A Phase II Study of Minocycline and Armodafinil for Reducing the Symptom Burden Produced by Chemoradiation Treatment for Esophageal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase II Study of Minocycline and Armodafinil for Reducing the Symptom Burden Produced by Chemoradiation Treatment for Esophageal Cancer (NCT01746043), a randomized phase 2 study. The primary outcome is Proportion of Patients Under Each Treatment Arm Who Experienced a Mean Symptom Increase of 2 Units or More From Baseline to 6 Weeks..

2. Clinical Question

3. Trial Design

This was a randomized, double-masked, parallel-group phase II study. Its stated purpose was to compare armodafinil and minocycline, used alone or together, for controlling symptoms associated with chemoradiation treatment for esophageal cancer. A placebo-only group was also included. The registry lists an enrollment of 22, while its primary efficacy description refers to 21 randomized patients; these are different figures reported for different aspects of the record, and no explanation for the difference is reported in the registry record.

The four treatment groups were armodafinil plus placebo, minocycline plus placebo, armodafinil plus minocycline, and placebos. The recorded drug schedules were armodafinil 150 mg by mouth once a day for 6 weeks and minocycline 100 mg by mouth 2 times a day for 6 weeks. Placebo was recorded as one capsule by mouth 2 times a day for 6 weeks. Treatment was to start on the day of chemoradiation or within 5 days of its start. Symptom questionnaires were completed before chemoradiation and then weekly during Weeks 1-10 of the study.

The registry lists the study as completed, with a start date of 2013-02 and a primary completion date of 2017-01. The reported primary efficacy population included 19 of the 21 randomized patients. Reasons that some randomized patients were not evaluable are not reported in the registry record. The registry record does not report more detail about allocation procedures, masking implementation, adherence, or departures from the planned treatment schedules.

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: Of the 22 participants enrolled 1 participant were excluded from the study before assignment to groups: Of the 21 randomized patients, 19 (90%) were evaluable for the primary efficacy analysis.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Of the 21 randomized patients , 19 (90% were evaluable for the primary efficacy analysis).

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Proportion of Patients Under Each Treatment Arm Who Experienced a Mean Symptom Increase of 2 Units or More From Baseline to 6 Weeks., measured over Baseline, 6 weeks.

The primary endpoint is a binary participant-level outcome: whether a patient experienced a mean symptom increase of 2 units or more between baseline and 6 weeks. The score is defined as the mean of five symptom ratings: fatigue, pain, disturbed sleep, lack of appetite, and drowsiness. The endpoint therefore reduces the symptom information to an event count for each treatment arm. It does not display the size or pattern of change for each symptom separately.

A threshold-based outcome can make a clinically specified degree of worsening easier to summarize than a set of separate symptom ratings. It also allows the outcome to be described as the number of participants meeting the threshold. However, dichotomizing a score discards information about changes that do not cross the threshold and about how far scores lie above it. The registry record does not report a rationale for selecting the 2-unit threshold or additional detail on how the mean score was constructed when individual symptom ratings were unavailable.

The endpoint timeframe is baseline to 6 weeks, matching the recorded treatment duration. The registry identifies 19 (90% were evaluable for the primary efficacy analysis) among 21 randomized patients. The posted arm results give event counts, rather than estimated proportions with confidence intervals, and the registry record does not report a statistical comparison between arms.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The posted endpoint is binary: a participant either met the specified increase threshold or did not. For a binary endpoint, the basic descriptive result is the count of participants with the event in each group, interpreted in relation to the number assessed in that group. The registry posts event counts and arm group counts. A proportion would convey the fraction meeting the endpoint, but the posted primary results do not include reported proportions or confidence intervals, and no new calculations are made.

What does a confidence interval tell the reader?

In general, a confidence interval expresses uncertainty around an estimated quantity under the assumptions of the statistical method used. For a treatment comparison, it can show which effect sizes remain reasonably compatible with the observed data and model. A narrow interval indicates greater precision than a wide one, while an interval spanning both potentially important benefit and harm may leave the direction or size uncertain. The registry record does not report confidence intervals for the primary outcome, so the precision of any difference between arms cannot be evaluated from the posted analysis.

Why are arm counts not enough for a comparison?

Counts describe the observed event burden but do not by themselves quantify the uncertainty in a contrast. A comparison would need a defined measure, such as a difference in proportions or a relative measure, and an appropriate method for estimating its uncertainty. The registry record does not report such comparative estimates or a primary-outcome significance test. It indicates that values were missing or not estimable for a two-arm comparison of the binary endpoint. Accordingly, no inferential treatment comparison is available to interpret.

How should the four groups be understood?

The four groups distinguish use of armodafinil, minocycline, both active drugs, or placebos. In principle, a factorial analysis could examine separate contributions of the two drugs and whether their joint effect differs from what would be expected from their individual effects. That is a general design concept, not a claim about the analysis conducted in this trial. The registry record does not report a factorial model, interaction analysis, or other statistical approach for separating these effects. It therefore does not support conclusions about individual drug contributions or interaction.

