novel endpoints to achieve broad adoption, their design must prioritize accuracy and accessibility. New tools should be rapid to administer, interoperable across technologies, and robust across languages and cultures. They must also be accessible to individuals with sensory limitations, such as hearing or vision loss; feasible in a range of clinical environments, ranging from specialized memory clinics to primary care and community‐based settings; and acceptable in communities with lower levels of trust in healthcare or limited internet connectivity. Remote testing approaches may help expand inclusion by reaching populations in regions with lower clinical trial capacity and may provide greater ecological validity, although they may also introduce challenges related to participant compliance. Importantly, successful implementation of new tools will require appropriate training for both participants and clinical staff. Incorporating these considerations from the outset will help ensure that promising measures extend beyond highly controlled research settings and into real‐world clinical practice.

3.2.4. Consider individual variability in endpoint design

Critical to endpoint design is a more nuanced approach to individual variability. Personalizing baseline assessments by accounting for cognitive reserve, demographic factors, and prior performance can help distinguish clinically meaningful change from expected fluctuation. Digital tools are especially well suited for this task, as they enable repeated, low‐burden assessments that provide a more granular view of individual trajectories over time. They also offer opportunities to reduce practice effects through adaptive tasks, varied stimuli, and real‐world performance measures—features that can improve sensitivity to early decline and treatment effects at the individual level. ²¹ , ²⁵ , ⁵⁰

3.2.5. Establish clear standards for next‐generation endpoints

Rethinking what constitutes the “gold standard” for assessing clinical status is central to the modernization of AD measurement. Legacy measures have provided an essential foundation for evaluating disease progression and treatment effects. However, as therapeutic goals evolve toward earlier detection and more nuanced characterization of disease trajectories, future endpoints should ideally surpass these legacy scales in sensitivity, relevance, and feasibility.

Greater regulatory clarity regarding expectations for demonstrating comparability across measures, including approaches for validating tools that assess novel domains or characterize established domains in new ways, will be critical to facilitating innovation while maintaining confidence in clinical trial outcomes. Establishing transparent standards for evidentiary thresholds, cross‐measure validation strategies, and integration of multidomain endpoints could help ensure that next‐generation assessment tools are both scientifically robust and clinically meaningful.

In parallel with clear regulatory standards, comparative observational studies will be needed to validate emerging tools and support their use in clinical trials. An observational study platform could generate independent, generalizable evidence on cross‐sectional agreement, longitudinal sensitivity to change, reliability, feasibility, equity, and cost across the disease continuum.

3.2.6. Initiate transparent collaboration with regulators and other stakeholders

Successful development and deployment of novel endpoints will depend on early and ongoing dialogue among key stakeholders. Regulators, trial sponsors, trial sites, technology developers, and clinician end users all bring different perspectives, incentives, and constraints. Without intentional consideration, promising endpoints risk remaining confined to early pilot studies rather than advancing into routine trial use and clinical decision‐making.

Central to this effort is proactive engagement with regulators to determine how novel endpoints should be benchmarked, validated, and ultimately incorporated into regulatory and clinical decision‐making. Clear and early dialogue is needed to align on evidentiary expectations and to establish pathways for evaluating new measures alongside legacy scales. We also advocate for transparent communication about the methods and validation approaches used to develop and evaluate novel measures, as well as sharing of implementation experience. Such open communication will not only help facilitate broad adoption of well‐validated measures but also help enable meaningful comparison of outcomes across trials.

Broader stakeholder engagement is also essential to successful implementation. Sponsors may be reluctant to act as early adopters of unfamiliar measures without confidence in regulatory acceptance, while trial sites and clinicians must balance workflow constraints, burden, and interpretability. Technology developers need clear signals about usability, validation requirements, and integration into clinical and trial infrastructures. Collaboration across these groups can help align expectations, reduce uncertainty, and smooth the transition from methodological innovation to scalable implementation.

