A Study to Evaluate Safety and Effectiveness of Cendakimab (CC-93538) in Participants With Moderate to Severe Atopic Dermatitis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate Safety and Effectiveness of Cendakimab (CC-93538) in Participants With Moderate to Severe Atopic Dermatitis (NCT04800315), a randomized phase 2 study. The primary outcome is Mean Percentage Change From Baseline in EASI at Week 16.

2. Clinical Question

The primary question concerns a continuous measure of atopic dermatitis severity, rather than a responder threshold. EASI combines the extent of affected skin across body regions with the intensity of signs of atopic dermatitis. A negative percentage change indicates a reduction from baseline in the measured score; the registry describes lower EASI scores as better. The posted group means are all negative, and the sponsor’s reported mean differences are also negative for the comparisons with placebo. Those directions describe the posted outcome values; they do not, by themselves, establish how certain or clinically important a difference is.

The study also reports secondary outcomes, including EASI responder measures, vIGA-AD response, pruritus, SCORAD, and body-surface-area change. Those outcomes address related but distinct aspects of disease severity or response. The primary comparison remains the mean percentage change in EASI at week 16, and the secondary findings should be read in that context rather than as interchangeable measurements.

3. Trial Design

The registry describes a phase 2, multicenter, global, randomized, double-blind, placebo-controlled, parallel-group study in adults with moderate to severe atopic dermatitis. Masking is recorded as quadruple. The study status is completed. The registry reports enrollment of 221, with an intervention period of 16 weeks in the listed cendakimab regimen descriptions.

There are three cendakimab treatment groups and a placebo group in the posted primary outcome. The schedule descriptions include cendakimab given subcutaneously once weekly, cendakimab given every other week with placebo on alternating weeks, and cendakimab given every other week alongside weekly matching placebo. The placebo group is described as receiving placebo subcutaneously each week. These schedules use placebo in different ways, including to maintain masking. The registry lists the primary outcome population as the modified Intent to Treat population.

The primary endpoint is the mean percentage change from baseline in EASI at week 16. The posted arm results report a mean and standard error for each group. The registry also reports sponsor analyses comparing each treatment group with placebo using ANCOVA and giving a mean difference, a 95% confidence interval, and a p-value. Recalculated comparisons based on the posted summary statistics are also available, but they are simple, unadjusted re-calculations rather than the sponsor’s analysis.

The record does not report the original sample-size justification, the protocol’s full statistical analysis plan, or details of interim analyses. It also does not report the precise masking procedures or the covariates used in the sponsor’s ANCOVA. Those details should not be inferred from the study title or from the analysis results. The registry’s description of randomized allocation and masking identifies important design features, but does not by itself supply all the information needed to assess conduct, missing-data handling, or the full analysis strategy.

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: 221 participants Randomized, 220 Participants Treated

Randomization

Reasons for not completing this period:

Treatment Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Modified Intent to Treat (mITT) Population

5. Baseline Characteristics

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

Population: All Randomized

6. Primary Endpoint

The registry lists the primary outcome as Mean Percentage Change From Baseline in EASI at Week 16, measured over From initial EASI measurement to week 16.

The primary endpoint is the mean percentage change from baseline in EASI at week 16. EASI, the Eczema Area and Severity Index, is an investigator-assessed composite measure. It considers the proportion of four body regions affected by atopic dermatitis and the intensity of four signs: erythema, induration or papulation, excoriation, and lichenification. The registry describes the intensity ratings on a scale from 0, none, to 3, severe, and states that the total score ranges from 0 to 72. Lower scores indicate less measured disease severity.

A percentage change expresses change relative to the baseline EASI measurement, so it can help describe the magnitude of change across participants whose starting scores may differ. The endpoint is continuous: it summarizes the group’s mean change rather than classifying each participant as a responder or nonresponder. The registry reports the mean and standard error for each group. The standard error describes the statistical uncertainty around an estimated group mean; it is not the spread of individual participants’ outcomes.

EASI is used because it combines affected area and signs of severity into a single investigator-assessed measure. This makes it possible to compare average changes between randomized groups. The endpoint does not capture every aspect of atopic dermatitis, and a group mean does not show how outcomes were distributed among individuals. The registry’s secondary outcomes include threshold-based EASI responses and measures of other symptoms or signs, which complement but do not replace the primary endpoint.

