A Study to Assess the Safety and Efficacy of Etrasimod in Subjects With Moderate-to-Severe Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Assess the Safety and Efficacy of Etrasimod in Subjects With Moderate-to-Severe Atopic Dermatitis (NCT04162769), a randomized phase 2 study. The primary outcome is Double-blind Treatment Period: Percent Change in Eczema Area and Severity Index (EASI) Score.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, triple-masked, parallel-group, placebo-controlled study sponsored by Arena Pharmaceuticals. The registry record gives the study status as completed and enrollment as 140. Its official title describes a multicenter, randomized, double-blinded, placebo-controlled 16-week study with a 52-week open-label extension. The posted primary endpoint, however, is explicitly measured from baseline to Week 12, and the double-blind treatment arms are labelled as a 12-week treatment period in their descriptions. The registry record does not explain the difference between the 16-week wording in the official title and the Week 12 outcome time point.

Participants in the double-blind period were assigned to etrasimod 1 mg, etrasimod 2 mg, or matching placebo. Triple masking means the registry classifies the study as masked at multiple levels, but the identities of the masked parties are not reported in the registry record. The allocation is identified as randomized; details of the randomization procedure, allocation concealment, stratification, and block structure are not reported in the registry record.

The study also included an open-label extension described as 52 weeks, with an etrasimod 2 mg arm. The posted safety data distinguish this extension period from the double-blind treatment period. The registry record does not report details of extension eligibility, timing of entry, continuation rules, or whether all participants in the extension had previously taken the same treatment.

Interpretation of the primary result should therefore center on the randomized, placebo-controlled comparison at Week 12. The extension is a different treatment setting and does not provide the same concurrent placebo comparison in the posted record. The study’s own detailed design assumptions, including its original target sample size rationale, 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.

Before assignment: A total of 140 participants were randomized into the 12-week Double-Blind Treatment Period and a total of 95 participants continued into the 52-Week OLE Period.

Double-blind Treatment (12-Week)

Reasons for not completing this period:

OLE Period (52-Week)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set 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: Full Analysis Set (FAS) included all randomized participants, irrespective of whether they received any study intervention.

6. Primary Endpoint

The registry lists the primary outcome as Double-blind Treatment Period: Percent Change in Eczema Area and Severity Index (EASI) Score, measured over Baseline (Day 1) and Week 12.

The primary endpoint is percent change in the Eczema Area and Severity Index score from baseline on Day 1 to Week 12. EASI combines assessments of clinical signs of atopic dermatitis with the extent of affected body surface area. The clinical signs described in the record are erythema, thickness-related changes, scratching-related changes, and lichenification. Assessments cover the head and neck, trunk, upper limbs, and lower limbs. Each region’s affected area contributes to the score, and the total EASI score ranges from 0 to 72, with higher scores indicating greater disease severity.

Percent change expresses how the score at follow-up differs relative to the baseline score. A negative percent change in the posted results indicates a reduction from baseline; it does not by itself establish that a difference between randomized groups is statistically clear or clinically important. The primary outcome is summarized as a least-squares mean, with standard error reported for each arm. Least-squares means are model-based group summaries, rather than simply the raw arithmetic means of individual observed changes.

EASI is used because it brings together the severity of several clinical signs and the area of skin affected into one outcome. That makes it possible to summarize changes in both dimensions in a single measure. The registry describes the endpoint and its scoring components, but the specific model covariates, handling of missing endpoint observations, and clinical threshold for an important difference are not reported in the registry record.

7. Results Posted in the Registry

• Double-blind Treatment Period: Etrasimod 1 Milligrams (mg): mean -58.68 versus -48.41 on control; difference -10.27 (95% CI -32.62 to 12.08), standardized effect d = -0.19; z-test p = 0.368.
• Double-blind Treatment Period: Etrasimod 2 mg: mean -57.18 versus -48.41 on control; difference -8.77 (95% CI -31.21 to 13.67), standardized effect d = -0.16; z-test p = 0.444.

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.

Double-blind Treatment Period: Percentage of Participants Achieving EASI-75

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Double-blind Treatment Period: Etrasimod 1 Milligrams (mg): 27.7% (13/47) versus 26.1% (12/46) on control; risk difference 1.6 percentage points (95% CI -16.4 to 19.6), risk ratio 1.06 (0.54–2.07); unadjusted two-proportion z-test p = 0.864.
  • Double-blind Treatment Period: Etrasimod 2 mg: 38.3% (18/47) versus 26.1% (12/46) on control; risk difference 12.2 percentage points (95% CI -6.6 to 31.0), risk ratio 1.47 (0.80–2.69); unadjusted two-proportion z-test p = 0.208.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind Treatment Period: Etrasimod 1 Milligrams (mg) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind Treatment Period: Etrasimod 2 mg and the control. The interval gives the range of effects that remain compatible with the data.

