← Clinical Trials
Idiopathic Pulmonary Fibrosis Phase 3 Terminated NCT04552899

STARSCAPE: Complete Statistical Analysis of Zinpentraxin Alfa in Idiopathic Pulmonary Fibrosis

An independent statistical analysis of the randomized, double-blind phase 3 STARSCAPE trial evaluating recombinant human pentraxin-2 (rhPTX-2; PRM-151), also called zinpentraxin alfa, versus placebo in participants with idiopathic pulmonary fibrosis.

Phase 3  ·  Randomized  ·  Double-blind  ·  Enrollment 665
Scope of this record

This page separates reported trial results from statistical interpretation. Numerical efficacy and safety results are taken from the ClinicalTrials.gov record. The registry reports that the study was terminated early following a futility analysis concluding that zinpentraxin alfa was unlikely to meet its primary endpoint.

Registry record: This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. View NCT04552899 on ClinicalTrials.gov.

1. Trial at a Glance

STARSCAPE was a randomized, double-blind, parallel-group phase 3 trial evaluating zinpentraxin alfa versus placebo in participants with idiopathic pulmonary fibrosis. The registry reports an enrollment of 665 participants, 1 registered primary endpoint, 16 posted outcome measures, and 4 posted statistical analyses.

665
Enrollment
Registry enrollment
2
Arms
Zinpentraxin alfa vs placebo
-20.83
FVC Difference
95% CI -87.94 to 46.29 mL
0.54
Primary P-value
Two-sided superiority analysis
FeatureSTARSCAPE
TrialSTARSCAPE
NCT IDNCT04552899
PhasePhase 3
ConditionIdiopathic Pulmonary Fibrosis
AllocationRandomized
Design modelParallel
MaskingDouble
Primary purposeTreatment
Enrollment665
InterventionsPRM-151 (Zinpentraxin Alfa) and placebo
Trial statusTerminated
Start2021-03-19
Primary completion2023-02-10
Lead sponsorHoffmann-La Roche
Sponsor typeIndustry

2. Clinical Question

The primary clinical question was whether zinpentraxin alfa could produce a different change in forced vital capacity compared with placebo from baseline up to Week 52 in participants with idiopathic pulmonary fibrosis.

Population

Participants with idiopathic pulmonary fibrosis enrolled in the phase 3 STARSCAPE trial.

Intervention

PRM-151 (zinpentraxin alfa).

Comparator

Placebo.

Primary question

Does zinpentraxin alfa produce a different absolute change in FVC from baseline through Week 52 compared with placebo?

3. Trial Design

01
Randomize 665 enrolled
02
Double-blind Two treatment groups
03
Parallel Zinpentraxin alfa vs placebo
04
Assess FVC through Week 52
05
Analyze Efficacy and safety
INTERVENTION

Zinpentraxin Alfa

  • PRM-151 (zinpentraxin alfa)
  • Compared with placebo in the randomized parallel-group design
COMPARATOR

Placebo

  • Placebo drug
  • Served as the randomized comparison group
Early termination is part of the statistical story. The registry states that the study was terminated early based on a futility analysis that concluded zinpentraxin alfa was unlikely to meet its primary endpoint. This is important when interpreting the posted results because the observed data arose from a trial that did not continue to its originally intended completion.

4. Trial Timing and Status

2021-03-19

Study start

The registry lists March 19, 2021 as the study start date.

2023-02-10

Primary completion

The registry lists February 10, 2023 as the primary completion date.

Terminated

Early termination

The study was terminated early following a futility analysis. The registry states that the futility analysis concluded that zinpentraxin alfa was unlikely to meet its primary endpoint.

5. Endpoints

EndpointRegistry definition / time frameType
Absolute Change in Forced Vital Capacity (FVC [mL]) From Baseline up to Week 52 Primary; continuous
Time to Disease Progression From Baseline up to 1 year Secondary; time-to-event
Time to First Respiratory-related Hospitalizations From Baseline up to 1 year Secondary; time-to-event
Time to First Acute Exacerbation of Idiopathic Pulmonary Fibrosis (IPF) From Baseline up to 1 year Secondary; time-to-event

The registry reports 16 outcome measures in total, but the ClinicalTrials.gov record contains formal numerical analyses for 1 primary endpoint and 3 secondary endpoints. The remaining posted outcome measures are therefore not described numerically here.

