Evaluating the Safety of Long Term Dosing of Romiplostim (Formerly AMG 531) in Thrombocytopenic Subjects With Myelodysplastic Syndromes (MDS): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluating the Safety of Long Term Dosing of Romiplostim (Formerly AMG 531) in Thrombocytopenic Subjects With Myelodysplastic Syndromes (MDS) (NCT00472290), a non-randomized na study. The primary outcome is Overall Summary of Adverse Events.

2. Clinical Question

This is a descriptive safety question rather than a test of whether romiplostim outperforms a control. The record reports participants with adverse events and selected serious adverse-event terms. It does not provide a randomized contrast or a causal estimate against another treatment.

3. Trial Design

This was an open-label extension study of romiplostim in thrombocytopenic participants with MDS. The registry identifies the allocation as non-randomized, the model as single group, and masking as none. It lists the study as completed and records an enrollment of 72. The listed phase is NA.

The study description emphasizes long-term safety, including overall adverse events, bleeding events, platelet transfusion utilization, and duration of platelet response. It also says the study would describe time to disease progression to acute myeloid leukemia and survival. The listed secondary outcomes include weekly bleeding events per 100 subject years, platelet transfusion events per 100 subject years, weeks with platelet response per year, time to first platelet response, and duration of platelet response. The record lists 7 outcome measures, but the posted results do not include an analysis for each of those questions.

The primary outcome is an overall summary of adverse events, with the unit recorded as participant. Its timeframe is the treatment period from first dose to the end-of-study visit, on average 56 weeks. The safety analysis population is defined as participants who took at least one dose of romiplostim.

Because the study has a single group and no masking, its design supports a description of events observed during treatment, not a like-for-like comparison with a control group. Details of the protocol, eligibility criteria beyond the study description, and the planned statistical analysis are not reported in the registry record. The registry also does not report an explanation for the sample-size choice.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Safety analysis includes subjects who took at least one dose of romiplostim.

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 Overall Summary of Adverse Events, measured over During treatment period from first dose of IP to End of Study visit, on Average 56 Weeks ..

The primary endpoint is titled “Overall Summary of Adverse Events.” It is measured in participants and covers the treatment period from the first dose of romiplostim through the end-of-study visit, which is described as lasting an average of 56 weeks. The stated safety population consists of participants who took at least one dose.

An adverse-event summary is used to describe the types and occurrence of unwanted medical events recorded during treatment. It can provide a broad account of what was observed in the study population over the stated observation period. Counts of participants with events are different from counts of all event episodes: a participant can have more than one event, and the registry summary does not establish how many episodes each participant experienced.

General statistical context: a participant-level summary is useful for showing how many people were affected, but it does not by itself show whether treatment caused an event. In a single-group, open-label study, events may reflect participants’ underlying illness, treatment, other care, or factors not distinguished by the reported summary. The endpoint can therefore support descriptive safety reporting, while causal attribution and comparison with an alternative require evidence not present in this record.

The registry gives an overall adverse-event endpoint title and separate serious and other adverse-event participant counts. Detailed endpoint definitions, collection rules, and the analysis method are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The design determines the kind of inference that is possible. This record describes a non-randomized, open-label, single-group extension study. Participants were not allocated at random to romiplostim or a comparator, and the registry identifies no comparator group. Consequently, the primary statistical task reflected by the posted results is description: identifying participants with reported adverse events and summarizing selected event terms.

What does a participant count represent?

The adverse-event summary uses participants as its unit. The registry reports 29 participants affected by serious adverse events and 66 affected by other adverse events, out of 72 at risk. A participant count answers how many people were recorded as experiencing an event category. It is not the number of separate episodes: repeated events in one person do not necessarily correspond to additional affected participants. The posted summary does not provide the number of event episodes or show whether serious and other categories overlap.

General statistical context: participant-level summaries are often easier to interpret than raw event totals when the question concerns how many people were affected. They nonetheless lose information about recurrence and event timing. Two studies could have the same number of affected participants but differ in how often events recur or how long they last. Those features cannot be reconstructed from the counts alone.

Why does the design matter?

Random allocation is a way to create comparable treatment groups, on average, by distributing measured and unmeasured characteristics across groups. That feature is absent. In a single-group study, the event experience observed during treatment has no concurrent randomized reference. Underlying MDS, other conditions, other care, and treatment exposure can all be relevant to observed events, but this record does not separate their contributions.

Open-label treatment also means masking was not used. General statistical context: awareness of treatment can affect reporting or assessment in some settings. The registry record does not say whether any particular event was affected in this way, so this is a design consideration rather than a claim about the posted events.

How should time in the study be considered?

The endpoint timeframe begins at the first dose and ends at the end-of-study visit; the record describes the average period as 56 weeks. A participant-level count does not adjust for differences in time under observation. If follow-up varies, people observed for longer have more opportunity to experience an event. Individual follow-up times and person-time analyses are not reported in the registry record.

