This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
CLARITY AD is a randomized, parallel phase 3 treatment trial evaluating lecanemab in participants with early Alzheimer's disease. The registry lists 1,906 participants, 9 arms, quadruple masking, and both core-study and extension-phase primary endpoints.
| Feature | CLARITY AD |
|---|---|
| Trial name | CLARITY AD |
| ClinicalTrials.gov identifier | NCT03887455 |
| Title | A Study to Confirm Safety and Efficacy of Lecanemab in Participants With Early Alzheimer's Disease |
| Phase | Phase 3 |
| Condition | Early Alzheimer's Disease |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Enrollment | 1,906 |
| Number of arms | 9 |
| Lead sponsor | Eisai Inc. |
| Sponsor type | Industry |
| Status | Active, not recruiting |
2. Clinical Question
The central clinical question is whether lecanemab can be evaluated for safety and efficacy in participants with early Alzheimer's disease using a randomized phase 3 design, with change from baseline in the CDR-SB at 18 months serving as the core-study primary efficacy endpoint.
Population
Participants with early Alzheimer's disease.
Interventions
Lecanemab IV, placebo, and lecanemab SC are the interventions listed in the registry.
Design
Randomized, parallel-group phase 3 trial with quadruple masking and a treatment primary purpose.
Primary question
What is the change from baseline in CDR-SB at 18 months, and how are safety, longer-term CDR-SB change, and lecanemab exposure evaluated during the extension phases?
3. Trial Design
The nine-arm structure is an important statistical feature. A trial with multiple arms can contain several scientifically relevant comparisons, but the presence of multiple arms also makes the definition of the primary comparison, analysis population, and multiplicity strategy especially important. The registry information identifies the overall design and interventions but does not provide a posted statistical analysis plan describing how the nine arms are to be compared.
Core study and extension phases
The primary endpoints span more than the core 18-month study. The registry separately identifies an extension-phase safety endpoint, an extension-phase CDR-SB endpoint extending to Month 66, and an extension Part B exposure endpoint extending from the 48th month of extension phase Part A through the end of Part B, up to 24 months.
| Phase | Primary statistical focus | Time frame |
|---|---|---|
| Core Study | Change from baseline in CDR-SB | Baseline, 18 months |
| Extension Phase | Treatment-emergent adverse events | From first dose through approximately 51 months, including 3 months follow-up |
| Extension Phase | Change from Core Study Baseline in CDR-SB | Baseline up to Month 66 |
| Extension Phase Part B | Participants exposed to lecanemab | From the 48th month in extension phase Part A to the end of Part B, up to 24 months |
4. Interventions and Treatment Structure
Lecanemab IV
- Listed as a drug intervention.
- Included within the 9-arm randomized parallel design.
Placebo
- Listed as a drug intervention.
- Included within the randomized parallel design.
Lecanemab SC
- Listed as a drug intervention.
- Included within the randomized parallel design.
Nine randomized arms
- The registry lists 9 arms overall.
- The available record identifies the three intervention types but does not provide arm-specific enrollment counts here.
The distinction between intervention type and arm count matters. Three intervention names do not imply three arms: the registry specifies nine arms. Without arm-specific sample sizes or a posted statistical analysis plan, the relative size of each randomized comparison cannot be reconstructed from the registry information presented here.
5. Primary Endpoints
Unlike a trial with a single primary efficacy outcome, the registry identifies four primary endpoints spanning the core study, the extension phase, and extension Phase Part B.
| Primary endpoint | Registry time frame | Endpoint type |
|---|---|---|
| Core Study: Change from Baseline in the CDR-SB at 18 Months | Baseline, 18 months | Continuous change-from-baseline efficacy measure |
| Extension Phase: Number of Participants Reporting One or More Treatment-emergent Adverse Events (TEAEs) | From first dose of study drug up to approximately 51 months, including 3 months follow up, for the extension phase | Participant-level safety event measure |
| Extension Phase: Change from Core Study Baseline in CDR-SB | Baseline up to Month 66 | Longitudinal change-from-baseline efficacy measure |
| Extension Phase Part B: Number of Participants Exposed to Lecanemab | From 48th month in extension phase Part A to the end of extension phase Part B, up to 24 months | Treatment-exposure measure |
CDR-SB endpoint
The CDR-SB is the central efficacy outcome named in both the core-study and extension-phase primary endpoints. The core endpoint is explicitly defined as change from baseline at 18 months. The extension endpoint instead evaluates change from Core Study Baseline over a longer period, through Month 66.
