A Study to Assess the Effects of PT003 and Placebo MDI on Specific Image Based Airway Volumes and Resistance in Subjects With Moderate to Severe COPD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Assess the Effects of PT003 and Placebo MDI on Specific Image Based Airway Volumes and Resistance in Subjects With Moderate to Severe COPD (NCT02643082), a randomized phase 3 study. The primary outcome is Specific Airway Volume (siVaw).

2. Clinical Question

This question concerns a quantitative airway imaging measure, rather than a patient-reported outcome or a clinical event. The endpoint is expressed in mL/L and is defined as specific airway volume (siVaw). The reported treatment ratio summarizes the modeled comparison; it should not be read as an absolute change in volume or as evidence about outcomes not measured by this endpoint. The study also posted secondary outcomes for airway resistance, airway volume, change from baseline in FEV1 at Day 15, and change from baseline in FRC at Day 15. The registry record does not report those outcomes’ values in the primary-result information used for this interpretation.

3. Trial Design

This was a Phase 3, randomized, triple-masked, single-center, two-treatment, two-period crossover study. The official title describes chronic dosing for 2 weeks. Participants received GFF MDI (PT003) and placebo MDI within the crossover study, so the design was intended to compare treatments under both treatment conditions rather than assign separate parallel groups to only one treatment. The registry record does not report the treatment sequence, period-specific results, washout details, or the method used to account for period and sequence effects.

The registry records enrollment of 20 and a completed study. The recorded start date is 2015-12-23, and the primary completion date is 2017-01-26. The primary outcome is specific airway volume at Day 15, described as the average across lobes adjusted for lobe volume. Its analysis population is the ITT Population. The endpoint summaries list 20 for GFF MDI and 19 for placebo MDI; those arm-level endpoint counts are not necessarily a full account of every participant’s sequence, exposure, or analysis status in a crossover trial.

Randomization and masking can reduce some forms of bias, but their presence in the design description does not establish that every source of bias was eliminated. In crossover studies, observations from the same participant under different treatments are related. Appropriate analysis therefore needs to account for within-participant correlation and may need to address treatment period, sequence, and possible carryover. The registry identifies the sponsor’s primary comparison method as mixed models analysis, but does not report enough detail to establish precisely which model terms or covariance assumptions were used.

The study is listed as completed, and the primary comparison is reported. The registry record does not report a protocol-level sample-size justification, interim analyses, or a full statistical analysis plan. Those absences limit what can be concluded about how design choices and analysis decisions were prespecified.

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: Subjects were randomized into 1 of 2 treatment sequences: subjects in Sequence 1 received GFF MDI in Treatment Period 1 followed by Placebo MDI in Treatment Period 2, and subjects in Sequence 2 received Placebo MDI in Treatment Period 1 followed by GFF MDI in Treatment Period 2. There was a a washout of 5-21 days,in between treatment periods.

Period 1

Reasons for not completing this period:

Washout

Period 2

Analysis population for the primary outcome, as defined in the registry: ITT Population

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: ITT Population - all subjects who were randomized to treatment. Treatment was assigned as randomized regardless of the treatment actually received.

6. Primary Endpoint

The registry lists the primary outcome as Specific Airway Volume (siVaw), measured over Day 15.

Specific airway volume (siVaw) is the primary endpoint. The registry defines it as a specific image-based airway volume, averaged across lobes and adjusted for lobe volume, measured at Day 15. Its stated unit is mL/L, and the reported arm summaries are geometric least-squares means with 95% confidence intervals. The endpoint therefore expresses a modeled imaging measure of airway volume in relation to lobe volume, rather than an unadjusted total volume.

Image-based airway measurements can offer a quantitative way to characterize airway structure. Adjusting for lobe volume makes the endpoint specific to the size of the lobe being assessed, while averaging across lobes provides the reported participant-level summary. Such a measure may be useful when the research question concerns airway characteristics that are not captured by a single conventional lung-function measure. That is a general explanation of the endpoint’s statistical role; the registry does not establish the endpoint’s clinical interpretation or threshold for meaningful change.

The registry identifies the primary endpoint population as ITT and gives Day 15 as the timeframe. It does not report the imaging acquisition or processing protocol, the handling of unavailable lobes, or a clinically important difference for siVaw. The endpoint’s statistical result should consequently be interpreted as a comparison on the stated imaging scale, not as a direct measure of symptoms, daily functioning, or long-term outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary result combines arm-level geometric least-squares means with a model-based treatment comparison. The reported method is mixed models analysis, and the comparison parameter is a least-squares mean ratio between treatments. A mixed model is a general statistical approach for data with related observations, such as repeated measurements from the same participant. In a crossover study, measurements under different treatments from one participant are not independent in the same way as observations from unrelated participants.

