A Proof of Concept Study to Assess Effect of Fluticasone Furoate (FF)/Levocabastine Fixed Dose Combination (FDC) Compared With Levocabastine and FF Alone in Subjects With Allergic Rhinitis (AR): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Proof of Concept Study to Assess Effect of Fluticasone Furoate (FF)/Levocabastine Fixed Dose Combination (FDC) Compared With Levocabastine and FF Alone in Subjects With Allergic Rhinitis (AR) (NCT01957202), a randomized phase 2 study. The primary outcome is Weighted Mean of the Total Nasal Symptom Score (TNSS) (0-4) Hours (h) Post Start of Allergen Chamber Challenge on Day 8 of Each Treatment Period.

2. Clinical Question

The study was a Phase 2 proof-of-concept investigation in participants with allergic rhinitis. Its stated rationale was to assess the combination relative to its components. The record reports least-squares means with 95% confidence intervals for the primary outcome and mixed-model ANOVA estimates for pairwise comparisons. Those results describe responses during a controlled allergen challenge, not necessarily symptoms in everyday settings.

3. Trial Design

This was a randomized, double-blind, placebo-controlled, three-way incomplete-block crossover study. The record lists enrollment of 71 and describes three treatment periods, with each participant assigned to a sequence of three treatments. Example sequences are ABC, BCD, and ACD. A washout period of 14-28 days was planned between treatment periods, and the expected study duration for an individual was a maximum of up to 20 weeks, including screening and follow-up.

The listed interventions were FF/levocabastine, FF, levocabastine, and placebo. The arm description specifies FF 100 μg, levocabastine 200 μg, and FF 100 μg/levocabastine 200 μg, administered as two sprays per nostril. Because the design is an incomplete block, not every participant necessarily received every listed intervention. The record does not give the full sequence allocation details or explain how treatment order was generated.

The primary outcome was measured during an allergen challenge in an Environmental Exposure Chamber. The primary timeframe is Day 8 of each treatment period, with the endpoint calculated over the 0-4 hours after challenge began. The registry lists primary completion in 2014-02 and status as completed. It identifies GlaxoSmithKline as sponsor.

The record names a mixed-model ANOVA for the reported pairwise analyses. More detail about the model specification, its treatment of period or sequence, and its assumptions is not reported in the registry record. A sample-size justification and any interim-analysis plan are also not reported in the registry record. The planning assumptions associated with illustrative sample-size calculations are not the same as the trial’s design justification or an observed-power estimate.

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: Participants who met the eligibility criteria at Screening were randomized to 1 of 18 treatment sequences. The treatment phase was comprised of three 8-day treatment periods, each separated by a 14- to 28-day washout period.

Only the first 6 of 18 groups are shown.

Treatment Period 1 (8 Days)

Washout Period 1 (14 to 28 Days)

Treatment Period 2 (8 Days)

Reasons for not completing this period:

Washout Period 2 (14 to 28 Days)

Analysis population for the primary outcome, as defined in the registry: Pharmacodynamic (PD) Population: all participants in the All Subjects Population (defined as all participants who received at least one dose of investigational product) and who also provided data from at least one PD assessment. Only those participants contributing data at the indicated time points were analyzed.

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.

6. Primary Endpoint

The registry lists the primary outcome as Weighted Mean of the Total Nasal Symptom Score (TNSS) (0-4) Hours (h) Post Start of Allergen Chamber Challenge on Day 8 of Each Treatment Period, measured over Day 8 of each treatment period (up to 80 days).

The primary endpoint is the weighted mean of the Total Nasal Symptom Score (TNSS) over 0-4 hours after the start of the allergen chamber challenge on Day 8 of each treatment period. TNSS adds ratings for nasal congestion, rhinorrhea, nasal itch, and sneezing. Each component is scored from 0 to 3, where 0 means none and 3 means severe, giving a total score ranging from 0 to 12.

Participants were assessed before the challenge and then every 15 minutes from 0 to 4 hours after challenge began. The record says the weighted mean was calculated as the area under the response-time curve over 0-4 hours, calculated by the trapezoidal rule, divided by the time interval of available data. In general, this kind of time-weighted summary represents symptom scores across the observation window rather than relying on a single measurement. The record’s stated reason for collecting symptoms during chamber exposure is to evaluate responses when aerosolized allergen is administered in a sealed environment.

