A Study of Sotatercept for the Treatment of Cpc-PH Due to HFpEF (MK-7962-007/A011-16): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Sotatercept for the Treatment of Cpc-PH Due to HFpEF (MK-7962-007/A011-16) (NCT04945460), a randomized phase 2 study. The primary outcome is Base Study: Change From Baseline in Pulmonary Vascular Resistance (PVR) at Week 24.

2. Clinical Question

The study was designed to evaluate efficacy, safety, and tolerability in adults with Cpc-PH due to HFpEF. PVR was the primary endpoint, and change in 6-minute walk distance was identified as a key secondary endpoint. Other listed secondary outcomes included hemodynamic measures, cardiac function measures, and time to first clinical worsening.

The primary result describes changes in PVR within each group and reported between-group analyses. It is not, by itself, a complete account of how patients felt or functioned, nor does it establish how results would apply outside the study population. Interpretation depends on the analysis method, missing-data procedures, uncertainty intervals, and the broader set of outcomes.

3. Trial Design

This was a Phase 2, randomized, double-blind, placebo-controlled, parallel-group study with triple masking recorded in the registry. The registry lists an enrollment of 165 and a completed study status. The study start date is 2021-12-29, and the primary completion date is 2025-09-05.

The placebo-controlled treatment period lasted 24 weeks. One sotatercept group received 0.3 mg/kg subcutaneously every 3 weeks. Another began at 0.3 mg/kg for 3 dosing visits, then escalated to 0.7 mg/kg from the fourth dosing visit and continued every 3 weeks for the remainder of the treatment period. The placebo group also received injections every 3 weeks. The registry lists extension-period safety groups, including participants who moved from placebo to sotatercept.

The primary endpoint was change from baseline in PVR at Week 24, measured by right heart catheterization. The primary analysis population included randomized participants in the Base Study with baseline data available. One participant in the Sotatercept 0.7mg/kg arm was randomized in error without study treatment and was excluded from that analysis.

The registry record identifies 6-minute walk distance as a key secondary endpoint and lists additional outcomes, including time to first clinical worsening, mean pulmonary arterial pressure, pulmonary capillary wedge pressure, and measures of right- and left-ventricular function. It does not report the sample-size justification, planned interim analyses, or a complete statistical analysis plan. The record's enrollment count should not be interpreted as the number analyzed for every endpoint; analysis populations and available data can differ by outcome.

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: Of 319 participants screened, 165 participants were randomized 1:1:1 in the placebo-control period to receive: placebo, 0.3mg/kg sotatercept, or 0.7mg/kg sotatercept. 144 participants continued into the optional extension period. Participants continued their sotatercept treatment as assigned in the placebo-controlled period. If the participant previously received placebo, then they were randomized into one of the two sotatercept treatments.

Only the first 6 of 7 groups are shown.

Base Study

Reasons for not completing this period:

Extension Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis population consisted of all randomized participants in the Base Study with data available at baseline. One participant in the Sotatercept 0.7mg/kg arm was randomized in error without study treatment and was excluded from analysis.

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.

6. Primary Endpoint

The registry lists the primary outcome as Base Study: Change From Baseline in Pulmonary Vascular Resistance (PVR) at Week 24, measured over Baseline and Week 24.

Pulmonary vascular resistance (PVR) is a hemodynamic measure of resistance to blood flow through the pulmonary circulation. In this study, it was measured by right heart catheterization, an invasive measurement method. The endpoint is the change from baseline to Week 24, so it focuses on how the measured value changed during the treatment period rather than only on the value at the end.

As a general statistical and clinical concept, a change-from-baseline measure can help describe whether a group’s measured status moved over time. For PVR, a decrease represents a reduction in the measured resistance; the clinical meaning of a change still depends on the disease context and the rest of the evidence. The primary endpoint does not directly measure symptoms, exercise ability, or clinical events. The registry separately identifies 6-minute walk distance and time to first clinical worsening among the secondary outcomes.

