Thymoglobulin and Cyclosporine in Patients With Aplastic Anemia or Myelodysplastic Syndrome: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Thymoglobulin and Cyclosporine in Patients With Aplastic Anemia or Myelodysplastic Syndrome (NCT00806598), a na phase 2 study. The primary outcome is Overall Response.

2. Clinical Question

The study also set out to examine safety. The record lists serious adverse events and other adverse events for the treated group. Those safety summaries address events observed during the study, but they do not by themselves establish that the treatment caused an event. In a single-group study, an event occurring after treatment cannot be separated from events related to the underlying conditions, other care, or other causes using a concurrent randomized comparison.

The primary clinical question is therefore descriptive: what responses and adverse events were recorded among participants treated with this combination? The registry contains a response definition and the number considered evaluable, but it does not report the actual primary endpoint result. There is no registry-based comparison with another regimen, and no basis in this record to attribute any outcome to one drug within the combination.

3. Trial Design

This was a phase II, single-group study sponsored by M.D. Anderson Cancer Center. It was listed as completed. The study start is recorded as 2005-05, and primary completion as 2012-06. The registry lists enrollment of 53. These registry counts describe different stages: the primary endpoint text states that 48 participants received treatment, and that 46 were evaluable for response. The reason that some enrolled or treated participants were not included in the response-evaluable population is not reported in the registry record.

The intervention was a combination of thymoglobulin, cyclosporine, methylprednisolone, and G-CSF. In the arm description, thymoglobulin and methylprednisone are described as intravenous treatment for 5 days, with the thymoglobulin dose stated as 3.5 or 2.5 mg/kg/day. Cyclosporin is described as oral treatment for 6 months following thymoglobulin, and G-CSF as subcutaneous treatment. The registry summary uses the names cyclosporine and methylprednisolone. No separate component groups are listed, so the study cannot distinguish the contribution of an individual drug from that of the combination.

The allocation field is NA, and the model is single group. There was no randomized comparator arm, and masking is listed as none. Thus, the design is not a like-for-like control comparison. The study can describe outcomes observed in treated participants, but the design does not provide a direct estimate of the difference between the combination and an alternative treatment.

The response timeframe begins at randomization according to the endpoint description, though allocation is listed as NA. The endpoint was assessed first at 3 months on study and then monthly up to 3 years, with follow-up described as continuing until first documented progression or death from any cause, whichever came first. How these timing descriptions were operationalized for this single-group study is 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.

Before assignment: A total of 53 participants were enrolled on the study, with four patients choosing alternative treatment; one participant died before administration of study treatment.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Of the 48 participants who received treatment 46 were evaluable for response.

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 Response, measured over From date of randomization until the date of first documented progression or date of death from any cause, whichever came first. Assessed first at 3 months on study, continuing monthly up to 3 years..

The primary endpoint, overall response, is a count of participants meeting a defined response category. Complete response required normalization of peripheral blood and bone marrow, including a marrow blast criterion and blood-count criteria. Partial response was defined using transfusion independence and specified blood-count thresholds. Hematologic improvement was defined as a clinically relevant increase in hemoglobin, platelets, or absolute neutrophil count. The registry presents these categories as the basis for judging response.

A response endpoint is useful in a phase II study because it describes whether participants meet clinically interpretable criteria after receiving the study regimen. The combined endpoint can include improvement that falls short of complete response, rather than counting only complete normalization. That broader definition may capture different degrees of hematologic change, but it also means that a reader needs the category-specific results and the exact analysis rule to understand what contributed to the overall count.

The registry records the endpoint unit as participants and identifies 46 treated participants as evaluable for response. It does not report how many met the overall response definition, how many were in each response category, or the corresponding uncertainty. It therefore offers the criteria but not the observed primary endpoint result. The timing description specifies assessment beginning at 3 months and continuing monthly up to 3 years; the response result by assessment time is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a single-group design estimate?

