A Study to Evaluate the Effects of SAGE-718 in Participants With Mild Cognitive Impairment or Mild Dementia Due to Alzheimer's Disease (AD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Effects of SAGE-718 in Participants With Mild Cognitive Impairment or Mild Dementia Due to Alzheimer's Disease (AD) (NCT05619692), a randomized phase 2 study. The primary outcome is Change From Baseline in the Wechsler Adult Intelligence Scale-IV (WAIS-IV) Coding Test Score.

2. Clinical Question

The study asked whether cognitive performance, measured by change in the WAIS-IV Coding Test score from baseline to Day 84, differed between participants receiving SAGE-718 and those receiving matching placebo. The registry describes the intended population as participants with mild cognitive impairment or mild dementia due to Alzheimer’s disease. The reported primary analysis compared least-squares mean changes between groups.

The central statistical question is about the size and precision of that between-group difference, rather than whether either group changed from baseline on its own. A positive change on the test indicates better processing speed, but the comparison between treatment groups is the relevant contrast for this randomized study. The registry reports an estimated difference of 1.5 points in LS means, with a 95% confidence interval from -0.64 to 3.73 and p = 0.164. The interval includes no difference, so the reported result does not establish a clear between-group difference in this outcome.

3. Trial Design

This was a randomized, parallel-group, placebo-controlled Phase 2 study. The registry records quadruple masking, meaning that the trial was described as masked at multiple levels; the specific roles covered by that label are not reported in the registry record. The study is listed as completed. Its recorded start was 2022-11-29, and primary completion was 2024-06-05.

The registry lists enrollment of 174. The primary efficacy analysis included participants in the FAS, defined as participants in the Safety Set who received at least one dose of investigational product and had baseline and at least one post-baseline efficacy evaluation. The number analyzed for an outcome reflects participants with data available for that analysis; the posted primary arm counts are 80 for placebo and 76 for SAGE-718. The safety groups have different denominators, so efficacy and safety counts should not be treated as interchangeable.

The active-treatment schedule changed during the treatment period: participants received SAGE-718 1.2 mg orally once daily for the first 6 weeks, then 0.9 mg once daily through Day 84. The placebo group received matching placebo once daily through Day 84. The primary outcome was change from baseline in WAIS-IV Coding score, with the stated timeframe of baseline and Day 84.

The listed secondary safety outcomes include participants with at least one treatment-emergent adverse event, participants with such events by severity, and participants who withdrew due to these events. The registry record does not report the trial’s original sample-size justification or the assumptions used to plan enrollment. Details of randomization procedures beyond the randomized designation, allocation concealment, and the specific masking roles are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: 174 participants were randomized to receive either SAGE-718 or placebo, of which 4 participants were not treated.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAS included all participants in the Safety Set (which included all participants who were administered at least one dose of the IP) who had baseline and at least 1 post-baseline efficacy evaluation. Overall number of participants analyzed indicates number of participants with data available for analysis at a specified timepoint.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Full analysis set (FAS) included all participants in the Safety Set (which included all participants who were administered at least one dose of the investigational product \[IP\]) who had baseline and at least 1 post-baseline efficacy evaluation.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in the Wechsler Adult Intelligence Scale-IV (WAIS-IV) Coding Test Score, measured over Baseline, Day 84.

The primary endpoint was change from baseline in the Wechsler Adult Intelligence Scale-IV (WAIS-IV) Coding Test score, assessed at baseline and Day 84. This test measures processing speed. Participants use a key to match symbols with numbers and write the corresponding symbols; the score is based on correctly completed codes during a 120-second test. The score ranges from 0 to 135, and higher scores indicate better processing speed. Positive change from baseline therefore represents improvement in the measured score.

The registry describes the test as a valid and sensitive measure of cognitive dysfunction that correlates with real-world functional outcomes, such as the ability to accomplish everyday tasks, and recovery from functional disability. That makes it relevant to the study’s focus on cognition, while still leaving the endpoint as a test score rather than a direct measure of everyday function.

The primary outcome is continuous: it can take a range of score values rather than simply classifying each participant into a category. The sponsor’s reported analysis used least-squares means from a mixed-effects model for repeated measures. These adjusted model estimates are intended to compare average change between groups while using the repeated measurements specified in the analysis. The registry record does not provide further detail on the model’s covariates or its handling of missing observations.

