A Study Evaluating Intensive Chemotherapy With or Without Glasdegib or Azacitidine With or Without Glasdegib In Patients With Previously Untreated Acute Myeloid Leukemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Evaluating Intensive Chemotherapy With or Without Glasdegib or Azacitidine With or Without Glasdegib In Patients With Previously Untreated Acute Myeloid Leukemia (NCT03416179), a randomized phase 3 study. The primary outcome is Intensive Study: Overall Survival (OS).

2. Clinical Question

3. Trial Design

This was a randomized, phase 3, parallel-group study with quadruple masking and placebo control. The official title describes randomization in a 1:1 ratio and a multicenter design. The registry record lists an overall enrollment of 730 participants. That overall figure covers a study with both intensive and non-intensive treatment populations; it should not be read as the number enrolled in the intensive comparison alone.

The intensive study compared glasdegib plus cytarabine and daunorubicin with placebo plus cytarabine and daunorubicin. The registry describes these regimens as glasdegib or placebo plus 7+3 induction. The primary outcome was overall survival. For that outcome, the reported arms contained 201 participants in the glasdegib group and 203 in the placebo group. The record defines the analysis population as all randomized participants in the FA set.

A separate non-intensive study compared glasdegib plus azacitidine with placebo plus azacitidine. The registry summary describes this group as adults with previously untreated acute myeloid leukemia who were not candidates for intensive induction chemotherapy. Results from that study are distinct from the intensive primary comparison and should not be merged into its treatment-effect estimate.

The record lists the study as completed and reports a start date of 2018-04-20 and a primary completion date of 2020-06-05. It lists 44 outcome measures and four analyses. Further information about the randomization procedure, stratification factors, treatment duration, adherence, and the original sample-size justification is not reported in the registry record. The record also does not report enough detail to reconstruct all aspects of how masking was maintained.

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: Inadvertently 1 participant was enrolled twice into the study resulting in enrollment number as 730. However, a total of 729 participants were randomized and received treatment in the study. Although participation of participants was terminated by the sponsor, the study was considered completed as participants were fully enrolled as planned.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FA set included all randomized participants.

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: Full analysis (FA) set included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Intensive Study: Overall Survival (OS), measured over Baseline up to 25 months.

The primary endpoint is overall survival, defined in the registry as time from randomization to death from any cause. A participant last known to be alive is censored at the date of last contact. The reported timeframe is baseline up to 25 months. The registry presents median overall survival for each intensive-study arm and a hazard ratio comparing the groups.

Overall survival is a time-to-event outcome: it uses both whether an event occurred and the timing of that event. Censoring allows participants without a recorded death by their last contact to contribute information up to that time, rather than treating their survival time as fully observed. This is a general explanation of the endpoint, not a claim about any additional analysis procedures.

The median is the time by which half of the participants in the relevant survival distribution have experienced the event, as estimated from the observed follow-up. It gives an interpretable summary of survival time, but it does not describe every point in the survival experience. The hazard ratio summarizes the relative event rate over time under the model or analysis used; it is not a ratio of the reported medians. The registry identifies Log Rank as the method for the primary comparison and reports a hazard ratio with a 95% confidence interval and p-value.

Overall survival is used because it captures death from any cause and does not require attribution of death to a particular cause. In this record, the endpoint’s definition and analysis population are stated, but detailed information about follow-up completeness, censoring patterns, and any additional assumptions is not reported in the registry record.

7. Results Posted in the Registry

• Intensive Study: Glasdegib + Cytarabine + Daunorubicin vs Intensive Study: Placebo + Cytarabine + Daunorubicin: hazard ratio 1.08 (95% CI 0.76–1.53), reported p = 0.658 (as reported in the registry).

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.

