A Study to Evaluate the Efficacy and Safety of Apremilast (CC-10004) in the Treatment of Behçet Disease: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Efficacy and Safety of Apremilast (CC-10004) in the Treatment of Behçet Disease (NCT00866359), a randomized phase 2 study. The primary outcome is Number of Oral Ulcers at Day 85.

2. Clinical Question

The clinical question is framed around a count of oral ulcers at a specified assessment time, rather than around a broad measure of disease status. The record describes counting oral ulcers at each visit and at the end of the treatment period, with baseline as the starting point. The primary comparison therefore concerns the estimated group means at Day 85. It does not, by itself, describe every aspect of disease activity, the course of individual participants’ symptoms, or the balance of benefits and harms.

Random allocation and masking are relevant to how the comparison is interpreted: the registry identifies the study as randomized and quadruple-masked. These design features are intended to support a fair comparison, although the details of the masking roles and implementation are not reported in the registry record. The endpoint result should be read as a comparison within this trial and its analysis population, not as a finding about all people with Behçet disease or about outcomes beyond those measured.

3. Trial Design

The registry describes a Phase 2, multi-center, randomized, double-blind, placebo-controlled, parallel-group study followed by an active-treatment extension. It records quadruple masking and a completed status. The study began on 2009-08-01, and primary completion was recorded as 2012-05-01. The sponsor is Amgen.

For the primary comparison, the reported arms were oral placebo twice daily and apremilast 30 mg orally twice daily. The arm sizes were 56 and 55 participants, respectively. The registry records total enrollment as 111. The primary endpoint was assessed at Day 85, and the result is reported as a least-squares mean for each arm with a standard error, alongside an ANCOVA comparison.

The study’s title indicates that an active-treatment extension followed the controlled phase. Safety information is also reported for Week 12 groups and for an apremilast group at Week 24. The extension’s detailed allocation, participant flow, and analysis structure are not reported in the registry record, so the Week 24 group should not be interpreted as a continuation of a particular randomized comparison without further detail.

The registry lists 24 outcome measures and 6 analyses, while 20 secondary outcomes are posted. The outcomes include ulcer counts and pain scores, as well as other measures of ulcer burden and disease activity. The full protocol, the rationale for the planned enrollment, and the original statistical analysis plan are not reported in the registry record. Consequently, the posted design information supports a description of the main comparison but does not establish all operational or analytic details.

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: Eligible participants were randomized 1:1 to receive study drug (apremilast or placebo). Since the incidence and severity of Behçet's Disease differ between males and females, randomization was stratified by gender.

Treatment Phase (Days 1 to 85)

Reasons for not completing this period:

Extension Phase (Day 86 to 169)

Reasons for not completing this period:

Observational Follow-Up Phase

Analysis population for the primary outcome, as defined in the registry: Intent to Treat (ITT) = all randomized participants with at least one oral ulcer evaluation (including the baseline visit). A Last Observation Carried Forward (LOCF) approach was applied for participants terminated early. If a participant had no post-baseline oral ulcer assessment, the baseline value was carried forward for calculation.

5. Baseline Characteristics

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

Population: The intent to treat (ITT) population was defined as all randomized participants with at least one oral ulcer evaluation (including the baseline evaluation)

6. Primary Endpoint

The registry lists the primary outcome as Number of Oral Ulcers at Day 85, measured over Day 85.

The primary endpoint is the number of oral ulcers at Day 85. The registry says oral ulcers were counted at each visit and at the end of the treatment period, with baseline identified as the starting point. A Day 85 count provides a defined-time assessment of oral-ulcer burden for comparison between the randomized groups. A count is directly interpretable as the number of ulcers recorded, although a single time point does not capture every change during follow-up.

The posted endpoint is expressed as a least-squares mean, with standard error as its reported measure of dispersion. A least-squares mean is an adjusted group estimate from a statistical model; the record names ANCOVA for the sponsor’s comparison. The covariates included in that model are not reported in the registry record. Accordingly, the estimates should be understood as the sponsor’s model-based group means, not necessarily as simple arithmetic averages of the raw participant counts.