The registry reports medians for the group-level changes and gives a full range based on the minimum and maximum of the 100 medians from the 100 imputed datasets. That is not the same as a conventional confidence interval around a single median. Missing baseline values led to exclusion from the primary analysis; missing Week 24 values were handled using specified procedures, with different ranks assigned for death and clinical worsening. These details affect what the endpoint estimates and how it should be interpreted.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Group summaries and the primary comparison. The primary endpoint is reported as a median change from baseline in PVR. The median is the middle of an ordered set of observations and is less influenced by extreme values than the arithmetic mean. The registry reports a full range defined as the minimum and maximum of the 100 medians from the 100 imputed datasets. This range describes variation across those imputed-data medians; it is not a confidence interval for the treatment effect and should not be read as the spread of individual participant measurements.

The reported between-group method is the Wilcoxon (Mann-Whitney) test, with a Hodges-Lehmann location shift estimate. As a general statistical explanation, the Wilcoxon method compares the relative ranks of observations across groups rather than relying on a comparison of arithmetic means. The Hodges-Lehmann estimate summarizes a location shift between the distributions. It is not necessarily equal to subtracting one displayed group median from another. This distinction matters when reading the reported estimates alongside the group medians.

A confidence interval gives a range of values compatible with the data and the assumptions of the method, at the stated confidence level. A confidence interval that includes zero does not establish that the groups are identical; it indicates uncertainty that includes no location shift under that interval procedure. A p value measures how unusual the observed test result would be under a specified null hypothesis and test assumptions. It does not measure the probability that the null hypothesis is true, the size of a clinical effect, or the probability that a result will replicate.

Missing primary-endpoint values. Participants missing a baseline PVR value were excluded from the primary analysis. For a missing Week 24 value, multiple imputation was used unless the missingness followed death or a clinical worsening event. Multiple imputation is a general method that fills in missing values repeatedly using a statistical model, then combines information across the completed datasets to reflect uncertainty from missingness. The registry record specifies 100 imputed datasets for the reported full-range definition, but does not report the imputation model or its assumptions.

The record assigns the worst-rank score to participants who died and the next-worst-rank score to participants missing Week 24 because of a clinical worsening event. As a general method, ranking such outcomes below observed measurements can incorporate clinically important events into an ordered analysis rather than treating all missing values alike. These assigned ranks are not measured PVR values. Their use also means the analysis is not simply a comparison of observed Week 24 readings.

Time-to-event analysis. Time to first clinical worsening was analyzed using the log-rank method, and the reported comparison measure is a hazard ratio. As a general statistical explanation, the log-rank method compares the timing of events between groups while accounting for participants whose event time is not observed during follow-up in the usual way for time-to-event analysis. A hazard ratio below 1 indicates a lower estimated event hazard in the first group under the comparison’s coding; interpretation remains dependent on the event definition, follow-up, and analysis assumptions. A hazard ratio is not an absolute risk difference and does not directly give an event-free time for an individual.

The registry reports that median time-to-event values were not estimable for the sotatercept groups and gives a lower confidence bound of 6.2 for the placebo group. Other median values and limits are not reported in the registry record. The reported hazard-ratio confidence intervals are important because they show uncertainty around the relative comparison; a point estimate alone would give an incomplete account.

Planning illustrations. Event-count planning illustrations are based on fractions of the observed effect and assumptions stated in the registry record. They are not observed power and do not establish what the study was designed to detect. As a general concept, planning calculations depend on assumptions about effect size, allocation, significance level, and analysis method. A planning illustration should not be mistaken for evidence that the trial had a particular probability of detecting an effect after the data were observed.

Multiplicity and interpretation. Multiple endpoints and comparisons create opportunities for apparently notable results to arise by chance. As a general statistical principle, multiplicity procedures can control the overall chance of false-positive conclusions across a family of tests, but the registry record does not report whether such procedures were used. The primary endpoint has a different role from secondary outcomes, which can add context but should be interpreted cautiously when adjustment is unreported. The registry record also does not report a complete statistical analysis plan, so details beyond those explicitly stated cannot be assumed.

10. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary between-group analyses using the Wilcoxon (Mann-Whitney) method and Hodges-Lehmann location shift estimates. For 0.3 mg/kg versus placebo, the reported estimate is -81.51, with a 95% confidence interval from -144.42 to -18.61 and p = 0.004. For 0.7 mg/kg versus placebo, it is -59.80, with a 95% confidence interval from -121.86 to 2.27 and p = 0.024.