Intensive Study: Percentage of Participants Who Improved in Fatigue Score Measured by the MD Anderson Symptom Inventory -Acute Myelogenous Leukemia/Myelodysplastic Syndrome (MDASI-AML/MDS) Questionnaire

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Intensive Study: Glasdegib + Cytarabine + Daunorubicin: 17.4% (35/201) versus 17.2% (35/203) on control; risk difference 0.2 percentage points (95% CI -7.2 to 7.6), risk ratio 1.01 (0.66–1.55); unadjusted two-proportion z-test p = 0.964.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intensive Study: Glasdegib + Cytarabine + Daunorubicin and the control. The interval gives the range of effects that remain compatible with the data.

Non-intensive Study: Percentage of Participants Who Improved in Fatigue Score Measured by the MDASI-AML/MDS Questionnaire at Week 12

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Non-intensive Study: Glasdegib + Azacitidine: 11.7% (19/163) versus 15.4% (25/162) on control; risk difference -3.8 percentage points (95% CI -11.2 to 3.7), risk ratio 0.76 (0.43–1.32); unadjusted two-proportion z-test p = 0.320.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Non-intensive Study: Glasdegib + Azacitidine and the control. The interval gives the range of effects that remain compatible with the data.

Intensive Study: Percentage of Participants With Complete Remission Without Negative Minimal Residual Disease (CRMRD-neg)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Intensive Study: Glasdegib + Cytarabine + Daunorubicin: 5.0% (10/201) versus 5.4% (11/203) on control; risk difference -0.4 percentage points (95% CI -4.8 to 3.9), risk ratio 0.92 (0.40–2.11); unadjusted two-proportion z-test p = 0.841.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Intensive Study: Glasdegib + Cytarabine + Daunorubicin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes across both the intensive and non-intensive populations. These include fatigue improvement measured with the MDASI-AML/MDS questionnaire and remission-related outcomes, including measures involving minimal residual disease and incomplete hematologic recovery. The timeframes differ by outcome and treatment population, so these results do not all represent the same clinical question or observation window.

Across the posted secondary endpoints, the estimates do not form a consistent pattern of clear separation between the glasdegib and placebo groups. The confidence intervals for the reported percentage outcomes generally encompass a range of possible group differences, and the available comparisons do not provide a firm basis for treating any one secondary finding as decisive. These secondary results are complementary to the primary overall-survival analysis, but they do not replace it. They measure different aspects of outcome and may be affected by different sources of missingness or measurement uncertainty.

Secondary analyses carry less weight when many outcomes are examined. In general, the more comparisons made, the greater the chance that at least one result will appear noteworthy by chance alone. Adjustments or other prespecified strategies can address multiplicity, but the approach used for the full set of secondary outcomes is not reported in the registry record. The record lists 44 outcome measures, which makes it especially important not to interpret isolated secondary p-values as if they were the sole planned test.

Some outcomes have sponsor-reported analyses, while the listed CRMRD-neg intensive outcome has no sponsor-reported analysis in the record. Where an inferential analysis is not posted, the arm summaries can describe the observed percentages but do not supply a complete sponsor-reported test of the contrast. Overall, the secondary endpoints offer additional descriptive evidence about fatigue and remission measures; they do not resolve the uncertainty in the primary survival comparison.

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

• Intensive Study: Percentage of Participants Who Improved in Fatigue Score Measured by the MD Anderson Symptom Inventory -Acute Myelogenous Leukemia/Myelodysplastic Syndrome (MDASI-AML/MDS) Questionnaire
• Non-intensive Study: Percentage of Participants Who Improved in Fatigue Score Measured by the MDASI-AML/MDS Questionnaire at Week 12
• Intensive Study: Percentage of Participants With Complete Remission Without Negative Minimal Residual Disease (CRMRD-neg)
• Non-intensive Study: Percentage of Participants With Complete Remission Without Negative Minimal Residual Disease (CRMRD-neg)
• Intensive Study: Percentage of Participants With Complete Remission Including Negative Minimal Residual Disease (CRMRD-neg)
• Non-intensive Study: Percentage of Participants With Complete Remission (CR) Including Negative Minimal Residual Disease (CRMRD-neg)
• Intensive Study: Percentage of Participants With Complete Remission With Incomplete Hematologic Recovery (CRi)
• Non-intensive Study: Percentage of Participants With Complete Remission With Incomplete Hematologic Recovery (CRi)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is time to death from any cause. For each participant, the time origin is randomization. Participants last known to be alive are censored at their last contact, as stated in the registry definition. This is a standard structure for survival analysis: some participants experience the event during observation, while others have incomplete event times because observation ends before a death is recorded.

