Recombinant EphB4-HSA Fusion Protein and Azacitidine or Decitabine for Relapsed or Refractory Myelodysplastic Syndrome, Chronic Myelomonocytic Leukemia, or Acute Myeloid Leukemia Patients Previously Treated With a Hypomethylating Agent: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Recombinant EphB4-HSA Fusion Protein and Azacitidine or Decitabine for Relapsed or Refractory Myelodysplastic Syndrome, Chronic Myelomonocytic Leukemia, or Acute Myeloid Leukemia Patients Previously Treated With a Hypomethylating Agent (NCT03146871), a na phase 2 study. The primary outcome is Assessment of Toxicity to Hypomethylating Agent (HMA).

2. Clinical Question

This is a pilot/safety study, rather than a comparative trial designed to estimate a treatment effect against a control group. The registry describes the disease setting as relapsed or refractory after prior treatment with a hypomethylating agent. It studies recombinant EphB4-HSA in combination with either azacitidine or decitabine, but the record does not report how many participants received each hypomethylating agent or provide separate outcome results for those regimens.

The principal statistical question is therefore descriptive: what toxicities were recorded among participants receiving the combination, and how were those events graded and counted? A single-group record can describe observed safety events, but it cannot by itself establish how those events compare with what would have occurred without recombinant EphB4-HSA or with a different regimen. The registry record does not report comparative efficacy results, nor does it report the primary toxicity endpoint values.

3. Trial Design

The registry classifies NCT03146871 as a phase 2, single-group study with allocation listed as not applicable and no masking. The planned treatment combined recombinant EphB4-HSA fusion protein with a hypomethylating agent, either azacitidine or decitabine. Enrollment is reported as 7. The sponsor is the University of Southern California. The trial status is terminated; the reason for termination is not reported in the registry record.

The treatment description specifies recombinant EphB4-HSA by intravenous infusion over 60 minutes at 15mg/kg on days 1 and 15 of a 28-day cycle. Azacitidine is described at 75mg/m2, administered intravenously or subcutaneously on days 1-7 or days 1-5 and 8-9 of a 28-day cycle. The listed decitabine regimen is incomplete in the registry record, so its dose and full schedule are not reported. The record does not say how the choice between azacitidine and decitabine was made.

The primary outcome is titled “Assessment of Toxicity to Hypomethylating Agent (HMA).” Toxicity was to be graded according to NCI CTCAE v4.0 after each cycle, over a timeframe of up to 13 months, described as up to 12 cycles plus 30 days from treatment start. The stated safety analysis includes participants who completed 2 cycles. The registry record does not report the number who met that population definition or the toxicity outcome counts and grades for that endpoint.

The reported serious adverse-event records list 7 participants at risk and 7 affected. These counts are safety information, not a control-group comparison and not, by themselves, the primary toxicity endpoint. The registry record does not report randomization, blinding, a separate comparator, or a statistical analysis plan. It also does not report why the study was terminated or whether termination affected the planned endpoint assessment.

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: The study has no pre-assignment. All participants were given the same treatment.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Safety analysis includes all participants who completed 2 cycles of treatment.

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 Assessment of Toxicity to Hypomethylating Agent (HMA), measured over Up to 13 months (up to 12 cycles + 30 days) from the start of treatment..

The primary endpoint is an assessment of toxicity associated with hypomethylating-agent treatment during the combination regimen. The registry says toxicity was assessed and graded using NCI CTCAE v4.0 after each cycle. The stated timeframe extends up to 13 months from treatment start, described as up to 12 cycles plus 30 days. The safety analysis population is defined as participants who completed 2 cycles of treatment.

Graded toxicity is useful in a safety study because it organizes adverse events by severity using a stated grading framework, rather than treating every event as equivalent. Assessing events after each cycle also describes the timing of planned monitoring. These are general features of toxicity assessment; the registry record does not report the actual grade distribution, the number of participants meeting a toxicity threshold, or the endpoint result.

