A Study of Galunisertib in Participants With Myelodysplastic Syndromes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Galunisertib in Participants With Myelodysplastic Syndromes (NCT02008318), a non-randomized phase 2, phase 3 study. The primary outcome is Percentage of Participants With Hematological Improvement (HI).

2. Clinical Question

The clinical question is therefore whether HI was observed among eligible participants receiving galunisertib with BSC, with response defined by the stated criteria and duration. The record includes placebo plus BSC as an arm, but non-random allocation limits what can be inferred from contrasts between arms. Baseline differences or other factors related to treatment assignment could also relate to outcomes. The registry record does not report the primary endpoint estimate, its confidence interval, or an analysis comparing the arms. It therefore does not establish the frequency or precision of HI in the study.

3. Trial Design

This was a completed Phase 2/3, parallel-group study with non-randomized allocation and no masking. The registry reports enrollment of 43 participants and lists myelodysplastic syndromes as the condition. The study began in 2014-03, and primary completion was recorded as 2016-03.

The listed arms were Phase 2 galunisertib plus BSC, Phase 3 placebo plus BSC, and Phase 3 galunisertib plus BSC. The galunisertib descriptions specify 150 milligrams orally twice daily for 14 days, followed by 14 days without study drug, in 28-day cycles; treatment was expected to last for 6 cycles. The placebo description gives the same on-and-off schedule. Participants received BSC according to institutional guidelines. The registry says participants may receive additional cycles in some circumstances, but the full arm descriptions are truncated in the record.

Non-random allocation is a central design feature. Randomization is generally used to reduce systematic differences between treatment groups at baseline. Without it, differences between groups may reflect participant selection, clinical decisions, or other factors, in addition to any treatment-related differences. No masking means treatment assignment was not concealed from participants or study personnel as a design feature. These characteristics make the trial’s structure different from a blinded randomized comparison.

The registry record does not report the allocation procedure, the planned group sizes, or a statistical plan for comparing Phase 2 and Phase 3 participants. It also does not describe how treatment assignment related to disease risk or other participant characteristics. The record lists both galunisertib and placebo interventions, but their presence does not by itself establish a randomized control comparison. The study’s primary endpoint is described for participants receiving galunisertib plus BSC.

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: This is a single-arm study (Galunisertib at 150 milligram \[mg\]); the Galunisertib at 80 mg was considered exploratory and only conducted in parallel with the main study, at one site in Spain.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Participants who received at least one dose of study drug.

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: All participants (including the Galunisertib at 80 mg exploratory arm) who received at least one dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Hematological Improvement (HI), measured over Baseline through end of study treatment (24 weeks).

The primary endpoint is the percentage of participants with hematological improvement (HI), assessed from baseline through the end of study treatment (24 weeks). Response was based on International Working Group 2006 criteria, with participants classified as responders only if the response lasted at least 8 weeks (56 days). The endpoint description applies to participants with very low-, low-, and intermediate-risk myelodysplastic syndromes treated with galunisertib plus BSC, with risk assessed using IPSS-R. The stated analysis population is participants who received at least one dose of study drug.

A percentage endpoint expresses how often a defined outcome occurred within a specified population. Requiring a response to persist for a minimum duration distinguishes a sustained response from a shorter-lived change. That can make the endpoint more clinically interpretable than a measure based only on whether any improvement occurred at any point. The confidence interval is intended to show the uncertainty around the estimated percentage; a wider interval generally indicates less precision.

The registry record gives the endpoint definition and a 95% confidence interval as its dispersion measure, but it does not report the observed percentage or interval. It also does not give the endpoint’s component response counts. Consequently, the magnitude and precision of the primary endpoint cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are easiest to interpret when the outcome, target population, comparison, and analysis procedure are all explicit. The registry record describes the primary outcome and an analysis population, but leaves several other elements unreported. The following concepts are general explanations; they do not imply that a particular method was used in this study.

