Dupilumab Effect on Pruritus Neuro-mechanisms in Patients With Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dupilumab Effect on Pruritus Neuro-mechanisms in Patients With Atopic Dermatitis (NCT04823130), a non-randomized phase 4 study. The primary outcome is Change From Baseline in Intraepidermal Nerve Fiber Density on Lesional Skin at Week 17.

2. Clinical Question

3. Trial Design

This was a completed, multicenter, exploratory Phase 4 study, with a parallel-group, non-randomized design and no masking. The record lists an enrollment of 54. It gives a study start date of 2021-04-22 and a primary completion date of 2022-06-30. The sponsor was Sanofi.

The two groups had different roles. Participants with moderate-to-severe AD received dupilumab, while matched healthy participants received no treatment and served as controls. Because allocation was not randomized and the control group did not consist of untreated participants with AD, comparisons between these groups cannot isolate a treatment effect from differences associated with disease status or other group characteristics. Matching on selected characteristics does not make the groups equivalent in all respects.

The treatment schedule began with a 600 mg subcutaneous injection on Day 1. Injections of 300 mg followed every 2 weeks from Week 3 to Week 15. The primary endpoint assessed lesional-skin intraepidermal nerve fiber density at baseline and Week 17. The record also lists secondary outcomes covering nerve fiber density and branching, itch measures, AD severity, and quality-of-life or patient-reported measures at specified follow-up times.

The registry record does not report a sample-size justification, detailed statistical analysis plan, or interim-analysis plan. The study's exploratory description signals that the aim included examining a biological measure and related clinical outcomes; it does not, by itself, establish that the design can support a definitive comparative treatment conclusion.

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: Healthy participants were considered as a reference comparator group and received no treatment. Healthy participants underwent only an additional 7-days observational period following collection of the skin biopsy on Day 1 (i.e., up to Day 8). Only safety data was collected for the healthy participants and no other outcome measures were assessed.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analyzed on modified intent-to-treat (mITT) population which included all AD participants who received at least 1 dose of IMP who had at least 1 skin biopsy performed, irrespective of compliance with study protocol and procedures, and who did not use prohibited therapies for AD from screening to end of study. 'Overall number of participants analyzed' = participants with available data for this outcome measure and 'number analyzed' = participants with available data for each specified category.

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: Analysis was performed on enrolled population.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Intraepidermal Nerve Fiber Density on Lesional Skin at Week 17, measured over Baseline, Week 17.

The primary endpoint was change from baseline in intraepidermal nerve fiber density on lesional skin at Week 17. The measure was expressed as F/mm^2 and summarized as a mean with standard deviation. Skin biopsies were used. Nerve fibers were visualized by staining for the pan-axonal marker protein gene product 9.5, and the basement membrane was visualized by staining for collagen type 4. Density was quantified by counting nerve fibers crossing the basement membrane per square millimeter.

The endpoint provides a tissue-based measure of nerve architecture in affected skin. Its use reflects the study's aim of assessing neuronal features in people with AD and chronic pruritus, alongside evaluation of itch and disease-related outcomes. As general statistical context, a mean describes the average measured value in a group, while a standard deviation describes how dispersed individual measurements are around that average. A change-from-baseline endpoint focuses on how measurements differ over time within the defined analysis population; it does not, by itself, establish that treatment caused the change.

The primary endpoint was not planned to be collected and analyzed for the healthy control group. Numerical results for the endpoint, including the mean change and its uncertainty, are not reported in the registry record. Therefore, the endpoint definition can be described, but the size and precision of any observed change cannot be assessed from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary endpoint summarize?

The primary measure is continuous: it is a numerical density of nerve fibers per unit area. The registry describes the outcome as a mean and standard deviation. In general, the mean gives a group's average measurement, while the standard deviation describes the spread of measurements around that average. Neither summary gives the full shape of the data distribution. The registry record does not report the endpoint's numerical summaries, so the mean change and the variability cannot be interpreted.

What does change from baseline mean?

A change-from-baseline outcome compares a participant's follow-up measurement with that participant's baseline measurement, then summarizes those changes across the analysis population. This can account for the starting value in describing how measurements evolve. A group mean change remains a descriptive statistic unless the analysis and design support a causal comparison. In this study, the primary endpoint was assessed in the AD population, while collection and analysis of this endpoint were not planned for the healthy control group. There is therefore no corresponding primary-endpoint change in the healthy group specified for a direct comparison.

Why does the design matter to statistical interpretation?

Randomization is a general method for making treatment groups comparable, on average, with respect to measured and unmeasured characteristics. This study was non-randomized. Its healthy controls were matched to selected AD participants on site, age, gender, race, and skin-area location, but matching on selected characteristics cannot ensure balance on every factor that may affect skin measurements or symptoms. The groups also differed by AD status and treatment exposure. Thus, any between-group contrast would not have the same interpretation as a randomized comparison of treatment strategies in people with AD.

No masking was used. In general, knowledge of treatment or group assignment can influence subjective reporting, clinical assessment, or study conduct. The primary endpoint uses biopsy measurements, but the record does not describe assessor masking or other procedures that would establish how measurement assessments were protected from knowledge of group assignment. Those details are not reported in the registry record.