The endpoint description names toxicity to the hypomethylating agent, while the treatment regimen combines that agent with recombinant EphB4-HSA. In a single-group combination study, an observed toxicity cannot be attributed to one component from the registry information alone. The record does not provide a separate assessment that distinguishes toxicity associated with azacitidine, decitabine, or recombinant EphB4-HSA.

The serious adverse-event records are reported separately from the primary endpoint description. They document serious-event counts and listed terms, but they do not substitute for the cycle-based CTCAE outcome. The primary endpoint's actual findings, including the analysis denominator and toxicity grades, are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical distinctions help interpret this record. They are general principles applied to the design and reported information; the registry record does not provide a full statistical analysis plan.

What does a single-group design allow?

A single-group study observes participants receiving the study regimen without a concurrent control group. It can describe which outcomes occurred in the observed group, how often they were recorded, and, when reported, their timing or severity. It cannot directly estimate a between-group treatment contrast because there is no separate group receiving a comparator regimen under the same study conditions. In this trial, allocation is listed as not applicable and masking as none. The record therefore offers no randomized comparison that could separate the effect of the combination from other influences.

This limitation matters for safety as well as efficacy. Toxicities may arise from a study treatment, the underlying disease, prior treatment, or other causes. A count of participants with a serious event does not identify which cause applies to each event. The registry record does not report attribution details that would allow those possibilities to be distinguished.

What is the role of the primary toxicity endpoint?

The primary endpoint uses toxicity grades assessed according to NCI CTCAE v4.0 after each cycle. A grading system gives structured categories for describing the severity of events; it is not the same as a comparison of treatment groups. The stated primary safety population includes participants who completed 2 cycles. The record does not report how many met that definition, the grades observed, or a summary measure for the endpoint. Consequently, the primary endpoint cannot be reconstructed from the available registry results.

Cycle-based assessment also raises a general issue of exposure time. Participants who remain on treatment for different lengths of time may have different opportunities for events to be detected. A count of affected participants does not by itself describe the amount of treatment exposure or the timing and recurrence of events. The registry's serious-event summary does not report person-time or an event-by-cycle breakdown.

How should counts and percentages be read?

The adverse-event record lists the number at risk and the number affected, alongside percentages for selected serious-event terms. Counts are direct descriptions of recorded participants; percentages provide a scale-free way to express occurrence in relation to a denominator. The listed serious terms include platelet count decreased and white blood cell decreased at 42.9% each, neutrophil count decreased at 28.6%, and febrile neutropenia and hypertension at 14.3% each. Those percentages should be quoted as recorded rather than treated as precise estimates of risk in a broader population.

In general, estimates based on a small group can vary substantially if the experience of only a few participants differs. Confidence intervals express uncertainty around an estimate under a specified statistical model. No confidence intervals are reported for the adverse-event percentages in the registry record. Their absence means precision is not quantified in the posted safety summary; it does not mean that the percentages are exact population risks.

What can a p-value or confidence interval show?

In general, a p-value evaluates how compatible observed data are with a specified statistical null model, while a confidence interval describes a range of values compatible with the data and method at a stated confidence level. Neither measure alone establishes causality or clinical importance. No p-value, confidence interval, or comparative effect estimate for the primary outcome is reported in the registry record. The absence of a reported comparison is especially important because the study has no separate control group.

What does over-dispersion mean for counts?

As a general statistical concept, over-dispersion occurs when count data vary more than a simple model assumes. It can make standard errors too small if the model does not allow for that extra variation. This can matter when analyzing repeated or clustered events, such as multiple events occurring in the same participant. The registry record does not report a count model, a repeated-event analysis, or an assessment of over-dispersion; no such method should be presumed.

How should planning calculations be distinguished from results?

The record describes a sample-size planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. Such calculations are planning illustrations, not observed power and not evidence that the study achieved a particular power. The record does not report a trial-specific sample-size justification or a comparative observed effect on which to base a substantive conclusion. Because this study is single-group, the stated equal-allocation assumption for the illustration should not be confused with the actual trial design.