What does the primary percentage represent?

The primary outcome is a binary response classification: a participant either met the HI response definition or did not. A percentage summarizes the frequency of response among a specified set of participants. Its interpretation depends on the denominator: which participants were eligible, treated, assessed, and included. The registry identifies participants who received at least one dose as the primary population, but does not report the number in that population or the number with a recorded endpoint.

The response rule includes both criteria and duration. It uses International Working Group 2006 criteria and requires HI to last at least 8 weeks (56 days). This operational definition helps make the outcome reproducible, but it also means the percentage concerns only responses meeting those rules. It should not be generalized to different definitions of improvement without additional evidence.

What does a confidence interval add?

In general, a confidence interval expresses the statistical uncertainty around an estimate. For a percentage, the interval gives a range of values compatible with the observed data under the method’s assumptions. It is not a guarantee that the true value lies in that range, and it does not address every source of uncertainty, such as selection bias or measurement differences. The registry names a 95% confidence interval for the primary endpoint, but does not report its bounds. Therefore, the precision of the primary estimate cannot be evaluated.

Why does allocation matter?

In general, random allocation helps balance known and unknown baseline characteristics across treatment groups on average. Non-random allocation does not provide that protection. Groups may differ in prognosis, disease characteristics, or other factors, and those differences can influence outcome rates. Statistical adjustment can sometimes address measured differences, but it cannot automatically remove bias from unmeasured factors. The registry describes allocation as non-randomized and does not report an adjustment method or a comparative analysis plan.

Masking is also relevant. In general, masking can reduce the possibility that knowledge of assignment influences care, assessment, or reporting. The study is listed as having no masking. That is a design descriptor, not proof that outcome assessment was biased; it means the registry does not describe masking as part of the study design.

How should a comparison be framed?

A treatment comparison requires a clearly defined contrast and an analysis appropriate to the design. The registry lists galunisertib plus BSC and placebo plus BSC arms, but non-random allocation means observed differences between them could reflect more than treatment. The primary endpoint description specifically refers to galunisertib plus BSC; the registry record does not state a primary comparative estimand. It also does not report an effect estimate comparing the listed arms.

In general, a confidence interval around a between-group difference would describe uncertainty in that contrast, while the choice of adjustment and analysis population would affect its interpretation. Such methods cannot be presumed from the registry record. No adjusted or unadjusted comparative result is reported there.

What can event counts tell us?

For a binary outcome, counts and percentages are complementary: counts show the number of participants with the event, while percentages relate that count to a denominator. Interpreting a percentage requires knowing how the denominator was defined and whether participants with missing observations were excluded or handled another way. The registry provides safety denominators for the listed galunisertib groups, but does not give primary endpoint denominators or response counts.

General statistical practice also distinguishes participant-level outcomes from event-level outcomes. A participant may experience more than one event, so counts of events and counts of participants with events are not interchangeable. The adverse-event summary is described in terms of participants affected for the broad categories, while the listed serious-event terms are percentages. The registry does not explain whether listed terms are mutually exclusive or give further details about their denominators.

What does the record say about analysis planning?

The registry record does not report the statistical analysis plan, a sample-size justification, interim analyses, or the method used to construct the primary confidence interval. It also does not report how multiplicity or missing data were handled. In general, a planning calculation is not an observed result, and a sample-size illustration based on an observed effect would not be the trial’s original design justification. No such design justification is reported in the registry record.

