A Study With a Initial Treatment Period Followed by a Randomized-withdrawal Period to Evaluate the Efficacy and Safety of Bimekizumab in Adult Subjects With Moderate to Severe Chronic Plaque Psoriasis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study With a Initial Treatment Period Followed by a Randomized-withdrawal Period to Evaluate the Efficacy and Safety of Bimekizumab in Adult Subjects With Moderate to Severe Chronic Plaque Psoriasis (NCT03410992), a randomized phase 3 study. The primary outcome is Percentage of Participants With a Psoriasis Area and Severity Index 90 (PASI90) Response at Week 16.

2. Clinical Question

The clinical question is about a defined short-term skin response under the study’s initial randomized comparison. It is not a question about every outcome related to psoriasis, nor does the Week 16 endpoint alone describe what happened throughout the later withdrawal and escape periods. The registry record describes the study as a comparison of efficacy and safety, but the primary endpoint is specifically the proportion meeting the PASI90 response threshold at Week 16.

Because the primary outcome is binary, each participant is classified as either meeting or not meeting the response definition. Participants with a missing Week 16 score were counted as nonresponders under the stated non-responder imputation approach. The primary population was the Randomized Set, rather than a population defined by completing treatment or meeting a post-randomization criterion. Details about the response thresholds used to determine later treatment pathways are not reported in the registry record.

3. Trial Design

The registry record describes a phase 3, multicenter, double-blind, placebo-controlled, randomized, parallel study with quadruple masking. It lists enrollment of 435 and study status as completed. The study began on 2018-02-05, and the primary completion date was 2018-12-28. The sponsor was UCB Biopharma SRL.

The design had an initial treatment period followed by a randomized-withdrawal period. In the bimekizumab cohort, participants received bimekizumab for 16 Weeks. Participants meeting predefined response criteria were re-randomized to bimekizumab or placebo through Week 56; participants who did not meet criteria entered the bimekizumab escape arm. In the placebo cohort, participants received placebo for 16 Weeks. Those meeting the response criteria continued placebo through Week 56, while those who did not meet them entered the bimekizumab escape arm. The escape-arm description says open-label bimekizumab was given for 12 weeks to participants who did not meet predefined criteria at Week 16 or later.

These later pathways matter when interpreting the overall study: the initial randomized comparison and the subsequent response-dependent treatment periods answer different questions. The primary endpoint is at Week 16, before the later withdrawal pathways are the focus. The exact response criteria that governed re-randomization and escape, the detailed visit schedule, the randomization ratio, and the original sample-size justification are not reported in the registry record. The record also does not state which roles or groups were masked despite recording quadruple masking.

The primary analysis population is the Randomized Set, defined as all randomized study participants. The registry lists 86 participants in the placebo group and 349 in the bimekizumab group for the primary endpoint. Those arm-specific populations are the basis for the reported Week 16 comparison; they should not be confused with the changing groups used later in the withdrawal and escape periods.

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: Study has a 2-5 weeks Screening Period, a 16 weeks Initial Period, a 40 weeks Randomized-Withdrawal Period (RWP) and a SFU Period (20 weeks after final dose). Participants who did not achieve a PASI90 response at Wk16 or who relapsed at Wk20/later during the RWP, entered 12 weeks of escape treatment. Participant Flow refers to the Randomized Set.

Only the first 6 of 11 groups are shown.

Initial Treatment Period: up to Wk16

Reasons for not completing this period:

Week 16 Assessment

Reasons for not completing this period:

Randomized-Withdrawal Period: up to Wk56

Reasons for not completing this period:

Escape Treatment: 12 Weeks

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Randomized Set (RS) consisted of all randomized study participants.

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: Baseline Characteristics refer to the Randomized Set (RS) which consisted of all randomized study participants.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With a Psoriasis Area and Severity Index 90 (PASI90) Response at Week 16, measured over At Week 16.

The primary endpoint is the percentage of participants with a PASI90 response at Week 16. A responder was defined as a participant with a 90% reduction from baseline in the Psoriasis Area and Severity Index (PASI) score. The PASI combines assessments of redness, thickness, and scaling across four body areas with the proportion of skin affected in those areas. The registry description gives a possible score range from 0, no disease, to 72, maximal disease.

