A Study for Evaluating the Efficacy and Safety of Zonisamide and Lamotrigine (Lamictal) for Subjects With Refractory Simple Partial, Complex Partial or Partial With Secondary Generalized Seizures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study for Evaluating the Efficacy and Safety of Zonisamide and Lamotrigine (Lamictal) for Subjects With Refractory Simple Partial, Complex Partial or Partial With Secondary Generalized Seizures (NCT00292461), a randomized phase 3 study. The primary outcome is The Percentage Change of Monthly Seizure Frequency at the End of the 16-week Treatment From Baseline.

2. Clinical Question

Statistically, this is a comparison of outcome distributions from a randomized, parallel-group trial, not a test of whether either arm changed from baseline in isolation. The reported primary summaries describe the observed percentage changes within each arm. Interpretation of the relative comparison also depends on the reported analysis and its assumptions, as well as on how missing outcomes and the ITT population were handled. Those details are not reported in the registry record.

3. Trial Design

The registry describes a Phase 3, randomized, parallel-group study with no masking. It lists zonisamide and lamotrigine as the interventions, both in tablet form, and reports enrollment of 64. The record states that the study is completed. Its listed start is 2006-03 and its primary completion is 2009-08. Eisai Inc. is identified as the sponsor.

The trial’s primary outcome was the percentage change in monthly seizure frequency at the end of the 16-week treatment period from baseline. The primary analysis population is described as ITT (intent-to-treat). The arm summaries list 24.0 for each group, with a median percentage change of -32.7 for Zonegran and -35.6 for lamotrigine. The reported full ranges were -100.0 to 180.0 and -100.0 to 444.4, respectively. These summaries describe the observed outcomes; they are not themselves an estimate of the difference between treatments.

The registry reports an ANOVA p-value of 0.240 for the primary comparison. It does not report a primary between-arm estimate or confidence interval, the detailed ANOVA specification, or the assumptions assessed for that analysis. Nor does it explain how the primary outcome was handled for participants without usable measurements. Randomization helps guard against systematic differences between assigned groups, but the record does not give the sequence-generation or allocation-concealment details. Masking is listed as none, so the registry does not describe masked treatment assignment or outcome assessment.

There are four outcome measures and four analyses listed overall, while three secondary outcomes are described in the registry summary. The record does not explain the relationship between those counts. Details of protocol amendments, sample-size justification, interim analyses, and a prespecified statistical analysis plan are not reported in the registry record. Those omissions limit how fully the design and analysis can be checked against the original intentions.

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 (intent-to-treat)

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 The Percentage Change of Monthly Seizure Frequency at the End of the 16-week Treatment From Baseline, measured over Baseline and 16 weeks.

The primary endpoint is percentage change in monthly seizure frequency from baseline to the end of the 16-week treatment period. The registry defines the calculation as (T-B)/B*100%, where T is total seizure frequency during the maintenance dose period divided by the maintenance dose period in weeks and multiplied by 4, and B is the monthly seizure frequency with one month prior to enrollment. Thus, the measure expresses the treatment-period frequency relative to the stated baseline frequency.

A negative percentage change represents a lower monthly frequency during the treatment period than at baseline; a positive value represents a higher frequency. The primary summaries use the median and full range, which describe the middle observed value and the span of observed values in each arm. A range can be wide when participants’ outcomes vary substantially, and it is sensitive to values at the extremes. The median gives a measure of location that is less influenced by extreme values than the mean, but it does not by itself describe the uncertainty in a between-arm comparison.

Monthly frequency puts seizure counts on a common time scale when the maintenance-period observation length can vary. The endpoint is used to summarize change from a pretreatment baseline over the stated treatment period. The registry record does not report whether additional rules were used for baseline values, incomplete observation periods, or missing measurements. Those details matter for interpreting percentage changes and cannot be inferred from the endpoint definition alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The following explains the statistical concepts relevant to the posted results in general, then relates them to what the registry reports. The registry identifies the trial as randomized and parallel-group, and names ANOVA for the primary comparison. It does not provide enough detail to reconstruct the analysis.

What does randomization contribute?

In general, randomization assigns participants to study groups by chance. Its purpose is to reduce systematic differences between groups at the point of assignment, including differences that investigators may not have measured. Randomization does not guarantee that the groups will be identical, especially in a limited trial, and it does not address bias arising after assignment, such as missing outcome data or differences in outcome assessment. The registry identifies the allocation as randomized but does not describe the sequence-generation or concealment methods.

