Phase 2 Study of Triheptanoin (UX007) for the Treatment of Glucose Transporter Type 1 Deficiency Syndrome (Glut1 DS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 2 Study of Triheptanoin (UX007) for the Treatment of Glucose Transporter Type 1 Deficiency Syndrome (Glut1 DS) (NCT01993186), a randomized phase 2 study. The primary outcome is Percent Reduction From Baseline to Week 8 in Frequency of Total Seizures (Normalized to a 4-Week Rate).

2. Clinical Question

The question is about a difference between assigned study groups over the treatment period, not whether seizures changed in an individual participant or whether UX007 has a general effect outside the study. The registry identifies efficacy and safety as primary objectives. For efficacy, the stated focus is reduction from randomization to Week 8 in seizure frequency. The posted primary outcome describes baseline and Week 8, with observable seizures assessed for 6 weeks after titration and absence seizures assessed overnight by EEG.

Frequency is expressed as a percent reduction, so negative values represent an increase rather than a reduction in seizure frequency. The posted group summaries and comparison estimate capture different aspects of the data: the group medians describe typical observed reductions, while the between-group estimate and its confidence interval address the comparative question. The interval is broad, so the data do not determine the direction or size of a comparative effect precisely.

3. Trial Design

This was a randomized, parallel-group Phase 2 study of UX007 and placebo in Glucose Transporter Type 1 Deficiency Syndrome. The registry lists enrollment as 36 and status as completed. The recorded study start was 2014-02-28, and the primary completion date was 2017-09-20. These dates describe the registry record; they do not provide additional detail about each participant’s treatment or follow-up.

The study is described as double-blind and quadruple-masked. These design labels indicate that masking was part of the study, but the registry record does not report which specific roles were included in the quadruple masking. Allocation was randomized, and the groups were parallel rather than a crossover in which participants receive both treatments.

The UX007 and placebo arm descriptions both report a 2-week dose-titration period aimed at a dose comprising up to 35% of total daily calories or the maximum tolerated dose level, followed by a 6-week treatment period maintained at the 35% total daily calorie dose level. The arm descriptions are truncated after beginning a subsequent statement, so additional details about what followed are not reported in the registry record. The primary outcome description says observable seizures were measured for 6 weeks after the 2-week titration period; absence seizures were measured overnight by EEG.

The efficacy population is described as randomized participants who received at least one dose of investigational product. The posted primary summaries show 25 participants for UX007 and 11 for placebo. The registry record does not report further detail on allocation procedures, concealment, treatment adherence, or the rationale for the enrollment target. The study’s stated primary objectives included both efficacy and safety, with safety assessed through adverse-event rates, laboratory values, and electrocardiogram values. Results for those safety domains are not fully described in the registry record presented for this analysis.

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: Beginning with the Screening visit, participants recorded seizure frequency during a 6-week Baseline Period. If the participant did not meet the seizure count criteria, the participant was considered a screen failure and was not randomized.

Double-Blind Placebo-Controlled Period

Reasons for not completing this period:

Open-Label Extension Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy Analysis Set: all randomized participants who received at least one dose of investigational product.

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.

6. Primary Endpoint

The registry lists the primary outcome as Percent Reduction From Baseline to Week 8 in Frequency of Total Seizures (Normalized to a 4-Week Rate), measured over Baseline, Week 8.

The primary endpoint is the percent reduction from baseline to Week 8 in frequency of total seizures, normalized to a 4-week rate. It combines two types of seizure measurement: observable seizures recorded by diary and absence seizures measured overnight by EEG. The diary categories include several observable seizure types; EEG assessment includes absence and indeterminate absence events during awake and sleep periods. The registry record describes the observable-seizure assessment as covering 6 weeks after the 2-week titration period.

Normalization to a 4-week rate expresses seizure frequency on a common time basis, which helps make frequency summaries interpretable when observations cover a defined follow-up interval. Percent reduction also describes change relative to baseline. A positive reduction indicates fewer recorded seizures relative to baseline; a negative value indicates an increase in frequency, as the registry explicitly states. It is therefore possible for a participant’s value to be negative, and the full range of posted values includes such increases.

