Exploring the Safety And Tolerability of Doses of E2007 up to a Maximum of 12 mg In Patients With Refractory Partial Seizures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exploring the Safety And Tolerability of Doses of E2007 up to a Maximum of 12 mg In Patients With Refractory Partial Seizures (NCT00416195), a randomized phase 2 study. The primary outcome is Percentage of Responders During the Maintenance Phase.

2. Clinical Question

The registry describes 48 participants, with two groups of 24 based on whether participants took concomitant inducing anti-epileptic drugs. Within each group, 18 were assigned to E2007 and six to placebo. These are the group sizes described in the study summary; the registry does not report the number included in each arm’s primary endpoint results.

The clinical question is therefore about a defined seizure-frequency response over a particular phase of treatment, not about every aspect of epilepsy or long-term outcomes. The record also lists percentage change in 28-day seizure frequency during maintenance using LOCF as a secondary outcome. It does not report the results for either efficacy outcome, so the relative responder outcomes and their precision cannot be assessed from the registry record.

3. Trial Design

This was a randomized, double-blind, placebo-controlled, parallel-group phase 2 study sponsored by Eisai Co., Ltd. The registry lists enrollment of 48 participants and a status of completed. The study start is recorded as 2006-12 and primary completion as 2008-02. Its stated purpose was to explore the safety and tolerability of E2007 doses up to a maximum of 12 mg in patients with refractory partial seizures.

Participants were divided into two groups according to concomitant inducing anti-epileptic drug use. The study summary describes 24 participants in each group. In each group, 18 were to receive E2007 and six placebo. Randomization is listed as the allocation method, masking as double, and the model as parallel-group. The registry does not report further details about the randomization procedure, allocation concealment, or who was masked.

The E2007 arm description lists once-daily doses of 2 mg for Days 1 to 14, 4 mg for Days 15 to 28, 6 mg for Days 29 to 42, 8 mg for Days 43 to 56, and 10 mg for Days 57 to 70, followed by a description that ends partway through the 12 mg schedule. The full final dosing description is not reported in the registry record. The placebo description specifies matching placebo once daily for 16 weeks, Days 1 to 112.

The primary endpoint concerns the percentage of responders during maintenance, with the timeframe specified as Day 85 through Day 112. This endpoint definition and timeframe should be distinguished from the treatment schedule: the endpoint is a participant-level response classification over a specified period, whereas the regimen describes assigned dosing. The registry does not report the efficacy results, participant flow by arm, or further operational details of the treatment and assessment schedule.

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: ITT population- all subjects in the Safety Population (all randomized subjects who took at least 1 dose of study drug) who had at least 2 weeks of baseline seizure frequency data and at least 1 week of seizure frequency data after baseline (LOCF - last observation carried forward)

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 Percentage of Responders During the Maintenance Phase, measured over Day 85 through Day 112.

The primary endpoint is the percentage of participants classified as responders during the maintenance phase, defined as Day 85 through Day 112. A participant counted as a responder if seizure frequency fell by 50% or greater from baseline. This converts each eligible participant’s change in seizure frequency into a binary outcome: responder or not responder. The endpoint is reported in the registry as a percentage of participants.

In general, a responder endpoint can make a clinical outcome easier to interpret than a group average alone: it expresses how many participants cross a prespecified threshold of change. But the classification discards information about the size of changes beyond that threshold. A participant just meeting the threshold and one with a larger reduction both count as responders, while a participant just below it is classified differently. The percentage therefore answers a threshold-specific question, not the full question of how seizure frequency changed across the group.

The registry defines the analysis population as all subjects in the Safety Population who had at least 2 weeks of baseline seizure-frequency data and at least 1 week of post-baseline data, using LOCF. The Safety Population is defined as all randomized subjects who took at least 1 dose of study drug. This means endpoint eligibility depends on both treatment exposure and seizure-frequency data availability. Responder counts, percentages, and uncertainty estimates are not reported in the registry record, so the observed endpoint cannot be interpreted from the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Randomization is intended, in general, to make assigned groups comparable at the start of a trial by using chance to determine treatment assignment. It can reduce systematic differences between groups, although it does not guarantee identical groups, especially in a study with limited enrollment. This record lists randomized allocation but does not report the randomization sequence, allocation concealment, or whether assignment was stratified by concomitant inducing anti-epileptic drug use.

