A Study to Evaluate the Safety and Tolerability of SAGE-718 in Participants With Huntington's Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Safety and Tolerability of SAGE-718 in Participants With Huntington's Disease (NCT05655520), a non-randomized phase 3 study. The primary outcome is Number of Participants With at Least One Treatment-Emergent Adverse Events (TEAEs).

2. Clinical Question

The study’s stated purpose was to evaluate the safety and tolerability of SAGE-718 in participants with Huntington’s disease. This is a safety-focused question: the primary outcome concerns whether a participant had at least one qualifying event, not the number of events each participant experienced or an estimate of treatment benefit. The registry record lists 153 participants as the enrollment, but does not report the primary outcome result. It also does not report a randomized comparison that could isolate a treatment effect from other differences between cohorts. The posted information therefore describes the study’s purpose and design, but does not establish how frequently participants experienced the primary outcome across the full study population or whether outcomes differed between cohorts.

3. Trial Design

This was a phase 3, multicenter, open-label safety study with a parallel model and non-randomized allocation. The intervention listed was SAGE-718. The registry record lists Huntington’s disease as the condition, an enrollment of 153, and a study status of terminated. The study start date was 2022-12-14, and the primary completion date was 2025-01-20. The reason for termination is not reported in the registry record.

The three cohorts were formed according to prior participation and timing of entry. Cohort 1, Direct Rollover, included participants from studies 718-CIH-201 and 718-CIH-202 who signed consent within 7 days after the corresponding parent study ended. Cohort 2, Gap Rollover, included participants from those studies who signed consent after a gap exceeding 7 days. The descriptions for both rollover cohorts specify oral SAGE-718 at 0.9 mg. Cohort 3, De Novo, included participants who had not previously been included in a SAGE-718 clinical study; its description states that treatment began on Day 1 onwards, without specifying a dose there.

Because allocation was not randomized and there was no masking, treatment assignment and knowledge of treatment were not controlled as they would be in a blinded randomized trial. The cohorts also differ by prior study participation and enrollment history. The registry record explicitly cautions that this is not a randomized comparison and that the arms are not a like-for-like control comparison. Cohort outcomes can be described separately, but differences between them cannot be interpreted as unbiased evidence of a treatment effect.

The primary completion date indicates when the primary outcome period was recorded as complete; the registry status is nevertheless listed as terminated. The registry record does not explain the relationship between these entries or provide a reason for termination. It also does not report detailed recruitment flow, cohort-specific follow-up duration, or whether every enrolled participant contributed to each outcome. Such details matter when assessing whether safety observations were captured over comparable periods.

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: A total of 165 participants were screened, of which, 153 participants were enrolled and received study treatment in either Cohort 1 (Direct rollover) or Cohort 2 (Gap rollover). The study was terminated early by the Sponsor and hence, no participants were enrolled in Cohort 3 (De novo) and therefore no data was collected and reported. No additional milestone.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Safety Set included all participants who were administered IP during the study.

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: The Safety Set included all participants who were administered investigational product (IP) during the study.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With at Least One Treatment-Emergent Adverse Events (TEAEs), measured over Up to 24 months.

The primary outcome is the number of participants with at least one treatment-emergent adverse event, with a timeframe of up to 24 months. It is a participant-level count: a participant qualifies if at least one event meets the registry’s definition. The endpoint does not, by itself, describe the total number of events, their severity, duration, or clinical consequences.

The registry defines an adverse event as an untoward medical occurrence in a participant administered a pharmaceutical product, whether or not it is considered causally related to that product. A treatment-emergent adverse event is one with onset after investigational product starts, or a worsening of a pre-existing medical condition or adverse event after treatment starts and throughout the study. This timing-based definition is not the same as a determination that treatment caused the event.