More generally, a sample-size calculation depends on assumptions about the target effect, variability or event rate, analysis, and design. Changing those assumptions changes the calculated sample size. A planning result does not establish that an estimate is precise, that an endpoint was completed, or that a treatment effect exists. The registry record does not report interim analyses, a stopping rule, or details of the statistical methods used for the posted outcomes.

What analysis details remain unclear?

The registry record does not report how missing toxicity assessments were handled, how participants who did not complete 2 cycles were treated in the primary analysis, or whether event grades were summarized by maximum severity, cycle, or another method. It also does not report adjustment for multiple outcomes or subgroup analyses. These details can affect how endpoint summaries are interpreted. They should be treated as unknown rather than inferred from the presence of an endpoint description or adverse-event listing.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 7 outcome measures but does not report a statistical analysis plan explaining how multiple outcomes were handled. It does not report whether a particular outcome hierarchy, adjustment for multiple testing, or separate confirmatory analysis was planned. When several outcomes are tested, the chance of at least one apparently notable result can increase; this is a general statistical concern. No claim about the trial's approach to that issue can be made from the registry information.

Missing-data handling is also not reported. In particular, the registry defines the primary safety analysis population as participants who completed 2 cycles, but does not report how many participants completed that amount of treatment, how participants outside that population were considered, or whether any toxicity assessments were missing. It does not describe an imputation method or a sensitivity analysis. The serious adverse-event section reports 7 participants at risk and 7 affected, but that safety summary does not resolve the missing details about the primary endpoint population.

The record does not report whether analysis was based on all enrolled participants, a treated population, or another population for endpoints other than the stated safety analysis. It also does not report the statistical method used to summarize graded toxicity. Those omissions prevent a full assessment of the analysis denominators and the consequences of incomplete follow-up.

General statistical interpretation depends on knowing who was included, what information was missing, and how outcome multiplicity was addressed. Here, those procedures are not reported in the registry record. The posted safety counts can be described as listed, but they do not fill those gaps or establish the results of the primary toxicity analysis.

11. 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

Serious events with the highest arm-level rate: platelet count decreased (42.9%), white blood cell decreased (42.9%), neutrophil count decreased (28.6%), febrile neutropenia (14.3%), hypertension (14.3%).

The serious adverse-event summary reports 7 participants at risk and 7 affected in the treatment group. The registry lists platelet count decreased and white blood cell decreased at 42.9% each, neutrophil count decreased at 28.6%, febrile neutropenia at 14.3%, and hypertension at 14.3% among the top serious terms. These figures should be read as the registry's event summary, not as a comparison with an untreated or alternative-treatment group.

The record also reports 7 at risk and 0 affected in a separate other-adverse-event field, and lists 1 death. It does not state the cause or timing of the death, whether it was considered treatment-related, or how it relates to the serious-event listing. No causal attribution should be inferred from the counts alone.

Serious adverse events are not interchangeable with the primary outcome, which was a cycle-based assessment of toxicity graded under NCI CTCAE v4.0. The record does not report the primary endpoint's grades or analysis denominator. The percentages for the serious terms should therefore not be taken as a complete summary of all toxicity or as a substitute for the unreported primary endpoint results.

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

The registry results establish that the study recorded serious adverse events in the treated group: 7 participants were listed as at risk and 7 as affected. They also identify selected serious terms and report 1 death. These are descriptive safety observations from a single-group study. They do not show what the event experience would have been under a different regimen or without the study treatment.

The central primary outcome was an assessment of toxicity after each cycle, with grading according to NCI CTCAE v4.0 and a stated safety population of participants who completed 2 cycles. The registry record does not report the resulting toxicity grades, the number included in that analysis population, or a summary of the primary endpoint. Thus the posted serious-event data do not answer the primary endpoint question by themselves.

The record also does not report comparative efficacy results. The study's single-group design, absence of a listed comparator, and lack of a reported treatment-effect estimate mean that the record cannot establish whether the combination changes outcomes relative to another treatment. Nor can event counts alone establish that an event was caused by a particular treatment component.