What does the hazard ratio summarize?

A hazard is the instantaneous event rate among participants who have not yet experienced the event, conditional on the analysis framework. A hazard ratio compares that rate between groups. The registry reports the estimate for glasdegib plus chemotherapy relative to placebo plus chemotherapy. The value is 1.08. It is not the difference between the two medians, and it should not be interpreted as a direct comparison of the chance of death by a fixed time. A hazard ratio is a relative time-to-event summary, whose interpretation depends on the analysis and the pattern of hazards over follow-up.

The record identifies Log Rank as the primary comparison method. In general, the Log Rank test compares event-time experience between groups while accounting for the timing of events and censoring. Its p-value addresses how compatible the observed comparison is with a null hypothesis of no group difference under the test’s assumptions. A p-value does not measure the size or clinical importance of an effect, nor does it give the probability that the null hypothesis is true.

How should the confidence interval be understood?

A confidence interval expresses uncertainty around an estimate under the method used to construct it. The reported 95% confidence interval for the hazard ratio is 0.755 to 1.532. Its range includes 1, so it includes values consistent with no relative hazard difference as well as values on either side. The width also indicates that the estimate is not precise enough to rule out a range of materially different relative hazards. The interval is about the hazard ratio, not the median survival in either arm.

The medians have their own reported uncertainty. The glasdegib median is 17.3 months, with a 95% confidence interval from 15.2 to 18.5 months. The placebo median is 20.4 months, with a lower confidence limit of 17.6 months; its upper limit is not reported in the registry record. Median confidence intervals and the hazard-ratio confidence interval concern different quantities, so their limits should not be compared as if they were the same effect measure.

What is censoring, and why does it matter?

Censoring occurs when a participant’s event time is not fully observed, for example because the participant was last known to be alive at the end of available contact. The participant contributes information up to that contact date. Survival methods can use such partial follow-up without treating the censoring date as a death. In general, valid interpretation depends on assumptions about censoring and follow-up. The registry definition explains when censoring occurs, but the detailed censoring distribution and checks of assumptions are not reported in the registry record.

How do the medians and hazard ratio fit together?

The median is a single point on the estimated survival distribution, whereas the hazard ratio summarizes relative event rates over time. The two summaries can provide complementary perspectives and need not tell an identical story. A comparison of medians does not recover the Log Rank result, and the hazard ratio cannot be obtained by dividing or subtracting the medians. The primary comparison should therefore be interpreted using the reported time-to-event analysis, alongside the median estimates as descriptive summaries.

The registry reports a 95% confidence interval and p-value for the hazard ratio and gives the Log Rank method. Further details such as the exact model specification, assumptions about proportional hazards, stratification, covariate adjustment, and sensitivity analyses are not reported in the registry record. Those details matter when assessing whether the reported summary adequately represents the time-varying pattern of events.

11. Confidence Intervals and Precision

The central primary estimate is a hazard ratio of 1.08, with a 95% confidence interval from 0.755 to 1.532. The interval spans values below and above 1. In general, 1 is the reference value for a hazard ratio indicating no relative difference. Thus, the data are compatible with a range that includes lower hazards in the glasdegib group, no relative difference, and higher hazards in that group. The interval does not identify a narrow range of plausible values.

Precision is distinct from statistical significance. The reported p-value is 0.658, and the confidence interval includes 1. Together, these indicate that the primary analysis does not provide strong evidence of a difference between the intensive-study groups. They do not prove the groups are equivalent. Demonstrating equivalence or non-inferiority would require a prespecified margin and a suitable analysis; such a margin is not reported in the registry record.