What does a geometric least-squares mean represent?

A least-squares mean is a model-based estimate of a group mean, adjusted for the terms included in the model. The adjective geometric indicates that the endpoint was summarized on a multiplicative rather than a simple additive scale, typically by modeling transformed values and expressing the result back on the original scale. This is a general explanation. The registry record names the geometric least-squares mean parameter but does not state the transformation, distributional assumptions, or adjustment terms used for the primary endpoint.

The reported estimates are 1.79 for GFF MDI and 1.02 for placebo MDI. They are accompanied by 95% confidence intervals of 1.48 to 2.16 and 0.85 to 1.24. A confidence interval describes the precision of an estimated quantity under the model and its assumptions. It reflects uncertainty about the modeled mean, not the range of individual participant values. Its interpretation depends on the analysis being appropriate for the design and data.

How should the treatment ratio be read?

The sponsor-reported mixed models analysis gives a least-squares mean ratio of 1.75, with a 95% confidence interval of 1.65 to 1.86. On a ratio scale, 1 is the value corresponding to equal modeled means. A ratio above 1 indicates that the modeled numerator is larger than the denominator according to the comparison’s stated ordering. The registry does not spell out the ordering in words, so the treatment labels and reported ratio should be retained when describing the result. The confidence interval is entirely above 1, and the reported p-value is p < 0.001.

The p-value measures how incompatible the observed result is with a specified null model, assuming the analysis assumptions hold. It is not the probability that the null hypothesis is true, nor does it quantify the chance that the finding is due to chance in ordinary language. It also does not measure clinical importance. A small p-value can accompany an estimate whose practical relevance depends on context and a meaningful-effect threshold, which is not reported for siVaw in the registry.

Why does the crossover structure matter?

In a crossover study, each participant receives more than one treatment across periods. This can improve precision by allowing comparisons within participants, because stable differences between participants need not dominate the treatment contrast. The trade-off is that analysis must respect the repeated-measures structure. A model that treats related observations as independent can give misleading uncertainty estimates. Mixed models are commonly used to account for this dependence; that is a general methodological point, not a claim about unreported model details in this trial.

Period effects occur when measurements differ by study period for reasons other than assigned treatment. Sequence effects may arise if treatment order matters, and carryover refers to an influence of an earlier treatment persisting into a later period. The registry identifies a mixed models analysis but does not report how period, sequence, or carryover were assessed or modeled. It also does not report the covariance structure, fitting method, or model diagnostics. These omissions prevent an independent assessment of the model’s implementation from the registry summary alone.

What does over-dispersion mean in general?

Over-dispersion occurs when observed variability is greater than a statistical model expects. It matters because standard errors and confidence intervals can be too narrow if the model understates variability. This is a general concept and is not a reported issue in this trial. The primary endpoint is summarized as a geometric least-squares mean, and the record does not describe an over-dispersion assessment.

How do planning illustrations differ from observed power?

Planning illustrations use assumptions to describe the sample size that might be needed for a future study to detect a chosen fraction of an observed effect. The stated assumptions for these illustrations are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. They use a fraction of the observed effect as the target effect. Such illustrations are prospective planning aids, not observed power, and they do not show the probability that this completed study detected its effect. They are not a substitute for the trial’s own design justification, which is not reported in the registry record.

For time-to-event outcomes, a general planning approach may count total events; for other outcomes, participant numbers per arm are used. The primary outcome in this trial is an imaging measure, not a time-to-event endpoint. The registry’s sample-size assumptions do not provide a protocol sample-size rationale, and the record does not report the planned effect size or power for the original study.

What can be inferred from the reported precision?

The confidence interval for the treatment ratio provides a range of values compatible with the data under the fitted model and its assumptions. Its finite width conveys that the estimate has uncertainty. Interpreting that uncertainty requires attention to the crossover design, the endpoint population, and the unreported details of the model. The reported interval does not account for uncertainty about whether siVaw is a clinically meaningful endpoint or whether other analyses would yield similar conclusions.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary comparison as a mixed models analysis of the least-squares mean ratio between treatments. The reported ratio is 1.75, with a 95% confidence interval of 1.65 to 1.86 and p < 0.001. The arm-level summaries are geometric least-squares means of 1.79 for GFF MDI and 1.02 for placebo MDI, each with its own 95% confidence interval.

