A Study to Evaluate the Efficacy and Safety of Reslizumab in Patients With Eosinophilic Asthma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Efficacy and Safety of Reslizumab in Patients With Eosinophilic Asthma (NCT01285323), a randomized phase 3 study. The primary outcome is Frequency of Clinical Asthma Exacerbations (CAEs) During 12 Months of Treatment.

2. Clinical Question

The clinical question is framed around exacerbations rather than a single measurement of lung function or symptoms. An exacerbation required specified worsening and treatment circumstances, with corroboration from at least one other measurement; events were adjudicated by committee. This definition aims to make the outcome more consistent than a count based only on a patient's report or an isolated change in a clinical measure.

The randomized comparison is intended to distinguish the observed treatment-group contrast from differences that might arise simply because patients entering one group differed from those entering the other. The registry identifies the trial as parallel-group and quadruple-masked, while its official title calls it double-blind. Which people were included in the masking designation is not reported in the registry record.

3. Trial Design

This was a phase 3, randomized, parallel-group study comparing reslizumab with placebo in patients with eosinophilic asthma. The official title describes a 12-month, double-blind, placebo-controlled study. The registry separately records masking as quadruple; it does not identify the people or roles covered by that designation. The study status is completed. The recorded start is 2011-03, and primary completion is recorded as 2014-04.

The intervention was reslizumab 3.0 mg/kg given intravenously once every 4 weeks, with a plus-or-minus 7-day window, for a total of 13 doses. The placebo group followed the same stated schedule and route. The registry lists enrollment as 464, and the primary outcome results include 232 patients in each group. These numbers describe the registry’s enrollment and outcome entries; the record does not explain in detail how every analysis handled participants with unavailable measurements.

The primary endpoint covered Day 1 to Month 12 and was the frequency of clinical asthma exacerbations. The primary population was the randomized set, defined as all patients assigned to a group at enrollment regardless of whether they took study drug. The endpoint definition allowed an event to qualify through either or both of two clinical pathways, with corroboration by at least one additional measurement. A committee adjudicated the events to support consistency in deciding which events met the definition.

The study also assessed lung function, asthma-related quality of life and control, time to first exacerbation, symptom utility, short-acting beta-agonist use, and blood eosinophil count. The registry lists 14 outcome measures and 10 analyses, and 12 secondary outcomes are posted. The sample-size justification, detailed allocation procedure, and full statistical analysis plan are not reported in the registry record. In particular, the record does not say how the enrollment target was selected or specify all planned handling of missing measurements and multiple secondary comparisons.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Randomized set includes all patients who were randomly assigned to a treatment group at enrollment, regardless of whether or not a patient took any study drug.

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: Randomized participants

6. Primary Endpoint

The registry lists the primary outcome as Frequency of Clinical Asthma Exacerbations (CAEs) During 12 Months of Treatment, measured over Day 1 to Month 12.

The primary endpoint is the frequency of clinical asthma exacerbations during 12 months of treatment, expressed as clinical asthma exacerbations in 52 weeks. It is a count-based outcome: it records qualifying events over a stated period rather than measuring the severity of a single episode or a continuous change in lung function. The registry reports adjusted means for the treatment groups and a rate ratio for the reslizumab-versus-placebo comparison.

An event counted as a clinical asthma exacerbation if the patient met either or both of the registry’s treatment-related criteria, and the worsening was corroborated with at least one other measurement. One criterion concerned systemic corticosteroid use or increased inhaled corticosteroid use for 3 or more days, including a specified increase for patients already using corticosteroids. The other concerned asthma-related emergency treatment or hospitalization. The event definition and committee adjudication were intended to establish a consistent basis for counting episodes.

This endpoint is useful for asking whether treatment groups differ in the frequency of clinically consequential worsening over the study period. It incorporates episodes that led to treatment escalation or urgent care, rather than treating every change in symptoms or lung function as an exacerbation. The endpoint does not, by itself, describe every aspect of asthma control, patient experience, or safety. Those questions require consideration of the secondary measures and adverse-event information as separate evidence.

The registry states that adjusted rates and confidence intervals were based on negative binomial regression adjusted for stratification factors. The stratification factors themselves are not reported in the registry record. The reported group estimates are adjusted means, so they should not be read as simple raw event counts.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a count-based rate, rather than a normally distributed continuous measurement. The registry says adjusted exacerbation rates and confidence intervals were based on negative binomial regression with adjustment for stratification factors. The listed comparison is a rate ratio, and the reported analysis method is Chi-squared. The registry does not explain the relationship between that method label and the regression-based estimation in further detail.