In a single-group study, outcomes are described among participants who receive the study regimen. The design does not randomly assign participants to a separate comparator, so it does not directly estimate the difference between treatment strategies. A response count can be informative as a description of what happened in the study group, but interpreting it as a treatment effect requires a comparison that this record does not contain. Differences between participants and outside groups, changes over time, and other aspects of care cannot be controlled through random assignment within this design.

Masking is listed as none. Masking can reduce the influence of knowledge about treatment assignment on some assessments or decisions, depending on the study procedures. In an unmasked study, that protection is not present. The registry does not describe whether response assessment was independently reviewed or what procedures were used to limit assessment bias, so those details are not reported in the registry record.

What does the binary endpoint mean?

The primary outcome is classified as binary at the participant level: a participant either meets the overall response definition or does not meet it under the analysis rule. The endpoint is expressed as a count of participants rather than a continuous measurement. The response categories themselves have clinical definitions, but the registry does not provide the number in each category or the rule for combining them beyond its descriptions of complete response, partial response, and hematologic improvement.

For a binary outcome, an analysis commonly presents the number meeting the endpoint in relation to the relevant analysis population, and may add an uncertainty interval. A confidence interval describes a range of values reasonably compatible with the observed data under the method's assumptions; it is not a guarantee that the true value lies within that range. The interval and its method are not reported in the registry record. Since the numerator is also absent, no response proportion or interval can be interpreted from this record.

What does an analysis population tell the reader?

The record distinguishes enrollment, treatment receipt, and response evaluability. These labels answer different questions. Enrollment describes participants recorded as enrolled; treatment receipt identifies those described as treated; evaluability indicates who was included in the response assessment population. A result based on an evaluable population may not answer the same question as an analysis of everyone enrolled. The registry does not describe the reasons for non-evaluability, a protocol-defined population, or an intention-to-treat analysis. Those analysis details are not reported in the registry record.

Missing outcome data can affect interpretation when the likelihood of missingness is related to prognosis or response. For example, excluding participants without an evaluable outcome can yield a different description from treating all enrolled participants according to a prespecified rule. That is a general methodological consideration, not a statement about why participants are absent from the evaluable group in this study. The record does not report missing-data assumptions, imputation, sensitivity analyses, or participant-level dispositions.

What is the role of a confidence interval or significance test?

A confidence interval communicates statistical uncertainty around an estimate, while a significance test evaluates compatibility of data with a specified null hypothesis under a chosen model. Neither replaces the need to state the endpoint, population, analysis method, and assumptions. For a single-group response estimate, a test against a benchmark would require a specified benchmark and analysis plan. Neither a benchmark nor a primary endpoint hypothesis test is reported in the registry record. The record lists no analyses for the outcome, so it is not possible to assess any test, adjustment, or model choice.

Statistical significance and clinical importance are different ideas. A p-value, if one were reported, would not by itself indicate the size or importance of a response. An estimate with an interval would better communicate both the observed magnitude and its precision, but neither is available for the primary endpoint in this record. General statistical principles cannot fill in missing outcome results.

How should planning calculations be read?

Sample-size calculations are prospective planning tools. They describe how many participants might be needed under assumptions about the effect to detect, significance level, allocation, and statistical model. They do not describe the result actually observed in a completed study and should not be called observed power. The stated planning assumptions for illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The targeted effect is expressed as a stated fraction of the observed effect. The registry record does not report the numerical illustration results or a sample-size justification from the original study design.

The two-arm assumptions are not a description of the enrolled trial, which is single group. Thus a hypothetical two-arm planning calculation should not be mistaken for a comparison performed in this study. A calculation based on an observed effect is especially dependent on that effect estimate and the assumptions used; it does not provide evidence that the original study was designed to detect that effect. The original protocol's sample-size rationale, power target, and design assumptions are not reported in the registry record.

Why can over-dispersion matter?

Over-dispersion is a general issue for count outcomes when observed variability exceeds what a simple probability model assumes. A model that ignores extra variability can produce uncertainty estimates that are too narrow. The primary outcome is a participant-level response classification, not a recurrent count of events, and the registry gives no statistical model for it. Over-dispersion should therefore not be assumed to have been assessed in this trial. For safety events, the record provides participant counts and event terms, not a model for repeated events or their dependence within participants. Any adjustment for event clustering or over-dispersion is not reported in the registry record.