Double masking means that more than one category of study participant or staff member is intended not to know the assigned treatment. In general, masking can reduce the influence of expectations on treatment delivery, reporting, or assessment. The registry labels masking as double but does not identify which roles were masked or describe how masking was maintained. A placebo comparator can help keep study procedures similar across assigned groups, but the record does not report whether participants or investigators correctly guessed assignments.

A parallel-group design assigns participants to a study group and compares outcomes across groups. Here, the record describes E2007 and matching placebo. The summary further separates participants by use or non-use of concomitant inducing anti-epileptic drugs and gives planned allocation within each of those groups. The registry does not provide outcome estimates within those subgroups or explain whether subgroup comparisons were prespecified. Consequently, no subgroup-specific result should be inferred from the design description alone.

What does the responder outcome estimate?

The primary outcome is a binary participant-level classification, summarized as a percentage. In general, a percentage of responders estimates the proportion of an analyzed group meeting a stated definition, while a between-group contrast compares those proportions. The registry gives the response threshold and timeframe, but it does not give the numbers of responders or the resulting percentages. It therefore does not allow an effect estimate, confidence interval, or significance test for this outcome to be described.

The secondary outcome is percentage change in 28-day seizure frequency from baseline during maintenance, with LOCF. This differs from the responder endpoint: it retains a numerical measure of change rather than classifying participants only by whether they cross a threshold. In general, a numerical outcome may reveal variation in the magnitude of changes, while the responder measure offers a threshold-based summary. The registry does not report the secondary outcome values, their spread, or the analysis comparing groups.

What do confidence intervals and p-values mean?

In general, a confidence interval communicates statistical uncertainty around an estimated effect under the chosen method and assumptions. A narrower interval indicates greater precision than a wider interval, but neither width nor location alone establishes clinical importance. A p-value, in general, describes how incompatible the observed data are with a specified null model, under its assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The registry record reports no primary endpoint confidence interval or p-value, so neither precision nor statistical compatibility with a null model can be evaluated.

The sample-size assumptions described for planning illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is expressed as a stated fraction of the observed effect. These are planning assumptions, not evidence about the effect observed in the trial. The record does not report a sample-size justification from the original protocol, and the planning assumptions should not be interpreted as proof that the trial had a particular probability of detecting an effect. Sample-size calculations are not observed power: they describe the sample size associated with chosen assumptions, while observed precision depends on the data and the analysis.

In general, over-dispersion occurs when variability is greater than a statistical model expects. When a model fails to account for such variability, its standard errors and uncertainty estimates can be misleading. This trial’s primary endpoint is binary, and no model specification or dispersion assessment is reported in the registry record. It is therefore not possible to say whether an over-dispersion adjustment was relevant to the sponsor’s analysis or used. This general concept should not be read as a claim that over-dispersion occurred in this trial.

For a binary outcome, the choice of effect measure also shapes interpretation. A difference in responder percentages, a ratio of those percentages, or another contrast answers a different statistical question. The registry record does not report the effect measure used for the primary comparison, nor does it provide the analysis result. No recalculated comparison is available from the posted primary endpoint data. Any numerical efficacy claim would require results that are not reported in the registry record.

Simple recalculations from summary statistics, when available, can be useful as checks, but they may not match a sponsor’s adjusted analysis or analysis population. If a discrepancy is identified, the sponsor’s own analysis takes precedence. In this record, there is no primary outcome analysis listed, and the registry lists zero analyses. Accordingly, there is no reported sponsor estimate to compare with a recalculation and no basis for adding an independent effect estimate.

