Combination of GTI-2040 and Cytarabine in the Treatment of Refractory and Relapsed Acute Myeloid Leukemia (AML): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Combination of GTI-2040 and Cytarabine in the Treatment of Refractory and Relapsed Acute Myeloid Leukemia (AML) (NCT00565058), a randomized phase 2 study. The primary outcome is Overall Response Rate of GTI-2040 Combined With HiDAC in Refractory or Relapsed AML.

2. Clinical Question

The study is identified as a Phase II multicenter trial of GTI-2040 combined with HiDAC in refractory or relapsed AML. Its stated purpose included evaluation of pharmacodynamic activity and overall response rate. The registry describes a randomized, parallel-group design with no masking. The timing contrast is clinically and statistically important to interpretation: the result does not compare the combination with an untreated group or with HiDAC alone. It compares two schedules of the same named combination agents. The registry does not report the sequence-generation method, allocation concealment, or the details of the pharmacodynamic investigation.

3. Trial Design

The registry identifies this as a Phase II, randomized, parallel-group, open-label, multicenter study. It enrolled 27 participants and is marked completed. The recorded study start is 2007-08, and primary completion is recorded as 2009-09. These record details describe the study status and timing; they do not establish how participants moved through screening, treatment, or follow-up.

The two groups differed in when GTI-2040 was added to HiDAC. In the Pilot group, GTI-2040 was delayed until 24 hours after HiDAC began. In the Phase II Arm, GTI-2040 was administered 24 hours before HiDAC was added. The registry labels these the Delayed GTI-2040 and Early GTI-2040 groups, respectively. Because each group received the combination with a different schedule, the comparison is about timing, not the presence versus absence of GTI-2040.

The official study title describes a Phase II Simon two-stage study and a pilot pharmacodynamic investigation. The registry also marks allocation as randomized and describes the groups as parallel. Details reconciling the Simon two-stage description with the randomized allocation, including any stage-specific conduct or decision rules, are not reported in the registry record. The original design assumptions, planned number of participants, and any stopping rules are also not reported in the registry record.

The primary outcome was overall response, defined as achieving CR or CRi while on study and assessed at 29-35 days. The posted result is described for the All Treated Patients population. The arm outcome denominators are 10 for Pilot and 15 for the Phase II Arm. The record's enrollment total is 27, so these outcome denominators do not account for all enrolled participants. The reasons and handling of participants not represented in these denominators are not reported in the registry record. A secondary outcome is a summary of treatment-emergent adverse events.

The record lists Aptose Biosciences Inc. as sponsor. It does not report masking, consistent with the open-label designation. Information on treatment adherence, protocol deviations, concomitant care, and follow-up beyond the stated primary-outcome timeframe is not reported in the registry record. Those omissions limit assessment of how closely the observed comparison corresponds to the intended treatment schedules.

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: Patients were enrolled in one of two treatment arms: "Pilot Arm" (Pilot PD Group) using delayed administration of GTI-2040, or a "Phase 2 Arm" (Phase II Group) using early administration of GTI-2040 with respect to High Dose Cytarabine (HiDAC).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All Treated Patients population

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: A total of 27 acute myeloid leukemia patients with either relapse (\> 6 months) after First complete remission (CR1) or refractory/early relapse CR1 duration \< 6 months were enrolled into the study from 6 sites.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate of GTI-2040 Combined With HiDAC in Refractory or Relapsed AML, measured over at 29-35 days.

The primary endpoint is the number of participants with an overall response, assessed at 29-35 days. Overall response was defined as achieving complete remission (CR) or CR with incomplete blood count recovery (CRi) while on the study. The endpoint combines these two categories into a binary outcome: a participant either met the definition or did not. The posted counts are 3 in the Pilot group and 4 in the Phase II Arm.

A binary response endpoint is useful for expressing how many participants meet a specified clinical response definition within a stated assessment period. Unlike a continuous measure, it does not describe the degree of change within each response category. The registry definition also makes the endpoint dependent on the assessment and classification of CR and CRi. Details about response adjudication, assessment procedures, or how missing assessments were classified are not reported in the registry record.

The endpoint's timeframe matters. The recorded assessment is at 29-35 days; the result therefore describes response status in that period, not longer-term outcomes. It does not directly summarize duration of response, survival, or other outcomes. Those measures should not be inferred from the response counts. The record's secondary outcome is a summary of treatment-emergent adverse events, a separate domain from the primary response endpoint.

The outcome is reported for the All Treated Patients population. The primary comparison is between the timing groups: GTI-2040 before HiDAC in the Phase II Arm, and GTI-2040 after HiDAC initiation in the Pilot group. This endpoint can describe the observed response counts in those groups, but interpretation of differences also depends on the size of the groups, uncertainty, and the completeness of outcome ascertainment.

7. Results Posted in the Registry

• Phase II Arm: 26.7% (4/15) versus 30.0% (3/10) on control; risk difference -3.3 percentage points (95% CI -39.5 to 32.8), risk ratio 0.89 (0.25–3.15); unadjusted two-proportion z-test p = 0.856.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is binary: each participant is classified according to whether CR or CRi was achieved while on study. For a binary endpoint, the observed response proportion is the number meeting the definition relative to the number assessed in the group. The registry posts the underlying counts and group denominators, as well as a comparison between the groups. This describes the observed data; it does not by itself establish that any difference is caused by the timing schedule.

