A Study Of PF-04449913 In Japanese Patients With Select Hematologic Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Of PF-04449913 In Japanese Patients With Select Hematologic Malignancies (NCT02038777), a non-randomized phase 1 study. The primary outcome is Number of Participants With Dose-limiting Toxicities (DLTs): Monotherapy Cohort.

2. Clinical Question

The study also evaluated safety, tolerability, efficacy, pharmacokinetics and pharmacodynamics. The registry lists 85 outcome measures, including pharmacokinetic measures for the monotherapy cohort. The primary endpoint, however, is specifically the count of participants with DLTs in that cohort. It should not be conflated with adverse-event counts across all cohorts or with measures of drug concentration.

3. Trial Design

This was a Phase 1, open-label, multi-center study with non-randomized allocation and a sequential model. The registry reports enrollment of 48 participants and a completed status. The recorded start date is 2014-03-25, and the primary completion date is 2021-02-12. These record-level dates describe the study timeline, not the timing of an individual participant’s treatment or follow-up.

The study included a monotherapy cohort and several combination cohorts. PF-04449913 was studied alone, with LDAC, with intensive chemotherapy using cytarabine and daunorubicin, and with azacitidine. The registry also lists a continuation cohort for one participant rolled over from another trial in the same project, and an expansion cohort combining PF-04449913 with LDAC to evaluate efficacy. These are different treatment settings, not randomized arms within a common comparison.

The intensive chemotherapy description specifies continuous PF-04449913 for 28 days and describes chemotherapy schedules, including 2-4 cycles of consolidation therapy. Those details describe that cohort’s listed regimen; the registry record does not establish that every cohort followed a common treatment schedule. The study’s sequential model and absence of masking are consistent with a cohort-based early-phase design, but the record does not report a detailed operational protocol or dose-escalation algorithm. That information is not reported in the registry record.

The record identifies the primary endpoint in the monotherapy cohort, while many of the listed regimens are combinations. Consequently, the primary endpoint should be understood in its stated cohort and not generalized automatically to the combination cohorts. No randomized comparison is available to separate treatment effects from differences in cohort composition or treatment context.

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: DLT evaluable analysis set included all enrolled participants who received at least 1 dose of study medication and who did not have major treatment deviations during first cycle (DLT observation period).

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: Safety analysis set included all enrolled participants who received at least 1 dose of study medication.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Dose-limiting Toxicities (DLTs): Monotherapy Cohort, measured over Day -5 up to Day 28 of Cycle 1 (33 days).

The primary endpoint is the number of participants with dose-limiting toxicities in the monotherapy cohort. The observation period runs from Day -5 through Day 28 of Cycle 1, a total of 33 days. The analysis population is described as participants who enrolled, received at least 1 dose of study medication, and had no major treatment deviations during the first-cycle DLT observation period.

The DLT definition covers specified toxicities and treatment-delivery problems. Examples include Grade 3 or higher non-hematologic toxicity, with stated exceptions; prolonged myelosuppression meeting the registry’s duration, blood-count and bone-marrow criteria; inability to deliver at least 80% of planned doses in a combination because of non-hematologic toxicity; and a delay of more than 28 days in the next scheduled cycle due to persisting non-hematologic toxicity. It also includes a specified process for confirming persistent Grade 3 QTc prolongation after correction of reversible causes.

As a general statistical concept, a DLT count can help describe how often participants meet prespecified tolerability criteria during an early treatment period. Interpretation depends on the definition, observation window and analysis population. The registry record provides those aspects but does not report the primary DLT count or its uncertainty. Therefore, the primary endpoint result cannot be quantified from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary endpoint as a count of participants with DLTs. For a binary participant-level event, the basic descriptive information is how many participants met the event definition and the number included in the relevant analysis population. The registry specifies the population criteria but does not report the resulting DLT count. Without that count, the event frequency cannot be described from the primary endpoint record.

What does a DLT count capture?

A DLT is not simply any adverse event. It is an event that meets the study’s stated criteria during a particular observation period. Here, the period extends from Day -5 through Day 28 of Cycle 1, or 33 days. The criteria include selected non-hematologic toxicities, prolonged myelosuppression, limits on delivering planned study doses, delays to a scheduled cycle, and a defined evaluation of persistent QTc prolongation. The registry also lists exceptions and additional conditions within these criteria. These details matter because changing the event definition or observation period can change which participants count as having a DLT.

The DLT-evaluable analysis set is defined by enrollment, receipt of at least 1 dose, and absence of major treatment deviations during the first-cycle observation period. This is not necessarily the same as everyone enrolled. The number included in that set is not reported in the registry record, and the number excluded for deviations or other reasons is also not reported. A transparent endpoint interpretation ordinarily distinguishes the number assessed from the number who met the event definition.

How should cohort counts be read?

