Study of AMG 330 in Combination With Pembrolizumab in Adult With Relapsed or Refractory Acute Myeloid Leukemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of AMG 330 in Combination With Pembrolizumab in Adult With Relapsed or Refractory Acute Myeloid Leukemia (NCT04478695), a non-randomized phase 1 study. The primary outcome is Number of Participants Who Experienced Dose-limiting Toxicities (DLTs).

2. Clinical Question

The study asked whether AMG 330 administered with pembrolizumab could be evaluated for safety and tolerability in adults with relapsed or refractory acute myeloid leukemia (R/R AML). The registry’s primary objective is safety and tolerability, with dose-limiting toxicities (DLTs) as the primary outcome. The study also listed remission, response duration, and pharmacokinetic outcomes.

3. Trial Design

The registry describes a Phase 1 study, while its official title calls it a Phase 1b study. It was non-randomized, unmasked, and used a sequential model. Its stated purpose was to assess safety, tolerability, pharmacokinetics, pharmacodynamics, and efficacy of AMG 330 with pembrolizumab in adults with R/R AML. The primary objective was safety and tolerability.

The registry records enrollment of 1 participant. Cohort 1 enrolled 1 participant, while Cohort 2 enrolled 0 participants. The study status is terminated. The recorded start date is 2020-09-29, and the primary completion date is 2020-10-23. The record does not explain the reason for termination or describe the circumstances of the enrollment.

Because allocation was non-randomized, the cohort labels should not be read as randomized treatment arms. The registry does not report a control intervention, and the cohort descriptions are blank. It therefore does not establish a controlled comparison of AMG 330 plus pembrolizumab against another treatment or against no treatment. The meaning of the sequential model for dose assignment or cohort progression is not reported in the registry record.

Dose levels, administration schedule, detailed eligibility criteria, and any dose-escalation rules are not reported in the registry record. Nor is a protocol-level statistical analysis plan reported. These omissions limit what can be said about the intended conduct of the study beyond its registered design descriptors and stated objectives.

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: Screening tests and procedures were performed within the 14 days preceding administration of the first dose.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Measured in the Dose-limiting Toxicity Evaluable Analysis Set, which was defined as DLT-evaluable participants in the Safety Analysis Set. The study was terminated with only 1 participant enrolled into Cohort 1 (0 participants were enrolled into Cohort 2). No data is reported here to maintain participant confidentiality.

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: The study was terminated with only 1 participant enrolled into Cohort 1 (0 participants were enrolled into Cohort 2). No data are reported here to maintain participant confidentiality.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants Who Experienced Dose-limiting Toxicities (DLTs), measured over 28 days.

The primary outcome is the number of participants who experienced dose-limiting toxicities (DLTs), assessed over a 28-day timeframe. The registry defines a DLT as an event during the DLT window unless it is clearly attributable to causes other than the investigational product. The listed events include any treatment-related death, Grade 4 neutropenia, Grade 3-4 non-hematologic toxicity not clearly resulting from the underlying leukemia (with some exceptions), cytokine release syndrome, and any Grade 5 toxicity.

DLT outcomes are commonly used in early-phase studies to characterize treatment-related toxicity within a specified observation window. A defined set of qualifying events makes the outcome more interpretable than an unspecified count of adverse events. The DLT-evaluable analysis set is described as DLT-evaluable participants in the Safety Analysis Set; further criteria for being DLT-evaluable are not reported in the registry record.

The registry gives no participant-level DLT information and no count for this primary outcome. It says no data are reported for the measure. Consequently, the endpoint definition can be described, but the number of participants meeting it cannot be assessed from the registry results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record supplies design descriptors and endpoint definitions, but does not report a statistical analysis plan or outcome analyses. The explanations below provide general statistical context; they do not describe analyses performed for this study unless explicitly identified as the registry entries.

What does a DLT count capture?

A count of participants experiencing DLTs records whether participants met a prespecified toxicity definition during the observation window. The definition matters because it determines which events count toward the endpoint. Here, the registry lists treatment-related death, specified grades of neutropenia and non-hematologic toxicity, cytokine release syndrome, and Grade 5 toxicity, with stated qualifications for attribution and exceptions. A DLT count is not interchangeable with a count of all adverse events: an adverse event may not meet the DLT criteria, and the registry does not indicate which events, if any, met them.

Why does the analysis population matter?