7. Results Posted in the Registry

• Treatment 1: mean -84.41 versus -62.65 on control; difference -21.76 (95% CI -36.46 to -7.06), standardized effect d = -0.55; z-test p = 0.004.
• Treatment 2: mean -76.03 versus -62.65 on control; difference -13.38 (95% CI -26.99 to 0.23), standardized effect d = -0.36; z-test p = 0.054.
• Treatment 3: mean -78.93 versus -62.65 on control; difference -16.28 (95% CI -30.29 to -2.27), standardized effect d = -0.43; z-test p = 0.023.

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.

Percentage of Responders With an vIGA-AD Score of 0 (Clear) or 1 (Almost Clear) and a Reduction ≥ 2 Points From Baseline at Week 16

• 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:
  • Treatment 1: 33.3% (18/54) versus 8.9% (5/56) on control; risk difference 24.4 percentage points (95% CI 9.8 to 39.0), risk ratio 3.73 (1.49–9.35); unadjusted two-proportion z-test p = 0.002.
  • Treatment 2: 23.6% (13/55) versus 8.9% (5/56) on control; risk difference 14.7 percentage points (95% CI 1.2 to 28.2), risk ratio 2.65 (1.01–6.93); unadjusted two-proportion z-test p = 0.036.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment 1 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment 2 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of EASI-75 Responders at Week 16

• 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:
  • Treatment 1: 50.0% (27/54) versus 26.8% (15/56) on control; risk difference 23.2 percentage points (95% CI 5.5 to 40.9), risk ratio 1.87 (1.12–3.10); unadjusted two-proportion z-test p = 0.012.
  • Treatment 2: 49.1% (27/55) versus 26.8% (15/56) on control; risk difference 22.3 percentage points (95% CI 4.7 to 39.9), risk ratio 1.83 (1.10–3.05); unadjusted two-proportion z-test p = 0.015.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment 1 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment 2 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of EASI-90 Responders at Week 16

• 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:
  • Treatment 1: 31.5% (17/54) versus 14.3% (8/56) on control; risk difference 17.2 percentage points (95% CI 1.8 to 32.6), risk ratio 2.20 (1.04–4.68); unadjusted two-proportion z-test p = 0.031.
  • Treatment 2: 23.6% (13/55) versus 14.3% (8/56) on control; risk difference 9.4 percentage points (95% CI -5.1 to 23.8), risk ratio 1.65 (0.74–3.68); unadjusted two-proportion z-test p = 0.209.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Treatment 1 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment 2 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several related dimensions: investigator-rated disease severity, EASI response thresholds, pruritus, SCORAD, and body-surface-area change. Across the posted responder outcomes, the treatment-group percentages generally sit above the placebo percentage, while the continuous symptom and severity measures provide complementary ways to describe change. This broad pattern is directionally consistent with the primary group means, but secondary measures do not establish or replace the primary comparison.

There are 12 posted secondary outcomes. The registry includes results and comparisons for some of them, while its summary list also names outcomes for which no result details appear in the available record. Missing details for an outcome should be read as not reported in the registry record, not as evidence that no effect occurred. The number and variety of outcomes also mean that the findings need to be interpreted as a set rather than selecting only those that appear most favorable.

Some secondary comparisons have sponsor-reported analyses, while the recalculated results are unadjusted. The registry does not report an overall multiplicity strategy for the secondary outcomes. In general, testing many outcomes increases the chance that some comparisons appear statistically persuasive by chance, even when each test is considered separately. This is why unadjusted secondary p-values carry less weight than a prespecified primary analysis and why consistency, effect size, interval width, and the outcome’s role in the study all matter.

The secondary outcomes can help assess whether the primary pattern is reflected in other measures, but they cannot resolve the limitations of the primary analysis or establish a definitive hierarchy among endpoints. The analysis plan specifying which secondary outcomes were confirmatory, how they were ordered, and whether any adjustment was planned is not reported in the registry record. Accordingly, the secondary findings are best treated as supportive descriptions of the posted data, not as independent proof of a treatment effect.

