This page separates reported registry information from statistical interpretation. The ClinicalTrials.gov record reports that results have been posted, but the ClinicalTrials.gov record does not contain formal statistical-analysis entries with treatment-effect estimates, confidence intervals, or p-values.
1. Trial at a Glance
KRONOS was a phase 3, randomized, quadruple-masked, parallel-group, multicenter study designed to assess the efficacy and safety of PT010, PT003, and PT009 compared with Symbicort Turbuhaler in COPD. The study used chronic dosing for 24 weeks.
| Feature | KRONOS |
|---|---|
| Trial name | KRONOS |
| ClinicalTrials.gov identifier | NCT02497001 |
| Phase | Phase 3 |
| Condition | COPD |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Enrollment | 1,902 |
| Results posted | Yes |
| Outcome measures posted | 13 |
| Lead sponsor | Pearl Therapeutics, Inc. |
| Sponsor type | Industry |
2. Clinical Question
The central statistical question was whether the investigational inhaled therapies could be evaluated for efficacy and safety against the active comparator, Symbicort Turbuhaler, in patients with COPD over the registered 24-week study period.
Population
The trial's condition was COPD. The ClinicalTrials.gov record does not provide additional population eligibility characteristics.
Interventions
BGF MDI 320/14.4/9.6 µg, GFF MDI 14.4/9.6 µg, and BFF MDI 320/9.6 µg.
Comparator
Symbicort Turbuhaler inhalation powder.
Primary question
How do the randomized treatment groups compare on the registered FEV1 efficacy endpoints at Week 24, while also considering safety?
3. Trial Design
The design is structurally important because it creates a randomized comparison among four treatment strategies while using quadruple masking to reduce the potential influence of treatment knowledge on participant behavior, investigator conduct, and outcome assessment.
BGF MDI
- BGF MDI 320/14.4/9.6 µg
- Drug intervention
- Serious adverse events: 55/639 affected/at risk
GFF MDI
- GFF MDI 14.4/9.6 µg
- PT003 drug intervention
- Serious adverse events: 68/625 affected/at risk
BFF MDI
- BFF MDI 320/9.6 µg
- PT009 drug intervention
- Serious adverse events: 21/314 affected/at risk
Symbicort Turbuhaler
- Symbicort Turbuhaler inhalation powder
- 400/12 µg in the ClinicalTrials.gov record
- Serious adverse events: 29/318 affected/at risk
Study timeline
Study start
The KRONOS study began on August 10, 2015.
Registered efficacy assessment
Both registered primary endpoints were assessed at Week 24.
Primary completion
The registry lists January 5, 2018 as the primary completion date.
4. Randomization and Masking
KRONOS used randomized allocation and a parallel-group design. Randomization is the feature that allows treatment groups to be compared without deliberately assigning patients according to prognostic characteristics.
The registry describes the study as quadruple-masked. Masking is a design feature rather than a statistical test: it is intended to reduce the opportunity for knowledge of treatment assignment to influence behavior, treatment administration, assessment, or reporting.
Why randomization matters
Randomization creates the basis for a causal treatment comparison because assignment is determined by the trial design rather than by observed patient characteristics.
Why masking matters
Masking can reduce differential expectations or assessment behavior that might otherwise introduce bias into a comparative trial.
5. Endpoints
| Registered primary endpoint | Time frame | Registry definition |
|---|---|---|
| FEV1 AUC0-4 | at Week 24 | FEV1 AUC0-4 (L) for the Efficacy Estimand: change from baseline in forced expiratory volume in 1 second (FEV1) area under the curve from 0 to 4 hours (AUC0-4). AUC was normalized for length of follow-up, typically 4 hours. |
| Change From Baseline in Morning Pre-dose Trough FEV1 | at Week 24 | Morning Pre-Dose Trough FEV1 (L) for the Efficacy Estimand. |
The registry lists 2 primary endpoints, with both assessed at Week 24. The ClinicalTrials.gov record classifies their endpoint type as “Other / unclear,” so this page retains the registry terminology rather than assigning a more specific formal endpoint category.
Understanding FEV1
FEV1 is a pulmonary function measurement representing forced expiratory volume in 1 second. In this trial, the two primary endpoints use FEV1 in different ways: one summarizes the change from baseline across the first 4 hours through an area-under-the-curve measure, while the other evaluates the change from baseline in morning pre-dose trough FEV1.