The design is non-randomized and sequential, with no masking. Participants in the monotherapy, LDAC, intensive chemotherapy, azacitidine, continuation and expansion cohorts did not form a randomized comparison. Their treatment regimens and cohort descriptions differ. In general, random allocation helps create groups that can be compared without systematically assigning participants according to clinical or operational decisions. In a non-randomized design, cohort differences may reflect participant characteristics, treatment setting, timing or other factors as well as treatment. The record does not report adjusted comparisons to account for such differences.

Counts from separate cohorts are therefore descriptive. A larger count in one cohort cannot, on its own, demonstrate a higher underlying risk than in another cohort. The numbers at risk may differ, and the cohorts may not represent the same population or treatment context. The registry’s serious adverse-event counts should also be kept separate from the primary DLT endpoint because the listed serious-event category does not necessarily match the protocol-defined DLT criteria.

What does a confidence interval contribute?

As a general statistical concept, a confidence interval expresses uncertainty around an estimate under a specified statistical procedure and its assumptions. A narrow interval indicates greater precision for that estimate than a wide interval, but it does not remove design limitations such as non-random allocation. No confidence interval for the primary DLT result is reported in the registry record. It is therefore not possible to describe the precision of that endpoint estimate from this record.

What is a planning illustration?

Sample-size calculations can illustrate how many participants per arm might be needed under chosen assumptions to detect a specified effect with a chosen significance level and power. The planning assumptions listed for these calculations use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a fraction of an observed effect. These are planning assumptions, not evidence that the trial was randomized or designed with that comparison. The study itself is non-randomized, and its cohorts are not like-for-like controls. Any such calculation is a planning illustration and must not be described as observed power.

General power is the probability, under a particular model and assumed effect, of detecting that effect according to a specified testing procedure. It depends on the assumptions used and does not describe the probability that a result is true. A planning calculation cannot repair missing endpoint results or make non-comparable cohorts exchangeable. The trial’s own sample-size justification is not reported in the registry record.

What statistical plan is available?

The registry records 85 outcome measures and 1 analysis, but that information does not provide a full analysis plan. The statistical methods used for the primary DLT endpoint, any prespecified hypotheses, handling of repeated or exploratory analyses, or planned adjustment for multiple testing are not reported in the registry record. Those omissions limit what can be inferred about how the study’s outcomes were selected and analyzed.

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("Monotherapy Cohort: PF-04449913 25 mg", "Monotherapy Cohort: PF-04449913 50 mg", "Monotherapy Cohort: PF-04449913 100 mg")
n      <- c(3, 3, 6)
events <- c(0, 0, 0)
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  Monotherapy Cohort: PF-04449913 25 mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Monotherapy Cohort: PF-04449913 50 mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Monotherapy Cohort: PF-04449913 100 mg  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 85 outcome measures, spanning safety, efficacy, pharmacokinetics and pharmacodynamics, and records 1 analysis. The number of listed outcomes alone does not show how many statistical hypotheses were tested, whether all outcomes were analyzed, or whether a hierarchy of primary and secondary analyses was used. A multiplicity strategy, such as a prespecified testing order or adjustment for multiple comparisons, is not reported in the registry record.

As a general statistical concept, examining many outcomes can increase the chance of finding apparently notable results by chance if each is tested separately. The implications depend on the analyses actually planned and conducted. Because the record does not provide a complete statistical analysis plan, it is not possible to determine whether or how multiplicity was addressed.

The primary endpoint population is defined as enrolled participants who received at least 1 dose and had no major treatment deviations during the DLT observation period. The record does not report the number of participants who met those criteria, the number excluded, or the handling of missing endpoint assessments. Rules for missing data, incomplete follow-up, and participants who did not complete the observation window are not reported in the registry record.

The registry’s DLT-evaluable set therefore needs to be distinguished from the total enrollment of 48. The available information does not establish that every enrolled participant contributed to the primary endpoint analysis. Details of other analysis populations, including those used for efficacy or pharmacokinetic outcomes, are not reported in the registry record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: enteritis infectious (33.3%), disease progression (25.0%), cerebral haemorrhage (25.0%), acute kidney injury (25.0%), pneumonia (16.7%).

Serious adverse-event figures are reported by cohort as counts of participants at risk and participants affected. In the 25 mg monotherapy cohort, 1 of 3 participants was affected; in the 50 mg monotherapy cohort, 3 of 4 were affected; and in the 100 mg monotherapy cohort, 1 of 6 was affected. In the LDAC combination cohort, 1 of 6 was affected; in the intensive chemotherapy cohort, 4 of 6 were affected; and in the azacitidine combination cohort, 1 of 6 was affected. The continuation cohort reports 0 of 1 affected, while the LDAC expansion cohort reports 9 of 15 affected.

The record separately lists deaths by cohort: 0, 2, 0, 5, 2, 4, 0 and 11, respectively, in the order. These are counts, not statements of cause or treatment attribution. The record does not report whether the deaths were considered related to treatment.

Listed serious-event terms and percentages include enteritis infectious, 33.3%; disease progression, 25.0%; cerebral haemorrhage, 25.0%; acute kidney injury, 25.0%; pneumonia, 16.7%; and femur fracture, 16.7%. Their denominator is not reported in the registry record. Cohort sizes and regimens differ, so the counts should be read descriptively rather than as a controlled safety comparison.

