Macitentan for the Treatment of Digital Ulcers in Systemic Sclerosis Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Macitentan for the Treatment of Digital Ulcers in Systemic Sclerosis Patients (NCT01474109), a randomized phase 3 study. The primary outcome is Incidence Rate of New Digital Ulcers (DUs) up to Week 16.

2. Clinical Question

This is a question about an average treatment-group comparison over a defined observation period, not a guarantee that every participant had the same course. The primary measure reflects newly recorded ulcers under the study’s definition; it does not encompass every possible feature of systemic sclerosis or ulcer experience. The registry reports separate comparisons for each macitentan dose against placebo. It does not report a direct comparison between the two dose groups.

3. Trial Design

DUAL-1 was a Phase 3, randomized, parallel-group, placebo-controlled study of macitentan in patients with systemic sclerosis and ongoing digital ulcers. The registry record describes the study as triple-masked and multicenter. It reports enrollment of 289 and lists Actelion as sponsor. The recorded study status is completed; the recorded start is 2011-12 and primary completion is 2013-11.

Participants were randomized in a 1:1:1 ratio to macitentan 3mg, macitentan 10mg, or matching placebo. Each was taken once daily. The described design had an initial fixed 16-week Period 1, followed by a variable-duration Period 2. Participants completing Period 1 were to continue their original randomized treatment into Period 2 until the last randomized participant had completed Period 1. The primary endpoint covers baseline to Week 16, so it concerns the fixed initial period rather than the variable extension.

The registry lists the primary endpoint as the incidence rate of new digital ulcers up to Week 16. It also lists secondary outcomes covering absence of new ulcers, digital-ulcer complications, and measures of hand function and disability. A time-to-first-complication objective is described for the entire treatment period. The registry record does not report detailed methods for the randomization sequence, allocation concealment, or masking procedures. It also does not report a sample-size justification or a complete statistical analysis plan. Those omissions prevent a fuller assessment of how design decisions were made and 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: A screening visit was performed between Day -14 and Day -1 of the study. Of the 327 patients screened for the study, 38 were screen failures.

Period 1: Baseline to Week 16

Period 2: Week 16 to End of Study

Analysis population for the primary outcome, as defined in the registry: Full analysis set

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: Full analysis set

6. Primary Endpoint

The registry lists the primary outcome as Incidence Rate of New Digital Ulcers (DUs) up to Week 16, measured over Baseline to week 16.

The primary endpoint is the incidence rate of new digital ulcers from baseline to Week 16. The registry defines the measure using the number of new ulcers relative to total observation days, with incidence adjusted for 16 weeks of observation. Relating the count to observation time can make the measure more informative when participants contribute different amounts of follow-up, though the registry does not provide participant-level observation time or detailed missing-observation handling.

The ulcer definition is specific. Ulcers from the proximal interphalangeal joint distally were recorded, including the dorsal and volar hand surfaces and digital tip. Location was noted, and newly observed locations were recorded at subsequent visits. Ulcers that occurred and healed between visits but were reported by participants were not recorded as new. An ulcer developing over a previously healed ulcer was counted as new. The assessment process used an experienced physician or trained rater, with the same rater for a participant at each visit whenever possible.

A count-based incidence rate is useful when the clinical question concerns the frequency of newly occurring events over a period, rather than only whether any event occurred. It preserves information about repeated new ulcers, whereas a binary outcome would classify participants only by whether they had at least one. The reported statistical comparison uses a negative binomial-2 regression estimate of new ulcers per participant. The registry does not report a dispersion estimate or enough detail to reconstruct the model from participant-level data.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Why use a count model?

The primary outcome is based on counts of new digital ulcers, with observation time incorporated into the incidence rate. Count models are designed for outcomes that take nonnegative integer values and can accommodate repeated events within a participant. A model can compare groups while accounting for the distribution of counts rather than reducing each participant’s experience to a simple yes-or-no indicator.

