Study to Evaluate the Safety and Efficacy of Luspatercept (ACE-536) in Participants With Beta-thalassemia (A536-04/MK-6143-002): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study to Evaluate the Safety and Efficacy of Luspatercept (ACE-536) in Participants With Beta-thalassemia (A536-04/MK-6143-002) (NCT01749540), a non-randomized phase 2 study. The primary outcome is Percentage of Non-transfusion Dependent (NTD) Participants With a Hemoglobin Increase of ≥1.5 g/dL From Baseline for ≥14 Days.

2. Clinical Question

The registry describes a broader primary objective as evaluating erythroid response in both non-transfusion-dependent and transfusion-dependent participants. For the former, response was based on a sustained hemoglobin increase; for the latter, the stated response definition was a reduction in red blood cell transfusion burden compared with pretreatment. The registered primary outcome analyzed in this record is specifically the hemoglobin response among non-transfusion-dependent participants.

This distinction matters when interpreting the question. The endpoint does not describe every participant’s response, and it is not a comparison with a concurrent control group. It is a threshold-based measure in a defined subgroup. Without the reported estimate and its uncertainty, the registry record does not establish how often participants met that threshold.

3. Trial Design

This was a Phase 2, open-label, ascending-dose study of luspatercept in participants with beta-thalassemia. The allocation was nonrandomized, masking was absent, and the registered model was parallel. The registry lists an enrollment of 64. The study started on 2013-02-28, reached primary completion on 2015-11-11, and is recorded as completed.

The dose cohorts were 0.2 mg/kg, 0.4 mg/kg, 0.6 mg/kg, 0.8 mg/kg, 1.0 mg/kg, and 1.25 mg/kg. In these cohorts, luspatercept was given as a subcutaneous injection every 3 weeks on day 1 of each cycle, for up to 5 cycles. The expansion cohort began at 0.8 mg/kg; for subsequent cycles, the dose could be titrated up to a maximum of 1.25 mg/kg based on safety review team recommendations.

Because participants were not randomized, cohort membership should not be interpreted as if it were assigned by chance. Dose cohorts in an ascending-dose study may differ in their composition or the circumstances under which participants entered them. The registry record does not report enough information to determine whether or how such differences were addressed in an analysis. In particular, comparisons across dose groups cannot by themselves isolate a dose effect.

There was no masked assessment, and no placebo or untreated arm is listed. Knowledge of treatment assignment may affect reporting, clinical decisions, or assessment, depending on the outcome and study procedures. The record does not describe measures used to limit those influences. The sponsor is listed as Acceleron Pharma, Inc., a wholly-owned subsidiary of Merck & Co., Inc., Rahway, NJ USA.

The registry lists 58 outcome measures. Although it describes the primary endpoint and several safety outcomes, the endpoint results and detailed analysis results are not reported in the registry record. A sample-size justification and the study’s own statistical analysis plan are not reported in the registry record.

4. Participant Flow and Accounting

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

Only the first 6 of 7 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants who received ≥1 dose of study treatment and received \<4 units of RBCs within 8 weeks prior to the first dose of study drug and had data available for the analyses

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 randomized participants who received ≥1 dose of study treatment

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Non-transfusion Dependent (NTD) Participants With a Hemoglobin Increase of ≥1.5 g/dL From Baseline for ≥14 Days, measured over Up to approximately 20 weeks.

The registered primary endpoint is the percentage of non-transfusion-dependent participants whose hemoglobin increased by at least 1.5 g/dL from baseline for at least 14 days, without a blood transfusion. It is a binary response measure at the participant level: each eligible participant either meets the stated response definition or does not. The reported percentage would describe the proportion meeting that definition in the analyzed population.

The endpoint uses a clinical threshold and a duration requirement rather than treating every change in hemoglobin as a response. Requiring the increase to persist helps distinguish a sustained change from a brief measurement fluctuation. Excluding measurements after transfusion is also relevant because transfusion can affect hemoglobin values independently of the study drug.

For non-transfusion-dependent participants, baseline hemoglobin was defined as the average of two or more measurements during screening. Measurements within 2 weeks after a red blood cell transfusion or 56 days after the last dose were excluded from the analysis. The population definition required fewer than 4 units of red blood cells in the 8 weeks before the first dose and data available for analysis.

