Comparing Cyclophosphamide and Abatacept With Standard of Care Treatment Following Stem Cell Transplantation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparing Cyclophosphamide and Abatacept With Standard of Care Treatment Following Stem Cell Transplantation (NCT03680092), a randomized phase 2 study. The primary outcome is Occurrence of Moderate and Severe Chronic GVHD at One Year Post Transplant.

2. Clinical Question

The registry frames the study around whether the prophylaxis regimens differ in the occurrence of moderate and severe chronic GVHD after transplantation. The endpoint description says chronic GVHD was to be diagnosed and staged according to National Institutes of Health consensus criteria. The comparison is randomized, but masking is listed as none, so the study is described as unmasked.

The question is clinically specific: it concerns one GVHD severity grouping at a stated follow-up horizon, rather than every possible transplant outcome. GVHD- and relapse-free survival by one year post transplant is also listed as a secondary outcome. That broader endpoint addresses a different outcome construct and should not be treated as interchangeable with the primary endpoint. The registry record does not report further eligibility details, such as age limits or transplant characteristics, so the results cannot be characterized more narrowly from this record.

3. Trial Design

This is a randomized, parallel, phase II trial with no masking. The registry lists an enrollment of 43 and a completed status. The recorded study start is 2019-11-26, and the primary completion date is 2023-06-15. The sponsor is Dimitrios Tzachanis, MD PhD. The registry identifies GVHD and hematologic neoplasms as conditions addressed.

The experimental arm is titled Cyclophosphamide and Abatacept. Its description specifies high dose cyclophosphamide on Days +3 and +4, followed by abatacept for 6 months, with abatacept at 10mg/kg administered on days +5, +14, +28, +56, +84, +112, +140 and +168. The comparator arm is titled Methotrexate and Tacrolimus. Its description specifies methotrexate on Days +1, +3, +6 and +11 and tacrolimus. The record describes that arm as standard of care.

The primary outcome is the occurrence of moderate and severe chronic GVHD at one year post transplant; its timeframe is also described as through study completion, an average of 1 year. The secondary outcome is GVHD- and relapse-free survival by one year post transplant. Posted primary outcome counts are 0 of 24 in the cyclophosphamide-and-abatacept arm and 8 of 14 in the methotrexate-and-tacrolimus arm. For the secondary outcome, the posted counts are 15 of 24 and 4 of 14, respectively. These endpoint denominators are reported separately from the overall enrollment. Reasons for the difference between enrollment and endpoint denominators are not reported in the registry record.

Random allocation supports a comparative interpretation, but the registry record does not report allocation concealment, detailed eligibility criteria, baseline characteristics, or methods for handling deviations from assigned treatment. It also does not give a fuller account of follow-up or outcome ascertainment beyond the endpoint descriptions. These omissions limit assessment of how representative the analyzed groups were and how the trial was conducted in practice.

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 Occurrence of Moderate and Severe Chronic GVHD at One Year Post Transplant, measured over through study completion, an average of 1 year.

The primary endpoint records participants with moderate or severe chronic GVHD at one year post transplant. The registry describes chronic GVHD diagnosis and staging according to National Institutes of Health consensus criteria. The endpoint is a participant count, so the posted results indicate how many participants in each reported arm experienced the defined event, rather than the severity distribution or timing of each event.

Such an endpoint is useful because it focuses the comparison on a clinically meaningful complication after transplantation and defines a specific observation horizon. The combined category of moderate and severe chronic GVHD captures the occurrence of either level included in the registry definition. It does not, by itself, show how many cases were moderate versus severe, when they began, how long they lasted, or what treatment they required. Those details are not reported in the registry record.

The posted primary counts are 0 in the cyclophosphamide-and-abatacept arm and 8 in the methotrexate-and-tacrolimus arm, with endpoint denominators of 24 and 14, respectively. The registry comparison reports a risk difference of 0.571, with a 95% confidence interval from 0.312 to 0.831 and p < 0.001. The contrast is expressed as the event risk in the methotrexate-and-tacrolimus arm relative to the cyclophosphamide-and-abatacept arm. These are the recorded results, not a description of outcomes beyond the stated endpoint or follow-up period.