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

• Percentage of Responders With an vIGA-AD Score of 0 (Clear) or 1 (Almost Clear) and a Reduction ≥ 2 Points From Baseline at Week 16
• Percentage of EASI-75 Responders at Week 16
• Percentage of EASI-90 Responders at Week 16
• Percent Change in Mean SCORAD Scores From Baseline at Week 16
• Percent Change From Baseline in Pruritus NRS at Week 16
• Percentage of Participants With a Response and Pruritus NRS Change of ≥ 4 Points From Baseline at Week 16
• Time to Achieve at Least 4 Points of Improvement in the Severity of Pruritus NRS Scale.
• Adjust Mean Percentage Change in BSA in Atopic Dermatitis From Baseline at Week 16

9. Statistical Interpretation

10. Statistical Methods Explained

What does a mean difference represent?

The primary outcome is continuous, so its group results are expressed as means. A mean summarizes the average percentage change in EASI for participants included in the analysis. A mean difference compares the treatment-group mean with the placebo-group mean. For this endpoint, a negative difference corresponds to a more negative mean percentage change in the treatment group than in placebo. It is a comparison of group averages, not a statement that every participant had that experience.

The registry gives standard errors alongside the arm means. In general, a standard error describes how precisely a sample mean estimates the corresponding population mean under the analysis assumptions. It is different from a standard deviation, which describes how spread out individual observations are. The posted arm results do not provide confidence limits around each arm mean, while the sponsor reports confidence intervals for the between-group comparisons.

What is ANCOVA?

The sponsor identifies ANCOVA as the method for the primary comparisons and reports mean differences in final values, 95% confidence intervals, and p-values. In general, analysis of covariance compares groups while accounting for one or more covariates in a statistical model. The registry record does not report which covariates were included, how the model was specified, or how missing observations were treated. It is therefore not possible to reconstruct the sponsor’s analysis from the method label alone.

The recalculated primary comparisons are described as simple unadjusted re-calculations from the posted summary statistics. They reproduce the reported mean differences but do not reproduce the sponsor’s full ANCOVA analysis. Their confidence intervals and p-values differ modestly from the sponsor-reported values. Because the discrepancy flag is true, the sponsor’s own analysis takes precedence when describing the trial’s inferential results.

How should the p-values be read?

A p-value summarizes how compatible the observed data are with a specified null hypothesis under the assumptions of the statistical test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The sponsor reports p-values for the primary comparisons, and the recalculated values are labeled unadjusted. The record does not state whether the sponsor’s p-values were adjusted for multiple comparisons; that information is not reported in the registry record.

In general, a confidence interval describes a range of values compatible with the data and the statistical model at the stated confidence level. For a difference, an interval that includes zero does not rule out no difference under that interval procedure. An interval that lies entirely on one side of zero indicates that the model-based interval does not include zero. These conventions do not establish clinical importance and do not remove the need to consider the analysis plan and multiplicity.

What do the binary secondary outcomes measure?

Some secondary outcomes classify participants according to whether they met a response definition, such as an EASI threshold or a vIGA-AD score criterion. For a binary outcome, the posted arm percentages describe the proportion meeting the definition in each group. A risk difference expresses the difference between two proportions in absolute terms. For example, a positive risk difference means that the event proportion is higher in the treatment group than in the comparator group, given the ordering used for that comparison. The recalculated secondary comparisons are unadjusted, and the registry also posts sponsor analyses for some outcomes.

For general interpretation, absolute differences and relative measures answer different questions. An absolute difference describes the separation between proportions; relative measures compare proportions as a ratio. The secondary results include recalculated risk differences and, for some comparisons, relative measures. These are not interchangeable and can convey different aspects of the same data. The posted secondary results should be considered exploratory unless the analysis plan establishes otherwise; that plan is not reported in the registry record.

Why do the analysis population and missing-data details matter?

The primary outcome is identified as using the modified Intent to Treat population. That label identifies the stated analysis population, but the registry record does not define the modification, list exclusion rules, or explain how participants with missing week 16 measurements were handled. Those details can affect which observations contribute to group estimates and the interpretation of the comparison. The record also does not report whether sensitivity analyses assessed alternative missing-data assumptions.

In general, missing-data methods rely on assumptions about why measurements are absent and what unobserved values might have been. Different assumptions can change estimates or their uncertainty. Without the protocol and analysis details, it is not possible to assess whether the approach was robust to missing outcomes. The analysis population label should therefore be reported as stated, without assuming it is equivalent to a full randomized population or to another commonly used analysis set.

