Efficacy and Safety Study of Benralizumab for Patients With Severe Nasal Polyposis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety Study of Benralizumab for Patients With Severe Nasal Polyposis (NCT03401229), a randomized phase 3 study. The primary outcome is Change From Baseline in Total NPS at Week 40.

2. Clinical Question

The clinical question is framed around a change in an endoscopic measure, rather than a direct measure of symptoms, surgery, or quality of life. Those broader dimensions appear among the listed secondary outcomes, including SNOT-22, disease severity score, and time to nasal polyp surgery and/or systemic corticosteroid use. The primary endpoint therefore addresses one defined aspect of nasal polyp burden at a specified follow-up time. It does not by itself capture every way patients may experience the condition or respond during the study.

3. Trial Design

The registry describes NCT03401229 as a completed, randomized, triple-masked, parallel-group phase 3 study sponsored by AstraZeneca. Its official title identifies a multicenter, randomized, double-blind, parallel-group, placebo-controlled design. Enrollment was 413. The study began on 2018-01-15, and primary completion was recorded as 2020-07-31. The registry status is completed.

Participants were assigned to benralizumab 30 mg subcutaneously plus mometasone furoate or matching placebo subcutaneously plus mometasone furoate. The comparison therefore concerns adding benralizumab rather than comparing mometasone furoate with no background treatment. The registry summary describes patients with severe bilateral nasal polyposis who remained symptomatic despite standard-of-care treatment, including current intranasal corticosteroids and prior surgery and/or systemic corticosteroid use.

The planned treatment period was 56 weeks. The registry summary also states that the first 200 patients completing the 56-week treatment period would have a six-month follow-up without dosing. The primary endpoint was change in total nasal polyps score from baseline to week 40, so the primary assessment occurred during the described treatment period rather than at its end.

Randomization and masking can reduce some forms of bias in treatment assignment and outcome assessment, while a placebo comparator helps isolate the contrast between the assigned study treatments. These are general design purposes, not claims about how successfully masking was maintained in this particular study. The specific randomization procedure, allocation concealment details, and masking roles are not reported in the registry record. The registry record also does not report the trial's original sample-size justification or detailed design assumptions.

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: In OSTRO, at the first visit, ie, the enrollment visit 1, patients were evaluated regarding the protocol mandated inclusion and exclusion criteria. After enrolment, eligible patients entered a 5-week screening/run in period on a stable dose of study provided Mometasone Furoate Nasal Spray (MFNS).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full analysis set: All patients randomized and receiving any IP are included in the full analysis set (FAS), irrespective of their protocol adherence and continued participation in the study. Patients are analyzed according to their randomized treatment. Patients who had not been rescued and whose post-baseline observation was missing at the timepoint of interest are excluded from this analysis.

5. Baseline Characteristics

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

Population: Full analysis set: All patients randomized and receiving any IP are included in the full analysis set (FAS), irrespective of their protocol adherence and continued participation in the study. Patients are analyzed according to their randomized treatment.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Total NPS at Week 40, measured over Baseline to week 40.

The primary endpoint was change from baseline in total nasal polyps score at week 40. It was defined as the week 40 value minus the baseline value, where baseline was the last valid value on or before randomization. Total score is the sum of the right and left nostril scores. Each nostril was assessed using a scale from 0 to 4, with a maximum total score of 8; higher values indicate heavier bilateral nasal polyp burden. Assessment used nasal endoscopy with central reading of the left and right scores.

A change score gives a within-participant measure of how the assessed polyp burden shifted between baseline and the specified follow-up. In the reported summaries, negative mean changes indicate a decline in score, while positive changes indicate an increase. The treatment comparison asks whether mean change differed between the randomized groups. The endpoint is useful statistically because it defines a common, quantifiable outcome and a specific assessment time for comparing groups.