What do the median and range describe?

In general, the median is the central value after observations are ordered. The full range gives the smallest and largest observed values. These are descriptive summaries, not a direct estimate of the uncertainty in the contrast between groups. The primary outcome is reported as a median with a full range. The ranges extend from -100.0 to 180.0 in the Zonegran arm and from -100.0 to 444.4 in the lamotrigine arm. The broad span indicates that individual percentage changes differed considerably. Ranges do not show how densely observations are distributed between their endpoints, and they are not confidence intervals.

How should the ANOVA result be understood?

In general, analysis of variance (ANOVA) tests whether group means are compatible with a common mean under the model used. Its validity depends on features of the outcome data and the analysis, including the model specification and the behavior of residuals. The registry names ANOVA and reports p = 0.240 for the primary comparison, but does not report whether the test was adjusted, what model terms were included, or what assumptions were checked. Because the descriptive endpoint statistic is the median, the record also does not clarify how the ANOVA result relates to the median summaries. No unreported analysis details should be assumed.

A p-value is a probability calculated under a null hypothesis and a statistical model; it is not the probability that the null hypothesis is true. A larger p-value does not establish equivalence or prove that groups are identical. To assess the size and precision of a treatment contrast, an effect estimate and confidence interval are useful. The registry does not report these for the primary comparison. General statistical interpretation therefore distinguishes the posted test result from conclusions about the magnitude of any difference.

What does a confidence interval add?

In general, a confidence interval describes a range of effect values compatible with the data and the method used, at a stated confidence level. Its width reflects precision: a wider interval allows a broader set of plausible effects. It is not a guarantee that the true value lies in that particular interval, nor does it describe the distribution of individual outcomes. The primary result has no posted effect interval, so its comparative precision cannot be judged from the p-value and arm ranges alone. For the secondary response-rate comparison, a recalculated interval is available, but it should be read as an unadjusted calculation rather than as a replacement for the sponsor’s analysis.

Why does the response-rate analysis use different methods?

The secondary response outcome is binary: a participant either met the defined response threshold or did not. In general, binary outcomes can be described as proportions and compared using a risk difference, risk ratio, odds ratio, or a test suited to the design. These measures answer related but different questions. A risk difference expresses the absolute contrast between proportions; a risk ratio compares their relative levels. The record reports a Cochran-Mantel-Haenszel analysis for the secondary comparison, but gives no stratification details. The recalculated unadjusted comparison uses the posted counts and is not the same analysis as a potentially stratified sponsor analysis.

Uncertainty matters even when the observed proportions match. A confidence interval for a difference can include effects in either direction when the data are imprecise. Therefore, an estimated difference of 0.0 does not establish that the underlying proportions are exactly equal. The response-rate interval extends from -0.282 to 0.282, and its width is relevant to the precision of that estimate. The registry also reports a sponsor p-value of 0.860 for that outcome, while the unadjusted recalculation gives p = 1.0. Different methods can yield different test results; the analysis specified by the study sponsor remains the reported trial analysis.

What does the sample-size illustration represent?

In general, a sample-size calculation estimates how many participants a future study might need under stated assumptions. The planning illustrations use the observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is a stated fraction of the observed effect. For non-time-to-event outcomes, the illustration is expressed as participants per arm for 80% power or another stated planning level. Such calculations are sensitive to the assumed effect and variability. They are planning illustrations, not observed power and not evidence that the completed trial had adequate or inadequate power. The registry does not report the trial’s own sample-size justification.

Finally, multiplicity and missing data affect interpretation across outcomes. Testing multiple endpoints can increase the chance of obtaining a nominally notable result by chance if no adjustment or prespecified hierarchy is used. The registry does not report a multiplicity strategy. Missing outcomes can also bias comparisons if their absence is related to outcome or treatment. The ITT label is reported for the primary population, but the record does not describe the analysis set in detail, imputation, or sensitivity analyses. General guidance about these issues cannot substitute for the missing study-specific methods.

10. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is an ANOVA with p = 0.240. The registry supplies arm-level medians and full ranges for the primary endpoint, but no primary between-arm estimate or confidence interval. The reported p-value should therefore be read as the sponsor’s statistical comparison, while the medians and ranges describe the arm outcomes. The record does not provide enough detail to reproduce or fully assess the ANOVA.