The sponsor’s comparison is the relevant result for the crossover treatment contrast because it is the reported model-based analysis. A ratio inferred by comparing arm summaries alone would not reproduce the model’s handling of repeated observations, adjustment terms, or within-participant correlation. The record does not report enough model detail to recreate that analysis fully. There is no noted discrepancy between the sponsor result and a recalculation, so no competing recalculated estimate takes precedence.

The reported interval and p-value describe statistical evidence under the sponsor’s model. They do not establish clinical importance, and they do not resolve questions about the model’s assumptions or missing measurements. The registry record does not report the full analysis plan, diagnostics, or detailed model specification needed to audit every aspect of the sponsor’s analysis.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures and 6 analyses, and it identifies 4 secondary outcomes: airway resistance (iRaw), airway volume (iVaw), change from baseline in FEV1 at Day 15, and change from baseline in FRC at Day 15. The primary endpoint is specific airway volume at Day 15. The presence of multiple outcomes and analyses raises the general issue of multiplicity: when several hypotheses are tested, the chance of at least one apparently positive result can increase unless the testing strategy accounts for it.

The registry record does not report a multiplicity adjustment, a hierarchy for testing endpoints, or which outcomes were intended to be confirmatory. It therefore does not establish how the reported p-value should be interpreted relative to other outcome analyses. The primary result can be described using its reported p-value, but the record does not support claims about family-wise error control across the full set of analyses.

The primary endpoint population is identified as ITT. The registry does not define the exact analysis-set rules, explain how missing imaging measurements were handled, or report the amount or pattern of missing data. The endpoint summaries list 20 for GFF MDI and 19 for placebo MDI, but the record does not explain the difference. In a crossover study, the number of observations contributing to a model can depend on whether measurements are available across periods and treatment conditions. Without further detail, the reason for the reported counts cannot be established.

The record also does not report whether missingness was assumed to be unrelated to the unobserved outcome, whether sensitivity analyses were conducted, or how protocol deviations were handled. These are general analytic considerations, not claims that a particular problem occurred. The absence of these details limits assessment of robustness.

The registry record does not report interim analyses or the full statistical analysis plan. It is therefore not possible to determine from this record whether the endpoint choices, multiplicity strategy, missing-data methods, and analysis population were prespecified in detail. These omissions should be kept distinct from the reported model estimate: they qualify interpretation but do not alter the figures stated in the registry.

12. 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: acute coronary syndrome (5.3%), acute myocardial infarction (5.0%).

The registry reports serious adverse events for the overall study and separately for the GFF MDI and placebo MDI groups. For the overall study, 2 serious adverse events are listed among 20 at risk. For GFF MDI, 1 is listed among 20 at risk; for placebo MDI, 1 is listed among 19 at risk. Deaths are recorded as 0 in the overall study and in each treatment group.

Other adverse events are listed for 15 among 20 at risk in the overall study, 12 among 20 at risk for GFF MDI, and 8 among 19 at risk for placebo MDI. These figures describe the registry’s reported event counts and at-risk denominators. They should not be interpreted as a causal comparison: the record does not provide event timing, participant-level exposure details, statistical comparisons, or enough context to establish whether events occurred during a particular treatment period.

The registry also lists Acute Coronary Syndrome at 5.3% and Acute Myocardial Infarction at 5.0% among top serious-event terms. The record does not specify a group attribution for those percentages in the information reported. They should not be assigned to a treatment arm or interpreted as group-specific rates. With a crossover design and limited reported safety detail, the figures are descriptive rather than a basis for a comparative safety conclusion.

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

The primary result shows a modeled difference between GFF MDI and placebo MDI on specific image-based airway volume at Day 15. The reported least-squares mean ratio is 1.75, with a 95% confidence interval of 1.65 to 1.86 and p < 0.001. The arm summaries are geometric least-squares means of 1.79 and 1.02, with their respective confidence intervals. These statistics describe the endpoint as defined in the registry and the comparison under the reported mixed model.

The result does not establish that the measured difference is clinically important. The registry does not report a clinically meaningful threshold for siVaw. Nor does the primary endpoint directly establish changes in symptoms, quality of life, exacerbations, or long-term disease outcomes. Those outcomes are not the primary measure.

The result also does not answer whether the treatments differ on all secondary outcomes. The registry lists airway resistance, airway volume, FEV1 change, and FRC change as secondary outcomes, but the primary result cannot substitute for their separate estimates. The record does not report a multiplicity strategy, so the statistical relationship among the several analyses remains unclear.

Because the study used a crossover design, its treatment estimate depends on analysis that appropriately handles repeated observations and period structure. The registry names mixed models analysis but does not provide the full model specification, sequence details, or carryover assessment. The reported estimate is therefore best read as the sponsor’s stated model-based comparison, with those design and reporting limitations in mind.