6. Statistical Methodology

Primary endpoint: Random Coefficient Regression Model

The primary endpoint, absolute change in FVC from baseline up to Week 52, was analyzed using a Random Coefficient Regression Model (RCRM). The registry identifies the outcome as continuous and reports the effect measure as the difference in change from baseline between zinpentraxin alfa and placebo.

Reported primary effect measure
Difference in Change from Baseline = Zinpentraxin Alfa − Placebo

The registry reports an estimated difference of -20.83 mL, with a two-sided 95% confidence interval of -87.94 to 46.29 mL.

Time-to-event endpoints: Log-rank testing

The three registry-reported secondary analyses were time-to-event outcomes. Each used a log-rank test and reported a hazard ratio as the effect measure. The analysis population was the full analysis set described in the registry.

Conceptual survival-analysis framework
HR < 1  →  lower estimated instantaneous event rate in the zinpentraxin alfa group

A hazard ratio compares estimated event rates over follow-up under the time-to-event analysis. It is not an absolute probability, a median survival time, or the proportion of participants who benefit.

Full analysis set

The registry analysis descriptions state that the full analysis set included all randomized participants who received at least one administration, either a full or partial dose, of study drug and used the grouping according to treatment assignment. The analysis text also identifies an intention-to-treat analysis concept.

Important terminology point: the registry definition is not identical to a simple statement of "all randomized participants." The full analysis set described here requires at least one administration of study drug, while the treatment grouping follows randomized assignment.

7. Primary Endpoint Result: Absolute Change in FVC

The registered primary endpoint was absolute change in forced vital capacity from baseline up to Week 52. The registry provides a formal statistical analysis using a Random Coefficient Regression Model.

Difference in change from baseline

-20.83 mL

95% CI: -87.94 to 46.29 mL   ·   P = 0.54

Comparison: zinpentraxin alfa vs placebo

FeaturePrimary FVC analysis
EndpointAbsolute Change in Forced Vital Capacity (FVC [mL])
Time frameFrom Baseline up to Week 52
Endpoint typeContinuous
Analysis methodRandom Coefficient Regression Model (RCRM)
Effect measureDifference in Change from Baseline
Estimate-20.83 mL
95% CI-87.94 to 46.29 mL
P-value0.54
Hypothesis typeSuperiority
Clinical Biostats interpretation

The estimated difference in change from baseline was -20.83 mL for zinpentraxin alfa relative to placebo. Because the effect is expressed as treatment minus placebo, the negative point estimate means the estimated change was 20.83 mL lower in the zinpentraxin alfa group under the reported model.

The estimate should not be interpreted as saying that every participant lost exactly 20.83 mL, nor does it represent an individual patient's predicted FVC change. It is a model-based comparison of the groups' change from baseline.

The 95% confidence interval of -87.94 to 46.29 mL spans both negative and positive values. Thus, the reported interval is compatible with a range of group differences in either direction rather than identifying a precise treatment effect.

The P-value of 0.54 is a measure of compatibility with the statistical testing framework under the null hypothesis; it is not a measure of effect size, clinical importance, or the probability that the treatment works. The magnitude and uncertainty of the estimated difference are better represented by the estimate and confidence interval.

Because the trial was terminated early after a futility analysis, the endpoint should also be interpreted in the context of the trial's early stopping. The registry's stated reason for termination was that zinpentraxin alfa was unlikely to meet the primary endpoint.

8. Secondary Endpoint Results

Time to Disease Progression

Hazard ratio

1.15

95% CI: 0.91–1.47   ·   P = 0.2512

Time frame: From Baseline up to 1 year

FeatureReported analysis
EndpointTime to Disease Progression
Time frameFrom Baseline up to 1 year
AnalysisLog Rank
Effect measureHazard Ratio (HR)
Estimate1.15
95% CI0.91–1.47
P-value0.2512
Hypothesis typeSuperiority
Clinical Biostats interpretation

An HR of 1.15 means that the estimated instantaneous rate of the analyzed progression event was higher in the zinpentraxin alfa group than in the comparison group under the reported time-to-event model. It does not mean that 15% more participants progressed, nor does it mean that an individual participant had a 15% higher probability of progression.