The registry gives a timeframe of up to approximately 20 weeks. It does not report the percentage meeting the endpoint, the number analyzed for this endpoint, or a confidence interval. Thus, the endpoint definition explains what was intended to be measured, but does not establish the observed response rate or its precision.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a percentage of participants meeting a binary response definition. Understanding the design and the endpoint helps clarify what a suitable analysis would describe, but the result itself is not reported in the registry record. The explanations below are general statistical context, not a claim about analyses performed in this study.

What does a percentage measure?

A percentage for a binary endpoint represents the share of analyzed participants who meet the specified response criterion. The denominator is the number of participants included in that outcome analysis, not necessarily everyone enrolled or treated. In this study, the outcome population is restricted to non-transfusion-dependent participants who had the required data. The endpoint-specific denominator is not reported in the registry record, so neither the percentage nor its sampling uncertainty can be reconstructed.

A percentage is easy to interpret, but it does not communicate precision on its own. The same observed percentage can be more uncertain when based on fewer observations than when based on more observations. A confidence interval expresses a range of values reasonably compatible with the data under the statistical method used. It is not a guarantee that the true value lies in that particular interval, and it does not account automatically for bias, design limitations, or missing data. Although the outcome description names a 95% confidence interval, the interval itself is not reported in the registry record.

Why define a threshold and duration?

A threshold turns a continuous measurement, such as hemoglobin, into a response classification. Here, the response definition requires a specified increase from baseline that lasts for a specified period and occurs without transfusion. Baseline is defined using the average of two or more screening measurements. Using repeated baseline measurements can reduce dependence on a single value, while a duration requirement seeks to distinguish a sustained change from a temporary fluctuation. These are general reasons for such endpoint choices; they do not show how participants’ results were distributed in this trial.

The registry also excludes certain measurements after transfusion and after the last dose. Such rules define which observations can contribute to the outcome. They can affect the analysis population and the available follow-up data. The registry record does not describe the number of measurements excluded or the effect of those exclusions on the result.

What does nonrandomized allocation imply?

Random assignment is a general method for balancing measured and unmeasured participant characteristics across groups, on average. This study used nonrandomized allocation, and the registry lists multiple dose cohorts rather than a control group. As a result, an observed difference across cohorts, if one were reported, could be associated with dose, but could also reflect differences in participant characteristics, timing, or other factors. Statistical adjustment may address measured differences under appropriate assumptions, but cannot guarantee removal of unmeasured confounding. The registry record does not report an adjusted dose comparison.

Open-label treatment means masking was absent. In general, knowledge of treatment can influence subjective reporting or outcome assessment and may also affect care or decisions during follow-up. The hemoglobin endpoint is based on laboratory measurements and prespecified thresholds, but the registry does not describe assessment procedures or safeguards against bias. No claim about the presence or size of bias can be made from the record alone.

What do planning calculations mean?

In general, sample-size planning uses assumptions about the outcome, the effect worth detecting, the significance level, the desired power, and the allocation between groups. The registry record does not report the trial’s own sample-size justification or statistical plan. Any sample-size values accompanying the posted results are planning illustrations based on assumptions, not observed power. Observed power is not a substitute for an effect estimate and confidence interval, and planning calculations do not show what the trial actually found.

The planning assumptions described for the posted calculations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the target effect is expressed as a fraction of the observed effect. Those assumptions are for illustrative planning, not evidence that the study used that design or analysis. This registry record does not report the observed primary endpoint effect needed to evaluate such an illustration.

Why does the absence of results matter?

A method cannot supply an outcome result that is not reported. Without the event count, endpoint denominator, estimated percentage, and confidence interval, readers cannot assess the magnitude or precision of the primary response. Nor can they evaluate whether an apparent difference among dose cohorts would be robust to confounding, missing outcome data, or alternative analysis decisions. The registry record describes the planned measurement, while leaving these result-level questions unanswered.