Double-blind Treatment Period: Percentage of Participants Achieving a Validated Investigator Global Assessment (vIGA) 0 or 1 Score and a Reduction From Baseline of >= 2 Points

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Double-blind Treatment Period: Etrasimod 1 Milligrams (mg): 14.9% (7/47) versus 13.0% (6/46) on control; risk difference 1.9 percentage points (95% CI -12.2 to 15.9), risk ratio 1.14 (0.42–3.14); unadjusted two-proportion z-test p = 0.797.
  • Double-blind Treatment Period: Etrasimod 2 mg: 29.8% (14/47) versus 13.0% (6/46) on control; risk difference 16.7 percentage points (95% CI 0.4 to 33.0), risk ratio 2.28 (0.96–5.43); unadjusted two-proportion z-test p = 0.049.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind Treatment Period: Etrasimod 1 Milligrams (mg) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Double-blind Treatment Period: Etrasimod 2 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Double-blind Treatment Period: Percent Change in Weekly Peak Pruritus Numerical Rating Scale (NRS) From an Itch Daily Diary

• 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:
  • Double-blind Treatment Period: Etrasimod 1 Milligrams (mg): mean -38.40 versus -24.37 on control; difference -14.03 (95% CI -37.21 to 9.16), standardized effect d = -0.31; z-test p = 0.236.
  • Double-blind Treatment Period: Etrasimod 2 mg: mean -35.04 versus -24.37 on control; difference -10.67 (95% CI -33.42 to 12.08), standardized effect d = -0.23; z-test p = 0.358.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind Treatment Period: Etrasimod 1 Milligrams (mg) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Double-blind Treatment Period: Etrasimod 2 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the assessment beyond the primary EASI percent-change measure. The registry lists response thresholds based on EASI, a validated investigator global assessment, itch-related measures, body surface area, and change in EASI during the open-label extension. The outcomes include binary measures, which classify participants by whether a threshold was reached, and continuous measures, which summarize the amount of change.

The posted results show a mixed pattern rather than a consistent separation across the displayed outcomes. The EASI-75 percentages are similar for the 1 mg and placebo arms, while the 2 mg estimate is higher; the confidence intervals for the group comparisons are broad. For the vIGA response, the 2 mg comparison has a reported nominal difference and p-value near the usual threshold, whereas the 1 mg comparison does not. The itch outcome estimates point toward more negative percent changes in both etrasimod arms than placebo, but the reported intervals include zero. These details are shown in the outcome tables and should be interpreted as secondary, supportive context rather than repeated as a separate set of conclusions.

This pattern does not replace the primary endpoint result. Secondary outcomes can help show whether an observed pattern is present across different measures, but they do not independently settle an inconclusive primary comparison. The measures also differ in what they capture: a threshold response is not equivalent to a mean change, and a patient-reported itch measure is not the same construct as an investigator-assessed skin score.

Multiple outcomes and comparisons create multiplicity: when many hypotheses are tested, the chance of at least one apparently notable result can increase. The registry record does not report whether or how the secondary tests were adjusted for multiplicity, or which secondary analyses were intended to be confirmatory. Accordingly, unadjusted secondary results carry less weight than a prespecified, multiplicity-controlled primary analysis. The posted record lists 17 outcome measures and 16 analyses, but does not report a complete multiplicity strategy. Results for secondary measures not detailed should not be inferred from the outcomes’ names alone.

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

• Double-blind Treatment Period: Percentage of Participants Achieving EASI-75
• Double-blind Treatment Period: Percentage of Participants Achieving a Validated Investigator Global Assessment (vIGA) 0 or 1 Score and a Reduction From Baseline of >= 2 Points
• Double-blind Treatment Period: Percent Change in Weekly Peak Pruritus Numerical Rating Scale (NRS) From an Itch Daily Diary
• Double-blind Treatment Period: Percentage of Participants With Improvement (Reduction) in Peak Pruritus NRS Greater Than or Equal to (>=)3 From an Itch Daily Diary
• Double-blind Treatment Period: Percentage of Participants Achieving EASI-50
• Double-blind Treatment Period: Percentage of Participants Achieving EASI-90
• Double-blind Treatment Period: Percent Change in Percent Body Surface Area (BSA)
• Open-label Extension (OLE) Period: Percent Change in EASI

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a continuous outcome: percent change in EASI. The posted arm summaries are least-squares means with standard errors, and the sponsor’s reported pairwise comparisons use ANCOVA. The page also presents unadjusted recalculations from the arm-level summaries. These approaches answer related but not identical questions, so their estimates and uncertainty should be read in context rather than combined as if they were one analysis.