10. Multiplicity, Missing Data and Analysis Populations

The registry states that the primary outcome has several categories or time points and is not reduced to one comparison. Multiple comparisons can arise when several outcome versions, time points, subgroups, or treatment contrasts are examined. In general, testing many comparisons without an appropriate strategy can increase the chance of at least one apparently notable result arising by chance. Methods such as prespecifying a primary comparison or adjusting for multiple testing can address some multiplicity concerns, but the registry record does not report the sponsor’s multiplicity strategy.

The record lists zero analyses. It does not provide a primary treatment contrast, statistical test, confidence interval, or p-value, and it does not report how the categories or time points were analyzed. It also does not report whether the analysis plan was prespecified, whether any interim analyses occurred, or whether the sponsor adjusted for multiple comparisons. Those details should not be inferred from the endpoint description or the stated sample-size assumptions.

Missing-data handling is relevant because the primary analysis population requires baseline and post-baseline seizure-frequency data. The record says post-baseline values were handled using LOCF, or last observation carried forward. In general, LOCF fills a later missing measurement with a participant’s last observed value. That approach can be problematic if outcomes change over time or if the reason for missing data is related to the outcome; it can understate or distort uncertainty by treating an imputed value as if it were observed. These are general limitations, not evidence that a particular bias occurred in this trial.

The record does not report the amount or pattern of missing seizure data, reasons for missingness, or sensitivity analyses using alternative assumptions. It defines the primary population using the Safety Population plus minimum data requirements: randomized participants had to take at least 1 dose, have at least 2 weeks of baseline data, and have at least 1 week of post-baseline data. The number meeting those requirements in each arm is not reported in the registry record. Thus, the effect of these eligibility rules on the analyzed groups cannot be assessed.

11. Safety Results

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

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: convulsions (10.0%), colonic polyp (2.6%).

The adverse-event record lists serious adverse events for 10 participants at risk in Placebo, with one affected, and for 38 participants at risk in Perampanel, also with one affected. For other adverse events, the corresponding counts are seven affected among 10 at risk in Placebo and 15 among 38 at risk in Perampanel. These are counts in the registry’s safety groups, not the planned allocation described in the study summary. No death count is reported; the death fields are null, which should not be read as a confirmed zero.

The record lists Convulsions at 10.0% and Colonic polyp at 2.6% among the top serious adverse-event terms. It does not report further details about timing, recurrence, severity beyond the serious-event classification, or participant-level overlap between event categories. The adverse-event group is named Perampanel, whereas the intervention is named E2007; the registry record does not explain that naming difference.

In general, adverse-event counts describe events recorded in a particular safety population and over a defined observation period. They do not by themselves establish that treatment caused an event. Comparisons also require attention to group definitions, follow-up, and uncertainty. The registry record does not report inferential comparisons or confidence intervals for these safety counts, so their precision and the statistical significance of any between-group difference cannot be assessed.

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

The record establishes that this was a randomized, double-blind, placebo-controlled, parallel-group phase 2 study designed to explore E2007 safety and tolerability, including doses up to a maximum of 12 mg, in patients with refractory partial seizures. It defines a primary responder outcome and a maintenance-phase timeframe. It also gives an operational description of the endpoint population and lists certain safety counts. These details help explain what the study set out to measure and how the registry describes its populations.

The record does not report the primary responder results or the secondary seizure-frequency results. It therefore cannot establish the observed size or direction of an efficacy difference between E2007 and placebo, or how precisely any difference was estimated. The planned enrollment and assigned group counts do not reveal how many participants contributed to each endpoint or what outcomes they experienced. No treatment conclusion should be drawn from the endpoint definition alone.

Likewise, the listed adverse-event counts are descriptive. They do not establish whether an event was caused by treatment, and the record supplies no uncertainty estimates or statistical comparisons for the safety groups. The safety-group label Perampanel differs from the intervention name E2007, a discrepancy the registry record does not explain. The null death fields do not provide a reported count of deaths.

In general, statistical evidence is assessed by considering the estimate, its uncertainty, the analysis population, and the assumptions used. A responder percentage is only one summary of seizure-frequency change; it does not show the full distribution of individual changes. Missing-data methods and choices among multiple categories or time points can also affect interpretation. Because the corresponding analysis details and outcome results are not reported in the registry record, those features cannot be evaluated for this study.