Because the scale measures symptom severity, a lower TNSS indicates less recorded nasal symptom burden during the specified challenge window. The endpoint is a continuous measure, and the registry reports least-squares means and 95% confidence intervals. The measure is specific to the challenge setting, assessment window, and treatment-period day; it should not be interpreted as a direct measure of symptoms outside that setting.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical explanation: a least-squares mean is a model-based mean estimate. It can differ from a simple arithmetic average when the analysis model accounts for features of the design or data. The registry identifies the primary summary as a least-squares mean and gives a 95% confidence interval for each treatment. The detailed model specification and the exact adjustments used to obtain those means are not reported in the registry record.

What does the confidence interval convey?

In general, a confidence interval describes uncertainty around an estimate under the assumptions of the statistical method. A narrower interval indicates greater precision than a wider interval, though precision does not by itself establish clinical importance. For the primary means, the posted intervals are 5.35 to 6.71 for placebo, 3.646 to 4.731 for FF 100 μg, 3.955 to 5.044 for levocabastine 200 μg, and 1.438 to 2.428 for the combination. The record labels these intervals as 95% confidence intervals.

For a contrast, an interval that includes zero is compatible with no difference under the contrast’s scale and model, while an interval wholly on one side of zero indicates that the estimated contrast is not centered on zero at that confidence level. This is a statistical description, not a judgment about clinical importance. The registry does not state a clinically meaningful difference threshold for TNSS.

How should the reported differences be read?

The registry reports mean differences from mixed-model ANOVA for six pairwise comparisons. The estimates are -2.255 for FF 100 μg and the combination, -2.566 for levocabastine 200 μg and the combination, -1.531 for placebo and levocabastine 200 μg, -1.842 for placebo and FF 100 μg, -4.097 for placebo and the combination, and -0.311 for FF 100 μg and levocabastine 200 μg. The record gives 95% confidence intervals and p-values for each. Since the registry does not further define the subtraction order, retain the reported signs and comparison labels rather than silently changing the direction of a contrast.

In general, a p-value quantifies how unusual results at least as extreme as the observed result would be under a specified null hypothesis and model. It is not the probability that the null hypothesis is true, nor a measure of effect size. The very small reported p-values for several comparisons do not indicate how important the difference is to participants. The p-value of 0.370 for the FF-versus-levocabastine comparison is not proof that the treatments are identical.

Why does the crossover design matter?

General explanation: crossover designs compare treatments within participants, which can reduce the influence of stable differences between participants. They also require attention to treatment period, order, and possible carryover effects. A washout period is intended to separate treatment periods, but its presence alone does not establish that carryover was absent. The record describes a washout of 14-28 days and a mixed-model ANOVA; whether and how the analysis addressed period, sequence, or carryover is not reported in the registry record.

The incomplete-block structure means each participant was assigned to a sequence of three treatments, rather than necessarily receiving every listed treatment. Consequently, treatment-specific counts need not match. The record’s primary analysis population additionally requires at least one dose and at least one pharmacodynamic assessment, with only participants who contributed data at the indicated time points included for those time points. This definition makes clear that the endpoint summaries depend on available assessments, but details of missing-data handling are not reported in the registry record.

What do the planning assumptions mean?

General explanation: sample-size planning illustrations use assumptions about the target effect, variability, allocation, and statistical test to estimate how many participants per arm might be needed under a chosen design. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. Such illustrations are planning calculations, not observed power and not evidence that the original trial was planned using those assumptions. Numerical planning results and the trial’s own sample-size justification are not reported in the registry record.

The posted record gives 24 analyses and lists three secondary outcomes in addition to the primary endpoint. It does not describe a multiplicity adjustment, a hierarchy for testing, a missing-data method, or a detailed model formula. Those omissions limit how fully an external reader can reproduce the inferential analysis or assess how the reported p-values relate to the full set of analyses.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s pairwise results as mixed-model ANOVA estimates, with 95% confidence intervals and p-values. These are the comparison estimates available for interpreting the posted analyses. The record does not report an independent recalculated set of comparison estimates, so there is no numerical recalculation to place alongside the sponsor’s analyses. The discrepancy indicator is false.

