A Study to Evaluate SAGE-217 in Participants With Parkinson's Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate SAGE-217 in Participants With Parkinson's Disease (NCT03000569), a non-randomized phase 2 study. The primary outcome is Percentage of Participants With Treatment-emergent Adverse Events (TEAEs) - Part A.

2. Clinical Question

3. Trial Design

The registry record describes NCT03000569 as a completed Phase 2 study with sequential, non-randomized allocation and no masking. Overall enrollment was 29. The study had two parts and assessed safety, tolerability, pharmacokinetics, and efficacy in participants with Parkinson's disease. The registry lists SAGE-217, levodopa, and antiparkinsonian agent(s) among the interventions.

In Part A, participants taking a stable morning dose of levodopa, including carbidopa-levodopa, as antiparkinsonian agent(s) received those agents from Days 1 to 3. The record then describes stopping levodopa and administering SAGE-217 as a 30 mg per day oral solution on Days 4 to 7, in the morning with food. The description of stable doses of other antiparkinsonian agents is truncated in the registry record, so additional schedule details are not reported in the registry record.

In Part B, participants on a stable dose of antiparkinsonian agent(s) received SAGE-217 capsules, up to 30 mg per day, on Days 1 to 7 in the evening with food. The record also reports follow-up periods for both parts. The posted adverse-event results identify 15 participants at risk during the Part A antiparkinsonian-agent period, 14 during its SAGE-217 period, and 15 during its follow-up. Part B reports 14 at risk during the SAGE-217 period and 14 during follow-up.

The study's non-randomized, unmasked structure does not provide a randomized control contrast. The sequence of periods is part of the design, but the registry record does not report a statistical plan for separating treatment effects from period, sequence, or other influences. The start date was 2016-11-30, and primary completion was 2017-09-11.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The safety population included all participants who were administered study drug (SAGE-217 or levodopa). Overall number of participants analyzed = Number of participants in the safety population that remained in the study during each period.

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 population included all participants who were administered study drug \[SAGE-217 or levodopa (Part A); or SAGE-217 or antiparkinsonian agent(s) (Part B)\].

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Treatment-emergent Adverse Events (TEAEs) - Part A, measured over Day 1 to Day 14.

The primary endpoint is the percentage of participants with treatment-emergent adverse events in Part A, over the timeframe Day 1 to Day 14. The registry defines an adverse event as an untoward medical occurrence associated with use of a drug in humans, regardless of whether it is considered drug-related. A TEAE is an adverse event occurring after the first administration of study drug. This timing-based definition captures events after exposure without asserting that study drug caused them.

Safety endpoints of this kind are used to describe the frequency of reported events during treatment and to identify events that warrant closer examination. They are useful because the definition does not restrict counting to events judged to be related to the drug. At the same time, the percentage is a descriptive measure: it does not establish causality, severity, or comparative risk on its own. Those questions require additional information, including event-level details and a suitable comparison design.

The registry states that the safety population included all participants administered study drug, SAGE-217 or levodopa. The number analyzed for a period was the number in the safety population who remained in the study during that period. Accordingly, period-specific denominators can differ. The endpoint is limited to Part A; it should not be read as an overall estimate for every participant or as the primary safety result for Part B.

The registry lists secondary outcomes involving MDS-UPDRS scores, TEAEs in Part B, and basophils in Part B. Numerical findings for those outcomes are not reported in the registry record presented for this analysis.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a binary safety outcome: whether a participant experienced a treatment-emergent adverse event during the stated period. In general, a binary outcome records whether an event occurred or did not occur for each participant. A percentage summarizes the number with the event relative to the number analyzed. Because participants may enter or remain in different study periods, the denominator must be interpreted for each period rather than assumed to be constant.

What does a confidence interval add?

In general, a confidence interval expresses the statistical uncertainty around an estimate. For a percentage, its width depends on factors such as the number of participants observed and how often the event occurs. A wider interval indicates less precision, while a narrower interval indicates greater precision under the method used. The registry record does not report confidence intervals for the primary endpoint, so its precision cannot be assessed from an interval estimate.

The posted adverse-event information gives counts at risk, counts affected by serious events, counts affected by other events, and deaths for the listed periods. The primary endpoint is defined as a percentage of participants with TEAEs in Part A from Day 1 to Day 14. These descriptions should not be conflated: counts for serious and other adverse events are not a substitute for a reported statistical analysis of the primary endpoint. The registry record lists zero analyses, and does not report a model, an estimate with uncertainty, or a hypothesis test for the primary endpoint.

Why does the design affect statistical interpretation?

In general, random allocation helps make treatment groups comparable with respect to both measured and unmeasured factors. Here, allocation is explicitly non-randomized, and masking is absent. The study is sequential, with the Part A antiparkinsonian-agent period followed by the SAGE-217 period. Consequently, changes between those periods may coincide with changes in treatment, time, or the set of participants observed. A period-to-period difference would not, by itself, isolate the effect of SAGE-217. Part B also does not establish an untreated control: participants received SAGE-217 alongside stable antiparkinsonian agent(s).