The median survival estimates also have uncertainty. Glasdegib plus chemotherapy has a reported median of 17.3 months with a 95% confidence interval of 15.2 to 18.5 months. Placebo plus chemotherapy has a reported median of 20.4 months and a lower confidence limit of 17.6 months; its upper limit is not reported in the registry record. The missing upper limit should not be inferred from the other interval or from the point estimates.

Confidence intervals describe sampling uncertainty under their analysis methods. They do not account automatically for all possible sources of uncertainty, such as incomplete reporting, departures from assumptions, or limits to generalizability. The registry record does not provide detailed information about censoring patterns or sensitivity analyses, so the effect of those factors cannot be evaluated from the reported summary alone.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations translate an assumed effect size into a hypothetical amount of information needed for a future comparison. For the primary time-to-event outcome, the figures are total events, not participants per arm. The event totals are calculated using a Schoenfeld approach. The other planning illustrations in the registry use participants per arm, but the primary survival comparison is the event-based one.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. The illustrations show planning for 100%, 75%, and 50% of the observed effect. Under these assumptions, the 80% power column gives the total number of events associated with an 80% chance of rejecting the null hypothesis if the assumed effect is true. The 90% power column gives the corresponding event total for a 90% chance. These are design-planning quantities, not observed trial outcomes.

Smaller assumed effects are harder to distinguish from random variation, so they require more information. The displayed event totals increase as planning shifts from the full observed effect toward a smaller fraction of it. This is a general feature of power calculations, not a claim that the trial was designed using these particular assumptions.

The original trial’s own sample-size justification and design assumptions are not reported in the registry record. The planning values are illustrations based on the observed effect and the stated assumptions, not a reconstruction of the trial’s original planning. They are not observed power: observed power is not an appropriate way to reinterpret the completed trial’s result, and these figures instead show hypothetical requirements under specified assumptions.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 1.080, so ln(HR) = 0.0770.
2. d = 4 × (1.96 + 0.842)² / [ln(1.080)]² = 5,300.6, rounded up to 5,301 events in total for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
hr <- 1.08                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(3416179)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 5,301 (80% power), 7,097 (90% power)
 75% of observed effect: events = 9,424 (80% power), 12,616 (90% power)
 50% of observed effect: events = 21,203 (80% power), 28,385 (90% power)
Simulated power with 5,301 events: 0.802 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The sponsor-reported primary analysis is the Log Rank hazard-ratio comparison. It reports a hazard ratio of 1.08, a 95% confidence interval from 0.755 to 1.532, and a p-value of 0.658. These are the primary comparison results used to interpret the intensive-study overall-survival outcome.

For this primary endpoint, there is no conflicting recalculated hazard ratio in the posted comparison. The figures reported by the sponsor are the relevant analysis, and the median values provide additional descriptive context rather than an alternative calculation of the hazard ratio. The registry record identifies the method, but it does not include detailed model specifications or all analysis documentation.

For some binary secondary outcomes, the record also contains sponsor-reported analyses and simple unadjusted calculations based on posted summary statistics. Those calculations may use a different method from the sponsor’s analysis, so their p-values need not match. Where a sponsor analysis is reported, it takes precedence for describing the sponsor’s own inferential result. The CRMRD-neg intensive outcome has no sponsor-reported analysis listed in the record. The calculations based on summary statistics are therefore descriptive illustrations, not replacements for a prespecified sponsor analysis.

These distinctions matter because statistical results depend on the method and analysis population, not just on the observed event counts or percentages. The registry does not report enough detail to establish why the methods differ for every secondary endpoint. Readers should keep the primary sponsor-reported time-to-event analysis separate from exploratory recalculations and from outcomes that lack a reported sponsor analysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple outcomes and analyses, including measures of overall survival, fatigue, remission, and other outcomes. When many hypotheses are tested, the chance of at least one apparently unusual result by chance increases. A prespecified multiplicity strategy can control or otherwise manage that issue, but the strategy for the outcome family is not reported in the registry record. Therefore, secondary p-values should not be read as though each arose from a single isolated confirmatory test.