Statistically, these are continuous change-from-baseline outcomes. The distinction between an 18-month endpoint and a longitudinal extension endpoint is important: the former focuses on a prespecified time point, while the latter describes change over a substantially longer follow-up period.
TEAE endpoint
The extension-phase safety endpoint is the number of participants reporting one or more treatment-emergent adverse events. The registry defines a TEAE as an adverse event that emerges during treatment or within 30 days of the last dose of study drug, was absent at pretreatment, reemerges during treatment after having stopped before treatment, or worsens in severity during treatment relative to the pretreatment state when the event was continuous.
The registry states that the number of participants with TEAEs, including serious and non-serious adverse events, will be reported based on regular measurement of vital signs, laboratory safety assessments, antidrug antibody assessments, suicidality assessments, magnetic resonance imaging, and electrocardiogram parameter values.
6. Planned Analysis
The ClinicalTrials.gov record does not contain posted statistical analyses or numerical trial results for the primary endpoints. The appropriate statistical framework therefore depends on the endpoint type rather than on observed estimates.
| Endpoint | Typical statistical approach | What the analysis estimates |
|---|---|---|
| Core-study change from baseline in CDR-SB at 18 months | A continuous-outcome comparison such as ANCOVA, or a longitudinal repeated-measures model when repeated observations are incorporated | Difference in mean change between randomized treatment groups, with uncertainty quantified by a confidence interval |
| Extension-phase change from Core Study Baseline in CDR-SB | Longitudinal repeated-measures modeling or a prespecified change-from-baseline comparison at relevant follow-up times | Difference in trajectories or changes over the extension period |
| Participants reporting one or more TEAEs | Descriptive safety summaries of participant counts and proportions; comparative methods may be used if prespecified | Frequency of participants experiencing at least one treatment-emergent adverse event |
| Participants exposed to lecanemab | Descriptive exposure accounting | Number of participants exposed during the specified extension Part B period |
Why the CDR-SB endpoint is statistically different from TEAEs
CDR-SB change is a quantitative outcome. A statistical analysis can therefore compare distributions or model mean changes between randomized groups. The TEAE endpoint is binary at the participant level: each participant either reports one or more qualifying TEAEs during the specified period or does not. The two endpoints require different statistical summaries even though both are designated primary endpoints in the registry.
Why the extension endpoint needs a longitudinal perspective
The extension CDR-SB endpoint covers baseline through Month 66 rather than stopping at the core-study 18-month assessment. Participants can therefore contribute measurements at multiple time points. A longitudinal analysis can use the repeated measurements jointly rather than treating every follow-up observation as an independent data point.
7. Statistical Methodology
Continuous change-from-baseline analysis
For the core-study CDR-SB endpoint, the statistical target is a change from baseline measured at 18 months. A common framework for this type of endpoint is an analysis of covariance, in which the follow-up value or change from baseline is compared across randomized groups while accounting for baseline CDR-SB.
The quantity of interest is a between-group difference in change. Its interpretation depends on the direction in which CDR-SB is defined and on the prespecified analysis convention.
Using baseline information can improve precision because participants enter the analysis with potentially different baseline CDR-SB values. Randomization remains the foundation of the treatment comparison; baseline adjustment is a precision-enhancing modeling step rather than a substitute for randomization.
Repeated-measures modeling
The extension endpoint follows CDR-SB from Core Study Baseline through Month 66. When multiple observations are available for the same participant, repeated-measures methods can account for within-participant correlation and estimate treatment differences over time.