The stated primary population includes randomized participants with at least one oral-ulcer evaluation, including baseline. The description also says that last observation carried forward was used for participants who terminated early. The remaining endpoint description is incomplete in the registry record, so no further imputation detail can be established. The endpoint answers a focused question about oral-ulcer counts at Day 85; it is not a complete summary of safety, disease activity, or participants’ overall experience.

7. Results Posted in the Registry

• Apremilast 30mg (Oral) BID: mean 0.40 versus 2.00 on control; difference -1.60 (95% CI -2.38 to -0.82), standardized effect d = -0.77; z-test p = <0.001.

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.

Pain of Oral Ulcers as Measured by Visual Analog Scale (VAS) at Day 85

• 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:
  • Apremilast 30mg (Oral) BID: mean 9.90 versus 36.70 on control; difference -26.80 (95% CI -35.85 to -17.75), standardized effect d = -1.10; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Apremilast 30mg (Oral) BID than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Area Under the Curve (AUC) for the Number of Oral Ulcers From Day 1 to 85

• 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:
  • Apremilast 30mg (Oral) BID: mean 67.74 versus 157.82 on control; difference -90.08 (95% CI -126.33 to -53.83), standardized effect d = -0.92; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Apremilast 30mg (Oral) BID than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Area Under the Curve (AUC) for the Number of Oral Plus Genital Ulcers From Day 1 to 85

• 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 deviations.
• Result:
  • Apremilast 30mg (Oral) BID: mean 65.79 versus 193.95 on control; difference -128.16 (95% CI -179.19 to -77.13), standardized effect d = -0.93; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Apremilast 30mg (Oral) BID than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes cover several related aspects of Behçet disease, including pain associated with oral or genital ulcers, ulcer counts over time, combined oral and genital ulcer burden, response categories, and a disease-activity score. This range complements the primary endpoint by looking beyond the number of oral ulcers at Day 85. It can help readers understand whether the primary finding sits within a broader set of measured outcomes, but the outcomes are not interchangeable and do not all represent the same clinical construct.

The posted continuous secondary results generally show lower estimated values in the apremilast arm than in the placebo arm for the selected pain and ulcer-burden measures. The corresponding sponsor analyses reported for selected outcomes also show statistically significant contrasts. Those results are directionally related to the primary oral-ulcer count, but they are secondary findings and should not be counted as independent confirmations without considering the overlap among outcomes and the analysis plan.

Several outcomes were assessed, and 20 secondary outcomes are posted. Testing many outcomes creates a multiplicity concern: even when no real differences exist, the chance of at least one small p-value can increase when many hypotheses are examined. General explanation: adjustment procedures can control this family-wise error or other error measures, but the registry record does not report whether a multiplicity procedure was prespecified or applied. Therefore, unadjusted secondary p-values carry less weight than the primary result and should be read as supporting or exploratory evidence rather than as a collection of definitive tests.

The record does not provide complete details for every secondary endpoint’s analysis population, missing-data handling, or model specification. Some listed outcomes are not represented among the posted results, and a detailed interpretation of unposted outcomes is not possible from the registry record. The secondary pattern can add context, but it does not remove uncertainty about multiplicity, analysis choices, or clinical importance.

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

• Pain of Oral Ulcers as Measured by Visual Analog Scale (VAS) at Day 85
• Pain of Genital Ulcers as Measured by Visual Analog Scale (VAS) Scores at Day 85
• Area Under the Curve (AUC) for the Number of Oral Ulcers From Day 1 to 85
• Area Under the Curve for the Number of Genital Ulcers From Day 1 to 85
• Area Under the Curve (AUC) for the Number of Oral Plus Genital Ulcers From Day 1 to 85
• Sum of the Number Oral Ulcers, Genital Ulcers or Oral Plus Genital Ulcers at Day 85
• Percentage of Participants Who Were Oral Ulcer-free (Complete Response), or Whose Oral Ulcers Were Reduced by ≥ 50%, (Partial Response)
• Change From Baseline in the Disease Activity as Measured by BD Current Activity Form/Index Score on Day 85

9. Statistical Interpretation

10. Statistical Methods Explained

The sponsor’s primary analysis is reported as ANCOVA, with the parameter described as a least-squares mean difference. ANCOVA is a model-based method for comparing group means while accounting for specified covariates. General explanation: adjustment can improve precision when relevant baseline variables are included and can account for initial differences in those variables. The registry record does not identify which covariates were used, how they were entered, or whether any additional model terms were included. Those details should not be inferred from the label ANCOVA alone.

