Erenumab - Comprehensive Assessment of Efficacy in (High-Frequency) Episodic Migraine: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Erenumab - Comprehensive Assessment of Efficacy in (High-Frequency) Episodic Migraine (NCT04252742), a randomized phase 4 study. The primary outcome is Change From Baseline in Mean Monthly Hours of at Least Moderate Headache Pain Intensity Over Months 1, 2, and 3.

2. Clinical Question

3. Trial Design

This was a phase 4, randomized, double-blind, parallel-group, placebo-controlled study. The official title describes it as global and identifies the intended population as people with high-frequency episodic migraine and at least one previously failed preventive treatment. The registry reports an enrollment of 512 and a completed study. The recorded study start date is 2020-09-15, and the primary completion date is 2023-10-26.

The comparison was between erenumab 140 mg SC Q4W and placebo. The registry describes a four-month double-blind treatment period with two parts: the main double-blind treatment phase, covering months 1 to 3, and an exploratory phase in month 4. The stated primary objective focuses on treatment benefit for headache duration at at least moderate pain intensity. Other listed study assessments include migraine functional impact, headache duration, and peak migraine pain intensity.

These design details do not establish how participants were randomized, how allocation was concealed, or what criteria governed eligibility beyond the population wording in the title. Those details are not reported in the registry record. The original sample-size justification, planned interim analyses, and full analysis plan are also not reported in the registry record. The counts used in the primary efficacy analysis are 251 for placebo and 253 for erenumab; these are the analysis-set counts, not a statement of the randomized count in each arm.

The main efficacy analysis set consisted of a subset of the full analysis set. It required at least one dose of investigational product, an observed baseline value, and at least one observed measurement during the main treatment phase. This distinction matters when interpreting the number of participants contributing to the primary outcome.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Eligible adult participants with high frequency episodic migraine (EP) who met specific eligibility criteria during a 2-week run-in period and 4-week baseline period entered the double-blind treatment period (DBTP). The DBTP included a 12-week main-DBTP (M-DBTP) and a 4-week exploratory DBTP (E-DBTP)

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The M-DBTP efficacy analysis set consisted of a subset the full analysis set (all participants randomized in the study) who received at least 1 dose of IP and had observed monthly hours of at least moderate headache pain intensity at baseline and at least 1 measurement during the M-DBTP.

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: The full analysis set included all participants who were randomized in the study.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Mean Monthly Hours of at Least Moderate Headache Pain Intensity Over Months 1, 2, and 3, measured over Baseline, Month 1, Month 2, and Month 3.

The primary endpoint is change from baseline in mean monthly hours of at least moderate headache pain intensity over months 1, 2, and 3. Headache intensity was recorded using a three-level scale, and moderate or severe ratings counted toward the endpoint. The measure therefore combines intensity classification with the duration of qualifying headache pain. It is expressed in hours per month.

The registry defines the monthly change as the difference from baseline for each month considered. The endpoint’s mean is based on the monthly changes with observed data, provided there was at least one observed monthly value. The primary analysis reports least-squares means, which are model-based estimates rather than simple unadjusted averages. A negative change indicates a reduction in mean monthly hours of at least moderate headache pain intensity from baseline.

This endpoint is relevant to the study’s stated primary objective because it captures time spent with headache pain at a specified intensity, rather than counting headache episodes alone. The listed secondary outcomes broaden the assessment to include functional impact and other pain characteristics. Whether the endpoint captures every aspect of headache burden is a broader measurement question; the registry record does not report a separate justification comparing this endpoint with alternative primary measures.

7. Results Posted in the Registry

• Erenumab 140 mg SC Q4W: mean -31.33 versus -23.38 on control; difference -7.95 (95% CI -11.43 to -4.47), standardized effect d = -0.40; 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.