The 95% CI of 0.91–1.47 includes 1.00, indicating substantial uncertainty about the direction and magnitude of the relative hazard. The P-value of 0.2512 is not an effect-size measure and should not be interpreted as the probability that the null hypothesis is true.

The analysis uses time-to-event methods, so censoring and the pattern of follow-up are integral to interpretation. The ClinicalTrials.gov record does not report median time to progression or a Kaplan-Meier curve, so those quantities are not presented here.

Time to First Respiratory-related Hospitalizations

Hazard ratio

1.01

95% CI: 0.51–1.97   ·   P = 0.9833

Time frame: From Baseline up to 1 year

FeatureReported analysis
EndpointTime to First Respiratory-related Hospitalizations
Time frameFrom Baseline up to 1 year
AnalysisLog Rank
Effect measureHazard Ratio (HR)
Estimate1.01
95% CI0.51–1.97
P-value0.9833
Hypothesis typeSuperiority
Clinical Biostats interpretation

An HR of 1.01 is very close to 1.00, so the point estimate indicates little difference in the estimated instantaneous hospitalization rate between the groups. It does not establish equivalence or prove that the two groups have identical risks.

The confidence interval is 0.51–1.97, which is wide and spans 1.00. This indicates considerable uncertainty around the estimated relative hazard. The P-value of 0.9833 does not measure the size or clinical importance of the effect.

Because this is a time-to-event endpoint, participants who do not experience a first respiratory-related hospitalization during the relevant observation period can contribute censored follow-up. The ClinicalTrials.gov record does not provide event counts, median time to hospitalization, or Kaplan-Meier estimates.

Time to First Acute Exacerbation of Idiopathic Pulmonary Fibrosis

Hazard ratio

0.86

95% CI: 0.40–1.86   ·   P = 0.7005

Time frame: From Baseline up to 1 year

FeatureReported analysis
EndpointTime to First Acute Exacerbation of Idiopathic Pulmonary Fibrosis (IPF)
Time frameFrom Baseline up to 1 year
AnalysisLog Rank
Effect measureHazard Ratio (HR)
Estimate0.86
95% CI0.40–1.86
P-value0.7005
Hypothesis typeSuperiority
Clinical Biostats interpretation

An HR of 0.86 corresponds to an estimated instantaneous event rate approximately 14% lower in the zinpentraxin alfa group relative to the comparison group, if the hazard-ratio interpretation is applied directly. This is a relative model-based quantity, not a 14% reduction in the proportion of participants experiencing an acute exacerbation.

The 95% CI of 0.40–1.86 is wide and includes 1.00. The estimate therefore has substantial statistical uncertainty, and the P-value of 0.7005 should not be interpreted as a probability that the treatment has no effect.

The registry does not provide median event time, event counts, or a survival curve for this analysis in the ClinicalTrials.gov record. Those quantities therefore cannot be inferred or reconstructed from the reported hazard ratio alone.

9. Results Summary

EndpointMethodEffect estimate95% CIP-value
Absolute Change in FVC, baseline to Week 52 Random Coefficient Regression Model Difference in change: -20.83 mL -87.94 to 46.29 mL 0.54
Time to Disease Progression, up to 1 year Log-rank HR 1.15 0.91–1.47 0.2512
Time to First Respiratory-related Hospitalizations, up to 1 year Log-rank HR 1.01 0.51–1.97 0.9833
Time to First Acute Exacerbation of IPF, up to 1 year Log-rank HR 0.86 0.40–1.86 0.7005

The ClinicalTrials.gov record therefore contain one formal primary-endpoint analysis and three formal secondary time-to-event analyses. The primary analysis reports a difference in FVC change of -20.83 mL with a 95% confidence interval extending from -87.94 to 46.29 mL. The secondary hazard-ratio estimates are 1.15, 1.01, and 0.86 for disease progression, respiratory-related hospitalization, and acute exacerbation, respectively.

