This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the ClinicalTrials.gov record.
1. Trial at a Glance
FREEDOM-C2 was a completed phase 3 randomized, parallel-group, double-masked treatment trial evaluating oral UT-15C SR versus placebo in pulmonary hypertension. The registry reports 310 enrolled participants, 2 treatment arms, 16 posted outcome measures, and 12 posted statistical analyses.
| Feature | FREEDOM-C2 |
|---|---|
| Trial name | FREEDOM-C2 |
| Brief title | Efficacy and Safety of Oral UT-15C Tablets to Treat Pulmonary Arterial Hypertension |
| Phase | Phase 3 |
| Status | Completed |
| Therapeutic area | Pulmonology |
| Condition | Pulmonary Hypertension |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 310 |
| Interventions | UT-15C SR (drug); Placebo (drug) |
| Results posted | Yes |
| Outcome measures posted | 16 |
| Statistical analyses posted | 12 |
| Lead sponsor | United Therapeutics |
2. Clinical Question
The central question was whether treatment with oral UT-15C SR produced a difference in the placebo-corrected change in 6-minute walk distance from Baseline to Week 16 in patients enrolled in this pulmonary hypertension trial.
Population
The registered condition is pulmonary hypertension, and the trial's brief title describes treatment of pulmonary arterial hypertension.
Intervention
UT-15C SR, administered as the study drug in the intervention arm.
Comparator
Placebo, administered as the comparator study drug.
Primary question
Does UT-15C SR differ from placebo in the placebo-corrected change in 6MWD from Baseline to Week 16?
3. Trial Design
The registry describes FREEDOM-C2 as randomized, parallel-group, double-masked, and treatment-focused. The ClinicalTrials.gov record reports 310 enrolled participants across 2 arms.
UT-15C SR
- Study intervention: UT-15C SR
- Drug intervention
- Compared with placebo
- Included in the randomized parallel-group comparison
Placebo
- Comparator intervention: placebo
- Drug intervention
- Compared with UT-15C SR
- Included in the randomized parallel-group comparison
4. Endpoints
The registry identifies one primary endpoint and reports results for that endpoint. The primary endpoint is the 6-minute Walk Distance (6MWD), assessed from Baseline to Week 16.
| Endpoint | Registry definition | Time frame |
|---|---|---|
| 6-minute Walk Distance (6MWD) | Placebo-corrected change in 6MWD from Baseline to Week 16, correlates with the current clinical standard for assessing patient functional status in the treatment of PAH and is considered an objective measure of patient functional status by the American Thoracic Society (ATS). The 6MWD was to be assessed between 3 and 6 hours after the morning dose of study drug and background therapy(ies). | Baseline and 16 weeks |
The statistical-analysis record classifies the primary endpoint as continuous, with outcome unit of meters. The primary analysis population is described as an intention-to-treat analysis.
5. Primary Result: 6-Minute Walk Distance
The posted primary analysis compared UT-15C SR with placebo using a non-parametric ANCOVA in the intention-to-treat population. The reported effect measure was a Hodges-Lehmann (H-L) estimate.
Placebo-corrected change in 6MWD
95% CI: −2.0 to 22.0 meters · P = 0.089
Analysis: non-parametric ANCOVA · Population: intention to treat · Hypothesis: superiority
| Primary endpoint | UT-15C SR vs placebo |
|---|---|
| Outcome | 6-minute Walk Distance (6MWD) |
| Time frame | Baseline and 16 weeks |
| Outcome unit | meters |
| Analysis population | Intention to treat analysis |
| Method | Non-parametric ANCOVA |
| Effect measure | Hodges-Lehmann (H-L) |
| Estimate | 10.0 |
| 95% CI | −2.0 to 22.0 |
| P-value | 0.089 |
| Hypothesis type | Superiority |
The reported H-L estimate of 10.0 meters describes the estimated treatment difference from the registry's non-parametric analysis, with UT-15C SR compared with placebo. It is a comparative estimate rather than a statement that every participant experienced a 10.0-meter change.