General statistical methods for repeated or sequential observations can account for some dependence between observations from the same participant, but the registry record does not report whether such methods were used. It also does not report covariate adjustment, a planned contrast, or a detailed analysis set for each efficacy outcome. The posted data should therefore be interpreted as descriptive rather than as evidence of an adjusted treatment comparison.

What do the planning assumptions mean?

The record describes sample-size illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It also describes detecting a chosen share of the observed effect. These are planning assumptions for illustrations, not a description of the actual non-randomized sequential design. Any resulting participant counts are planning quantities, not observed trial outcomes and not observed power. The registry record does not report the trial's own sample-size justification.

In general, power is a design property associated with a specified effect, variability, significance level, and analysis. It is not a measure of whether a completed study has demonstrated an effect. Retrospective calculations based on an observed estimate do not repair the absence of a reported analysis or a randomized comparison. The record's planning description should not be used to infer that the study was designed to test a particular treatment contrast.

Why does the event definition matter?

The TEAE definition is based on timing after the first study-drug administration, not on a judgment that the drug caused the event. This operational rule makes the endpoint suitable for systematic safety description, but it includes events that may have other explanations. The count of participants with an event also differs from the count of event episodes: the registry endpoint is framed as participants affected, and the posted summaries do not provide a complete account of recurrent events or event-level timing.

In general, sparse event data require cautious interpretation. A count of zero serious events means none were reported in the listed period; it does not establish that the underlying risk is zero. Small observed groups can also yield imprecise estimates, and the record does not provide interval estimates to express that uncertainty. These are general considerations; no additional numerical estimate is warranted from the registry record.

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("Part A: Antiparkinsonian Agent(s) Day 1 to Day 3", "Part A: SAGE-217 Day 4 to Day 7", "Part A: Follow-up")
n      <- c(15, 14, 15)
events <- c(4, 14, 2)
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  Part A: Antiparkinsonian Agent(s) Day 1 to Day 3  15  4  26.7  7.8  55.1  10.9  52.0  11.0  52.4
ARM  Part A: SAGE-217 Day 4 to Day 7  14  14  100.0  76.8  100.0  78.5  100.0  78.2  99.8
ARM  Part A: Follow-up  15  2  13.3  1.7  40.5  3.7  37.9  4.0  38.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 119 outcome measures and a set of secondary outcomes, including MDS-UPDRS measures, Part B TEAE measures, and basophils. It does not report a multiplicity strategy, such as a hierarchy of tests or an adjustment for making multiple statistical comparisons. Without a stated plan and reported analyses, it is not possible to determine how the study intended to control the chance of false-positive findings across outcomes. The registry record lists no analyses, so no inferential results for those outcomes are available for interpretation.

Missing-data methods are not reported in the registry record. The primary outcome description says that the number analyzed for a period was the number in the safety population who remained in the study during that period. The period-specific at-risk counts differ, which makes it important to read each count with its own denominator. The record does not describe reasons for participants not remaining in a period, how missing event information was handled, or whether any sensitivity analysis was planned.

The safety population included participants administered study drug, SAGE-217 or levodopa. This is an exposure-based population definition, not necessarily the analysis population for every secondary outcome. The registry record does not report a separate analysis population or missing-data approach for each outcome. It also does not state whether efficacy analyses followed the same population definition.

These omissions limit what can be concluded from the posted summaries. They do not show that a particular missing-data or multiplicity method was used; that information is not reported in the registry record. General statistical practice is to specify outcome priorities, analysis populations, and missing-data handling in advance, especially when many outcomes are collected, but the study's own procedures cannot be inferred from that general principle.

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 non-serious adverse event terms

The serious adverse-event counts are 0 for each listed period: Part A antiparkinsonian agent(s), Part A SAGE-217, Part A follow-up, Part B SAGE-217, and Part B follow-up. The posted data also report 0 deaths in each period. These are observed counts in the participants at risk during those periods, not estimates that the underlying chance of a serious event or death is zero.

The counts for other adverse events are reported separately from serious events. For Part A, the posted counts are 4 of 15 during the antiparkinsonian-agent period, 14 of 14 during the SAGE-217 period, and 2 of 15 during follow-up. For Part B, the count is 8 of 14 during SAGE-217 treatment and 0 of 14 during follow-up. These are participant counts, not event totals, and the sequential, non-randomized design does not support treating the periods as equivalent comparison groups.

The record does not provide serious-event types or further detail on severity for these periods. The absence of reported serious events in these observed groups is descriptive and does not rule out uncommon events or establish comparative safety.