The endpoint is not a complete measure of disease experience. Symptoms, quality of life, rescue treatment, and surgery are distinct outcomes and are represented among the listed secondary endpoints. The registry also specifies special handling of observations after nasal polyp surgery or systemic corticosteroids for nasal polyposis, so the reported change is interpreted in the context of those analysis rules rather than as a simple summary of every observed score.

7. Results Posted in the Registry

• Benra 30 mg: mean -0.36 versus 0.17 on control; difference -0.53 (95% CI -0.82 to -0.24), standardized effect d = -0.37; 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.

Change From Baseline in SNOT-22 at Week 40

• 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:
  • Benra 30 mg: mean -15.20 versus -10.70 on control; difference -4.50 (95% CI -10.72 to 1.72), standardized effect d = -0.14; z-test p = 0.156.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Benra 30 mg and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in DSS at Week 40

• 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:
  • Benra 30 mg: mean -0.34 versus -0.16 on control; difference -0.18 (95% CI -0.32 to -0.04), standardized effect d = -0.26; z-test p = 0.013.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Benra 30 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in NPS at Week 56

• 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:
  • Benra 30 mg: mean -0.22 versus 0.18 on control; difference -0.40 (95% CI -0.75 to -0.05), standardized effect d = -0.25; z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Benra 30 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes span symptoms and quality of life, disease severity, nasal polyp scores at a later assessment, and time to surgery and/or systemic corticosteroid use. The posted continuous-outcome results do not all show the same degree of statistical clarity. The SNOT-22 comparison at week 40 has confidence intervals that include zero, whereas the disease severity score comparison at week 40 and the nasal polyp score comparison at week 56 have intervals below zero in both the unadjusted calculations and reported ANCOVA analyses. This broad pattern is directionally compatible with the primary nasal polyp score result for some measures, but it is not uniform across outcomes.

These secondary outcomes address related but distinct questions. SNOT-22 is a patient-reported measure, disease severity score is another clinical measure, and later nasal polyp score concerns a different assessment time. A consistent direction across related outcomes may add context, but differences in measurement, timing, missingness, and analysis mean they should not be treated as identical evidence. The registry lists time-to-event outcomes, but numerical results for those outcomes are not reported in the registry record in the information available for this analysis.

Unadjusted secondary results carry less inferential weight when many outcomes are assessed because the chance of observing at least one apparently notable result can increase across multiple comparisons. Whether and how the trial controlled multiplicity is not reported in the registry record. The secondary analyses are therefore best read as supportive or exploratory context rather than as a collection of independent confirmations of the primary finding. Their interpretation should also distinguish the registered ANCOVA results from simple unadjusted recalculations.

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

• Change From Baseline in SNOT-22 at Week 40
• Time to First NP Surgery and/or SCS Use for NP to Week 56
• Time to the First NP Surgery up to Week 56
• Change From Baseline in DSS at Week 40
• Change From Baseline in NPS at Week 56
• Change From Baseline in NBS at Week 56
• Change From Baseline in SNOT-22 at Week 56
• Change From Baseline in DSS at Week 56

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant has a numerical change score, and the registry reports means and standard deviations. A mean summarizes the central value of a group, while a standard deviation describes the spread of participant measurements around that mean. These are different quantities. A group mean can be estimated with some uncertainty, and the uncertainty in a between-group contrast is conveyed by its confidence interval.

What the unadjusted comparison means

The recalculated mean difference compares the two posted group means for the endpoint. The placebo mean is -0.17? No: the posted placebo mean is 0.17, and the benralizumab mean is -0.36; the recalculated contrast is -0.53. This unadjusted calculation uses the arm-level summary statistics and does not adjust for baseline values or other covariates. It is a transparent descriptive comparison, but it is not a substitute for the sponsor's specified analysis. The registry record lists ANCOVA as the method for the reported primary analysis.

A p-value summarizes how incompatible the observed data are with a statistical null model of no group difference, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The recalculated unadjusted p-value is < 0.001; the registry's reported ANCOVA p-value is < 0.001. These values belong to different analyses and should be read alongside their corresponding estimates and confidence intervals.