The 95% confidence interval of −2.0 to 22.0 meters expresses uncertainty around the estimated treatment difference. Because the interval extends below zero as well as above zero, the registry result is compatible with a small difference in either direction as well as a larger positive difference within the interval.
The P-value of 0.089 addresses the statistical evidence against the null hypothesis under the specified testing framework. It does not measure the size or clinical importance of the estimated effect. A P-value is therefore not a substitute for the estimate and its confidence interval.
Because the analysis is described as non-parametric ANCOVA, the result should be interpreted according to that analysis rather than treated as an ordinary unadjusted mean difference. The ClinicalTrials.gov record does not provide the covariates or full model specification used in the non-parametric ANCOVA.
6. Secondary Endpoint Results
The registry contains four secondary statistical analyses with estimates, confidence intervals, and P-values. Three use the Wilcoxon rank sum test, while Clinical Worsening Assessment uses Fisher Exact.
| Secondary endpoint | Method | Estimate | 95% CI | P-value |
|---|---|---|---|---|
| Borg Dyspnea Score | Wilcoxon rank sum test | 0.0 | −1.0 to 0.0 | 0.22 |
| World Health Organization (WHO) Functional Class | Wilcoxon rank sum test | 0.0 | 0.0 to 0.0 | 0.43 |
| Dyspnea Fatigue Index | Wilcoxon rank sum test | 0.0 | 0.0 to 1.0 | 0.30 |
| Clinical Worsening Assessment | Fisher Exact | Not reported | Not reported | 1.00 |
Borg Dyspnea Score
All subjects with a Baseline Borg Score were included in the analysis. One subject in the placebo group did not have a Baseline Borg Score recorded. The reported Hodges-Lehmann estimate was 0.0, with a two-sided 95% CI of −1.0 to 0.0 and P = 0.22.
WHO Functional Class
The analysis included subjects with a WHO functional class assessment at Week 16. The reported Hodges-Lehmann estimate was 0.0, with a two-sided 95% CI of 0.0 to 0.0 and P = 0.43.
Dyspnea Fatigue Index
Subjects without a Dyspnea Fatigue Index score at Baseline were not included in the analysis. The reported Hodges-Lehmann estimate was 0.0, with a two-sided 95% CI of 0.0 to 1.0 and P = 0.30.
Clinical Worsening Assessment
The Clinical Worsening Assessment was analyzed in the intention-to-treat population using Fisher Exact. The registry reports a P-value of 1.00, but no effect estimate or confidence interval is reported in the ClinicalTrials.gov record.
These secondary results should be read as separate statistical comparisons rather than as interchangeable measurements of the same outcome. The Borg Dyspnea Score and Dyspnea Fatigue Index are continuous or scale-based outcomes analyzed with a rank-based method, while WHO Functional Class is ordinal in nature despite being classified as binary in the ClinicalTrials.gov record. Clinical Worsening Assessment was analyzed categorically using Fisher Exact.
The P-values of 0.22, 0.43, 0.30, and 1.00 do not quantify effect size. The corresponding estimates and confidence intervals provide the more informative description of the observed statistical comparison where those quantities were reported.
7. Additional Post Hoc Analyses
The registry also reports seven post hoc analyses examining 6MWD across selected PAH etiology, background-therapy, and time-since-diagnosis groups. These analyses are distinct from the primary endpoint analysis and should not be treated as if they were the prespecified primary comparison.