Why use a count model?

Counts of events often have a different distribution from continuous measurements such as lung function. A simple count model may assume that the variability is closely linked to the mean. In some settings, event counts vary more than that assumption allows, a feature called over-dispersion. A negative binomial model is a general way to accommodate extra variation in count data. This is a general explanation of the method; the registry does not report a diagnostic assessment of over-dispersion for this trial or the model specification beyond the stated adjustment.

Adjustment for stratification factors means that the analysis accounts for factors used to form strata in the randomized comparison. This can preserve the intended structure of the randomized design in the analysis. The factors and the full adjustment details are not reported in the registry record, so their contribution to the estimate cannot be assessed from the posted summary.

How to read the rate ratio and interval

A rate ratio compares the estimated event rates between groups. A value of 1 represents equal rates; a value below 1 represents a lower estimated rate in the numerator group, which is reslizumab relative to placebo. The 95% confidence interval describes uncertainty around that relative estimate under the analysis assumptions. Because the reported interval, 0.282 to 0.586, lies below 1, the interval is consistent with a lower rate in the reslizumab group under the fitted analysis. The p-value, p < 0.001, is evidence against the null comparison of no difference under the test used. A p-value is not the probability that the null hypothesis is true, and it does not measure clinical importance.

The group-level adjusted means are estimates of event frequency over the stated period. They are not raw totals. Their own confidence intervals express uncertainty for each group estimate; the rate ratio interval addresses the relative comparison. A reader should not infer a comparison’s uncertainty by looking only at whether the separate group intervals overlap. The reported comparison directly quantifies the relative contrast.

Why secondary recalculations can differ

For selected continuous secondary outcomes, the posted recalculations use group summaries to form simple, unadjusted comparisons. They are approximate because standard deviations are estimated from the reported standard errors, and they do not reproduce the sponsor’s repeated-measures model. The recalculated intervals and p-values therefore answer a simpler question under different assumptions. They are useful as a check on scale and uncertainty, but they are not substitutes for the sponsor’s model-based results.

For example, the Week 16 lung-function recalculation gives a mean difference of 0.101 liters, with an approximate 95% confidence interval from -0.023 to 0.225 and an unadjusted p-value of 0.109. The reported repeated-measures analysis gives a 95% confidence interval from 0.023 to 0.179 and p = 0.011. The contrast in uncertainty reflects the different calculations and analysis approaches; the registry’s sponsor analysis is the result to use when describing the trial’s reported analysis.

In general, a standard error quantifies uncertainty in an estimated mean or contrast. A confidence interval is built using an estimate and its uncertainty, along with assumptions about the sampling distribution. The interval’s width depends on the amount of information and the variability represented by the model. A non-significant recalculated p-value does not establish that there is no effect; it indicates that the simple calculation does not distinguish the contrast from zero at conventional thresholds. Conversely, a significant p-value does not say whether the magnitude is important to patients.

Planning illustrations are not observed power

The sample-size illustrations use the observed secondary-outcome effects as planning inputs under stated assumptions: a two-sided significance level of 0.05, equal allocation, and a normal approximation. Rows consider planning for 100%, 75%, or 50% of the observed effect, and report participants per arm for 80% power and participants per arm for 90% power. These are hypothetical planning calculations, not the trial’s original sample-size justification and not observed power. The sample-size assumptions and the trial’s own design justification are separate matters; the latter is not reported in the registry record.

10. Analyses Reported by the Sponsor

The registry’s sponsor-reported analyses use mixed model repeated measures for the posted continuous outcomes. The recalculations are simple, unadjusted comparisons derived from the reported group summaries. They are not attempts to recreate the repeated-measures model, and their intervals and p-values can differ because their assumptions and available information differ.

For the Week 16 FEV1 outcome, the reported mean difference is 0.101 liters, with a 95% confidence interval from 0.023 to 0.179 and p = 0.011. The recalculation gives 0.101, with an approximate interval from -0.023 to 0.225 and p = 0.109. For FEV1 over 16 weeks, the reported mean difference is 0.093 liters, with a 95% confidence interval from 0.030 to 0.155 and p = 0.004; the recalculation gives 0.093, with an approximate interval from -0.021 to 0.207 and p = 0.109. For AQLQ, the reported mean difference is 0.209, with a 95% confidence interval from 0.025 to 0.393 and p = 0.026; the recalculation gives 0.21, with an approximate interval from -0.110 to 0.530 and p = 0.199.