Why ANCOVA is relevant

ANCOVA is a regression-based method for comparing outcomes between groups while accounting for selected covariates. In a trial with a baseline measure, an adjusted analysis can account for baseline outcome values, among other covariates if included. That is a general description of the method; the covariates included in this study's model are not reported in the registry record. The reported ANCOVA estimate is a net mean difference, while the simple recalculation is based on the posted group means. Their estimates can differ because adjustment changes the model-based comparison and its precision.

The posted primary description specifies how unusual post-randomization events and missing values were handled. Data collected after nasal polyp surgery and/or systemic corticosteroids for nasal polyposis were set to missing. For summary statistics, the registry describes worst-possible values after surgery and worst-observation carried forward after systemic corticosteroids. For ANCOVA, it describes a hybrid approach using worst-possible values after surgery, worst-observation carried forward after systemic corticosteroids, and multiple imputation assuming missing at random to build complete imputation datasets. These rules are part of the estimand and analysis; they affect what the reported treatment contrast represents.

Missingness and imputation in general

Missing data can matter because the observed outcomes may not represent all randomized participants if the reason for missingness is related to outcome or treatment. Multiple imputation is a general method that creates multiple plausible completed datasets under stated assumptions, analyzes each, and combines the resulting estimates to reflect imputation uncertainty. The registry specifies a missing-at-random assumption for the multiple-imputation component, but further implementation details are not reported in the registry record. A missing-at-random assumption means, in general, that after accounting for variables in the imputation model, the probability of missingness does not depend on the unobserved value itself. The validity of that assumption cannot be judged from the endpoint summary alone.

Worst-observation carried forward and worst-possible imputation are distinct handling rules. The former uses a participant's worst observed value as a replacement after the specified rescue event; the latter assigns the worst possible outcome under the endpoint's scoring framework. These rules can be conservative in some settings, but their implications depend on the endpoint and data pattern. The record gives the rules, but does not report enough detail to assess their practical impact for individual participants.

Spread, standard errors, and over-dispersion

The arm standard deviations characterize outcome variability. Standard errors, in contrast, characterize the uncertainty of an estimated quantity such as a group mean or difference. General statistical reasoning: greater variability can make an estimate less precise when other features are held constant. If outcomes are more variable than a model assumes, standard errors may be understated and confidence intervals too narrow; this is one reason over-dispersion matters in analyses whose variance assumptions do not fit the data. The registry endpoint is a continuous score analyzed using ANCOVA, and the registry record does not report an over-dispersion assessment. No claim about over-dispersion in this study is therefore warranted.

The primary analysis population was the full analysis set, defined in the registry as all randomized participants who received any investigational product, irrespective of protocol adherence and continued participation, analyzed according to randomized treatment. This is a treatment-policy style assignment principle in the sense that participants remain grouped by randomized treatment; that general characterization does not resolve how every post-randomization event is handled. The endpoint description supplies specific rescue and imputation rules. Further population definitions, protocol deviations, and analysis-set counts are not reported in the registry record.

Finally, statistical significance and precision answer different questions. A small p-value indicates evidence against the null model under the analysis assumptions. The confidence interval gives a range of effect values compatible with the data and model at the stated confidence level. Neither alone tells whether an effect is important to patients. A clinical-importance threshold for total nasal polyps score is not reported in the registry record.

11. Confidence Intervals and Precision

The primary ANCOVA estimate is -0.570, with a 95% confidence interval from -0.852 to -0.289. In general, a confidence interval is a range of values compatible with the observed data and the statistical model at the stated confidence level. It conveys uncertainty about the estimated group contrast; it is not a range containing a fixed proportion of individual participant outcomes, nor does it mean there is a 95% probability that the true effect lies in this particular interval.

Because the primary interval is entirely below zero, the registry's adjusted analysis is consistent with a negative mean difference rather than no difference. Its endpoints also convey that the magnitude is not known exactly. The unadjusted recalculation has a 95% interval from -0.822 to -0.238. The two intervals are similar in direction and overlap, but their estimates and widths arise from different analyses. The recalculated interval uses posted group summaries; the registry's interval comes from ANCOVA.