This endpoint is used because seizure frequency is the trial’s stated primary efficacy focus. Combining diary-recorded observable seizures with EEG-measured absence seizures gives the endpoint a broader scope than either measurement alone. At the same time, the combined measure depends on both recording approaches and on the baseline-to-Week 8 comparison. The registry record does not report additional details about how missing diary entries or unavailable EEG measurements were handled, or the exact rules used to combine the component counts.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary comparison used a Wilcoxon rank-sum test, and the registry reports a Hodges-Lehmann estimate with a 90% confidence interval. These methods are suited to comparing two independent groups when the outcome is summarized by ranks rather than relying on a normal-distribution model for the raw values. The registry reports medians and full ranges for the groups, consistent with describing an outcome that can be highly variable and may include extreme values.

What does a rank-sum test compare?

As a general statistical explanation, a rank-sum test orders observations across groups and compares their rank distributions. It is not simply a test of whether the group medians differ. Its interpretation depends on the shape and spread of the distributions as well as their location. When the groups have similarly shaped distributions, a location shift can be a useful way to understand the comparison; when shapes differ, the result is more broadly about differences in distributions. The registry record does not report distribution plots or additional diagnostics for this endpoint.

How should the Hodges-Lehmann estimate be read?

As a general explanation, the Hodges-Lehmann estimate is a rank-based estimate of the location of differences between observations in two groups. It provides a point estimate in the outcome’s units, while the confidence interval represents uncertainty around that estimate. Here, the registry reports an estimate of 13.45 and a 90% confidence interval from -38.63 to 80.95. Since the interval extends across zero, the data are compatible with effects in either direction under the reported analysis. The interval does not mean there is a 90% probability that the true effect lies within these particular limits; it describes the coverage property of the method over repeated applications under its assumptions.

What does the p-value add?

The reported p-value is 0.581. In general, a p-value measures how unusual data at least as inconsistent with a specified null hypothesis would be if that null hypothesis and the analysis assumptions held. It is not the probability that the null hypothesis is true, nor a measure of the probability that UX007 is effective. A p-value above a conventional significance threshold does not demonstrate equivalence or rule out an important effect. The registry record does not report a prespecified significance threshold for this analysis, so no conclusion about the study’s formal decision rule should be inferred.

Why does outcome spread matter?

The primary outcome’s full ranges extend widely in both groups, and negative values indicate increases in seizure frequency. As a general statistical point, substantial spread makes group comparisons less precise, particularly when the number of observations is limited. Extreme values can also have a strong influence on means, which is one reason rank-based methods and medians can be useful descriptions. Rank methods are not immune to variability; broad overlap and heterogeneous outcomes still limit how precisely the groups can be distinguished.

What are the planning illustrations?

The accompanying planning illustrations are not observed power calculations and do not describe the trial’s original design justification. They use observed effects as planning inputs, with the stated assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation. General statistical interpretation: such calculations illustrate how a future study’s required size can depend on the effect chosen for detection and the desired power. They are sensitive to the observed effect, which may be unstable in a small study, and should not be read as a guarantee that a future study will attain a particular result. The registry record does not report the original sample-size justification or its assumptions.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary between-group analysis as a Wilcoxon rank-sum test, with a Hodges-Lehmann estimate of 13.45, a 90% confidence interval from -38.63 to 80.95, and p = 0.581. The group summaries are medians with full ranges. The comparison estimate and its interval should be read together: the point estimate is only one possible summary of the observed contrast, while the interval conveys substantial uncertainty about its direction and size.

No discrepancy is flagged between the registry’s reported primary analysis and a recalculation, so there is no identified conflict requiring a recalculated result to replace the sponsor’s analysis. The reported sponsor analysis is the primary inferential result. Recalculated comparisons for selected secondary outcomes are unadjusted illustrations based on posted summaries; they are not substitutes for the sponsor’s primary analysis or for any sponsor analysis that is not reported.

