A Placebo-controlled Study of Efficacy & Safety of 2 Trough-ranges of Everolimus as Adjunctive Therapy in Patients With Tuberous Sclerosis Complex (TSC) & Refractory Partial-onset Seizures: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Placebo-controlled Study of Efficacy & Safety of 2 Trough-ranges of Everolimus as Adjunctive Therapy in Patients With Tuberous Sclerosis Complex (TSC) & Refractory Partial-onset Seizures (NCT01713946), a randomized phase 3 study. The primary outcome is Core Phase: European Medicine Agency (EMA): Seizure Frequency Response Rate.

2. Clinical Question

Statistically, the primary outcome is a binary response: each participant either met the stated reduction threshold during the specified period or did not. The comparison therefore concerns group response percentages, rather than the average number of seizures or the size of each participant’s individual change. The registry record describes the study as Phase 3, randomized, quadruple-masked, and parallel-group. Its listed primary outcome is an EMA seizure-frequency response rate with 95% confidence intervals. The results report each target group and placebo separately, as well as sponsor-reported comparisons with placebo. These results address the specified response definition and analysis period; they do not, by themselves, describe every aspect of seizure burden, quality of life, safety, or outcomes during later study phases.

3. Trial Design

The registry record describes a Phase 3, randomized, quadruple-masked, parallel-group, placebo-controlled study sponsored by Novartis Pharmaceuticals. It lists the study as completed, with a start date of 2013-04-29 and a primary completion date of 2017-10-25. Enrollment is listed as 366. The study evaluated two everolimus trough targets as adjunctive therapy for patients with tuberous sclerosis complex who had refractory partial-onset seizures.

The randomized groups were assigned to everolimus with a low trough target of 3 to 7 ng/mL, everolimus with a high trough target of 9 to 15 ng/mL, or placebo. Each arm description includes 1 to 3 antiepileptic drugs alongside the assigned study treatment. The registry record describes the primary analysis population as the Full Analysis Set, comprising all patients assigned study treatment by randomization.

The study description outlines a baseline phase from Screening Week -8 to randomization at Week 0, a core phase from randomization to Week 18, an extension phase lasting until 48 weeks after the last patient completed the core phase, and a post-extension phase. The primary outcome timeframe is the baseline 8-week period before randomization and the Week 7 to 18 maintenance period of the core phase. The outcome was the percentage meeting a threshold of at least a 50% reduction in partial-onset seizure frequency.

These design details establish the randomized comparison and the period to which the main result applies. They do not establish how the original sample size was chosen, what data-monitoring procedures were used, or how every outcome and missing observation was handled. Those details are not reported in the registry record.

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: The Full Analysis Set (FAS) comprised all patients to whom study treatment was assigned by randomization

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Full Analysis Set: The Full Analysis Set (FAS) comprised all patients to whom study treatment was assigned by randomization.

6. Primary Endpoint

The registry lists the primary outcome as Core Phase: European Medicine Agency (EMA): Seizure Frequency Response Rate, measured over Baseline (8-week period before randomization), Week 7 to 18 (12-week maintenance period of the core phase).

The primary endpoint is the percentage of participants who met the study’s seizure-frequency response definition during the maintenance period of the core phase. A response meant at least a 50% reduction from baseline in partial-onset seizure frequency. The registry record defines baseline as the 8-week period before randomization and specifies Week 7 to 18 as the 12-week maintenance period.

This binary endpoint classifies each participant according to whether the threshold was reached. Group percentages provide a direct description of how often the defined response occurred in each randomized arm. They do not show the full range of individual seizure-frequency changes: participants just below the threshold and participants with little or no reduction are both classified as not meeting the response criterion, while the size of reduction among responders is not represented by the response percentage alone.

In general, a threshold-based endpoint can make a comparison straightforward to communicate, but the selected threshold and observation window determine what the result means. In this study, the result pertains to the specified reduction threshold and the core-phase maintenance period. It is not a measure of seizure freedom, because the primary definition requires a reduction from baseline rather than no seizures. Seizure-free status appears as a separate secondary outcome in the registry record.