Because the primary outcome estimate and analysis methods are absent, the reader cannot verify whether the prespecified endpoint definition was applied as intended, assess the estimate’s precision, or evaluate the effect of analytic choices. Those are limits on interpretation of the registry record, not evidence that any particular analysis was or was not conducted outside it.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Galunisertib at 150 mg", "Arm: Galunisertib at 80 mg")
n      <- c(41, 2)
events <- c(13, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Galunisertib at 150 mg  41  13  31.7  18.1  48.1  19.6  47.0  19.6  47.1
ARM  Arm: Galunisertib at 80 mg  2  0  0.0  0.0  84.2  0.0  65.8  0.8  70.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists the primary HI endpoint and 9 outcome measures, including secondary outcomes concerning fatigue, quality of life, cytogenetic response, hospitalization, population pharmacokinetics, overall survival, and bone marrow fibrosis grading. The registry record does not report a multiplicity strategy, such as a hierarchy for testing outcomes or an adjustment for multiple comparisons. Without such information, the intended interpretation of any collection of statistical tests cannot be assessed. No outcome analyses are listed in the record.

Missing data handling is also not reported in the registry record. The primary endpoint population is described as participants who received at least one dose of study drug, but the registry does not give the number assessed for HI, the number with missing assessments, or the rule used for incomplete follow-up. These details matter because the denominator and assumptions about unobserved outcomes can influence a percentage estimate. In general, excluding participants with missing data can produce a different result from treating missing outcomes as nonresponse or using a model-based method; no approach can be attributed to this study from the registry record.

The record does not state whether analyses followed an intention-to-treat principle, a treated-population approach beyond the endpoint population definition, or another analysis population. It does not report protocol deviations or the extent of follow-up for the primary endpoint. The absence of these details prevents evaluation of how participants who discontinued treatment or lacked endpoint data may have affected the reported result. Since the primary result itself is not reported, neither its sensitivity to missingness nor the treatment of multiple outcomes can be examined.

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: cardiac failure (4.9%), respiratory failure (4.9%), febrile neutropenia (2.4%), lymphadenopathy (2.4%), crohn's disease (2.4%).

The safety summary lists two galunisertib groups. For Galunisertib at 150 mg, 41 participants were at risk for the serious-event summary and 8 were affected; 41 were at risk for the other-event summary and 33 were affected. For Galunisertib at 80 mg, 2 participants were at risk for the serious-event summary and 0 were affected; 2 were at risk for the other-event summary and 2 were affected. Deaths are recorded as null for both groups, so death counts are not reported in the registry record.

The listed serious-event terms and percentages are Cardiac failure, 4.9%; Respiratory failure, 4.9%; Febrile neutropenia, 2.4%; Lymphadenopathy, 2.4%; Crohn's disease, 2.4%; and Retroperitoneal haemorrhage, 2.4%. These are registry-listed safety observations. The record does not provide further details about the terms’ denominators, whether a participant could appear under more than one term, or whether events were considered related to study treatment.

In general, safety counts describe reported events in the observed participants and do not by themselves establish causation. The small listed group at 80 mg also means its counts should be read in the context of the group size, rather than as a stable estimate of event frequency. The safety summary does not provide a randomized comparison.

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

The registry record establishes the study’s primary endpoint definition: HI under International Working Group 2006 criteria, lasting at least 8 weeks (56 days), assessed through the end of study treatment (24 weeks). It also identifies the population as participants with very low-, low-, or intermediate-risk myelodysplastic syndromes who received at least one dose of study drug. Those details clarify what a reported primary percentage would mean.

The registry record does not report that percentage, its 95% confidence interval, or an HI responder count. It therefore does not show how frequently participants met the endpoint, how precise an estimate was, or whether results differed between groups. The listing of an endpoint is not itself evidence that participants achieved it, and the lack of an outcome estimate is not evidence that no response occurred.

The design further limits comparative interpretation. Allocation was non-randomized and masking was absent. Although galunisertib plus BSC and placebo plus BSC arms are listed, any apparent difference between them could be associated with differences in who entered each arm or how care and assessments occurred. The registry does not report a primary comparative analysis or methods to address such differences. Accordingly, the arm listings alone do not establish a causal treatment effect.

The safety entries describe affected participants and selected serious-event terms in the galunisertib groups. They should be read as descriptive observations, not as proof that treatment caused or prevented those events. Death counts are not reported in the registry record.