| Post hoc analysis | Analysis population | Estimate | 95% CI | P-value |
|---|---|---|---|---|
| 6MWD by PAH Etiology: IPAH / HPAH | Subjects with IPAH/HPAH | 14.0 | 0.0 to 28.0 | 0.058 |
| 6MWD by Background PAH Therapy: PDE-5i Only | Subjects receiving only a PDE-5i at Baseline | 15.0 | −1.0 to 29.0 | 0.054 |
| 6MWD by Background PAH Therapy: ERA + PDE-5i | Subjects receiving both an ERA and a PDE-5i at Baseline | 4.0 | −16.0 to 24.0 | 0.674 |
| 6MWD by Time to PAH Diagnosis: 0 - 0.9 Years | Subjects diagnosed with PAH between 0 to 0.9 years prior to Baseline | 28.0 | 1.0 to 59.0 | 0.059 |
| 6MWD by Time Since PAH Diagnosis: 0.9 - 1.74 Years | Subjects diagnosed with PAH between 0.9 and 1.74 years prior to Baseline | 10.0 | −10.0 to 31.0 | 0.22 |
| 6MWD by Years Since PAH Diagnosis: 1.8 - 3.5 Years | Subjects diagnosed with PAH between 1.8 and 3.5 years prior to Baseline | 3.0 | −23.0 to 28.0 | 0.99 |
| 6MWD by Time Since PAH Diagnosis: 3.6 - 26.4 Years | Subjects diagnosed with PAH for 3.6 to 26.4 years prior to Baseline | −2.0 | −25.0 to 22.0 | 0.84 |
All seven post hoc analyses used ANCOVA and reported a Hodges-Lehmann estimate with a two-sided 95% confidence interval. The registry classifies these analyses as superiority comparisons.
8. Statistical Methodology
Non-parametric ANCOVA for the primary endpoint
The primary 6MWD analysis was reported as non-parametric ANCOVA. ANCOVA combines an outcome comparison with adjustment for one or more covariates. In this trial, the registry identifies the method but does not provide the full covariate specification in the ClinicalTrials.gov record.
The "non-parametric" qualifier is important. A conventional ANCOVA is a linear-model approach that relies on assumptions concerning the outcome distribution and model residuals. A non-parametric ANCOVA is intended to provide a rank-based or distributionally less restrictive approach while retaining the general objective of comparing groups after accounting for relevant information.
The exact model used in FREEDOM-C2 cannot be reconstructed from the registry summary because the complete model specification is not provided.
Hodges-Lehmann estimation
The primary and several secondary/post hoc analyses report a Hodges-Lehmann (H-L) estimate. This is a non-parametric measure of location shift associated with rank-based comparisons. It complements the P-value by providing an estimate of the magnitude of the treatment difference.
For the primary 6MWD analysis, the H-L estimate was 10.0 meters. The corresponding 95% confidence interval was −2.0 to 22.0 meters.
Wilcoxon rank sum test
The Borg Dyspnea Score, WHO Functional Class, and Dyspnea Fatigue Index analyses used the Wilcoxon rank sum test. This is a non-parametric two-group comparison based on the ordering or ranks of observations rather than requiring a normal-distribution assumption for the raw outcome.
Fisher Exact test
Clinical Worsening Assessment was analyzed with Fisher Exact. Fisher's exact test is designed for categorical comparisons, particularly when exact inference is preferred or when expected cell counts may make a large-sample approximation less appropriate.
Intention-to-treat analysis
The primary 6MWD analysis and the Clinical Worsening Assessment analysis are explicitly identified as intention-to-treat analyses. The ITT principle preserves the randomized treatment assignment when comparing groups, rather than redefining treatment groups according to subsequent exposure or adherence.
For the other secondary analyses, the registry supplies endpoint-specific analysis populations. For example, the Borg analysis included subjects with a Baseline Borg Score, while the Dyspnea Fatigue Index analysis excluded subjects without a Baseline score.
9. Confidence Intervals and P-values
The primary result illustrates why an effect estimate, confidence interval, and P-value should be read together.
Estimate
The H-L estimate of 10.0 meters describes the estimated treatment difference reported by the non-parametric analysis.
Confidence interval
The 95% CI of −2.0 to 22.0 meters describes uncertainty around the estimated treatment difference under the analysis framework.
P-value
The P-value of 0.089 quantifies evidence against the null hypothesis under the specified testing framework; it does not measure the size of the treatment effect.