4. CONCLUSION

AD clinical trials must evolve to conform to a changing standard of care that features include earlier detection, DTTs, and patient‐centered models of care. Two strategies to achieve this are improving how existing measures are communicated and developing novel endpoints that capture patient and care partner priorities and better align clinical research with clinical practice. To implement the recommendations provided in this report, more dialogue and commitment to improving AD measurement among trial sponsors, clinicians, regulators, and advocates will be required. Each stakeholder group has a distinct role in this effort, and coordinated efforts across these groups are necessary to improve understanding and decision‐making in this new era of AD therapeutics.

CONFLICT OF INTEREST STATEMENT

P.N.T. reports receiving consulting fees from AbbVie, AC Immune, Acadia, Athira, Axsome, Biogen, Bristol Myers Squibb, Cognition Therapeutics, Cognito, Corium, CuraSen, Eisai, EMA Wellness, Genentech, Immunobrain, Janssen, Lundbeck, MapLight, Merck and Co., Neumora, Novartis, Novo Nordisk, Otsuka Pharmaceuticals, ONO Pharma, Roche, and T3D Therapeutics, and a speaking fee from Novo Nordisk. R.P. has received honoraria for advisory boards and speaker engagements from Roche, EISAI, Eli Lilly, Biogen, Janssen‐Cilag, AstraZeneca, Schwabe, Grifols, Novo Nordisk, GSK, BMS, and Tabuk and holds shares of Medotrax GmbH and Vistim Labs Ltd. S.E.S. has not received any financial compensation from pharmaceutical or diagnostics companies since 2024, when she received honoraria for serving on scientific advisory boards on biomarker testing and education for Eisai and Novo Nordisk and speaking fees from Eisai, Eli Lilly, and Novo Nordisk. S.E.S has recently received honoraria for educational presentations from Medscape, PeerView, the Academy for Continued Healthcare Learning, Med Learning Group, and the Global Learning Collaborative and has provided unpaid scientific advising to Acumen, Biogen, Cognito Therapeutics, Danaher, Eisai, Eli Lilly, Johnson and Johnson Innovative Medicine, Sanofi, and Siemens. At the time of this work, R.P was an employee of UsAgainstAlzheimer's, which receives donations unrelated to this work from several pharmaceutical and diagnostics companies. L.H. is employed by the Global Alzheimer's Platform Foundation, has served as a consultant to Ixico, Dementia Discovery Fund, Violet Therapeutics, Therini Bio, Kesmalea, Aerska, and Endlyz, and serves on the advisory board of Oxford Brain Diagnostics. J.K. has received consulting fees from Biogen and Darmiyan and serves on an advisory board for Linus Health. J.K. has participated as an investigator in industry‐sponsored clinical studies, including the Generation Program (Novartis), the A4 Study (Eli Lilly), the AHEAD Study (Eisai), and the ENGAGE, EMERGE, and ADVANCE studies (Biogen). J.D. is President of the Global Alzheimer's Platform Foundation. J.T. is the founder and CEO of Voices of Alzheimer's. K.V.N. serves as a consultant for Genentech, EMD Serono, Eli Lilly, Global CEO Initiative on Alzheimer's Disease, Eisai, and PhRMA and has received speaking fees from Sanofi, Amgen, MJH Life Sciences, Eli Lilly, and the American Academy of Neurology. G.V. is the Convener of CEOi. B.T. is the Executive Director of CEOi. D.P. and E.S. are paid consultants for CEOi. N.L. is an employee and shareholder of ki elements GmbH. C.G. is an employee and shareholder of Bristol Myers Squibb. S.P.D. is an employee of Pentara Corporation, a firm that provides consulting services and has consulted for 80+ companies operating in the Alzheimer's diseases space. Author disclosures are available in the Supporting Information.

Supporting information

Supporting Information: alz71951‐sup‐0001‐ICMJE.pdf