Change From Baseline in Mean Monthly Physical Function Domain Score as Measured by the Migraine Functional Impact Questionnaire (MFIQ) Over Months 1, 2, and 3

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Erenumab 140 mg SC Q4W: mean -30.28 versus -22.92 on control; difference -7.36 (95% CI -10.80 to -3.92), standardized effect d = -0.38; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Erenumab 140 mg SC Q4W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Mean Monthly Usual Activities Domain Score as Measured by the MFIQ Over Months 1, 2, and 3

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Erenumab 140 mg SC Q4W: mean -31.08 versus -23.98 on control; difference -7.10 (95% CI -10.33 to -3.87), standardized effect d = -0.39; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Erenumab 140 mg SC Q4W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Mean Monthly Emotional Functioning Domain Score as Measured by the MFIQ Over Months 1, 2, and 3

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Erenumab 140 mg SC Q4W: mean -29.83 versus -22.77 on control; difference -7.06 (95% CI -10.76 to -3.36), standardized effect d = -0.34; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Erenumab 140 mg SC Q4W than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists six secondary outcomes, covering physical function, usual activities, emotional and social functioning, duration of headache pain, and peak migraine pain intensity. The detailed posted analyses include results for physical function, usual activities, and emotional functioning. For each of these reported MFIQ domains, the sponsor’s model-based comparison is negative, the 95% confidence interval is below zero, and the reported p-value is < 0.001. The broad pattern is therefore directionally consistent with the primary comparison among the secondary analyses with detailed results.

These secondary measures address related but distinct aspects of migraine impact. The primary outcome measures monthly hours with at least moderate headache pain, while the MFIQ domains concern reported functional impact. Consistency across related measures may add context to the primary finding, but it does not make the outcomes interchangeable. The listed social functioning, headache-duration, and peak-pain outcomes do not have detailed numerical results in the registry record included for this analysis.

The registry does not report whether multiplicity was controlled across the outcome family, or whether a hierarchy or other confirmatory testing strategy was specified. When many outcomes are tested, the chance of at least one apparently notable result can increase unless the analysis plan accounts for multiple testing. For that reason, secondary p-values without a reported multiplicity strategy carry less confirmatory weight than a prespecified primary result. The directionally similar posted secondary analyses can be described, but should not be treated as independent confirmation free from the uncertainty created by multiple outcomes.

The secondary results also should not be used to infer outcomes that were not measured or to establish persistence beyond the observed assessment period. The registry describes the main treatment phase as months 1 to 3; the detailed secondary estimates correspond to that period. Conclusions about the month 4 exploratory phase are not supported by the detailed secondary results described in the record.

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

• Change From Baseline in Mean Monthly Physical Function Domain Score as Measured by the Migraine Functional Impact Questionnaire (MFIQ) Over Months 1, 2, and 3
• Change From Baseline in Mean Monthly Usual Activities Domain Score as Measured by the MFIQ Over Months 1, 2, and 3
• Change From Baseline in Mean Monthly Emotional Functioning Domain Score as Measured by the MFIQ Over Months 1, 2, and 3
• Change From Baseline in Mean Monthly Social Functioning Domain Score as Measured by the MFIQ Over Months 1, 2, and 3
• Change From Baseline in Mean Monthly Average Duration of at Least Moderate Headache Pain Intensity in Migraine Attacks Occurring Over Months 1, 2, and 3
• Change From Baseline in Mean Monthly Average Peak Migraine Pain Intensity as Assessed by the 11-point Numeric Rating Scale (NRS) Over Months 1, 2, and 3

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it measures hours per month, and the reported treatment effects are means and mean differences. A mean gives a summary of the observed or estimated values across participants, while also concealing variation between individuals. The registry reports least-squares means with standard errors for each arm. Least-squares means are estimates from a statistical model that accounts for the covariates and repeated measurements specified in that model; they are not necessarily identical to raw averages.