7. Results Posted in the Registry

• SAGE-718: mean 5.30 versus 3.80 on control; difference 1.50 (95% CI -0.66 to 3.66), standardized effect d = 0.22; z-test p = 0.174.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Participants Who Withdrew From Study Due to TEAEs

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • SAGE-718: 2.4% (2/84) versus 2.3% (2/86) on control; risk difference 0.1 percentage points (95% CI -4.5 to 4.6), risk ratio 1.02 (0.15–7.10); unadjusted two-proportion z-test p = 0.981.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between SAGE-718 and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes concerning participants with at least one treatment-emergent adverse event, adverse events by severity, and withdrawals due to treatment-emergent adverse events. These outcomes provide safety and tolerability context; they address different questions from the primary cognitive-performance endpoint and do not directly confirm or refute its result.

The registry record does not report detailed results for the two listed outcomes concerning participants with at least one treatment-emergent adverse event and the severity of those events. Results for those endpoints are therefore not reported in the registry record. The withdrawal endpoint has posted group counts, but it is a sparse binary outcome and its unadjusted comparison is not a substitute for the primary model-based analysis.

Secondary findings generally carry less inferential weight when the registry does not report a prespecified testing hierarchy or adjustment for multiple comparisons. In general, testing several outcomes without accounting for multiplicity can increase the chance of a nominally positive result occurring by chance. Conversely, a non-significant secondary comparison does not establish that groups are the same, especially when events are uncommon and estimates may be imprecise.

The overall relationship is therefore complementary rather than confirmatory: the primary outcome concerns a continuous cognitive test score, while the secondary outcomes concern adverse events and withdrawals. The available registry information does not establish a pattern across all listed secondary outcomes, because detailed results for some are not reported. These outcomes should be interpreted separately and with attention to their descriptive or exploratory status.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Participants With at Least One Treatment-Emergent Adverse Event (TEAE)
• Number of Participants With at Least One TEAE by Severity
• Number of Participants Who Withdrew From Study Due to TEAEs

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a continuous score, and the registry reports least-squares means for change from baseline. A least-squares mean is a model-based estimate of an average. It can differ from a simple arithmetic mean when the model accounts for repeated observations or other terms. The registry states that the sponsor calculated LS means using a mixed-effects model for repeated measures (MMRM). Further details of the model specification, including covariates and covariance structure, are not reported in the registry record.

What does a repeated-measures model do?

In general, a repeated-measures model analyzes observations collected from the same participants at different times, accounting for the fact that observations from one participant are related rather than independent. A mixed-effects model can represent this within-participant structure while estimating group patterns over time. The registry identifies the method as MMRM, but does not report enough detail to determine precisely how all model terms were specified or how missing measurements were handled.

The endpoint description gives baseline and Day 84 as the timeframe, while the method is identified as a repeated-measures model. The number and timing of all measurements entered into the primary model are not reported in the registry record. No additional visits or model details should be inferred from the method label alone.

How should the reported estimate be read?

The reported difference in LS means is the estimated contrast between the groups for change in score. Its positive direction indicates a higher estimated change for SAGE-718 than for placebo. The estimate is expressed in points on the test scale. The 95% confidence interval gives a range of values compatible with the data and the model under the usual repeated-sampling interpretation. Because that interval includes zero, the analysis does not rule out no difference at the conventional significance threshold. The p-value, 0.164, is also above 0.05; the registry does not state a different threshold.

A p-value is calculated under a null hypothesis of no group difference and the assumptions of the statistical procedure. It describes how unusual data at least as discrepant as those observed would be under that null model; it is not the probability that the null hypothesis is true. A p-value does not say whether an effect is clinically important. The confidence interval and the point estimate should be considered together.

What is the role of the recalculation?

The unadjusted recalculation uses the posted arm-level means, standard errors, and analysis counts to estimate a simple group comparison. It produces a mean difference of 1.5 and p = 0.174. Because it is derived from summary statistics, it is not a reconstruction of the sponsor’s MMRM. In particular, it cannot account for model terms or the within-participant correlation structure unless those are available in the underlying data. The small difference between the recalculated and sponsor-reported p-values is therefore not a reason to replace the sponsor’s analysis.

