Evaluation of AMG 714 for Vitiligo: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of AMG 714 for Vitiligo (NCT04338581), a randomized phase 2 study. The primary outcome is Proportion of Participants Achieving a ≥ 35% Improvement From Baseline in the Facial Vitiligo Area Scoring Index (F-VASI35) at Week 24.

2. Clinical Question

3. Trial Design

This was a Phase 2a randomized, double-blind, placebo-controlled trial with a parallel-group model. The registry identifies masking as quadruple, though it does not specify which people or roles were masked. The study is titled Evaluation of AMG 714 for Vitiligo and was sponsored by the National Institute of Allergy and Infectious Diseases (NIAID). Its status is listed as completed.

The registry reports a total enrollment of 60. For the primary endpoint, the modified Intent-to-Treat population included randomized participants who received at least one dose of either AMG 714 or placebo. The reported arm sizes for that outcome are 40 for AMG 714 and 19 for placebo. Safety results use different at-risk group sizes: 41 for AMG 714 and 18 for placebo. These outcome-specific populations should be kept distinct rather than treated as interchangeable.

Participants assigned to AMG 714 were scheduled to receive 300 mg at Week 0 and every 2 weeks through Week 10, for a total of 6 doses, with each dose delivered as 2 subcutaneous injections of 1.5 mL each. The placebo group was scheduled to receive 2 injections of 1.5 mL on the same schedule. Although nbUVB phototherapy is listed as an intervention, its allocation and role in the study are not described in the registry record.

The registry lists a study start date of 2020-12-11 and a primary completion date of 2024-10-16. Beyond the listed outcomes and the primary-endpoint missing-data rule, details such as the full analysis plan and the rationale for the planned enrollment 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 Modified Intent-to-Treat (mITT) population included all randomized participants who received at least one dose of either AMG 714 or placebo.

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 Modified Intent-to-Treat (mITT) population included all randomized participants who received at least one dose of either AMG 714 or placebo.

6. Primary Endpoint

The registry lists the primary outcome as Proportion of Participants Achieving a ≥ 35% Improvement From Baseline in the Facial Vitiligo Area Scoring Index (F-VASI35) at Week 24, measured over Week 24.

The primary endpoint is the proportion of participants achieving at least 35% improvement from baseline in the Facial Vitiligo Area Scoring Index (F-VASI) at Week 24. F-VASI assesses the area of the face affected by vitiligo. It combines the estimated percentage of body surface area affected with the estimated degree of depigmentation. The investigator estimates affected area using hand or thumb units and records it to the nearest 0.1% of body surface area. Depigmentation is estimated using the categories listed in the registry record. F-VASI ranges from 0 to 3.5, with higher scores indicating more severe disease.

The endpoint turns change on a continuous severity score into a responder outcome: each participant either meets the stated improvement threshold or does not. This can make the result straightforward to interpret as a proportion of participants meeting a defined criterion, but it also discards information about the size of improvement among responders and non-responders. The threshold is one selected level of change, not a complete account of all changes in facial vitiligo.

For the primary endpoint, missing Week 24 F-VASI assessments are counted as non-response. That rule makes the outcome definition explicit, but the resulting proportion reflects both observed responses and the prespecified classification of missing assessments. The registry record does not describe a separate sensitivity analysis for this missing-data rule.

7. Results Posted in the Registry

• AMG 714: 0.0% (0/40) versus 0.0% (0/19) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0); unadjusted two-proportion z-test p = n/a.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary: each participant is classified as a responder or non-responder under the Week 24 F-VASI35 definition. For a binary outcome, the arm-level proportion is the estimated fraction of participants in that group classified as responders. The registry reports these proportions with 95% confidence intervals. A confidence interval describes uncertainty in an estimate under the method used to construct it; it is not a range that contains a fixed percentage of individual participant outcomes, nor does it assign a probability to a treatment effect.

What the arm-level estimates describe

The AMG 714 estimate is 0.15, with a 95% confidence interval from 0.057 to 0.298. The placebo estimate is 0.053, with a 95% confidence interval from 0.001 to 0.26. Each interval belongs to its own arm-level proportion. Neither one directly estimates the difference between the two groups. A contrast requires an analysis that compares the arms and accounts for the sampling uncertainty of that comparison. The reported inferential analysis is a Fisher Exact test, but the registry does not provide an interpretable recalculated contrast estimate in the comparison fields.