What does a least-squares mean represent?

A least-squares mean is a model-adjusted estimate of a group mean. In a randomized trial, the raw group means and model-adjusted means may be close, but they are not necessarily identical. The posted primary arm estimates are 2.0 for placebo and 0.4 for apremilast, each with a standard error of 0.28. The sponsor reports the adjusted contrast as -1.6. The model’s adjustment details are not reported in the registry record, so the adjusted result cannot be independently reconstructed from the posted arm summaries alone.

The standard error describes the estimated uncertainty of an estimate under its analysis. It is not the same as the spread of individual participant outcomes. The primary outcome’s reported dispersion is a standard error, not a standard deviation. General explanation: a standard error is typically smaller when an estimate is based on more information and reflects uncertainty in the estimated mean, whereas a standard deviation describes variability among observations. Confusing these quantities can lead to a mistaken impression of how variable participants’ ulcer counts were.

How should the two p-values be understood?

The sponsor’s ANCOVA reports p < 0.001. The unadjusted recalculation from posted summary statistics reports p < 0.001. The latter is a separate, simple comparison based on the summary data rather than a reproduction of the sponsor’s model. A p-value measures how incompatible the observed data are with a specified null hypothesis under the assumptions of the test; it is not the probability that the null hypothesis is true, and it does not measure clinical importance. The sponsor analysis takes precedence as the reported trial analysis.

The registry-based recalculation is useful as a check on the direction and rough scale of the contrast. Its estimate is -1.6, matching the sponsor’s reported point estimate. Its inferential result should not replace the ANCOVA result, because the analysis specifications may differ. The record marks no discrepancy between the sponsor and recalculated findings. More detailed model outputs, including covariate coefficients and diagnostic checks, are not reported in the registry record.

What does a confidence interval add?

A confidence interval expresses uncertainty around an estimated contrast as a range. General explanation: under repeated use of the same statistical procedure and its assumptions, intervals constructed in this way would cover the target value at the stated rate. A 95% confidence interval is not a statement that there is a 95% probability that this particular interval contains the true value. The sponsor’s primary interval is -2.4 to -0.9 ulcers. Because it excludes zero, it is consistent with the sponsor’s reported test against no difference at the stated level. The interval does not address unmeasured bias, the correctness of the model, or the practical importance of the effect.

The posted secondary outcomes include both least-squares means with standard errors and an outcome reported as means with standard deviations. These measures of dispersion have different interpretations. For outcomes with least-squares means, the registry-based calculation estimates a pooled standard deviation using the posted standard errors; where this conversion is approximate, the recalculated standardized effect and planning illustration inherit that approximation. The posted AUC for oral plus genital ulcers reports standard deviations directly. General explanation: a standard deviation can describe substantial participant-to-participant spread even when a group mean is estimated with reasonable precision.

Why can over-dispersion matter for count outcomes?

General explanation: count data can be more variable than a simple model assumes, a property often called over-dispersion. When variation exceeds the model’s assumptions, standard errors and confidence intervals from an unsuitable model may be too narrow or too wide. The registry record identifies ANCOVA for the sponsor’s primary comparison but does not report diagnostic checks, a count-specific model, or an assessment of over-dispersion. It is therefore not possible to say from the posted record whether this issue was assessed or how it affected the analysis.

The primary population is described as all randomized participants with at least one oral-ulcer evaluation, including baseline. The description states that last observation carried forward was applied to participants who terminated early. General explanation: this method carries a prior observed value forward to a later assessment and relies on assumptions about what an unobserved value would have been. Its appropriateness depends on the outcome and reasons for missingness. The registry does not report sensitivity analyses that test alternative assumptions, nor does it fully describe how participants without any post-baseline assessment were handled. The protocol’s original analysis plan and the sample-size justification are not reported in the registry record.