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

The record establishes that this was a completed, open-label Phase 1 study using sequential, non-randomized cohorts of Japanese patients with select hematologic malignancies. It identifies a primary endpoint: the number of participants with DLTs in the monotherapy cohort during the first-cycle observation period. It also defines the DLT-evaluable population and provides selected adverse-event counts by cohort.

It does not report the primary DLT count, the size of the DLT-evaluable population, or an uncertainty interval for the primary endpoint. Thus, the primary endpoint’s magnitude and precision cannot be stated from the registry record. Nor does the record support a randomized comparison between monotherapy and combination treatment, or among the different combination cohorts.

The serious adverse-event counts describe participants affected within each cohort. They do not establish that PF-04449913 caused an event, and the death counts do not establish treatment attribution. The listed serious-event terms and percentages provide additional descriptive information, but their denominator is not reported in the registry record. Serious adverse events also should not be treated as interchangeable with protocol-defined DLTs.

In general, early-phase cohort data can describe observed events in the studied participants and treatment settings. Such descriptions do not by themselves establish comparative effects, causal relationships, or outcomes in populations not represented by the study. The record’s non-randomized allocation and differing cohort regimens are especially important when interpreting between-cohort counts.

The planning illustrations are not observed power, and their assumptions do not convert this study into a controlled trial. The registry record supports a careful account of design, endpoint definition and selected safety observations. It does not provide enough primary endpoint results or analysis detail to determine the observed frequency of DLTs or make a comparative efficacy conclusion.

14. Why This Trial Matters Statistically

This trial illustrates why early-phase results need to be read in the context of their cohort structure and endpoint definitions. The registry describes a sequential, non-randomized study with several regimens, but its primary endpoint concerns DLTs in the monotherapy cohort. Keeping the endpoint’s population and observation window explicit prevents safety counts from other cohorts from being mistaken for the primary result.

The DLT criteria combine clinical events with treatment-delivery constraints, including specified toxicity thresholds, prolonged myelosuppression, dose delivery and cycle delay. Such operational definitions shape the event count. They also make the analysis population important: participants had to receive at least 1 dose and avoid major treatment deviations during the observation period to be DLT-evaluable. Yet the registry does not report the number analyzed or the primary count.

The record also demonstrates a distinction between descriptive safety information and comparative evidence. Serious adverse-event counts are available for cohorts with different doses and combinations, but the study did not randomly allocate participants to comparable arms. Those counts are informative as cohort descriptions, not as estimates of the causal difference between regimens.

Finally, the registry lists many outcome measures but limited analysis information. A count of outcomes and analyses does not reveal the full statistical plan, handling of missing data or multiplicity procedures. Recognizing those gaps is part of sound statistical interpretation: conclusions should track what the record actually reports, while uncertainty and unavailable results remain explicit.

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.

Several limitations specific to this record constrain interpretation. The primary endpoint is a DLT count for the monotherapy cohort, but the primary count and the number of DLT-evaluable participants are not reported in the registry record. The event definition is detailed, yet the absence of the observed result prevents an estimate of how frequently participants met those criteria.

The study used non-randomized, sequential cohorts and no masking. Cohorts differ by treatment regimen and include monotherapy, LDAC, intensive chemotherapy, azacitidine, a continuation cohort and an LDAC expansion cohort. The record does not provide a like-for-like control comparison. Differences in their event counts cannot be isolated from differences in cohort composition or treatment context.

The total enrollment is 48, but the record does not state how many participants contributed to each endpoint analysis beyond the cohort-specific safety counts and the DLT-evaluable population definition. It does not report missing-data procedures, a full statistical analysis plan, multiplicity handling, or the trial’s sample-size justification. These details are not reported in the registry record.

The serious-event percentages are listed without a reported denominator, and the record does not provide attribution details for deaths. The DLT endpoint and serious adverse-event summaries measure different things and should not be substituted for one another. Finally, registry descriptions do not establish treatment causality or support broad generalization beyond the study population and treatment settings described.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study Of PF-04449913 In Japanese Patients With Select Hematologic Malignancies — NCT02038777.
• Registry record last updated 2025-02-26; status last verified 2025-02.
• Results first posted 2022-03-18.
• 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

NCT02038777 is a completed Phase 1, open-label, multi-center study sponsored by Pfizer. It enrolled 48 participants and used sequential, non-randomized cohorts. The study evaluated oral PF-04449913 (glasdegib) alone and with LDAC, intensive chemotherapy, or azacitidine in Japanese patients with select hematologic malignancies.

The primary endpoint was the number of participants with DLTs in the monotherapy cohort from Day -5 through Day 28 of Cycle 1 (33 days). The registry defines the DLT-evaluable population but does not report the primary endpoint count. Cohort-level serious adverse-event and death counts are recorded, but the study does not provide a randomized comparison or establish event attribution.