The primary descriptive statistic is the median, whereas ANOVA is generally associated with comparing means. The registry does not explain how the ANOVA was specified for this endpoint. That is a reporting limitation, not grounds to replace the sponsor’s analysis with an assumed alternative. No discrepancy is flagged in the record.

For the secondary response-rate outcome, the sponsor reports a Cochran-Mantel-Haenszel p-value of 0.860. A recalculated unadjusted risk difference is 0.0, with an interval from -0.282 to 0.282, and the unadjusted p-value is 1.0. These recalculations use the posted summaries and are not necessarily identical to the sponsor’s method. The registry does not describe the stratification used for the Cochran-Mantel-Haenszel analysis. The sponsor’s reported result is the trial analysis; a simple unadjusted recalculation is supplementary and should not be treated as a correction.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and three secondary outcomes, and reports four outcome measures and four analyses overall. It does not describe a prespecified hierarchy among the endpoints, a familywise error strategy, or another adjustment for multiple testing. These procedures are not reported in the registry record.

In general, when multiple outcomes or comparisons are tested, interpreting each p-value as if it were the only planned test can increase the chance of at least one apparently notable result arising by chance. A multiplicity adjustment or a prespecified testing sequence can address that issue, but the appropriate approach depends on the study’s objectives and analysis plan. Without such details, secondary results are best interpreted cautiously and should not be treated as independent confirmation of the primary result.

The primary population is identified as ITT (intent-to-treat). That label ordinarily refers to analyzing participants according to their randomized assignment, but it does not by itself describe which participants contributed outcome data or how missing outcomes were handled. The registry does not report the missing-data method, imputation, sensitivity analyses, or reasons for missing primary measurements. It also does not explain why the primary arm summaries list 24.0 in each group when total enrollment is 64. Those numbers should not be reconciled by assumption.

Missing outcome data can affect both precision and validity. In general, a complete-case analysis may be biased if the availability of measurements is related to treatment assignment or the outcome. Imputation and sensitivity analyses can address some concerns, but they depend on assumptions that should be described. Since the registry does not state which approach was used, it is not possible to assess how the primary analysis accounts for missingness or whether its conclusions are robust to alternative assumptions.

The registry also does not report the trial’s sample-size justification, interim-analysis plan, or full statistical analysis plan. These omissions make it difficult to assess whether the posted analyses follow prespecified decisions or how the study was intended to control error across outcomes. They are limitations of the registry record and do not establish that the trial team lacked such plans.

12. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: lumber contusion (3.3%), nystagmus (3.3%).

The serious adverse event summary lists 34 participants at risk in the Zonegran group and 30 in the lamotrigine group. It reports no serious adverse events among the Zonegran group and one among the lamotrigine group. The registry lists “Lumber contusion” and “nystagmus” among the top serious-event terms, each at 3.3%. These are registry safety summaries and should be read in the context of the stated groups at risk.

The record also reports other adverse events in 20 participants in each group, with 34 at risk for Zonegran and 30 at risk for lamotrigine. These summaries do not describe event severity, timing, duration, recurrence, or whether an event was considered related to treatment. The registry does not report deaths in the posted safety data.

Counts alone do not establish a comparative safety difference. The groups at risk are not identical, and the record does not provide a statistical comparison, exposure-adjusted rates, or additional context for the listed serious-event terms. Serious adverse events are distinct from the broader other-adverse-event counts. The registry’s figures describe what was recorded; they do not by themselves show causation or support a clinical recommendation.

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

The posted primary results describe median percentage changes in monthly seizure frequency from baseline in each treatment group. Both arm medians are negative, and the ranges include both negative and positive changes. The registry’s ANOVA result is p = 0.240. Together, these entries describe the available summaries and reported test, but they do not provide a primary between-arm effect estimate or a confidence interval.

Accordingly, the record does not quantify how large a comparative difference might be or how precisely it was estimated. A p-value alone cannot establish that the treatments are equivalent, that their effects are identical, or that either treatment has a particular clinical effect. Nor do within-arm changes alone isolate the effect of treatment from other influences over time. The trial’s randomized comparison is the relevant basis for a between-group question, but its primary comparative reporting is incomplete.

The secondary response-rate results offer another perspective, but the reported percentages are the same in the two arms and the recalculated interval remains compatible with differences in either direction. The sponsor’s analysis and the unadjusted recalculation use different methods, and the registry does not provide the details needed to explain the sponsor’s stratification. The global-assessment results are not reported in the registry record.