How should uncertainty be assessed?

Statistical uncertainty is commonly conveyed with a confidence interval and, depending on the question, a test statistic or p-value. A confidence interval communicates the range of values reasonably compatible with the data under a specified model and analysis. Its width reflects both the amount of information and variability in the measurements. A p-value addresses compatibility with a particular null hypothesis under model assumptions; it is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry record does not report a primary endpoint estimate, confidence interval, or p-value, so neither magnitude nor precision can be assessed.

The standard deviation listed in the endpoint definition is a descriptive measure of person-to-person spread, not the same as the uncertainty around the mean. General statistical context: standard errors and confidence intervals quantify uncertainty in estimates, whereas a standard deviation describes variation among observations. One cannot infer the endpoint's confidence interval from the outcome definition alone.

What do planning calculations mean?

Sample-size planning uses assumptions about a target effect, variability, significance level, allocation, and statistical method to estimate the number of participants needed for a planned analysis. The record describes planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such illustrations are not observed power and do not show that an effect was detected. They are hypothetical planning exercises. The registry record does not include the numerical observed effect in the endpoint results, so those assumptions should not be mistaken for a documented design justification or for evidence about the study's result.

Why can variability and model assumptions matter?

Continuous measurements can vary between participants and may not follow the distribution assumed by a particular analysis. General statistical context: when data are more variable than a model allows, uncertainty estimates can be too optimistic if that extra variability is ignored. This is sometimes described as over-dispersion, especially for count-like outcomes. The primary endpoint is described as a continuous density measure; the registry record does not report its distribution or a fitted model, so no claim about over-dispersion or model adequacy can be made.

Biopsy-based measures may also be affected by measurement procedures and by whether a follow-up sample is available. The endpoint description explains the staining and counting approach, but the record does not report measurement reliability, missing-biopsy handling, or sensitivity analyses. These omissions prevent assessment of how robust the estimate would be to measurement variation or incomplete observations.

What can be concluded from the registry analysis record?

The record defines the primary endpoint and analysis population but lists no analyses. It does not report the statistical model, covariate adjustment, handling of repeated measures, or a detailed plan for inference. These methods can affect the estimate and its uncertainty, particularly when there are multiple follow-up measures or incomplete data. Since these details are not reported in the registry record, readers should distinguish the endpoint definition from a fully described statistical analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 16 outcome measures. In addition to the primary Week 17 nerve fiber density endpoint, the listed secondary outcomes include nerve fiber density and branching at Weeks 3 and 21, itch measures, EASI and SCORAD severity scores, PROMIS-itch, POEM, and DLQI at specified follow-up times. Assessing multiple outcomes or time points creates a multiplicity issue: as the number of statistical tests increases, the chance of at least one apparently notable result can increase when all null hypotheses are true. General statistical approaches include prespecifying a hierarchy of outcomes or adjusting inference for multiple testing. The registry record does not report a multiplicity adjustment or a testing hierarchy.

Missing-data methods are also not reported in the registry record. The primary population definition requires at least 1 skin biopsy and excludes participants who used prohibited AD therapies from screening to the end of the study. The record does not describe how missing baseline or Week 17 measurements were handled, whether participants with incomplete follow-up were included in each analysis, or whether sensitivity analyses were conducted. These details matter because an analysis limited to participants with available measurements can be biased if missingness is related to outcome or treatment experience.

The primary population is described as a modified intent-to-treat population, including AD participants who received at least 1 dose, had at least 1 biopsy, and did not use prohibited therapies. The registry record does not report the number analyzed for the primary endpoint or the detailed analysis procedures. It also specifies that the primary endpoint was not planned for the healthy control group. These population and collection rules should be kept distinct from the broader enrollment figure and from the safety counts.

11. 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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pulmonary embolism (3.2%).

The registry lists serious adverse events for the healthy control group as 0 affected among 13 at risk, and for the AD dupilumab group as 1 affected among 31 at risk. It lists 0 deaths in both groups. Pulmonary Embolism is identified as a top serious term at 3.2%. The registry also reports other adverse events for the groups, with 0 affected among 13 healthy controls and 24 among 31 participants with AD receiving dupilumab.

These are reported counts, not estimates of a causal treatment effect. The study was non-randomized, and the groups differ in both disease status and treatment exposure. The record does not provide enough event-level detail to determine timing, severity beyond the category reported, relationship to treatment, or whether the serious term corresponds to the serious event count in a particular way. A count in a group does not establish that treatment caused the event; a zero count in a group does not establish that the risk is absent. The small group counts also make comparisons difficult to interpret. Safety findings should therefore be read as registry-recorded events in the listed groups, not as a controlled estimate of relative safety.

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

The study addresses a biological question: whether lesional-skin intraepidermal nerve fiber density changes from baseline to Week 17 in AD participants receiving dupilumab. It also lists secondary measures related to nerve architecture, itch, AD severity, and quality of life. These aims connect tissue measurement with symptoms and disease assessments, but the registry record does not report numerical outcome results. It is therefore not possible to state the direction, size, or precision of the primary change from the record.