The registry identifies a linear mixed model for the primary analysis. The listed fixed terms are treatment, visit, the treatment-by-visit interaction, and baseline value. Treatment represents the comparison between study groups, visit represents the assessment occasions, and the interaction permits the treatment contrast to vary across visits. Baseline value is included as a covariate. The registry says the model used a first-order autoregressive covariance structure. In general, a covariance structure describes how repeated measurements from the same participant are related. An autoregressive structure represents a pattern in which the association between measurements depends on their order or separation; the registry does not provide further detail about the implementation.

Repeated monthly measurements on the same participant are not independent observations. A mixed model can account for that within-participant structure while estimating treatment and visit effects. It also allows the analysis to combine available repeated measurements under the chosen model. The registry specifies the model and describes which monthly values contribute to the endpoint, but the detailed handling of missing measurements and the assumptions used for that handling are not reported in the registry record. It would therefore be unwarranted to infer a particular missing-data procedure or claim that missingness had no effect.

The primary endpoint averages monthly changes from baseline over months 1, 2, and 3 when observed values are available, with at least one observed monthly value required. The efficacy population also required a baseline measurement and at least one measurement during the main phase, in addition to receipt of at least one dose. These requirements define the analysis set used for the posted efficacy result. The registry identifies the full analysis set as all randomized participants, then describes the efficacy set as a subset of that group. The exact reasons individual participants did not enter the efficacy set are not reported in the registry record.

A standard error describes the estimated uncertainty of a mean estimate under the analysis model. It is not a measure of the spread of individual participants’ outcomes. The reported primary arm standard errors are 1.26 for placebo and 1.25 for erenumab. A confidence interval for the between-group effect expresses a range of values compatible with the data and the model at the stated confidence level. The sponsor’s 95% interval is from -11.45 to -4.46 hours/month. The interval addresses sampling uncertainty under the analysis; it does not capture every possible source of uncertainty, such as uncertainty about the model or the generalizability of the study population.

The sponsor reports a mixed-model least-squares mean difference, while the unadjusted recalculation uses posted arm-level means and standard errors to approximate a comparison. Such a recalculation does not reproduce the participant-level repeated-measures model. It is useful as a check on the broad direction and scale of the result, but the sponsor’s model-based analysis is the relevant estimate for the stated primary analysis. Here the recalculated point estimate agrees with the sponsor’s point estimate, while the confidence limits differ slightly.

The reported p-value is a measure of how incompatible the observed data are with a specified null model, under the assumptions of the analysis. The primary sponsor-reported p-value is p < 0.001. A small p-value does not quantify the size or practical importance of an effect, and it is not the probability that the null hypothesis is true. The estimate and its confidence interval provide information about direction, magnitude, and precision alongside the p-value.

Over-dispersion is a general concern when data vary more than a statistical model expects. If present and not addressed, uncertainty estimates can be too small and inferential results can appear more precise than warranted. The primary endpoint is continuous and analyzed with a linear mixed model; the registry does not report a separate over-dispersion assessment or a correction for it. This general issue should not be taken as evidence that over-dispersion occurred in this study.

The registry reports a superiority analysis for the sponsor’s primary comparison. That label describes the stated analysis framework; it does not by itself establish clinical importance. The protocol’s detailed hypothesis wording, significance-level plan for the primary outcome, and multiplicity strategy are not reported in the registry record. The planning illustrations use a stated two-sided significance level of 0.05, but those calculations are not evidence of the trial’s original design assumptions.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is a least-squares mean difference of -7.95 hours/month, with a 95% confidence interval from -11.45 to -4.46. In general, a confidence interval represents a range of effects reasonably compatible with the observed data and the statistical model at its stated confidence level. Its width communicates precision: a narrower interval indicates less sampling uncertainty, while a wider interval indicates more. The interval does not represent a range containing every individual participant’s response.

Both limits of the sponsor-reported interval are negative. For this endpoint, negative values correspond to fewer hours of at least moderate headache pain in the erenumab group relative to placebo for the modeled change from baseline. The interval also communicates that the estimated difference is not precisely pinned to one value: it extends from -11.45 to -4.46 hours/month. Interpretation should consider the full range, rather than treating -7.95 as an exact effect.

