A Study of Bomedemstat (IMG-7289/MK-3543) With and Without ATRA, in Participants With Advanced Myeloid Malignancies (IMG-7289-CTP-101/MK-3543-001): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Bomedemstat (IMG-7289/MK-3543) With and Without ATRA, in Participants With Advanced Myeloid Malignancies (IMG-7289-CTP-101/MK-3543-001) (NCT02842827), a non-randomized phase 1, phase 2 study. The primary outcome is Number of Participants Who Experienced an Adverse Event (AE).

2. Clinical Question

The study asks about the safety and tolerability of bomedemstat, given alone or with tretinoin, in people with advanced myeloid malignancies. Its stated interests also include pharmacokinetics and pharmacodynamic effects across bomedemstat doses and treatment durations. The primary endpoint is a count of participants who experienced an adverse event during study participation.

The question is not framed as a confirmatory test of whether bomedemstat, with or without tretinoin, improves a disease outcome relative to a control. The registry describes a safety and dose/duration investigation, with pharmacokinetic and pharmacodynamic interests. Treatment cohorts therefore provide information about participants observed under different regimens, rather than a randomized estimate of comparative treatment effect.

3. Trial Design

This is a completed, multicenter, open-label study with Phase 1 and Phase 2 components. The registry records a parallel model, non-randomized allocation, and no masking. It lists 45 participants enrolled. The study began on 2016-10-06, and primary completion was recorded as 2018-10-30.

The listed cohorts cover bomedemstat monotherapy at 0.75 mg/kg/day, 1.5 mg/kg/day, 3 mg/kg/day, and 6 mg/kg/day. Other cohorts combine bomedemstat at 3 mg/kg/day or 6 mg/kg/day with tretinoin 45 mg/m^2/day. Regimens include 14-day cycles with 7 days of treatment and 7 days of rest; the record also describes a 21-day cycle with 14 days of treatment and 7 days of rest, and a 28-day cycle with 21 days of treatment and 7 days of rest. The number of cycles and treatment duration differ across the listed cohorts. For the 28-day regimen, the record states that participants receive at least 1 cycle and that subsequent cycles are at the investigator’s discretion; further details are not reported in the registry record.

Because treatment assignment was not randomized and the study was open label, cohort membership may reflect features of the study’s dose exploration rather than comparable groups created to estimate a treatment contrast. The cohort descriptions do not establish that participants were exchangeable across regimens. Consequently, differences in observed event counts across cohorts should be read as descriptions of those cohort records, not as evidence that a regimen caused more or fewer events.

The registry summary identifies safety and tolerability, pharmacodynamic effects, and pharmacokinetics as study interests. The primary outcome record specifically defines an adverse event and reports the number of participants who experienced one. Detailed information on the dose-escalation procedure, cohort assignment rules, eligibility criteria, and sample-size justification 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.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants who received ≥1 dose of study 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.

Population: All participants who received ≥1 dose of study treatment

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Who Experienced an Adverse Event (AE), measured over Up to approximately 24 months.

The primary endpoint is the number of participants who experienced an adverse event, reported for participants who received at least 1 dose of study treatment. The registry defines an adverse event broadly as an undesirable physical, psychological, or behavioral effect during participation in an investigational study, in conjunction with use of the drug or biologic, whether or not related to study treatment. The time window runs from the first dose until completion of the study, with the endpoint timeframe described as up to approximately 24 months.

This kind of endpoint is used to describe the extent to which participants experienced events while enrolled and receiving study treatment. A participant count is a binary outcome at the individual level: for this endpoint, a participant is counted as having experienced an adverse event or not. It is different from counting the total number of event episodes, since a participant may experience more than one event and the endpoint description focuses on participants.

The broad definition is important for interpretation. An event’s occurrence during the study does not establish that study treatment caused it. The endpoint definition explicitly includes events whether or not they are related to the study drug. The record does not provide an adjudicated causal estimate or a comparison with an untreated or randomized control group.

The endpoint is appropriate for reporting observed safety experience in a dose-exploration study, but by itself it does not characterize severity, duration, or clinical consequences for every event. Those details require other outcome definitions and results. The registry also lists serious adverse event information separately; those counts should not be treated as interchangeable with the primary count of participants with any adverse event.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a count-based, binary participant outcome. Each participant contributes to the count if an adverse event was recorded during the specified observation period. This is distinct from an event-rate analysis, which would require information about the amount of time each participant was observed, and from an event-count analysis, which could count multiple episodes for one participant. The primary endpoint description identifies the number of participants with an event, not total event episodes.