For binary outcomes, such as whether a participant withdrew due to a treatment-emergent adverse event, the outcome is a count of participants with or without the event. The posted comparison uses a risk difference. In general, a risk difference expresses the difference between the two group proportions, with zero representing equal proportions. The posted recalculation for that endpoint is unadjusted and is not the same analysis as the sponsor’s primary continuous-outcome model. The registry record does not list a sponsor-reported inferential analysis for that secondary endpoint.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variability in count or binary data is greater than a statistical model expects. If a model understates variability, uncertainty intervals may be too narrow and tests may appear more decisive than warranted. This is a general consideration for analyses of event outcomes; the registry record does not report an over-dispersion assessment for these secondary outcomes, so none should be presumed.

Finally, multiple outcomes and comparisons create a multiplicity concern. If many hypotheses are tested, the chance of at least one nominally significant result can increase unless the analysis plan addresses that issue. The registry record does not report a multiplicity adjustment or a hierarchy for secondary endpoints. Consequently, unadjusted secondary comparisons should be treated as exploratory rather than as independent confirmation of the primary result.

11. Confidence Intervals and Precision

The sponsor’s 95% confidence interval for the primary difference in LS means extends from -0.64 to 3.73 points. Its width indicates substantial uncertainty around the point estimate of 1.5 points. The interval includes zero and values on either side of zero, so the data are compatible with no group difference as well as with a positive or negative difference under the reported model.

A confidence interval is not a range that contains a fixed proportion of individual participants’ results. In general, it describes the uncertainty in an estimated population quantity under repeated use of the same method and its assumptions. It does not guarantee that the true effect lies within the particular interval, and it does not account for sources of bias that the model does not address.

The posted standard errors for the group mean changes are 0.77 for placebo and 0.79 for SAGE-718. A standard error describes the precision of an estimated mean, not the spread of individual participant scores. It should not be read as a standard deviation. The sponsor’s confidence interval for the between-group contrast is the more direct expression of uncertainty about the primary comparison.

The sample analyzed for the primary outcome was smaller than the recorded enrollment, and the registry defines the FAS using availability of baseline and post-baseline efficacy data. The reason for any unavailable data and its potential effect on precision are not reported in the registry record. Nor is a clinically meaningful threshold reported, so statistical uncertainty cannot be translated into a judgment about whether the range of possible differences matters clinically.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the trial’s original design calculations. For the continuous primary endpoint and the binary withdrawal endpoint, the figures are participants per arm. They are not total events. The registry record does not report the trial’s original sample-size justification or its original design assumptions.

The listed assumptions for these illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect used for planning is the stated fraction of the observed effect. Separate rows describe planning for 100%, 75%, and 50% of the observed effect. These are hypothetical target effects rather than additional observed findings.

The 80% power column gives the estimated participants per arm needed, under those assumptions, to detect the specified effect with 80% power. The 90% power column gives the corresponding estimate for 90% power. In general, power is the chance that a study will meet its statistical criterion when the specified effect is real and the assumptions hold. Greater power generally requires more participants, and smaller effects need more participants to detect than larger effects.

These sample-size figures are planning illustrations, not observed power. They do not show the probability that the completed study’s finding is true, and they cannot change the uncertainty in the observed estimate. In particular, calculations based on a trial’s observed effect should be treated cautiously because that estimate itself is uncertain. The registry record does not report the original assumptions used to determine enrollment.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 1.500; pooled standard deviation σ = 6.887.
2. n = 2 × (1.96 + 0.842)² × 6.887² / 1.500² = 330.9, rounded up to 331 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 1.5                      # observed difference in means
sigma <- 6.88707522790373                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(5619692)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 331 (80% power), 444 (90% power)
 75% of observed effect: n per arm = 589 (80% power), 788 (90% power)
 50% of observed effect: n per arm = 1,324 (80% power), 1,773 (90% power)
power.t.test: 331.9 
Simulated power at n = 331 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor-reported primary analysis gives a difference in LS means of 1.5, with a 95% confidence interval from -0.64 to 3.73 and p = 0.164. The registry identifies the analysis method as MRMM and describes the primary LS means as calculated using a mixed-effects model for repeated measures. The sponsor’s reported analysis is the main inferential result for the primary endpoint.