This is a conceptual representation rather than a claim about the exact model used in the registry. The treatment-by-time component is particularly useful when the scientific question concerns differences in change over follow-up.
Binary safety endpoint
The TEAE endpoint counts participants reporting one or more treatment-emergent adverse events. For each participant, the endpoint can be represented as a binary indicator: one if at least one qualifying TEAE occurs during the specified period and zero otherwise.
Safety results are often summarized using counts and percentages by treatment group. Depending on the prespecified statistical analysis plan, between-group risk differences, risk ratios, odds ratios, or confidence intervals may also be reported. The registry information available here does not specify which comparative safety statistic is planned.
Exposure endpoint
The extension Phase B endpoint is a count of participants exposed to lecanemab during a defined period. This is fundamentally descriptive. Unlike a CDR-SB change score, it does not by itself quantify a difference in an outcome between randomized groups.
8. Statistical Methods Explained
Why is CDR-SB change treated as a continuous endpoint?
The registry defines the core efficacy endpoint as change from baseline in CDR-SB at 18 months. A continuous outcome preserves information about the magnitude of change rather than reducing participants to a binary classification such as improved versus not improved. This allows a treatment comparison to use the numerical structure of the CDR-SB measure.
Why might ANCOVA be used for the 18-month CDR-SB endpoint?
ANCOVA can compare treatment groups while accounting for baseline CDR-SB. This can reduce residual variability when baseline values are associated with follow-up values. The resulting treatment contrast is still anchored to the randomized groups.
Why might a repeated-measures model be useful for the extension CDR-SB endpoint?
The extension endpoint runs from baseline through Month 66. A participant can therefore contribute multiple CDR-SB measurements. A repeated-measures model can use those observations jointly and account for correlation among measurements from the same participant.
How is a TEAE endpoint different from CDR-SB change?
The TEAE endpoint asks whether a participant experienced at least one qualifying treatment-emergent adverse event during the specified period. It is therefore a participant-level binary outcome, whereas CDR-SB change is a quantitative outcome. The appropriate summary statistics and modeling strategies are consequently different.
Why is the number of exposed participants mainly descriptive?
The extension Phase B endpoint is explicitly the number of participants exposed to lecanemab during a specified period. A count can describe the extent of exposure, but it is not itself an effect estimate comparing clinical outcomes between randomized treatment groups.
Why does the nine-arm design matter statistically?
With 9 arms, there can be multiple possible pairwise or grouped comparisons. Every additional confirmatory comparison can affect the overall type I error unless the statistical plan specifies how the comparisons are organized. The registry information identifies the number of arms but does not provide the multiplicity strategy or arm-specific comparison hierarchy.
9. Analysis Populations and Missing Data
The registry information identifies the randomized design but does not provide posted definitions for an intention-to-treat population, per-protocol population, modified intention-to-treat population, or safety population. Those distinctions can materially affect interpretation of a clinical trial.
Randomized population
Randomization provides the primary basis for comparing treatment groups without introducing post-randomization selection into the allocation process.
Safety population
Safety analyses commonly depend on actual treatment exposure, but the specific population definition for this trial is not stated in the available registry information.
Missing CDR-SB values
Longitudinal trials can have missing assessments because participants discontinue, miss visits, or otherwise do not contribute a measurement at a scheduled time.
Imputation
The registry information does not specify a missing-data or imputation method for the CDR-SB endpoints.
For a continuous longitudinal endpoint, the assumptions made about missing observations can influence the estimated treatment effect. Common strategies include likelihood-based repeated-measures methods and sensitivity analyses under alternative missing-data assumptions. The choice should be prespecified because different assumptions correspond to different interpretations of what would have happened after missingness.