What does the planning information mean?

In general, a sample-size calculation estimates how many participants would be needed under stated assumptions to detect a specified effect with a chosen error rate and power. The registry's planning description specifies a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the planning effect is described as a stated fraction of the observed effect. Such an illustration is not an observed analysis and does not establish the power achieved by the completed trial. The registry record does not report the trial's own design justification, its original target effect, or a sample-size rationale. No planning count should be read as an observed result.

General statistical methods for binary outcomes also depend on adequate event information and suitable handling of sparse data. When counts are small or events are uncommon, estimates may be unstable and standard approximations may be unreliable; exact or other methods may be considered depending on the analysis plan. This is a general methodological point, not a description of a method used in this study. The registry record does not report the selected test or model, its assumptions, or any adjustment strategy.

The record reports 19 evaluable patients among 21 randomized for primary efficacy, but the reasons for missing evaluation and the handling of missing outcome data are not reported in the registry record. In general, excluding participants without outcome data can affect interpretation if missingness is related to treatment or symptom status. Statistical methods cannot resolve that concern without information about missingness and a transparent analysis strategy. No sensitivity analysis or imputation method is reported in the registry record.

Finally, statistical significance and clinical importance are different concepts. A test addresses compatibility with a null hypothesis under a model; an effect estimate and its uncertainty address magnitude and precision. A thresholded outcome can be useful for a specified event definition but does not preserve all information in the underlying symptom scores. Since the registry record presents counts without comparative estimates, intervals, or tests, the primary results can be read descriptively but cannot establish the size or precision of a treatment contrast.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Armodafinil + Placebo", "Minocycline + Placebo", "Armodafinil + Minocycline", "Placebos")
n      <- c(5, 4, 5, 5)
events <- c(0, 2, 1, 2)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Armodafinil + Placebo  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Minocycline + Placebo  4  2  50.0  6.8  93.2  15.0  85.0  14.7  85.3
ARM  Armodafinil + Minocycline  5  1  20.0  0.5  71.6  3.6  62.4  4.3  64.1
ARM  Placebos  5  2  40.0  5.3  85.3  11.8  76.9  11.8  77.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and no posted secondary results. It reports 0 analyses, and no statistical comparison for the primary binary endpoint is available. The registry record does not report a multiplicity strategy, such as how testing across treatment groups, outcomes, or follow-up times would be controlled. Without a reported analysis plan and results, it is not possible to determine whether multiplicity was considered or how it would affect interpretation.

The primary efficacy description states that 19 (90% were evaluable for the primary efficacy analysis) among 21 randomized patients. The record does not report why the other randomized patients were not evaluable, whether missingness was related to treatment or symptoms, or how missing baseline or 6-week measurements were handled. The registry lists enrollment of 22 but does not explain the relationship between that figure and the randomized and evaluable populations.

The analysis population is described only as evaluable for the primary efficacy analysis. Whether the primary analysis followed an intention-to-treat approach, a per-protocol approach, or another prespecified definition is not reported in the registry record. No imputation, sensitivity, or missing-data analysis is described. These omissions matter because analysis-population definitions and missing-outcome handling can influence the meaning of a binary event count.

12. 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 0% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: fever (25.0%), fatigue (20.0%), dysphagia (20.0%), peripheral neuropathy (20.0%), headache (20.0%).

The adverse-event record lists serious adverse events for each treatment group. It records 1 affected participant in each of the armodafinil, minocycline, armodafinil-plus-minocycline, and placebo groups. It records 0 participants with other adverse events in each group and 0 deaths in each group. These are the reported counts, not comparative safety estimates.

The serious-event terms with posted percentages are Fever at 25.0%, Fatigue at 20.0%, Dysphagia at 20.0%, Peripheral Neuropathy at 20.0%, and Headache at 20.0%. The registry record does not provide a group-specific breakdown for these terms in the listed summary. The percentages should be read as reported; the record does not state additional details about their calculation in this summary.

These figures do not establish whether an event was caused by a study drug, by chemoradiation, or by another factor. Nor do they permit a reliable comparison of safety across groups: the record does not report comparative analyses, uncertainty estimates, or detailed event timing and attribution. The listed counts describe what the registry reports, not a causal assessment.

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

The primary results describe how many participants in each arm crossed the prespecified symptom-increase threshold from baseline to 6 weeks. The counts were 0 for armodafinil plus placebo, 2 for minocycline plus placebo, 1 for armodafinil plus minocycline, and 2 for placebos. This is the direct descriptive content of the posted results. The registry record does not report a comparative estimate, confidence interval, or test that would quantify the uncertainty around differences between these groups.

Consequently, the counts do not establish that one treatment strategy is better or worse than another. They also do not establish that the groups are equivalent. A finding of no statistically significant difference cannot be inferred because no primary-outcome test is reported. In general, absence of a reported test or estimate is not evidence that treatment effects are absent; it means that the registry entry does not provide the comparative analysis needed to evaluate them.