The registry record reports percentages and 95% confidence intervals for the arms. Those intervals communicate uncertainty around the estimated response percentages; they do not change the endpoint definition or establish what occurred beyond the specified period.

7. Results Posted in the Registry

• Everolimus LT Target of 3 to 7 ng/mL: 28.2% (33/117) versus 15.1% (18/119) on control; risk difference 13.1 percentage points (95% CI 2.7 to 23.5), risk ratio 1.86 (1.11–3.12); unadjusted two-proportion z-test p = 0.015.
• Everolimus HT Target of 9 to 15 ng/mL: 40.0% (52/130) versus 15.1% (18/119) on control; risk difference 24.9 percentage points (95% CI 14.3 to 35.5), risk ratio 2.64 (1.64–4.25); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Core Phase: Percentage of Patients With at Least a 25% Reduction in Seizure Frequency

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Everolimus LT Target of 3 to 7 ng/mL: 52.1% (61/117) versus 37.8% (45/119) on control; risk difference 14.3 percentage points (95% CI 1.8 to 26.9), risk ratio 1.38 (1.03–1.84); unadjusted two-proportion z-test p = 0.027.
  • Everolimus HT Target of 9 to 15 ng/mL: 70.0% (91/130) versus 37.8% (45/119) on control; risk difference 32.2 percentage points (95% CI 20.4 to 43.9), risk ratio 1.85 (1.43–2.39); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Everolimus LT Target of 3 to 7 ng/mL than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Everolimus HT Target of 9 to 15 ng/mL than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry record lists a range of secondary outcomes, including seizure-free status, threshold reductions in seizure frequency, distributions of seizure-frequency change, seizure-free days, time on treatment, and age-specific quality-of-life measures. It also lists 20 posted secondary outcomes. The detailed secondary results available for interpretation include the percentage meeting a threshold of at least a 25% reduction in seizure frequency during the core-phase maintenance period.

For that posted threshold outcome, the arm percentages are 52.1 for low-target everolimus, 70.0 for high-target everolimus, and 37.8 for placebo. The recalculated risk differences versus placebo are 0.143 and 0.322, with unadjusted p-values of 0.027 and p < 0.001. The registry also reports odds ratios of 1.77 and 3.82 for the respective comparisons, with confidence intervals but no p-values shown. This pattern is directionally consistent with the primary response percentages, but the secondary threshold is different from the primary threshold.

That consistency is a supporting description, not an independent replacement for the primary result. Secondary endpoints can differ in definition, analysis, and completeness of reporting. The record does not give a multiplicity method for these reported secondary comparisons, so their p-values should not be treated as if they necessarily controlled the overall chance of false-positive findings across all outcomes and comparisons. Results from numerous secondary analyses generally carry less confirmatory weight than the prespecified primary endpoint, particularly when adjustment procedures are unreported.

Although many secondary outcomes are listed, their presence alone does not establish their results or a uniform pattern across them. Where detailed results or analysis information are not given, those details are not reported in the registry record. The primary result remains specific to its own response threshold, analysis, and maintenance period.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Percentage of Seizure-free Patients During the Maintenance Period of the Core Phase
• Core Phase: Percentage of Patients With at Least a 25% Reduction in Seizure Frequency
• Core Phase: Distribution of Reduction From Baseline in Seizure Frequency
• Core Phase: Changes From Baseline in Number of Seizure-free Days
• Core Phase: Probability That a Patient Remains On-treatment up to a Specified Time Point
• Core Phase: Change From Baseline in the QOLCE Overall Quality-of-life Score for Patients <11 Years
• Core Phase: Change From Baseline in the QOLIE-AD-48 Overall Quality-of-life Score for Patients >=11 to 18 Years
• Core Phase: Change From Baseline in the QOLIE-31-P Overall Quality-of-life Score for Patients Aged >=18 Years

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a binary percentage: participants are counted according to whether they met the response definition. The definition is at least a 50% reduction in partial-onset seizure frequency from baseline during the maintenance period. This type of outcome is different from a continuous measure such as the mean change in seizure frequency. It records whether a threshold was reached, not the exact amount of change for every participant.