11. Confidence Intervals and Precision

The sponsor’s primary estimate is a least-squares mean difference of -1.6 ulcers, with a 95% confidence interval from -2.4 to -0.9. The interval gives a range of plausible values for the model-based contrast under its assumptions. Its endpoints are on the same side of zero, supporting the reported evidence of a difference between groups in this analysis. The interval is still a range, not a guarantee that the true contrast lies within it.

The interval’s width reflects sampling uncertainty as represented by the sponsor’s model. It does not account for every possible source of uncertainty. For example, it cannot by itself resolve whether the model was correctly specified, whether missing-data assumptions were adequate, or whether the result would generalize beyond the trial population and period. Those questions require additional design and analysis information that is not reported in the registry record.

The unadjusted recalculation uses summary statistics and produces an estimate of -1.6, while the sponsor’s adjusted interval is the primary inferential result. Since the sponsor’s analysis and the recalculation are different methods, their intervals should not be substituted for one another. The sponsor reports the interval as -2.4 to -0.9; the recalculated interval uses more detailed decimal values and is not needed to interpret the sponsor’s reported result.

Precision should also be distinguished from clinical importance. The registry record does not state a threshold for a clinically important difference in oral-ulcer count. The confidence interval can describe statistical uncertainty, but it cannot determine whether the estimated difference is important to participants or clinicians without an external interpretive standard.

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 planning figures are participants per arm because the illustrated outcome is a continuous measure, not a time-to-event outcome. The calculations assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They use the observed effect as a starting point and show planning for 100%, 75%, and 50% of that observed effect. These are illustrative assumptions for a planning exercise; the trial’s original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the estimated number of participants per arm for a design with an 80% chance of detecting the stated effect if that effect were the true one and the planning assumptions held. The 90% power column makes the same kind of illustration for a 90% chance. These are design calculations, not guarantees about the outcome of a future study. For the primary endpoint, planning for 100% of the observed effect gives 27 participants per arm for 80% power and 36 per arm for 90% power. Planning for 75% gives 48 and 64 per arm; planning for 50% gives 107 and 143 per arm.

Smaller target effects generally require more participants to detect, because the groups’ expected results are closer together relative to the uncertainty in the estimates. The figures are not observed power: they do not measure the probability that the completed trial’s result is correct or establish what power the trial actually had. They are post hoc planning illustrations based on an effect observed in this trial, and can be optimistic if that observed effect is larger than the effect a future study should be designed to detect.

These calculations do not reproduce the trial’s original design process. The registry does not report the original target effect, assumed variability, dropout allowance, allocation details for the original calculation, or any other sample-size rationale. The planning figures should therefore be read as conditional illustrations under the stated assumptions, not as evidence that the trial was designed around those participant counts.

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 = -1.600; pooled standard deviation σ = 2.086.
2. n = 2 × (1.96 + 0.842)² × 2.086² / -1.600² = 26.7, rounded up to 27 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 <- -1.6                      # observed difference in means
sigma <- 2.08603922135408                     # 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(866359)
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 = 27 (80% power), 36 (90% power)
 75% of observed effect: n per arm = 48 (80% power), 64 (90% power)
 50% of observed effect: n per arm = 107 (80% power), 143 (90% power)
power.t.test: 27.7 
Simulated power at n = 27 per arm: 0.789 (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 analysis is ANCOVA. It estimates a least-squares mean difference of -1.6 ulcers for apremilast relative to placebo, with a 95% confidence interval of -2.4 to -0.9 and p < 0.001. That is the analysis to use when describing the trial’s reported primary result.

A separate recalculation from the posted summary statistics gives the same point estimate, -1.6, and an unadjusted p-value of <0.001. It is a simple, unadjusted check rather than an attempt to reproduce ANCOVA. The recalculation does not have access to the sponsor model’s covariates or other analytic details, which are not reported in the registry record.