Clinical scale
Because the endpoint is measured in meters, the estimate and confidence interval retain the original outcome unit and are therefore directly interpretable on the 6MWD scale.
A confidence interval is not a prediction interval for individual patients. The interval −2.0 to 22.0 meters describes statistical uncertainty around the estimated comparative effect; it does not mean that individual patients' responses must fall inside that range.
Likewise, a P-value is not the probability that the treatment has no effect, and it is not the probability that the observed result occurred "by chance." It is a probability calculated under a specified null model and testing procedure.
The most complete statistical description of the primary finding is therefore the combination of the 10.0-meter H-L estimate, its 95% CI of −2.0 to 22.0 meters, the P-value of 0.089, the intention-to-treat population, and the non-parametric ANCOVA method.
10. Analysis Populations and Missing Baseline Information
The ClinicalTrials.gov record shows that analysis populations were defined separately for different endpoints. This matters because the population actually analyzed can affect both precision and interpretation.
| Endpoint | Analysis population described by the registry |
|---|---|
| 6MWD | Intention to treat analysis |
| Borg Dyspnea Score | All subjects with a Baseline Borg Score; one placebo-group subject did not have a Baseline Borg Score recorded |
| WHO Functional Class | Subjects with a WHO functional class assessment at Week 16 |
| Dyspnea Fatigue Index | Subjects without a Baseline Dyspnea Fatigue Index score were not included |
| Clinical Worsening Assessment | Intention to treat analysis |
These definitions show why a single label such as "the trial population" is insufficient for interpreting every posted analysis. Different outcomes can have different evaluable populations.
11. Multiplicity and the Role of Post Hoc Analyses
The registry reports 1 primary-endpoint analysis, 4 secondary analyses, and 7 post hoc analyses, for a total of 12 statistical analyses posted. The ClinicalTrials.gov record classifies the hypothesis type as superiority for each listed analysis.
| Analysis category | Number reported | Interpretive role |
|---|---|---|
| Primary endpoint analysis | 1 | Primary treatment comparison |
| Secondary endpoint analyses | 4 | Additional prespecified outcome comparisons as reported |
| Post hoc analyses | 7 | Additional exploratory comparisons identified as post hoc |
| Total statistical analyses posted | 12 | All statistical analyses represented in the ClinicalTrials.gov record |
This distinction is particularly important for the seven post hoc 6MWD analyses. Examining multiple subgroups creates more opportunities for apparently notable estimates or P-values to arise through ordinary sampling variability. The appropriate interpretation is descriptive unless a formal prespecified multiplicity and interaction framework is documented.
12. Safety Results
The ClinicalTrials.gov record reports serious adverse events by treatment arm using affected participants over participants at risk.
| Safety measure | UT-15C SR | Placebo |
|---|---|---|
| Serious adverse events, affected / at risk | 23 / 157 | 23 / 153 |
Serious adverse events
UT-15C SR vs placebo, respectively
The reported counts show the number of participants affected by serious adverse events alongside the number at risk in each arm. The affected counts are 23 in both groups, while the denominators are 157 and 153.
These safety data should not be converted into an efficacy conclusion. Safety and efficacy answer different questions and are analyzed on different outcome scales. The ClinicalTrials.gov record does not provide a formal statistical comparison of serious adverse events, so no inferential P-value or effect estimate is assigned here.
13. Trial Timeline
Trial start
The ClinicalTrials.gov record identifies June 2009 as the study start.
Primary completion
The ClinicalTrials.gov record identifies July 2011 as the primary completion date.
Registry status
The trial is listed as completed, with results posted on ClinicalTrials.gov.
14. Statistical Methods Explained
Why was a non-parametric ANCOVA used for 6MWD?
The registry specifically reports a non-parametric ANCOVA for the primary 6MWD endpoint. ANCOVA is useful when a treatment comparison is intended to account for covariate information, while the non-parametric formulation reduces reliance on conventional distributional assumptions. The ClinicalTrials.gov record does not specify the complete covariate set, so the exact model cannot be reconstructed here.
