A Study of TAK-994 in Adults With Type 1 and Type 2 Narcolepsy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of TAK-994 in Adults With Type 1 and Type 2 Narcolepsy (NCT04096560), a randomized phase 2 study. The primary outcome is Part A: Number of Participants Who Experience at Least One Treatment Emergent Adverse Event (TEAE) During the Study.

2. Clinical Question

The trial asked whether oral TAK-994 could be studied across doses in adults with narcolepsy, with or without cataplexy, while assessing safety, tolerability, drug concentrations, and symptoms. The registry record describes a Phase 2 randomized, double-blind, placebo-controlled study and includes participants with narcolepsy type 1 and type 2.

The registry record also describes symptom outcomes in Parts B and C: change in Epworth Sleepiness Scale total score and weekly cataplexy rate at Week 8. These are secondary outcomes, so they address additional questions rather than replacing the primary safety question. Numerical results for Part D are not reported in the registry record.

3. Trial Design

This was a Phase 2, randomized, quadruple-masked, parallel-group, placebo-controlled study of oral TAK-994. The registry lists enrollment of 97. Its brief description divides the study into four parts, and participants could join only one part. The posted outcome data are organized by part, dose, and placebo group; the registry record does not provide a single combined treatment comparison across all parts.

Part A enrolled participants with type 1 narcolepsy and compared placebo with TAK-994 120 mg or 180 mg for 28 days. The primary endpoint was the number of participants with at least one treatment-emergent adverse event during the study, from first dose through follow-up up to Day 35. Part B, also in type 1 narcolepsy, compared placebo with TAK-994 30 mg, 90 mg, or 180 mg over 56 days. Part C was described as a study in Chinese participants with type 1 narcolepsy receiving TAK-994 or placebo over 56 days; the posted arm descriptions list a TAK-994 180 mg arm. Part D was described for type 2 narcolepsy, with TAK-994 or placebo over 28 days.

The secondary outcomes include pharmacokinetic measures in Part A, such as maximum observed plasma concentration and time to that concentration, and symptom measures in Parts B and C. The registry record reports results for the Epworth Sleepiness Scale and weekly cataplexy rate, along with treatment-emergent adverse events in Parts B and C. The primary safety results are specifically from Part A; they should not be treated as the primary endpoint results for the full four-part program.

The study started on 2020-05-27, had primary completion on 2021-11-05, and is listed as terminated. The reason for termination is not reported in the registry record. Nor is it reported whether termination changed the planned conduct or analysis of any part.

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: Participants with Narcolepsy Type 1 or Type 2 were planned to be enrolled in study in Parts A, B, C and D to receive TAK-994 or placebo. Due to the early termination of the study, data was not collected for Part C: placebo and was insufficient for Part C: TAK-994 180 mg to allow the pre-planned analyses. Similarly, for Part D, the prespecified sample size at Week 4 was not reached and therefore the pre-planned analyses cannot be adequately interpreted.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Safety Analysis Set included all participants who were randomized and received at least 1 dose of study drug.

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: Safety Analysis Set included all participants who were randomized and received at least 1 dose of study drug.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Part A: Number of Participants Who Experience at Least One Treatment Emergent Adverse Event (TEAE) During the Study, measured over First dose of study treatment to end of study follow-up (up to Day 35) in Part A.

The primary endpoint counts participants in Part A who experienced at least one treatment-emergent adverse event during the defined safety follow-up. The registry defines an adverse event as an unfavorable or unintended medical occurrence temporally associated with study drug use, whether or not it is considered related to treatment. A treatment-emergent adverse event is one whose onset occurs after study drug is received. The endpoint timeframe runs from first dose through the end of follow-up, up to Day 35.

This kind of endpoint is used to describe the frequency of participants who had one or more newly arising events during treatment and follow-up. It is a participant count: someone with multiple events is still counted once for this endpoint. It does not, by itself, show how many events occurred, how severe they were, how long they lasted, or whether investigators judged them to be caused by treatment. Those questions require other safety summaries.