The reported method is negative binomial-2 regression, abbreviated NB-2 in the registry record. In general, a negative binomial model is useful when count data vary more than a basic Poisson model would expect. This is called over-dispersion: the observed variability is greater than the mean-variance relationship assumed by the Poisson model. Allowing extra variability can avoid overly narrow uncertainty estimates when that assumption is appropriate. The registry does not report the estimated dispersion parameter, diagnostics, covariates, or full model specification, so the adequacy of the model cannot be assessed from the record alone.

How should the estimate and interval be read?

The registry reports an NB-2 estimate of new ulcers per participant for each macitentan-versus-placebo comparison, along with a 95% confidence interval and p-value. In general, a confidence interval represents a range of values compatible with the observed data and the model’s assumptions at a stated confidence level. It describes estimation uncertainty; it is not a range containing a known proportion of individual participant outcomes. The reported intervals, 0.663 to 1.834 for macitentan 3mg and 0.763 to 2.106 for macitentan 10mg, include 1. For a relative rate comparison, 1 represents equal estimated rates. The intervals therefore do not isolate a difference in one direction.

A p-value summarizes how unusual data at least as incompatible with a null hypothesis would be under that hypothesis and the model assumptions. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The reported p-values are 0.706 and 0.360. They do not by themselves show equivalence, and they should be interpreted alongside the estimates and intervals. The registry record does not identify a significance threshold or explain whether these tests were adjusted for multiple comparisons.

How does the primary measure differ from a binary outcome?

An incidence rate retains information about repeated new events and observation time. A binary endpoint, such as whether a participant had no new ulcer, treats each participant as having one of two classifications and does not distinguish one event from multiple events. The registry reports both kinds of outcomes: the count-based primary endpoint and binary secondary measures. They answer related but different questions, so their estimates need not agree exactly. The primary outcome’s rules for ulcers occurring and healing between visits also mean that its result depends on the specified ascertainment process.

What do the recalculations represent?

Some additional comparisons are simple, unadjusted recalculations from posted summary statistics. For binary outcomes, a risk difference describes the contrast in observed proportions between groups; a confidence interval expresses uncertainty around that contrast. These calculations do not reproduce the sponsor’s chi-squared analyses or adjust for other variables. They are useful as an independent descriptive view of the posted counts, but should not be treated as replacements for the sponsor’s stated analyses. For the primary count endpoint, the posted record does not include the participant-level information needed to rebuild the NB-2 model.

Planning illustrations based on observed effects are not observed power. In general, a sample-size illustration asks how many participants per arm might be needed to detect an assumed effect under specified design assumptions. Such figures depend on the assumed effect and statistical method; they do not change what the completed trial observed. The registry record does not report that these illustrations were the trial’s original design rationale.

What information is needed for a fuller statistical review?

A fuller review would require the prespecified analysis plan, the model specification and diagnostics, the treatment of missing outcome data, and details on how the full analysis set was defined. It would also require the planned approach to multiplicity and the handling of any protocol deviations. These items are not reported in the registry record. Without them, the posted estimates can be described and interpreted, but some assumptions and analysis choices cannot be independently checked.

10. Analyses Reported by the Sponsor

The registry reports sponsor analyses for the two primary comparisons using negative binomial-2 regression. The reported estimates are 1.103 for macitentan 3mg versus placebo and 1.268 for macitentan 10mg versus placebo, with 95% confidence intervals of 0.663 to 1.834 and 0.763 to 2.106, respectively. The p-values are 0.706 and 0.360. These are the principal reported inferential results for the primary count outcome.

The posted arm-level incidence figures are descriptive summaries, and they are not the same quantity as a model-based comparison. Simple unadjusted recalculations based on summary statistics can offer a check on basic contrasts for outcomes where the relevant counts are posted, but they cannot reproduce the sponsor’s primary regression model. The record does not contain the participant-level outcome data or full model specification required to do so.

