Safety and Efficacy of Milnacipran in Pediatric Patients With Primary Fibromyalgia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Milnacipran in Pediatric Patients With Primary Fibromyalgia (NCT01328002), a randomized phase 2 study. The primary outcome is Time to First Loss of Therapeutic Response (LTR) Following Randomization to Milnacipran or Placebo..

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, parallel-group, quadruple-masked study of milnacipran and placebo in pediatric patients with primary fibromyalgia. Its official title describes a multicenter, double-blind, placebo-controlled withdrawal study. The registry lists an enrollment of 116 and a status of terminated. The sponsor is Forest Laboratories. The record gives a study start date of 2011-04-30 and a primary completion date of 2012-08-31; a reason for termination is not reported in the registry record.

The design had an open-label period followed by a double-blind treatment period. Patients in the open-label period received milnacipran. The registry states that 20 patients out of 116 enrolled had at least a 50% reduction in pain from baseline, were classified as responders, and were randomized at Visit 7. The randomized groups were milnacipran and placebo, each with oral dosing twice daily. The outcome population is thus a selected responder population: the randomized comparison addresses continuation versus withdrawal to placebo among patients who first responded during open-label treatment. It does not directly describe outcomes for all patients who began the open-label period.

The registry identifies the design as randomized and parallel, with quadruple masking. It does not specify in the record which roles were masked. The listed secondary outcome is Patient Global Impression of Severity (PGIS). There are two outcome measures posted, but no analyses are listed. Additional details of the randomization procedure, allocation concealment, site-level conduct, and protocol deviations are not reported in the registry record. The record also does not explain why the study was terminated. Those omissions limit the ability to reconstruct the conduct and analysis from the registry alone.

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: 116 patients took at least 1 dose of open-label investigational product; 20 patients were randomized to receive double-blind treatment.

Open-Label Period

Reasons for not completing this period:

Double-Blind Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Open-Label Safety Population consists of 116 patients who took at least 1 dose of open-label milnacipran. 20 patients who were randomized to a treatment group (Randomized Population) took at least 1 dose of double-blind treatment (Double-blind Safety Population) and were analyzed as randomized (Double-blind Intent-to-Treat \[ITT\] Population).

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: All demographic data was based on the Open-Label Safety Population, which consists of 116 patients who took at least 1 dose of open-label milnacipran.

6. Primary Endpoint

The registry lists the primary outcome as Time to First Loss of Therapeutic Response (LTR) Following Randomization to Milnacipran or Placebo., measured over Change from Visit 7 (Week 8) to Visit 10 (Week 16).

The primary endpoint is time to first loss of therapeutic response after randomization to milnacipran or placebo. The stated measurement unit is days, and the reported summary parameter is the mean with standard deviation. The timeframe is from Visit 7 (Week 8) to Visit 10 (Week 16). The registry defines loss of response using three possible events: worsening fibromyalgia requiring alternate treatment, a rise in the one-week mean of daily pain ratings to greater than 70% of baseline, or withdrawal from the study except for withdrawals due to extenuating circumstances.

Time to event endpoints can capture both whether an event occurs and how long participants remain without that event. This can be useful when the timing of deterioration matters, rather than only whether a participant meets a threshold at a single visit. In general, time-to-event analyses need clear event definitions and methods for handling participants whose event is not observed during follow-up. The registry’s definition identifies the possible triggers, but the handling of censoring, follow-up after withdrawal, and the analysis method is not reported in the registry record.

The posted mean for milnacipran is 7.0 days, with a standard deviation of 1.4 days, among 14 patients. No primary endpoint value is reported for the 6 patients in the placebo arm. The absence of that arm’s value prevents a direct estimate of the randomized-group difference from the posted summary statistics.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean and standard deviation describe?

The primary outcome is labeled as a continuous measure summarized by a mean and standard deviation. In general, the mean gives the arithmetic average of observed values, and the standard deviation describes how much the observations vary around that average. They answer different descriptive questions: the mean locates the group’s average, while the standard deviation characterizes spread. Neither is a measure of uncertainty in the estimated mean; uncertainty is commonly expressed using a confidence interval. No confidence interval for the primary endpoint is reported in the registry record.

The registry reports a milnacipran mean of 7.0 days and a standard deviation of 1.4 days for 14 patients. It gives no corresponding placebo mean or standard deviation for 6 patients. A comparison of group averages requires information from both arms, so the missing placebo summary prevents an unadjusted recalculation of the difference. The record lists no analysis for this outcome.

How does a time-to-event outcome differ from a visit-based score?

In general, a time-to-event outcome records the time until a defined event occurs. Here, the event is the first loss of therapeutic response, with several qualifying event conditions specified in the registry. Time-to-event methods can use information about when an event occurred and can account for participants whose event was not observed during follow-up, provided the analysis rules are known. Handling such participants usually depends on whether their follow-up ended before an event and on the analysis plan.