What does the risk difference describe?

The risk difference expresses the difference between the two groups' response proportions on an absolute scale. The registry comparison is oriented as Phase II Arm relative to Pilot, and the posted estimate is negative and close to zero. A negative value means the Phase II Arm's observed response proportion is lower under that orientation. Its confidence interval crosses zero, so the posted data are compatible with no difference as well as with differences in either direction. The estimate is unadjusted.

How should ratio measures be read?

The risk ratio compares the response proportions as a ratio, while the odds ratio compares the response odds. A ratio of one represents no difference on the corresponding scale. The posted confidence intervals for both measures span that no-difference value, and they are wide. These measures offer alternate descriptions of the same binary outcome; they do not add independent participants or resolve the uncertainty caused by the small group sizes. Their interpretation also depends on which group is the reference and on the event definition.

What does the p-value say?

The unadjusted p-value is 0.856. General statistical explanation: a p-value measures how incompatible the observed data are with a specified null model, under the assumptions of the test. It is not the probability that the null hypothesis is true, and it is not the probability that the result occurred by chance. A relatively large p-value does not demonstrate equivalence or rule out a meaningful difference. No equivalence margin or non-inferiority framework is reported in the registry record.

Why does the confidence interval matter?

General statistical explanation: a confidence interval describes the range of estimates reasonably compatible with the data and the method's assumptions. It communicates precision and the direction and scale of uncertainty, which a single p-value does not. In this comparison, the risk-difference interval crosses the no-difference value and reaches in both directions. The ratio intervals likewise include their no-association value. This indicates that the small study does not pin down the between-group contrast precisely. The confidence intervals do not account for possible sources of bias or problems not represented in the statistical model.

What does the planning illustration represent?

The planning illustration uses the observed effect as its starting point and applies stated fractions of that effect. It gives the number of participants per arm for 80% power and for 90% power under its stated assumptions. General statistical explanation: power is the probability that a study design would detect an effect of a specified size at a specified significance level, if that effect and the model assumptions were true. The illustration is a planning exercise, not a description of the power achieved by the completed trial. Because the starting effect is estimated imprecisely from small groups, planning quantities anchored to it should not be treated as a stable estimate of what a future study would require.

For binary outcomes, the estimated response proportions and the variance of those estimates depend on the numbers of participants and outcome frequencies in each group. General statistical explanation: small samples usually yield less precise estimates, particularly when events are uncommon. A confidence interval may consequently include a broad range of plausible effects even when the observed point estimate is near zero. Larger samples can improve precision, but the benefit depends on the outcome frequency, the effect to be detected, and the analysis assumptions.

The record labels the study randomized and parallel, which identifies broad design features. It does not report the statistical test used for the posted comparison, the confidence-interval construction method, covariate adjustment, or details of any prespecified analysis. The p-value is explicitly unadjusted. There are no posted formal analysis records in the registry, although the primary outcome and comparison values are reported. Details of any interim analyses, stagewise decision rules, or handling of repeated looks are not reported in the registry record. Those methods should not be inferred from the study title.

10. Confidence Intervals and Precision

The main precision message is that the observed group comparison is uncertain. The risk-difference estimate is near zero, but its confidence interval reaches to both sides of zero. The interval therefore includes the possibility of a lower response proportion in the Phase II Arm, no difference, or a higher response proportion. The ratio-based confidence intervals also include their no-association value. Each interval is broad relative to its central estimate, so the record does not tightly determine the size or direction of the underlying contrast.

General statistical explanation: a confidence interval is a range generated by a statistical procedure that reflects sampling uncertainty under that procedure's assumptions. It is not a range containing every possible value, nor does it account automatically for bias, missing outcome information, or limitations in study conduct. A broad interval indicates lower precision; it does not mean that every value in the interval is equally likely.

The outcome groups are small, with 10 participants in Pilot and 15 in the Phase II Arm. The response counts are 3 and 4, respectively. With few observations, each participant's outcome can have a substantial influence on the estimated group proportions. That makes the point estimate sensitive to the limited sample and leaves the comparison imprecise. The registry enrollment of 27 is not the denominator used in the posted arm-specific outcome counts; the reasons for the difference are not reported in the registry record.

The reported p-value, 0.856, should be read alongside the confidence intervals rather than as a measure of precision or proof of no effect. It reflects the posted unadjusted test, while the intervals show how uncertain the effect estimates are. The record does not report a clinically important difference threshold, so statistical uncertainty cannot be translated into a judgment about whether a particular difference would matter clinically.

11. 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 sample-size figures are planning illustrations based on the observed effect, not the trial's original design calculations. They show participants per arm for 80% power and participants per arm for 90% power, rather than a total number of events. The entries apply the observed effect and smaller fractions of that effect. The values are shown as participants per arm, so they are not totals across both groups.