General statistical context: event rates per unit of person-time can address unequal observation time when the required event and follow-up data are available. Such rates are distinct from the proportion of participants affected. The registry lists some secondary outcomes as events per 100 subject years, but that wording alone does not supply the results or methods for those outcomes.

What does uncertainty add?

A confidence interval, in general, describes a range of values compatible with data under a specified statistical procedure and assumptions. It helps show the precision of an estimate; it does not remove bias from a design without a comparator. The registry summary does not report confidence intervals or a statistical test for the adverse-event counts. The number of analyses is recorded as 0, so the posted results do not provide an inferential analysis to interpret alongside the counts.

For rare or relatively infrequent events, estimates can be uncertain, particularly when based on a limited group. General statistical context: uncertainty depends on the number of observations and the analysis structure, and event data can also exhibit over-dispersion when variability exceeds a model’s assumptions. Over-dispersion matters if a count model is used because standard errors and intervals can otherwise be too optimistic. This record does not report a count model, an over-dispersion assessment, or model-based estimates; these concepts are general and should not be mistaken for methods used in this study.

What do the planning illustrations mean?

The sample-size section is a planning illustration based on an observed effect, not a description of the trial’s design calculation. Its stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. General statistical context: such calculations explore the sample size that would be required for a hypothetical comparison under chosen assumptions. They are not observed power, do not show that the present study was adequately powered, and do not turn a single-group study into a randomized comparison. The trial’s own sample-size justification and statistical plan are not reported in the registry record.

In short, the posted participant counts can be read directly as descriptive summaries. Any stronger interpretation would require more information about event definitions, follow-up, missing observations, analysis procedures, and a valid comparison strategy.

10. Multiplicity, Missing Data and Analysis Populations

The study description names several areas of interest beyond the primary overall adverse-event summary, including bleeding, platelet transfusions, platelet response, disease progression to acute myeloid leukemia, and survival. The registry lists secondary outcome measures, but the posted record reports no analyses; the analysis count is 0. It is therefore not possible to assess which outcomes were analyzed, whether all planned outcomes were reported, or how the investigators handled multiple statistical tests.

Multiplicity arises when many outcomes or comparisons are examined. General statistical context: testing more hypotheses can increase the chance of at least one apparently notable result by chance, so a prespecified analysis plan and transparent reporting help readers interpret a collection of results. The statistical plan, any multiplicity adjustment, and any hierarchy among outcomes are not reported in the registry record. Since no inferential analyses are posted here, there are no reported p-values or confidence intervals to evaluate for multiplicity.

The registry identifies the safety analysis population as participants who took at least one dose of romiplostim. The adverse-event summary lists 72 at risk. The record does not describe how missing event information, incomplete follow-up, withdrawals, or other missing data were handled. It also does not report whether the listed outcome summaries use any population other than the stated safety population.

General statistical context: missing data can affect a safety summary if participants with incomplete follow-up differ from those with complete observation. The direction and size of any effect depend on why information is missing and how the analysis handles it. Without a reported missing-data approach or participant-level follow-up information, the consequences cannot be assessed for this record. These are reporting gaps, not evidence that data were missing or that any particular analysis was performed.

11. 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: pneumonia (4.2%), haematoma (4.2%), anaemia (2.8%), febrile neutropenia (2.8%), bronchitis (2.8%).

The registry lists 72 participants at risk in the romiplostim group. It records serious adverse events affecting 29 participants and other adverse events affecting 66 participants. Read these as participant-level summaries during the treatment period from first dose to end-of-study visit, on average 56 weeks, rather than as totals of event episodes.

The serious terms highlighted are pneumonia and haematoma, each listed as 4.2%; anaemia, febrile neutropenia, bronchitis, and urinary tract infection, each listed as 2.8%. The record does not give further detail about individual events, their timing, duration, severity grading, outcome, or attribution to treatment. It also does not explain the calculation or denominator for the percentages in the listed terms, so they should be quoted as reported rather than reinterpreted.

General statistical context: serious adverse events and other adverse events are categories that can include different levels of clinical consequence, but event definitions and classification rules matter. A participant may have more than one event, and the categories may not be mutually exclusive. The summary does not clarify overlap. These figures describe what the registry reports in the treated group; without a comparator, they cannot establish that romiplostim caused an event or quantify how the event experience compares with another treatment.

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

The posted results mean that adverse events were recorded among participants in an open-label romiplostim extension study. The registry reports 72 participants at risk, with serious adverse events affecting 29 participants and other adverse events affecting 66 participants. It also highlights six serious adverse-event terms with their listed percentages. These details provide a descriptive account of the treated group during the reported treatment period.