The enrollment was 20, and the primary endpoint summaries list 20 for GFF MDI and 19 for placebo MDI. Those summary counts do not explain the participant flow or missing data. The registry record also does not provide enough information to assess the precision of the estimate beyond its stated confidence interval or to evaluate whether the model’s assumptions were met. The findings are specific to this completed, single-center study and its Day 15 imaging endpoint; extrapolation beyond that setting and measure is not established by the reported statistics.

14. Why This Trial Matters Statistically

This trial is statistically notable for its crossover structure and imaging-based primary endpoint. In a crossover study, participants contribute observations under different treatments, creating within-participant dependence. The treatment comparison can use that structure, but the analysis must represent it correctly. The registry identifies mixed models analysis, which is a framework suited to related observations in general, while leaving the exact model specification unavailable.

The primary measure is specific airway volume, an image-based endpoint averaged across lobes and adjusted for lobe volume. It illustrates how clinical studies may use quantitative physiological or imaging outcomes alongside conventional measures such as FEV1. Such outcomes can answer a different measurement question, but their interpretation depends on understanding the scale, adjustment, and clinical relevance. The registry reports the scale and construction of siVaw but does not provide a clinically meaningful threshold.

The treatment ratio and its confidence interval are more informative than a p-value alone because they convey both direction on the ratio scale and precision. The p-value of p < 0.001 indicates strong statistical evidence under the reported model, while the interval of 1.65 to 1.86 shows the model-based range of estimates compatible with the data. Neither statistic establishes practical importance or resolves model assumptions.

The record also highlights common reporting questions: the role of secondary outcomes, multiplicity, missing endpoint data, analysis-set definitions, and crossover-specific effects. The registry lists several outcomes but does not report how the analyses were coordinated or how missing measurements were handled. For readers assessing a trial, recognizing those gaps is part of distinguishing a precisely reported estimate from a fully transparent analysis.

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 gives the primary endpoint definition, arm summaries, and sponsor-reported treatment comparison, but it does not provide a full statistical analysis plan. The model terms, covariance structure, fitting method, diagnostics, and handling of period or sequence effects are not reported. The crossover design makes those details particularly relevant because measurements from the same participant are related and treatment order may matter.

The record does not report the treatment sequence, washout details, or an assessment of carryover. These omissions do not show that a design or analysis problem occurred; they mean the registry summary cannot establish how those issues were addressed. It also does not explain why the endpoint summaries list 20 for GFF MDI and 19 for placebo MDI, or describe the availability of measurements across periods.

Missing-data methods and sensitivity analyses are not reported. The record identifies the ITT Population for the primary endpoint but does not define its operational rules or state how incomplete imaging measurements were handled. No protocol-level sample-size justification or interim-analysis information is reported. The planning assumptions described in the registry are illustrations based on the observed effect, not evidence of the study’s original design rationale.

The registry lists secondary outcomes and analyses but does not report a multiplicity procedure or testing hierarchy. It also does not provide a clinically meaningful threshold for the imaging endpoint, imaging-process details, or participant-level results. As a result, the reported effect can be described statistically, but its clinical importance, robustness to analytic choices, and relationship to other outcomes cannot be fully assessed from the registry record alone.

The study is identified as single-center and includes enrollment of 20. The record does not establish how broadly its result applies beyond the enrolled population, study setting, treatment schedule, and Day 15 endpoint. These limits concern the scope of interpretation rather than a contradiction of the reported primary result.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Assess the Effects of PT003 and Placebo MDI on Specific Image Based Airway Volumes and Resistance in Subjects With Moderate to Severe COPD — NCT02643082.
• Registry record last updated 2019-03-19; status last verified 2019-02.
• Results first posted 2019-03-19.
• 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

This completed Phase 3 study was randomized, triple-masked, single-center, and used a two-treatment, two-period crossover design with chronic dosing for 2 weeks. It compared GFF MDI (PT003) with placebo MDI in subjects with moderate to severe COPD. The registry records enrollment of 20, a start date of 2015-12-23, and a primary completion date of 2017-01-26.

The primary endpoint was specific airway volume at Day 15, defined as an image-based measure averaged across lobes and adjusted for lobe volume, in the ITT Population. The sponsor-reported mixed models analysis gives a least-squares mean ratio between treatments of 1.75, with a 95% confidence interval of 1.65 to 1.86 and p < 0.001. The registry also lists secondary outcomes and adverse-event summaries.

The record does not report the full model specification, missing-data methods, multiplicity approach, or a clinical-importance threshold for the primary endpoint. The result is therefore best understood as a model-based comparison on the stated imaging measure, with uncertainty and reporting limitations.