Why a Fisher Exact test is used

Fisher's Exact test is a method for comparing categorical outcomes between groups. It evaluates how compatible the observed responder and non-responder counts are with a null hypothesis of no association between treatment group and response. It is commonly used when group sizes or event counts are small, because its calculation uses the discrete distribution of possible tables rather than relying on a large-sample approximation. The registry gives the test name and p-value, but does not report additional details about the testing procedure, such as whether the reported test was one-sided or two-sided.

The reported p-value is 0.411. A p-value is calculated under a null model and describes the probability of obtaining data at least as incompatible with that model as the observed data, according to the test's definition. It is not the probability that the null hypothesis is true, and it does not quantify the chance that a treatment effect exists. It also does not express the size of a difference. Interpretation therefore requires the arm estimates, their uncertainty, and the design and outcome definitions, not the p-value alone.

What over-dispersion means in general

General statistical context: over-dispersion occurs when observed variability is greater than a model assumes. It can make standard errors too small and confidence intervals too narrow if a model does not account for the extra variation. Over-dispersion is especially relevant to some count or clustered-data models. The primary endpoint is a binary responder outcome, and the registry identifies a Fisher Exact comparison rather than a count-model analysis. An over-dispersion adjustment is not reported in the registry record.

How missing assessments enter the endpoint

The primary endpoint specifies that participants missing the Week 24 F-VASI assessment are imputed as non-responders. This is a conservative classification in the limited sense that a missing assessment cannot be counted as a response under this rule; it is not necessarily conservative for every possible comparison or missingness pattern. The resulting analysis depends on both observed assessments and the specified treatment of missing outcomes. The registry does not report the number of missing Week 24 assessments or a sensitivity analysis under alternative assumptions.

The modified Intent-to-Treat population is defined as all randomized participants who received at least one dose of AMG 714 or placebo. This differs from a population that includes every randomized participant regardless of treatment received. The registry's arm-level outcome sample sizes are 40 and 19. Details of the analysis population for other endpoints may differ and are not described in full in the registry record.

What the posted analysis does not establish

The registry does not report the confidence-interval construction method for the arm proportions, the full statistical analysis plan, or a model-based adjusted analysis for the primary endpoint. It also does not report how potential protocol deviations or other intercurrent events were handled. Consequently, interpretation should stay close to the reported outcome definition and stated analysis. The available information supports describing the posted proportions, intervals, and Fisher Exact p-value, but not inferring unreported modeling choices or additional statistical procedures.

10. Confidence Intervals and Precision

The 95% confidence interval for the AMG 714 response proportion extends from 0.057 to 0.298 around a point estimate of 0.15. For placebo, the interval extends from 0.001 to 0.26 around a point estimate of 0.053. These intervals are wide relative to the point estimates, indicating that the observed data leave substantial uncertainty about each arm's underlying response proportion. The registry does not report the method used to construct these intervals, so their exact statistical properties cannot be assessed from the record alone.

Confidence intervals for the two separate arm proportions are not the same as a confidence interval for the treatment contrast. Even if arm intervals overlap or do not overlap, that visual comparison is not a replacement for a directly estimated between-group effect and its uncertainty. The registry-reported Fisher Exact test gives an inferential comparison, but the registry does not provide a usable recalculated contrast estimate and interval in its comparison fields. The arm-level intervals should therefore be interpreted only as precision statements about their respective proportions.

Group size affects precision: with fewer observations, each outcome can have a greater influence on the estimated proportion. The primary outcome results are reported for 40 participants in the AMG 714 group and 19 in the placebo group. These sizes, together with the observed response proportions, are relevant context for the broad intervals. They do not justify inferring an unreported sample-size calculation or the probability that the study would detect a particular effect.

The intervals concern the endpoint as defined, including the rule that missing Week 24 assessments count as non-responses. They do not quantify uncertainty arising from choices not represented in the interval, such as alternative missing-data assumptions, the selection of one threshold among multiple outcomes, or the applicability of the result beyond the trial population.