Precision is influenced by the amount of information and the variability of the outcome, as well as the analysis model. The reported endpoint groups each include 187 participants, and their standard deviations are 1.66 and 1.18. These counts and spreads help contextualize the estimate but do not, by themselves, explain all features of the ANCOVA interval. Details such as model covariates and imputation implementation are not reported in the registry record.

The confidence interval addresses statistical uncertainty in the mean contrast, not the range of likely responses for an individual participant. It also does not establish a threshold for clinical importance. Such a threshold is not reported in the registry record. The interval should therefore be read as evidence about the direction and precision of the estimated group-level score difference, alongside the endpoint definition and missing-data approach.

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 effect, not the original design justification for the trial. For this continuous endpoint, the figures are participants per arm, not total events. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each row plans for a stated fraction of the observed effect: 100%, 75%, or 50% of that observed effect.

The 80% power column gives the illustrated number of participants per arm for an analysis to have an 80% chance of detecting the assumed effect under those planning assumptions, if that effect and the assumptions were correct. The 90% power column makes the same type of illustration for a 90% chance. These are planning probabilities, not guarantees about a future study. The registry record's own original sample-size assumptions and justification are not reported.

The table shows that planning for smaller effects requires more participants per arm. For example, planning for 100% of the observed effect gives 116 participants per arm for 80% power and 156 for 90% power, while planning for 50% of the observed effect gives 464 and 621, respectively. The larger requirements arise because a smaller effect is harder to distinguish from random variation under the same general assumptions.

These values do not describe the probability that the completed trial found a real effect. This is not observed power. They are hypothetical design illustrations using the observed effect as a starting point and should not be read as evidence that the trial's original enrollment target was chosen by this calculation.

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 = -0.530; pooled standard deviation σ = 1.440.
2. n = 2 × (1.96 + 0.842)² × 1.440² / -0.530² = 115.9, rounded up to 116 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 <- -0.53                      # observed difference in means
sigma <- 1.44013888219158                     # 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(3401229)
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 = 116 (80% power), 156 (90% power)
 75% of observed effect: n per arm = 207 (80% power), 276 (90% power)
 50% of observed effect: n per arm = 464 (80% power), 621 (90% power)
power.t.test: 116.9 
Simulated power at n = 116 per arm: 0.797 (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 registry's primary analysis is the sponsor-reported ANCOVA result: net mean difference -0.570, 95% confidence interval -0.852 to -0.289, p < 0.001. The unadjusted recalculation from posted arm summaries is -0.53, with a 95% confidence interval from -0.822 to -0.238 and p < 0.001. Both analyses point in the same direction and both intervals exclude zero.

The values differ because the recalculation is a simple unadjusted contrast of group means, whereas ANCOVA is a model-based adjusted analysis. The registry does not report the covariates included in ANCOVA, so the reason for the precise difference cannot be determined from the record. The recalculation is useful as a check on the basic arm-level pattern, but it does not reproduce the full analysis approach, including the stated handling of rescue events and missing data.

The registry also marks the ANCOVA analysis as a superiority analysis. That label identifies the reported comparison framework; it does not mean the result alone answers every question about clinical importance, patient experience, or safety. The posted registry analysis is the relevant primary inferential result, while the recalculation provides a separate, unadjusted reading of the summary statistics.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 35 outcome measures and 22 analyses overall, and 33 secondary outcomes are posted. Testing multiple endpoints or time points raises a multiplicity question: when many hypotheses are examined, the chance of at least one statistically notable result can increase. General statistical practice may address this through a prespecified hierarchy, adjustment of significance levels, or other procedures. The multiplicity strategy for this trial is not reported in the registry record, so it is not possible to determine whether the secondary p-values were adjusted or how they were intended to support confirmatory claims.