The appropriate interpretation is limited to the information recorded: trial structure, endpoint definitions, and descriptive safety figures. The record does not supply the results needed to judge the primary comparison or its precision. Absence of a reported result is not evidence of no effect, just as the trial’s design and objective are not evidence of an effect.

13. Why This Trial Matters Statistically

This trial illustrates why a registry’s endpoint definition and outcome results need to be read separately. A clear responder threshold and timeframe explain what the endpoint was intended to capture, but they do not reveal how many participants met the definition. Without the numerator, analyzed denominator, and uncertainty around the estimate, the result cannot be reconstructed from enrollment or planned assignment counts.

It also highlights the distinction between an outcome’s measurement scale and its interpretation. A responder percentage reduces seizure-frequency change to a threshold classification. The secondary outcome retains percentage change in 28-day seizure frequency, using LOCF. In general, these approaches summarize different features of the data, and one cannot stand in for the other. Since neither outcome result is reported in the registry record, their findings cannot be compared.

The record also shows why the analysis population matters. The primary endpoint population is described through treatment exposure and minimum data availability, while the safety record lists different group sizes and uses a group name that differs from the intervention name. Each denominator answers a specific question about a defined analysis population; planned assignment counts, safety denominators, and endpoint denominators should not be treated as interchangeable.

Finally, the statement that the primary outcome has multiple categories or time points, without one comparison being identified, makes multiplicity and prespecification important considerations. The registry does not report how those comparisons were handled. General statistical principles cannot fill that gap: an analysis plan and outcome estimates are needed to understand how the evidence was generated and how much uncertainty remains. This is a useful example of why design descriptions alone cannot substitute for posted results.

14. 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 trial-specific limitations follow from what the registry record does and does not report. The primary endpoint is defined, but responder counts, arm-specific percentages, treatment contrasts, confidence intervals, and p-values are absent. Results for the listed secondary outcome are also not reported. The absence of these results prevents assessment of efficacy magnitude, direction, and precision.

The record describes an analysis population based on the Safety Population, baseline data availability, and post-baseline data availability, with LOCF for the latter. It does not report the number of participants excluded under these criteria, the amount or pattern of missing data, the reasons data were missing, or sensitivity analyses. The impact of these features on the endpoint estimate therefore cannot be assessed.

Although the primary outcome is noted to include several categories or time points, the registry does not describe the analysis strategy for those comparisons. It also does not report the statistical plan, protocol-level sample-size justification, interim analyses, or multiplicity handling. The planning assumptions described for illustrative sample-size calculations do not fill these gaps and should not be confused with the trial’s original design rationale.

The safety record uses a Perampanel group label while the intervention is named E2007, and the record does not explain the difference. The listed safety denominators also differ from the allocation counts in the study summary. Without clarification of the group definitions and participant flow, the relationship between planned groups and safety populations is uncertain. Null death fields do not establish a reported zero count. These limitations restrict interpretation to the descriptive details explicitly recorded.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Exploring the Safety And Tolerability of Doses of E2007 up to a Maximum of 12 mg In Patients With Refractory Partial Seizures — NCT00416195.
• Registry record last updated 2014-07-11; status last verified 2013-08.
• Results first posted 2012-11-22.
• 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.

18. Record Summary

The registry describes a completed randomized, double-blind, placebo-controlled, parallel-group phase 2 study of E2007 in people with refractory partial seizures. It lists enrollment of 48 and planned groups based on concomitant inducing anti-epileptic drug use, with E2007 and placebo allocation within each group. The primary endpoint was the percentage of responders during Day 85 through Day 112, using a 50% or greater reduction in seizure frequency from baseline.

The registry does not report the primary or secondary efficacy results, a treatment contrast, or its uncertainty. Safety counts are listed for Placebo and Perampanel groups, but the reason for the Perampanel label is not reported. The record gives no death count, and does not describe how multiple endpoint categories or time points were analyzed. Conclusions about efficacy or comparative safety cannot be drawn from the reported details alone.