What does a participant count tell the reader?

A participant count paired with a denominator describes how many people in a defined group met the outcome definition. It is a direct description of the observed study group, not necessarily a reliable estimate of what would occur in a broader population. General statistical principle: estimates based on fewer observations tend to be less precise, and a small change in the observations can make a larger difference to the apparent proportion than it would in a much larger sample. No confidence interval for the primary endpoint is reported in the registry record.

Counts also depend on outcome definitions. The registry defines an adverse event broadly and includes events regardless of whether they were considered related to study drug. Thus, the primary endpoint combines different kinds of events and does not separate treatment-related events from unrelated events. The serious adverse event listing addresses a more specific category, but still does not by itself establish causality.

Why do cohort comparisons need caution?

In a randomized study, random assignment is intended to make treatment groups comparable on average, including for factors that may affect outcomes. This study is explicitly non-randomized. General statistical principle: when groups are formed without random assignment, observed differences can reflect treatment, differences among participants, or other differences in care and observation. The cohort counts therefore describe each group as observed; they do not isolate the effect of a dose or combination.

The parallel structure means that multiple cohorts are represented, but parallel cohorts are not automatically equivalent to randomized arms. Here, the regimens differ in dose, cycle structure, and use of tretinoin. The registry record does not report an adjustment model or an analysis designed to account for differences between cohorts. A like-for-like comparative interpretation is not justified by the design information recorded.

What would uncertainty measures add?

A confidence interval is a range of values compatible with the observed data under a specified statistical model and method. General statistical principle: it expresses uncertainty around an estimate, rather than a guarantee that the true value lies in the interval. For a binary participant outcome, an interval around a group proportion would help communicate how imprecisely that proportion is estimated. A comparative interval would additionally require a defined contrast and analysis. Neither a confidence interval nor a comparative estimate for the primary endpoint is reported in the registry record.

A p-value, when used, summarizes how compatible data are with a specified null hypothesis under the assumptions of a test. It is not the probability that the null hypothesis is true and it does not measure the size or clinical importance of an effect. No primary-endpoint p-value is reported in the registry record. In a non-randomized study with small, unequal cohorts, a p-value alone would not resolve confounding or establish causation.

Why can sample-size illustrations be misleading?

Sample-size calculations depend on an assumed effect, outcome variability, significance threshold, allocation, and statistical model. The posted planning description states a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; it also describes detecting a stated fraction of an observed effect. These are planning assumptions, not evidence that the study was designed or conducted as a randomized two-arm comparison. A planning illustration is not observed power, and it cannot turn non-randomized cohort data into a controlled treatment-effect estimate.

General statistical principle: power is a design property calculated under assumptions about the effect and data-generating process. Once study results are observed, a retrospective power calculation does not add evidence beyond the estimate and its uncertainty. The registry’s protocol-specific sample-size justification is not reported in the registry record. The numerical outputs of any planning illustrations are not necessary for interpreting the posted adverse-event counts.

Why does over-dispersion matter for some outcomes?

General statistical principle: count outcomes are sometimes modeled using a probability distribution that assumes a particular relationship between the mean and variability. Over-dispersion means that observed variability is greater than that model expects. If ignored, standard errors may be too small and uncertainty may be understated. This concept is relevant to analyses of event counts, but the primary endpoint is the number of participants with an event, not the total number of events. The registry does not report a model for over-dispersion or enough detail to determine whether such modeling was relevant to other analyses.