This endpoint is used to express a substantial change from each participant’s baseline disease score as a clear yes-or-no response. Reporting the percentage who cross a prespecified threshold can be easier to interpret than reporting only an average score change, because it directly describes how many participants met the defined response criterion. However, a threshold also simplifies a continuous score: it does not show how far from the threshold nonresponders were, or how much improvement occurred among people who did not reach PASI90.

Missing Week 16 scores were counted as nonresponses under the registry’s NRI approach. That rule keeps missing scores from being treated as successful responses, but it also means the endpoint combines observed failures to reach the threshold with participants whose Week 16 score was missing. The registry record does not report additional sensitivity analyses for missing outcomes or further details about the baseline score assessment.

General: A binary endpoint is useful for comparing response proportions between randomized groups. The estimated difference in proportions describes the absolute separation between those groups, while an odds ratio describes the relative odds of the binary outcome. These measures use the same underlying response classifications but express the comparison in different ways.

7. Results Posted in the Registry

• Bimekizumab 320 mg Q4W (RS): 90.8% (317/349) versus 1.2% (1/86) on control; risk difference 89.7 percentage points (95% CI 85.9 to 93.4), risk ratio 78.11 (11.13–548.40); unadjusted two-proportion 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.

Percentage of Participants With a PASI100 Response 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:
  • Bimekizumab 320 mg Q4W (RS): 68.2% (238/349) versus 1.2% (1/86) on control; risk difference 67.0 percentage points (95% CI 61.6 to 72.4), risk ratio 58.65 (8.35–412.15); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Bimekizumab 320 mg Q4W (RS) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With a IGA Clear Response 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:
  • Bimekizumab 320 mg Q4W (RS): 69.6% (243/349) versus 1.2% (1/86) on control; risk difference 68.5 percentage points (95% CI 63.1 to 73.8), risk ratio 59.88 (8.52–420.78); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Bimekizumab 320 mg Q4W (RS) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With a PASI75 Response at Week 4

• 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:
  • Bimekizumab 320 mg Q4W (RS): 75.9% (265/349) versus 1.2% (1/86) on control; risk difference 74.8 percentage points (95% CI 69.7 to 79.8), risk ratio 65.30 (9.30–458.72); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Bimekizumab 320 mg Q4W (RS) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes cover additional measures of skin clearance, symptoms, and disease areas. They include PASI100 at Week 16, IGA Clear at Week 16, PASI75 at Week 4, symptom diary responses for pain, itch, and scaling at Week 16, scalp IGA response among participants with scalp psoriasis at baseline, and a later PASI90 response among Week 16 PASI90 responders. The registry lists 17 secondary outcomes.

Among the secondary results with numerical comparisons reported, PASI100 and IGA Clear at Week 16 and PASI75 at Week 4 all show higher response percentages in the bimekizumab group than in the placebo group, with unadjusted p-values of <0.001. This overall pattern is directionally consistent with the primary Week 16 PASI90 result, while addressing different response thresholds, assessment measures, or time points. The outcome-specific values should be interpreted separately rather than treated as interchangeable measures.

Detailed numerical results for the other listed secondary outcomes are not reported in the registry record. Their names alone do not establish the size, precision, or direction of any group differences. In particular, a Week 56 outcome among Week 16 responders concerns a selected subgroup and a later treatment period; it does not have the same interpretation as the initial randomized comparison.

Unadjusted secondary results carry less weight as confirmatory evidence when several outcomes are tested. General: The more statistical comparisons made, the greater the chance of observing small p-values by chance somewhere among them. Prespecification, a testing hierarchy, or multiplicity adjustment can help control that risk. The registry record does not report whether such procedures were used. The secondary findings can therefore add context to the primary result, but their confirmatory status and family-wise error control cannot be determined from the registry record.

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

• Percentage of Participants With a PASI100 Response at Week 16
• Percentage of Participants With a IGA Clear Response at Week 16
• Percentage of Participants With a PASI75 Response at Week 4
• Percentage of Participants With a Patient Symptom Diary Response for Pain at Week 16
• Percentage of Participants With a Patient Symptom Diary Response for Itch at Week 16
• Percentage of Participants With a Patient Symptom Diary Response for Scaling at Week 16
• Percentage of Participants With Scalp IGA Response (Clear or Almost Clear) at Week 16 for Participants With Scalp Psoriasis (PSO) at Baseline
• Percentage of Participants With a PASI90 Response at Week 56 Among Week 16 PASI90 Responders

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: each participant either met the PASI90 definition at Week 16 or did not. The posted arm-level percentages are 1.2% for placebo and 90.8% for bimekizumab 320 mg Q4W. The primary population is the Randomized Set, defined as all randomized study participants. The registry gives arm sizes of 86 and 349, respectively, for this outcome.