In Part A, the posted results list 2 participants in the placebo arm, 6 in the TAK-994 120 mg arm, and 7 in the TAK-994 180 mg arm as experiencing at least one treatment-emergent adverse event. The corresponding arm sizes are 7, 7, and 8 participants. These are descriptive counts from the safety analysis set, defined in the registry as randomized participants who received at least one dose of study drug.

The endpoint is appropriate to a safety-focused primary question, but it is not a symptom-efficacy endpoint. It also does not establish causation or provide a complete accounting of safety. Serious events, event types, and results from other study parts require separate interpretation.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The reported statistical analyses need to be read in light of the outcome type, study part, and analysis method. The primary Part A measure is a count of participants with at least one treatment-emergent adverse event. The secondary symptom outcomes are continuous change scores or event-rate outcomes. Different methods answer different statistical questions, so their estimates should not be treated as interchangeable.

What does a confidence interval describe?

General explanation: A confidence interval expresses uncertainty around an estimate under the assumptions of the analysis. A narrower interval indicates greater precision than a wider interval, though the interval does not remove possible bias from study design, missing data, or model assumptions. The registry’s reported symptom comparisons include 95% confidence intervals. For the Epworth Sleepiness Scale, the intervals are for model-estimated mean differences. For weekly cataplexy rate, the intervals accompany incidence rate ratios. The scale and interpretation differ between these outcomes.

The reported Epworth analyses use MMRM, a mixed model for repeated measures, and give least-squares mean differences. In general, MMRM models repeated measurements over time and estimates group differences while accounting for the structure of repeated observations. The registry record does not report the model’s full specification, including its covariates, covariance structure, or missing-data assumptions. Those details are important for reproducing or fully assessing the analysis and are not reported in the registry record.

The posted weekly cataplexy rate analyses also identify MMRM as the method, but report incidence rate ratios. An incidence rate ratio compares modeled event rates between groups; it is a ratio measure, not a difference in mean change scores. Its interpretation depends on the model and the event-rate scale. The analysis details needed to judge those assumptions are not reported in the registry record.

How do the recalculated comparisons differ?

The simple recalculations accompanying some results use posted summary statistics to form unadjusted comparisons. They are not the sponsor’s repeated-measures model analyses. For example, the registry reports an MMRM least-squares mean difference for each posted Epworth comparison, while the recalculated comparison uses arm-level change summaries and an approximate standard deviation. The recalculated confidence intervals and p-values therefore need not match the reported model results. They address a simpler, unadjusted comparison and should not replace the reported analysis.

For weekly cataplexy rate, the recalculated mean differences are also unadjusted comparisons from summary data. They are not the same estimand as the registry’s reported incidence rate ratios. In general, rate models can be useful for event counts because they represent rates on a ratio scale, while a mean difference represents an absolute difference on the measured scale. The registry’s reported model results should be interpreted according to the measure named in each analysis, not converted into another measure without the underlying data.

Why can over-dispersion matter?

General explanation: Over-dispersion occurs when event counts vary more than a model’s assumed distribution allows. If ignored, uncertainty estimates can be too small and tests can appear more decisive than warranted. Weekly cataplexy counts can vary substantially among participants, and the posted arm summaries include standard deviations. The registry record does not say whether or how over-dispersion was assessed or handled in the reported model.

The posted planning illustrations use observed effects, a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They show participants per arm for 80% or 90% power when planning for 100%, 75%, or 50% of the observed effect. These are hypothetical planning calculations, not the trial’s own sample-size justification and not observed power. They do not show that the trial was adequately powered for its primary endpoint or its secondary outcomes. The trial’s sample-size assumptions and justification are not reported in the registry record.

Unadjusted p-values are also limited. They quantify compatibility with a null hypothesis under the specific analysis assumptions; they do not measure the probability that a treatment effect is true, the clinical importance of an estimate, or the chance of bias. They should be considered alongside effect estimates, interval width, outcome definitions, and the number of analyses performed.