The registry gives a primary outcome and identifies the primary analysis population as the FA set, including all randomized participants. This supports a clear description of the primary endpoint’s intended population. However, details about the statistical analysis plan, planned hierarchy among secondary outcomes, and handling of multiplicity are not reported in the registry record. The record does not establish whether secondary outcomes were confirmatory, exploratory, or subject to a formal testing sequence.

Missing outcome data can affect estimates, particularly for questionnaire-based outcomes that depend on participants completing assessments. The registry lists fatigue improvement outcomes and their timeframes, but detailed completion rates, reasons for missing assessments, and the method used to handle missing questionnaire data are not reported in the registry record. Likewise, the extent of missing data for remission-related outcomes is not described in the information.

For overall survival, the registry specifies censoring at the last contact for participants last known alive. It does not report the amount or pattern of censoring, follow-up completeness, or sensitivity analyses for censoring assumptions. The record also does not provide detail about deviations from randomized assignment or other analysis-set decisions beyond the statement that the FA set included all randomized participants.

These reporting gaps limit assessment of robustness. They do not show that the analysis was flawed; they mean that the registry summary does not allow readers to evaluate these particular design and analysis choices. The reported estimates can be interpreted as posted, while conclusions about multiplicity control, missing-data assumptions, or additional analysis populations remain unavailable.

16. 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: pneumonia (22.5%), febrile neutropenia (14.8%), disease progression (13.8%), sepsis (8.6%), pyrexia (6.9%).

For the intensive glasdegib group, serious adverse events were recorded for 86 of 198 participants at risk; for the intensive placebo group, they were recorded for 92 of 201 participants at risk. The record also lists other adverse events for 194 of 198 and 195 of 201 participants at risk, respectively. These are counts of participants with recorded events in the stated groups, not estimates of a causal treatment effect.

The registry lists pneumonia at 22.5%, febrile neutropenia at 14.8%, disease progression at 13.8%, sepsis at 8.6%, pyrexia at 6.9%, and electrocardiogram QT prolonged at 6.6% among the leading serious adverse event terms. These listed terms should be read as event reporting, not as a comparison between randomized intensive arms. Detailed arm-specific event rates for each term, event timing, severity grades, and attribution are not reported in the registry record.

Deaths are listed as 90 in the intensive glasdegib group and 88 in the intensive placebo group. These counts are not a substitute for the time-to-event analysis: they do not by themselves account for follow-up duration, censoring, or timing of deaths. Safety counts should therefore be interpreted descriptively and separately from the primary hazard-ratio result.

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

The posted primary results describe overall survival in the randomized intensive-study comparison of glasdegib plus cytarabine and daunorubicin versus placebo plus the same chemotherapy. The median estimates were 17.3 months and 20.4 months, respectively, and the reported hazard ratio was 1.08 with a 95% confidence interval from 0.755 to 1.532. The interval includes 1, and the reported p-value was 0.658. Taken together, the primary analysis does not show a clearly distinguishable overall-survival difference between these groups in the posted results.

That interpretation is not the same as proving that the regimens are equivalent. The confidence interval leaves room for a range of relative hazard values, and the registry does not report an equivalence or non-inferiority margin. Nor does a non-significant result establish that there is no effect. It indicates that the analysis does not provide strong evidence against the no-difference hypothesis under the reported method.

The result applies to the intensive-study population, treatment comparison, and follow-up described in the registry. The trial also included a distinct non-intensive study of glasdegib or placebo with azacitidine. Its results should not be assumed to be interchangeable with the intensive comparison. The registry record does not provide enough detail to extend the primary estimate to other populations, treatment combinations, or settings.

Overall survival captures death from any cause and is a meaningful time-to-event endpoint, but the available summary does not show the full survival experience over time. Median estimates alone cannot describe the shape of that experience. The registry summary also does not provide detailed information about censoring, missing assessments, or sensitivity analyses. These limits affect how fully the result can be evaluated, without changing what the posted primary estimate says.