What does the risk difference express?

General: A risk difference subtracts one group’s response proportion from the other group’s response proportion. It is an absolute measure on the probability scale and is often useful for describing how far apart two response rates are. The simple unadjusted re-calculation from the posted summary statistics gives a risk difference of 0.897, with a 95% confidence interval from 0.859 to 0.934. The confidence interval conveys uncertainty in the estimated difference; it is not a range containing a known fixed treatment effect with certainty.

The re-calculation uses the posted arm-level response information without adjustment for other variables or strata. It therefore need not match an analysis that accounts for the trial’s design in a different way. For the primary result, the registry also reports a Cochran-Mantel-Haenszel analysis of the odds ratio. The registry record does not describe the stratification variables or provide further detail about the model implementation.

What does the odds ratio express?

General: An odds ratio compares the odds of response in one group with the odds in another. Odds are not the same as probabilities, and an odds ratio should not be interpreted as a percentage-point difference. When response is very uncommon in one arm and common in the other, the odds ratio can be numerically large. For the primary endpoint, the sponsor’s reported odds ratio is 496.318, with a 95% confidence interval from 82.798 to 2975.086 and p <0.001.

The simple re-calculation also gives relative measures, including a very large unadjusted odds ratio. Those recalculated measures are not substitutes for the sponsor’s reported Cochran-Mantel-Haenszel result. The registry record does not flag a discrepancy between the sponsor analysis and the re-calculation. The two approaches should still be read as distinct because they report different methods and may use different assumptions.

How should the p-value be read?

General: A p-value describes 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, and it does not measure the size or clinical importance of a result. Both the unadjusted comparison and the sponsor-reported primary analysis give p <0.001. This indicates strong statistical evidence against equal response rates under the respective analyses, but interpretation also depends on design, analysis choices, precision, and the full set of outcomes examined.

The primary analysis method is reported as Cochran-Mantel-Haenszel, with the parameter identified as an odds ratio and the analysis type identified as superiority. General: This method can combine information across strata when estimating an association between treatment group and outcome. The registry record does not identify the strata or describe any additional covariate adjustment, so those details should not be inferred.

How does missing-data handling affect the endpoint?

The registry states that participants with a missing Week 16 PASI score were counted as nonresponders, using NRI. General: This rule makes a conservative classification relative to treating missing participants as responders, but it is not the only possible assumption about missing outcomes. The consequences depend on how many scores were missing and why they were missing. The number of missing Week 16 scores and sensitivity analyses under alternative assumptions are not reported in the registry record.

The primary population includes all randomized participants according to the registry definition. This supports an interpretation tied to randomized assignment rather than only participants who completed the treatment period. The record does not provide further details on protocol deviations, exclusions, or whether other analysis populations were used for the primary result.

What can the secondary analyses establish?

Several other binary outcomes are listed, including PASI100, IGA Clear, and PASI75, as well as symptom diary and other outcomes. The detailed results for PASI100, IGA Clear, and PASI75 all report p <0.001 in their unadjusted comparisons. General: Testing many outcomes increases the chance that some results appear statistically significant by chance, even when each individual test uses a conventional threshold. A multiplicity adjustment or a prespecified testing hierarchy can address this issue, but such procedures are not reported in the registry record.

The record lists 19 outcome measures and 12 analyses. It does not report a full statistical analysis plan, the sample-size justification, interim analyses, or the multiplicity approach. These omissions limit assessment of whether the reported tests were confirmatory, how the analysis choices were set in advance, and how the endpoint results should be ranked beyond the stated primary designation.

11. Confidence Intervals and Precision

The primary risk-difference re-calculation is 0.897, with a 95% confidence interval from 0.859 to 0.934. The interval gives a range of values that are compatible with the observed data under the assumptions of that calculation. It is an estimate of uncertainty, not a guarantee that the true difference lies within those limits.