The registry record does not flag a discrepancy between the sponsor result and the recalculation. Their agreement in the point estimate supports consistency in the direction and scale of the reported contrast, but the inferential results remain method-specific. In particular, the sponsor’s confidence interval and p-value should not be replaced by a recalculated interval or p-value. The adjusted analysis is the primary reported analysis; the summary-statistic calculation is supplementary context.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes, including the primary oral-ulcer count and a broad set of secondary measures. When many outcomes are tested, the chance of finding at least one apparently unusual result by chance can rise. General explanation: multiplicity procedures are methods for controlling the interpretation of several statistical tests, for example by adjusting significance thresholds or reported p-values. The registry record does not report a prespecified multiplicity strategy, whether secondary outcomes were arranged into families, or whether any secondary p-values were adjusted. The reported secondary tests should therefore be treated cautiously.

The primary analysis population is described as all randomized participants with at least one oral-ulcer evaluation, including baseline. The record also states that last observation carried forward was applied for participants who terminated early. This gives some information about missing primary endpoint data, but it is not a complete account of missingness. The number of participants with missing Day 85 observations, reasons for missingness, and sensitivity analyses under alternative assumptions are not reported in the registry record. The handling of participants without a post-baseline oral-ulcer assessment is not fully specified in the available description.

General explanation: missing-data methods rely on assumptions about how observed and unobserved outcomes are related. Last observation carried forward assumes that carrying an earlier value forward is an acceptable representation of a later unobserved value. If outcomes change after the last observation, this method can misrepresent the final value. The record does not report whether the sponsor examined the robustness of the primary estimate to other missing-data approaches, or whether results differed in a population restricted to participants with observed endpoint data.

The registry names ANCOVA but does not describe its covariates, model diagnostics, or other specification details. It does not report the original statistical analysis plan, the trial’s sample-size justification, or interim analyses. The reasons for the analysis population definition and the details of protocol deviations are also not reported in the registry record. These omissions do not negate the posted result, but they limit independent assessment of the analysis choices and the sensitivity of the result to them.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: behçet's syndrome (3.6%), diplegia (1.8%), anal fissure (1.8%), haemorrhoids (1.8%), pyrexia (1.8%).

The safety figures are reported for Week 12 placebo and apremilast groups and a Week 24 apremilast group. Serious adverse events were recorded for 3 of 56 participants in the Week 12 placebo group, 2 of 55 in the Week 12 apremilast group, and 6 of 100 in the Week 24 apremilast group. Other adverse events were reported for 48 of 56, 47 of 55, and 82 of 100 participants, respectively. These are counts of participants affected within the reported groups, not estimates of comparative treatment effects.

The Week 24 group is not directly comparable with the Week 12 placebo group on the information in the registry record. The record does not give enough detail about the extension’s allocation and exposure periods to interpret these groups as a randomized comparison. The listed serious-event terms include Behçet's Syndrome at 3.6%, and Diplegia, Anal Fissure, Haemorrhoids, and Pyrexia at 1.8% each, with Influenza at 1.0%. These percentages should be read as reported terms, not as a complete description of event severity or causal attribution.

Death counts are not reported in the registry record. Definitions, exposure-adjusted rates, and detailed event narratives are also not reported. The figures describe reported adverse events in the listed groups and do not establish whether an event was caused by treatment.

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

The primary result is a statistical comparison of Day 85 oral-ulcer counts among participants in a randomized, placebo-controlled study. The sponsor’s ANCOVA estimated a lower least-squares mean count in the apremilast arm than in the placebo arm, with a 95% confidence interval that excludes zero and a reported p-value of < 0.001. The posted estimates and uncertainty describe the trial’s primary endpoint under the stated analysis, not every possible outcome for each participant.

This finding does not by itself establish that every participant experienced fewer ulcers, nor does it describe how quickly any change occurred or how it varied across visits. It does not answer all questions about genital ulcers, pain, overall disease activity, safety, or longer-term outcomes. The registry lists related secondary measures, but those have distinct definitions and are subject to multiplicity and reporting limitations.

Statistical evidence and clinical importance are different. The result indicates evidence of a difference under the sponsor’s model and its assumptions. The registry record does not state a threshold for a clinically important change in oral-ulcer count, so the clinical meaning of the estimated size cannot be determined from the statistical result alone. A confidence interval conveys uncertainty in the estimated contrast but does not resolve this interpretive question.