The simple recalculation from the posted group summaries also gives a mean difference of 1.5, with p = 0.174. It is an unadjusted comparison based on summary statistics, rather than a reproduction of the sponsor’s model. The numerical agreement in the estimated difference does not make the procedures equivalent; model-based estimates and unadjusted estimates can differ in their assumptions and use of the repeated data.

The recalculated result is useful as a basic consistency check, but it has less information than the sponsor’s analysis and should not supersede it. The discrepancy flag is false, and the reported summaries do not indicate a conflicting estimate of the direction or size of the primary contrast. The sponsor-reported confidence interval and p-value remain the appropriate figures to use when describing the trial’s primary analysis.

The registry record does not report a sponsor-reported inferential analysis for the posted withdrawal endpoint. Its simple comparison should therefore be read as a descriptive, unadjusted result, not as an independently confirmed finding.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study examines several outcomes, time points, or treatment contrasts. In general, the more statistical tests conducted, the greater the possibility that at least one will appear significant by chance if each is judged separately. Common approaches include defining a hierarchy of hypotheses or adjusting the significance threshold, but the registry record does not report whether either approach was used for this study. It also does not report a multiplicity adjustment for the secondary endpoints.

The primary outcome is identified, and the sponsor reports a primary analysis. The statistical plan beyond the stated use of a mixed-effects model for repeated measures is not reported in the registry record. Details such as the prespecified estimand, covariate adjustment, model covariance structure, missing-data sensitivity analyses, or interim analyses are not reported in the registry record. No assumptions about these procedures should be inferred.

The registry defines the FAS as participants in the Safety Set who received at least one dose and had baseline and at least one post-baseline efficacy evaluation. It also says that the overall number analyzed indicates participants with data available for analysis. This definition makes the analysis population dependent on having efficacy measurements, but the registry record does not explain why data might be unavailable or how missing measurements were handled by the model.

Missing data can affect both the amount of information and the validity of an estimate. In general, a repeated-measures model relies on assumptions about the relationship between observed and missing measurements; different assumptions can lead to different conclusions. The registry record does not report sensitivity analyses that would show how robust the primary result is to alternative missing-data assumptions.

The safety outcomes use safety-group denominators, whereas the primary efficacy analysis uses the FAS. Those populations serve different analytic purposes. The registry record does not report a complete statistical analysis plan, the original sample-size justification, or the details of any interim monitoring. The absence of those details limits assessment of the full analysis strategy.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: fall (1.2%), hip fracture (1.2%), upper limb fracture (1.2%), atrial fibrillation (1.2%), cellulitis (1.2%).

The registry reports serious adverse event counts by safety group. In the placebo group, 5 participants were affected among 86 at risk; in the SAGE-718 group, 4 were affected among 84 at risk. These are counts of participants with at least one serious adverse event, not counts of all serious-event episodes.

The registry also records 1 death in the placebo group and 0 in the SAGE-718 group. The listed serious-event terms include Fall, Hip fracture, Upper limb fracture, Atrial fibrillation, Cellulitis, and Carotid artery stenosis, each reported as 1.2%. The registry does not specify an arm for each listed term in the information summarized in the record, so those terms should not be assigned to a group.

These figures describe events recorded during the trial; they do not by themselves establish that an event was caused by treatment. The registry also lists treatment-emergent adverse events and withdrawals as secondary safety outcomes, but detailed results for all of those outcomes are not reported in the registry record. Group sizes and exposure duration should be considered when interpreting event counts, and a numerical comparison of counts alone is not a causal assessment.

17. What the Result Does, and Does Not, Mean

The primary result is an estimate of the difference between groups in change on a processing-speed test. The sponsor-reported estimate is positive, but its confidence interval includes zero and the reported p-value is 0.164. The appropriate conclusion is that the primary analysis did not establish a clear difference between groups, while retaining uncertainty about the size and direction of the underlying difference.

This is not proof that the treatments have identical effects. A result that does not meet a conventional significance threshold can reflect an effect near zero, an effect that the study estimated imprecisely, or both. The confidence interval conveys a range of values compatible with the data and model; it does not establish equivalence or rule out clinically meaningful effects. The registry record does not report an equivalence margin or a threshold for clinical importance.