The interval is on the absolute probability-difference scale. It should not be confused with the sponsor’s reported odds-ratio interval, which is from 82.798 to 2975.086 around an odds ratio of 496.318. The scales differ, so the endpoints cannot be compared directly. The sponsor’s odds-ratio interval is wide, reflecting limited precision for that relative measure despite the strong separation between the observed response proportions. The registry also reports p <0.001 for the sponsor analysis.

General: Confidence intervals help readers distinguish the estimated effect from its uncertainty. A narrow interval indicates greater precision on the chosen scale; a wide interval indicates less precision. Precision depends on the outcome frequency, sample size, and analysis method, among other factors. A confidence interval does not show the probability that a treatment effect is clinically important, nor does it account automatically for multiplicity or bias.

The arm sizes and the very different response proportions shape the precision of the estimates. The posted summaries do not include a separate interval for each arm’s response percentage, and the registry does not report further interval methods for the sponsor’s analysis. The confidence interval for the recalculated risk difference comes from a simple unadjusted calculation, whereas the sponsor’s analysis used a Cochran-Mantel-Haenszel method.

Precision is also endpoint- and time-specific. These intervals concern PASI90 at Week 16 in the Randomized Set. They do not quantify uncertainty for the withdrawal period, escape treatment, safety outcomes, or the other listed endpoints.

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 illustrations based on the observed effect, not the original design calculations. The 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 set to the stated fraction of the observed effect. For this binary endpoint, the figures are participants per arm, not total event counts.

The 80% power column gives the number of participants per arm estimated to detect the selected effect with 80% power under those assumptions. In plain language, power is the chance that a specified statistical procedure would identify an effect of that size if the assumed effect were the true one. The 90% power column gives the corresponding planning number for a higher chance of detection under the same assumptions. These are planning quantities, not descriptions of the participants enrolled in the study.

The planning illustrations consider 100%, 75%, and 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so it requires more participants per arm to reach the same target power. The figures are based on a simplified two-arm comparison and do not establish the sample-size assumptions used by the trial investigators. The original sample-size justification and design assumptions are not reported in the registry record.

This is not observed power. It is a forward-looking planning illustration calculated using an effect estimated from the posted results. It should not be used to claim that the completed study had a particular probability of detecting its result, or to assess whether the original trial was adequately planned. The assumptions may not reflect the full trial design, especially the later response-dependent treatment pathways.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.012; treatment proportion p₁ = 0.908; average p̄ = 0.460.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7048 = 1.3815.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3079 = 0.2591.
4. n = (1.3815 + 0.2591)² / (0.8967)² = 3.3, rounded up to 4 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
p0 <- 0.0116279069767442                          # control proportion
p1 <- 0.908309455587393                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3410992)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 4 (80% power), 4 (90% power)
 75% of observed effect: n per arm = 7 (80% power), 9 (90% power)
 50% of observed effect: n per arm = 13 (80% power), 17 (90% power)
power.prop.test: 3.3 
Simulated power at n = 4 per arm: 0.942 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 primary analysis reports a Cochran-Mantel-Haenszel odds ratio of 496.318, with a 95% confidence interval from 82.798 to 2975.086 and p <0.001. The simple unadjusted re-calculation reports a risk difference of 0.897, with a 95% confidence interval from 0.859 to 0.934, and an unadjusted p-value of <0.001. These results describe the same Week 16 comparison using different effect measures and analysis approaches.

The risk difference is an absolute comparison of response proportions. The odds ratio compares odds, not probabilities, and its numerical size should not be translated into a percentage-point difference. Because the sponsor’s analysis is a Cochran-Mantel-Haenszel analysis, it may reflect stratification, but the registry record does not identify the strata. The re-calculation is explicitly unadjusted and should be treated as a check based on the posted summaries rather than as a replacement for the sponsor’s analysis.

The registry does not flag a discrepancy between the analyses. The estimates are not expected to be numerically identical because they use different scales and methods. Both show a marked contrast in Week 16 responses, with p-values below 0.001. The registry does not provide enough methodological detail to reconstruct the sponsor’s calculation or determine the exact contribution of any stratification.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies PASI90 at Week 16 as the primary endpoint and lists 19 outcome measures and 12 analyses overall. It also provides multiple secondary outcomes at different times, including response thresholds, symptom diary measures, and outcomes for participants with scalp psoriasis. This breadth makes the approach to multiple testing relevant when interpreting the collection of p-values.