11. Confidence Intervals and Precision

Confidence intervals show the uncertainty around an estimated comparison under the statistical model used. They are more informative than a p-value alone because they indicate both the direction and range of effect sizes compatible with the analysis. A 95% confidence interval is not a guarantee that the true effect lies within the interval; in general, it describes the long-run coverage of the interval procedure when its assumptions hold.

For Treatment 1 versus placebo, the sponsor reports a mean difference of -21.76 and a 95% confidence interval from -35.81 to -7.70. For Treatment 3 versus placebo, the reported mean difference is -16.28, with a 95% confidence interval from -30.88 to -1.68. These intervals lie below zero, though they cover a range of possible differences rather than a single precise value. For Treatment 2 versus placebo, the mean difference is -13.38 and the 95% confidence interval runs from -27.19 to 0.43. It includes zero, so the sponsor’s interval does not distinguish that estimated difference from zero at the stated confidence level.

The recalculated intervals differ from the sponsor’s intervals, which is why the sponsor’s results take precedence. The posted standard errors for the arm means also give context for the uncertainty in the group estimates, but they are not the same as confidence intervals for treatment contrasts. The registry reports group sizes between 54 and 56 for the primary outcome arms. Those sizes and the standard errors help describe the amount of information in the posted estimates, but do not eliminate uncertainty.

Precision is not the same as importance. A relatively narrow interval may still include effects of different practical relevance, and a result whose interval excludes zero is not automatically clinically important. The registry record does not provide a threshold for a clinically meaningful EASI difference. It also does not report multiplicity adjustment for the three primary comparisons, which affects how the individual interval and p-value results should be interpreted as a family of tests.

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 figures are planning illustrations based on the observed effects, not the original enrollment rationale for this trial. For these comparisons, the figures are participants per arm. The registry record does not report the trial’s own original sample-size assumptions or justification. The planning assumptions listed are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect used for each illustration is the stated fraction of the observed effect.

Each row indicates how many participants per arm would be needed under those assumptions to target the stated effect and power. The 80% power column represents a design that, under the planning assumptions and if the target effect and model assumptions hold, would have an 80% chance of rejecting the null hypothesis at the stated significance level. The 90% power column uses a higher target chance of detecting that same assumed effect. These are prospective planning interpretations, not descriptions of what happened in this trial.

The rows for 100%, 75%, and 50% of the observed effect illustrate how the required per-arm sample size changes when the target effect is smaller. A smaller assumed effect requires more participants to achieve the same target power under otherwise unchanged assumptions. For example, the posted primary comparison planning figures increase across those rows. The figures are specific to each comparison and should not be treated as universal requirements for another trial.

This is not observed power. Observed power is not a useful substitute for reading the estimated effect and its confidence interval after a study has been completed. The planning figures instead illustrate how the observed effect, or a fraction of it, would translate into a hypothetical future design under the stated assumptions. They do not establish that the original trial was designed using these assumptions, and they do not account for details absent from the registry record, such as its original analysis plan or missing-data assumptions.

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 = -21.760; pooled standard deviation σ = 39.412.
2. n = 2 × (1.96 + 0.842)² × 39.412² / -21.760² = 51.5, rounded up to 52 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 <- -21.76                      # observed difference in means
sigma <- 39.4119501289008                     # 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(4800315)
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 = 52 (80% power), 69 (90% power)
 75% of observed effect: n per arm = 92 (80% power), 123 (90% power)
 50% of observed effect: n per arm = 206 (80% power), 276 (90% power)
power.t.test: 52.5 
Simulated power at n = 52 per arm: 0.790 (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 own primary analyses use ANCOVA and report the mean difference in final values, a 95% confidence interval, and a p-value for each cendakimab group compared with placebo. Those analyses are the relevant reported inferential results. The record also contains simple unadjusted re-calculations from posted arm summary statistics. They reproduce the sponsor’s mean differences but yield different intervals and p-values.