A participant-level endpoint is useful in safety monitoring because it captures whether an individual experienced at least one qualifying event. It can provide a broad description of event occurrence, but it treats participants with one event and participants with multiple events alike for this particular count. Other outcomes are needed to describe specific event types, seriousness, recurrence, or severity. The registry lists 27 outcome measures, but the primary outcome result itself is not reported in the registry record. Its analysis method, estimate, and precision are therefore not available for interpretation.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a non-randomized, open-label, parallel study and lists no outcome analyses. This combination sets important limits on statistical interpretation. The primary outcome is a count of participants who experienced at least one event, rather than a continuous measurement or a time-to-event estimate. The registry record does not report the method used to analyze that outcome, an estimate, a confidence interval, or a hypothesis test.

What does a participant count represent?

For a binary participant-level outcome, each person is classified according to whether the event occurred at least once during the defined observation period. A count reports how many participants met that definition in the relevant analysis group. To interpret such a count as a frequency, the number of participants assessed for the outcome and the completeness of event follow-up also matter. The primary outcome population is defined as the Safety Set, meaning participants administered investigational product, but the primary outcome count and its analysis denominator are not reported in the registry record.

This endpoint differs from an event count. If one participant has several qualifying events, that participant still contributes once to the primary participant-level count. Separate event summaries can convey which events occurred and how often, but they answer a different question. Likewise, a count of participants with serious adverse events is not interchangeable with a count of participants with any treatment-emergent adverse event. Definitions and groupings must be kept distinct.

Why does randomization matter?

Random allocation is intended to balance measured and unmeasured characteristics across treatment groups on average. Here, allocation is listed as non-randomized. The cohorts were formed according to prior participation and timing of enrollment, so they are not equivalent by design. A difference in observed counts could reflect the cohort histories, differences in time observed, other participant characteristics, or chance. Without a randomized control group and a suitable analysis, a direct causal interpretation is not supported. The registry itself warns that the arms are not a like-for-like control comparison.

Open-label status means treatment masking was absent. In safety studies, reporting and classification of events can involve judgment; awareness of treatment may affect how events are noticed, described, or attributed. The registry’s adverse-event definition explicitly does not require a causal relationship to treatment. Event timing therefore identifies treatment-emergent events but does not show that the investigational product caused them.

What does a confidence interval add?

In general, a confidence interval describes the statistical precision of an estimate under the assumptions of the analysis method. A narrow interval usually indicates greater precision than a wide one, while an interval can still be compatible with a range of possible underlying values. An interval does not repair design limitations such as non-random allocation, nor does it establish causation. No confidence interval for the primary outcome is reported in the registry record, so its statistical precision cannot be assessed.

In general, over-dispersion matters when event counts vary more than a simple statistical model expects. A model that assumes too little variability can produce uncertainty estimates that are too small. The primary outcome is described as a count of participants with an event, and the registry record does not report a model or an over-dispersion assessment. It would not be appropriate to infer that a particular count model or variance adjustment was used.

How should planning illustrations be read?

The stated planning assumptions describe illustrations based on a stated fraction of an observed effect, a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. For most outcomes, the illustration is expressed as participants per arm for 80% or 90% power; for time-to-event outcomes, it is expressed as a total number of events. Such calculations are planning illustrations, not observed power and not results from the trial. They depend on assumptions about the effect and analysis. The registry record does not report an endpoint-specific illustration value in the registry record for this study, nor does it report that these assumptions were the trial’s own design basis.

Equal allocation between two arms is an assumption for a planning illustration, not a description of this study’s enrollment structure. This study has three non-randomized cohorts, and the registry cautions against treating them as control comparisons. A planning calculation cannot make the cohort structure randomized or remove its potential for confounding.

What can be inferred from the listed analyses?