Why the two primary endpoints are different
The two endpoints describe different aspects of lung-function response. The FEV1 AUC0-4 endpoint summarizes the trajectory over a defined post-dose interval, whereas morning pre-dose trough FEV1 focuses on a measurement obtained before the next dose. These quantities should therefore not be treated as interchangeable measurements.
6. Statistical Methodology
The ClinicalTrials.gov record does not contain entries in the posted results. Consequently, there are no registry-reported treatment-effect estimates, confidence intervals, or p-values that can be reproduced here.
For continuous pulmonary-function endpoints such as the two registered FEV1 measures, a randomized parallel-group analysis would ordinarily compare treatment groups using a model appropriate for a continuous outcome. Depending on the prespecified statistical analysis plan, this might involve an analysis of covariance or a mixed-effects model incorporating baseline FEV1 and relevant repeated measurements. The exact model, covariance structure, multiplicity strategy, and estimand definition cannot be established from the ClinicalTrials.gov record.
In a simple ANCOVA framework, the treatment coefficient estimates the adjusted difference between randomized groups after accounting for baseline measurement. This equation is educational rather than a reconstruction of the KRONOS statistical analysis plan.
FEV1 AUC0-4
An area-under-the-curve endpoint summarizes information across multiple post-dose measurements rather than relying on a single time point. The registry specifies that the AUC is normalized for the length of follow-up, typically 4 hours. The primary endpoint is explicitly defined as change from baseline in FEV1 AUC0-4.
Morning pre-dose trough FEV1
The trough endpoint focuses on the morning pre-dose FEV1 measurement. Because it is defined as change from baseline, the statistical comparison would ordinarily account for baseline FEV1 when estimating the treatment contrast, although the registry extract does not specify the actual model.
Four-arm comparison
With four randomized arms, statistical interpretation becomes more complicated than a single two-group comparison. If multiple pairwise hypotheses were tested, the analysis plan would normally need to specify how the familywise type I error or another multiplicity criterion would be controlled. The ClinicalTrials.gov record does not state which pairwise comparisons were confirmatory or what multiplicity procedure was used.
7. Planned Analysis
ClinicalTrials.gov reports that results have been posted and that both registered primary endpoints have results posted for them. However, the ClinicalTrials.gov record contains no formal statistical-analysis records. The appropriate interpretation is therefore to distinguish what was measured from what can be statistically concluded from the registry-reported extract.
| Endpoint | What was measured | Typical statistical approach | Formal comparison in the ClinicalTrials.gov record |
|---|---|---|---|
| FEV1 AUC0-4 | Change from baseline in FEV1 AUC0-4 at Week 24 | Continuous-outcome comparison, potentially ANCOVA or another prespecified longitudinal model | Not in the posted results |
| Morning pre-dose trough FEV1 | Change from baseline in morning pre-dose trough FEV1 at Week 24 | Continuous-outcome comparison, typically adjusted for baseline and potentially other prespecified covariates | Not in the posted results |
The registry's posting of outcome measures should not be confused with the presence of a formal comparative statistical result in the registry record. A result value for an endpoint and a documented inferential analysis are different pieces of information.
What a complete primary analysis would ordinarily report
- The analysis population and treatment assignment rule.
- The model used for the continuous FEV1 endpoint.
- The estimated treatment contrast for each prespecified comparison.
- A confidence interval around each treatment contrast.
- A p-value or other prespecified inferential measure.
- The handling of missing Week 24 measurements.
- The multiplicity procedure for the two primary endpoints and any multiple treatment comparisons.
- The prespecified estimand and the relationship between treatment discontinuation, rescue treatment, and the analysis.
8. Safety Results
The ClinicalTrials.gov record provides serious adverse events by treatment arm as affected patients over the number at risk. These counts can be reported directly without converting them into percentages.
| Treatment arm | Serious adverse events affected / at risk |
|---|---|
| BGF MDI 320/14.4/9.6 µg | 55/639 |
| GFF MDI 14.4/9.6 µg | 68/625 |
| BFF MDI 320/9.6 µg | 21/314 |
| Symbicort Turbuhaler 400/12 µg | 29/318 |
The affected-patient counts alone should not be interpreted as comparative safety risks because the arms have different numbers at risk. A formal comparison would generally use the appropriate safety population and prespecified definitions, and would distinguish the number of affected patients from the number of adverse-event episodes.