For Treatment 1, the sponsor reports -21.76, a 95% confidence interval from -35.81 to -7.70, and p = 0.003; the recalculated p-value is 0.004. For Treatment 2, the sponsor reports -13.38, a 95% confidence interval from -27.19 to 0.43, and p = 0.057; the recalculated p-value is 0.054. For Treatment 3, the sponsor reports -16.28, a 95% confidence interval from -30.88 to -1.68, and p = 0.029; the recalculated p-value is 0.023.

The direction and broad pattern are similar, but the numerical differences mean the recalculations should not be substituted for the sponsor’s results. The discrepancy flag is true, and the sponsor’s own analysis takes precedence. The registry record does not describe the ANCOVA covariates or other analysis details needed to explain precisely why the estimates’ uncertainty and p-values differ. Those details should be treated as not reported in the registry record rather than inferred.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint compares three cendakimab groups with a shared placebo group. The registry reports separate sponsor p-values and confidence intervals for those comparisons, but a multiplicity procedure is not reported in the registry record. In general, when several hypotheses are tested, the probability of at least one chance finding across the family can increase. Procedures such as controlling a family-wise error rate or using a prespecified testing sequence can address that issue, but no such approach should be attributed to this trial without registry documentation.

The same concern applies to the secondary outcomes, which span several different measures and response definitions. The registry lists 13 outcome measures overall and 12 posted secondary outcomes. The analysis hierarchy, any prespecified secondary testing order, and any adjustment for multiple secondary comparisons are not reported in the registry record. Individual unadjusted results therefore require caution, especially when considered alongside many other outcomes.

The primary analysis population is identified as the modified Intent to Treat population. The registry record does not explain how that population was defined, whether all randomized participants were included, or what exclusions were made. The number and reasons for missing week 16 EASI measurements, the imputation method, and sensitivity analyses for missing data are also not reported in the registry record.

These omissions matter because analysis population definitions and missing-data assumptions can affect estimated group differences and their uncertainty. General statistical practice distinguishes between the randomized population, the population with available measurements, and a protocol-defined analysis set; the record’s label alone does not show how these relate in this trial. Without the protocol and full statistical analysis plan, the risk of bias from exclusions, missing measurements, or selective emphasis cannot be fully evaluated. No claim about interim analyses or the original analysis plan is supported by the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cellulitis (1.9%), pulmonary hypertension (1.9%), deep vein thrombosis (1.9%), covid-19 (1.8%), device dislocation (1.8%).

The registry lists serious adverse events affecting 2 participants in Treatment 1, 1 in Treatment 2, 2 in Treatment 3, and 4 in the placebo group. It reports no deaths in any of the four groups. These counts describe participants recorded as affected by a serious adverse event; they are not counts of every event or a complete account of all safety outcomes.

The record also lists the most frequent serious-event terms as cellulitis and pulmonary hypertension at 1.9% each, and deep vein thrombosis, COVID-19, device dislocation, and angioedema at 1.8% each. The listed terms should not be attributed to a particular group without group-specific information. The registry summary does not provide enough detail to assess event timing, severity beyond the serious-event classification, relatedness to study treatment, or how events were adjudicated.

The counts are descriptive and do not establish that the groups have equivalent safety profiles. Differences in observed counts can reflect uncertainty, and the record does not supply a formal comparative safety analysis. The registry also reports other adverse events affecting participants, but these summaries do not replace a complete review of event types and clinical context. Safety interpretation should therefore remain limited to the posted counts and descriptions.

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

The primary result describes average EASI change over the study’s 16-week period in the modified Intent to Treat population. The posted means are negative in each group, including placebo, and the sponsor reports between-group estimates that are negative for all three treatment comparisons. The sponsor’s intervals exclude zero for Treatment 1 and Treatment 3, while the interval for Treatment 2 includes zero. These are statistical descriptions of the registry results; they do not by themselves establish how important the observed differences are to participants.

The result does not show that every participant improved, that the average change applies to every adult with atopic dermatitis, or that the differences persisted beyond week 16. Group means can conceal variation in individual outcomes, and the registry record does not report longer-term results in the registry record described. The primary endpoint measures EASI, not every symptom, functional outcome, or experience relevant to participants. Secondary measures broaden the outcome picture but do not remove those limits.

The result also does not establish that cendakimab is safe or unsafe. Serious adverse event counts are descriptive, and no comparative safety conclusion can be drawn from those counts alone. The record does not provide details about event relatedness, complete safety analyses, or missing outcome handling sufficient for broader conclusions.