What does a Hodges-Lehmann estimate of 10.0 mean?
The H-L estimate is the reported non-parametric estimate of the treatment difference. Here it is 10.0 on the meter scale used for 6MWD. It is an estimate for the randomized-group comparison, not a statement that each participant improved by 10.0 meters.
Why does the confidence interval include negative values?
The 95% CI extends from −2.0 to 22.0 meters. This means the statistical uncertainty around the estimate includes values slightly below zero as well as positive values. The interval therefore provides a broader picture than the point estimate alone.
Why doesn't P = 0.089 measure treatment effect size?
A P-value describes the strength of evidence against a null hypothesis under a specified statistical model and testing procedure. It does not describe how large the treatment effect is. Effect size is addressed by the H-L estimate, while precision is addressed by the confidence interval.
Why use the Wilcoxon rank sum test for several secondary endpoints?
The Wilcoxon rank sum test compares the ordering of observations between two groups and does not require the same distributional assumptions as a standard two-sample mean comparison. FREEDOM-C2 used this method for the Borg Dyspnea Score, WHO Functional Class, and Dyspnea Fatigue Index analyses.
Why use Fisher Exact for Clinical Worsening Assessment?
Fisher Exact is an exact method for categorical data. It evaluates the treatment-group association with the categorical outcome without relying on the same large-sample approximation used by a conventional chi-square test. The registry reports Fisher Exact for Clinical Worsening Assessment and a P-value of 1.00.
Why should post hoc analyses be interpreted cautiously?
The seven additional 6MWD analyses were explicitly classified as post hoc. They examine selected subsets defined by PAH etiology, background therapy, or time since diagnosis. Because they were not identified as the primary confirmatory analysis, their estimates are better viewed as additional exploratory evidence rather than as replacements for the primary endpoint analysis.
15. What the Primary Result Does — and Does Not — Mean
The primary analysis reports an H-L estimate of 10.0 meters for the UT-15C SR versus placebo comparison on the 6MWD endpoint.
It does not mean that every patient experienced a 10.0-meter improvement, nor does it establish an identical treatment effect for every participant or every clinical outcome.
The 95% CI of −2.0 to 22.0 meters quantifies uncertainty around the reported treatment-effect estimate under the statistical framework used for the analysis. It does not describe individual patient responses.
The P-value of 0.089 describes the evidence against the null hypothesis under the specified superiority test. It does not measure clinical importance, probability of benefit, or probability that the null hypothesis is true.
16. Limitations
- Limited model detail: the registry identifies the primary method as non-parametric ANCOVA but the ClinicalTrials.gov record does not provide the full model specification or covariate set.
- Secondary analysis populations differ: several secondary outcomes required Baseline or Week 16 assessments, so their analysis populations are not necessarily identical to the primary ITT population.
- Missing-data strategy: the ClinicalTrials.gov record does not describe a formal imputation strategy.
- Post hoc analyses: seven additional 6MWD analyses were identified as post hoc and therefore should not be interpreted as equivalent to the primary endpoint.
- Multiplicity information: the ClinicalTrials.gov record does not specify a multiplicity-adjustment or alpha-allocation strategy for the secondary and post hoc analyses.
- No formal safety comparison reported: serious adverse events are reported as affected/at-risk counts by arm, but the ClinicalTrials.gov record does not report an inferential comparison.
- Endpoint classification: the registry identifies the primary endpoint type as "Other / unclear" while the statistical analysis classifies 6MWD as continuous. The analysis here follows the endpoint definition and statistical-analysis record reported.
- Registry-level evidence: this page is restricted to the ClinicalTrials.gov record and does not add numerical results from publications or other external sources.