The endpoint is a standardized test score related to processing speed, not a direct measure of every aspect of cognition or daily functioning. Although the registry describes an association between the test and functional outcomes, the reported analysis does not itself quantify changes in everyday activities. The study result should therefore be interpreted in terms of the specified WAIS-IV Coding score and its measurement timeframe.

The findings also have limits in scope. They concern the study population and the specified dosing schedule through Day 84. The registry record does not provide enough information to assess how the result might apply to people outside the enrolled population, to different treatment schedules, or to outcomes beyond those reported. No treatment recommendation follows from these statistical results alone.

18. Why This Trial Matters Statistically

This trial illustrates why an estimated effect, its uncertainty, and the analysis method should be read together. The point estimate is positive, but the confidence interval spans values on both sides of no difference. Looking only at the point estimate would overstate certainty; looking only at whether the p-value passes a threshold would omit the range of effects compatible with the data.

It also demonstrates the difference between a sponsor-reported model analysis and a recalculation from summary statistics. Both give a mean difference of 1.5, while their p-values are 0.164 and 0.174. The recalculation is a useful check, but the registry identifies a repeated-measures model for the primary endpoint, and the underlying participant-level data and full model details are not available in the record. The model-based analysis is therefore the primary inferential result.

The trial also connects efficacy and safety questions without treating them as interchangeable. A cognitive test score is a continuous outcome, whereas adverse events and withdrawals are event counts. The methods and interpretation differ, and sparse event data can be particularly uncertain. Secondary outcomes can add context, but they do not automatically provide confirmation of the primary finding, especially when multiplicity procedures are not reported.

Finally, the planning illustrations show the dependence of sample-size requirements on the target effect and desired power. They are not evidence that the completed study had a particular probability of success. Separating planning assumptions from observed results helps keep design calculations, statistical inference, and clinical interpretation distinct.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record does not report the original sample-size justification or the assumptions used to plan the study. The planning illustrations based on observed effects cannot reconstruct that original rationale. It also does not provide the complete statistical analysis plan, including the model’s covariates, covariance structure, or any prespecified sensitivity analyses.

The primary analysis population requires baseline and at least one post-baseline efficacy evaluation. The registry does not report the reasons for unavailable measurements or enough detail to assess the assumptions used for missing data. This limits assessment of whether missingness could affect the estimated treatment contrast.

The primary endpoint is a test score, and the registry does not state a threshold for a clinically meaningful score difference. Although the test is described as related to functional outcomes, the reported result does not quantify daily functioning. The confidence interval is compatible with effects in either direction, which limits precision about the true contrast.

There are also limits to interpretation of secondary outcomes. Detailed registry results for two listed treatment-emergent adverse-event outcomes are not reported in the registry record, and the record does not describe multiplicity control. Serious-event counts and the listed terms are descriptive information; the information summarized does not assign each term to a treatment group or establish causation.

Finally, the primary timeframe is baseline to Day 84, and the active-treatment dose changed after the first 6 weeks. The registry information does not support conclusions about longer-term outcomes or treatment schedules other than the one described.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Evaluate the Effects of SAGE-718 in Participants With Mild Cognitive Impairment or Mild Dementia Due to Alzheimer's Disease (AD) — NCT05619692.
• Registry record last updated 2025-09-15; status last verified 2025-06.
• Results first posted 2025-06-18.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT05619692 is a completed randomized, quadruple-masked, parallel-group, placebo-controlled Phase 2 study of SAGE-718 in participants with mild cognitive impairment or mild dementia due to Alzheimer’s disease. The recorded enrollment was 174. SAGE-718 was administered orally once daily, with the dose changing from 1.2 mg during the first 6 weeks to 0.9 mg through Day 84; the comparator was matching placebo.

The primary outcome was change from baseline in WAIS-IV Coding score. The sponsor-reported difference in LS means was 1.5, with a 95% confidence interval from -0.64 to 3.73 and p = 0.164. The interval includes zero, so the primary analysis did not establish a clear between-group difference. The registry record does not report the original sample-size justification, full statistical analysis plan, or detailed results for all listed secondary outcomes.