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: Placebo", "Part A: TAK-994 120 mg", "Part A: TAK-994 180 mg")
n      <- c(7, 7, 8)
events <- c(2, 6, 7)
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: Placebo  7  2  28.6  3.7  71.0  8.2  64.1  8.5  65.1
ARM  Part A: TAK-994 120 mg  7  6  85.7  42.1  99.6  48.7  97.4  47.3  96.8
ARM  Part A: TAK-994 180 mg  8  7  87.5  47.3  99.7  52.9  97.8  51.8  97.2

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple outcomes across study parts, including safety, pharmacokinetic, sleepiness, and cataplexy measures. Several doses and treatment comparisons are represented in the posted results. When many hypotheses are tested, the chance of at least one apparently strong result can increase. Statistical approaches such as family-wise error control, false-discovery control, or a prespecified testing sequence can address multiplicity, but the registry record does not report whether any such method was planned or used.

The record includes p-values for reported secondary analyses, but does not state whether they are adjusted for multiple comparisons. They should therefore be read as the reported p-values, without assuming that they control the overall false-positive risk across endpoints, doses, or parts. The protocol’s statistical analysis plan, including any endpoint hierarchy or multiplicity procedure, is not reported in the registry record.

Missing-data handling is also not reported. The outcome summaries include differing analysis counts across arms, and the Part C symptom outcomes are not estimable in the posted results. The registry record does not explain why results are unavailable, whether participants discontinued, how missing measurements were handled, or what assumptions were made by the MMRM analyses. These details matter because conclusions from a repeated-measures model can depend on the pattern and assumptions surrounding missing observations.

The primary endpoint population is described as the safety analysis set: randomized participants who received at least one dose of study drug. The record does not report the complete analysis-population definitions for each secondary outcome or whether every randomized participant was included in each efficacy analysis. It also does not report interim analyses, stopping rules, or the reason the trial was terminated. Without these details, it is not possible to determine whether the posted analyses followed a prespecified plan or how termination affected their interpretation.

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: hepatitis acute (5.3%).

Serious adverse events are a separate safety category from the primary endpoint’s count of participants with any treatment-emergent adverse event. The adverse-event results list no serious adverse events in the Part A placebo, 120 mg, or 180 mg groups. In Part B, the placebo, 30 mg, and 90 mg groups list none; the 180 mg group lists 1 serious adverse event. Part C lists none in its TAK-994 180 mg group. The record lists zero deaths in each posted group.

The registry also names Hepatitis acute as a top serious-event term and gives 5.3%. The record does not explain the denominator or additional context for that percentage in the information available for these results, so it should not be treated as a rate for a particular arm without that information. Serious-event counts are small and do not, on their own, establish whether an event was caused by study treatment. They also do not capture non-serious adverse events, which are reported separately and were more numerous in the posted groups.

These safety figures span different study parts and treatment periods. They are descriptive counts, not a formal comparative safety analysis. The registry record does not report a statistical test for serious adverse events or detailed causality assessments. The counts therefore support a description of what was recorded, but not a definitive comparison of risk between treatments.

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

The posted results establish what the registry reports for the measured outcomes in the analyzed groups. In Part A, participants with at least one treatment-emergent adverse event are counted by arm. In Parts B and C, the registry reports selected symptom and safety outcomes, including model-based comparisons for sleepiness and cataplexy measures. These are observations from this trial, under the recorded study conditions.

They do not establish that TAK-994 caused or prevented any adverse event. An adverse event is not necessarily considered related to treatment, and the primary endpoint counts participants with events rather than assigning causation. Nor do the primary safety counts answer whether symptoms improved. The secondary symptom analyses address different outcomes and study parts, and do not prove that any observed change is clinically important. The registry record does not report a threshold for clinical importance.

The posted results also do not establish a complete picture across every planned part. Part D was described in the study summary, but numerical results for it are not reported in the registry record. Part C has a posted TAK-994 arm, but its symptom results are not estimable, and a numerical placebo result is not reported in the registry record. The record does not explain the reason the trial was terminated or the effect of termination on the analyses.