7. Results Posted in the Registry

• Methotrexate and Tacrolimus: 57.1% (8/14) versus 0.0% (0/24) on control; risk difference 57.1 percentage points (95% CI 31.2 to 83.1); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

GVHD- and Relapse- Free Survival by One Year Post Transplant

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Methotrexate and Tacrolimus: 28.6% (4/14) versus 62.5% (15/24) on control; risk difference -33.9 percentage points (95% CI -64.5 to -3.3), risk ratio 0.46 (0.19–1.11); unadjusted two-proportion z-test p = 0.044.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Methotrexate and Tacrolimus than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists GVHD- and relapse-free survival by one year post transplant as the secondary outcome. Its comparison points toward a higher recorded count of participants with that outcome in the cyclophosphamide-and-abatacept arm than in the methotrexate-and-tacrolimus arm. This direction is broadly consistent with the primary outcome pattern, in which fewer moderate or severe chronic GVHD events were recorded in the cyclophosphamide-and-abatacept arm. The endpoints nevertheless differ: one is a defined chronic GVHD event, while the other combines GVHD and relapse in a survival outcome. They are not interchangeable measures of the same clinical event.

The secondary result is reported with an unadjusted p-value of 0.044, and the registry lists no reported analyses for it. A result from a secondary endpoint generally carries less weight than a primary result when its role in the analysis plan, multiplicity handling, and interpretation hierarchy are not documented. Testing several outcomes can increase the chance of observing a nominally notable result by chance. The registry record does not report whether any adjustment or prespecified hierarchy was used, so the secondary finding should be viewed as supportive or exploratory rather than conclusive.

Its confidence interval also conveys uncertainty about the size of the observed contrast. The posted result can inform interpretation of the overall pattern, but it cannot independently establish a treatment effect or resolve limitations in the primary analysis. In particular, the record does not provide a detailed endpoint definition, censoring approach, or time-to-event analysis for the secondary outcome. Further, the broader combined endpoint can be affected by its component outcomes in ways that are not visible from the summary count alone.

Taken together, the secondary outcome is directionally compatible with the primary comparison, but its unadjusted status and limited reporting mean it should not be treated as a separate confirmatory demonstration. Its value is in adding context to the endpoint pattern while leaving the primary outcome as the main registry comparison.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• GVHD- and Relapse- Free Survival by One Year Post Transplant

9. Statistical Interpretation

10. Statistical Methods Explained

What does a risk difference describe?

A risk difference compares the proportion of participants with an event in one arm with the corresponding proportion in another. It is an absolute contrast, expressed on the event-proportion scale. A value of zero represents equal observed risks; positive or negative signs indicate which arm has the larger risk according to the order used for the comparison. For the primary outcome, the registry comparison is oriented as methotrexate and tacrolimus relative to cyclophosphamide and abatacept. The reported positive estimate therefore corresponds to more primary events in the former arm. The secondary comparison is also oriented with methotrexate and tacrolimus relative to cyclophosphamide and abatacept, but its negative estimate reflects fewer participants with the secondary event in the former arm.

A risk difference is distinct from a risk ratio or odds ratio. It focuses on an absolute difference in event proportions, whereas relative measures express one risk in relation to another or compare odds. The registry reports a risk difference for both outcomes. For the secondary outcome, it also lists a risk ratio and odds ratio with intervals. Those alternative measures answer related but not identical questions and should not be substituted for the primary risk-difference comparison.

How should the p-value be read?

A p-value describes how unusual the observed data, or data more extreme, would be under a statistical model in which the comparison has no underlying difference. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. The primary comparison has a reported unadjusted p-value of < 0.001. The secondary comparison has an unadjusted p-value of 0.044. Both are posted as unadjusted; the registry record does not report a multiplicity adjustment. A p-value near a conventional cutoff should not be treated as a sharp dividing line between reliable and unreliable evidence.

Why do confidence intervals matter?

A confidence interval gives a range of values that are reasonably compatible with the observed data under the interval's statistical method. It presents uncertainty about the estimated effect more directly than a p-value alone. The primary risk-difference interval runs from 0.312 to 0.831. The secondary interval runs from -0.645 to -0.034. Each interval is a measure of statistical precision, not a guarantee that the true effect lies within those endpoints. It also does not incorporate every source of uncertainty, such as possible selection, measurement, or reporting limitations.