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("Luspatercept 0.2 mg/kg", "Luspatercept 0.4 mg/kg", "Luspatercept 0.6 mg/kg", "Luspatercept 0.8 mg/kg", "Luspatercept 1.0 mg/kg", "Luspatercept 1.25 mg/kg")
n      <- c(6, 6, 5, 3, 2, 1)
events <- c(0, 0, 0, 2, 1, 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  Luspatercept 0.2 mg/kg  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Luspatercept 0.4 mg/kg  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Luspatercept 0.6 mg/kg  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Luspatercept 0.8 mg/kg  3  2  66.7  9.4  99.2  20.8  93.9  19.4  93.2
ARM  Luspatercept 1.0 mg/kg  2  1  50.0  1.3  98.7  9.5  90.5  9.4  90.6
ARM  Luspatercept 1.25 mg/kg  1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 58 outcome measures and zero analysis entries. It does not report a multiplicity procedure, a testing hierarchy, or a primary-versus-secondary testing strategy beyond identifying the primary outcome. When a study evaluates many outcomes, repeated statistical tests can increase the chance of at least one apparently notable finding by chance. Methods such as a prespecified hierarchy or adjustment can address that issue, but whether any such method was planned or used is not reported in the registry record.

The record does not report how missing outcome data were handled, whether any participants lacked endpoint measurements, or how missingness was assessed. The primary outcome population is defined as participants who met the non-transfusion-dependent criterion and had data available for analysis. Restricting analysis to participants with available data can be informative, but its interpretation depends on why data are missing and whether missingness relates to response. No imputation method or sensitivity analysis is reported in the registry record.

Some measurement exclusions are specified: hemoglobin measurements within 2 weeks after a red blood cell transfusion or 56 days after the last dose were excluded. That rule does not, by itself, explain the handling of other missing observations or establish whether all eligible participants contributed to the endpoint analysis.

The registry record also does not report an endpoint-specific analysis population count, a protocol-deviation approach, or a detailed statistical analysis plan. Accordingly, readers cannot verify from this record how the analysis population was selected, whether multiplicity was controlled, or how conclusions might change under different assumptions about missing data.

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: bone marrow failure (16.7%).

The safety record lists serious adverse events and other adverse events by luspatercept group. For serious adverse events, the counts were 0 among 6 participants in the 0.2 mg/kg group, 1 among 6 in the 0.4 mg/kg group, 0 among 6 in the 0.6 mg/kg group, 0 among 6 in the 0.8 mg/kg group, 0 among 6 in the 1.0 mg/kg group, 0 among 5 in the 1.25 mg/kg group, and 0 among 29 in the expansion cohort. The record lists 0 deaths in each group.

For other adverse events, the recorded counts were 5 among 6 participants in the 0.2 mg/kg group, 6 among 6 in the 0.4 mg/kg group, 5 among 6 in the 0.6 mg/kg group, 6 among 6 in the 0.8 mg/kg group, 6 among 6 in the 1.0 mg/kg group, 5 among 5 in the 1.25 mg/kg group, and 25 among 29 in the expansion cohort. These are descriptive counts; they do not establish that an event was caused by luspatercept.

Bone marrow failure is listed as the top serious-event term at 16.7%. The registry record does not give further context for interpreting that percentage in this summary. Small group sizes and the absence of a control arm limit comparisons between groups. The counts should not be treated as precise estimates of risk or as proof that a dose is safer or less safe than another.

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

The study record establishes that luspatercept was evaluated in a Phase 2, open-label, nonrandomized, ascending-dose study in participants with beta-thalassemia. It defines a primary hemoglobin-response endpoint for non-transfusion-dependent participants and gives the criteria for classifying a response. It also records adverse-event counts by treatment group.

What the record does not establish is the observed primary response rate. It does not report the number of eligible participants included in that analysis, the percentage meeting the response definition, or a confidence interval. Without those results, the magnitude and precision of the endpoint cannot be judged. The absence of a reported result should not be interpreted as a result of zero responses or as evidence that the endpoint was not assessed.

The design also does not support a randomized comparison of doses or a comparison with no treatment. Participants were not randomly allocated, treatment was not masked, and no control arm is listed. A difference between dose cohorts would not, on its own, show that dose caused the difference. Such an interpretation would depend on the comparability of the cohorts and the analysis methods, neither of which is sufficiently described in the registry record.