An endpoint result is meaningful only in relation to the people included in its analysis. The registry describes the primary endpoint population as the DLT-evaluable analysis set, defined as DLT-evaluable participants in the Safety Analysis Set. It does not report the criteria for DLT evaluability, the size of the analysis set, or whether the enrolled participant was included in it. Enrollment and endpoint analysis populations are not necessarily identical in a trial; reasons for any difference should be specified rather than assumed. Those details are not reported in the registry record.

What does a confidence interval add?

General statistical context: a confidence interval communicates uncertainty in an estimated quantity, such as a proportion, under the model and sampling assumptions used. It is not a guarantee about the result for an individual, and it does not correct for limitations in study design or incomplete reporting. No DLT estimate or confidence interval is reported, so readers cannot assess the precision of a toxicity estimate. With no primary outcome result, there is also no observed effect from which to form a recalculated estimate.

Why are randomization and a comparator relevant?

Random allocation can help balance measured and unmeasured factors between groups, while a concurrent comparator can provide a reference for interpreting differences. These are general principles, not claims about a planned analysis in this study. The registry labels this study non-randomized and does not describe a control group. It records 1 participant in Cohort 1 and 0 in Cohort 2. The cohort structure therefore does not provide the basis for a like-for-like controlled comparison, and the record does not report a comparative treatment estimate.

What can the planning assumptions establish?

The registry’s planning illustration is described as using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; it describes the target effect as a stated fraction of the observed effect. For time-to-event outcomes, the illustration is described as giving total events, and for other outcomes as participants per arm. These are planning assumptions, not evidence that the trial used those methods in its own design or analysis. The record does not report a sample-size justification or an observed effect for the primary endpoint. Any sample-size illustration based on an observed effect is not observed power and cannot replace a missing endpoint result.

What is known about the analysis process?

The registry reports 0 analyses. It does not report an analysis method for the primary outcome, an estimation procedure, a confidence interval, a hypothesis test, adjustment for covariates, or interim analyses. It also does not report a detailed statistical plan. In the absence of those details and the endpoint data, no recalculated result can be made from the posted summary information. There is no reported sponsor analysis to compare with a recalculation, and no discrepancy between a reported analysis and a recalculation can be assessed.

General statistical context: sparse data can make estimates unstable and uncertainty wide, while missing outcome data can prevent even a basic estimate. Methods for handling missing observations require information about why values are missing and what assumptions are appropriate. The registry record does not provide individual DLT observations or a missing-data method. Any account of such a method would therefore be speculative.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary DLT outcome and secondary outcomes that include complete remission without minimal residual disease, complete remission, complete remission with incomplete recovery, morphological leukemia-free state, partial remission, duration of response, plasma concentration of AMG 330, and half-life of AMG 330. It reports 15 outcome measures overall and 0 analyses. The outcomes listed indicate a broad set of safety, response, and pharmacokinetic interests, but the record does not report results for them in the available results information.

Multiplicity arises when multiple outcomes or statistical tests are considered: interpreting many tests can increase the chance of a seemingly notable finding that is not robust. General approaches include prespecifying a primary comparison or adjusting inference across tests. The registry record does not report a multiplicity strategy, an analysis hierarchy for the secondary outcomes, or results of statistical tests. No adjustment method should be inferred.

The primary endpoint is defined for a DLT-evaluable analysis set within the Safety Analysis Set, but the registry does not provide the analysis-set size, participant-level endpoint data, or reasons for any exclusions. It reports enrollment of 1 participant in Cohort 1 and 0 in Cohort 2, but does not establish whether the enrolled participant was DLT-evaluable. No missing-data handling method is reported.

The registry says no data are reported for the primary measure. That absence prevents assessment of endpoint completeness and prevents calculation of a DLT estimate. The adverse-event summary’s death entry does not resolve the missing primary endpoint because its treatment relationship, timing, and DLT classification are not reported. Thus neither multiplicity handling nor missing-data assumptions can be evaluated from the registry record.

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

The registry establishes that this was a terminated, non-randomized study of AMG 330 with pembrolizumab in adults with R/R AML, with a primary focus on safety and tolerability. It defines the DLT endpoint and a 28-day timeframe, and records enrollment of 1 participant in Cohort 1 and 0 in Cohort 2. These details describe the study’s scope and registered outcome, not the result of that outcome.

The record does not report the number of participants who experienced DLTs. It therefore does not establish whether the enrolled participant experienced a DLT, and it cannot support an estimate of DLT frequency or precision. The separate adverse-event summary records 1 death in Cohort 1, but does not report whether that death was treatment-related, occurred in the DLT window, or met the DLT definition. It cannot be interpreted as a DLT result on its own.