10. Multiplicity and Multiple Primary Endpoints
CLARITY AD lists four primary endpoints across the core study and extension phases. That structure makes multiplicity an important design consideration.
| Endpoint family | Primary endpoint | Statistical issue |
|---|---|---|
| Core efficacy | Change from Baseline in the CDR-SB at 18 Months | Continuous treatment comparison |
| Extension safety | Participants reporting one or more TEAEs | Participant-level safety frequency |
| Extension efficacy | Change from Core Study Baseline in CDR-SB | Longitudinal treatment comparison |
| Extension exposure | Participants Exposed to Lecanemab | Descriptive exposure accounting |
Multiplicity is not simply a matter of counting endpoints. It depends on which hypotheses are intended to be confirmatory, how they are ordered, whether they are tested jointly or hierarchically, and which analysis family shares a type I error budget. The ClinicalTrials.gov record does not describe an alpha-allocation or hierarchical testing procedure.
11. Extension-Phase Statistical Structure
The extension portion of CLARITY AD introduces a different statistical problem from the core 18-month efficacy assessment. Participants are followed for substantially longer periods, and the registry defines separate safety, CDR-SB, and exposure endpoints.
Trial start
The registry lists March 27, 2019 as the study start date.
18-month CDR-SB endpoint
The core-study primary endpoint is change from baseline in CDR-SB at 18 months.
Safety through approximately 51 months
The registry specifies TEAE reporting from first dose through approximately 51 months, including 3 months of follow-up for the extension phase.
Longitudinal CDR-SB endpoint
The extension-phase CDR-SB endpoint evaluates change from Core Study Baseline up to Month 66.
Lecanemab exposure
Participants exposed to lecanemab are counted from the 48th month in extension phase Part A through the end of extension phase Part B, up to 24 months.
Primary completion date
The registry lists June 30, 2029 as the primary completion date.
12. Safety Endpoint: Treatment-Emergent Adverse Events
The extension-phase safety endpoint is the number of participants reporting one or more TEAEs. The registry explicitly includes serious and non-serious adverse events in this endpoint.
| Safety component | Registry definition or assessment |
|---|---|
| TEAE | An adverse event that emerges during treatment or within 30 days of the last dose, was absent at baseline, reemerges after having stopped before treatment, or worsens in severity during treatment when continuous. |
| Serious and non-serious events | Both are included in the reported participant count. |
| Vital signs | Included among regular safety measurements. |
| Laboratory tests | Included among safety assessments. |
| Antidrug antibodies | Included among safety assessments. |
| Suicidality | Included among safety assessments. |
| Magnetic resonance imaging | Included among safety assessments. |
| Electrocardiogram parameters | Included among safety assessments. |
A participant-level "one or more TEAEs" endpoint deliberately collapses multiple adverse events within a participant into a single indicator for that endpoint. Consequently, this measure answers the question of how many participants experienced at least one qualifying event, not how many total events occurred or how frequently individual events recurred.
What a safety comparison can and cannot establish
If comparative rates are eventually reported, the difference between treatment groups can describe the observed frequency of participants with at least one TEAE. It would not, by itself, characterize event severity, duration, recurrence, timing, or the distribution of specific adverse-event categories. Those require additional safety summaries.
13. Extension CDR-SB: Why Long-Term Change Requires Careful Interpretation
The extension-phase CDR-SB endpoint is defined as change from Core Study Baseline through Month 66. This is not simply another 18-month endpoint. It extends the observation window and therefore raises additional questions about repeated measurements, treatment exposure, discontinuation, and missing observations.
Repeated observations
A participant may contribute multiple CDR-SB assessments over the extension period, creating correlated observations within participants.
Longer follow-up
Extending assessment to Month 66 allows the analysis to examine change over a much longer period than the core 18-month endpoint.
Attrition
Longer follow-up creates more opportunity for missed assessments or treatment discontinuation, making missing-data assumptions more important.
Interpretation
A long-term change estimate summarizes the trajectory represented by the participants and measurements included under the prespecified analysis framework.
14. Exposure Endpoint in Extension Phase Part B
The registry defines a primary endpoint for extension Phase B as the number of participants exposed to lecanemab from the 48th month in extension phase Part A through the end of extension phase Part B, up to 24 months.
Exposure endpoint
Defined period: from the 48th month in extension phase Part A through the end of extension phase Part B, up to 24 months.