9. Statistical Methods Explained
Why was randomization important?
Randomization determines treatment assignment independently of the clinical characteristics that investigators might otherwise use to select therapy. In a parallel-group trial, this provides the foundation for comparing outcomes across treatment groups. The validity of the resulting comparison depends on the integrity of randomization, follow-up, outcome measurement, and the prespecified analysis.
Why was quadruple masking useful?
Masking is intended to reduce the influence of treatment knowledge on trial conduct and outcome assessment. The ClinicalTrials.gov record identifies the KRONOS study as quadruple-masked but does not specify the four parties included in that designation. Therefore, the appropriate conclusion is that masking was a formal design feature, not that a particular set of four individuals or groups was necessarily masked.
What does FEV1 AUC0-4 measure?
FEV1 AUC0-4 summarizes FEV1 information over the first 4 hours after the relevant treatment period. An area-under-the-curve endpoint can capture the overall response across a time interval rather than depending on one measurement. KRONOS defines the endpoint as change from baseline in FEV1 AUC0-4, with the AUC normalized for length of follow-up.
Why is baseline adjustment useful for FEV1?
When the outcome is a change in a continuous measurement, baseline values contain information about the patient's starting point. An ANCOVA-style analysis can use baseline FEV1 as a covariate, producing an adjusted comparison that accounts for baseline differences. This is a general statistical explanation; the registry-reported KRONOS data do not establish that ANCOVA was the actual prespecified model.
Why does a four-arm trial raise multiplicity issues?
With four treatment groups, there can be several scientifically plausible pairwise comparisons. Testing many hypotheses without an appropriate error-control strategy can increase the probability of obtaining at least one apparently positive result by chance. A rigorous analysis therefore specifies which comparisons are confirmatory and how the overall type I error is controlled.
Why should the posted results not be confused with formal statistical analyses?
A registry can contain outcome-measure results without the ClinicalTrials.gov record exposing every component of the inferential analysis. A reported endpoint value answers “what was observed,” while an analysis record should additionally establish the comparison, model, uncertainty, and hypothesis-testing framework. In the registry record, no formal statistical analyses were posted, so those inferential details cannot be reproduced here.
10. Understanding the Two Primary Endpoints Together
FEV1 AUC0-4
This endpoint summarizes change from baseline in FEV1 across the 0-to-4-hour interval. It provides an integrated measure of pulmonary-function response over time.
Morning trough FEV1
This endpoint evaluates change from baseline in morning pre-dose trough FEV1. It represents a different time-point perspective from the AUC endpoint.
Because the endpoints summarize different aspects of FEV1 response, agreement between them can provide complementary information, while disagreement would require careful consideration of timing and the estimand represented by each endpoint.
Conceptually, the area under the FEV1-time curve represents the accumulated measurement across the specified interval. The KRONOS registry definition additionally specifies normalization for length of follow-up.
11. Analysis Populations and Missing Data
the ClinicalTrials.gov record identifies an “Efficacy Estimand” in the definitions of both primary endpoints, but it does not provide the complete estimand framework or the exact analysis population definition.
Likewise, the ClinicalTrials.gov record does not specify how missing FEV1 measurements were handled. For a 24-week randomized trial, missing outcome measurements can arise through discontinuation, missed visits, or other forms of incomplete follow-up. The statistical consequences depend on the amount and pattern of missingness and on the assumptions and methods specified in the statistical analysis plan.
What can be stated
The registered endpoints were defined for an efficacy estimand and assessed at Week 24.
What the registry record does not include
The ClinicalTrials.gov record does not state the imputation method, missing-data assumptions, sensitivity analyses, or exact efficacy analysis population.
12. Multiplicity and the Four-Arm Structure
KRONOS had four treatment arms and two registered primary endpoints. This combination creates a potentially complex hypothesis-testing structure. However, the registry extract does not specify which treatment contrasts were primary, whether the two endpoints were tested sequentially or jointly, or what multiplicity adjustment was applied.
| Feature | What the ClinicalTrials.gov record establishes | Statistical implication |
|---|---|---|
| Number of arms | 4 | More than one treatment comparison may be possible. |
| Primary endpoints | 2 | Multiple primary hypotheses may require prespecified error control. |
| Results posted | Yes | Endpoint results exist in the registry record. |
| Statistical analyses reported | None | Formal p-values, confidence intervals, and comparison definitions cannot be reproduced from this extract. |
A statistically rigorous interpretation therefore requires separating the design structure from the unreported inferential details. The presence of four arms and two primary endpoints tells us that multiplicity is an important design topic, but it does not tell us which multiplicity procedure was actually used.