The registry gives the outcome unit as days and identifies the timeframe as Change from Visit 7 (Week 8) to Visit 10 (Week 16). It does not report the detailed time-to-event method, the censoring rules, or how withdrawal and alternate treatment were represented in the analysis. It also does not describe whether the reported mean is based on event times for every participant or on another convention. These omissions make it difficult to interpret the mean as a complete time-to-event analysis. No hazard ratio or other time-to-event contrast is reported.

What does a randomized withdrawal design address?

In general, a randomized withdrawal design first gives participants an active treatment and then randomizes a selected group to continue active treatment or switch to placebo. This approach asks about outcomes among people who initially responded or met another selection criterion during the first treatment period. It does not estimate the same quantity as a trial that randomizes all eligible participants at the outset.

In this study, the registry says that 20 patients out of 116 enrolled achieved at least a 50% pain reduction from baseline during open-label treatment and were randomized. The comparison is therefore relevant to that responder group. The record does not report a placebo-controlled comparison among all 116 patients, nor does it give outcome results for those who were not randomized. This design context is important when considering how broadly the results could apply.

What can be learned from the planning assumptions?

The record describes planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It says that the effect to detect is a stated fraction of the observed effect. These are planning assumptions, not observed outcomes or evidence that the study had a particular probability of finding a result. Sample-size illustrations are design calculations; they are not observed power.

For time-to-event outcomes, the record says planning tables count total events; for other outcomes, they report participants per arm. The primary outcome is a time-to-event endpoint, but the registry does not report a usable two-arm effect estimate or an analysis result. The planning assumptions alone do not establish the trial’s own design rationale. The sample-size justification and the trial’s original statistical plan are not reported in the registry record.

Why does missing outcome information matter?

In general, a group comparison depends on the observed outcome information in each group and on appropriate uncertainty estimates. A sample count without outcome values cannot establish the group’s average or the contrast with another group. Here the placebo count is reported, but its primary outcome value is absent; the record explicitly identifies the two-arm continuous comparison as not estimable. No recalculated comparison is therefore presented.

Different approaches to missing data can lead to different estimates, particularly when missingness is related to participants’ outcomes. The registry does not report a missing-data strategy, reasons for absent endpoint values, or sensitivity analyses. Those details should not be inferred from the fact that the primary outcome lacks a placebo summary. The posted figures are descriptive and incomplete for comparing the randomized groups.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Milnacipran")
n   <- c(14)
m   <- c(7)
sp  <- c(1.4)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Milnacipran  14  7.000  1.400  6.192  7.808

11. Multiplicity, Missing Data and Analysis Populations

The record lists Patient Global Impression of Severity (PGIS) as a secondary outcome and says that two outcome measures are posted. It does not report the statistical plan for controlling multiplicity across outcomes, time points, or other analyses. Multiplicity matters in general because evaluating several outcomes or comparisons can increase the chance of at least one apparently notable result by chance. Without the planned testing hierarchy or adjustment method, the registry record cannot establish how any collection of tests was intended to be interpreted. No analyses are listed, so the record gives no multiplicity-adjusted or unadjusted inferential results.

Important details about missing data are also not reported in the registry record. The placebo arm has a reported group count but no primary endpoint value, and the record marks the two-arm continuous comparison as not estimable. Reasons for the missing value, participant-level completeness, handling of withdrawals, and sensitivity analyses are not described. General statistical practice distinguishes the reasons data are missing and examines whether conclusions depend on assumptions about those data; no such assessment is reported.

The record identifies an open-label safety population of 116 patients who took at least one dose of milnacipran. It also states that 20 patients were randomized and took at least one dose of double-blind treatment, forming the double-blind safety population. Those randomized patients were analyzed as randomized, according to the record. The primary endpoint description concerns patients randomized after response during the open-label period. The precise analysis population for the primary endpoint, beyond the listed arm counts, is not further described. The registry does not report an intention-to-treat definition, per-protocol population, or additional exclusions for the efficacy analysis.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: suicidal ideation (7.1%).

The registry separates safety information by treatment period. During open-label milnacipran treatment, the safety population included 116 patients; 0 serious adverse events are listed, and 62 patients are listed with other adverse events. During double-blind treatment, the placebo group included 6 patients, with 0 serious adverse events and 4 patients with other adverse events. The double-blind milnacipran group included 14 patients, with 1 serious adverse event and 6 patients with other adverse events. These are counts as posted, not estimates of a causal treatment effect.

The open-label period differs from the randomized comparison: everyone in that period received milnacipran, while the double-blind period compares the two randomized groups. The periods also involve different populations, so their counts should not be treated as directly comparable measures of risk. The registry lists suicidal ideation as the top serious event term at 7.1%, but does not specify its treatment-group context in that listing. Death counts are not reported in the registry record.

In general, adverse-event counts need context about definitions, follow-up, and how events were collected. The record’s counts alone do not show whether events were related to treatment, their timing or duration, or whether any person had more than one event. They should be read as registry safety summaries, not as proof of cause or as a complete safety assessment.