The trial’s scope also matters. The primary analysis population included randomized participants with at least one oral-ulcer evaluation, including baseline, and the description states that last observation carried forward was used for early termination. Missing-data details and sensitivity analyses are not reported in the registry record. The record describes a Phase 2 study and an active-treatment extension, but the detailed extension design and longer-term comparative results are not fully specified. Conclusions should therefore remain tied to the posted comparison, endpoint, population, and treatment period.

The result is not a substitute for a full assessment of harms and benefits. Safety counts are reported for several groups, but event definitions, causal attribution, and detailed exposure information are not. The statistics presented can characterize the reported findings; they cannot by themselves establish a complete clinical judgment.

18. Why This Trial Matters Statistically

This trial illustrates the distinction between a randomized treatment comparison and a model-based estimate of that comparison. The posted arm estimates are least-squares means, and the sponsor’s principal contrast comes from ANCOVA. Recognizing that these are adjusted estimates helps prevent readers from treating them as raw averages or assuming that the model can be reproduced from the arm summaries alone.

It also shows why a confidence interval is useful alongside a p-value. The sponsor reports both the estimated contrast and an interval, allowing the result to be read in terms of direction and statistical precision rather than as a binary significance statement. The interval remains conditional on the analysis assumptions and does not establish clinical importance or generalizability.

The record provides an opportunity to compare the sponsor’s analysis with a simple recalculation from posted summary statistics. The two approaches agree on the primary point estimate, while differing in method and reported p-value. This is a useful lesson in how summary-statistic checks can support transparency without replacing a trial’s own adjusted analysis.

The secondary outcomes and sample-size illustrations add further statistical lessons. Multiple outcomes require care in interpretation when adjustment details are absent, and planning calculations based on an observed effect should not be confused with the trial’s original design rationale. Likewise, safety counts from different follow-up groups should not be treated as direct comparative rates without the relevant exposure and design information. Together, these issues demonstrate why sound interpretation depends on both numerical results and the details of how those results were generated.

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.

The primary endpoint description is incomplete in the registry record. It identifies an ITT population with at least one oral-ulcer evaluation, including baseline, and says last observation carried forward was used after early termination, but the remaining description is truncated. The record does not fully specify how participants with no post-baseline assessment were handled.

The sponsor’s ANCOVA is named, but its covariates and model details are not reported. The original analysis plan, sample-size justification, and any interim analyses are also not reported in the registry record. That limits independent reconstruction of the primary analysis and assessment of whether all analysis decisions were prespecified.

Missing-data information is limited. Counts and reasons for missing Day 85 data, along with sensitivity analyses using other assumptions, are not reported. The record therefore does not establish how robust the primary estimate is to the handling of missing outcomes.

Although the trial title describes an active-treatment extension, detailed information about extension allocation, participant flow, and exposure is not reported. The Week 24 apremilast safety group should not be assumed to represent a direct randomized comparison with the Week 12 placebo group. Death counts are not reported in the registry record, and adverse-event definitions and exposure-adjusted rates are not included in the posted details.

Finally, multiple secondary outcomes are listed, but a multiplicity plan is not reported. Some outcome results and analyses are not posted, and the record does not provide a clinical-importance threshold for the primary contrast. These limits constrain interpretation of the broader evidence beyond the primary statistical comparison.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Evaluate the Efficacy and Safety of Apremilast (CC-10004) in the Treatment of Behçet Disease — NCT00866359.
• Registry record last updated 2020-06-19; status last verified 2020-06.
• Results first posted 2014-08-27.
• 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, quadruple-masked, parallel-group study of apremilast and placebo in Behçet disease, with an active-treatment extension named in the study title. Total enrollment was 111. The primary endpoint was the number of oral ulcers at Day 85.

The sponsor’s ANCOVA reported a least-squares mean difference of -1.6 ulcers for apremilast relative to placebo, with a 95% confidence interval from -2.4 to -0.9 and p < 0.001. A summary-statistic recalculation gives the same point estimate, but it is not a replacement for the sponsor’s adjusted analysis.

Secondary outcomes and safety counts provide additional context, but interpretation is limited by unreported multiplicity procedures, incomplete missing-data details, and limited information about the extension. The registry record does not report the original sample-size justification or full statistical analysis plan.