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

The posted results describe adverse-event counts among participants observed during specified treatment and follow-up periods. They show that other adverse events were recorded in each of the active treatment periods and in some follow-up periods, while the serious adverse-event count and death count were 0 in every listed period. These findings characterize what was recorded in this study; they are not a causal estimate of SAGE-217's effects.

The design matters to interpretation. Participants were not randomized, the study was unmasked, and the periods were sequential. Part A's antiparkinsonian-agent period and subsequent SAGE-217 period differ in treatment and in the number at risk. Part B involved SAGE-217 alongside stable antiparkinsonian agent(s), rather than a concurrent untreated comparison group. The counts cannot therefore be read as if they came from exchangeable treatment groups.

The primary endpoint uses a timing-based TEAE definition. An event could count after study-drug administration even if it was not considered drug-related. Aggregate counts do not show which events occurred, their severity, how long they lasted, whether participants experienced repeated events, or how investigators judged causality. The registry record does not report statistical analyses or uncertainty intervals for the endpoint, so the precision of an estimated percentage or any treatment contrast is not available.

The record also lists efficacy and other secondary outcomes, but numerical results for those outcomes are not reported in the registry record. The adverse-event summaries cannot answer efficacy questions. More generally, a completed study record is not equivalent to a definitive comparative analysis: the design, outcome definitions, follow-up, and completeness of reporting all constrain what can be learned from the observed data.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of why study design and endpoint definition matter when reading clinical safety data. The registry labels the study Phase 2, sequential, non-randomized, and unmasked. Its Part A sequence makes exposure and period coincide: participants move from an antiparkinsonian-agent period to a SAGE-217 period. A change in recorded events across these periods cannot be attributed to treatment alone without assumptions or analyses that address other differences over time.

The endpoint also illustrates a distinction central to safety interpretation. A TEAE is defined by occurring after the first administration of study drug, not by a finding that the drug caused it. This inclusive definition supports systematic event collection, while also requiring readers to avoid interpreting every counted event as drug-related. Serious events, other events, and deaths are presented as distinct categories, and a zero count in an observed group is not proof of zero underlying risk.

Period-specific denominators are important because the number at risk changes across periods. The safety population is exposure-based, and the record defines each period's analyzed number according to who remained in the study. That makes clear reporting of analysis populations and missing-data handling especially relevant. The record does not report those methods in detail, nor does it list statistical analyses or multiplicity procedures despite reporting multiple outcome measures.

The study therefore offers descriptive safety observations, but the registry record does not support a precise causal comparison or a conclusion about efficacy. Recognizing that boundary is a core biostatistical skill: the design and reporting determine what questions the observed counts can answer.

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 most important limitation for interpretation is the non-randomized, unmasked sequential design. The registry explicitly cautions that the arms are not a like-for-like control comparison. In Part A, the antiparkinsonian-agent period precedes the SAGE-217 period, so treatment is intertwined with time and sequence. Part B combines SAGE-217 with stable antiparkinsonian agent(s), and does not provide a randomized comparator. The registry record does not report analyses designed to address these limitations.

The total enrollment was 29, and the posted period-specific at-risk counts are 15 or 14. These limited observed groups constrain how informative event counts can be, particularly for uncommon outcomes. No confidence intervals or other uncertainty estimates are reported for the primary endpoint. Zero serious adverse events in a listed period should be read as an observed count, not as evidence that no such risk exists.

The registry describes the primary endpoint definition and safety population, but does not report detailed event types, event-level timing, or a full account of severity in the listed results. The description of stable doses of other antiparkinsonian agents in Part A is truncated. Missing-data procedures, multiplicity control, and the trial's own sample-size justification are not reported in the registry record.

The record lists many outcome measures, but no statistical analyses. Numerical results for the listed secondary efficacy outcomes are not reported in the registry record. As a result, the posted safety counts cannot answer efficacy questions, and there is no reported inferential framework with which to assess the precision or comparative meaning of the primary endpoint.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate SAGE-217 in Participants With Parkinson's Disease — NCT03000569.
• Registry record last updated 2023-11-29; status last verified 2023-11.
• Results first posted 2020-11-23.
• 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

NCT03000569 is a completed Phase 2, two-part study of SAGE-217 in participants with Parkinson's disease. The design was sequential, non-randomized, and unmasked; overall enrollment was 29. Part A involved an antiparkinsonian-agent period followed by SAGE-217 and follow-up. Part B involved SAGE-217 alongside stable antiparkinsonian agent(s), followed by follow-up.

The primary endpoint was the percentage of Part A participants with TEAEs from Day 1 to Day 14. The posted adverse-event summaries report period-specific counts, including 0 serious adverse events and 0 deaths in each listed period. The registry record lists 119 outcome measures but no statistical analyses. It does not report inferential estimates, missing-data methods, multiplicity procedures, or the trial's own sample-size justification. These results are descriptive and do not constitute a randomized treatment comparison.