The posted treatment means provide context for the contrasts, but they are not a substitute for the model-based comparison estimates in a crossover design. The sponsor’s reported contrast estimates should therefore be read with their stated confidence intervals, p-values, endpoint definition, and treatment-pair labels. Their statistical interpretation remains limited by information not reported in the registry record, including the detailed model specification and whether any adjustment was made for multiple comparisons.

The sponsor’s analyses address the recorded Day 8 TNSS endpoint in the allergen challenge setting. They do not by themselves establish the importance of a difference outside that setting, and a small p-value does not quantify clinical relevance. The record does not state a threshold for a clinically meaningful TNSS difference or provide a separate equivalence analysis for FF compared with levocabastine.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome, three secondary outcomes, and 24 analyses. It also reports six pairwise comparisons for the primary endpoint. The record does not state whether the analyses were arranged in a prespecified hierarchy, whether family-wise error or another multiplicity approach was used, or whether the reported p-values were adjusted. In general, when multiple hypotheses are tested, the chance of at least one false-positive result can increase unless an appropriate strategy is used. The absence of a reported adjustment means the displayed p-values should be read as reported, without assuming they account for the full set of analyses.

Missing-data procedures are not reported in the registry record. The primary population is described as participants who received at least one dose and provided at least one pharmacodynamic assessment, with only participants contributing data at the indicated time points included. This tells readers that endpoint data availability affected inclusion at the relevant times, but it does not explain how missing measurements were handled in the model, whether any imputation was used, or how sensitivity to missingness was assessed.

The record describes a crossover design with treatment sequences and washout periods. The analysis method is named as mixed-model ANOVA, but model terms and assumptions are not fully specified. Details on adjustment for period or sequence, assessment of carryover, and the treatment of incomplete treatment sequences are not reported in the registry record. A sample-size justification and interim-analysis plan are also not reported in the registry record. These gaps matter when judging the robustness and reproducibility of the statistical conclusions, especially given the several outcomes and comparisons recorded.

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 0% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event summary reports zero serious adverse events in each treatment group: placebo, FF 100 μg, levocabastine 200 μg, and FF 100 μg/levocabastine 200 μg. The corresponding serious-event denominators are 34, 54, 53, and 68. These are group-level figures in the registry record; the crossover design means they should not be added to infer a count of unique participants.

Other adverse events are reported for 2 participants in the placebo group, 1 in the FF 100 μg group, 1 in the levocabastine 200 μg group, and 3 in the combination group. The record does not provide serious adverse-event terms, and no top serious terms are listed. Death counts are not reported in the registry record; a blank or absent death figure should not be read as a reported count of zero.

These figures are descriptive and do not establish comparative safety. The registry record does not provide event-level details, severity or relatedness for the other events, or a statistical comparison of adverse events between treatments. The group denominators and follow-up context should be considered when reading the counts, and the summary does not characterize the full safety profile.

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

The posted primary outcome estimates show different weighted mean TNSS values across the treatment groups during an allergen chamber challenge on Day 8. The combination group’s reported least-squares mean was 1.933, compared with 4.189 for FF 100 μg, 4.5 for levocabastine 200 μg, and 6.03 for placebo. Since TNSS assigns higher scores to more severe symptoms, the combination’s lower posted mean indicates a lower measured symptom score in that specified assessment window. The pairwise results also report estimates and uncertainty for these comparisons.

These results do not establish how participants would respond in ordinary environmental conditions, how long any difference would persist beyond the stated assessment, or whether the observed score differences are clinically important. The trial used an allergen challenge chamber and a defined symptom scale. The record does not state a threshold for clinical importance, and the statistical results alone cannot supply one.

The findings also do not demonstrate equivalence between FF and levocabastine. Their reported comparison has a confidence interval that includes zero, but the registry does not report an equivalence margin or an equivalence design. Likewise, the combination-versus-component p-values do not on their own show that all possible outcomes favor the combination or that effects extend to other patient groups.