10. Statistical Methods Explained

Why was a Random Coefficient Regression Model used for FVC?

FVC is a continuous measurement observed over time. A Random Coefficient Regression Model can account for repeated measurements within participants while allowing individual participants to have their own trajectory characteristics. In this trial, the registry specifically reports RCRM as the method for the primary FVC analysis.

The important distinction is between the observed repeated measurements and the estimated treatment difference in change from baseline. The reported -20.83 mL is the treatment comparison generated by the model, not a raw difference calculated from two single observations.

What does a difference in change from baseline of -20.83 mL mean?

The effect measure is defined as the difference in change from baseline between zinpentraxin alfa and placebo. A negative value means the estimated change was lower in the zinpentraxin alfa group under the reported model. Here, the estimate is -20.83 mL.

The estimate should not be converted into a percentage without additional baseline information. The ClinicalTrials.gov record does not provide the baseline FVC values needed for such a calculation.

Why does the confidence interval matter?

The 95% confidence interval for the primary difference is -87.94 to 46.29 mL. This interval describes statistical uncertainty around the estimated group difference under the analysis framework. It does not describe the range of individual participant responses.

Because the interval includes zero, the data are compatible with both a negative and a positive treatment difference within the uncertainty represented by the interval.

What does a hazard ratio of 1.15 mean?

For the time-to-disease-progression analysis, the reported HR is 1.15. In a conventional hazard-ratio interpretation, this represents an estimated instantaneous event rate 15% higher in the zinpentraxin alfa group relative to the comparator. It is not equivalent to saying that 15% more participants experienced disease progression.

Hazard ratios summarize relative event rates over time. They depend on the time-to-event framework, including censoring and the model assumptions underlying the reported effect estimate.

Why does an HR of 0.86 not mean that 14% fewer participants had an acute exacerbation?

The HR of 0.86 for time to first acute exacerbation represents a relative comparison of instantaneous event rates. A simple subtraction from 1 gives 0.14, or 14%, but that should be described as a relative hazard interpretation rather than as a 14-percentage-point reduction in the proportion of participants with an event.

An absolute risk difference would require the corresponding cumulative event probabilities or other absolute measures, which are not reported in the ClinicalTrials.gov record.

Why does the P-value not measure treatment effect size?

A P-value reflects the degree of incompatibility between the observed data and a specified null hypothesis under the statistical testing framework. It is affected by both the magnitude of the observed effect and the amount of information available.

For example, the primary P-value is 0.54, but that number does not tell us that the treatment effect is "54%" or that there is a 54% probability that the treatment is ineffective. The estimated difference and its confidence interval provide the direct quantitative description of the observed treatment comparison.

Why is early termination important for interpretation?

The registry states that the study was terminated early after a futility analysis concluded that zinpentraxin alfa was unlikely to meet its primary endpoint. Early stopping changes the amount of information accumulated and can affect the precision and interpretation of the observed treatment effect.

Most importantly, the termination statement is itself part of the documented trial record. The posted results should therefore be understood as results from a trial that ended early rather than as results from a fully completed trial with all originally planned information.

11. Intention-to-Treat and Analysis Population

The ClinicalTrials.gov record identifies the analysis population as the full analysis set. The registry description states that this set included all randomized participants who received at least one administration, full or partial, of study drug, with grouping according to treatment assignment. The analysis text also identifies intention-to-treat analysis as a concept used in the analysis.

Population featureRegistry-supported description
RandomizationParticipants were randomized.
Minimum treatment exposureAt least one administration, full or partial dose, of study drug.
Treatment groupingAccording to treatment assignment.
Analysis conceptIntention-to-treat analysis is identified in the analysis text.

The distinction matters because treatment assignment and treatment exposure answer different questions. An assignment-based analysis retains the randomized comparison, while requiring at least one administration introduces an exposure condition into the definition of the full analysis set.

12. Time-to-Event Analysis

Three of the four registry-reported formal statistical analyses are time-to-event analyses. Each uses a log-rank method and reports a hazard ratio.