For the binary secondary outcomes, the sponsor reports chi-squared analyses as odds ratios, while the simple recalculations express contrasts as risk differences. Those measures use different scales and methods, so their numerical estimates and p-values are not expected to be identical. The registry does not flag a discrepancy between the posted analyses and the recalculations. The sponsor’s analyses remain the reported trial analyses; the recalculations are supplementary descriptive checks, not an alternative prespecified analysis.

11. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how multiplicity was handled. The trial included two dose comparisons for the primary outcome and lists multiple secondary outcomes, with additional analyses reported for some of them. When many hypotheses are tested, the chance of at least one nominally significant result can rise. A prespecified multiplicity strategy can control that risk, but without a reported strategy it is not possible to tell whether the stated p-values were adjusted or how they were intended to support confirmatory claims.

The registry identifies the primary analysis population as the full analysis set, but does not define the population’s inclusion rules in detail. It does not report how missing primary endpoint observations were handled, whether any imputation or sensitivity analyses were used, or how participants with incomplete follow-up contributed to the incidence calculation. Because the primary rate uses observation days, details about follow-up time and missing visits could matter to interpretation. Those details are not reported in the registry record.

The record also does not report an interim-analysis plan, a sample-size justification, or the full statistical analysis plan. The absence of those details does not establish that such procedures were absent; it means they are not documented in the registry record. In particular, the posted sample-size illustrations based on observed effects are planning illustrations, not evidence of the trial’s original assumptions or observed power.

These omissions place limits on a registry-based assessment. The posted point estimates, intervals, and p-values can be described, but the record does not support a complete review of prespecification, analysis-population construction, missing-data assumptions, or family-wise error control. Those questions would require the study protocol, statistical analysis plan, and fuller trial reporting.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gangrene (3.2%), skin ulcer (3.2%), infected skin ulcer (3.1%), pneumonia (2.1%), ventricular extrasystoles (2.1%).

The registry reports serious adverse events for participants at risk in each group. The serious-event counts were 17 in the macitentan 3mg group, 14 in the macitentan 10mg group, and 13 in the placebo group. Other adverse-event counts were 61, 73, and 69, respectively. These are reported counts and should be read alongside their listed at-risk group counts: 94, 97, and 97. They are not, by themselves, adjusted comparisons or evidence that one group had a greater underlying risk.

Serious-event terms listed in the record include GANGRENE and SKIN ULCER at 3.2%, INFECTED SKIN ULCER at 3.1%, PNEUMONIA and VENTRICULAR EXTRASYSTOLES at 2.1%, and BRONCHOPNEUMONIA at 1.1%. A listed term does not establish that the event was caused by treatment. The registry record does not provide a causal assessment for these terms.

Deaths are not reported in the registry record. They should not be interpreted as zero. The available adverse-event summary does not provide the detailed timing, duration, severity gradations, or participant-level context needed for a complete safety assessment. These figures describe reported events during the trial and are not medical advice.

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

The primary result answers a focused statistical question: how the modeled incidence of newly recorded digital ulcers through Week 16 compared between each macitentan dose and placebo in the study population. The reported estimates and confidence intervals do not establish a clear difference in either comparison. They also do not demonstrate that the groups were equivalent, because the intervals include values on both sides of the equal-rate reference.

The findings apply to the endpoint as defined in the registry, the assessment procedures used, the study period, and the participants included in the analysis. The endpoint counts only ulcers meeting the stated location and recording rules. It is not a complete measure of every aspect of ulcer burden, hand function, systemic sclerosis, or a participant’s overall experience. The outcomes concerning hand functionality and disability are separate measures, and their numerical results are not reported in the registry record.

A lack of a statistically clear difference is not proof that there is no effect. Conversely, an estimate or a point estimate favoring one direction would not alone establish a meaningful or reliable effect. The confidence intervals express uncertainty under the analysis assumptions, while the p-values summarize compatibility with a null hypothesis under those same assumptions. Neither replaces consideration of the model, outcome definition, missing-data handling, or the broader set of analyses.