For binary outcomes, the main modeling issues instead include the definition of the analysis population, the handling of missing outcome status, the choice of effect measure, and adjustment for design features or confounders. The registry record does not report the inferential model for the primary endpoint. Without those details, the safest statistical reading is descriptive: the posted counts document observed participants with events in each cohort, with no reported randomized treatment contrast.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort 1a: Bomedemstat 0.75 mg/kg/Day", "Cohort 1b: Bomedemstat 1.5 mg/kg/Day", "Cohort 1c: Bomedemstat 3 mg/kg/Day", "Cohort 1d: Bomedemstat 6 mg/kg/Day", "Cohort 1x3: Bomedemstat 3 mg/kg/Day Orally Plus Tr", "Cohort 1x6: Bomedemstat 6 mg/kg/Day Plus Tretinoin")
n      <- c(3, 6, 4, 6, 5, 9)
events <- c(3, 6, 4, 6, 5, 9)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort 1a: Bomedemstat 0.75 mg/kg/Day  3  3  100.0  29.2  100.0  43.9  100.0  39.8  99.4
ARM  Cohort 1b: Bomedemstat 1.5 mg/kg/Day  6  6  100.0  54.1  100.0  61.0  100.0  59.0  99.6
ARM  Cohort 1c: Bomedemstat 3 mg/kg/Day  4  4  100.0  39.8  100.0  51.0  100.0  47.8  99.5
ARM  Cohort 1d: Bomedemstat 6 mg/kg/Day  6  6  100.0  54.1  100.0  61.0  100.0  59.0  99.6
ARM  Cohort 1x3: Bomedemstat 3 mg/kg/Day Orally Plus Tr  5  5  100.0  47.8  100.0  56.6  100.0  54.1  99.6
ARM  Cohort 1x6: Bomedemstat 6 mg/kg/Day Plus Tretinoin  9  9  100.0  66.4  100.0  70.1  100.0  69.2  99.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 27 outcome measures and 6 analyses. Several listed secondary outcomes concern maximum percent change from baseline in blood counts, blasts, and hemoglobin after bomedemstat administration. The presence of multiple outcomes and analyses creates the general issue of multiplicity: when many hypotheses are tested, the chance of at least one apparently unusual result can increase if each test is assessed separately. Whether and how to control that issue depends on the planned hierarchy and analysis strategy.

A prespecified multiplicity procedure, such as a defined testing order or adjustment method, is not reported in the registry record. It is also not reported whether the analyses were exploratory or confirmatory, how endpoints were prioritized, or which comparisons were planned before data review. Therefore, a statistical significance claim across multiple outcomes cannot be reconstructed from the registry information described.

Missing-data methods are also not reported in the registry record. The record does not describe how participants with incomplete follow-up, unavailable measurements, or treatment discontinuation were handled in the analyses. The primary endpoint population is defined as all participants who received at least 1 dose of study treatment, but the endpoint-specific number analyzed and the handling of missing adverse-event status are not reported in the registry record.

Analysis populations beyond that primary safety definition are not reported in the registry record. Details such as intention-to-treat, per-protocol, pharmacokinetic, or pharmacodynamic analysis sets should not be assumed from the study’s phase or stated objectives. The registry’s definitions and reported counts support a descriptive reading, but not a full reconstruction of the statistical analysis plan.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 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 (50.0%), cellulitis (40.0%), ear pain (33.3%), malaise (33.3%).

The serious adverse event figures are participant counts by cohort. For example, the registry lists 3 of 3 participants with a serious adverse event in Cohort 1a and 5 of 6 in Cohort 1d. The other listed cohorts have 6 of 6, 4 of 4, 5 of 5, 9 of 9, 4 of 4, and 8 of 8 participants with serious adverse events, respectively. These denominators are small, so the counts should be read as descriptions of the participants enrolled in those cohorts, not stable estimates of population risk.

The record separately lists counts for other adverse events, as well as deaths. These categories should not be conflated: experiencing an adverse event, experiencing a serious adverse event, and dying are distinct recorded outcomes. The adverse-event definition includes events whether or not related to treatment. Attribution of the serious events or deaths to study treatment is not reported in the registry record.

Term percentages for serious adverse events are listed for febrile neutropenia, sepsis, cellulitis, ear pain, malaise, and pyrexia. The denominator for those percentages is not reported in the registry record. The values therefore cannot be treated as cohort-specific proportions or compared directly with the cohort counts. They also do not indicate the number of episodes per participant.

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

The posted data show that adverse events were recorded among participants across the listed treatment cohorts, with cohort-specific counts for serious and other adverse events. Those figures describe the safety experience captured during this study. They can help readers understand what was recorded and how the recorded burden varied across the cohort listings, while keeping the small denominators and differing regimens in view.

The results do not establish that bomedemstat, tretinoin, or their combination caused any particular adverse event. The primary definition explicitly includes events regardless of relation to study drug. Nor can the counts establish that one regimen is safer than another: allocation was non-randomized, masking was absent, the regimens varied, and a randomized control comparison is not reported.