The group medians and the reported location-shift estimates answer related but different questions. A simple recalculation from group summary statistics, if shown, would not reproduce the sponsor’s rank-based analysis: the Hodges-Lehmann estimate uses the underlying comparison structure rather than only subtracting the displayed medians. The registry record does not provide a separate recalculated primary estimate to report. It does not flag a discrepancy, so there is no conflicting recalculation to set against the sponsor’s analysis.

The second comparison’s confidence interval includes zero even though its listed p value is 0.024. The record does not explain this pairing. Both statistics should be retained as reported, and the interval should not be silently replaced or treated as though it excludes zero. In general, sponsor analyses defined by the registered statistical method take precedence over simple unadjusted recalculations when the estimates differ.

11. Multiplicity, Missing Data and Analysis Populations

The primary analysis population consisted of randomized participants in the Base Study with baseline data available. Participants missing baseline PVR were excluded. The record also states that one participant in the Sotatercept 0.7mg/kg arm was randomized in error without study treatment and excluded from the primary analysis. This means the primary result is not necessarily based on every randomized participant or on the same available-data set used for other outcomes.

For missing Week 24 PVR, the registry specifies multiple imputation, except for missingness due to death or clinical worsening. Death was assigned the worst-rank score, and missingness due to clinical worsening was assigned the next-worst-rank score. These choices matter because missing data can be related to a participant’s health status. As a general statistical point, an imputation analysis depends on assumptions about how missing values relate to observed information; the imputation model and its assumptions are not reported in the registry record.

The registry lists numerous secondary outcomes and analyses, but a multiplicity adjustment is not reported. When many tests are conducted, the chance of at least one apparently positive result can increase. This general issue makes unadjusted secondary p values less secure as confirmatory evidence. The registry record does not report a hierarchy of testing, a family of hypotheses, or a procedure for controlling multiplicity.

The record also does not report the sample-size justification, interim analyses, or a complete statistical analysis plan. It is therefore not possible to determine from the registry record which analyses were planned in detail, whether any decisions were made after reviewing accumulating data, or how the trial’s enrollment target was justified. These points should be treated as unreported, not inferred from the observed enrollment or the posted results.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cardiac failure (13.6%), hypertransaminasaemia (8.7%), acute kidney injury (8.7%), acute respiratory failure (8.7%), angina pectoris (4.5%).

During the placebo-controlled treatment period, the registry reports serious adverse events for 11 participants in the Sotatercept 0.3 mg/kg group, 18 in the Sotatercept 0.7 mg/kg group, and 12 in the placebo group. Other adverse events were reported for 45, 44, and 37 participants in those respective groups. Deaths reported for the same period were 0, 1, and 2. These are counts, not estimates of comparative risk, and no percentages have been calculated.

The registry also gives separate extension-period counts. Those groups include continued sotatercept and participants who moved from placebo to sotatercept. Extension-period counts should not be combined with placebo-controlled-period counts because the exposure and follow-up contexts differ. Serious adverse event terms listed among the more frequent terms include Cardiac failure at 13.6%, Hypertransaminasaemia at 8.7%, Acute kidney injury at 8.7%, Acute respiratory failure at 8.7%, Angina pectoris at 4.5%, and Atrial fibrillation at 4.5%.

As a general safety principle, event counts alone do not show whether a treatment caused an event. They also do not summarize severity, duration, recurrence, or individual clinical circumstances. The registry record’s counts should be interpreted alongside the study design, exposure period, group sizes at risk, and the limitations of the reported safety information.

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

The primary result describes a difference in PVR change between treatment groups during a 24-week placebo-controlled period. The group medians moved in different directions: both sotatercept groups had negative median changes, while the placebo group had a positive median change. The reported rank-based comparison for 0.3 mg/kg had a confidence interval entirely below zero. For 0.7 mg/kg, the registry’s confidence interval includes zero, while its listed p value is 0.024; the record does not reconcile those statistics.

These results do not, by themselves, show how every participant’s PVR changed. Group summaries conceal individual variation, and the reported analysis includes imputation and ranking rules for missing values and clinical events. Nor does the PVR endpoint directly measure symptoms, quality of life, exercise performance, or survival. The study listed other outcomes relevant to those dimensions, but most secondary results are not reported in the registry record.

The findings also do not establish that either regimen works or does not work for all adults with Cpc-PH due to HFpEF. A Phase 2 study is one part of an evidence base, and estimates remain subject to uncertainty, endpoint-specific limitations, and the analysis assumptions. The registry record does not provide a complete statistical analysis plan or multiplicity procedure, which limits how strongly multiple reported comparisons can be interpreted together.