What do the arm percentages describe?

Each arm percentage estimates the proportion of participants in that group who met the response definition. The posted results give the low-target group as 28.2, the high-target group as 40.0, and placebo as 15.1. Each estimate has a 95% confidence interval. In general, a confidence interval is a range of values compatible with the data and the statistical method used, subject to the method’s assumptions. It communicates precision: a narrower interval generally means a more precise estimate, while a wider interval indicates greater uncertainty. An interval is not a guarantee that the true value lies within its endpoints.

What is a risk difference?

A risk difference compares the response proportions in two groups by subtracting one group’s proportion from the other’s. The recalculated results report risk differences of 0.131 for low-target everolimus versus placebo and 0.249 for high-target everolimus versus placebo. Their respective 95% confidence intervals are 0.027 to 0.235 and 0.143 to 0.355. These are unadjusted re-calculations from the posted summary statistics. They are not the sponsor’s primary reported comparison estimates.

General interpretation of a risk difference depends on the group ordering and the outcome definition. In this comparison, the stated direction is everolimus arm versus placebo, and the endpoint is meeting the response threshold. The estimate describes the difference between group response proportions for the specified endpoint. It does not describe the size of the reduction among responders or the outcome for any particular participant.

What is an odds ratio?

An odds ratio compares the odds of meeting the response definition between groups. It is a relative measure and is not the same as a risk difference or a ratio of response percentages. The sponsor reported odds ratios of 2.21 for the low-target comparison and 3.93 for the high-target comparison. Those estimates have 95% confidence intervals and p-values, and the registry record identifies the comparisons as using the Bonferroni-Holm method. Because the registry record flags a discrepancy between the sponsor analysis and the simple recalculation, the sponsor’s own analysis takes precedence.

How should p-values be read?

A p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the statistical procedure. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The sponsor-reported p-values are p = 0.008 for low-target everolimus versus placebo and p < 0.001 for high-target everolimus versus placebo. The recalculated p-values are unadjusted; the sponsor’s analyses, not the unadjusted recalculations, are the relevant reported comparisons.

Why can multiplicity matter?

When several hypotheses are tested, the chance of obtaining at least one apparently notable result can increase if each test is assessed without accounting for the others. General multiplicity procedures adjust how evidence is assessed across a family of comparisons. The sponsor’s primary comparisons are reported with Bonferroni-Holm. The registry record does not report the full definition of the hypothesis family or the complete analysis plan. The method listed for a primary comparison should therefore not be assumed to describe every secondary outcome or every analysis in the study.

What can the posted group counts establish?

The primary Full Analysis Set is defined in the registry record as all patients assigned treatment by randomization. The posted arm counts and response percentages describe the reported analysis. General statistical interpretation also depends on how missing outcome observations, treatment discontinuations, and protocol deviations were handled. Those details are not reported in the registry record. The stated analysis population should not be taken to imply a particular missing-data method.

Finally, the sample-size planning figures are illustrations based on observed effects, not a report of the original trial design. They use the assumptions stated in the registry record and should be read as planning scenarios rather than as evidence about the trial’s achieved power. The original design assumptions are not reported in the registry record.

11. Confidence Intervals and Precision

The 95% confidence intervals show the uncertainty around the reported primary response percentages. For low-target everolimus, the interval is 20.3 to 37.3 around the estimate of 28.2. For high-target everolimus, it is 31.5 to 49.0 around 40.0. For placebo, it is 9.2 to 22.8 around 15.1. These intervals are estimates for the arm-level response percentages, not prediction intervals for individual participants.