General: When many hypotheses are tested, the chance of at least one false-positive finding across the set can rise if each test is interpreted independently. A prespecified testing sequence or a multiplicity adjustment can control error across related tests. The registry record does not report whether a multiplicity procedure was used, how the outcomes were ordered for testing, or which analyses were intended as confirmatory.

The endpoint description states that participants with missing Week 16 PASI scores were classified as nonresponders using NRI. The registry record does not report the number of missing scores, reasons for missingness, or sensitivity analyses using alternative assumptions. NRI specifies how missing scores were classified for this endpoint, but it does not demonstrate that missingness had no effect on the estimate.

The primary population is the Randomized Set, defined as all randomized participants. Further information about exclusions, protocol deviations, treatment adherence, or other analysis populations is not reported in the registry record. The record also does not report the original sample-size justification, interim analyses, or a complete statistical analysis plan. These omissions limit assessment of whether all analysis choices were prespecified and how the primary and secondary inferences were protected against selective interpretation.

The trial’s later randomized-withdrawal and escape periods use response-dependent groups. Those stages differ from the initial randomized comparison and may involve selected participants. The registry record describes the pathways but does not provide enough detail in the available results to evaluate the full analysis strategy for those later periods.

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: diarrhoea (2.0%), ischaemic cardiomyopathy (1.2%), coronary artery disease (1.2%), duodenal ulcer haemorrhage (1.2%), type 2 diabetes mellitus (1.2%).

The registry reports serious adverse event counts for several analysis groups, including the initial placebo and bimekizumab groups and groups in the later withdrawal or escape periods. In the initial groups, serious adverse events were reported for 2 of 86 participants receiving placebo and 6 of 349 receiving bimekizumab 320 mg Q4W. The registry reports zero deaths in each listed group.

Other adverse events are separately counted in the record. For the initial groups, the counts are 15 of 86 for placebo and 66 of 349 for bimekizumab 320 mg Q4W. These counts should not be treated as directly comparable with serious adverse event counts: they refer to a different category, and the registry does not provide detailed event definitions in the registry record. Later groups have different treatment pathways and group sizes, so their counts should be read in the context of their specific periods rather than combined with the initial groups.

The listed serious-event terms include diarrhoea at 2.0%, ischaemic cardiomyopathy at 1.2%, coronary artery disease at 1.2%, duodenal ulcer haemorrhage at 1.2%, type 2 diabetes mellitus at 1.2%, and syncope at 1.2%. The registry record does not provide enough detail to establish the timing, attribution, or comparative risk for each listed term. These descriptive counts do not establish whether an event was caused by treatment, and they do not by themselves support a clinical safety conclusion.

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

The primary result describes a large difference in the observed Week 16 PASI90 response proportions between the randomized groups. The posted results report 90.8% for bimekizumab 320 mg Q4W and 1.2% for placebo. The simple unadjusted risk-difference estimate is 0.897, with a 95% confidence interval from 0.859 to 0.934; the sponsor-reported Cochran-Mantel-Haenszel odds ratio is 496.318, with a 95% confidence interval from 82.798 to 2975.086 and p <0.001.

These statistics address the defined endpoint in the Randomized Set at Week 16. They do not describe every possible measure of psoriasis, the degree of improvement among participants who did not reach PASI90, or the durability of response after the initial period. They also do not show outcomes for all secondary endpoints, because detailed results for several listed outcomes are not reported in the registry record.

General: A statistically clear difference is not the same as a complete account of clinical benefit and harm. The endpoint threshold, the uncertainty around the estimates, missing-data handling, the multiplicity strategy, and the study population all matter to interpretation. The safety counts are descriptive and do not establish causation or comparative safety on their own.

The study’s later withdrawal and escape periods are important to its overall design, but the primary Week 16 comparison should not be extended automatically to those periods. After Week 16, treatment assignments and eligibility for pathways depended on response criteria. The record describes these pathways but does not report enough detail to make the primary result a summary of the full 56-week experience.