Safety counts give a descriptive account of reported events in distinct treatment periods, not proof of causation or a complete assessment of comparative safety. The placebo-controlled and extension periods have different exposure contexts, and their counts should remain separate. Overall, the registry results support a careful reading of the measured outcomes and their uncertainty, rather than a broad conclusion about patient benefit or individual treatment decisions.

14. Why This Trial Matters Statistically

This trial is statistically informative because it links a hemodynamic primary endpoint to a randomized, placebo-controlled comparison and also reports a clinical-worsening analysis. Randomization is a general design feature intended to make treatment groups comparable on average, while masking can reduce some sources of bias in outcome assessment and participant behavior. The registry records randomized allocation and triple masking, although it does not identify which parties were masked.

The primary endpoint’s analysis illustrates why the named summary statistic and method matter. The group medians are not the same quantity as the Hodges-Lehmann location shift, and the rank-based comparison cannot be reconstructed from the displayed medians alone. The missing-data procedures also shape the estimand: observed Week 24 PVR values, imputed values, and ranks assigned for death or clinical worsening are handled differently.

Uncertainty is important to the interpretation. The two primary comparisons have different confidence-interval patterns, and the registry reports a p value for the 0.7 mg/kg comparison that does not align with whether its confidence interval includes zero under the usual two-sided interpretation. Acknowledging this feature is more informative than reducing the result to a threshold label.

The record also highlights the distinction between a primary result and a broader outcome set. Secondary outcomes may contribute useful context, but their results are not all reported, and multiplicity handling is not described. These limitations matter when deciding how much weight to give any single finding. The trial therefore offers a useful case study in reading endpoint definitions, analysis populations, missing-data rules, effect estimates, and uncertainty together.

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 does not report the full statistical analysis plan, sample-size justification, interim analyses, or a multiplicity procedure. Without those details, it is not possible to assess how every analysis was prespecified or how the family of endpoint comparisons was handled. Several secondary outcomes are listed, but their results are not reported in the registry record.

The primary endpoint uses multiple imputation for some missing Week 24 values, while death and clinical worsening are handled through assigned ranks. The imputation model and its assumptions are not reported in the registry record. The primary analysis also excludes participants without baseline PVR and identifies one participant randomized in error without treatment as excluded. These features limit the interpretation of the primary analysis as a summary of all randomized participants.

The primary outcome is a hemodynamic measure, not a direct measure of symptoms or daily functioning. The registry identifies 6-minute walk distance as a key secondary outcome, but its result is not reported in the registry record. The lack of reported results for many secondary measures prevents an assessment of whether findings across different domains were consistent.

The registry reports a confidence interval that includes zero for the 0.7 mg/kg PVR comparison alongside p = 0.024, without explaining the difference. The time-to-event median estimates are also largely unavailable, and the record does not provide enough detail to assess all aspects of follow-up and censoring. Finally, the listed adverse-event counts are descriptive and do not establish causation. These issues call for a qualified interpretation of the posted results rather than a single broad conclusion.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Sotatercept for the Treatment of Cpc-PH Due to HFpEF (MK-7962-007/A011-16) — NCT04945460.
• Registry record last updated 2026-09-23; status last verified 2026-08.
• Results first posted 2026-09-23.
• 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

NCT04945460 is a completed, randomized, double-blind, placebo-controlled Phase 2 study of sotatercept for adults with Cpc-PH due to HFpEF. The placebo-controlled treatment period lasted 24 weeks and included two sotatercept regimens and placebo, given subcutaneously every 3 weeks. The registry lists enrollment of 165.

The primary endpoint was change from baseline in PVR at Week 24, measured by right heart catheterization. The record reports group medians, full ranges across imputed-dataset medians, and Wilcoxon-based comparisons with Hodges-Lehmann estimates. It also reports time-to-first-clinical-worsening hazard ratios and safety counts across treatment periods.

Important interpretation details include the missing-data and ranking procedures, an unreported multiplicity strategy, secondary outcome results that are not reported in the registry record, and a difference between the 0.7 mg/kg confidence interval and listed p value that the record does not explain. These details temper conclusions drawn from any single estimate.