Overall, the study record supports a description of its design, response criteria, analysis population counts, and listed safety information. It does not provide enough primary outcome analysis detail to evaluate an estimate, its precision, or a model-based inference.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and 0 analyses. It does not report a primary response estimate or an inferential result. It also does not describe a statistical analysis plan, a hierarchy of hypotheses, or an adjustment for multiple comparisons. Whether additional outcomes or subgroup comparisons were considered in the study protocol is not reported in the registry record.

Multiplicity arises when several hypotheses, endpoints, time points, or subgroups are examined. The more opportunities there are to identify an apparently unusual result, the more important it becomes to prespecify which comparisons are confirmatory and how error rates will be controlled. This is a general statistical concern. The response description includes complete response, partial response, and hematologic improvement, but the record does not state whether they were separately analyzed, combined according to a prespecified rule, or treated as exploratory categories. The registry's outcome definition alone does not establish a multiplicity plan.

Missing-data handling is also not described. The endpoint record states that 46 participants were evaluable for response after 48 received treatment, while enrollment is listed as 53. The reasons for these differences, the outcomes of participants not evaluated, and the denominator used for any response analysis are not reported in the registry record. No imputation method, missing-at-random assumption, sensitivity analysis, or analysis of all enrolled participants is described.

Consequently, the reader cannot determine how the primary result would change under alternative assumptions about participants without evaluable responses. Nor can the record establish whether all planned analyses were reported. This uncertainty is about the documentation available in the registry record; it is not evidence that a particular analysis was or was not performed outside the registry.

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: infection (6.2%), death (4.2%), fever (2.1%), syncope (2.1%), neutropenic fever (2.1%).

The safety record lists one treatment group, with 48 participants at risk for serious adverse events and 8 participants affected. It separately lists 48 participants at risk for other adverse events and 48 affected. These are participant counts in the registry's safety summary. They should not be added together: the categories may capture different event groupings, and a participant may experience events in more than one category. The record does not report a detailed event-level accounting or clarify overlap.

The serious adverse event terms with listed percentages are Infection, 6.2%; Death, 4.2%; Fever, 2.1%; Syncope, 2.1%; Neutropenic Fever, 2.1%; and Confusion, 2.1%. The registry's serious-event group information gives no death count. These percentages are reported as event-term percentages; they should not be converted to counts or interpreted as mutually exclusive categories. The registry does not report event timing, duration, severity grades beyond the serious-event designation, or attribution to treatment for each term.

Adverse events observed during treatment do not automatically establish that treatment caused them. With no control group, the registry cannot provide a comparative estimate of excess risk against another regimen. These figures describe recorded safety events in the study group and should be interpreted alongside the limitations in event attribution and reporting.

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

The record establishes that a phase II study of a four-drug combination was conducted in a single group and that overall response was designated as the primary endpoint. It defines complete response, partial response, and hematologic improvement, and specifies assessments beginning at 3 months with monthly follow-up up to 3 years. It also reports the size of the treated and response-evaluable populations and provides a safety summary.

The key missing element is the primary outcome result. Without the number meeting the response definition, the record cannot show the frequency of response, the distribution across response categories, or the precision of an estimate. The protocol definition clarifies what would count as response; it does not show whether participants met that definition. The absence of an analysis in the registry record also means there is no reported primary endpoint test or confidence interval to interpret.

Even if a response count were available, the single-group design would support a description of outcomes in treated participants rather than a controlled estimate of treatment effect. Other factors, including the underlying conditions and care outside the study intervention, could contribute to observed outcomes. The record provides no randomized comparison that would separate those influences from the effect of the combination.

The safety terms and counts are similarly descriptive. They show adverse events recorded in the group, but do not identify causation or comparative risk. The record's serious-event list includes Infection and Death among its terms, but does not provide a death count in the group summary. No conclusion about the balance of benefits and harms can be drawn from the safety list alone, particularly when the primary response result is absent.