11. Analyses Reported by the Sponsor

The registry reports a Fisher Exact test for the primary comparison, with p = 0.411. That is the reported analysis result to use when describing the trial's inferential comparison. The recalculated comparison fields are not interpretable: they report zero-valued estimates, interval bounds, and event counts, despite nonzero arm-level response proportions and reported group sizes. They cannot provide a meaningful independent estimate of the contrast.

The arm-level results remain interpretable as posted: 0.15 for AMG 714, with a 95% confidence interval from 0.057 to 0.298, and 0.053 for placebo, with a 95% confidence interval from 0.001 to 0.26. These estimates do not repair the unusable recalculated contrast, and they should not be combined by informal arithmetic to create a new effect estimate. The registry's own Fisher Exact result is the available reported inferential comparison.

The p-value is not a measure of effect size or precision. It should be read alongside the arm proportions and their intervals, while recognizing that the registry does not provide a usable contrast interval. The record does not explain the zero-valued recalculated fields or provide further detail that would resolve them.

12. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is one of a broader set of outcomes. The registry lists 21 outcome measures and reports 20 secondary outcomes. Several listed secondary outcomes use different F-VASI response thresholds, assess total-body vitiligo with T-VASI thresholds, or measure response at multiple follow-up times. The registry also records 80 analyses. When many outcomes, thresholds, and time points are examined, the chance of at least one apparently notable result can increase. A multiplicity adjustment can address some such concerns, but the approach used for these analyses is not reported in the registry record.

The primary endpoint is clearly identified as F-VASI35 at Week 24, and its missing-assessment rule is explicit: participants missing the Week 24 assessment are imputed as non-responders. The record does not report how many assessments were missing, by group, or whether the same missing-data rule was used for secondary endpoints. It also does not report sensitivity analyses exploring alternative assumptions about missing outcomes. Without those details, readers cannot assess how dependent the primary result is on the handling of missing assessments.

The primary analysis population is the modified Intent-to-Treat population, defined as randomized participants who received at least one dose of either study treatment. This definition excludes randomized participants who did not receive a dose. The registry reports 40 and 19 participants in the two primary-outcome arms, but does not give a fuller account of participant flow or explain the difference between the listed overall enrollment and the endpoint-specific group sizes.

Details of the statistical plan, including any prespecified multiplicity strategy, are not reported in the registry record. Nor are interim analyses or a sample-size justification reported. These omissions do not by themselves invalidate the posted results, but they limit evaluation of how the primary endpoint was selected and how secondary findings should be interpreted alongside it.

13. 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 non-serious adverse event terms

The safety record reports 0 serious adverse events among 41 participants at risk in the AMG 714 group and 0 among 18 participants at risk in the placebo group. It reports 0 deaths in both groups. These are the posted counts for the specified safety categories; they do not demonstrate that no adverse events of any kind occurred.

For other adverse events, the registry reports 23 affected participants among 41 at risk in the AMG 714 group and 9 among 18 at risk in the placebo group. These are counts, not a complete description of the types, timing, duration, severity, or relatedness of events. The registry does not list top serious adverse event terms.

The safety group sizes differ from the primary-outcome group sizes, which are 40 and 19. The denominators should therefore be read as outcome-specific rather than assumed to represent the same analysis population. Zero reported serious events also does not establish that the underlying risk is zero; with limited group sizes, rare outcomes may not be observed. The record does not provide enough detail to characterize comparative safety beyond the posted counts.

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

The primary result describes the proportion of participants meeting a defined facial vitiligo response threshold at Week 24, under the registry's rule that a missing assessment counts as non-response. It gives an outcome-specific summary for the modified Intent-to-Treat population, not a description of every randomized participant or every aspect of disease change. The posted arm proportions are 0.15 for AMG 714 and 0.053 for placebo, with broad 95% confidence intervals.

The result does not establish the size of a between-group effect with precision. The registry-reported Fisher Exact p-value is 0.411, while the recalculated contrast fields do not provide a usable estimate. A p-value is not a verdict about efficacy, and a difference between arm point estimates is not, by itself, proof of a treatment effect. The posted information supports a cautious account of the observed proportions and the uncertainty surrounding them, rather than a definitive conclusion about the comparative effect.

The endpoint captures whether an individual met the threshold, not how far that person's score changed beyond the threshold or how change evolved over time. It concerns facial involvement, not the total-body endpoint. The registry lists additional F-VASI and T-VASI thresholds and follow-up visits, but those outcomes are separate questions and should not be inferred from the Week 24 primary result.