General statistical interpretation requires attention to uncertainty, selection of outcomes, analysis methods, missing data, and the number of comparisons. The registry provides some confidence intervals and p-values for secondary analyses, but does not report the full analysis plan or missing-data assumptions. The estimates should therefore be interpreted as trial-specific results, not as a comprehensive or definitive account of treatment effects. They do not by themselves support conclusions about other populations, settings, durations, or outcomes not studied or reported.

14. Why This Trial Matters Statistically

This trial illustrates why outcome labels and study parts matter when interpreting clinical results. Its primary endpoint is a safety count in Part A, while the posted sleepiness and cataplexy outcomes are secondary measures in Parts B and C. Treating all results as if they answered one common question would confuse safety with symptoms and mix distinct populations, treatment periods, and analysis methods.

It also shows why raw counts require context. The Part A arms contain small numbers of participants, and the primary endpoint is reported as counts without confidence intervals or a statistical comparison. A difference in observed counts is not, by itself, a precise estimate of a treatment effect. The secondary analyses provide model estimates and intervals, but their interpretation depends on model specification and missing-data assumptions that are not reported in the registry record.

Multiple doses and outcomes make multiplicity relevant. A collection of unadjusted p-values cannot be read as though it came from one prespecified comparison. The registry record does not report a testing hierarchy or adjustment procedure, limiting the strength of confirmatory claims from the secondary analyses.

Finally, the record’s four-part design and terminated status underline the importance of transparent reporting. Numerical results are not reported for every described part, and the reason for termination is not given. These gaps do not erase the posted findings, but they constrain the questions those findings can answer. Statistically careful reading keeps the reported estimates tied to their particular endpoint, analysis, and population rather than extending them beyond the available evidence.

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.

Several limitations are specific to the record. The primary safety endpoint is reported as participant counts for small Part A arms, with no confidence intervals or formal comparison listed. This limits assessment of precision. The record defines the safety analysis set, but does not supply detailed event-level information for the primary endpoint, such as event duration, severity distribution, or investigator assessment of relatedness for each participant.

The study is divided into parts with different populations, doses, durations, and outcomes. The Part A primary endpoint should not be combined with adverse-event counts from Parts B and C as though they were collected in the same comparison. Part C is described as enrolling Chinese participants with type 1 narcolepsy, but its posted numerical symptom results are not estimable. Numerical results for the type 2 narcolepsy Part D are not reported in the registry record.

Some secondary comparisons are recalculated from arm-level summaries, while the registry also reports model-based results. These are not necessarily the same estimand or method. The registry record does not provide the full MMRM specification, missing-data approach, sample-size justification, or multiplicity plan. It is therefore not possible to fully reproduce the reported analyses or evaluate all assumptions.

The record lists the trial as terminated and does not report why. It does not state whether termination affected enrollment, follow-up, or outcome availability. The posted data should be interpreted with these reporting limits in mind, especially when considering generalizability beyond the studied groups or drawing conclusions about outcomes that were not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of TAK-994 in Adults With Type 1 and Type 2 Narcolepsy — NCT04096560.
• Registry record last updated 2024-10-29; status last verified 2024-10.
• Results first posted 2024-10-29.
• 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

The registry record describes a Phase 2 randomized, quadruple-masked, parallel-group, placebo-controlled study of oral TAK-994 in adults with type 1 or type 2 narcolepsy. Enrollment is listed as 97. The study started on 2020-05-27, had primary completion on 2021-11-05, and is listed as terminated.

The primary endpoint was the number of participants in Part A with at least one treatment-emergent adverse event during follow-up up to Day 35. The registry also posts secondary safety and symptom outcomes, including Epworth Sleepiness Scale and weekly cataplexy rate results in Parts B and C. The reported secondary analyses include MMRM estimates and 95% confidence intervals.

The reason for termination, full statistical analysis plan, multiplicity handling, and missing-data methods are not reported in the registry record. Numerical outcomes for Part D are also not reported in the registry record.