Efficacy and Safety of Eltrombopag In Patients With Severe and Very Severe Aplastic Anemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Eltrombopag In Patients With Severe and Very Severe Aplastic Anemia (NCT01703169), a na phase 2 study. The primary outcome is Proportion of Participants With Platelet Response.

2. Clinical Question

The investigators also described hypotheses about platelet counts, transfusions, bleeding events, toxicity, and the relationship between eltrombopag levels in peripheral blood and platelet counts. These are hypotheses in the study summary, not reported findings. The registry lists secondary outcomes covering platelet count, hematology laboratory measures, adverse events, and eltrombopag pharmacokinetics.

Because the study has no control group, its main clinical question is descriptive: what platelet response was observed among participants receiving the study intervention under the study conditions? It cannot, by itself, establish how that response compares with what would have happened without eltrombopag or with another treatment. The registry record does not report the primary endpoint result, so the response proportion and its uncertainty cannot be described from this record.

3. Trial Design

This was a completed phase 2 study sponsored by the University of Utah. The registry records a start in 2012-11 and primary completion in 2016-06. Enrollment was 13. The study model is listed as single group, allocation as not applicable, and masking as none. The arm description calls it a single-arm study with dose escalation.

All participants were assigned to the eltrombopag group in the registry description; there is no like-for-like control arm. This matters for interpretation: results can describe outcomes observed during the study, but changes over time cannot be cleanly separated from other influences using a concurrent randomized comparison. Neither the participant-level dosing details nor the escalation rules are reported in the registry record.

The primary outcome was the proportion of participants with a platelet response, assessed up to 12 weeks. The response definition required a stable platelet count of 50,000/μl or more during any four-week period within the possible 12 weeks while on study. The registry description also refers to maximal platelet counts achieved in participants with moderate to very severe aplastic anemia.

Listed secondary outcomes include platelet count twice baseline, hematology laboratory measures, the number of participants with adverse events for toxicity assessment using NCI CTCAE, and characterization of the pharmacokinetic profile using AUC, Cmax, Cmin, and Tmax. The record lists five outcome measures and one analysis, but the detailed analysis plan is not reported in the registry record. The design information does not establish how participants were selected for particular doses or how outcomes were handled across dose levels.

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.

Overall Study

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 Proportion of Participants With Platelet Response, measured over up to 12 weeks.

The primary endpoint is a proportion: the number of participants meeting the platelet-response definition relative to the participants assessed. A response required a stable platelet count of 50,000/μl or more during any four-week period within up to 12 weeks while on study. The description additionally says that maximal platelet counts achieved in participants with moderate to very severe aplastic anemia were included. The registry record does not clarify further how this wording was applied in analysis.

A proportion is useful for expressing how often a defined event or response occurred in a study group. Here, it turns a platelet-count criterion over a specified observation period into a participant-level outcome. The accompanying 95% confidence interval is intended to communicate uncertainty around the estimated proportion. In general, a confidence interval describes a range of values compatible with the observed data under the analysis method; it is not a guarantee that the true value lies within that range.

The record gives the endpoint definition and confidence-interval type, but does not report the observed number of responders, the analyzed denominator, the estimated proportion, or its interval. Therefore, the response rate and its precision cannot be interpreted from this registry record. The endpoint also does not, on its own, describe transfusions, bleeding, toxicity, or pharmacokinetic outcomes; those appear separately among the study’s stated outcomes and hypotheses.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary participant-level response: each participant either meets the stated platelet-response definition or does not. The registry specifies that the response is a stable platelet count of 50,000/μl or more during any four-week period within up to 12 weeks while on study. It also refers to maximal platelet counts achieved. The record does not provide further operational detail about stability, measurement timing, or how the maximal-count wording was applied.

What a proportion describes

In general, a proportion summarizes the share of assessed participants who meet a definition. It is a descriptive summary of the study group, not a direct comparison with a different treatment. To interpret a reported proportion, a reader needs both the number meeting the definition and the number assessed. The registry record does not report either for the primary outcome, nor does it report the resulting estimate.

Why the confidence interval matters

The primary outcome lists a 95% confidence interval. In general, a confidence interval conveys the uncertainty associated with an estimate under a particular statistical method. A narrow interval indicates greater precision than a wide interval, while neither width nor coverage can be determined without the reported estimate and interval. The statistical method used to construct the interval is not reported in the registry record. The interval type alone does not reveal the analysis details.

What the single-group design permits

The study is recorded as single group, with no masking and no allocation to a comparator. In general, a single-group study can document outcomes among participants receiving an intervention, but it cannot make the same direct treatment comparison as a study with a concurrent control group. Without a control group, observed changes cannot be separated from other influences by a randomized between-group comparison. The registry record does not report a comparative efficacy analysis.