The unadjusted recalculation gives a 95% confidence interval from -11.43 to -4.471. Its point estimate is -7.95, and its p-value is < 0.001. The sponsor’s interval, from -11.45 to -4.46, differs slightly from that recalculated interval. This is expected to be read as a small numerical difference between the sponsor’s model-based analysis and an approximation from posted summaries, not as two independent trial findings. The sponsor’s mixed-model result is the primary reported analysis.

Confidence intervals quantify uncertainty under the model and analysis assumptions. They do not address all uncertainty about trial conduct, missing data, measurement, or applicability to people outside the study population. The registry does not report enough detail to assess all of those sources. The interval should therefore be read as statistical precision for the posted model-based comparison, not as a complete measure of uncertainty about every possible interpretation.

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 planning figures are participants per arm for a continuous-outcome comparison, not total events. Their stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. The original design assumptions for this trial, including its sample-size justification, are not reported in the registry record, so these planning figures should not be read as the trial’s planned enrollment target.

The 80% power column gives the illustrative number of participants per arm under which the calculation would have an 80% chance of detecting the chosen effect if that effect and the stated assumptions held. The 90% power column gives the corresponding illustration for a 90% chance. These are planning concepts, not guarantees: actual power depends on the true effect, variability, analysis, and other design features.

The displayed scenarios use 100%, 75%, and 50% of the observed effect. For the primary endpoint, planning for 100% of the observed effect gives 99 participants per arm for 80% power and 132 for 90% power. Planning for 75% gives 176 and 235 participants per arm, respectively. Planning for 50% gives 395 and 528 participants per arm, respectively. The pattern illustrates that smaller effects require more participants to detect under the same stated assumptions.

These figures are planning illustrations based on the observed effect, not observed power. They do not tell whether the completed trial was well powered according to its original design, because the registry record does not report that design justification. Nor do they revise the primary result: the trial’s analysis and its uncertainty are described by the posted estimates, not by a retrospective power calculation.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -7.950; pooled standard deviation σ = 19.922.
2. n = 2 × (1.96 + 0.842)² × 19.922² / -7.950² = 98.6, rounded up to 99 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
delta <- -7.95                      # observed difference in means
sigma <- 19.9221917744911                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(4252742)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 99 (80% power), 132 (90% power)
 75% of observed effect: n per arm = 176 (80% power), 235 (90% power)
 50% of observed effect: n per arm = 395 (80% power), 528 (90% power)
power.t.test: 99.5 
Simulated power at n = 99 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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 sponsor reports a mixed-model analysis with an erenumab-versus-placebo least-squares mean difference of -7.95 hours/month. Its 95% confidence interval is from -11.45 to -4.46, and the reported p-value is < 0.001. This analysis uses the model specified in the registry for the repeated primary measurements.

The unadjusted recalculation from the posted arm summaries also gives a mean difference of -7.95 and p < 0.001, with a 95% confidence interval from -11.43 to -4.471. The point estimates agree, and the interval limits are close but not identical. The recalculation is an approximation from summary statistics and does not recreate the sponsor’s participant-level mixed model, its repeated-measure covariance structure, or other details of the fitted analysis.

For interpretation of the trial’s primary comparison, the sponsor’s reported model-based result is the principal estimate. The recalculation serves as a transparent, unadjusted comparison using the posted summaries. The small difference in interval endpoints does not change the direction of either estimate. The registry record does not report a discrepancy requiring a different reading of the sponsor’s result.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports a primary outcome and multiple secondary outcomes. It gives the primary mixed-model result and detailed numerical results for several MFIQ domains. However, the multiplicity strategy is not reported in the registry record. In particular, the record does not describe whether testing across secondary outcomes was adjusted, whether the primary test was protected by a testing hierarchy, or whether some outcomes were designated exploratory for inferential purposes. Without that information, the secondary p-values should be read as reported results rather than as a family-wise controlled set of confirmatory tests.