The recalculated risk-difference intervals are 0.027 to 0.235 for low-target everolimus versus placebo and 0.143 to 0.355 for high-target everolimus versus placebo. The sponsor-reported odds-ratio intervals are 1.16 to 4.20 and 2.10 to 7.32, respectively. These intervals concern different statistical measures and are not directly comparable by their numerical widths. The sponsor’s odds-ratio analyses take precedence over the recalculated summaries because the registry record flags a discrepancy.

In general, confidence intervals help distinguish the point estimate from the range of uncertainty around it. They do not account for every possible source of uncertainty, such as limitations in the study design or unreported analysis choices. Nor do they show the probability that a specific value is true. Precision should be considered together with the endpoint definition, the randomized comparison, the analysis method, and the limits of the registry record.

The widths of the reported intervals indicate that the estimates are not exact. A single response percentage should therefore not be treated as a perfectly precise description of the underlying group response rate. The intervals also do not establish that outcomes beyond the core-phase maintenance period would have the same pattern.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations, not a statement of the original trial’s design rationale. The registry record states that the planning calculation uses the observed effect, a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. For these binary outcomes, the listed figures are participants per arm, not total events.

The 80% power column gives the planned number of participants per arm under the stated assumptions for a design with 80% power to detect the specified effect. In general, power is the probability of detecting an effect of the specified size under the assumptions of the calculation, if that effect is present. The 90% power column gives the corresponding planning figure for a design with 90% power under those assumptions. Neither column describes the probability that the study result is correct.

The rows labeled 100%, 75%, and 50% of the observed effect represent planning scenarios that assume progressively smaller effects than the observed effect. Smaller effects are harder to distinguish from chance variation, so they require more participants per arm under the same planning assumptions. The entries are illustrations of that relationship and are not evidence that any particular effect size will recur in another study.

These values are not observed power. They do not revise or establish the trial’s actual sample-size justification, which is not reported in the registry record. The original design assumptions and the planning illustration are distinct: the latter is a recalculation based on the posted observed effect and stated planning assumptions.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.151; treatment proportion p₁ = 0.282; average p̄ = 0.217.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5826 = 1.1419.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5752 = 0.4841.
4. n = (1.1419 + 0.4841)² / (0.1308)² = 154.6, rounded up to 155 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.151260504201681                          # control proportion
p1 <- 0.282051282051282                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1713946)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 155 (80% power), 207 (90% power)
 75% of observed effect: n per arm = 261 (80% power), 348 (90% power)
 50% of observed effect: n per arm = 550 (80% power), 736 (90% power)
power.prop.test: 154.6 
Simulated power at n = 155 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The primary comparisons reported by the sponsor are odds ratios analyzed with the Bonferroni-Holm method. For low-target everolimus versus placebo, the sponsor reports an odds ratio of 2.21, a 95% confidence interval from 1.16 to 4.20, and p = 0.008. For high-target everolimus versus placebo, the sponsor reports an odds ratio of 3.93, a 95% confidence interval from 2.10 to 7.32, and p < 0.001. Both are labeled superiority analyses in the registry record.

The simple recalculations use posted summary statistics to estimate unadjusted risk differences and their intervals and p-values. They are on a different scale from the sponsor’s odds ratios and do not reproduce the sponsor’s complete analysis. The registry record flags a discrepancy, so the sponsor’s own analysis takes precedence. The recalculated figures can help describe the observed data on a risk-difference scale, but they should not replace the sponsor-reported primary inferential results.

The reported estimates should be read with their confidence intervals and the endpoint definition. The p-values are not measures of effect size, and agreement in the general direction of the arm percentages does not erase differences between analysis methods or scales. The registry record does not explain the source of the discrepancy, so no cause should be inferred.

15. Multiplicity, Missing Data and Analysis Populations

The sponsor’s primary comparisons are reported using the Bonferroni-Holm method. In general, this procedure addresses multiplicity by adjusting how evidence is evaluated across a defined set of tests. The registry record does not specify the full family of hypotheses to which this adjustment was applied, so the method should not be assumed to cover every outcome, subgroup, or analysis in the study.