They do not establish the probability of an event for a broader MDS population, nor do they show the causal effect of romiplostim. There is no randomized comparator, and the registry reports no inferential analysis for the primary endpoint. The record also does not give a confidence interval or another measure of uncertainty for the participant counts. The posted data therefore do not quantify how precisely a population-level event frequency has been estimated.

The results are also not a complete account of every clinical question named in the study description. Although the record identifies bleeding, platelet transfusion use, platelet response, disease progression to acute myeloid leukemia, and survival as areas of interest, corresponding analyses are not reported in the results. Their absence from the posted analysis cannot be read as evidence of a particular outcome.

General statistical context: an observed event in a treatment study is not automatically an event caused by treatment. Causal interpretation depends on an appropriate comparison and assumptions about how groups differ. The single-group design documents events during treatment but cannot separate treatment effects from the underlying condition, other care, or other influences using the information in this record.

Accordingly, the appropriate reading is limited but useful: the record describes the safety events observed in this study population, under the recorded treatment and follow-up conditions. It does not establish comparative safety, causation, or outcomes beyond those reported.

13. Why This Trial Matters Statistically

This study matters statistically because long-term safety questions require attention to both what is counted and how people are observed. The registry identifies a participant-level adverse-event endpoint, defines a safety population based on receiving at least one dose, and specifies a treatment-period window from first dose to end-of-study visit. Those elements clarify the intended scope of the summary.

The record also illustrates the limits of descriptive evidence. A single-group extension can document events occurring during treatment, but without a concurrent randomized comparator it cannot distinguish treatment-associated events from events related to MDS or other influences. The open-label design and absence of a posted inferential analysis further shape interpretation. Reporting the design alongside the counts is essential; a numerical summary alone can invite a stronger conclusion than the study supports.

Another statistical issue is the distinction between people and episodes. Participant counts show how many people were affected, while event totals, recurrence, and person-time rates answer different questions. The primary endpoint uses participants, and the record does not supply event-level or individual follow-up detail. This makes it important not to treat the summary as a rate or as an account of event burden per person.

The registry’s stated interest in bleeding, transfusions, platelet response, disease progression, and survival also highlights the need for transparent outcome reporting. The record lists several secondary measures but does not post analyses for them. Readers cannot determine which were completed or how they were analyzed from this record alone.

Finally, planning illustrations must be kept separate from evidence observed in the trial. A sample-size calculation based on a hypothetical two-arm comparison does not change the single-group design or measure observed power. Keeping design, descriptive results, uncertainty, and hypothetical planning distinct makes the interpretation more faithful to what the registry actually reports.

14. 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.

This record has several trial-specific limitations for interpreting the posted safety summaries. The study is non-randomized, single group, and open label, with no comparator reported. Consequently, the event counts cannot support a controlled comparison or isolate the contribution of romiplostim from the underlying condition and other influences.

The primary outcome is a participant-level overall adverse-event summary, but the record does not provide detailed event definitions, coding rules, event-level records, recurrence, timing, duration, or outcomes. It reports the serious and other affected-participant counts without clarifying whether the categories overlap. The selected serious-event terms are listed with percentages, but the denominator and calculation details are not reported in the registry record.

Follow-up is described as an average of 56 weeks, but individual exposure and observation times are not reported. The counts therefore cannot be interpreted as rates standardized for time at risk. The registry does not describe the handling of withdrawals, incomplete observation, or missing data, nor does it report a sensitivity analysis.

The registry lists multiple safety and clinical outcomes of interest, but the analysis count is 0 and no inferential analysis is posted for the primary endpoint. The protocol-level statistical plan, multiplicity approach, sample-size justification, and interim-analysis procedures are not reported in the registry record. The absence of these details prevents assessment of whether analyses followed prespecified methods; it is not evidence that any particular method was or was not used.

These limitations constrain causal and comparative conclusions. They do not negate the descriptive value of documenting reported adverse events, but they set clear boundaries on what the counts can establish.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Evaluating the Safety of Long Term Dosing of Romiplostim (Formerly AMG 531) in Thrombocytopenic Subjects With Myelodysplastic Syndromes (MDS) — NCT00472290.
• Registry record last updated 2017-12-29; status last verified 2017-12.
• Results first posted 2013-07-15.
• 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.

18. Record Summary

The registry record describes a completed, open-label extension study of romiplostim in people with MDS and thrombocytopenia. It identifies a non-randomized, single-group design with no masking and records enrollment of 72 participants. The safety analysis includes participants who took at least one dose, and the primary outcome is an overall summary of adverse events during treatment from first dose to the end-of-study visit, on average 56 weeks.

The posted adverse-event summary lists 29 participants with serious adverse events and 66 with other adverse events, among 72 at risk. It also lists selected serious adverse-event terms and percentages. No inferential analyses are reported in the registry record. The findings are therefore descriptive and do not provide a randomized comparison or establish causation. The registry’s statistical plan, missing-data procedures, and sample-size justification are not reported.