The safety counts document events recorded within the listed groups, but they do not prove causation or give a controlled estimate of adverse-event risk. The small group denominators also mean that each recorded event can have a substantial influence on the descriptive group figures. The record does not provide enough information to determine the full context of event assessment or follow-up.

Overall, the registry provides a study design, endpoint definitions, and selected safety counts, but not the efficacy results needed to answer the primary question. The statistical interpretation must therefore remain limited to what was registered and recorded, without drawing conclusions about treatment benefit or lack of benefit.

14. Why This Trial Matters Statistically

This study illustrates why an endpoint definition and an endpoint result are not interchangeable. The registry defines a sustained hemoglobin increase, specifies how baseline is formed, and states which measurements are excluded. Those details make the intended outcome more interpretable. However, without the number analyzed, the response percentage, and its confidence interval, readers cannot judge how often the threshold was met or how precisely the response rate was estimated.

The study also demonstrates the importance of design when interpreting dose groups. An ascending-dose structure can support evaluation across dose levels, but nonrandomized allocation means group differences may reflect more than dose. A control arm and random allocation would address different questions and reduce some risks of confounding; neither is listed in this registry record. Statistical methods cannot fully repair design limitations when relevant differences are unmeasured or unreported.

Safety summaries require similar care. Event counts are useful for describing what the registry records, but they are not causal estimates. A count in a small cohort can change substantially with an additional event, and a comparison across groups may be misleading when allocation is nonrandomized. Reporting denominators alongside counts helps readers see the scale of the data, but does not remove uncertainty or bias.

Finally, the record lists many outcome measures and no analysis entries. This makes reporting of analysis populations, missing-data decisions, multiplicity handling, and confidence intervals especially relevant. Their absence limits independent evaluation of the findings. The statistical importance of this study record is therefore not a conclusion about efficacy, but a clear example of how transparent results and methods are needed to connect a clinical question to a defensible estimate.

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 primary limitation for interpreting efficacy is that the registry record does not report the primary endpoint result. The analyzed population size, response count, response percentage, and confidence interval are also not reported. Consequently, neither the observed response frequency nor its precision can be assessed from this record.

The endpoint applies only to non-transfusion-dependent participants meeting the stated red blood cell use criterion and having data available. The number of participants meeting those conditions is not reported in the registry record. The overall enrollment should not be taken as the endpoint denominator, and the safety group counts do not supply that missing information.

Nonrandomized allocation, open-label treatment, and the absence of a listed control group limit causal interpretation. Comparisons between dose cohorts could be affected by differences in participants or study circumstances. The record does not report an adjusted analysis or enough detail to evaluate whether such differences were considered.

The record does not provide the study’s own statistical analysis plan, sample-size justification, missing-data methods, or multiplicity approach. It also does not explain whether any endpoint analyses were altered from their original plans. These gaps prevent assessment of the prespecification and robustness of the analysis.

Safety counts are descriptive, and the group sizes are limited. The registry record does not give a controlled comparison of event risk or establish that events were caused by treatment. Its listed safety figures therefore should not be generalized beyond the participants and follow-up recorded in the study.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study to Evaluate the Safety and Efficacy of Luspatercept (ACE-536) in Participants With Beta-thalassemia (A536-04/MK-6143-002) — NCT01749540.
• Registry record last updated 2024-07-18; status last verified 2024-02.
• Results first posted 2024-07-18.
• 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

NCT01749540 is a completed Phase 2, open-label, nonrandomized, parallel, ascending-dose study of luspatercept in participants with beta-thalassemia. The registry records enrollment of 64 and lists six dose cohorts plus an expansion cohort. Treatment was administered by subcutaneous injection, with scheduled dosing every 3 weeks in the dose cohorts.

The primary outcome is the percentage of non-transfusion-dependent participants with a hemoglobin increase of at least 1.5 g/dL from baseline for at least 14 days without transfusion, assessed over up to approximately 20 weeks. The registry record describes the endpoint but does not report its result or an endpoint-specific analysis population.

Serious adverse-event, other adverse-event, and death counts are recorded by group. These are descriptive and do not provide a controlled estimate of treatment effects or risks. The record does not report detailed analysis methods, missing-data handling, multiplicity procedures, or the study’s own sample-size justification.