Bardoxolone Methyl in Patients With Connective Tissue Disease-associated Pulmonary Arterial Hypertension - CATALYST: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Bardoxolone Methyl in Patients With Connective Tissue Disease-associated Pulmonary Arterial Hypertension - CATALYST (NCT02657356), a randomized phase 3 study. The primary outcome is Change From Baseline in Six-minute-walk Distance (6MWD) Relative to Placebo at Week 24.

2. Clinical Question

The clinical question is framed as a comparison between treatment strategies, rather than as a test of whether either group changed from its own baseline. The relevant estimate is the difference in estimated change between the two assigned groups. The primary endpoint is described as change from baseline through 24 weeks after the first dose. The registry record also lists safety and a recommended dose range among the study aims, but the posted primary statistical result concerns six-minute-walk distance.

The trial enrolled 202 participants and was terminated. The registry record does not report why it was terminated. The posted results therefore answer a narrower question than the full stated study aims: they describe the recorded comparison for the primary measure and the available secondary and safety summaries. They do not by themselves explain the decision to terminate the study or establish how the recorded findings should guide care.

3. Trial Design

CATALYST is a Phase 3, randomized, double-blind, parallel-group study comparing bardoxolone methyl capsules with placebo capsules in patients with connective tissue disease-associated pulmonary arterial hypertension. The registry lists Biogen as the sponsor. Enrollment was 202 participants, with 102 in the placebo group and 100 in the bardoxolone methyl group in the primary analysis. The primary population is the ITT population: all randomized patients categorized by assigned treatment regardless of treatment exposure.

Both groups were assigned once-daily oral capsules for 24 weeks. The bardoxolone methyl regimen started at 5 mg, with escalation to 10 mg at Week 4 unless clinically contraindicated. The placebo group also received capsules once a day for 24 weeks. The registry records a study start date of 2016-10-04 and a primary completion date of 2020-05-07; the study status is terminated.

Randomization and double masking are design features intended to support a fair comparison. Random assignment helps balance measured and unmeasured characteristics on average, while masking is intended to reduce influences on participant experience, treatment delivery, and outcome assessment. These design features do not remove uncertainty from the estimated treatment contrast, and the record does not describe the specific procedures used to implement masking or allocation.

The record identifies the primary outcome as change from baseline in six-minute-walk distance relative to placebo at Week 24. It also lists time to first persistent clinical improvement event as a secondary outcome. The registry record does not report the original sample-size justification, planned interim analyses, or the full prespecified statistical plan. The reason for termination is also not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT population (all randomized patients categorized by their assigned treatment group regardless of treatment exposure)

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: Intent to treat (ITT) population (all randomized patients categorized by their assigned treatment group regardless of treatment exposure)

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Six-minute-walk Distance (6MWD) Relative to Placebo at Week 24, measured over Baseline through 24 weeks after participant receives the first dose.

The primary endpoint is change from baseline in six-minute-walk distance, measured in meters, through 24 weeks after the participant receives the first dose. The six-minute-walk test records the distance a participant can walk during a fixed assessment period. As a continuous measure, it can retain more information about individual differences than an outcome recorded only as an event or category.

For a randomized comparison, the central quantity is the difference in estimated change between the bardoxolone methyl and placebo groups. A between-group contrast helps distinguish the treatment-group pattern from changes that may also occur over time in the comparison group. The registry reports least-squares means for the groups and a net least-squares-mean difference. The posted primary analysis uses the ITT population, classified by assigned group regardless of treatment exposure.

The endpoint is used to assess change in walking capacity over the study period. It is one measure of function, not a complete account of a participant’s health or experience. The posted result should therefore be interpreted as an estimate for this particular measure and time frame. It does not, on its own, describe every clinical outcome or establish how the observed difference relates to outcomes not reported in the registry.

7. Results Posted in the Registry

• Bardoxolone Methyl Capsules: mean -3.61 versus 9.25 on control; difference -12.86 (95% CI -26.59 to 0.87), standardized effect d = -0.26; z-test p = 0.066.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Time to First Persistent Clinical Improvement Event

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Bardoxolone Methyl Capsules: mean 6.62 versus 17.90 on control; difference -11.28 (95% CI -14.06 to -8.50), standardized effect d = -1.11; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Bardoxolone Methyl Capsules than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists one secondary outcome, time to first persistent clinical improvement event. Its posted material contains group summaries in weeks and an unadjusted mean contrast, as well as a reported hazard-ratio analysis. These are different statistical summaries of the secondary outcome and should be interpreted on their own scales. The registry’s reported analysis gives a hazard ratio of 1.984, a 95% confidence interval from 1.360 to 2.895, and p < 0.001. The record does not fully describe the hazard-ratio reference direction or event coding, so the ratio alone cannot establish which group had the higher event rate in the relevant sense.