The near agreement in some point estimates does not make the analyses interchangeable. The reported model is the sponsor’s analysis of the trial data; the recalculations are approximate checks based on summary statistics. The registry does not flag a discrepancy for this record. The reported model-based estimates should therefore be used when describing the sponsor’s findings, while the recalculated results illustrate how conclusions can depend on the analysis method.

11. Multiplicity, Missing Data and Analysis Populations

The registry identifies the primary endpoint population as the randomized set, including all patients assigned to a group at enrollment regardless of whether they took study drug. This definition is important: it describes the population targeted by the primary analysis and preserves assignment as the basis of comparison. The record does not give a detailed account of the analysis population used for every secondary outcome or explain how missing measurements were handled in each model.

Missing data can affect both precision and interpretation. If a measurement is unavailable, the analysis must use an approach that makes assumptions about the unobserved value or the process leading to missingness. Different approaches can yield different estimates, especially if missingness is related to treatment, health status, or the outcome. The registry record does not report the missing-data assumptions, imputation procedures, sensitivity analyses, or numbers of missing observations for the endpoints. It is therefore not possible to assess from this record how robust the results are to alternative missing-data assumptions.

Multiplicity is another unresolved issue for the secondary outcomes. When several hypotheses are tested, interpreting each p-value as if it were the only test can overstate the evidence across the full set. Statistical plans may control the family-wise chance of false-positive findings, use a hierarchical testing sequence, or designate some outcomes as descriptive. The registry record does not report which approach, if any, was planned for the secondary endpoints. The posted secondary p-values should therefore be interpreted individually and cautiously, not as proof that every measured domain changed.

The sample-size justification, any interim analyses, and the detailed statistical analysis plan are not reported in the registry record. There is also no basis in the record to infer whether interim monitoring or additional decision rules were used. These omissions limit assessment of how the analyses were planned before outcome data were available. They do not erase the reported primary result, but they constrain independent evaluation of the full analytic strategy.

12. 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: pneumonia (2.6%), asthma (2.6%), road traffic accident (1.3%), anaphylactic reaction (0.9%), contusion (0.9%).

The registry’s adverse-event entries list serious adverse events affecting 23 patients in the placebo group and 18 patients in the reslizumab 3.0 mg/kg group. The corresponding groups each have 232 patients at risk in the posted entries. For other adverse events, the record lists 160 patients affected in the placebo group and 123 in the reslizumab group, with 232 at risk in each group. These counts describe the posted safety entries; they do not establish that every event was caused by study treatment.

The listed serious-event terms include pneumonia and asthma at 2.6% each, road traffic accident at 1.3%, anaphylactic reaction and contusion at 0.9% each, and extrasystoles at 0.4%. These are the registry’s listed percentages for selected serious-event terms. The record does not give enough information to infer the basis of each percentage, the number of episodes per patient, or whether one patient could contribute to multiple terms. Death counts are not reported in the registry record.

Safety counts should be considered separately from the primary efficacy endpoint. They do not, on their own, quantify the comparative risk of each event or establish causation. The registry record also does not provide a complete narrative of event timing, severity, relatedness assessment, or exposure-adjusted safety comparisons in the information.

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

The primary result is evidence about the frequency of adjudicated clinical asthma exacerbations in the randomized groups during the study’s treatment period, as estimated by the registry’s adjusted analysis. The reported rate ratio and its confidence interval favor a lower estimated exacerbation rate in the reslizumab group than in the placebo group. The small reported p-value indicates that the observed comparison is difficult to reconcile with a no-difference hypothesis under the stated analysis assumptions.

That conclusion is narrower than saying that treatment works for every patient or improves every aspect of asthma. The trial’s primary endpoint is one outcome, in a defined population and study setting. It does not determine an individual patient’s likely course, and it does not directly answer questions about every symptom, lung-function measure, quality-of-life outcome, or adverse event. The secondary outcomes provide additional evidence, but they have different definitions and timeframes and lack a reported multiplicity plan.

The adjusted means and rate ratio are estimates, not exact values known without uncertainty. Their confidence intervals describe uncertainty under the fitted model, and other reasonable assumptions or analysis choices could affect estimates. The registry does not report the full model specification, the stratification factors, a detailed missing-data strategy, or sensitivity analyses. Those gaps limit independent assessment of robustness.