13. Interim Analysis, Crossover, and Bayesian Methods
The ClinicalTrials.gov record does not report an interim-analysis strategy, crossover procedure, Bayesian analysis, or non-inferiority margin.
Interim analysis
No interim-analysis method is provided in the ClinicalTrials.gov record.
Crossover
No crossover procedure is reported in the ClinicalTrials.gov record.
Bayesian methods
No Bayesian method is identified in the ClinicalTrials.gov record.
Non-inferiority margin
No non-inferiority margin is provided; the ClinicalTrials.gov record does not classify the primary question as a non-inferiority analysis.
These omissions matter because they prevent the statistical framework from being reconstructed beyond the information explicitly provided. In particular, absence of a reported method should not be treated as evidence that a method was not used; it means only that the method is not available in the ClinicalTrials.gov record.
14. Limitations
- No formal statistical-analysis entries: no formal statistical analyses were posted, so treatment-effect estimates, confidence intervals, and p-values cannot be reproduced.
- Limited endpoint context: the registry definitions identify the two primary FEV1 measures, but the ClinicalTrials.gov record does not provide the complete statistical analysis plan.
- Multiplicity cannot be reconstructed: the trial has four arms and two primary endpoints, but the ClinicalTrials.gov record does not specify the confirmatory comparison hierarchy or adjustment method.
- Missing-data handling is unspecified: no imputation method, missingness assumption, or sensitivity-analysis strategy is provided.
- Analysis populations are incompletely described: the endpoints refer to an efficacy estimand, but the ClinicalTrials.gov record does not define the complete analysis population.
- Safety interpretation: serious adverse-event counts are available as affected/at-risk values, but the registry-reported extract does not provide a full safety-analysis framework.
- No baseline table: baseline demographic and clinical characteristics are not included in the ClinicalTrials.gov record.
- No subgroup results: the ClinicalTrials.gov record does not provide subgroup analyses or interaction tests.
- No long-term outcome analysis: the registered primary endpoint time frame is Week 24, and the ClinicalTrials.gov record does not provide additional efficacy follow-up results.
15. Why This Trial Matters Statistically
KRONOS is a useful teaching example because it combines several fundamental clinical-trial concepts in one design: randomized allocation, quadruple masking, four parallel treatment groups, continuous pulmonary-function endpoints, change-from-baseline analysis, an area-under-the-curve endpoint, a trough endpoint, and the potential for multiplicity across treatment comparisons and primary outcomes.
| Concept | How it appears in KRONOS |
|---|---|
| Randomization | The trial uses randomized allocation across four treatment arms. |
| Blinding | The study is described as quadruple-masked. |
| Parallel-group design | The four treatment groups are evaluated in a parallel design. |
| Continuous outcome | Both primary endpoints are based on FEV1 measurements in liters. |
| Change from baseline | Both primary endpoint definitions specify change from baseline. |
| AUC analysis | FEV1 AUC0-4 summarizes FEV1 over a 0-to-4-hour interval. |
| Trough measurement | Morning pre-dose trough FEV1 provides a different time-point measure. |
| Multiple arms | Four treatment arms create multiple possible comparative hypotheses. |
| Multiple primary endpoints | Two primary endpoints are registered at Week 24. |
| Safety comparison | Serious adverse events are reported by randomized treatment arm. |
16. Interpreting Safety Counts Correctly
The serious-adverse-event data illustrate an important distinction between counts and rates. The registry extract gives both the number of affected patients and the number at risk for each arm:
Reported serious adverse-event counts
BGF MDI · GFF MDI · BFF MDI · Symbicort Turbuhaler
The denominator is essential. A raw count cannot be interpreted as a risk without considering the population at risk. Even after obtaining rates, a complete safety comparison would still require attention to exposure time, event definitions, recurrent events, treatment discontinuation, and the prespecified safety population.