The arm description says dose escalation, but the record does not report dose levels, dose assignment rules, or analyses by dose. It would therefore be inappropriate to infer which dose was received by any participant or how dose was incorporated into the primary analysis. The study’s one listed analysis does not, by itself, describe the statistical model or analysis population.

Interpreting precision in a small study

Enrollment was 13. In general, with a small study group, a small change in the number of participants meeting an endpoint can materially affect a proportion. Confidence intervals help show the resulting uncertainty, but the actual precision cannot be assessed without the posted estimate and interval. The record does not report how many participants had sufficient follow-up for the primary outcome or how missing observations were treated.

The response definition allows a qualifying period within an observation window of up to 12 weeks. In general, outcome definitions that consider whether a threshold was reached at any qualifying time can depend on the timing and availability of measurements. The registry description does not give the measurement schedule or explain how incomplete observation was addressed. Those details would be important for reproducing the analysis.

Planning illustrations are not trial results

The sample-size section presents planning illustrations based on an observed effect, not the study’s own design. The stated assumptions include a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. These are illustration assumptions, not evidence that the trial was planned as a two-arm study. The actual study was single group.

In general, a sample-size calculation describes how many participants a hypothetical design might require under specified assumptions. It does not show the probability that this completed study detected an effect, and it is not observed power. The registry record does not report the trial’s sample-size justification. Any planning illustration should therefore be read separately from the observed study design and from the missing primary endpoint result.

The record does not identify the statistical test, model, confidence-interval construction method, covariate adjustment, or handling of repeated measurements. These methods should not be inferred from the endpoint type or from the fact that a confidence interval was planned. Without the numerical result and analysis details, the posted registry information supports a description of the question and design, but not an independent reconstruction of the primary analysis.

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("Eltrombopag")
n      <- c(13)
events <- c(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  Eltrombopag  13  0  0.0  0.0  24.7  0.0  22.8  0.2  23.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists five outcome measures, including the primary platelet-response outcome and several secondary measures. It also records one analysis. The detailed statistical analysis plan, including any prespecified hierarchy or adjustment for testing multiple outcomes, is not reported in the registry record. The number of listed outcomes alone does not establish whether or how multiplicity was addressed.

In general, examining multiple outcomes or multiple analyses can increase the chance of finding an apparently notable result by chance if no appropriate strategy is used. A prespecified outcome hierarchy or multiplicity adjustment can help control that issue. The available record does not establish that such a strategy was used or omitted, so no claim about its presence should be made.

The analysis population is not described in the registry record. It does not say whether all enrolled participants were included in the primary analysis, whether any participants were excluded, or whether the analysis was based on participants with available measurements. The primary endpoint denominator and responder count are also not reported.

Missing-data handling is not reported in the registry record. There is no description of how absent platelet measurements, incomplete follow-up, withdrawal, or other unavailable outcomes were treated. In general, this matters because the interpretation of a proportion depends on which participants are included and how unknown outcomes are handled. With an endpoint assessed over up to 12 weeks, information about follow-up completeness and measurement availability would be useful, but it cannot be inferred from the stated timeframe.

Accordingly, neither the effect of multiplicity nor the influence of missing observations can be assessed from the registry information. These are reporting limitations, not proof that the original investigators did not consider the issues.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: congestive heart failure (7.7%), cholecystitis (7.7%), syncope (7.7%).

The posted safety figures list 13 participants at risk for serious adverse events, with 3 affected. The figures also list 13 at risk for other adverse events, with 13 affected, and 2 deaths. These are descriptive counts in the eltrombopag group; they are not comparisons against a control group. The registry does not report enough detail to determine whether events were judged related to treatment.

The three leading serious adverse-event terms shown are congestive heart failure, cholecystitis, and syncope, each reported as 7.7%. The record does not provide event-level detail or explain whether participants could contribute more than one term, so these percentages should not be added or treated as mutually exclusive categories. Nor should the listed terms be interpreted as a complete account of all safety findings.

The death count is listed without causes, timing, or attribution in this summary. The registry record also does not explain how adverse events were collected or how the toxicity outcome was analyzed. The study summary’s hypothesis about an acceptable toxicity rate <3% is a proposed expectation, not an observed safety result. These figures describe what is posted; they do not establish causality or comparative risk.

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

This study addresses a clinically relevant measurement question: how often participants met a defined platelet-response threshold during a specified period while receiving eltrombopag. The registry provides the response definition, the single-group design, and several listed secondary outcomes. It also provides some adverse-event counts. These details help readers understand what the investigators intended to assess.

