To Evaluate The Efficacy and Safety of E2007 (Perampanel) Given as Adjunctive Therapy in Subjects With Refractory Partial Seizures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for To Evaluate The Efficacy and Safety of E2007 (Perampanel) Given as Adjunctive Therapy in Subjects With Refractory Partial Seizures (NCT00699582), a randomized phase 3 study. The primary outcome is Percent Change in the 28-day Seizure Frequency From Baseline to the End of the Double-blind Phase (Titration and Maintenance Phases).

2. Clinical Question

3. Trial Design

This was a Phase 3 randomized, double-blind, placebo-controlled, parallel-group study of perampanel as adjunctive therapy in subjects with refractory partial seizures. Its title describes a dose-escalation study. The registry lists three outcome arms: placebo, Perampanel 8mg, and Perampanel 12mg. It records total enrollment as 389 and study status as completed. The study start is listed as 2008-05 and primary completion as 2010-12.

The primary endpoint timeframe is baseline (pre-randomization) through Week 19. The endpoint description says seizure frequency per 28 days was derived from information recorded in subject diaries. The outcome population is identified as the Full Intent-to-Treat (ITT) Analysis Set, defined in the record as randomized subjects who received study drug and had any seizure-frequency data during the double-blind phase.

The registry record does not report the allocation ratio, detailed dose-escalation schedule, exact timing of visits, or the full arm descriptions. It also does not report the randomization procedure or details of how masking was maintained. Those design features should not be inferred from the trial title or from the fact that the study was randomized and double-blind. The posted outcome data present the arms separately; the registry does not supply an estimable two-arm comparison for this outcome. The tables and plots on this page display the reported arm-level results rather than replacing them with a calculated treatment-effect estimate.

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: Full Intent-to-Treat (ITT) Analysis Set - group of subjects who were randomized to study drug, received study drug, and had any seizure frequency data during the Doubleblind Phase. For Placebo arm, 2 subjects were randomized but not treated. For Perampanel 8mg arm, 1 subject was randomized but not treated.

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 Change in the 28-day Seizure Frequency From Baseline to the End of the Double-blind Phase (Titration and Maintenance Phases), measured over Baseline (Pre-randomization) through Week 19.

The primary endpoint measures the percent change in seizure frequency over 28 days, comparing baseline with the end of the double-blind phase. The registry specifies that the phase includes titration and maintenance and that the timeframe runs from baseline (pre-randomization) through Week 19. Seizure frequency was derived from subject diary information.

A frequency-based endpoint summarizes how often seizures occurred, while expressing change relative to the participant’s baseline can make changes easier to interpret across participants with different starting frequencies. That is a general explanation of this type of measure, not a statement about the trial’s design rationale, which is not reported in the registry record. Negative posted changes correspond to reductions in frequency under the endpoint’s change-from-baseline convention; positive values correspond to increases.

The registry reports the median percent change for each arm and labels the dispersion as the full range. A median describes the middle of the observed distribution rather than its arithmetic average. The full range conveys the span of posted values, not the uncertainty around the median or a confidence interval. The registry record does not report a confidence interval for the arm medians or a treatment-effect estimate for a comparison between perampanel and placebo. Thus, this endpoint description helps identify what the posted numbers summarize, but does not by itself provide a precise estimate of the difference between groups.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a median as the summary statistic for the primary endpoint and a full range as its measure of dispersion. It describes the endpoint as percent change in 28-day seizure frequency from baseline to the end of the double-blind phase. No further analysis-method details for this endpoint are reported in the registry record. The explanations below distinguish what the posted summaries mean in general from what can be concluded about this particular trial.

What does a median summarize?

In general, the median is the central value after observations are ordered. It is less affected by extreme observations than the arithmetic mean, which can be useful when a distribution is skewed or contains unusually large values. The registry reports medians for each arm, not means. A median does not tell us how many participants had a change close to that value, how the remaining observations are distributed, or whether the arms differ beyond what could arise from sampling variation.

Here, the posted arm medians are -9.72 for placebo, -30.52 for Perampanel 8mg, and -17.57 for Perampanel 12mg. A comparison of these descriptive summaries is not itself a formal test or an estimate with a stated precision. The registry does not report a confidence interval or p-value for the primary endpoint, and it flags the two-arm comparison values as missing or not estimable. No alternative inferential procedure can be attributed to the study from the information in the registry record.

How should the full range be read?

In general, the full range gives the smallest and largest observed values. It shows the overall span of the observations, but gives little information about how values are concentrated within that span. The range is particularly affected by extreme observations. It is not the same as a confidence interval: a confidence interval represents uncertainty in an estimated quantity under a statistical method, while the range describes observed endpoints in the data.