General time-to-event framework
Survival function: S(t) = P(T > t)

A time-to-event analysis uses both whether an event occurred and when it occurred. Participants who have not experienced the event by the end of observed follow-up may contribute censored information rather than being treated as if an event occurred.

Log-rank testing

The log-rank test compares event-time distributions between treatment groups. It is particularly useful when participants have different lengths of follow-up and some observations are censored.

In STARSCAPE, the registry analyses use the log-rank method for time to disease progression, time to first respiratory-related hospitalization, and time to first acute exacerbation of IPF.

Hazard ratios

The reported hazard ratio provides a relative comparison of event rates. An HR below 1 favors a lower estimated event rate in the zinpentraxin alfa group; an HR above 1 favors a higher estimated event rate in that group.

EndpointHRDirection of point estimate
Time to Disease Progression1.15Higher estimated event rate with zinpentraxin alfa
Time to First Respiratory-related Hospitalizations1.01Very close to the comparator
Time to First Acute Exacerbation of IPF0.86Lower estimated event rate with zinpentraxin alfa

These directions describe the point estimates only. They do not establish that the treatment effects differ from zero on an absolute scale or from a hazard ratio of 1.00 on a statistical-testing scale. The confidence intervals must be considered alongside each estimate.

13. Confidence Intervals and Precision

The four formal analyses provide confidence intervals with substantially different widths. This illustrates why the point estimate alone is insufficient for interpreting clinical-trial results.

EndpointPoint estimate95% CIInterpretive feature
FVC change-20.83 mL-87.94 to 46.29 mLIncludes both negative and positive differences
Disease progressionHR 1.150.91–1.47Includes 1.00
Respiratory-related hospitalizationHR 1.010.51–1.97Wide interval; includes 1.00
Acute exacerbation of IPFHR 0.860.40–1.86Wide interval; includes 1.00

The width of a confidence interval reflects statistical precision. The hospitalization and acute-exacerbation hazard-ratio intervals are especially broad relative to their point estimates, indicating substantial uncertainty around the estimated relative hazards.

A confidence interval that crosses the null value should not be described as proving that there is no treatment effect. It indicates that the interval includes values consistent with the null as well as values on either side of it.

14. Superiority Testing

The analyses posted on ClinicalTrials.gov identify superiority as the hypothesis type. This is important because the statistical question is whether the treatment groups differ under a superiority framework, rather than whether the treatment is merely not worse than placebo by a prespecified non-inferiority margin.

Primary endpoint

The primary FVC analysis tested a superiority hypothesis using the reported difference in change from baseline.

Secondary endpoints

The three time-to-event analyses also identify superiority as the hypothesis type and use hazard ratios as effect measures.

No non-inferiority margin is reported in the ClinicalTrials.gov record. Accordingly, no non-inferiority interpretation is made on this page.

15. Multiplicity and Multiple Endpoints

The trial registry reports 1 primary endpoint and 16 posted outcome measures, with 4 statistical analyses reported in the ClinicalTrials.gov record. The ClinicalTrials.gov record does not describe an alpha-allocation strategy or a formal multiplicity adjustment across the primary and secondary endpoints.

Analysis roleEndpoint(s)What the ClinicalTrials.gov record establishes
PrimaryAbsolute Change in FVCOne formal superiority analysis is posted.
SecondaryTime to Disease ProgressionFormal log-rank analysis posted.
SecondaryTime to First Respiratory-related HospitalizationsFormal log-rank analysis posted.
SecondaryTime to First Acute Exacerbation of IPFFormal log-rank analysis posted.
Multiplicity caution: the ClinicalTrials.gov record does not report a formal multiplicity-adjustment strategy for the secondary analyses. Their P-values should therefore be interpreted according to the statistical information actually registry-reported rather than automatically treating all four analyses as independent confirmatory tests.

16. Futility and Early Stopping

The registry provides a specific reason for termination: the study was stopped early based on a futility analysis that concluded that zinpentraxin alfa was unlikely to meet its primary endpoint.

What a futility analysis is intended to do
Accumulating evidence → assessment of whether continuing is likely to achieve the prespecified objective

A futility assessment is a forward-looking decision tool. It is distinct from a conventional superiority test: its purpose is to determine whether continuing the trial is unlikely to produce the desired result under the information available at the interim assessment.