The secondary result is statistically distinct from the primary walking-distance comparison. The primary result is a continuous difference in estimated change, with a confidence interval that includes zero; the secondary result is reported using a ratio for an event outcome, with a confidence interval that excludes 1. The two results therefore do not contradict one another: they address different outcomes, use different scales, and may capture different aspects of participants’ experience. Nor should the secondary finding be used to rewrite the primary result.

Unadjusted secondary results generally carry less confirmatory weight than a prespecified primary result, especially when a study evaluates multiple outcomes. Repeated testing can increase the chance of observing an apparently strong result by chance. The registry does not report a multiplicity adjustment or a hierarchy for interpreting secondary outcomes. It is therefore appropriate to describe the secondary analysis as a reported statistical signal while keeping its exploratory or confirmatory status uncertain.

The secondary endpoint’s title, mean summaries, and hazard-ratio analysis do not provide a complete account of how the event was defined or analysed. The registry record does not report the event criteria, censoring rules, or full analysis plan. Those omissions limit the strength of any broader conclusion based on this outcome.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Time to First Persistent Clinical Improvement Event

9. Statistical Interpretation

10. Statistical Methods Explained

For a continuous endpoint, a group mean summarizes the average outcome in that group, while a between-group mean difference describes the contrast between the two group summaries. In this trial, the primary endpoint is a change from baseline, and the reported group values are least-squares means. Least-squares means are model-based estimates rather than simple raw averages; they are calculated within the specified model. The registry identifies the reported method as mixed models analysis, but does not give the full model specification, covariates, or handling of missing observations.

What does the mean difference represent?

The primary reported contrast is -12.86 meters for bardoxolone methyl relative to placebo. The comparison is expressed as the intervention group’s estimated change minus the comparator group’s estimated change. A negative value therefore places the intervention estimate lower than the placebo estimate on the change scale. This is a contrast between groups, not a claim that every participant changed by that amount and not a comparison of the groups’ baseline measurements.

The recalculated mean difference uses the posted group summaries and is labeled unadjusted. It is a simple calculation from those summary statistics, rather than a reproduction of the mixed model. The recalculation’s standard deviation is approximate, as noted in the analysis details. It can serve as a check on the scale and direction of the estimate, but it does not replace the registry’s model-based analysis. Small differences between recalculated and reported confidence limits or p-values can arise because the analyses use different methods and assumptions.

How should the standard error be understood?

A standard error describes the sampling uncertainty of an estimate under the method used: estimates with larger standard errors are less precise, all else being equal. It is not the spread of individual participant outcomes. The registry reports standard errors alongside the primary group estimates. The posted group-level standard errors are useful for understanding the uncertainty of those estimated means, but the registry does not provide enough model detail to reconstruct the mixed-model analysis completely.

The recalculated comparison uses a pooled standard deviation estimated from the posted summaries. A standard deviation describes variation among observations, whereas a standard error describes uncertainty in an estimated mean. The recalculation treats the comparison as an unadjusted contrast and relies on an approximate standard deviation. This is why it should be distinguished from the trial’s reported mixed-model result, even though the resulting estimates are close.

What do the confidence interval and p-value say?

A confidence interval gives a range of values compatible with the observed data and the analysis assumptions at the stated confidence level. It conveys both the estimated direction and the precision of the contrast. For the registry’s primary analysis, the 95% confidence interval extends from -26.69 to 0.98 meters. Because this range includes zero, the data do not distinguish the estimated contrast from zero at the corresponding conventional two-sided significance threshold. The interval also includes effects on either side of zero, so the result is not evidence that the groups are equivalent.

A p-value measures how surprising data at least as inconsistent with a specified null hypothesis would be under that hypothesis and the analysis assumptions. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The registry reports p = 0.068 for the primary mixed-model comparison. The unadjusted recalculation gives p = 0.066. They are close but represent different analyses; the reported mixed-model result is the trial’s own analysis.

Why can over-dispersion matter?

Over-dispersion is a general statistical issue in which the observed variability is greater than a model’s assumed variability. When a model understates variation, standard errors can be too small and confidence intervals can be too narrow, which can make evidence appear more precise than it is. Appropriate modelling of variation is therefore important, particularly for outcomes whose variability does not follow the model’s simplest assumptions. The registry does not report whether over-dispersion was assessed or relevant to the primary continuous endpoint, so no trial-specific conclusion on that point is warranted.