The registry lists zero analyses. This means no outcome analysis records are listed there; it does not establish what analyses may have been conducted outside the registry. In the absence of registered estimates and methods, readers cannot assess adjustment, confidence intervals, statistical tests, or model assumptions for the primary outcome. The general statistical principles above describe how such analyses are ordinarily interpreted, not procedures confirmed for this study.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort 1 (Direct Rollover)", "Cohort 2 (Gap Rollover)")
n      <- c(72, 81)
events <- c(46, 46)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort 1 (Direct Rollover)  72  46  63.9  51.7  74.9  52.4  74.0  52.3  74.0
ARM  Cohort 2 (Gap Rollover)  81  46  56.8  45.3  67.8  45.9  67.0  45.9  67.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 27 outcome measures and zero analyses. It does not report a multiplicity strategy, such as whether any outcomes were designated for confirmatory testing, whether tests were adjusted for multiple comparisons, or how outcomes were prioritized beyond identifying the primary outcome. When many outcomes are assessed, the chance of obtaining an apparently notable result by chance can increase if the analyses are treated as independent confirmatory tests. A prespecified strategy helps readers distinguish primary inference from exploratory description. No such strategy is reported in the registry record.

The primary outcome population is the Safety Set, defined as all participants administered investigational product during the study. This definition is relevant to safety assessment, but the number in the Safety Set is not separately reported alongside a primary outcome result. The registry enrollment of 153 is not necessarily the analysis denominator, and the registry record does not give a participant flow accounting for enrollment, treatment administration, follow-up, and inclusion in each outcome.

Missing-data handling is not reported in the registry record. In a safety outcome with a timeframe of up to 24 months, incomplete follow-up could affect whether an event is observed. The record does not describe how participants with partial observation, discontinuation, or missing event information were handled, whether follow-up time was comparable across cohorts, or whether any sensitivity analyses were conducted. No assumptions about these matters should be inferred from the listed counts.

Because no analyses are listed, there are no reported adjusted estimates or inferential results to assess. The registry record also does not report a statistical analysis plan, sample-size justification, or interim analysis approach. The absence of those details in the registry does not establish whether corresponding documents or analyses exist elsewhere; it means their methods cannot be evaluated from this record.

12. 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: psychotic disorder (1.4%), suicidal ideation (1.4%), immune thrombocytopenia (1.4%), schizoaffective disorder (1.2%), suicidal behaviour (1.2%).

The registry lists serious adverse-event counts for two cohorts. In Cohort 1, 3 participants were affected among 72 listed at risk; in Cohort 2, 6 were affected among 81 listed at risk. It lists zero deaths for each of those cohorts. The other-adverse-event entries are 46 affected among 72 at risk in Cohort 1 and 22 affected among 81 at risk in Cohort 2. These figures should be read as separate descriptive categories, not combined into a single count.

The record’s serious-event term list gives psychotic disorder, suicidal ideation, and immune thrombocytopenia at 1.4% each; schizoaffective disorder, suicidal behaviour, and suicide attempt are listed at 1.2% each. The denominator used for these term percentages is not reported in the registry record. The terms and percentages do not by themselves specify the number of events, whether a participant had more than one event, or the duration and outcome of each event.

The group figures are not comparative evidence of treatment effects. Cohorts were not randomized, and the record does not list corresponding safety group figures for Cohort 3. The zero death entries apply to the two cohorts for which those figures are reported; they should not be extended beyond those entries. The registry’s definition of adverse events also does not imply that every event was caused by SAGE-718.

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

The study was designed to describe safety and tolerability during SAGE-718 exposure in participants with Huntington’s disease. The registry identifies a primary treatment-emergent adverse-event outcome and gives a definition for it, along with a Safety Set population and a timeframe of up to 24 months. Those details clarify the question the study was intended to address.

The record does not report the primary outcome result. It therefore cannot show from the primary endpoint how many participants experienced at least one treatment-emergent adverse event, nor how precisely that outcome was estimated. The listed serious and other adverse-event counts offer additional descriptive information for Cohorts 1 and 2, but do not fill in the missing primary outcome result. Their denominators and cohort context should be retained when reading them.

The listed counts do not establish whether SAGE-718 caused the reported events. The adverse-event definition allows events to be recorded without a causal relationship, and treatment-emergent timing alone is not a causal test. The study was non-randomized and open-label, with cohorts defined by prior study history and timing of entry. Thus, differences among cohort counts cannot be attributed to treatment as though they came from a randomized, like-for-like comparison.