The widths of intervals depend on the amount of information in the data and on the observed outcomes. With relatively small endpoint denominators, a few outcomes can have substantial influence on the estimated proportions. In particular, observing no primary events in one arm gives a clear description of this trial's recorded data, but does not eliminate uncertainty about the event risk that might be observed in a larger population. The confidence interval for the difference is important context for that zero count.

What are the planning illustrations?

The posted planning figures are illustrations based on the observed secondary outcome effect. They assume a two-sided significance level 0.05, equal allocation between the two arms, and a normal approximation. The effect used in each scenario is a stated fraction of the observed effect. The figures are participants per arm for 80% power and participants per arm for 90% power. They are not the trial's original design calculation and are not observed power. They do not establish that the trial was adequately sized for its primary outcome or that a future trial will reproduce the observed effect.

Illustrations based on an observed effect can be optimistic when the effect estimate is imprecise or unusually large. Using a smaller fraction of that observed effect in the planning scenarios represents a more conservative target effect, but it still does not substitute for a design based on clinically justified assumptions. The registry record does not report the trial's own sample-size justification, target effect, or power calculation. It also does not report whether any planned interim analysis was conducted.

What can be inferred from the listed analyses?

The registry lists no reported analyses for the primary outcome and an empty reported-analyses list for the secondary outcome, while it also posts outcome counts and comparisons. These record features do not provide a detailed description of the statistical procedures used to generate each comparison. The posted estimates can be interpreted as the registry's arm-level results, but the record does not describe adjustment for baseline factors, missing observations, or departures from the assigned regimen.

In general, randomization is intended to create comparable groups at allocation, while analysis choices determine how the observed data are used to estimate a contrast. An intention-to-treat analysis, for example, generally analyzes participants according to randomized assignment, while a per-protocol analysis uses a different population definition. The registry record does not state which analysis population was used. These distinctions matter because exclusions or missing outcome data can affect the estimate, particularly when group sizes are limited. No method for addressing missing data is reported.

Because masking is listed as none, knowledge of assigned regimen may have been available during the trial. Unmasked treatment assignment can be relevant to care, assessment, and reporting, although the registry record does not establish that any such influence occurred. The stated diagnostic and staging criteria give a framework for the primary endpoint, but details such as independent adjudication or assessor masking are not reported. Statistical interpretation therefore relies on the posted counts and comparisons while recognizing that important operational and analytic details remain unspecified.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.571, with a 95% confidence interval from 0.312 to 0.831. The entire interval is above zero, consistent with the direction of the observed difference favoring a higher recorded primary event risk in the methotrexate-and-tacrolimus arm. The interval communicates uncertainty around the estimate: values within its range are compatible with the data under the analysis method, but the range does not guarantee where the underlying difference lies.

The secondary risk-difference estimate is -0.339, with a 95% confidence interval from -0.645 to -0.034. This interval is entirely below zero and corresponds to fewer recorded participants with the secondary outcome in the methotrexate-and-tacrolimus arm. Its lower and upper limits are relatively close to zero on one side and extend farther in the other direction, illustrating that the magnitude remains uncertain. The unadjusted p-value is 0.044; the interval and p-value are related summaries of the same comparison, not independent confirmations.

Precision depends partly on how many observations contribute to an estimate and how outcomes are distributed across arms. The primary outcome denominators are 24 and 14, and the secondary denominators are also 24 and 14. These are the endpoint denominators recorded for the comparisons, not a statement that every enrolled participant contributed to each outcome. Reasons for any difference between enrollment and the endpoint denominators are not reported in the registry record.

Confidence intervals describe sampling uncertainty under statistical assumptions. They do not account automatically for possible outcome misclassification, missing-data bias, multiple outcome testing, or limited generalizability. The intervals are also not prediction intervals for a future trial or a future population. The relatively small endpoint denominators and the absence of reported analysis details counsel against treating the point estimates as exact measures of the effects that would be expected elsewhere.