How does the secondary analysis differ?

The registry lists time to first persistent clinical improvement event as a secondary outcome. A time-to-event analysis typically considers both whether an event occurred and the time until it occurred, while accounting for participants whose event time is not observed within the study period. The registry reports a hazard ratio and identifies the analysis method as Chi-squared. It does not give enough detail about event definitions, censoring, model specification, or the reference group to fully reconstruct the analysis. A hazard ratio is not a difference in mean weeks, and it should not be read as though it were the primary mean difference in meters.

The secondary summary also gives group means in weeks and an unadjusted mean difference. These summaries are on a different scale from the reported hazard ratio. The two forms of summary should not be collapsed into a single estimate: one compares the posted means, while the reported hazard ratio comes from the event analysis. Their interpretation depends on their respective methods and definitions, which are not fully detailed in the registry record.

The registry does not report a complete statistical analysis plan, original sample-size justification, or interim analysis plan. Consequently, the posted methods support a focused explanation of the reported estimates, but not a full audit of whether every modelling decision followed a prespecified protocol.

11. Confidence Intervals and Precision

The primary registry estimate is -12.86 meters, with a 95% confidence interval from -26.69 to 0.98. This interval expresses uncertainty around the estimated between-group contrast. Its lower limit is a sizeable negative contrast, while its upper limit is close to zero and slightly positive. The range therefore includes both a difference favoring the placebo estimate and a small difference in the opposite direction, using the reported contrast orientation.

Precision concerns the width and location of the interval, not simply whether a p-value crosses a threshold. The interval is broad enough to leave substantial uncertainty about the size of the primary contrast. A point estimate alone would hide that uncertainty; the interval makes clear that the data do not identify one narrowly defined effect. The fact that zero lies in the interval means the reported analysis does not establish a nonzero contrast at the corresponding conventional threshold. It does not show that the groups are identical.

The unadjusted recalculation has a confidence interval from -26.59 to 0.869. It is close to the registry interval but is based on a simpler calculation from group summaries, including an approximate standard deviation. The registry’s mixed-model confidence interval should be used when describing the trial’s reported primary analysis. Neither interval should be interpreted without its underlying assumptions, and the registry does not provide the complete model detail needed to examine those assumptions fully.

The secondary reported hazard ratio has a 95% confidence interval from 1.360 to 2.895. That interval excludes 1, the null value for a ratio measure, and the reported p-value is < 0.001. This is a distinct result on a different outcome and scale, with event coding and reference direction not fully described in the record. It does not make the primary confidence interval more precise or alter what the primary estimate says.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size display is a planning illustration based on observed effects, not the trial’s original design calculation. For a continuous-outcome calculation, entries are participants per arm. A time-to-event planning calculation instead reports the total number of events; the unit matters because an event total is not a participant count per group. The primary six-minute-walk-distance endpoint is continuous, so its planning figures are read as participants per arm.

The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Each row uses a stated fraction of the observed effect as the effect to detect. The assumptions describe the illustration, not necessarily the design used when the trial was planned. The original sample-size justification and its assumptions are not reported in the registry record.

The 80% power column means the illustrated number of participants per arm for a design that would have an 80% probability of detecting the specified effect under the planning assumptions, if that effect and those assumptions were correct. The 90% column describes the corresponding planning target at 90% probability. These are design probabilities for hypothetical repeated studies under assumptions; this is not observed power.

The rows using smaller fractions of the observed effect require more participants per arm in the planning illustration. That is a general feature of study design: smaller differences are harder to distinguish from random variation, so more information is usually needed. These figures should not be treated as proof that the completed trial was adequately or inadequately sized, because they are based on the observed effect and do not reconstruct the original design rationale.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -12.860; pooled standard deviation σ = 49.780.
2. n = 2 × (1.96 + 0.842)² × 49.780² / -12.860² = 235.2, rounded up to 236 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -12.86                      # observed difference in means
sigma <- 49.7796991067644                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2657356)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 236 (80% power), 315 (90% power)
 75% of observed effect: n per arm = 419 (80% power), 560 (90% power)
 50% of observed effect: n per arm = 941 (80% power), 1,260 (90% power)
power.t.test: 236.2 
Simulated power at n = 236 per arm: 0.795 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry’s reported primary analysis is a mixed-model analysis of the least-squares-mean difference. It gives -12.86 meters, a 95% confidence interval from -26.69 to 0.98, and p = 0.068. This is the study’s reported analysis and should be the main reference when describing the trial’s primary result.