The design is exploratory, non-randomized, and unmasked. The untreated healthy group offers a reference group, but it is not an untreated AD comparator. Accordingly, the design cannot by itself distinguish a treatment-associated change from changes over time, measurement variability, or differences between people with AD and healthy participants. Matching on selected characteristics does not remove all potential differences.

General statistical context: even a precisely estimated change would need to be interpreted in light of the design, endpoint's measurement properties, and clinical meaning. Statistical evidence about a measured change and evidence that a treatment caused that change are separate questions. A biological measurement also does not automatically establish a change in symptoms or patient experience. The registry lists related outcomes, but does not report their numerical results.

The safety record reports a serious adverse event in the dupilumab group and none in the healthy control group, along with no deaths in either group. These observations are important descriptions of what was recorded, but cannot establish causation or quantify a reliable comparative safety difference. The record also does not describe event-level timing or attribution.

In short, the registry record identifies a study question and measurement framework, and reports limited safety counts. It does not provide the outcome estimates needed to determine what the study observed for the primary endpoint or how precisely it estimated a change. The design and reporting limitations also preclude a definitive causal conclusion from group contrasts.

13. Why This Trial Matters Statistically

This trial is statistically notable because it combines a tissue-based endpoint with symptom and disease assessments in an exploratory study of AD and chronic pruritus. The primary measure quantifies nerve fibers crossing the basement membrane in lesional skin, rather than relying only on reported itch or clinical severity. Such measurements can help characterize a proposed biological process, while the secondary outcomes address related patient and disease measures.

The design also illustrates why a study's question and its comparison structure must be considered together. The healthy group was matched on selected characteristics, but was untreated and did not have AD. It can provide context for selected measurements, but it cannot serve as a randomized untreated AD group. The non-randomized, unmasked design therefore limits causal interpretation even if an association or change is reported elsewhere.

From a reporting perspective, defining an endpoint is not equivalent to reporting its result. The registry record names the primary outcome, measurement procedure, time point, and analysis population, but does not include numerical primary estimates or uncertainty. Without those elements, readers cannot evaluate effect size, precision, or statistical compatibility with different values. The record's list of multiple secondary outcomes also raises questions about multiplicity, while missing-data handling and a detailed statistical plan are not reported.

The study is therefore useful as an example of how mechanistic endpoints and clinical outcomes can be framed in a clinical trial record, and how design and reporting determine the strength of conclusions available to readers. Its registry entries support careful description of the planned question and reported safety counts, but not a quantitative assessment of the primary result.

14. 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.

Several limitations are specific to this study record. The study was non-randomized and unmasked, and the control group comprised healthy participants who received no treatment rather than untreated participants with AD. This makes the groups unsuitable as a like-for-like treatment comparison. Matching on site, age, gender, race, and skin-area location addresses selected characteristics only; the record does not establish comparability on other factors.

The primary endpoint was not planned for collection and analysis in the healthy group. Thus, the record does not describe a parallel healthy-control estimate for the primary outcome. In addition, numerical primary endpoint results, uncertainty measures, and detailed analysis methods are not reported in the registry record. The listed outcome measures include multiple endpoints and assessment times, but a multiplicity strategy is not reported.

The record defines a modified intent-to-treat population using treatment exposure, biopsy availability, and prohibited-therapy criteria. It does not report the number analyzed for the primary endpoint, the handling of missing measurements, or sensitivity analyses. It also does not report measurement reliability or assessor masking procedures for biopsy quantification. These omissions limit evaluation of selection, missing-data, and measurement effects.

Finally, the registry describes the study as exploratory and does not report a sample-size justification or detailed statistical analysis plan. The available safety counts do not include sufficient event-level information for causal attribution. These limitations do not negate the study's descriptive purpose, but they constrain how strongly its results could be interpreted even if numerical estimates were available.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Dupilumab Effect on Pruritus Neuro-mechanisms in Patients With Atopic Dermatitis — NCT04823130.
• Registry record last updated 2025-09-16; status last verified 2025-09.
• Results first posted 2023-07-11.
• 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.

18. Record Summary

The registry record describes a completed, multicenter, exploratory Phase 4 study of dupilumab in participants with moderate-to-severe AD and chronic pruritus, alongside an untreated healthy control group. The design was parallel, non-randomized, and unmasked. The primary endpoint was change from baseline in lesional-skin intraepidermal nerve fiber density at Week 17, measured from biopsies and expressed as F/mm^2. The primary analysis population was a defined modified intent-to-treat group of AD participants.

The record lists secondary outcomes involving nerve measures, itch, disease severity, and patient-reported measures. It reports group-level adverse-event counts, including 1 serious adverse event-affected participant in the dupilumab group and none in the healthy group, with no deaths listed. Numerical results for the primary endpoint, uncertainty estimates, multiplicity procedures, missing-data methods, and a detailed statistical analysis plan are not reported in the registry record. The study design does not support a like-for-like randomized comparison with the healthy controls.