The registry’s description of the primary endpoint says the monthly change is averaged over months with observed data, provided at least one monthly value was observed. The efficacy analysis set also required a baseline value and at least one observed measurement during the main phase. These details identify what observations qualified for the endpoint and analysis population, but they do not fully specify how missing data were handled in the model or what assumptions were made about missingness. Further missing-data methods are not reported in the registry record.

The primary analysis used a linear mixed model with treatment, visit, treatment-by-visit interaction, and baseline value, and a first-order autoregressive covariance structure. This is useful information about the model, but it does not establish all aspects of the analysis plan. The protocol’s handling of intercurrent events, deviations from assigned treatment, and sensitivity analyses for missing data are not reported in the registry record.

The full analysis set is described as all randomized participants. The M-DBTP efficacy analysis set was a subset requiring at least one dose and specified observed measurements. The registry reports the analysis counts as 251 for placebo and 253 for erenumab for the primary outcome. The record does not explain all differences between the full randomized population and this efficacy set. That limits assessment of how exclusions or unavailable measurements may have affected the estimated treatment comparison.

The registry record also does not report the original sample-size justification or interim analysis plan. Consequently, the planning illustrations based on the observed effect cannot be used to reconstruct the original design or its prespecified assumptions. These gaps are limitations in what can be assessed from the registry record; they do not by themselves establish that the analyses were inappropriate.

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

Serious events with the highest arm-level rate: covid-19 pneumonia (0.4%), psychogenic tremor (0.4%).

The registry reports serious adverse events for 2 of 256 participants at risk in the placebo group and 0 of 254 participants at risk in the erenumab group. The serious-event terms listed are COVID-19 pneumonia and psychogenic tremor, each reported as 0.4%. The record does not indicate in this summary whether those events were considered related to study treatment, and no causal attribution should be inferred from the group counts alone.

For other adverse events, the reported counts are 0 of 256 participants at risk for placebo and 0 of 254 for erenumab. The registry reports no deaths in either group. These counts describe the safety results recorded for the listed groups and should be interpreted with their denominators in view. The serious-adverse-event denominators differ from the primary efficacy analysis counts, because they refer to the safety groups rather than the efficacy analysis set.

These safety counts do not establish that an event cannot occur, nor do they provide a complete characterization of risk beyond the events and observation recorded. Further detail about event definitions, timing, severity, relatedness, and the duration of safety follow-up is not reported in the registry record.

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

The posted primary analysis estimates a more negative change from baseline in mean monthly hours of at least moderate headache pain intensity for erenumab than for placebo during the main treatment phase. The sponsor’s model-based difference is -7.95 hours/month, with a 95% confidence interval from -11.45 to -4.46 and p < 0.001. The endpoint definition gives this direction a straightforward statistical interpretation: negative change represents reduced monthly time at moderate or severe headache pain intensity, and the between-group estimate is also negative.

This result is an average group comparison. It does not indicate that every participant experienced a reduction, that the effect was the same for all participants, or that an individual result can be predicted from the group estimate. The confidence interval describes uncertainty in the model-based average difference, not the range of individual outcomes. The registry does not report individual participant-level distributions in the registry record used for this analysis.

The result applies to the endpoint, assessment period, analysis model, and efficacy population described in the registry. It does not by itself establish effects on every aspect of migraine, on outcomes not measured, or beyond the main phase. The month 4 phase is described as exploratory, and detailed results for it are not reported in the registry record. Nor does the trial result answer questions about people who do not match the population described in the official title.

Statistical evidence and clinical importance are different considerations. A p-value describes the data in relation to a statistical null model; it is not a measure of benefit magnitude or patient-perceived importance. The registry record does not provide a threshold for a clinically important difference on the primary endpoint. The numerical estimate and confidence interval therefore describe the statistical result, while a separate clinical judgment about importance cannot be established from these registry entries alone.