The unadjusted recalculation from the posted group summaries also gives a mean difference of -12.86 meters, with a confidence interval from -26.59 to 0.869 and p = 0.066. It is a simple summary-statistic calculation, uses an approximate standard deviation, and is not the same as the reported mixed model. Its role is to provide a comparison of scale and direction, not to replace the registry analysis.

The two estimates are close and point in the same direction, while their confidence limits and p-values are not identical. Differences are expected when an unadjusted calculation is compared with a model-based analysis that may use additional structure. The registry does not provide the model details needed to attribute the small differences to a particular modelling choice. The discrepancy indicator is false, so no discrepancy between the recalculation and reported analysis is flagged in the record.

The registry does not give a separate sponsor narrative explaining the statistical result. The interpretation should therefore rest on the reported mixed-model estimate, its uncertainty, and the limits of the available methodological detail rather than on an inferred rationale.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is change from baseline in six-minute-walk distance at Week 24. The registry also lists time to first persistent clinical improvement event as a secondary outcome. When more than one outcome or analysis is tested, the chance of at least one small p-value can increase. Statistical plans sometimes address this issue by defining a testing hierarchy or adjusting significance thresholds. The registry record does not report whether either approach was planned or used.

Without a documented multiplicity strategy, the primary result and secondary result should not be treated as if they had equal confirmatory status. The primary analysis is the central comparison identified in the record. The secondary analysis can add context, but its p-value should be read with awareness that the record does not explain how it fits into a broader testing plan. No multiplicity-adjusted secondary result is reported.

The record identifies the primary population as all randomized patients categorized by assigned group regardless of treatment exposure. This is an intention-to-treat approach to group assignment, which preserves the randomized comparison in the analysis population. It does not, by itself, explain how missing outcome measurements were handled, whether participants with no Week 24 measurement contributed to the model, or what assumptions the model made about missingness. Those details are not reported in the registry record.

The registry names mixed models analysis for the primary endpoint but does not give the model specification, covariates, covariance structure, or missing-data procedures. It also does not report the original sample-size justification, planned interim analyses, or a complete statistical analysis plan. The termination reason is not reported in the registry record. These omissions make it difficult to assess whether the analyses and decisions were prespecified, and they limit independent reproduction of the reported results.

These reporting gaps are distinct from the numerical uncertainty in the confidence interval. Even a precisely reported estimate depends on knowing the population, endpoint definition, analysis method, and handling of incomplete data. The available record supports a description of the posted results but not a full evaluation of every analytic choice.

16. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pulmonary arterial hypertension (4.0%), pneumonia (2.9%), breast cancer (2.0%), cardiac failure acute (1.0%), cardiac failure congestive (1.0%).

The registry lists serious adverse events for 18 of 102 participants in the placebo group and 16 of 100 participants in the bardoxolone methyl group. It lists other adverse events for 79 of 102 participants in the placebo group and 85 of 100 participants in the bardoxolone methyl group. The deaths recorded are 1 in the placebo group and 0 in the bardoxolone methyl group. These are descriptive group counts; they do not by themselves establish that a treatment caused or prevented an event.

The record’s listed serious adverse event terms include pulmonary arterial hypertension at 4.0%, pneumonia at 2.9%, breast cancer at 2.0%, acute cardiac failure at 1.0%, congestive cardiac failure at 1.0%, and right ventricular failure at 1.0%. These terms and percentages describe reported events in the registry. They are not a complete clinical assessment of severity, timing, relatedness, or comparative risk.

Group counts can help orient a reader to the posted safety information, but interpretation also depends on how events were collected, classified, and attributed, and on exposure and follow-up. The registry record does not provide enough detail to make causal comparisons. The counts should therefore be read as descriptive safety reporting, not as a standalone estimate of treatment-related harm.

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

The primary result means that the registry’s mixed-model analysis estimated a lower change in six-minute-walk distance for bardoxolone methyl relative to placebo, with a net difference of -12.86 meters. The 95% confidence interval, from -26.69 to 0.98, includes zero and leaves considerable uncertainty about the size and direction of the true contrast under the model. The primary p-value is 0.068. Together, these statistics describe an estimate that does not meet the conventional 0.05 threshold and is not precise enough to exclude a range of plausible effects.

This does not mean the groups were proven identical. A nonsignificant result is not an equivalence finding, and an equivalence conclusion requires an appropriate design and prespecified margin. The registry does not report such a margin. The result also does not establish that the intervention caused a particular individual’s outcome; group-level estimates describe averages and uncertainty across the analysed population.