These limitations concern interpretation of the registry record, not a claim about what may be available elsewhere. The central statistical message is that design and endpoint definitions are documented, while the primary outcome estimate and its uncertainty are not reported in the registry record.

13. Why This Trial Matters Statistically

This study illustrates why a trial's design and its reported results must be read together. A clearly defined endpoint is necessary for interpretation, but a definition alone cannot answer whether participants met the endpoint. Here, overall response has explicit clinical criteria, and the registry identifies the response-evaluable population. The primary result itself is not reported in the registry record, leaving the endpoint's observed frequency and precision unknown.

The distinction between enrolled, treated, and evaluable participants is also statistically important. These populations determine what a result describes and which participants contribute to an estimate. The record lists enrollment of 53, treatment receipt for 48, and response evaluability for 46. Because the reasons for the differences and the missing-data approach are not reported, a reader cannot assess possible selection effects or reproduce the intended analysis population.

The study is a phase II, single-group evaluation, rather than a randomized comparison. That design can characterize treatment experience and generate information for future research, but it does not directly estimate a comparative treatment effect. The absent control group is central to interpretation, not a minor reporting detail. Similarly, listed adverse events provide an account of observed safety terms but do not provide a comparison of risk or establish causation.

For statisticians and clinical researchers, the record underscores the need to report endpoint counts, analysis populations, uncertainty intervals, missing-data handling, and the prespecified analysis plan. Those elements allow readers to distinguish descriptive outcomes from comparative claims and to judge how precisely an estimate is known. The registry entries several design and endpoint features, but these key outcome-analysis details are not reported in the registry record.

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.

The primary limitation for statistical interpretation is that the registry record does not report the overall response count, category-specific response counts, an estimate of response frequency, or a confidence interval. Although the endpoint is defined and 46 participants are described as evaluable, the primary result cannot be reconstructed from the information in the record.

The single-group, unmasked design has no concurrent control group. It therefore cannot directly estimate a difference between this combination and another treatment, and the record cannot separate treatment effects from the course of disease, other care, or assessment influences. The number enrolled, number treated, and number evaluable are all recorded, but why these populations differ is not reported in the registry record. Missing-data procedures and the handling of participants who were not evaluable are also not reported.

The endpoint description gives several response criteria, but the registry does not report the number satisfying each category or how the categories contributed to any reported overall result. The timing description states an initial assessment at 3 months followed by monthly assessments up to 3 years; outcome results by time point are not reported. The endpoint text refers to time from randomization, while allocation is listed as NA. How this wording maps to the single-group study is not reported in the registry record.

The registry lists no outcome analyses and does not describe the statistical analysis plan, sample-size justification, multiplicity handling, or sensitivity analyses. The serious adverse event summary includes affected-participant counts and event terms, but not detailed event-level information or treatment attribution. The record does not include a death count in the serious adverse event group information. Together, these omissions limit assessment of efficacy, precision, safety attribution, and robustness.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Thymoglobulin and Cyclosporine in Patients With Aplastic Anemia or Myelodysplastic Syndrome — NCT00806598.
• Registry record last updated 2013-03-12; status last verified 2013-03.
• Results first posted 2013-03-06.
• 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

NCT00806598 is a completed phase II, single-group study sponsored by M.D. Anderson Cancer Center. It enrolled 53 participants and evaluated thymoglobulin, cyclosporine, methylprednisolone, and G-CSF for aplastic anemia or myelodysplastic syndrome. The registry states that 48 participants received treatment and 46 were evaluable for response.

Overall response was the primary endpoint, with criteria for complete response, partial response, and hematologic improvement. The primary response result and endpoint analyses are not reported in the registry record. The safety summary lists 8 participants with serious adverse events and 48 with other adverse events among 48 at risk for each category, along with serious-event terms and percentages.

Because the study had no control group and the primary result is absent, the registry supports description of the design and outcome definitions, but not an estimate of comparative treatment effect or the precision of the response outcome.