What does ANCOVA contribute?

ANCOVA is a regression-based method that compares group outcomes while accounting for specified covariates. In a randomized trial, a model can use baseline information and other prespecified variables to estimate an adjusted group contrast. The result is commonly expressed as a least-squares mean difference. Adjustment can improve precision when the model and covariates are appropriate, but the exact covariates and model details are not reported in the registry record. The reported ANCOVA result is therefore the sponsor’s posted analysis, while the recalculation is a simpler summary-based comparison.

For the 1 mg comparison, the sponsor reports a least-squares mean difference of -10.28, a 95% confidence interval from -23.228 to 2.674, and p = 0.120. For the 2 mg comparison, the corresponding values are -8.77, -21.544 to 4.002, and p = 0.178. The negative differences are in the direction of a more negative EASI percent change relative to placebo. The confidence intervals include zero; the p-values do not provide strong evidence against a zero difference under the reported analyses.

Why show a recalculation as well?

A simple recalculation offers a transparent comparison based on the posted group summaries. It can help readers see how the arm estimates relate to a between-group contrast. It is not a reconstruction of the trial’s complete statistical model. The posted arm-level standard errors are used to derive approximate spread information for these recalculations, and the calculation is explicitly unadjusted. It does not account for model covariates or other features that may have informed the sponsor analysis. In this record, the recalculated estimates are -10.27 for 1 mg and -8.77 for 2 mg, with wider intervals than the sponsor’s reported intervals. The displayed p-values also differ. That is a reason to keep the analyses distinct, not to choose whichever value looks more favorable.

How should a confidence interval be understood?

In general, a confidence interval describes the uncertainty around an estimated contrast under the method used to construct it. Its width reflects the precision of that estimate. An interval that includes zero means the data and analysis do not rule out a zero contrast at the confidence level shown. It does not mean that zero is the only plausible value, nor does it establish that the groups are equivalent. The interval also does not directly express the probability that the true effect lies within its bounds.

The sponsor’s intervals are 95% confidence intervals. The recalculated intervals are also presented as 95% intervals. They differ because the calculations and assumptions differ. A model-based interval can be narrower or wider than a summary-based interval depending on the information and assumptions used; the registry record does not provide enough model detail to explain the exact difference in this case.

What does the p-value say?

In general, a p-value summarizes how incompatible the observed result is with a specified null hypothesis under a statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. The sponsor’s reported p-values are associated with its ANCOVA comparisons. The recalculated p-values are unadjusted and based on summary statistics. Neither type should be read without the corresponding estimate and interval.

Secondary endpoints include both binary outcomes, such as achievement of specified EASI response thresholds, and continuous outcomes, such as itch-score change. Binary outcomes are commonly summarized as proportions, while continuous outcomes are summarized using means or model-based means. The registry’s posted comparisons use methods including Cochran-Mantel-Haenszel analyses for binary outcomes and mixed-model analyses for one continuous itch outcome. The precise models, covariates, and missing-data procedures are not reported in the registry record.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variation in outcome data is greater than a model’s assumed variation. If a model underestimates variability, its standard errors may be too small and its intervals too narrow. If variability is modelled poorly in the other direction, precision may be understated. This is a general consideration for statistical modelling; the registry record does not report an over-dispersion assessment for the primary EASI analysis.

The endpoint summaries also need to be distinguished from participant-level observations. A group least-squares mean does not show the distribution of individual responses, and a standard error is not the same as a standard deviation. The arm-level summaries cannot show whether responses were concentrated around the mean or varied substantially across participants. The posted intervals and analyses are therefore the appropriate basis for discussing uncertainty, with the limitations of the registry-level information kept in view.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data locate the between-group estimate. For the primary endpoint, the sponsor’s 95% confidence interval for etrasimod 1 mg versus placebo ranges from -23.228 to 2.674. For etrasimod 2 mg versus placebo, it ranges from -21.544 to 4.002. Both intervals include zero and extend to values on either side of it. The estimates are therefore uncertain enough that the posted results do not distinguish a zero contrast from a range of possible differences.