The registry-reported serious-adverse-event figures establish the number of affected patients and the corresponding at-risk population in each arm. They do not by themselves establish whether the treatment groups differ statistically in serious-adverse-event risk.
17. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The study was randomized, parallel-group, quadruple-masked, and included two registered primary continuous FEV1 endpoints assessed at Week 24. The registry-reported extract does not provide the formal inferential analyses needed to quantify treatment differences.
Clinical interpretation
The trial was designed to evaluate efficacy and safety of three investigational inhaled drug interventions against Symbicort Turbuhaler in COPD. Clinical meaning depends on the magnitude and uncertainty of the endpoint differences, which cannot be reconstructed from the registry-reported statistical-analysis field.
This distinction is important. A clinical trial can have a clearly defined randomized design and registered endpoints even when the statistical-analysis details available in a particular data extract are incomplete.
18. A Framework for Reading the KRONOS Results
| Question | Why it matters |
|---|---|
| What is the treatment contrast? | The direction and magnitude of the difference define the estimated effect. |
| What is the confidence interval? | It describes statistical uncertainty around the estimated treatment contrast. |
| What is the p-value? | It evaluates evidence against a specified null hypothesis; it does not measure clinical importance or effect size. |
| Which analysis population was used? | The population definition determines which participants contribute to the estimate. |
| How were missing measurements handled? | Different missing-data assumptions can produce different estimates. |
| How were multiple comparisons handled? | Four treatment arms and two primary endpoints can create multiple hypothesis-testing opportunities. |
| What does the absolute difference mean clinically? | A statistically detectable difference may or may not represent a clinically meaningful change. |
For KRONOS, these questions are particularly important because the registered outcomes are continuous FEV1 measures rather than time-to-event outcomes. The appropriate effect measure is therefore generally a difference in the continuous endpoint or another prespecified contrast, rather than a hazard ratio.
19. What Is Known — and What Is Not — From the ClinicalTrials.gov record
| Information | Status in the ClinicalTrials.gov record |
|---|---|
| Trial completed | Yes |
| Phase | Phase 3 |
| Enrollment | 1,902 |
| Number of arms | 4 |
| Randomized design | Yes |
| Parallel design | Yes |
| Quadruple masking | Yes |
| Primary endpoints | 2 |
| Primary endpoint time frame | Week 24 for both |
| Results posted | Yes |
| Formal statistical-analysis entries reported | No |
| Primary treatment-effect estimates reported | No |
| Primary confidence intervals reported | No |
| Primary p-values reported | No |
| Serious adverse events by arm | Yes |
| Subgroup estimates | No |
| Interim-analysis details | No |
| Bayesian methods | No |
| Crossover details | No |
| Non-inferiority margin | No |
This distinction is the central methodological point of the page: The registry record establishes the design and registered outcomes, while the ClinicalTrials.gov record does not establish the numerical inferential results.
20. Related Tutorials
Learn more about the methods used in this trial:
21. Related Statistical Calculators
22. Sources
- ClinicalTrials.gov: NCT02497001 — KRONOS.
- PubMed record: PMID 41237914.
- PubMed record: PMID 40340809.
- PubMed record: PMID 39691156.
- PubMed record: PMID 39103901.
- PubMed record: PMID 35815359.
Continue through the Clinical Biostats statistical pathway
Use the trial design and endpoint concepts here as a starting point for deeper study of randomized trials, masking, and statistical analysis.
23. Record Summary
KRONOS was a completed phase 3 randomized, quadruple-masked, parallel-group trial enrolling 1,902 participants across four treatment arms. Its two registered primary endpoints were change from baseline in FEV1 AUC0-4 and change from baseline in morning pre-dose trough FEV1, both assessed at Week 24. The ClinicalTrials.gov record also reports serious adverse events by treatment arm and confirms that results have been posted.
From a statistical perspective, the trial provides a useful example of how continuous pulmonary-function endpoints can be embedded in a multi-arm randomized design. The AUC and trough endpoints represent different summaries of FEV1 response, while the four-arm structure creates a potential multiplicity problem that must be addressed by the prespecified testing strategy.
The principal limitation of the registry-reported statistical record is that no formal statistical analyses were posted. The appropriate statistical interpretation is to describe the registered estimands and explain the analyses that would ordinarily be required, while keeping the distinction between documented trial information and general statistical methodology explicit.