These limits do not make the safety listing meaningless: it documents the serious-event experience recorded for the treatment group and identifies the terms appearing in the registry. But interpretation beyond those descriptive details requires details that are not reported, including event attribution, follow-up completeness, primary endpoint results, and the reason for termination. Any broader conclusion about comparative benefit, causal safety, or treatment effectiveness would go beyond the registry record.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of the difference between a safety description and a comparative treatment estimate. A single-group study can document adverse events during exposure to a regimen, but it does not provide a concurrent control group for judging whether those events are more or less frequent than under another treatment. That distinction is important when interpreting counts that may appear precise but have no internal comparator.

The study also illustrates why the endpoint definition and the reported results need to be considered separately. The registry specifies cycle-based toxicity grading and a safety analysis population based on completing 2 cycles. However, the primary toxicity results and the size of that analysis population are not reported. Serious adverse-event counts are available, but they do not substitute for the planned graded endpoint. Knowing the endpoint is not the same as knowing its result.

Small enrollment and incomplete outcome reporting make careful use of denominators especially important. The adverse-event summary lists participants at risk and affected, while the primary endpoint refers to a defined subset. The record does not establish that these populations are identical. Keeping those denominators distinct prevents the serious-event data from being mistaken for a complete account of the primary analysis.

Finally, the record shows why status and design details should not be overinterpreted. The study is listed as terminated, but the reason is not reported. Without the endpoint results, analysis methods, and termination context, statistical conclusions should remain limited to the outcomes explicitly recorded. This is a useful case in separating documented observations from unanswered questions.

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 registry record has important trial-specific reporting limitations. Enrollment is reported as 7, but the number included in the primary safety population is not reported. That population is defined as participants who completed 2 cycles, and the record does not state how many completed those cycles or whether participants who did not complete them contributed to other safety summaries.

The primary endpoint result is not reported. Although the record describes CTCAE v4.0 grading after each cycle and a timeframe of up to 13 months, it does not provide grades, endpoint counts, or a summary estimate. The listed serious adverse events cannot fill this gap because they are a separate safety summary.

The study has no separate comparator group, and the registry lists no random allocation or masking. Consequently, the observed events cannot be compared within the study with outcomes under another regimen. The record does not report attribution of events to individual components, the reason for termination, or details of death timing and cause. These omissions constrain causal and comparative interpretation.

The intervention record is also incomplete for decitabine: its full dose and schedule are not reported. The record does not report how participants were assigned to azacitidine or decitabine, or outcomes by agent. Statistical analysis details—including missing-data procedures, multiplicity handling, and methods for summarizing toxicity—are not reported. These uncertainties should be kept distinct from the serious adverse-event counts that are explicitly listed.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Recombinant EphB4-HSA Fusion Protein and Azacitidine or Decitabine for Relapsed or Refractory Myelodysplastic Syndrome, Chronic Myelomonocytic Leukemia, or Acute Myeloid Leukemia Patients Previously Treated With a Hypomethylating Agent — NCT03146871.
• Registry record last updated 2026-06-08; status last verified 2025-12.
• Results first posted 2026-06-08.
• 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

NCT03146871 is a phase 2, single-group, unmasked pilot/safety study sponsored by the University of Southern California. It enrolled 7 participants and is listed as terminated; the reason for termination is not reported. The study evaluated recombinant EphB4-HSA fusion protein with azacitidine or decitabine in patients with relapsed or refractory myelodysplastic syndrome, chronic myelomonocytic leukemia, or acute myeloid leukemia after prior hypomethylating-agent treatment.

The primary outcome assessed toxicity after each cycle using NCI CTCAE v4.0, with a stated safety population of participants who completed 2 cycles. The primary endpoint results are not reported in the registry record. The serious adverse-event summary lists 7 at risk and 7 affected, selected serious-event terms, and 1 death. With no separate comparator and no reported primary toxicity results, the posted information supports descriptive reading of the listed safety events rather than a comparative treatment conclusion.