Finally, the registry does not report the original sample-size rationale, full statistical analysis plan, missing-data approach, or multiplicity procedure. The sponsor’s ANCOVA results take precedence over the simple recalculations, but the absent model details limit independent appraisal. The findings are best understood as estimates from this randomized phase 2 study, with uncertainty and reporting limitations, rather than as a definitive answer for all patients, settings, or treatment durations.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares three cendakimab regimens with placebo using a common continuous primary endpoint. The randomized parallel-group design and masking described in the registry provide a framework for comparing groups, while the endpoint captures average change in an investigator-assessed measure of disease severity. The arm-level means, standard errors, and sponsor-reported comparisons make it possible to consider both the estimated differences and their uncertainty.

The results also illustrate why an estimate, confidence interval, and p-value should be read together. The treatment comparisons have negative point estimates, but the intervals differ in whether they include zero. The recalculated values are not identical to the sponsor’s ANCOVA results, demonstrating that the method and analysis details matter even when posted group means reproduce the same mean differences. Since a discrepancy is flagged, the sponsor’s analysis is the primary reported inference.

The study also shows the interpretive challenge of multiple outcomes and comparisons. Three primary treatment comparisons and a broad set of secondary measures create a family of statistical questions. The registry record does not report multiplicity handling, so isolated unadjusted findings warrant restraint. Similarly, the planning illustrations show how assumed effect size affects hypothetical future sample sizes, but they do not reveal the original design rationale or the power of the completed study.

These features make the record useful for statistical education: it distinguishes continuous outcomes from responder outcomes, sponsor analyses from simple recalculations, and planning assumptions from trial results. It also highlights what cannot be evaluated when model, missing-data, and multiplicity details are absent. A careful reading preserves the numerical findings while keeping their uncertainty and reporting context visible.

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 trial-specific limitations follow from the registry information. The study is phase 2 and reports a 16-week primary endpoint, so the posted result addresses the measured period and does not establish longer-term outcomes. The registry describes adults with moderate to severe atopic dermatitis; conclusions should not be extended beyond that population without other evidence. The record does not report demographic or clinical characteristics needed to assess how closely enrolled participants represent other populations.

The sponsor’s primary analyses use ANCOVA, but the covariates and full model specification are not reported in the registry record. The simple recalculations differ from those sponsor results, and the discrepancy flag directs readers to prioritize the sponsor’s analysis. Without the analysis plan, it is not possible to independently determine which model details explain the differences.

The record identifies a modified Intent to Treat population but does not define its inclusion rules. Missing week 16 outcome counts, reasons for missingness, methods for handling missing measurements, and sensitivity analyses are not reported. These omissions limit assessment of how missing data or analysis-set selection may affect the findings.

Multiplicity procedures are also not reported despite three primary comparisons and numerous secondary outcomes. Individual p-values therefore cannot be assumed to control the chance of false-positive findings across all comparisons. The registry’s sample-size illustrations are based on observed effects and stated planning assumptions, not the original design justification. Finally, the safety summaries provide counts and selected serious-event terms but not enough detail to assess causality, timing, or comparative safety. These constraints call for cautious interpretation rather than extrapolation beyond the posted results.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Evaluate Safety and Effectiveness of Cendakimab (CC-93538) in Participants With Moderate to Severe Atopic Dermatitis — NCT04800315.
• Registry record last updated 2024-01-03; status last verified 2023-12.
• Results first posted 2024-01-03.
• 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 describes a completed phase 2, randomized, quadruple-masked, parallel-group study of cendakimab in adults with moderate to severe atopic dermatitis. It reports enrollment of 221 and a primary endpoint of mean percentage change from baseline in EASI at week 16 in the modified Intent to Treat population. Three cendakimab groups were compared with weekly placebo.

The sponsor’s ANCOVA results report negative mean differences versus placebo for all three treatment groups. The sponsor-reported 95% confidence intervals exclude zero for Treatment 1 and Treatment 3 and include zero for Treatment 2. Simple recalculations from the posted summaries differ, and the sponsor’s own analysis takes precedence. The record also posts secondary outcomes and descriptive adverse-event counts, but does not report multiplicity handling, missing-data methods, the original sample-size justification, or key details of the ANCOVA model.