The registry reports ranges of -91.8 to 404.3 for placebo, -94.0 to 234.3 for Perampanel 8mg, and -100.0 to 858.3 for Perampanel 12mg. These are the reported full ranges, not interval estimates for the arm medians or a difference between arms. They show that values varied widely within each arm, including increases as well as decreases under the endpoint’s change convention. The record does not provide quartiles, standard deviations, or other distribution summaries for this outcome.

What do the analysis counts represent?

The posted outcome counts are 136.0 for placebo, 129.0 for Perampanel 8mg, and 121.0 for Perampanel 12mg. The stated outcome population is a modified ITT definition in the practical sense that inclusion requires randomization, receipt of study drug, and at least some seizure-frequency data during the double-blind phase. The registry calls it the Full Intent-to-Treat (ITT) Analysis Set; this page retains that registry terminology rather than assigning a different population label.

The population description notes that 2 subjects in the placebo arm were randomized but not treated. It also begins a note about a randomized participant in the Perampanel 8mg arm, but the available registry text is incomplete. The record does not report a complete account of exclusions by arm or how participants with incomplete diary data were handled. The outcome counts should therefore be read alongside the stated population definition, not assumed to represent all enrolled participants.

What can be inferred from the displayed comparison?

Randomization is a design feature intended, in general, to support fair comparisons by assigning participants to groups through a chance process. Double masking is generally intended to limit the influence of treatment knowledge on care or assessment. Those general purposes do not supply a missing statistical comparison. The registry presents outcomes by arm but does not give an estimable two-arm contrast for this outcome. Without an effect estimate and its uncertainty, readers cannot assess the precision of the difference from the posted summary statistics alone.

The sample-size section on the page is a planning illustration based on the observed effect, not the trial’s own design calculation. Its stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. Such figures are hypothetical planning quantities, not observed power and not evidence that the original trial was designed with those assumptions. The registry does not report its own sample-size justification or the assumptions used to choose the enrollment target.

Other methods details—including the statistical model, handling of repeated or diary-based observations, approach to missing outcome data, covariate adjustment, and any interim analyses—are not reported in the registry record. These omissions limit reconstruction of the analysis. They do not justify filling in a likely method based on common practice; conclusions should stay with the posted summaries and explicit record descriptions.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists the primary endpoint as percent change in 28-day seizure frequency and lists two secondary outcomes: responder rate, and percent change in 28-day complex partial plus secondarily generalized seizure frequency from baseline to the end of the double-blind phase. It also reports three outcome measures and no posted analyses. The registry record does not describe how the study handled multiple endpoints, multiple dose comparisons, or any planned hierarchy among outcomes. Any multiplicity adjustment or testing sequence is therefore not reported in the registry record.

The primary outcome population is described as randomized participants who received study drug and had any seizure-frequency data during the double-blind phase. The record notes that 2 placebo participants were randomized but not treated, and begins a note about a randomized participant in the Perampanel 8mg arm that is incomplete in the available text. The outcome denominators are 136.0, 129.0, and 121.0 for placebo, Perampanel 8mg, and Perampanel 12mg, respectively.

The registry does not report how missing diary entries or other missing outcome data were handled, whether any imputation was used, or what assumptions underpinned missing-data analyses. It also does not give a complete participant flow by arm in the information. Since the endpoint depends on seizure-frequency information recorded in subject diaries, the availability and completeness of those records are relevant, but the registry does not report details that would allow their impact to be assessed.

The stated analysis set requires some seizure-frequency data; it does not describe how participants without such data were treated in every analysis. The registry’s record does not provide a separate per-protocol analysis definition or sensitivity analyses for missing data. These gaps matter when interpreting an arm-level summary: the data included in the analysis and the approach to unavailable observations can affect the result. No assumptions about those procedures should be made from the outcome values alone.

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: convulsion (1.6%), facial bones fracture (0.8%), accidental overdose (0.8%), head injury (0.8%), radius fracture (0.8%).

The posted safety groups report serious adverse events in 7 of 136 participants in the placebo group, 10 of 129 in the Perampanel 8mg group, and 12 of 121 in the Perampanel 12mg group. Other adverse events are reported in 55 of 136, 90 of 129, and 90 of 121 participants, respectively. These are counts of participants affected within the listed groups. They are not measures of event severity beyond the serious-adverse-event classification, and they do not establish that a group difference is statistically reliable.

The registry lists serious adverse-event terms and percentages including Convulsion, 1.6%; Facial bones fracture, 0.8%; Accidental overdose, 0.8%; Head injury, 0.8%; Radius fracture, 0.8%; and Dizziness, 0.8%. These entries describe reported terms; the record does not provide enough detail to infer cause or compare each event’s risk between arms. Death counts are not reported in the registry record.

As a general point, safety counts need context: how events were defined, how long participants were followed, whether a participant could experience more than one event, and how events were assessed. Details needed to interpret those features are not reported in the registry record. The counts should be read as descriptive summaries, not as proof of safety or harm.