The ClinicalTrials.gov record does not report the futility boundary, conditional-power calculation, predictive probability, information fraction, or interim sample size. Those quantities are therefore not presented or reconstructed here.

The early termination also means that the observed estimates should not be interpreted as though the trial had necessarily accumulated the same amount of information as a fully completed study.

17. Safety Results

The ClinicalTrials.gov record reports serious adverse events by arm. The figures are presented as affected participants divided by participants at risk.

Safety measureZinpentraxin AlfaPlacebo
Serious adverse events, affected / at risk46 / 33140 / 329

Zinpentraxin Alfa

46 / 331 participants were affected by serious adverse events among those at risk in the ClinicalTrials.gov record.

Placebo

40 / 329 participants were affected by serious adverse events among those at risk in the ClinicalTrials.gov record.

These figures are reported as affected participants over participants at risk. The ClinicalTrials.gov record does not provide a formal statistical comparison, confidence interval, or P-value for serious adverse events, so none is inferred.

Safety interpretation: safety and efficacy address different statistical questions. A serious-adverse-event count describes an observed safety outcome among participants at risk; it does not by itself establish causality, relative risk, or an overall benefit-risk conclusion.

18. Missing Data and Censoring

The ClinicalTrials.gov record does not report a specific missing-data or imputation strategy for the primary FVC analysis. They also do not specify a detailed imputation procedure for the repeated FVC measurements used in the Random Coefficient Regression Model.

For the time-to-event endpoints, censoring is inherently part of survival analysis. Participants who have not experienced the event by the end of their observed follow-up can contribute information up to the point of censoring. The ClinicalTrials.gov record, however, do not provide enough detail to describe the exact censoring rules used for each endpoint.

Do not infer an imputation method. Because the ClinicalTrials.gov record does not identify a specific missing-data procedure, this page does not assume last observation carried forward, multiple imputation, mixed-model-based imputation, or any other particular strategy.

19. Proportional-Hazards Considerations

The secondary analyses report hazard ratios, which are commonly interpreted as relative event-rate comparisons over time. A single hazard ratio is most straightforward to interpret when the relative hazards are reasonably stable over the relevant follow-up.

The registry-reported STARSCAPE data do not provide a proportional-hazards diagnostic, a time-varying hazard-ratio analysis, or a Kaplan-Meier curve from which the assumption could be assessed. Therefore, no conclusion about whether the proportional-hazards assumption held is made here.

Hazard ratio is not a universal risk ratio
HR ≠ risk ratio ≠ absolute risk difference

These measures answer different questions. The STARSCAPE registry analyses specifically report hazard ratios for the time-to-event endpoints, so those estimates should not be converted into absolute risks without additional data.

20. What the Primary Result Does — and Does Not — Mean

Effect estimate

The primary estimate of -20.83 mL is the reported difference in change from baseline in FVC between zinpentraxin alfa and placebo. It is a group-level model-based estimate, not an individual prediction.

Confidence interval

The 95% CI of -87.94 to 46.29 mL quantifies uncertainty around the estimated difference under the reported statistical framework. It does not represent the range of FVC changes that individual participants experienced.

P-value

The P-value of 0.54 is a hypothesis-testing quantity. It does not mean there is a 54% probability that zinpentraxin alfa is ineffective, and it does not measure clinical importance.

Early stopping

The registry states that the trial was terminated after a futility analysis concluded that zinpentraxin alfa was unlikely to meet its primary endpoint. That documented design history is essential context for interpreting the posted primary result.

21. Interpreting the Secondary Hazard Ratios Together

The three secondary hazard ratios point in different directions at the level of their point estimates: 1.15 for disease progression, 1.01 for respiratory-related hospitalization, and 0.86 for acute exacerbation of IPF.

This pattern should not be compressed into a single overall conclusion. Each endpoint represents a different clinical event, and each has its own uncertainty. The confidence intervals for all three secondary hazard ratios include 1.00.