Interpretation is further bounded by what the record does not report: the detailed model specification, missing-data procedures, multiplicity strategy, and sample-size justification. The endpoint analysis population is defined by treatment exposure and availability of pharmacodynamic assessment data. Consequently, the posted estimates should be understood as results from the specified analysis population and challenge-based endpoint, not as a complete account of every potential benefit or risk.

14. Why This Trial Matters Statistically

Statistically, the study is notable for using an incomplete-block crossover design to compare a fixed-dose combination with its component treatments and placebo under a controlled allergen challenge. The design allows treatment-period responses to be observed within assigned sequences, while the reported analysis uses mixed-model ANOVA. This makes the model specification and handling of period, sequence, and incomplete treatment exposure important to interpretation.

The primary endpoint is a time-weighted continuous symptom measure rather than a single symptom assessment. Repeated measurements during the 0-4-hour challenge window contribute to the weighted mean, which is intended to reflect the response across that window. The registry describes the trapezoidal-rule area-under-the-curve calculation and reports least-squares means with confidence intervals, allowing readers to consider both estimated scores and their precision.

The record also illustrates why statistical significance and clinical importance are different questions. Several pairwise comparisons have small reported p-values, but the registry does not define a clinically meaningful TNSS difference. The FF-versus-levocabastine comparison has a confidence interval that includes zero, but that does not establish equivalence. The multiple outcomes and comparisons also make the absence of a reported multiplicity strategy relevant.

Finally, group-specific analysis counts and the defined pharmacodynamic population show why a crossover trial’s treatment summaries should not be treated as independent parallel-arm samples. Assessing the results requires attention to who contributed endpoint data, what the model included, and what is not described. The record gives useful outcome estimates but leaves several elements needed for a complete statistical audit unreported.

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.

This record gives limited detail for independently assessing the analysis. It names mixed-model ANOVA but does not report the model formula, covariance structure, treatment of period and sequence, or assessment of carryover. Although the study included washout periods of 14-28 days, the record does not establish whether any carryover was present or how it was evaluated.

The primary population definition depends on receiving at least one dose and contributing pharmacodynamic data. The record says that only participants contributing data at the indicated time points were included, but does not explain how missing measurements were handled or whether sensitivity analyses were conducted. Treatment-specific analysis counts vary, and the record does not fully explain the reasons for those differences.

There are multiple outcomes and pairwise analyses, but a multiplicity procedure or testing hierarchy is not reported. A sample-size justification, interim-analysis plan, and clinically meaningful TNSS threshold are also not reported in the registry record. These omissions constrain interpretation of the p-values, precision, and clinical relevance.

The study’s chamber-based endpoint measures symptoms during a controlled allergen exposure, so the results are specific to that environment and the recorded assessment window. The record does not provide enough information to extend the estimates to other settings or populations. The safety summary is also limited: serious-event counts are reported, but deaths are not reported in the registry record, and the adverse-event information does not give detailed event descriptions or comparative statistical analyses.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Proof of Concept Study to Assess Effect of Fluticasone Furoate (FF)/Levocabastine Fixed Dose Combination (FDC) Compared With Levocabastine and FF Alone in Subjects With Allergic Rhinitis (AR) — NCT01957202.
• Registry record last updated 2017-01-09; status last verified 2016-11.
• Results first posted 2014-10-08.
• 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

NCT01957202 is a completed Phase 2 randomized, double-blind, placebo-controlled, incomplete-block crossover study sponsored by GlaxoSmithKline. It enrolled 71 participants with allergic rhinitis and compared a fixed-dose combination of FF and levocabastine with the component treatments and placebo across treatment sequences.

The primary outcome was the weighted mean TNSS over 0-4 hours of allergen challenge on Day 8 of each treatment period. The registry reports least-squares means and 95% confidence intervals, plus mixed-model ANOVA estimates for pairwise comparisons. The combination group had the lowest posted mean score, and several reported pairwise comparisons had confidence intervals that did not include zero. The FF-versus-levocabastine comparison had a confidence interval that included zero.

The study’s chamber-based endpoint, crossover design, varying treatment-specific analysis counts, and unreported details about missing data, multiplicity, and model specification are important context. Serious adverse-event counts were zero in the listed groups; deaths are not reported in the registry record.