The registry’s secondary findings provide related context for selected functional measures, but the multiplicity strategy is not reported. Safety counts are also limited to the events recorded in the registry summary. Taken together, the record supports a description of the observed model-based comparison and its precision, but not broader claims about unreported outcomes, causation for specific adverse events, or long-term effects.

18. Why This Trial Matters Statistically

This trial illustrates how a clinical endpoint can be expressed as a continuous measure of time spent with symptoms rather than only as a count of episodes. The primary outcome combines a specified pain-intensity threshold with headache duration and summarizes monthly change from baseline. That design makes the outcome’s interpretation depend on both the intensity scale and the method used to aggregate monthly observations.

It also illustrates why the analysis model matters. The primary result is not just a contrast of raw arm means: it is a least-squares mean difference from a linear mixed model that includes treatment, visit, their interaction, and baseline value, with a first-order autoregressive covariance structure. Repeated observations from the same participant have a correlation structure, and modeling that structure is part of estimating the treatment contrast and its uncertainty.

The comparison between the sponsor’s result and the unadjusted recalculation is informative about the role of model-based analysis. Both yield a point estimate of -7.95, while their confidence limits differ slightly. This highlights that summary-statistic recalculations can be useful checks, but are not substitutes for the reported repeated-measures analysis.

The record also shows why a statistically small p-value is not a complete interpretation. Effect magnitude, confidence interval, population definition, missing-data handling, and multiplicity all matter. Here the primary confidence interval is negative, but several relevant design and analysis details—including the original sample-size justification and multiplicity strategy—are not reported in the registry record.

Finally, the efficacy and safety summaries use different analysis populations and denominators. Keeping those populations distinct prevents an inappropriate direct comparison of participant counts and reminds readers that each result pertains to its own analysis set.

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.

The registry record gives a concise description of the primary model, but does not report the original sample-size justification, detailed randomization procedures, allocation concealment, or interim analysis plan. The planning illustrations based on the observed effect cannot fill those gaps or establish the assumptions used when the trial was designed.

The primary efficacy analysis uses a subset of the randomized population, with requirements for treatment exposure and observed baseline and follow-up measurements. Although the registry defines those requirements, the reasons for missing observations or exclusions from the efficacy set are not reported in the registry record. The detailed missing-data assumptions, sensitivity analyses, and methods for handling intercurrent events are also not reported.

The registry does not report a multiplicity adjustment or a testing hierarchy for the secondary outcomes. In addition, detailed secondary numerical analyses are available for only some of the listed outcomes. This restricts the strength of conclusions that can be drawn across the full outcome set. The available secondary results can be described as a pattern, but the record does not establish their confirmatory status.

The main treatment phase spans months 1 to 3, while month 4 is identified as an exploratory phase. The detailed results do not support conclusions about the exploratory period or longer-term outcomes. The registry record also does not report a clinical-importance threshold for the primary endpoint, so the statistical estimate cannot alone establish how important the difference is to patients.

For safety, the summary provides serious adverse event counts, listed terms, other adverse event counts, and deaths, but does not give event timing, treatment relatedness, severity detail, or the full follow-up context. These limitations constrain interpretation of the registry safety summary and should not be filled in by assumption.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Erenumab - Comprehensive Assessment of Efficacy in (High-Frequency) Episodic Migraine — NCT04252742.
• Registry record last updated 2026-03-31; status last verified 2026-03.
• Results first posted 2024-08-30.
• 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 describes NCT04252742 as a completed phase 4 randomized, double-blind, parallel-group study of erenumab and placebo in high-frequency episodic migraine. Enrollment was 512. The primary endpoint was change from baseline in mean monthly hours of at least moderate headache pain intensity over months 1, 2, and 3.

The sponsor’s mixed-model analysis reports an erenumab-versus-placebo difference of -7.95 hours/month, with a 95% confidence interval from -11.45 to -4.46 and p < 0.001. The result describes the model-based comparison in the defined efficacy analysis population. The original sample-size justification, multiplicity plan, and detailed missing-data strategy are not reported in the registry record.