The registry record also does not provide the full statistical plan, missing-data methods, or multiplicity approach. As a result, conclusions should remain limited to the reported comparisons. The results do not establish effects beyond the studied population or period, nor do the safety counts establish causation. They are a record of observed and analyzed outcomes, not a recommendation about treatment.

14. Why This Trial Matters Statistically

This trial illustrates why endpoint choice matters. A rate of newly recorded ulcers captures repeated events in relation to observation time, whereas a binary outcome records whether a participant crossed a yes-or-no threshold. These measures describe different aspects of the same clinical problem and can produce different patterns without being mathematically inconsistent.

It also shows why estimates should be read together with uncertainty. A point estimate gives a compact summary of a comparison, but its confidence interval indicates how much precision the data and model support. In the primary analyses, the intervals span values below and above the equal-rate reference. That limits the strength of any directional conclusion, even though the reported estimates are specific numerical results.

The registry’s use of negative binomial-2 regression highlights a common feature of clinical count outcomes: repeated events may vary substantially across participants. A model that allows over-dispersion can be more suitable than a model with a restrictive mean-variance assumption, provided its assumptions fit the data. Yet the registry record does not report the model diagnostics or full specification, so the appropriateness of those choices cannot be independently judged from the posted summary.

Finally, the record demonstrates the distinction between a primary analysis and supporting outcomes. Secondary measures can add context, but multiplicity and incomplete reporting constrain how strongly they can support claims. For statistical readers, the key issues are not only which estimates were reported, but also how endpoints were defined, how observation time and missingness were treated, and whether the analysis plan was prespecified. Several of these details are not reported in the registry record.

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 record gives a concise account but lacks details needed to audit the primary analysis fully. It identifies NB-2 regression without reporting the complete model specification, covariates, dispersion estimate, or model diagnostics. It also does not provide participant-level event data or observation times, limiting independent reconstruction of the count analysis.

The primary endpoint relies on a particular ulcer-assessment process. Ulcers occurring and healing between visits, when reported by participants, were not recorded as new; ulcers over a previously healed ulcer were counted as new. These rules define the estimand represented by the result, but the record does not quantify the potential effect of ascertainment differences or describe inter-rater assessment in detail. It notes that the same rater was used whenever possible, not that this was achieved for every assessment.

The full analysis set is named but not defined, and missing-data handling is not reported in the registry record. The analysis population and incomplete follow-up could affect how much information contributed to the incidence rates. The record also does not report the multiplicity strategy, the original sample-size justification, interim analyses, or the complete statistical analysis plan.

Only some secondary outcomes have numerical results in the record. The listed hand-function and disability outcomes therefore cannot be assessed numerically from this information. The safety summary gives counts and selected serious-event terms, but not detailed participant-level context or a causal assessment. These limitations do not invalidate the reported results; they restrict how comprehensively they can be evaluated.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Macitentan for the Treatment of Digital Ulcers in Systemic Sclerosis Patients — NCT01474109.
• Registry record last updated 2015-01-06; status last verified 2015-01.
• Results first posted 2015-01-06.
• 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

NCT01474109 is a completed Phase 3 randomized, triple-masked, parallel-group study of macitentan for digital ulcers in systemic sclerosis. The registry describes 289 enrolled participants assigned to macitentan 3mg, macitentan 10mg, or matching placebo, each taken once daily. The primary endpoint was the incidence rate of new digital ulcers from baseline to Week 16.

The reported NB-2 estimates versus placebo were 1.103 for macitentan 3mg and 1.268 for macitentan 10mg. Their 95% confidence intervals included 1, and the reported p-values were 0.706 and 0.360. The primary results do not establish a clear difference between either dose and placebo, and they do not establish equivalence. Important details on multiplicity, missing data, and model specification are not reported in the registry record.