12. Multiplicity, Missing Data and Analysis Populations

The registry posts a primary endpoint and a secondary endpoint, with an unadjusted p-value for each comparison. It does not report a multiplicity adjustment or a prespecified testing hierarchy. When multiple outcomes are examined, the chance of at least one apparently notable result can increase. Adjustment procedures or a clear hierarchy can address that issue, but the record does not say whether either was used. The secondary result therefore requires particular caution, and the primary result should also be interpreted in the context of the full analysis plan, which is not reported.

The registry record does not report a missing-data method, reasons for missing endpoint observations, or a definition of the analysis population. Endpoint denominators are 24 and 14 for each posted outcome, while overall enrollment is listed as 43. The record does not explain the relationship between enrollment and those endpoint denominators. It would be inappropriate to infer why records are not represented in the endpoint counts or how their outcomes were handled.

Similarly, the record does not state whether participants were analyzed according to randomized assignment, whether treatment discontinuations or crossovers occurred, or whether any per-protocol analysis was performed. It does not report interim analyses, stopping rules, or the trial's own sample-size justification. The posted planning illustration is based on the observed secondary effect and is explicitly not the trial's original design calculation. It should not be used as evidence of observed power.

These omissions do not negate the posted results, but they limit assessment of robustness and risk of bias. Without a stated analysis population and missing-data approach, it is not possible to determine from the registry record whether the comparisons account for all randomized participants in the intended way. Without multiplicity information, the evidentiary status of the secondary test is also uncertain. These are reporting limitations rather than proof that the analyses were conducted improperly.

13. 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: febrile neutropenia (66.7%), sepsis (46.7%), lung infection (20.0%), c diff infection (20.0%), oral mucositis (13.3%).

The safety record reports serious adverse events for groups with 25 participants at risk in the cyclophosphamide-and-abatacept arm and 15 at risk in the methotrexate-and-tacrolimus arm. Serious adverse events affected 9 and 13 participants, respectively. Other adverse events affected 5 and 10 participants, and deaths are recorded as 2 and 3, respectively. These safety denominators differ from the endpoint denominators used for the efficacy comparisons and should be read separately.

The listed serious adverse-event terms include febrile neutropenia at 66.7%, sepsis at 46.7%, lung infection at 20.0%, C diff infection at 20.0%, oral mucositis at 13.3%, and diarrhea at 13.3%. These percentages are registry entries for listed terms; the record does not provide enough detail to determine overlap among terms or interpret every percentage's denominator beyond the safety group information.

Adverse event counts describe events recorded in the trial, not necessarily events caused by the assigned regimen. The registry record does not report event definitions, exposure time, attribution rules, severity breakdowns beyond the serious-event category, or comparative statistical testing for safety. The counts should therefore be used as descriptive safety information, not as a standalone estimate of relative risk or causal effect.

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

The results describe what was recorded in this randomized phase II trial over the endpoint follow-up stated in the registry. The primary outcome counts and comparison show a clear observed difference between the arms for moderate or severe chronic GVHD. The confidence interval and p-value quantify statistical evidence and uncertainty under the analysis used. They do not make the estimate exact, establish why the difference occurred, or show that the same pattern would occur in every transplant setting.

The secondary result points in a compatible direction for GVHD- and relapse-free survival, but it is unadjusted and is less persuasive as confirmatory evidence. It should not be combined conceptually with the primary result as if both measured the same outcome. The registry provides a limited summary of outcomes, without detailed time-to-event methods, analysis populations, or missing-data handling. As a result, the posted comparisons support a focused reading of the reported endpoints rather than a complete assessment of all possible outcomes.

A randomized design is useful for comparing assigned strategies, but randomization does not eliminate uncertainty in a small trial or solve all problems of outcome reporting and analysis. The unmasked design and the absence of reported details on assessment and analysis add context to interpretation. The registry record also does not provide baseline characteristics or sufficient eligibility information to assess how closely the participants resemble other populations.

These findings do not establish that either regimen is universally preferable, and the registry information is not a basis for individual treatment decisions. They describe a statistical comparison in this specific study, with the limitations of its reported sample, endpoint definitions, and available analysis details. The results may motivate further study, but the record alone cannot determine whether the observed estimates will persist with larger or different groups.