Safety counts add descriptive context but should not be used alone to attribute individual events to treatment or to compare risks without appropriate analysis. The results are statistical evidence from this trial record, not medical advice or a recommendation about treatment.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares an added study drug against placebo while keeping the intensive chemotherapy backbone the same in the two randomized groups. Randomization is intended, in general, to support a fair comparison by balancing prognostic factors across groups on average. The registry identifies the design as randomized, parallel, placebo-controlled, and quadruple-masked, though it does not report the detailed procedures used to implement these features.

The primary endpoint also illustrates why time-to-event methods are useful in clinical research. Participants may have different lengths of follow-up, and some may be alive at their last contact. Survival analysis can incorporate their observed time without treating them as if an event had occurred. The hazard ratio and median survival offer related but distinct summaries of the outcome.

The reported estimate is not especially precise: the hazard-ratio confidence interval includes 1 and spans values on either side. This makes it important to distinguish an inconclusive comparison from proof of no effect. It also shows why point estimates alone are not enough; the uncertainty interval and analysis method are central to interpretation.

The study’s two treatment populations and its many outcome measures create additional statistical considerations. The intensive primary result should remain separate from the non-intensive comparison, and secondary findings require caution when multiplicity handling is not reported. Finally, planning illustrations based on observed effects should not be mistaken for the original study’s design rationale. The registry record does not report that rationale, so the illustrations serve only to explain how information requirements change under different assumed effect sizes.

19. 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 summary does not report the original sample-size justification, planned power, or assumptions used to design the trial. Consequently, the planning illustrations based on the observed effect cannot establish whether the completed study met its original design targets. They should be treated as hypothetical planning information rather than evidence about the trial’s intended operating characteristics.

Important analysis details are also absent from the registry record. These include any stratification factors or covariate adjustment, the exact model specification, checks of proportional-hazards assumptions, and sensitivity analyses. The record identifies the Log Rank method and reports the hazard ratio, but does not provide enough information to assess these other choices.

The placebo median’s upper confidence limit is not reported in the registry record. The censoring distribution and completeness of follow-up are not described, even though survival estimates depend on observed event times and censoring. These omissions limit assessment of the precision and robustness of the survival summaries.

For secondary outcomes, the record does not report a comprehensive multiplicity strategy, detailed missing-data methods, or questionnaire completion rates. The listed CRMRD-neg intensive outcome has no sponsor-reported analysis shown. These limitations matter when interpreting multiple endpoints and any secondary p-values, but do not invalidate the primary result by themselves.

The overall enrollment of 730 covers both intensive and non-intensive populations, while the primary survival comparison concerns the intensive groups only. The registry summary also does not report enough treatment-exposure or adherence detail to assess how closely participants followed the assigned regimens. These are limitations of the available record, not claims about unreported study conduct.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study Evaluating Intensive Chemotherapy With or Without Glasdegib or Azacitidine With or Without Glasdegib In Patients With Previously Untreated Acute Myeloid Leukemia — NCT03416179.
• Registry record last updated 2023-04-21; status last verified 2023-04.
• Results first posted 2021-06-25.
• 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.

23. Record Summary

NCT03416179 is a completed randomized, phase 3, parallel, quadruple-masked, placebo-controlled study with overall enrollment of 730. It evaluated glasdegib in intensive chemotherapy and azacitidine treatment populations with previously untreated acute myeloid leukemia.

The primary outcome discussed is overall survival in the intensive study, comparing glasdegib plus cytarabine and daunorubicin with placebo plus the same chemotherapy. The registry reports median survival of 17.3 months versus 20.4 months and a Log Rank hazard ratio of 1.08, with a 95% confidence interval of 0.755 to 1.532 and p-value of 0.658. The interval includes 1, leaving substantial uncertainty about the relative hazard.

Secondary outcomes include fatigue and remission measures. The registry also reports serious adverse event counts and selected serious event terms. Details about the original design assumptions, multiplicity handling, missing-data methods, and several analysis features are not reported in the registry record.