The recalculated intervals are wider: -32.62 to 12.08 for the 1 mg comparison and -31.21 to 13.67 for the 2 mg comparison. They reflect unadjusted re-calculations from posted arm summaries, not the sponsor’s ANCOVA model. The difference in widths is a reminder that interval estimates depend on the analysis and its assumptions. These recalculations are useful as a supplementary view, not as replacements for the sponsor’s reported analysis.

In general, a confidence interval is a range of estimates that are reasonably compatible with the observed data under the method used. It is not a range of individual participant outcomes and does not show the probability that a particular value is true. A wide interval signals limited precision; it does not automatically mean that the study was poorly conducted. Precision depends on factors such as sample size, outcome variability, and analysis choices. The registry record does not report the original sample-size justification or enough model detail to attribute the observed width to any one cause.

It is also important not to interpret an interval containing zero as proof of no effect. The interval includes values consistent with no difference as well as values in both directions. Nor does a point estimate on one side of zero establish a definitive difference. Consider the estimate, interval, analysis method, endpoint meaning, and trial design together. The sponsor-reported and recalculated results both point to substantial uncertainty in the primary comparison.

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 sample-size display is a planning illustration based on the observed effect, not a reconstruction of the trial’s original design. Each row considers planning for a stated fraction of the observed effect, and the columns show participants per arm for 80% power and participants per arm for 90% power. These figures are per arm because this is a non-time-to-event outcome; the table does not show total event counts.

The stated planning 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 the stated fraction of the observed effect. The 80% power column describes the planned number of participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if that effect and model assumptions were correct. The 90% power column gives the corresponding planning quantity for 90% power. Neither column describes the achieved power of this completed trial.

The rows for 100%, 75%, and 50% of the observed effect illustrate how the planning requirement changes as the target effect changes. Smaller effects need more participants to detect under the same assumptions. For example, the displayed primary-outcome planning figures increase across the rows as the target fraction becomes smaller. The exact values are shown in the sample-size table; they should be read as scenario-based planning quantities, not as evidence that the trial was designed around those effects.

This is not observed power. The trial’s own original sample-size justification, target effect, and planning assumptions are not reported in the registry record. The illustration is therefore useful for understanding how assumptions about effect size relate to planning, but it cannot establish whether the completed study met its original design target or whether a result should be considered conclusive.

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 = -10.270; pooled standard deviation σ = 54.937.
2. n = 2 × (1.96 + 0.842)² × 54.937² / -10.270² = 449.2, rounded up to 450 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 <- -10.27                      # observed difference in means
sigma <- 54.9373345631284                     # 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(4162769)
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 = 450 (80% power), 602 (90% power)
 75% of observed effect: n per arm = 799 (80% power), 1,070 (90% power)
 50% of observed effect: n per arm = 1,797 (80% power), 2,406 (90% power)
power.t.test: 450.2 
Simulated power at n = 450 per arm: 0.801 (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’s reported primary comparisons use ANCOVA and give least-squares mean differences, 95% confidence intervals, and p-values. Those are the registry’s own posted analyses. The recalculations use the group-level summaries to form simple, unadjusted comparisons. The recalculated estimates are close in direction and, for the 2 mg comparison, the same stated value to the displayed precision, but the intervals and p-values differ from the sponsor’s results.

For etrasimod 1 mg, the sponsor reports -10.28 with a 95% interval from -23.228 to 2.674 and p = 0.120; the recalculation is -10.27 with an interval from -32.62 to 12.08 and an unadjusted p-value of 0.368. For etrasimod 2 mg, the sponsor reports -8.77 with an interval from -21.544 to 4.002 and p = 0.178; the recalculation is -8.77 with an interval from -31.21 to 13.67 and an unadjusted p-value of 0.444.

These are not competing versions of an identical analysis. The sponsor’s method is model-based and the recalculation is unadjusted; the registry does not provide enough detail to reproduce the model. The discrepancy flag is false, but that does not make the analyses interchangeable. The sponsor’s posted analysis should be identified as the trial’s reported inferential result, while the recalculation is a supplementary check based on summary statistics. Neither provides a basis for ignoring the uncertainty shown by its interval.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a complete statistical analysis plan, so several features needed to evaluate the inferential framework are unavailable. In particular, the planned handling of multiplicity across the primary and secondary outcomes is not reported in the registry record. The record lists multiple outcome measures and analyses, including more than one treatment-versus-placebo comparison. When many outcomes or comparisons are examined, some results can appear notable by chance even if there is no underlying difference. In general, multiplicity procedures control or describe this risk, but the registry does not state which procedure, if any, was used for the posted secondary analyses.