The primary endpoint is clearly identified as change from baseline in total nasal polyps score at week 40, and the registry reports an ANCOVA result for it. The registry record does not report a complete statistical analysis plan, including details of any interim analyses, endpoint hierarchy, or multiplicity control. Those aspects should not be inferred from the existence of listed outcomes or the reported p-values.

Missing-data handling is described specifically for the primary endpoint. Values after nasal polyp surgery and/or systemic corticosteroids for nasal polyposis were set to missing. For descriptive summaries, worst-possible values after surgery and worst-observation carried forward after systemic corticosteroid use were applied. For ANCOVA, the registry describes a hybrid approach using those rules together with multiple imputation assuming missing at random. The imputation model details, the amount and pattern of missing data, and sensitivity analyses are not reported in the registry record.

The full analysis set includes all randomized participants who received any investigational product, regardless of protocol adherence or continued participation, and analyzes participants according to randomized treatment. This definition supports an analysis organized by randomized assignment, but the record does not provide enough detail to judge how exclusions, deviations, or missing outcomes affected the final endpoint dataset. The analysis population definition and missing-data rules should be read together rather than assuming every randomized participant contributed an observed week 40 score.

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: pericarditis (1.0%), gastritis (1.0%), lymphadenopathy (0.5%), arrhythmia (0.5%), rhegmatogenous retinal detachment (0.5%).

The registry lists serious adverse events for 207 participants at risk in the benralizumab group, with 25 affected, and for 203 participants at risk in the placebo group, with 19 affected. Deaths are recorded as 0 in each group. These counts describe the posted safety summaries; they do not establish that an event was caused by treatment. The registry information does not provide a causal assessment for each event.

The listed serious adverse event terms and percentages include pericarditis at 1.0%, gastritis at 1.0%, lymphadenopathy at 0.5%, arrhythmia at 0.5%, rhegmatogenous retinal detachment at 0.5%, and appendicitis perforated at 0.5%. These are registry-listed terms; the record summary does not establish that every term occurred in both groups. The denominators and further term-by-group breakdown are not reported in the registry record in this information.

Safety counts should be interpreted in relation to the number at risk and the way events were collected and classified. Counts alone do not describe duration, severity beyond the serious-event classification, or attribution. The registry also lists other adverse events affecting 64 participants at risk in the benralizumab group and 83 in the placebo group. No deaths were recorded in either group. These descriptive figures should not be converted into conclusions about comparative safety without fuller event definitions and analysis details.

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

The primary result describes a between-group difference in mean change on a centrally read endoscopic score at week 40. The registry's ANCOVA estimate and confidence interval indicate a negative mean contrast for benralizumab versus placebo, with the interval not including zero. The unadjusted calculation from the posted group summaries points in the same direction. This is evidence about the average score contrast under the reported analysis, not a prediction for every participant.

A group mean can conceal variation in individual outcomes. The registry reports standard deviations, but those do not show the full distribution or how many participants experienced improvement, no change, or worsening. The mean difference also does not state how participants valued a given score change. A threshold for clinical importance is not reported in the registry record, so statistical evidence should not be equated automatically with a specific level of patient-perceived benefit.

The primary endpoint is one component of a wider set of outcomes. Symptoms, quality of life, disease severity, later score changes, and use of surgery or systemic corticosteroids are separate domains. Some posted secondary continuous results have intervals that include zero, while others do not; the registry does not report numerical results for every listed time-to-event outcome in this information. The primary result cannot stand in for all of those questions.

The analysis also depends on how outcomes after rescue treatment and missing observations were handled. The record describes specific imputation rules for the primary endpoint, but does not report the full data pattern or sensitivity analyses needed to assess how alternative assumptions might change the estimate. The result should therefore be read as the trial's reported estimate under its stated analysis approach, not as a guarantee that the same contrast would arise under every possible outcome definition or missing-data assumption.

These statistics do not provide medical advice and do not establish a treatment decision for an individual. They describe comparative trial data for a defined population and endpoint. Broader conclusions require attention to the totality of outcomes, uncertainty, safety information, and the limits of what the registry record reports.