This endpoint has a different statistical role from the CDR-SB endpoint. It documents treatment exposure rather than estimating a change in cognitive or functional status. The number of exposed participants can help characterize the population contributing to long-term treatment experience, but exposure itself should not be interpreted as an efficacy effect.
15. ClinicalTrials.gov Status and Study Timing
| Registry field | Value |
|---|---|
| Status | Active, not recruiting |
| Start date | 2019-03-27 |
| Primary completion date | 2029-06-30 |
| Enrollment | 1,906 |
| Phase | 3 |
| Lead sponsor | Eisai Inc. |
| Sponsor type | Industry |
The long primary-completion horizon is relevant when interpreting the registry as a living record. A trial can have a substantial enrollment and a clearly defined core endpoint while still having extension-phase objectives that remain active over a longer period.
16. Important Limitations and Interpretation Issues
- No posted outcome estimates: the registry does not provide numerical results for the listed primary endpoints.
- No posted statistical analysis output: the available record does not provide formal endpoint-specific statistical results or treatment-effect estimates.
- Nine-arm structure: the registry identifies 9 arms but does not provide arm-specific enrollment counts or the comparison hierarchy needed to understand every planned treatment contrast.
- Multiple primary endpoints: four endpoints are designated primary across the core and extension phases, but the available record does not specify the multiplicity strategy.
- Missing-data methodology: the available record does not specify the imputation or missing-data assumptions for the CDR-SB endpoints.
- Analysis populations: the available record does not provide the detailed efficacy and safety population definitions needed to distinguish all analysis sets.
- Longitudinal follow-up: the extension CDR-SB endpoint extends to Month 66, creating additional issues involving repeated measurements, discontinuation, and incomplete follow-up.
- Safety interpretation: the TEAE endpoint counts participants with at least one qualifying event and does not by itself describe the complete clinical profile of individual adverse events.
- Exposure interpretation: the Part B exposure endpoint is a participant count and should not be interpreted as a treatment-effect estimate.
17. Why This Trial Matters Statistically
CLARITY AD is a useful statistical teaching case because it combines a randomized parallel phase 3 design with a continuous core-study efficacy endpoint, long-term repeated CDR-SB assessment, a participant-level safety endpoint, and an exposure endpoint.
| Concept | How it appears in CLARITY AD |
|---|---|
| Randomization | The trial uses randomized allocation. |
| Parallel-group design | The registry identifies a parallel design model with 9 arms. |
| Masking | The study is quadruple masked. |
| Continuous endpoint | Core-study change from baseline in CDR-SB at 18 months. |
| Longitudinal analysis | Extension-phase change from Core Study Baseline in CDR-SB through Month 66. |
| Safety analysis | Number of participants reporting one or more TEAEs during the extension phase. |
| Exposure accounting | Number of participants exposed to lecanemab during extension Phase B. |
| Multiplicity | Four registry-designated primary endpoints span core and extension phases. |
| Missing data | Longitudinal CDR-SB assessment creates the need to specify assumptions for incomplete observations. |
| Analysis populations | Interpretation requires clear definitions of randomized, efficacy, and safety populations. |
18. Statistical Interpretation: What to Look For When Results Are Reported
For the 18-month CDR-SB endpoint, the central quantity would ordinarily be a between-group difference in change from baseline, accompanied by a confidence interval. The sign of that difference must be interpreted using the direction of the CDR-SB scale and the prespecified analysis convention.
A 95% confidence interval around a treatment difference describes statistical uncertainty around the estimated population-level contrast under the specified model and sampling framework. It does not describe the range of individual patient responses.
A p-value addresses the compatibility of the observed data with a specified null hypothesis under the statistical model. It does not measure the size of the treatment effect, the probability that a treatment works, or the magnitude of benefit experienced by an individual participant.
For the extension safety endpoint, the clinically relevant comparison would be the proportion of participants reporting at least one qualifying TEAE in each relevant treatment group. A participant with multiple TEAEs still contributes one participant to this particular endpoint.