The registry record lists 22 outcome measures and 18 analyses. It also lists 20 posted secondary outcomes. These counts indicate that the record includes multiple measures and analyses, but do not by themselves show which were prespecified, how they were prioritized, or whether all were included in a common multiplicity adjustment. Those details are not reported in the registry record. The secondary comparison entries do not identify an adjustment method or provide p-values for the threshold-reduction outcome.

The primary analysis population is described as the Full Analysis Set, comprising all patients assigned study treatment by randomization. That definition is useful for understanding who the primary analysis is intended to include. It does not state how missing seizure-frequency data were handled, whether participants with incomplete follow-up were included in the endpoint calculation, or what assumptions were used for any missing observations. These details are not reported in the registry record.

Likewise, the record does not report the handling of treatment discontinuations, protocol deviations, or other departures from the intended analysis. It does not report the trial’s original sample-size justification, interim analyses, or complete statistical analysis plan. The posted planning illustrations use stated recalculation assumptions, but those assumptions do not fill in missing details about the trial’s original design or analysis. Interpretation should therefore distinguish the reported primary comparison from additional analyses for which adjustment and missing-data methods are unclear.

16. 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: pneumonia (13.7%), seizure (6.0%), status epilepticus (6.0%), pyrexia (4.3%), gastroenteritis (4.3%).

The safety figures are reported for groups labeled low-target everolimus, high-target everolimus, Everolimus Start Ext, and Everolimus All. In the low-target group, the registry lists 50 serious adverse events among 117 at risk and 110 other adverse events among 117 at risk; deaths are listed as 1. In the high-target group, it lists 49 serious adverse events among 130 at risk and 126 other adverse events among 130 at risk; deaths are listed as 2.

The Everolimus Start Ext group has 38 serious adverse events among 114 at risk, 109 other adverse events among 114 at risk, and 1 death. The Everolimus All group has 137 serious adverse events among 361 at risk, 345 other adverse events among 361 at risk, and 4 deaths. The record lists pneumonia as the most frequent serious adverse event term at 13.7%, followed by seizure and status epilepticus at 6.0% each, pyrexia and gastroenteritis at 4.3% each, and tonsillitis at 2.6%.

These are safety counts and listed event terms, not a randomized comparison of risk between treatment and placebo. A placebo safety-group summary is not reported in the registry record. The group labels and at-risk denominators should be retained when interpreting the figures; the Everolimus All group is not a separate randomized target arm. These summaries alone do not establish causation or provide a complete account of severity, timing, or individual outcomes.

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

The posted primary results describe how often participants in each randomized group met a specific seizure-frequency response threshold during a defined part of the core phase. The observed response percentages are 28.2 for low-target everolimus, 40.0 for high-target everolimus, and 15.1 for placebo. The sponsor-reported comparisons against placebo use odds ratios and the Bonferroni-Holm method, with both comparisons labeled superiority in the registry record. These statistics characterize the group results for the stated analysis.

They do not mean that every participant in an everolimus group had the response, or that a particular participant would have the same outcome. A response meant at least a 50% reduction from baseline in partial-onset seizure frequency; it did not mean seizure freedom. The primary endpoint also does not describe the exact amount of reduction for every participant, changes after the maintenance period, or outcomes in populations not described by the study.

The result does not by itself establish an overall balance of benefit and harm. The registry record contains safety summaries, but the listed safety information is not a full randomized comparison against placebo. Nor do the primary results answer the secondary questions about quality of life, seizure-free days, time on treatment, or longer-term outcomes. Each requires its own complete results and appropriate interpretation.

Confidence intervals convey uncertainty in estimates, and the sponsor-reported p-values describe evidence under the specified analysis. Neither is a measure of clinical importance or a guarantee about future results. The discrepancy flag also means the simple unadjusted recalculation should not replace the sponsor’s primary analysis. In general, a clinical trial result applies most directly to its studied population, endpoint definition, comparator, and observation period. It should not be extended beyond those boundaries without additional evidence.