The record also does not establish efficacy. While pharmacodynamic and pharmacokinetic effects were study interests and multiple secondary outcomes are listed, the registry record does not report results for those measures. No disease-response estimate or comparative clinical benefit is reported in the registry record.

Statistical precision is also limited by what is available. The record gives participant counts but does not report confidence intervals or a primary-endpoint comparative estimate. Term percentages lack a reported denominator. The counts can be interpreted as observed data, but they should not be generalized as precise estimates of future risk or treated as evidence of causal differences among regimens.

Thus, the appropriate conclusion is narrow: this completed, non-randomized study records adverse-event experience across different bomedemstat regimens, with some cohorts also receiving tretinoin. The data are descriptive and do not answer a controlled comparative question about treatment effects.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of the distinction between dose-exploration safety data and a comparative treatment-effect study. It includes multiple cohorts with different doses, schedules, and use of tretinoin. The design can describe what happened among participants treated under each regimen, but non-randomized assignment does not provide the balance that random allocation is intended to create.

The primary outcome is a participant-level safety count, an important distinction from a count of every event episode. In a study with a broad adverse-event definition, it is also important to separate recording an event from attributing it to study treatment. The registry’s definition explicitly allows events that are not considered related to the study drug. Reading the outcome as a causal treatment effect would therefore go beyond its definition and design.

The cohort sizes make denominator-aware interpretation essential. Counts such as all participants affected within a cohort can look definitive, but small groups provide limited precision and are sensitive to individual observations. Confidence intervals and an explicit comparison model could help quantify uncertainty, but they are not reported for the primary endpoint in the registry record.

Finally, the study has multiple outcomes and analyses, while the registry does not report how multiplicity, missing data, or analysis populations were handled. These omissions matter when interpreting any inferential claims. The record supports a careful descriptive account of observed safety experience, while highlighting why design, endpoint definition, uncertainty, and analysis planning all shape what can be concluded.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary limitation for interpretation is the non-randomized design. Cohorts may differ in participant characteristics or other factors, and the record does not describe methods to adjust for such differences. Their event counts are not like-for-like treatment comparisons. Open-label conduct also means masking was absent; details of any independent outcome review are not reported in the registry record.

The cohort denominators are small, making observed counts imprecise as estimates of broader risk. The primary endpoint covers adverse events over a timeframe of up to approximately 24 months, but cohort-specific follow-up duration and completeness are not reported in the registry record. Differences in observation time could affect the opportunity to record events.

The adverse-event definition includes events whether or not related to study treatment. The record does not provide an overall causal assessment, event duration, recurrence, or severity detail for each outcome. The denominator for the listed serious-event term percentages is not reported in the registry record, and the relationship of the recorded deaths to study treatment is not established.

Details needed to reproduce or fully evaluate the analysis are also not reported in the registry record, including the primary-endpoint statistical model, confidence intervals, missing-data handling, multiplicity procedure, and analysis populations beyond the stated safety population. The record does not report a protocol-specific sample-size justification or detailed rules for cohort assignment. These limitations constrain causal and comparative interpretation, but do not negate the descriptive value of the recorded cohort counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Bomedemstat (IMG-7289/MK-3543) With and Without ATRA, in Participants With Advanced Myeloid Malignancies (IMG-7289-CTP-101/MK-3543-001) — NCT02842827.
• Registry record last updated 2023-07-11; status last verified 2023-07.
• Results first posted 2023-07-11.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT02842827 is a completed, multicenter, open-label Phase 1 and Phase 2 study of bomedemstat, alone or with tretinoin, in advanced myeloid malignancies. The registry records non-randomized allocation, a parallel model, no masking, and 45 participants enrolled. The primary endpoint counts participants with an adverse event among those who received at least 1 dose of study treatment, over up to approximately 24 months.

Listed regimens vary by bomedemstat dose, tretinoin use, and treatment schedule. Cohort-level serious and other adverse-event counts and death counts are reported. Because the study was non-randomized and the cohort groups are not like-for-like controls, the counts are descriptive rather than causal comparisons. The registry record does not report a primary-endpoint comparative estimate or confidence interval, nor does it report results for the study’s stated pharmacokinetic and pharmacodynamic interests.