EndpointHR95% CIKey statistical point
Time to Disease Progression1.150.91–1.47Point estimate above 1; interval includes 1
Time to First Respiratory-related Hospitalizations1.010.51–1.97Point estimate near 1; interval is wide
Time to First Acute Exacerbation of IPF0.860.40–1.86Point estimate below 1; interval is wide

The correct statistical lesson is that treatment effects should be interpreted endpoint by endpoint, with the point estimate, confidence interval, testing framework, analysis population, censoring, and trial stopping history considered together.

22. Limitations

23. Why This Trial Matters Statistically

STARSCAPE is a useful teaching case because it combines a longitudinal continuous primary endpoint with multiple time-to-event secondary endpoints and an early futility termination. It illustrates why clinical-trial interpretation requires more than reading a single P-value.

ConceptHow it appears in STARSCAPE
RandomizationThe trial used randomized allocation to zinpentraxin alfa or placebo.
BlindingThe study was double-blind.
Parallel-group designThe trial used a parallel design with 2 arms.
Continuous endpointAbsolute change in FVC was the registered primary endpoint.
Repeated-measures modelingThe primary analysis used a Random Coefficient Regression Model.
Intention-to-treat conceptThe analysis text identifies intention-to-treat analysis.
Time-to-event analysisThree secondary endpoints were analyzed as time-to-event outcomes.
Log-rank testEach registry-reported secondary time-to-event analysis used a log-rank method.
Hazard ratioEach secondary time-to-event analysis reported an HR.
Confidence intervalsAll four formal analyses provide 95% confidence intervals.
Superiority testingThe analyses posted on ClinicalTrials.gov identify superiority as the hypothesis type.
Futility analysisThe study was terminated early after a futility analysis.
Safety analysisSerious adverse events are reported by treatment arm.

24. A Statistical Reading of the Trial

A useful way to read STARSCAPE is to move from the estimand to the uncertainty and then to the design context.

1. Start with the endpoint

The primary endpoint is absolute change in FVC from baseline through Week 52, not survival time or an event rate.

2. Identify the model

The registry reports a Random Coefficient Regression Model for the primary continuous endpoint.

3. Read the estimate

The reported difference in change from baseline is -20.83 mL for zinpentraxin alfa versus placebo.

4. Read the interval

The 95% CI extends from -87.94 to 46.29 mL, demonstrating uncertainty around the point estimate.

5. Read the P-value

The P-value is 0.54. It is a hypothesis-testing quantity, not a measure of treatment magnitude.

6. Add trial context

The study was terminated early after a futility analysis concluded that the primary endpoint was unlikely to be met.

25. Related Tutorials

Learn more about the methods used in this trial:

26. Related Statistical Calculators

27. Sources

28. Record Summary

STARSCAPE provides a useful example of how different endpoint types require different statistical approaches within the same randomized clinical trial. The registered primary endpoint was absolute change in FVC from baseline up to Week 52, analyzed with a Random Coefficient Regression Model. The reported treatment difference was -20.83 mL, with a 95% CI of -87.94 to 46.29 mL and a P-value of 0.54.

The registry-reported secondary analyses used log-rank methods for three time-to-event endpoints and reported hazard ratios of 1.15 for time to disease progression, 1.01 for time to first respiratory-related hospitalization, and 0.86 for time to first acute exacerbation of IPF. Their corresponding 95% confidence intervals were 0.91–1.47, 0.51–1.97, and 0.40–1.86, respectively.

The most important design consideration is the trial's early termination. ClinicalTrials.gov states that the study was terminated based on a futility analysis concluding that zinpentraxin alfa was unlikely to meet its primary endpoint. This means that interpretation should account not only for the numerical estimates and P-values, but also for the amount of information ultimately accumulated and the statistical decision to stop the trial.

Clinical Biostats methodology: A trial-results page should distinguish the reported statistical estimate from the interpretation placed around it. For STARSCAPE, that means reporting the FVC difference and its uncertainty, treating the secondary hazard ratios as time-to-event measures rather than simple risk ratios, and incorporating the documented early futility termination into the statistical interpretation without inventing unreported analyses.

Continue with the statistical methods behind STARSCAPE

Explore the core clinical-trial and survival-analysis concepts used to interpret randomized treatment effects, confidence intervals, and time-to-event endpoints.