18. Why This Trial Matters Statistically

This trial offers a useful example of how a three-arm randomized comparison can distinguish two target ranges of an adjunctive treatment from placebo while keeping the primary endpoint definition explicit. The outcome is not simply seizure frequency in the abstract; it is a binary response based on at least a 50% reduction from baseline during the maintenance period of the core phase. That definition is central to interpreting both the percentages and the comparisons.

It also illustrates why the statistical scale matters. The posted arm percentages communicate observed response frequency. Risk differences describe absolute differences between proportions, while odds ratios express a relative contrast in odds. Those measures answer related but non-identical questions, and their values should not be treated as interchangeable. The registry record’s discrepancy flag makes it especially important to distinguish the unadjusted recalculation from the sponsor-reported Bonferroni-Holm analysis.

The study record also highlights the role of uncertainty and multiplicity. Confidence intervals qualify the apparent precision of each estimate, while adjustment procedures matter when testing several comparisons. A listed method for primary analyses does not automatically clarify how all secondary outcomes were handled. Clear reporting of analysis populations, missing-data methods, and prespecified hypotheses is necessary for a complete statistical assessment; several such details are not reported in the registry record.

Finally, the posted planning illustrations show how required sample sizes change under different assumed effect sizes, while remaining separate from the trial’s original design assumptions. Together, these features make the record useful for examining endpoint choice, comparison scale, inferential procedures, and the limits of what registry results alone can establish.

19. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific analytical details unclear. Although it defines the Full Analysis Set for the primary outcome, the handling of missing seizure-frequency observations is not reported in the registry record. The record also does not describe how treatment discontinuations or protocol deviations were addressed in the primary analysis. These omissions limit assessment of how the reported percentages relate to all randomized participants.

The record flags a discrepancy between sponsor-reported primary analyses and the simple unadjusted recalculations, but does not explain its source. The sponsor’s analysis takes precedence; without an explanation, the recalculated risk differences and p-values should be understood as descriptive, unadjusted alternatives rather than a resolution of the discrepancy.

The registry lists many outcomes and analyses, but does not report the full statistical analysis plan, the complete hypothesis family for multiplicity control, or the adjustment method for the secondary outcome comparisons described. The original sample-size justification and any interim analyses are not reported in the registry record. The planning illustrations therefore should not be used to infer the trial’s original design choices.

The safety information lists event counts and selected serious adverse event terms, but a placebo safety-group summary is not reported in the registry record. The record also does not supply enough information to use these summaries as a full comparative safety analysis. Finally, the primary endpoint concerns a threshold response during the core-phase maintenance period; it does not alone characterize longer-term outcomes, seizure-free status, quality of life, or the distribution of seizure-frequency changes. Those limitations define the scope of conclusions that can be drawn from the posted primary result.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Placebo-controlled Study of Efficacy & Safety of 2 Trough-ranges of Everolimus as Adjunctive Therapy in Patients With Tuberous Sclerosis Complex (TSC) & Refractory Partial-onset Seizures — NCT01713946.
• Registry record last updated 2018-11-07; status last verified 2018-11.
• Results first posted 2018-11-07.
• 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.

23. Record Summary

The registry record describes a completed Phase 3, randomized, quadruple-masked, parallel-group, placebo-controlled study sponsored by Novartis Pharmaceuticals. It evaluated low- and high-trough everolimus targets as adjunctive therapy, compared with placebo, in patients with tuberous sclerosis complex-associated refractory partial-onset seizures. Enrollment is listed as 366.

The primary outcome was the percentage meeting a definition of at least a 50% reduction in partial-onset seizure frequency during the core-phase maintenance period. The posted response percentages were 28.2 for low-target everolimus, 40.0 for high-target everolimus, and 15.1 for placebo. Sponsor-reported comparisons used odds ratios and Bonferroni-Holm, and the record flags a discrepancy with simple recalculations; the sponsor’s analyses take precedence.

The record lists additional seizure, treatment-duration, and quality-of-life outcomes, along with safety summaries. Important details about missing-data handling, the complete analysis plan, and the original sample-size justification are not reported in the registry record.