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

The posted primary endpoint result is a descriptive mean for the milnacipran arm: 7.0 days, with a standard deviation of 1.4 days, among 14 patients. The placebo arm has 6 patients listed, but no primary endpoint mean or standard deviation. Because the registry marks the two-arm comparison as not estimable and lists no analyses, the record does not provide a direct estimate of the difference between randomized treatments, its uncertainty, or a statistical test.

This is not evidence that the groups had the same outcome, nor does the missing comparison establish that either group did better. It means the posted results do not contain the outcome information needed to quantify the contrast. A single arm’s mean cannot answer the randomized comparative question on its own. The endpoint is also defined through several possible forms of loss of response, while the event-type breakdown and analytic handling are not reported.

The population under study matters for interpretation. Patients entered an open-label milnacipran period, and those meeting the response criterion were randomized to continue milnacipran or receive placebo. The primary comparison therefore pertains to the selected responders described in the record. It is not a direct estimate for every pediatric patient with primary fibromyalgia, nor does the record provide the result for patients who did not reach the response criterion.

The safety counts are likewise descriptive. The record separates open-label and double-blind periods and reports serious and other adverse-event counts, but it does not provide enough context to attribute events to treatment or compare risk reliably. The study status is terminated; the reason is not reported in the registry record. Overall, the registry provides design and population details and one arm’s primary endpoint summary, but not a complete comparative efficacy analysis.

14. Why This Trial Matters Statistically

This trial is statistically notable because its randomized withdrawal design defines a specific estimand: outcomes after randomization among participants who first met a response criterion during open-label treatment. That is a different question from whether starting treatment benefits an unselected population. Recognizing the selection step helps prevent the result from being generalized beyond the population the randomized comparison was designed to address.

The endpoint also illustrates why an outcome definition and an analysis method need to be read together. Loss of therapeutic response could be triggered by worsening requiring alternate treatment, a pain-rating threshold, or qualifying withdrawal. Those events may represent different clinical pathways. A time-to-event analysis can use the timing of events, but understanding its result depends on details such as censoring rules and follow-up. Those methods are not reported in the registry record.

The primary result is incomplete for a randomized comparison: a mean and standard deviation appear for milnacipran, but not for placebo, and the record states the comparison is not estimable. This is a useful reminder that a trial’s existence, randomization, and posted group counts do not guarantee that the registry contains a usable comparative estimate. Without the second arm’s outcome and inferential results, readers cannot judge the magnitude or precision of the between-group contrast from this record.

Finally, the safety information is separated by treatment period, and the open-label and randomized periods are not interchangeable. Distinguishing their populations and designs is necessary before interpreting event counts. Taken together, the record demonstrates the value of checking which participants were randomized, what endpoint was measured, and whether both arms have analyzable results before drawing conclusions.

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.

The registry record leaves several trial-specific uncertainties. Although it identifies a randomized, parallel, quadruple-masked study, it does not describe the randomization procedure, allocation concealment, or which roles were masked. It reports that the study was terminated but gives no reason. The original protocol and statistical analysis plan are not reported in the registry record, so planned analyses and any deviations from them cannot be assessed.

The primary endpoint is called time to first loss of therapeutic response, but the record summarizes it using a mean and standard deviation. It does not describe the method used to handle participants without an observed event, the censoring rules, or how the alternate-treatment and withdrawal criteria entered the analysis. Nor does it provide participant-level event times or a breakdown by trigger. The placebo outcome summary is absent, and the record says the two-arm comparison is not estimable. No effect estimate, confidence interval, or p-value is listed.

The randomized population consists of patients identified as responders after open-label treatment. This selection limits the scope of the primary question to that group. The record does not report endpoint outcomes for patients who did not meet the response criterion. It also does not give reasons for missing endpoint information, the missing-data method, or sensitivity analyses.

For safety, the record reports counts separately for open-label and double-blind periods but does not explain event attribution, timing, duration, or repeated events. The serious-event term listing does not identify its group context. These gaps mean the registry offers only a partial account of efficacy and safety; they should not be filled by assumptions about the protocol or conduct.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Safety and Efficacy of Milnacipran in Pediatric Patients With Primary Fibromyalgia — NCT01328002.
• Registry record last updated 2019-05-14; status last verified 2019-03.
• Results first posted 2019-05-14.
• 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

NCT01328002 is a phase 2 randomized, parallel, quadruple-masked, placebo-controlled withdrawal study of milnacipran in pediatric patients with primary fibromyalgia. The registry lists an enrollment of 116, a terminated status, and Forest Laboratories as sponsor. Patients received open-label milnacipran; responders were randomized to twice-daily oral milnacipran or placebo. The record states that 20 patients out of 116 enrolled met the response criterion and were randomized.

The primary endpoint was time to first loss of therapeutic response from Visit 7 (Week 8) to Visit 10 (Week 16). For 14 patients in the milnacipran arm, the registry reports a mean of 7.0 days and a standard deviation of 1.4 days. The placebo arm has 6 patients listed, but no primary endpoint value; the two-arm comparison is marked not estimable, and no analyses are listed. The record also reports safety counts for open-label and double-blind periods. A reason for termination, detailed analysis methods, and the missing placebo outcome are not reported in the registry record.