17. Why This Trial Matters Statistically
FREEDOM-C2 is a useful teaching example because its posted results illustrate a different statistical pathway from a conventional time-to-event trial. The primary endpoint is a continuous functional measure, the primary comparison uses non-parametric ANCOVA, and several secondary outcomes use rank-based or exact categorical methods.
| Concept | How it appears in FREEDOM-C2 |
|---|---|
| Randomization | The trial is registered as randomized. |
| Parallel design | The design model is parallel. |
| Blinding | The masking is double. |
| Intention-to-treat analysis | The primary 6MWD analysis uses an ITT population. |
| ANCOVA | The primary analysis is reported as non-parametric ANCOVA; post hoc 6MWD analyses use ANCOVA. |
| Hodges-Lehmann estimation | The primary endpoint and several other analyses report H-L estimates. |
| Wilcoxon rank sum test | Used for Borg Dyspnea Score, WHO Functional Class, and Dyspnea Fatigue Index. |
| Fisher Exact | Used for Clinical Worsening Assessment. |
| Confidence intervals | 95% two-sided confidence intervals are reported for the primary and several secondary/post hoc estimates. |
| Multiplicity | 12 statistical analyses are posted, including 1 primary, 4 secondary, and 7 post hoc analyses. |
| Post hoc analysis | Seven analyses examine selected PAH etiology, background-therapy, and diagnosis-duration groups. |
| Safety analysis | Serious adverse events are reported by arm as affected participants over participants at risk. |
18. A Practical Statistical Reading of FREEDOM-C2
A disciplined reading of the trial begins with the primary endpoint rather than with the collection of secondary and post hoc P-values. The primary analysis compares UT-15C SR with placebo using a non-parametric ANCOVA in the ITT population. Its estimated treatment difference is 10.0 meters, but the 95% confidence interval ranges from −2.0 to 22.0 meters and the two-sided P-value is 0.089.
The next step is to examine whether the secondary outcomes tell a consistent statistical story. The registry reports P-values of 0.22 for Borg Dyspnea Score, 0.43 for WHO Functional Class, 0.30 for Dyspnea Fatigue Index, and 1.00 for Clinical Worsening Assessment. These results should be considered separately from the primary endpoint and with awareness that the registry does not provide a multiplicity strategy in the ClinicalTrials.gov record.
Finally, the seven post hoc 6MWD analyses should be treated as exploratory. Their estimates range from −2.0 to 28.0, and their confidence intervals vary substantially. That variation is not, by itself, evidence that treatment effects truly differ across the selected subgroups.
19. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The primary analysis estimated a 10.0-meter treatment difference in 6MWD, with a two-sided 95% CI of −2.0 to 22.0 meters and P = 0.089. The analysis was a non-parametric ANCOVA in the ITT population.
Clinical interpretation
6MWD is described in the registry as an objective measure of patient functional status and a clinical standard for assessing functional status in PAH. The ClinicalTrials.gov record does not establish a separate clinical-importance threshold for the 10.0-meter estimate.
20. Related Tutorials
Learn more about the statistical methods used in this trial:
21. Related Calculators
22. Sources
- ClinicalTrials.gov: FREEDOM-C2, NCT00887978.
- PubMed record: PMID 41627369.
- PubMed record: PMID 28090293.
- PubMed record: PMID 23669822.
Continue with the underlying statistical methods
Use the related tutorials and calculators to explore ANCOVA, non-parametric comparisons, confidence intervals, P-values, randomization, and intention-to-treat analysis.
23. Record Summary
FREEDOM-C2 provides a useful example of statistical analysis for a randomized phase 3 trial with a continuous functional endpoint. The primary 6MWD analysis used a non-parametric ANCOVA in the intention-to-treat population and reported a Hodges-Lehmann estimate of 10.0 meters, a two-sided 95% confidence interval of −2.0 to 22.0 meters, and a P-value of 0.089.
The broader registry record includes rank-based secondary analyses, an exact categorical analysis, and seven post hoc 6MWD comparisons. Interpreting the trial therefore requires more than reading individual P-values: the analysis population, statistical method, effect estimate, confidence interval, endpoint role, and post hoc status all affect how the evidence should be understood.