19. What This Analysis Can Teach About Endpoint Selection
The four primary endpoints illustrate why a clinical trial can require several distinct statistical perspectives. A cognitive or functional score such as CDR-SB produces a quantitative change measure. A TEAE endpoint produces a participant-level safety frequency. An exposure endpoint documents treatment experience. These outcomes should not be reduced to a single generic statistical test.
Efficacy
CDR-SB change asks whether the treatment groups differ in a quantitative clinical measure over time.
Safety
The TEAE endpoint asks how many participants experience at least one qualifying treatment-emergent adverse event.
Durability
The extension CDR-SB endpoint follows change from Core Study Baseline through Month 66.
Exposure
The Part B endpoint documents how many participants were exposed to lecanemab during the specified extension period.
20. Why the Randomized Design Matters
Randomization is the principal design feature supporting causal comparison between treatment groups. When participants are randomized, baseline prognostic factors are expected to be balanced in expectation across treatment assignments, although finite randomized samples can still differ by chance.
The parallel design means that treatment groups are followed concurrently rather than having the same participant serve sequentially as both treatment and control. Quadruple masking is an additional design feature intended to limit knowledge of treatment assignment among relevant trial participants and personnel.
The statistical interpretation still depends on the prespecified analysis population, endpoint definition, missing-data strategy, and handling of multiple comparisons. Randomization establishes the foundation for causal inference; it does not remove the need for appropriate statistical analysis.
21. Trial Timeline
Study begins
The registry lists March 27, 2019 as the start date.
18-month primary efficacy assessment
Change from baseline in CDR-SB is the core-study primary endpoint at 18 months.
Safety and longitudinal CDR-SB
The extension includes TEAE assessment through approximately 51 months and CDR-SB change from Core Study Baseline through Month 66.
Lecanemab exposure assessment
The registry counts participants exposed to lecanemab from the 48th month in extension phase Part A through the end of Part B, up to 24 months.
Primary completion
The registry lists June 30, 2029 as the primary completion date.
22. Statistical Concepts in This Trial
The principal statistical concepts represented by CLARITY AD are randomized parallel-group comparison, continuous change-from-baseline analysis, longitudinal repeated-measures analysis, participant-level safety summaries, exposure accounting, missing-data methodology, and multiplicity across multiple primary endpoints.
| Concept | Application to this trial |
|---|---|
| Randomized comparison | Participants are allocated randomly within a phase 3 parallel design. |
| Change from baseline | The core primary endpoint measures change in CDR-SB from baseline at 18 months. |
| Longitudinal modeling | The extension primary endpoint follows CDR-SB from Core Study Baseline through Month 66. |
| Binary safety outcome | The extension TEAE endpoint identifies participants reporting one or more qualifying events. |
| Exposure analysis | Extension Phase B counts participants exposed to lecanemab during the defined period. |
| Multiplicity | Four primary endpoints require a prespecified framework for interpreting multiple hypotheses. |
| Missing-data assumptions | Longitudinal CDR-SB assessment requires an explicit strategy for incomplete measurements. |
| Analysis populations | Efficacy and safety interpretation depends on clearly defined analysis sets. |
23. Sources
- ClinicalTrials.gov: NCT03887455 — CLARITY AD.
Continue through the Clinical Biostats trial library
Explore additional clinical trial analyses and statistical methods across randomized clinical research.
24. Record Summary
CLARITY AD is a randomized, parallel phase 3 trial of lecanemab in participants with early Alzheimer's disease, with 1,906 participants and 9 arms. The registry identifies four primary endpoints spanning the core study and extension phases: change from baseline in CDR-SB at 18 months; participants reporting one or more treatment-emergent adverse events during the extension; change from Core Study Baseline in CDR-SB through Month 66; and the number of participants exposed to lecanemab during extension Phase B.
Statistically, the core CDR-SB endpoint is a continuous change-from-baseline outcome, while the extension CDR-SB endpoint is longitudinal and the TEAE endpoint is participant-level safety information. The exposure endpoint is descriptive. The nine-arm design and four primary endpoints also make the prespecified comparison structure and multiplicity strategy important to interpretation.