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

The posted primary outcome summarizes median percent change in 28-day seizure frequency within each arm over the specified baseline-to-Week-19 timeframe. It shows a median of -9.72 in the placebo arm, -30.52 in the Perampanel 8mg arm, and -17.57 in the Perampanel 12mg arm. The full ranges show substantial spread in the observed values, with both negative and positive values in each arm. These are descriptive results from the analysis population defined in the registry.

The results do not, on their own, establish the magnitude or precision of a perampanel-versus-placebo effect. The registry record does not provide an estimable two-arm comparison, confidence interval, or p-value for this primary endpoint. A more negative arm median is not equivalent to a randomized treatment-effect estimate with uncertainty. The posted values also cannot show whether differences among arms reflect sampling variation, differences in outcome distributions, or other features of the analysis.

The record does not report the statistical analysis plan, a sample-size justification, missing-data methods, or multiplicity procedures. Those details would be relevant to a fuller evaluation of the strength and robustness of the findings. The planning figures shown elsewhere on the page are illustrations based on stated assumptions and the observed effect; they are not observed power and do not describe the trial’s original design process.

The safety summaries show serious and other adverse-event counts by arm, as well as selected serious adverse-event terms. They are descriptive counts, not comparative safety estimates. Taken together, the registry supports a limited account of what was posted: arm-level endpoint summaries, outcome counts, and safety counts. It does not support claims beyond those data about clinical importance, certainty, or outcomes not reported in the record.

13. Why This Trial Matters Statistically

This trial illustrates why a registry record’s outcome summaries should be separated from a formal comparison. The primary endpoint is reported as a median and full range for each arm, which describes observed data but does not quantify the difference between treatment and placebo. Although the design is randomized and double-blind, the posted results do not include an estimable two-arm contrast or its uncertainty. Design features cannot replace missing inferential results.

The endpoint also highlights the distinction between a central summary and the spread of individual observations. The median gives a concise description, while the full range shows that observations extend widely around it. Neither alone describes the distribution in enough detail to assess the shape, precision, or statistical significance of a between-arm difference. The registry’s diary-based frequency measure and stated analysis population are essential context for understanding what the summaries cover.

The record further illustrates the importance of transparent methods reporting. It does not report the statistical plan, missing-data methods, multiplicity handling, or the trial’s own sample-size justification. Without these details, readers cannot fully evaluate how the analysis was specified or how robust it may be to analytic choices. Planning illustrations based on observed effects answer a different question from the original design rationale and should not be mistaken for observed power.

Finally, the adverse-event counts demonstrate why safety summaries also need careful interpretation. Counts by arm identify what was posted, but without additional detail they do not establish causation or a precise comparative risk. The statistical lesson is not a judgment about treatment, but the need to distinguish arm-level description from comparative evidence and to recognize what the registry does not report.

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.

The primary-outcome results are limited to arm-level medians and full ranges. The registry marks values as missing or not estimable for a two-arm comparison, and it does not report a confidence interval, p-value, or effect estimate for that comparison. As a result, the precision and statistical evidence for differences between arms cannot be evaluated from the posted primary outcome summaries alone.

The registry does not report the analysis plan, the statistical model, any interim analyses, or the trial’s sample-size justification. It also does not describe multiplicity procedures or how missing diary data were handled. The stated analysis population requires participants to have received study drug and to have some seizure-frequency data during the double-blind phase, so it is important not to equate these outcome summaries automatically with all enrolled participants.

The record’s description of the analysis population includes a note that 2 placebo participants were randomized but not treated. Its note about a randomized participant in the Perampanel 8mg arm is incomplete in the available text. Full participant flow and a detailed explanation of exclusions are not reported in the registry record.

For safety, the record provides counts and selected serious adverse-event terms, but detailed event-level context and death counts are not reported in the registry record. The registry summaries do not supply enough information to determine causality or make a precise comparative safety assessment. These limitations are specific to the information posted for this trial; they do not imply that unreported analyses or procedures did or did not occur.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: To Evaluate The Efficacy and Safety of E2007 (Perampanel) Given as Adjunctive Therapy in Subjects With Refractory Partial Seizures — NCT00699582.
• Registry record last updated 2014-07-11; status last verified 2012-10.
• 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 record describes a completed, randomized, double-blind, parallel-group Phase 3 study of adjunctive perampanel for refractory partial seizures. Enrollment was 389. The primary endpoint was median percent change in 28-day seizure frequency from baseline through Week 19, with the full range reported for each arm. The posted medians were -9.72 for placebo, -30.52 for Perampanel 8mg, and -17.57 for Perampanel 12mg.

The record does not provide an estimable two-arm comparison for the primary endpoint, and it does not report a confidence interval or p-value for that comparison. It reports serious and other adverse-event counts by arm, but further details needed for comparative interpretation are not reported in the registry record. The available record therefore supports descriptive reading of the posted summaries, with substantial limits on inference.