Nor should the primary finding be interpreted as a direct comparison of the number of patients who experienced an event. The endpoint measures event frequency, so repeated events may contribute to a patient’s outcome. The reported adjusted rates are not raw event totals. The registry’s event criteria and committee adjudication define what counted, but the summary alone does not describe every clinical circumstance surrounding those events.

In short, the posted analysis supports a difference in the primary exacerbation-rate endpoint between the randomized groups. It does not by itself establish the overall balance of benefit and harm, a universal treatment effect, or effects outside the study’s population and follow-up. Those broader judgments require evidence beyond this endpoint and registry summary.

14. Why This Trial Matters Statistically

This trial illustrates why the choice of endpoint and statistical model matters in clinical research. Clinical asthma exacerbations are repeated events over time, not a single continuous measurement. The registry therefore reports adjusted rates and a rate ratio, with a negative binomial regression model for the primary endpoint. A model suited to event counts can represent the outcome differently from a comparison of group averages on a continuous scale.

The primary result also demonstrates the value of separating a treatment-group estimate from its uncertainty. The rate ratio gives the relative comparison, while its confidence interval indicates the precision of that estimate under the analysis. The p-value addresses compatibility with a no-difference hypothesis, not clinical importance or the probability that the treatment is effective. Reading these statistics together avoids relying on a single numerical summary.

The trial’s secondary results show another important statistical issue: analysis choices can affect apparent precision. The recalculations based on group summaries are approximate and unadjusted, while the sponsor-reported continuous-outcome analyses use mixed models for repeated measures. Even when point estimates are similar, intervals and p-values can differ because the methods use information and assumptions differently. This is why a transparent distinction between a sponsor analysis and a simple recalculation matters.

Finally, a set of secondary outcomes raises questions about multiplicity, missing data, and prespecification. Without a reported analysis plan, readers cannot determine how the secondary tests were prioritized or whether their error rates were controlled. The record supports a clear description of the posted primary comparison, while also making the boundaries of what can be inferred visible. That combination—interpreting the main estimate while acknowledging uncertainty and incomplete analytic detail—is central to sound biostatistical reading.

15. 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 summary does not report the full statistical analysis plan, the sample-size justification, or any interim analyses. It also does not identify the stratification factors used in the primary negative binomial model. These omissions limit assessment of how the analysis was planned, whether all decisions were prespecified, and how adjustment affected the primary estimate.

Missing-data methods and endpoint-specific analysis populations are not detailed for the secondary outcomes. The primary population is defined as the randomized set, but the record does not explain the handling of missing follow-up data or provide sensitivity analyses for alternative assumptions. This is especially relevant to repeated measurements, for which available observations can differ across visits and participants.

The registry includes selected secondary results and lists a larger set of outcome measures. A multiplicity strategy is not reported, so the level of confirmatory evidence for secondary comparisons is uncertain. The sponsor’s repeated-measures results and the approximate unadjusted recalculations are also not identical analyses; the recalculations should not be treated as independent replications or as replacements for the model-based estimates.

The posted safety summary gives counts of patients affected and selected serious-event terms, but not a complete account of event timing, duration, severity, relatedness, or exposure-adjusted comparisons. Death counts are not reported in the registry record. Consequently, these entries cannot alone establish causality or fully characterize comparative safety.

Finally, the registry’s summary results do not provide individual-level data, detailed baseline characteristics, or enough modeling detail to reproduce the adjusted estimates. These limitations do not negate the reported primary result, but they constrain how far the findings can be independently checked or generalized beyond the trial population and endpoint.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate the Efficacy and Safety of Reslizumab in Patients With Eosinophilic Asthma — NCT01285323.
• Registry record last updated 2021-11-09; status last verified 2021-11.
• Results first posted 2016-06-27.
• 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.

19. Record Summary

The registry describes a completed phase 3 randomized, parallel-group, placebo-controlled study of reslizumab 3.0 mg/kg in patients with eosinophilic asthma. Treatment was administered intravenously once every 4 weeks, with a plus-or-minus 7-day window, for a total of 13 doses. The primary endpoint was the frequency of clinical asthma exacerbations during 12 months of treatment, analyzed in the randomized set.

The adjusted mean was 2.115 exacerbations in 52 weeks for placebo and 0.859 for reslizumab. The reported rate ratio was 0.406, with a 95% confidence interval from 0.282 to 0.586 and p < 0.001. The registry also posts secondary outcome results and adverse-event counts. Interpretation of secondary comparisons is limited by the absence of a reported multiplicity strategy, and the record does not describe detailed missing-data methods, the sample-size justification, or the full analysis plan.