The reported secondary p-values should therefore be treated cautiously, especially where they are not accompanied by a reported adjustment strategy. The primary outcome remains the clearest focus for the trial’s main question, but the registry record does not provide enough information to confirm the hierarchy of endpoints, testing sequence, or whether the primary p-value was protected from other testing decisions.

Missing-data handling is also not reported in the registry record. The primary analysis is identified as the Full Analysis Set Population, but that label does not by itself explain how missing Week 12 measurements, discontinuations, or other incomplete observations were handled. The record does not describe imputation, model assumptions for missing observations, sensitivity analyses, or how participants with no post-baseline result were treated. These details can matter because different approaches rely on different assumptions about the unobserved outcomes.

The posted primary arm counts are 47, 47, and 46, and the registry gives enrollment as 140. The record does not explain the relationship between those analysis counts and all randomized or treated participants. It also does not identify the analysis populations for every secondary endpoint in the information. The primary population label should not be assumed to apply to all outcomes without an explicit statement.

Interim analyses, stopping boundaries, protocol deviations, and a detailed sample-size justification are not reported in the registry record. These omissions do not demonstrate that such procedures were or were not used; they mean that they cannot be assessed from the posted information. This limits how fully readers can evaluate prespecification, robustness, and the effect of missing observations on the reported estimates.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: extradural abscess (1.1%), staphylococcal bacteraemia (1.1%), tooth abscess (1.1%), pneumonia (1.1%), sepsis (1.1%).

The serious adverse event summary reports zero affected participants in each double-blind treatment arm: etrasimod 1 mg, etrasimod 2 mg, and placebo. In the open-label extension arm, the record reports 3 affected participants among 95 at risk. The posted death count is zero in each of the listed groups. These are registry counts for the stated periods; they do not establish whether an event was caused by treatment.

The record also lists other adverse events affecting 19 participants in the 1 mg double-blind arm, 31 in the 2 mg arm, and 22 in the placebo arm. In the extension arm, 53 participants were reported as affected. These counts are separate from the serious-event figures and should not be conflated with them. The registry lists serious-event terms including extradural abscess, staphylococcal bacteraemia, tooth abscess, pneumonia, sepsis, and arthropod bite, each shown as 1.1%. The record does not establish from these listings alone that any event was treatment-related.

Safety comparisons require context about event definitions, observation time, participant exposure, ascertainment, and the handling of repeated events. Those details are not fully. The extension also differs from the randomized, placebo-controlled treatment period, so its counts should not be interpreted as a direct comparison with the placebo arm. Small trial groups and event counts limit the conclusions that can be drawn about uncommon harms.

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

The posted primary outcome describes change in EASI from baseline to Week 12. All three arms have negative least-squares mean percent changes, and the sponsor’s treatment-versus-placebo estimates are also negative. On this scale, that direction corresponds to a greater reduction from baseline in the etrasimod arms than in placebo. However, the sponsor’s 95% confidence intervals include zero for both dose comparisons, and the recalculated unadjusted intervals are wider and also include zero. The posted evidence therefore leaves substantial uncertainty about the size and direction of the between-group differences.

This result does not establish that the groups are identical. A confidence interval that crosses zero can include a range of potentially meaningful differences in either direction. It also does not demonstrate that an apparent difference is clinically important. The registry record does not report a threshold for clinical importance, and the statistical estimates alone do not supply one.

The secondary outcomes provide additional descriptions of disease signs, investigator assessment, itch, and other measures, but the registry does not report a complete multiplicity strategy. A nominally notable secondary comparison should therefore be treated as exploratory context, not as confirmation that the primary question has been resolved. Nor should outcomes with no detailed results in this summary be assumed to support a particular conclusion.

The safety counts describe events recorded in the listed groups and periods. They are not proof that an event was caused by the study drug, and zero recorded serious events in a group does not demonstrate absence of risk. The open-label extension lacks a posted placebo comparator and should be read separately from the randomized double-blind comparison.

Overall, the page supports a careful account of what the registry reports: group-level changes, model-based contrasts, uncertainty, secondary outcome patterns, and safety counts. It does not support medical advice, a decision for an individual, or a broad conclusion about treatment effectiveness beyond the outcomes and follow-up described in this record.