However, the primary numerical outcome is not reported in the registry record. Without the responder count, denominator, estimated proportion, and confidence interval, readers cannot determine the observed response frequency or how precise its estimate was. The investigators’ stated hypotheses about platelet counts, transfusions, bleeding, toxicity, and drug levels should not be mistaken for evidence that those outcomes occurred as hypothesized.

The single-group design also limits causal interpretation. Even if a change in platelet counts were reported, a study without a concurrent control group would not distinguish the intervention’s contribution from other possible influences in the way a controlled comparison can. The record does not present a randomized comparison. Therefore, it does not support a comparative estimate against another treatment or no treatment.

The safety figures provide a descriptive account of recorded events, including serious adverse events and deaths, but do not specify causes or treatment attribution in the information. They should not be used to infer that eltrombopag caused or did not cause an event. The proposed toxicity threshold in the study summary is a hypothesis, not the reported toxicity rate.

In general, a registry record is useful for identifying a study’s design and planned outcome definitions, while interpretation of results depends on the actual outcome data and analysis details. Here, the record allows a careful account of the question asked and the limitations of the available information; it does not provide the primary result needed to quantify platelet response.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of how design and reporting determine what can be learned from a clinical study. Its primary endpoint is defined as a participant-level proportion over a bounded observation period. The registry also specifies a 95% confidence interval as the measure of dispersion, signaling that precision is relevant to interpretation. Yet the numerical endpoint result and interval are not reported in the registry record.

The single-group design makes the distinction between description and comparison especially important. A single-arm study can show what outcomes were recorded among participants receiving eltrombopag, but it does not supply a concurrent counterfactual group. Readers should not treat a within-study outcome as a controlled estimate of treatment effect. This is a general statistical distinction, not a judgment about the clinical value of the study.

The study also illustrates why outcome definitions need to be read closely. The platelet endpoint depends on a threshold, a stability condition, a qualifying period, and an observation window of up to 12 weeks. The record’s reference to maximal platelet counts adds wording that would benefit from further operational detail. Measurement schedules and rules for incomplete follow-up are not reported in the registry record.

Finally, the planning illustrations should be kept separate from the trial’s actual design. Their assumptions describe hypothetical planning, whereas the recorded study was single group and enrolled 13 participants. Planning calculations are not observed power and cannot substitute for outcome estimates or uncertainty intervals. Clear separation of design, planned hypotheses, observed results, and hypothetical calculations is essential for sound statistical reading.

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 is the absence of the primary endpoint result in the registry record. The outcome definition is available, but the responder count, denominator, estimated proportion, and 95% confidence interval are not reported. This prevents readers from assessing the observed response or its precision.

The study enrolled 13 participants and had no concurrent control group. The design is listed as single group, allocation not applicable, and masking none. These features constrain comparative and causal interpretation. The record does not report dose levels or escalation rules, so response patterns by dose cannot be evaluated from this information.

Several implementation details are also not reported in the registry record: the primary analysis population, handling of missing measurements, follow-up completeness, measurement schedule, and the detailed statistical analysis method. The wording about a stable platelet count and maximal platelet counts does not provide enough operational detail to reproduce the endpoint assessment fully.

Safety reporting in the available summary is limited to counts, deaths, and selected serious-event terms. Causes, timing, attribution, and participant-level overlap are not described. The investigator hypothesis about toxicity <3% is not an observed rate. The registry also does not report the numerical findings for the listed secondary outcomes.

These limitations concern what can be concluded from the registry information. They do not establish what information may have been collected elsewhere, nor do they show that an unreported analysis was or was not performed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Efficacy and Safety of Eltrombopag In Patients With Severe and Very Severe Aplastic Anemia — NCT01703169.
• Registry record last updated 2017-10-19; status last verified 2017-09.
• Results first posted 2017-10-19.
• 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

The registry record describes a completed phase 2, single-group, dose-escalation study of eltrombopag sponsored by the University of Utah. Enrollment was 13; masking was none. The record lists moderate, severe, and very severe aplastic anemia among the conditions studied.

The primary outcome was the proportion of participants with a stable platelet count of 50,000/μl or more during any four-week period within up to 12 weeks on study. The record specifies a 95% confidence interval but does not report the numerical primary result in the information.

Safety figures list 3 of 13 participants with serious adverse events, 13 of 13 with other adverse events, and 2 deaths. Congestive heart failure, cholecystitis, and syncope are each listed at 7.7% among the leading serious-event terms. These figures are descriptive and do not establish event causality. Details of the analysis plan, missing-data handling, and numerical secondary outcomes are not reported in the registry record.