15. Why This Trial Matters Statistically

This trial is statistically informative because it reports a randomized comparison of two GVHD prophylaxis strategies and a specific primary endpoint. The outcome is defined as moderate or severe chronic GVHD by one year post transplant, giving the comparison a clear clinical focus. Reporting arm-level counts alongside an estimated risk difference, confidence interval, and p-value allows readers to consider both the direction of the observed difference and its uncertainty.

The record also illustrates why a statistically striking result should not be read without context. The primary event count is zero in one arm, and the endpoint denominators are limited. A zero count is important evidence about the observed trial, but it is not proof that the event cannot occur under that regimen. The interval is therefore central to interpretation, as are the trial's phase II status and the absence of a complete statistical analysis description.

The secondary outcome offers a related perspective, but its unadjusted status demonstrates the importance of distinguishing a primary test from secondary findings. Multiple outcomes can give a fuller picture, yet they also require a planned approach to interpretation. The registry does not report whether multiplicity was addressed or how the primary and secondary analyses were prioritized.

Finally, the record shows the value of transparent reporting beyond point estimates. Details on the analysis population, missing outcomes, endpoint assessment, and original sample-size justification would help readers judge the credibility and generalizability of the findings. The posted planning illustration is useful as a scenario, but it is not a substitute for the trial's design rationale. Statistical importance comes from reading the observed pattern together with these limits, rather than treating a p-value as a complete answer.

16. 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 registry results have important trial-specific reporting limitations. Although the study is listed as randomized, the record does not describe sequence generation or allocation concealment. Masking is listed as none, and no further detail is given about who assessed chronic GVHD or whether endpoint assessors were independent. The endpoint description names National Institutes of Health consensus criteria, but does not provide assessment procedures or adjudication details.

Overall enrollment is listed as 43, while the primary and secondary outcome comparisons use arm denominators of 24 and 14. Safety denominators are 25 and 15. The registry record does not explain these differing denominators or specify how missing observations, withdrawals, or treatment deviations were handled. It also does not identify the analysis population. This makes it difficult to assess whether the reported comparisons include all randomized participants or how incomplete follow-up might have affected results.

The record provides no baseline characteristics, detailed inclusion and exclusion criteria, or transplant-specific descriptors beyond the broad study context. That limits assessment of group comparability and applicability to other settings. Although the trial is completed, the registry record does not provide a detailed account of follow-up duration for each participant, event timing, censoring, or competing events. Such details are particularly relevant to interpreting an outcome described as survival.

Finally, the statistical plan, primary sample-size justification, and any interim analyses are not reported in the registry record. The secondary outcome has an unadjusted comparison, and the record does not state whether multiplicity was handled. These limitations constrain assessment of precision, robustness, and the status of secondary findings. They do not show that the trial was improperly conducted, but they leave questions that cannot be resolved from the registry results alone.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Comparing Cyclophosphamide and Abatacept With Standard of Care Treatment Following Stem Cell Transplantation — NCT03680092.
• Registry record last updated 2025-10-02; status last verified 2025-09.
• Results first posted 2024-10-02.
• 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.

20. Record Summary

NCT03680092 is a completed, randomized, unmasked, parallel phase II trial comparing cyclophosphamide and abatacept with methotrexate and tacrolimus for GVHD prophylaxis after hematopoietic stem cell transplantation. The registry lists enrollment of 43, a study start of 2019-11-26, and primary completion on 2023-06-15.

The primary outcome is moderate or severe chronic GVHD at one year post transplant. The posted counts are 0 of 24 in the cyclophosphamide-and-abatacept arm and 8 of 14 in the methotrexate-and-tacrolimus arm. The reported risk difference is 0.571, with a 95% confidence interval from 0.312 to 0.831 and p < 0.001. The listed secondary outcome is GVHD- and relapse-free survival by one year post transplant, with an unadjusted comparison.

The registry does not report the analysis population, missing-data approach, multiplicity handling, or the trial's original sample-size justification. Endpoint and safety denominators are reported separately, and their relationship to overall enrollment is not explained.