The findings apply to the trial population and conditions studied as recorded. The registry record does not report enough information to determine how representative participants were of all adults with chronic plaque psoriasis or how the estimates would translate to settings with different populations or care. The result is evidence about a specified randomized comparison, not a guarantee of an individual participant’s response.

18. Why This Trial Matters Statistically

This trial illustrates how a binary endpoint can communicate a treatment-group contrast in a form that is directly tied to a defined response threshold. PASI90 classifies participants according to whether their PASI score fell by 90% from baseline by Week 16. Reporting response proportions makes the threshold-based outcome transparent, while the risk difference supplies an absolute comparison between groups.

The result also demonstrates why effect measures must be interpreted on their own scales. The sponsor reports an odds ratio, while the simple unadjusted re-calculation reports a risk difference. Both describe the same primary comparison, but one is an odds-based relative measure and the other an absolute difference in response proportions. Their numerical values cannot be compared directly or substituted for one another.

Precision is another key lesson. The primary risk-difference interval is relatively compact on its scale, while the sponsor’s odds-ratio interval spans a broad range. General: Estimates can appear differently precise depending on the effect measure used, especially when outcomes are sparse in one group. A p-value alone does not convey the range of plausible effect sizes.

The study design also makes time and population essential to interpretation. The initial randomized comparison precedes response-dependent re-randomization and escape treatment. Week 16 results do not automatically answer questions about outcomes during later periods, when participants follow different pathways based on response.

Finally, the record identifies a primary endpoint but does not report multiplicity procedures, the original sample-size rationale, or a full statistical analysis plan. That distinction matters: a strongly separated primary result can coexist with uncertainty about how a broader set of secondary tests was controlled or prioritized. Statistical interpretation requires attention to both the numerical findings and the limits of the reported analysis details.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the original sample-size justification, interim analyses, or a complete statistical analysis plan. It also does not state whether multiplicity adjustment or a prespecified testing hierarchy was used for the many listed outcomes. These gaps limit evaluation of the planned inferential framework and the confirmatory status of secondary results.

The primary endpoint uses non-responder imputation for missing Week 16 PASI scores, but the number of missing scores, their reasons, and sensitivity analyses are not reported in the registry record. The reported response proportions therefore depend on a stated missing-data rule whose impact cannot be assessed from the available details.

The sponsor’s reported primary odds ratio uses the Cochran-Mantel-Haenszel method, but the registry record does not identify the strata or give enough detail to reproduce the analysis. The unadjusted re-calculation is a different analysis and should not be taken as a reconstruction of the sponsor’s method. The odds-ratio confidence interval is broad, and relative measures can be difficult to interpret when response is rare in one group.

The design includes response-dependent treatment changes after Week 16, but the criteria for re-randomization and escape are not described in detail. Later groups therefore should not be treated as equivalent to the initial randomized groups. Detailed results for many secondary outcomes are not reported in the registry record, limiting assessment of the full pattern across outcomes and time points.

Finally, the registry information does not establish how representative the study participants were of the broader population or provide detailed event attribution for safety outcomes. These are limits on interpretation of the record, not evidence that the reported analyses are invalid.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study With a Initial Treatment Period Followed by a Randomized-withdrawal Period to Evaluate the Efficacy and Safety of Bimekizumab in Adult Subjects With Moderate to Severe Chronic Plaque Psoriasis — NCT03410992.
• Registry record last updated 2026-04-15; status last verified 2026-04.
• Results first posted 2022-03-04.
• 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

This completed phase 3 randomized, double-blind, placebo-controlled study evaluated bimekizumab in adults with moderate to severe chronic plaque psoriasis. The design included an initial 16-week treatment period followed by response-dependent withdrawal and escape pathways. The primary endpoint was PASI90 response at Week 16 in the Randomized Set, with missing Week 16 scores counted as nonresponses.

The registry reports PASI90 responses of 90.8% with bimekizumab 320 mg Q4W and 1.2% with placebo. The sponsor’s Cochran-Mantel-Haenszel analysis reports an odds ratio of 496.318, a 95% confidence interval from 82.798 to 2975.086, and p <0.001. The simple unadjusted risk-difference re-calculation is 0.897, with a 95% confidence interval from 0.859 to 0.934.

Interpretation is specific to this endpoint and time point. The registry record does not report the original sample-size justification, multiplicity procedures, or a complete statistical analysis plan.