The secondary hazard-ratio result is statistically stronger by its reported p-value, but it concerns a different outcome and a different analysis. The event definition and reference direction are not fully available in the record, and no multiplicity approach is reported. It should be considered alongside, not substituted for, the primary result. The record’s safety counts are likewise descriptive and do not establish causality.

The trial’s terminated status is relevant context, but the registry does not report the reason for termination. It would be inappropriate to infer that termination was caused by the primary result, safety, enrollment, or any other factor. The record also lacks the full statistical plan, original sample-size justification, interim analysis details, and missing-data methods. These limitations constrain what can be concluded from the posted estimates.

Overall, the posted data describe a primary estimate with substantial uncertainty and a distinct secondary statistical signal. They do not, by themselves, settle the broader clinical question or support conclusions about outcomes not reported in the registry.

18. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation should begin with the estimand and outcome scale. The primary question concerns a between-group difference in change in six-minute-walk distance, measured in meters. The reported estimate is a model-based contrast, while the recalculation is an unadjusted summary-statistic comparison. Both have a similar direction and point estimate, but they are not interchangeable analyses. Keeping that distinction clear prevents a convenient calculation from being mistaken for the trial’s own analysis.

The confidence interval also matters as much as the threshold-based p-value. The primary interval includes zero and spans a range of possible contrasts, demonstrating uncertainty that a binary label such as statistically significant or not significant would conceal. The result is not evidence of equivalence, and its lack of precision should be considered when discussing what the study can establish.

The secondary result highlights a separate issue: different outcomes can use different measures and methods. A mean difference in weeks and a hazard ratio are not the same quantity, and a secondary p-value does not replace the primary estimate. When the registry does not report a multiplicity strategy, the strength of secondary evidence is harder to assess.

Finally, the trial’s termination and missing methodological details limit interpretation beyond the posted statistics. The original sample-size rationale, interim analyses, complete model specification, and reason for termination are not reported in the registry record. Statistical analysis therefore requires both careful reading of the numerical results and restraint about what the available record cannot establish.

19. 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 primary result is based on the registry’s reported mixed-model analysis, but the record does not provide the full model specification. The covariates, covariance structure, assumptions, and missing-data procedures are not reported in the registry record. This limits independent reproduction and makes it difficult to assess how robust the estimate is to alternative modelling choices.

The registry identifies the ITT population as all randomized patients categorized by assigned treatment regardless of treatment exposure. That definition does not reveal how many participants had an observed Week 24 measure, how missing measurements were treated, or whether additional analysis populations were examined. Those details are not reported in the registry record.

The study status is terminated, but the reason for termination is not reported in the registry record. The original sample-size justification and planned interim analyses are also not reported. The sample-size illustrations based on observed effects are not substitutes for the trial’s design rationale. They cannot establish what assumptions guided the original enrollment target.

The secondary outcome is described as time to first persistent clinical improvement event, yet the registry also presents group means in weeks and a reported hazard ratio. The event definition, coding, reference group, and censoring details are not fully reported. This makes it difficult to reconcile the summaries or independently interpret the hazard ratio’s direction.

The registry does not report a multiplicity plan, so the confirmatory status of the secondary analysis is uncertain. The safety summaries give counts and selected serious event terms, but the record does not provide enough information to establish causality or a full comparative safety profile. These limitations do not erase the posted findings; they narrow the conclusions that can be drawn from them.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Bardoxolone Methyl in Patients With Connective Tissue Disease-associated Pulmonary Arterial Hypertension - CATALYST — NCT02657356.
• Registry record last updated 2025-06-04; status last verified 2025-05.
• Results first posted 2023-05-19.
• 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.

23. Record Summary

CATALYST (NCT02657356) was a Phase 3, randomized, double-blind, parallel-group study of bardoxolone methyl versus placebo in patients with connective tissue disease-associated pulmonary arterial hypertension. The registry lists enrollment of 202, a study start date of 2016-10-04, a primary completion date of 2020-05-07, and a terminated status.

The primary endpoint was change from baseline in six-minute-walk distance at Week 24. The registry’s mixed-model analysis reported a net least-squares-mean difference of -12.86 meters, with a 95% confidence interval from -26.69 to 0.98 and p = 0.068. The record also reports a secondary hazard-ratio analysis and descriptive safety counts.

The primary estimate is uncertain, and the secondary outcome uses a different measure and analysis. The reason for termination, original sample-size justification, interim analysis plan, full statistical analysis plan, and detailed missing-data procedures are not reported in the registry record.