18. Why This Trial Matters Statistically

This trial is statistically informative because it places the primary outcome in a randomized, masked, placebo-controlled comparison rather than relying only on change within treated groups. The distinction matters: participants in all arms had negative mean percent changes, so within-arm change alone cannot separate treatment-period patterns from changes that also occur with placebo or from other influences.

The posted primary estimates illustrate why effect estimates and uncertainty belong together. Both sponsor-reported contrasts point in the same direction, but their confidence intervals include zero. Reporting only the point estimates would make the comparisons seem more definite than the intervals support. The simple recalculations also show that a comparison derived from arm summaries can have different uncertainty from the sponsor’s model-based ANCOVA result.

The outcome set includes both continuous measures and response thresholds. That variety can describe different aspects of atopic dermatitis, but it also makes a prespecified analysis hierarchy and multiplicity approach important. The registry record does not report enough detail to determine how those issues were handled. This is a practical reminder that transparent reporting of analysis plans, missing-data procedures, and endpoint relationships is part of interpreting trial results, not an administrative extra.

The sample-size illustrations offer a further statistical lesson: planning requirements depend strongly on the effect a study aims to detect. Smaller target effects require more participants under the stated assumptions. Those illustrations are not the original design rationale and do not reveal achieved power. The record’s limited reporting of original design assumptions means readers should not reverse-engineer the intended target from the observed results.

Finally, the extension and safety summaries show why study periods and comparisons must be kept distinct. An open-label arm can contribute descriptive follow-up information, but it is not the same as a concurrent randomized placebo comparison. Reading the trial with those distinctions intact helps avoid overinterpreting either positive-looking secondary results or sparse event counts.

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.

Several limitations specific to this registry record constrain interpretation. The official title describes a 16-week double-blind study, while the posted primary endpoint is measured at Week 12 and the arm descriptions refer to a 12-week double-blind treatment period. The record does not explain this difference. The primary finding should be tied to the stated Week 12 endpoint rather than assuming that the posted result summarizes every part of the study period.

The primary group sizes are modest, and the reported confidence intervals are broad. That makes the treatment contrasts imprecise. The record does not report the trial’s original sample-size assumptions or justification, so the posted planning illustrations cannot be used to judge whether the study met its original target.

The endpoint analysis is identified as using the Full Analysis Set Population, but the record does not describe the analysis population in more detail or explain how missing observations were handled. The exact ANCOVA covariates and model specification are also not reported. This prevents independent reconstruction of the sponsor’s analysis from the registry summary. The summary-based recalculations are necessarily unadjusted and use approximate spread information derived from the posted standard errors.

Many outcomes and comparisons are listed, but the registry record does not report a complete multiplicity strategy or testing hierarchy. That limits the weight that can be assigned to nominal secondary findings. The detailed results shown for some secondary outcomes should not be taken as a complete account of all listed measures, and no results should be inferred for outcomes whose results are not present in the record.

The extension is open-label and its listed arm receives etrasimod 2 mg. The registry summary does not report a concurrent placebo comparison for that period, nor does it provide enough detail about entry and follow-up to make it directly comparable with the double-blind phase. Finally, safety counts are descriptive and do not identify event causality. These limitations do not erase the posted results, but they define what those results can reasonably support.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Assess the Safety and Efficacy of Etrasimod in Subjects With Moderate-to-Severe Atopic Dermatitis — NCT04162769.
• Registry record last updated 2022-11-04; status last verified 2022-10.
• Results first posted 2022-11-04.
• 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 record describes a completed phase 2 randomized, triple-masked, parallel-group, placebo-controlled study sponsored by Arena Pharmaceuticals, with enrollment of 140. The double-blind comparison included etrasimod 1 mg, etrasimod 2 mg, and matching placebo. The primary endpoint was percent change in EASI from baseline to Week 12 in the Full Analysis Set Population.

The sponsor’s ANCOVA comparisons reported negative least-squares mean differences for both etrasimod doses versus placebo, with 95% confidence intervals that included zero. Recalculations from posted arm summaries are unadjusted and also show substantial uncertainty. Secondary outcomes cover additional skin, investigator, and itch measures; the registry does not report a complete multiplicity strategy. Serious adverse event counts were zero in the three double-blind groups and 3 among 95 participants in the listed open-label extension group. Missing-data procedures, detailed model specifications, and the original sample-size rationale are not reported in the registry record.