In general, a result’s meaning depends on its endpoint definition, population, uncertainty, missing-data handling, and design. Here, endpoint and design information are available, while the primary numerical result and several analysis details are not. The record supports a description of what the study set out to assess and what safety counts it lists, but it does not support a quantitative conclusion about primary efficacy or a causal comparison between arms.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of why endpoint definitions and study design must be read together. The primary outcome is not simply any reported improvement: it applies International Working Group 2006 criteria and requires the response to last at least 8 weeks (56 days). That duration rule shapes the event being counted and determines what a primary percentage would represent.

The non-randomized, unmasked, parallel design also matters. The study includes both galunisertib plus BSC and placebo plus BSC arms, but allocation was not randomized. In general, arm comparisons in a non-randomized study may be affected by baseline differences or treatment-selection processes. The existence of a placebo arm does not remove that limitation or make the contrast equivalent to a randomized comparison.

The registry record illustrates a further distinction between a clearly stated outcome and a reported result. It specifies the primary endpoint and names a 95% confidence interval as the dispersion measure, yet it does not report the percentage, interval, or responder count. Without those values, readers cannot assess either the magnitude or statistical precision of the endpoint.

The record also makes the safety denominator visible for two galunisertib dose groups and lists selected serious-event terms. That information can be described, but its interpretation remains constrained by the absence of detail about term denominators, overlap, and attribution. Overall, the study record is useful for understanding the design, outcome definition, and limits of the available analysis information; it is not sufficient for estimating an HI rate or a treatment effect.

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.

Several limitations specific to this registry record constrain interpretation. First, it reports no primary endpoint estimate, confidence interval, or responder count, despite describing the endpoint and its assessment period. The record lists 0 analyses, so the primary result cannot be evaluated from the registry entry.

Second, allocation was non-randomized and the study was unmasked. Any arm-level outcome contrast could reflect differences in participant characteristics or care, as well as treatment assignment. The registry record does not describe the allocation procedure, adjustment methods, or a primary comparative estimand. The placebo arm therefore should not be read as establishing a randomized control comparison.

Third, the record does not provide the primary analysis denominator or describe missing-data handling. It defines the primary population as participants who received at least one dose, but does not state how many met that definition or how incomplete assessments were handled. These omissions prevent assessment of the outcome’s denominator and potential sensitivity to missing information.

Fourth, the registry record does not report a statistical analysis plan, sample-size justification, interim analyses, or multiplicity strategy. The listed secondary outcomes cover several kinds of measures, but the record does not explain how their analyses were prioritized or interpreted together.

Finally, the arm descriptions are truncated in the registry record, and the safety summary gives limited detail for the listed serious-event terms. Death entries are null, and further information on event attribution and term denominators is not reported. These gaps limit both efficacy and safety interpretation without establishing that the study lacked further information outside the registry.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Galunisertib in Participants With Myelodysplastic Syndromes — NCT02008318.
• Registry record last updated 2019-09-11; status last verified 2019-08.
• Results first posted 2019-03-07.
• 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

NCT02008318 is a completed, non-randomized, unmasked Phase 2/3 parallel study of galunisertib in participants with myelodysplastic syndromes. The registry reports enrollment of 43, a start date of 2014-03, and primary completion in 2016-03. Listed arms include galunisertib plus BSC and placebo plus BSC.

The primary endpoint was the percentage of participants with HI under International Working Group 2006 criteria, with response lasting at least 8 weeks (56 days), assessed through the end of treatment (24 weeks). The registry record does not report the primary endpoint result, its confidence interval, or an analysis comparing arms.

Safety entries include affected-participant counts for galunisertib at 150 mg and 80 mg and selected serious-event terms. Death counts are not reported in the registry record. Statistical planning, missing-data handling, multiplicity methods, and a sample-size justification are also not reported in the registry record.