The findings are specific to the trial design, its adult vitiligo population, the AMG 714 and placebo regimens, and the outcome definition. They do not establish results for other populations, treatment schedules, or unmeasured outcomes. The record's limited detail on missingness, multiplicity, and the full analysis plan further constrains how broadly the posted result can be interpreted.

15. Why This Trial Matters Statistically

This trial illustrates why a responder proportion should be read with its uncertainty and its definition. The Week 24 endpoint uses a specific F-VASI threshold and an explicit rule for missing assessments. Those choices determine what counts as a response and therefore shape the quantity being estimated. A proportion is not a universal summary of clinical change: it answers the threshold-based question selected for the endpoint.

The arm estimates and the reported comparison answer related but distinct statistical questions. Each confidence interval describes uncertainty in one group's response proportion. The Fisher Exact test evaluates a comparison between groups. Neither the arm intervals nor the p-value gives a complete account of effect magnitude and precision. The unusable recalculated contrast fields underscore why a direct, interpretable estimate of the between-group effect and its interval is useful alongside a test result.

The registry also lists many outcomes and analyses. This makes transparent reporting of the primary endpoint, analysis population, missing-data approach, and multiplicity strategy important. The primary endpoint and its missing-data rule are described, while several other details are not reported in the registry record. Recognizing that boundary helps distinguish what the results can answer from what remains uncertain.

For biostatisticians and clinical researchers, the record is a practical example of keeping descriptive estimates, inferential tests, safety counts, and design information separate. It also shows why outcome-specific denominators matter: the primary endpoint and safety summaries use different reported group sizes. Careful interpretation depends on tracking which participants, outcome, time point, and method each number describes.

16. 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 provides only limited information for evaluating the primary statistical analysis beyond the endpoint definition, analysis population, confidence intervals, and Fisher Exact test. The method used to construct the arm-level confidence intervals and the sidedness of the reported test are not reported in the registry record. The recalculated comparison fields contain zero-valued quantities that do not yield an interpretable contrast, and the record does not explain those values.

The primary outcome sizes are 40 for AMG 714 and 19 for placebo, while the registry's overall enrollment is 60. The record does not provide a participant-flow explanation that reconciles the overall enrollment with the endpoint-specific groups. Safety denominators are different again, at 41 and 18. These distinct populations limit direct comparison across outcome summaries unless their definitions are kept in view.

The registry specifies that missing Week 24 F-VASI assessments count as non-response, but does not report how many assessments were missing or whether sensitivity analyses were performed. The full statistical analysis plan, the sample-size justification, and any interim analyses are not reported in the registry record. A multiplicity strategy for the many outcome measures and analyses is also not reported.

Although nbUVB phototherapy is included in the intervention list, its role and use in the trial arms are not described. The registry identifies the study as quadruple-masked but does not specify which roles were masked. Finally, the posted primary endpoint focuses on a single facial response threshold and time point. It does not, by itself, characterize response at other thresholds, later follow-up visits, or total-body vitiligo outcomes.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Evaluation of AMG 714 for Vitiligo — NCT04338581.
• Registry record last updated 2026-04-21; status last verified 2026-04.
• Results first posted 2025-11-17.
• 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.

20. Record Summary

NCT04338581 is a completed Phase 2a randomized, parallel-group, quadruple-masked, placebo-controlled trial titled Evaluation of AMG 714 for Vitiligo. The registry reports 60 enrolled participants and lists a study start date of 2020-12-11 and a primary completion date of 2024-10-16.

The primary endpoint was the proportion achieving at least 35% improvement from baseline in F-VASI at Week 24. The reported proportions were 0.15 for AMG 714 and 0.053 for placebo; the registry reports a Fisher Exact p-value of 0.411. The confidence intervals around both arm-level estimates are broad, and the recalculated contrast fields are not interpretable.

The primary analysis population included randomized participants who received at least one dose. Missing Week 24 assessments were classified as non-responses. The registry also lists secondary outcomes across F-VASI and T-VASI thresholds and follow-up times. Details of multiplicity, the full analysis plan, and the rationale for planned enrollment are not reported in the registry record.