The registry record does not report efficacy results, so the safety information should not be used to make claims about treatment benefit. It also does not report the reason the study was terminated, missing-data procedures, or a statistical analysis plan. These absences limit interpretation but do not establish that a particular event was more or less likely, that a treatment effect exists, or that an unreported analysis was or was not performed. The most defensible reading is descriptive: the registry states the study’s safety objective and lists certain safety counts, while leaving the primary outcome result and key analysis details unreported.

14. Why This Trial Matters Statistically

Safety studies ask a different statistical question from trials designed primarily to compare efficacy. A count of participants with at least one treatment-emergent adverse event can describe how many participants met a broad event definition, while serious-event summaries provide information about more specific safety categories. Keeping these outcomes separate prevents a broad participant-level measure from being mistaken for a complete account of event type, seriousness, or recurrence.

This record also illustrates why study design must accompany event counts. The three cohorts reflect different enrollment histories, and treatment allocation was non-randomized. The registry explicitly warns that the cohorts are not like-for-like control groups. A simple contrast between their counts would not account for differences in prior study participation, entry timing, follow-up, or other factors. Statistical methods can describe uncertainty, but cannot by themselves recreate the balance expected from randomization.

Completeness of registry reporting is another statistical concern. The record specifies the primary outcome and Safety Set definition but does not report the primary outcome result, an analysis method, or a confidence interval. It lists 27 outcomes and zero analyses. Without estimates and details on missing data and multiplicity, readers cannot assess the precision or inferential status of results across outcomes.

These features make the record useful for understanding the intended safety question and some reported cohort-level safety observations, while also showing the limits of what can be concluded. Careful interpretation distinguishes a registry-listed count from a comparative treatment effect, and distinguishes event occurrence after treatment begins from evidence that treatment caused the event.

15. 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 primary outcome result is not reported in the registry record, so the main safety endpoint cannot be evaluated quantitatively. The registry lists serious and other adverse-event counts for Cohorts 1 and 2, but does not list equivalent safety group figures for Cohort 3. The listed cohort counts therefore do not represent a complete outcome account for all three cohorts.

The non-randomized, open-label design limits causal interpretation. Cohort membership is linked to prior SAGE-718 study participation and timing of enrollment, so the cohorts may differ in ways that affect observed safety counts. The registry itself warns against treating them as like-for-like controls. It does not report adjusted comparisons that would address differences in participant characteristics or observation.

The primary outcome timeframe is stated as up to 24 months, but cohort-specific follow-up and the handling of incomplete observation are not reported. The missing-data approach, analysis denominators for the primary outcome, and participant flow are also not reported in the registry record. The serious-event term percentages are listed without their denominator or calculation details.

The registry lists 27 outcome measures but no analyses. It does not report the statistical analysis plan, multiplicity approach, sample-size justification, interim analyses, or reason for termination. These omissions prevent assessment of how the primary result was to be analyzed and how uncertainty across outcomes was handled. They should be treated as limits of the registry information, not as evidence that a particular analysis or procedure did not occur outside the registry.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate the Safety and Tolerability of SAGE-718 in Participants With Huntington's Disease — NCT05655520.
• Registry record last updated 2025-12-17; status last verified 2025-08.
• Results first posted 2025-12-17.
• 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.

19. Record Summary

This phase 3 study evaluated the safety and tolerability of SAGE-718 in participants with Huntington’s disease. It was multicenter, open-label, non-randomized, and parallel, with direct-rollover, gap-rollover, and de novo cohorts. The registry lists enrollment of 153, status as terminated, a start date of 2022-12-14, and a primary completion date of 2025-01-20.

The primary outcome was the number of participants with at least one treatment-emergent adverse event over a timeframe of up to 24 months. Its population was the Safety Set, defined as participants administered investigational product. The primary outcome result and its analysis are not reported in the registry record. Serious and other adverse-event counts are listed for the direct-rollover and gap-rollover cohorts, with zero deaths recorded for each; these are descriptive entries, not a randomized comparison. The record lists 27 outcomes and zero analyses. It does not report the reason for termination or key statistical details such as missing-data handling and multiplicity methods.