The endpoint represents a mean of fatigue, pain, disturbed sleep, lack of appetite, and drowsiness, converted to a threshold event. It does not show the course of each symptom separately, the distribution of score changes, or whether participants experienced improvement in some symptoms alongside worsening in others. Although questionnaires were scheduled weekly during Weeks 1-10, secondary outcome results are not posted. The primary endpoint is limited to the baseline-to-6-week comparison.

Interpretation is also limited by the reported analysis population. The registry identifies 19 evaluable participants among 21 randomized patients but does not describe reasons for missing outcome data or how those data were handled. The posted group counts should therefore be understood within the registry's stated primary efficacy context, rather than assumed to represent every randomized patient. The enrollment figure of 22 is also distinct in the record, with no explanation of the difference.

Overall, the record provides descriptive primary event counts and adverse-event counts. It does not provide enough inferential detail to determine the magnitude, precision, or statistical reliability of treatment contrasts. These registry results should not be extended to outcomes, populations, or analyses that are not reported.

14. Why This Trial Matters Statistically

This trial is statistically notable for its four-arm comparison of two drugs, their combination, and placebo, while using a single thresholded primary symptom outcome. Such a structure raises questions about which contrasts are of primary interest and whether separate drug effects or a joint effect are to be assessed. Those questions require a prespecified statistical plan. The registry record does not report that plan or a completed comparative analysis, so the design's potential contrasts cannot be interpreted as tested findings.

The endpoint combines five symptoms into a mean and then classifies participants using a threshold. This approach offers a compact event-based summary, but it necessarily compresses the underlying symptom information. A participant-level event count cannot indicate which symptoms changed, how much they changed, or how scores evolved over the weekly questionnaire schedule. Those distinctions are relevant when interpreting what an endpoint captures, independent of any treatment effect.

The record also illustrates the importance of a clearly described analysis population. It reports 21 randomized patients, 19 evaluable for primary efficacy, and enrollment of 22, without explaining the relationship among those figures or the reasons for non-evaluability. Transparent accounting and missing-data methods are essential for understanding how observed endpoint counts relate to the randomized groups.

Finally, the gap between descriptive counts and comparative inference is central. Counts are useful for reporting observed events, but comparative estimates and uncertainty are needed to assess treatment differences. The registry does not post those estimates for the primary outcome. Thus, the statistical significance of the study lies in identifying the information needed for interpretation, as much as in the observed counts themselves.

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 primary endpoint is based on the mean of five symptoms and a threshold of 2 units or more. The registry record does not report a rationale for the threshold, details on the scoring procedure beyond the stated mean, or how incomplete symptom questionnaires were treated. As a result, the effect of missing symptom items on the endpoint cannot be assessed.

The primary efficacy description refers to 19 evaluable patients among 21 randomized, while the record lists enrollment of 22. It does not explain the relationship between enrollment and randomization, why participants were not evaluable, or whether the posted arm counts include all participants with an available endpoint. No missing-data sensitivity analysis is reported.

The primary outcome results are counts by arm. Comparative effect estimates, confidence intervals, and significance tests are not reported, and the record indicates that two-arm values were missing or not estimable for the binary endpoint. This prevents assessment of comparative precision and leaves the statistical uncertainty of treatment contrasts unresolved.

The record lists weekly symptom questionnaires during Weeks 1-10 but reports no secondary results. It therefore does not provide outcome trajectories or results for individual symptoms. The adverse-event summary similarly provides counts and selected terms but no causal attribution or comparative safety analysis. Further limitations include the absence of reported details on allocation implementation, masking procedures, adherence, the statistical analysis plan, and reasons for any departures from treatment or assessment schedules.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Phase II Study of Minocycline and Armodafinil for Reducing the Symptom Burden Produced by Chemoradiation Treatment for Esophageal Cancer — NCT01746043.
• Registry record last updated 2018-06-06; status last verified 2018-05.
• Results first posted 2018-06-06.
• 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

NCT01746043 is a completed randomized, double-masked, parallel-group phase II study concerning symptom burden during chemoradiation for esophageal cancer. The four groups were armodafinil plus placebo, minocycline plus placebo, both active drugs, and placebos. The primary endpoint was the count of participants with a mean symptom increase of 2 units or more from baseline to 6 weeks.

The record lists enrollment of 22 and states that 21 patients were randomized, with 19 evaluable for the primary efficacy analysis. Primary event counts by arm were 0, 2, 1, and 2, respectively. Comparative estimates, confidence intervals, and statistical tests are not reported in the registry record. Reasons for non-evaluability and missing-data methods are also not reported.

Adverse-event counts and selected serious-event terms are recorded, but no comparative safety analysis or causal attribution is given. The record lists no posted secondary results. The start date is 2013-02, the primary completion date is 2017-01, and the study status is completed.