Safety counts are also descriptive. They do not, by themselves, establish whether an event was caused by treatment or whether one treatment is safer than the other. The record lacks details such as event timing and exposure-adjusted comparisons.

In general, sound interpretation weighs effect size, uncertainty, study design, analysis methods, missing data, and the clinical context together. Here, the registry supports a limited description of the recorded outcomes and sponsor analyses. It does not support a definitive claim of superiority, equivalence, or a treatment recommendation.

14. Why This Trial Matters Statistically

This trial is statistically useful as an example of the difference between descriptive summaries and comparative inference. The primary outcome is reported with medians and full ranges, while the named comparison is ANOVA. Because the registry does not explain how the analysis was specified, readers cannot fully connect the descriptive measure to the inferential test. That distinction matters whenever a reported p-value is read alongside summaries based on a different statistic.

The record also illustrates why an effect estimate and its uncertainty are central to interpretation. A p-value of 0.240 does not show the size of the primary between-arm contrast, and no primary confidence interval is posted. Without those quantities, readers cannot directly assess the range of comparative effects compatible with the data. The arm-level ranges describe participant outcomes, not uncertainty in the treatment contrast.

Reporting of analysis populations and missing data is another important issue. The primary population is labeled ITT, yet the relationship between enrollment and the arm-level endpoint counts is not explained. The record does not describe how missing outcomes were handled. These details affect confidence in whether the posted comparison reflects the intended randomized population.

Finally, the secondary response result demonstrates why recalculations should be distinguished from sponsor analyses. The unadjusted calculation and the sponsor’s Cochran-Mantel-Haenszel result are not the same method, and details of the sponsor’s analysis are absent. Clear reporting of prespecified methods, multiplicity, missing-data handling, effect estimates, and intervals would make the evidence easier to evaluate. These are general statistical reporting principles, not conclusions about unreported study procedures.

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 available registry record. The primary outcome is summarized by medians and full ranges, but the registry reports ANOVA for the comparison without describing its model or explaining how it relates to the median summaries. No primary between-arm estimate or confidence interval is posted, limiting assessment of comparative magnitude and precision.

The reported ranges are broad, particularly the upper endpoint of 444.4 for lamotrigine, but the record does not provide the distribution of individual outcomes or diagnostics. It is not possible to determine from the range alone whether extreme observations influenced the analysis, or whether assumptions were checked. The registry does not describe any transformation or alternative analysis.

The enrollment is 64, while the arm-level primary summaries list 24.0 for each treatment. The registry does not explain the difference, the exact analysis-set derivation, or the number and reasons for missing endpoint measurements. Although ITT is named as the primary population, the record does not describe missing-data handling or sensitivity analyses.

The allocation is identified as randomized, but sequence generation and allocation concealment are not described. Masking is listed as none. The record does not explain whether outcome assessment was standardized or whether the lack of masking could have affected assessment. No dosing or treatment-duration details beyond the stated 16-week endpoint timeframe are reported in the registry record.

Secondary results are incomplete: numerical results for the physician and participant global assessments are not reported in the registry record. The multiplicity approach, sample-size justification, interim analyses, and full statistical analysis plan are also not reported in the registry record. These omissions restrict independent evaluation but do not show what procedures may have been documented elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study for Evaluating the Efficacy and Safety of Zonisamide and Lamotrigine (Lamictal) for Subjects With Refractory Simple Partial, Complex Partial or Partial With Secondary Generalized Seizures — NCT00292461.
• Registry record last updated 2013-05-20; status last verified 2009-08.
• Results first posted 2011-08-15.
• 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 NCT00292461 as a completed, randomized, Phase 3, parallel-group study with no masking, sponsored by Eisai Inc. Enrollment was 64. The study compared zonisamide, listed as Zonegran, with lamotrigine in subjects with refractory partial seizures.

The primary endpoint was percentage change in monthly seizure frequency from baseline to the end of the 16-week treatment period. The reported medians were -32.7 for Zonegran and -35.6 for lamotrigine, with full ranges of -100.0 to 180.0 and -100.0 to 444.4, respectively. The sponsor-reported ANOVA p-value was 0.240. A primary comparative effect estimate and confidence interval are not reported in the registry record.

The available secondary numerical result concerns response rate; the global-assessment results are not reported in the registry record. Safety summaries include serious and other adverse-event counts, but provide limited event context. Missing-data methods, multiplicity procedures, the sample-size justification, and detailed analysis methods are not reported in the registry record.