The registry record does not report additional details about the primary test’s prespecified decision rule, handling of missing endpoint data, or any model-based sensitivity analyses. Those omissions limit how far the analysis can be independently evaluated. The reported result should therefore be interpreted as the registry’s stated comparison, with its broad confidence interval and the outcome variability kept in view.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple outcome measures and analyses, including a primary seizure-frequency endpoint and several secondary measures. It does not report a multiplicity adjustment or a hierarchy of testing across the outcomes. Without that information, the secondary p-values and comparisons should be treated as exploratory rather than as a family of confirmatory tests. As a general statistical point, evaluating many endpoints increases the chance of at least one apparently unusual result, so an isolated unadjusted result is less persuasive than a prespecified primary comparison.

The registry identifies the efficacy analysis population as randomized participants who received at least one dose of investigational product. This definition is not identical to an analysis of every randomized participant regardless of treatment received. The record does not report a separate intention-to-treat analysis, the number of participants with complete primary endpoint data, or the reasons for any exclusions from a particular outcome analysis. The posted primary group summaries list 25 for UX007 and 11 for placebo, but the registry record does not provide a full participant-flow account in the registry record.

The primary outcome combines diary-based observable seizures with EEG-measured absence seizures. The registry describes the measurement approaches and timeframe, but does not report how missing diary entries, incomplete EEG recordings, or other unavailable measurements were handled. It also does not report imputation methods, sensitivity analyses for missing data, or assumptions about whether missingness was related to outcomes. These details matter because the validity of a treatment comparison can depend on how unavailable observations are addressed.

The registry record also does not report whether interim analyses were planned or conducted, the original sample-size justification, or a complete statistical analysis plan. Those points should be regarded as not reported in the registry record, rather than inferred from the posted final status or from the observed results. In the absence of these details, interpretation should remain focused on the reported primary comparison and acknowledge the limits on evaluating prespecification, missingness, and multiplicity.

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: status epilepticus (5.9%), subcutaneous haematoma (2.9%), seizure (2.9%).

During the placebo-controlled period, the registry reports serious adverse events for 1 participant in the UX007 group and none in the placebo group. Other adverse events are reported for 21 participants in the UX007 group and 9 in the placebo group. The registry also reports an extension-period UX007 group, with serious adverse events for 2 participants and other adverse events for 31 participants. No deaths are reported in these groups.

The registry lists Status Epilepticus, Subcutaneous Haematoma, and Seizure among the top serious adverse-event terms, with the percentages recorded for each term. These terms provide event-level context, but the record excerpt does not specify which study period each listed term belongs to or provide a full account of event timing, duration, severity, relatedness, or clinical management.

These are descriptive safety counts, not a controlled estimate of comparative risk. The extension-period group is presented separately and should not be compared directly with the placebo-controlled groups as though assignment and observation conditions were the same. Small group sizes and the limited detail in the registry record constrain interpretation. The reported counts alone do not establish whether an event was caused by treatment, and the registry’s stated safety assessments also included laboratory values and electrocardiogram values, for which results are not reported in the registry record.

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

The primary result describes a comparison of Week 8 seizure-frequency reduction between UX007 and placebo in the registry’s efficacy analysis set. The posted medians differ descriptively, and the rank-based estimate is positive. However, the confidence interval is broad and spans values in both directions, while the reported p-value does not provide strong evidence against the null hypothesis under the stated test. The result is therefore uncertain, not a demonstration that the treatments are equivalent and not proof that UX007 has no effect.

Nor does the point estimate establish that UX007 reduces seizures for a particular participant. Group summaries describe distributions across the analyzed participants, and individual outcomes varied widely. The endpoint combines diary and EEG measurements and applies to the study’s specified treatment period and Week 8 assessment. It does not answer questions about longer-term outcomes, other populations, or treatment outside the study conditions.

The secondary outcomes offer additional context but do not supply a consistent basis for a stronger conclusion. Their interpretation is limited by the number of outcomes, the unadjusted nature of recalculated comparisons, and the fact that results are not reported for every listed outcome in the record details available for this analysis. Safety summaries are also descriptive and do not by themselves establish treatment causation or provide a complete comparison of harms.

In general, a confidence interval is useful because it shows the range of effects that remain compatible with the observed data and analysis. Here, its breadth means that a range of possible comparative effects remains unresolved. A non-significant p-value is not evidence that all clinically important effects have been ruled out. The appropriate conclusion is bounded: the registry reports an imprecise primary comparison, with substantial observed variability, and the available results do not settle the size or direction of the comparative effect.