18. Why This Trial Matters Statistically

This trial illustrates why the definition of an endpoint matters as much as the headline comparison. Total nasal polyps score is a structured, bilateral endoscopic measure, with central reading and a specified scoring scale. The treatment contrast therefore has a clear measurement target. Readers can interpret the sign of the estimate only after understanding that higher scores mean heavier polyp burden and that the outcome is change from baseline.

It also shows the distinction between an unadjusted summary-statistic comparison and the trial's reported ANCOVA analysis. The estimates are close in direction and both intervals lie below zero, yet they are not interchangeable. The recalculation helps describe the raw arm-level contrast, while the sponsor-reported model is the relevant adjusted result. Understanding that difference prevents a simple reanalysis from being mistaken for a reproduction of the trial's full method.

The handling of rescue treatment and missing values is another important statistical feature. Surgery and systemic corticosteroid use are clinically meaningful post-randomization events, and the registry's analysis rules explicitly incorporate them into the outcome construction. This makes clear that trial estimates depend not only on observed scores but also on how intercurrent events and missing observations are treated.

Finally, the wider set of secondary outcomes demonstrates the importance of multiplicity and cautious interpretation. Related endpoints can provide context without being equivalent tests of the primary claim. When the full hierarchy and adjustment strategy are not reported in the registry record, secondary findings should not be counted as multiple independent confirmations. This is a useful example of reading estimates, confidence intervals, analysis populations, and missing-data rules together rather than focusing on a p-value alone.

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 registry record does not report the original sample-size justification, detailed trial design assumptions, or the statistical analysis plan. The planning illustrations based on the observed effect are not a replacement for those original documents. In particular, the record does not provide enough information to assess the assumptions used to choose the study's enrollment target.

Important details of the primary ANCOVA model are not reported, including the covariates and their specification. Although the endpoint description gives a hybrid missing-data approach, it does not provide the imputation model, the number or pattern of missing values, or sensitivity analyses under alternative assumptions. The practical influence of the surgery and systemic corticosteroid rules therefore cannot be evaluated from the posted summaries alone.

The registry lists many outcomes, but a multiplicity strategy and endpoint hierarchy are not reported. This limits interpretation of secondary p-values as confirmatory evidence. Numerical results for all listed outcomes, including the listed time-to-event endpoints, are not reported in the registry record in the information summarized for these results. The apparent pattern among the posted continuous secondary outcomes should not be generalized to endpoints without numerical results.

The primary endpoint is a score-based measure, and the registry does not report a clinical-importance threshold for interpreting a given mean difference. Arm means and standard deviations do not reveal the full distribution of participant responses or the proportion meeting a responder definition. Detailed adverse-event definitions and attribution information are likewise not reported in the registry record. These limitations do not negate the posted statistical result, but they constrain claims about clinical meaning, robustness, and the full benefit-risk profile.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Efficacy and Safety Study of Benralizumab for Patients With Severe Nasal Polyposis — NCT03401229.
• Registry record last updated 2021-10-12; status last verified 2021-10.
• Results first posted 2021-10-12.
• 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

NCT03401229 is a completed randomized, triple-masked, parallel-group phase 3 study sponsored by AstraZeneca, with enrollment of 413. It compared benralizumab 30 mg subcutaneously plus mometasone furoate with matching placebo plus mometasone furoate in patients with severe nasal polyposis.

The primary endpoint was change from baseline in total nasal polyps score at week 40. The registry-reported ANCOVA mean difference was -0.570, with a 95% confidence interval from -0.852 to -0.289 and p < 0.001. The unadjusted recalculation was -0.53, with a 95% confidence interval from -0.822 to -0.238 and p < 0.001.

The registry describes special handling of outcomes after surgery and systemic corticosteroid use, along with multiple imputation for the ANCOVA analysis. The original sample-size justification, multiplicity strategy, detailed model specification, and several missing-data diagnostics are not reported in the registry record.