These data also do not show whether the combination is effective or ineffective. The registry lists remission and pharmacokinetic outcomes, but their results are not reported in the record information. The study has no described randomized comparator, and Cohort 2 has no enrolled participants, so a comparative treatment effect cannot be inferred. The termination status alone does not explain why the study ended or what the outcome would have been under continued enrollment.

General statistical context: absence of a reported estimate is not evidence of no effect or no toxicity. It means the relevant quantity cannot be evaluated from the reported results. The same distinction applies to the listed secondary outcomes: their presence in the registry indicates that they were specified as outcomes, not that results were observed or demonstrated. The conclusions should remain bounded by the available information.

12. Why This Trial Matters Statistically

This record illustrates an important distinction in early-phase trial interpretation: a carefully defined safety endpoint is not the same as a reported safety result. The registry specifies DLT criteria and a 28-day timeframe, which identify what the primary outcome was intended to capture. But it reports no value for that outcome. Without the endpoint result, the definition cannot answer the clinical or statistical question by itself.

It also shows why enrollment and analysis populations need to be distinguished. The registry records 1 participant in Cohort 1 and 0 in Cohort 2, while the primary outcome is defined for a DLT-evaluable analysis set. The record does not state whether the enrolled participant qualified for that set. Interpreting a result requires knowing both the outcome and who contributed to it.

The non-randomized design and absence of a described control group constrain comparative interpretation. Cohort labels should not be treated as evidence of a controlled contrast. This matters particularly when outcomes are intended to describe safety or efficacy: an observed outcome, even if reported, would not automatically establish how it compares with another treatment or with a different course of care.

Finally, the adverse-event summary’s death entry underlines the need to distinguish event categories. A death is not automatically a DLT; its timing and attribution relative to the endpoint definition matter, and those details are not reported. Statistical interpretation depends on definitions, analysis populations, and transparent results together. In this record, the endpoint definition is available, but the primary outcome result and the details needed to interpret the death entry are not.

13. 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 principal limitation is that the registry reports no result for the primary DLT outcome. Without that result, there is no estimate to interpret and no way to assess its precision. The enrollment record is also very limited: 1 participant is recorded in Cohort 1 and 0 in Cohort 2. The primary endpoint’s analysis population is defined, but its size and relationship to enrollment are not reported.

The study was non-randomized and does not describe a control group. The registry’s sequential model and cohort names do not provide enough detail to infer how participants were assigned, how doses were selected, or how cohorts differed. Cohort descriptions are blank. Dose levels, treatment schedule, and detailed eligibility criteria are not reported in the registry record.

The adverse-event summary records 1 death in Cohort 1, but does not report its cause, treatment relationship, timing, or DLT classification. It therefore cannot resolve the missing primary endpoint. The record also lists no top serious terms. The meaning of the zero counts in the adverse-event categories cannot be evaluated without additional information on event collection and the applicable analysis population.

The registry reports 0 analyses and does not provide a statistical analysis plan, multiplicity strategy, missing-data approach, sample-size justification, interim-analysis information, or the reason for termination. Secondary outcomes are listed, but results are not reported in the record information. These gaps limit assessment of both the study’s conduct and its statistical findings.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Study of AMG 330 in Combination With Pembrolizumab in Adult With Relapsed or Refractory Acute Myeloid Leukemia — NCT04478695.
• Registry record last updated 2024-03-08; status last verified 2023-06.
• Results first posted 2024-03-08.
• 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.

17. Record Summary

The registry record describes NCT04478695, a Phase 1 study whose official title identifies it as Phase 1b. It evaluated AMG 330 with pembrolizumab in adults with relapsed or refractory acute myeloid leukemia. The design was non-randomized, unmasked, and sequential, with a stated primary objective of assessing safety and tolerability.

The primary outcome was the number of participants experiencing DLTs during a 28-day timeframe. The registry records enrollment of 1 participant in Cohort 1 and 0 in Cohort 2, and the study status is terminated. It reports no primary endpoint result and lists 0 analyses. A separate adverse-event summary records 1 death in Cohort 1, without details establishing whether it met the DLT definition.

Remission, response-duration, and pharmacokinetic outcomes are also listed, but their results are not reported in the registry record information. The record does not establish a controlled comparison or explain the reason for termination.