14. Why This Trial Matters Statistically

This trial is statistically instructive because the primary outcome is a change in seizure frequency with substantial variability, including negative values that represent increased frequency. Such an outcome makes it important to distinguish the typical group summaries from the full distribution and from the comparative estimate. A median alone cannot convey how broadly individual outcomes vary, while a point estimate without its interval can suggest more certainty than the data support.

The analysis also illustrates why the method used matters. The registry reports medians and full ranges alongside a rank-based test and a Hodges-Lehmann estimate. These approaches offer a way to compare groups without treating the raw endpoint as normally distributed, but they do not eliminate uncertainty arising from small groups and wide outcome variation. Interpretation requires attention to both the estimand and the confidence interval, not just a p-value.

The endpoint’s combination of diary-recorded and EEG-measured events raises practical measurement questions. As a general statistical matter, outcomes assembled from different measurement processes can be difficult to interpret when details about missing observations and component handling are not reported. That makes transparent definitions, data-quality procedures, and sensitivity analyses important for evaluating future evidence.

Finally, the registry includes several secondary outcomes and only selected reported comparisons. This is a reminder that a broad outcome list does not automatically produce confirmatory evidence across every measure. Multiplicity, prespecification, complete reporting, and the distinction between exploratory recalculation and sponsor-reported analysis all matter. The trial’s statistical value lies not in a simple yes-or-no label, but in showing how uncertainty, measurement, analysis choices, and reporting shape the conclusions that can responsibly be drawn.

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 estimate is imprecise, with a broad 90% confidence interval, and the full ranges show substantial variation in seizure-frequency change. These features limit confidence in the size and direction of the comparison. The posted primary group summaries include 25 participants for UX007 and 11 for placebo, so estimates from this analysis should not be treated as highly precise descriptions of a broader population.

The registry’s endpoint definition combines observable seizures recorded by diary with absence seizures assessed overnight by EEG. The record describes seizure categories and the assessment period, but does not report how the two components were combined in detail, how missing measurements were handled, or whether sensitivity analyses addressed alternative assumptions. These omissions make it harder to evaluate the robustness of the endpoint result.

The available record does not include a complete statistical analysis plan, original sample-size justification, multiplicity strategy, interim-analysis information, or full participant-flow details. These items are not reported in the registry record and should not be inferred. The record also does not provide results for every listed secondary outcome in the registry record for this analysis, limiting assessment of the full outcome pattern.

The secondary recalculations are unadjusted, and planning illustrations based on observed effects are not evidence of the original study’s design assumptions or of observed power. The safety counts are descriptive; extension-period observations are not directly comparable to placebo-controlled observations, and details such as event relatedness and timing are not fully reported. These limitations qualify interpretation without changing the reported numbers or the registry’s primary analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Phase 2 Study of Triheptanoin (UX007) for the Treatment of Glucose Transporter Type 1 Deficiency Syndrome (Glut1 DS) — NCT01993186.
• Registry record last updated 2020-06-19; status last verified 2020-06.
• Results first posted 2020-05-29.
• 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 describes a completed, randomized, quadruple-masked, parallel-group Phase 2 study of UX007 versus placebo for Glucose Transporter Type 1 Deficiency Syndrome. Enrollment is listed as 36. The primary efficacy endpoint was percent reduction from baseline to Week 8 in total seizure frequency, normalized to a 4-week rate, combining observable seizures measured by diary and absence seizures measured by EEG.

The posted primary analysis used a Wilcoxon rank-sum test and reported a Hodges-Lehmann estimate of 13.45, with a 90% confidence interval from -38.63 to 80.95 and p = 0.581. The group medians were 12.6 for UX007 and 0.0 for placebo, with wide full ranges. The comparison remains imprecise and does not establish the direction or size of a comparative effect.

Secondary outcomes cover seizure responses and cognitive measures, but the available results are incomplete and unadjusted recalculations are exploratory. Safety summaries report serious and other adverse events by study period; they are descriptive and do not establish causation. Several important analysis details, including missing-data handling, multiplicity procedures, and the original sample-size justification, are not reported in the registry record.