The stated assumptions for the illustration are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The original trial's own design assumptions, including any sample-size justification, are not reported in the registry record. The study title refers to a Simon two-stage design, but the details of its planned sample size or decision rules are not reported in the registry record.

In plain language, the 80% power column represents the participant count per arm for a hypothetical study designed to have an 80% chance of detecting the selected effect under the stated assumptions. The 90% power column represents the count per arm for a 90% chance under those same assumptions. They are prospective planning quantities, not observed power for this completed trial.

The listed scenarios use 100%, 75%, and 50% of the observed effect. Smaller effects need more participants to detect under the same general assumptions, which is why the required counts rise as the effect fraction gets smaller. Because the observed effect comes from small groups and is itself uncertain, the resulting planning figures should be treated as illustrations rather than definitive requirements for a future study. The curve should be read in the same planning context: it relates the assumed effect size to the participant requirement, rather than describing additional observed outcomes.

12. 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

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.300; treatment proportion p₁ = 0.267; average p̄ = 0.283.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6373 = 1.2490.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6368 = 0.5360.
4. n = (1.2490 + 0.5360)² / (-0.0333)² = 2,867.6, rounded up to 2,868 participants per arm 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
p0 <- 0.3                          # control proportion
p1 <- 0.266666666666667                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(565058)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 2,868 (80% power), 3,839 (90% power)
 75% of observed effect: n per arm = 5,144 (80% power), 6,886 (90% power)
 50% of observed effect: n per arm = 11,675 (80% power), 15,628 (90% power)
power.prop.test: 2867.6 
Simulated power at n = 2,868 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, overall response at 29-35 days, and one secondary outcome, a summary of treatment-emergent adverse events. It reports a single unadjusted p-value for the posted primary comparison. No multiplicity adjustment or prespecified hierarchy for testing outcomes is reported in the registry record. The record does not state whether other analyses were planned or conducted, or whether any adjustment was intended for multiple comparisons, endpoints, or interim looks.

The registry includes no posted formal analysis records. The result values nevertheless include group counts and an unadjusted comparison. This does not establish which statistical methods were prespecified, whether the displayed comparison was the protocol's primary analysis, or whether other analyses were performed. Details of an analysis plan, including handling of covariates or departures from the intended analysis, are not reported in the registry record. Interim analyses and any statistical stopping boundaries are likewise not reported in the registry record.

The primary endpoint is described for the All Treated Patients population. The record does not explain the population definition in further detail or report an intention-to-treat analysis. Enrollment is 27, while the outcome counts are based on group denominators of 10 and 15. Reasons for the difference, participant flow, and whether any participants were excluded after enrollment are not reported in the registry record.

Missing outcome data and their handling are not reported in the registry record. It is therefore not possible to determine whether outcome status was unavailable for any participant, how unavailable assessments were treated, or whether a sensitivity analysis was performed. Response assessment procedures and any adjudication are also not described. These omissions matter because the interpretation of a binary endpoint depends on how participants with no recorded assessment are handled.

Multiplicity and missingness are different concerns. Multiplicity concerns how interpretation changes when several statistical questions are examined; missingness concerns the absence of outcome information for some participants. The record does not provide enough detail to assess either issue fully. The posted p-value should consequently be understood as an unadjusted result, not as evidence that all relevant analyses or missing-data concerns have been addressed.

14. Safety Results

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

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: disease progression (20.0%), pneumonia (20.0%), febrile neutropenia (10.0%), acute respiratory failure (10.0%), hypoxia (10.0%).

The safety results are reported separately for the Pilot and Phase II Arm. Serious adverse events affected 2 participants among 10 at risk in the Pilot group and 2 among 15 at risk in the Phase II Arm. Other adverse events affected 10 of 10 and 15 of 15 participants, respectively. These are counts of participants recorded with events, not a complete account of event frequency, duration, severity, or causal attribution.

The registry lists serious-event terms including Disease Progression and Pneumonia, each reported as 20.0%, and Febrile Neutropenia, Acute Respiratory Failure, Hypoxia, and Pulmonary oedema, each reported as 10.0%. The record does not provide enough detail to determine whether participants could have experienced more than one term, how events were coded, or the precise denominator used for each listed percentage. These terms should not be treated as mutually exclusive categories or added together.

Deaths are not available in the posted safety figures; a blank death count does not establish that no deaths occurred. The secondary outcome is described as a summary of treatment-emergent adverse events, but event-level timing, severity grades, relatedness, and discontinuations are not reported in the registry record. The safety counts allow a limited description of what is posted, not a comprehensive comparative safety assessment.

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

The observed response counts are 3 in the Pilot group and 4 in the Phase II Arm, with group denominators of 10 and 15. The posted statistical comparison is close to no difference, has a broad confidence interval spanning both directions, and has an unadjusted p-value of 0.856. Taken together, the results do not precisely establish the size or direction of a difference between the timing groups.

That uncertainty is not the same as evidence that the schedules are equivalent. Demonstrating equivalence or non-inferiority requires a prespecified margin and an analysis designed for that question; neither is reported in the registry record. Similarly, a non-significant comparison does not prove that there is no effect. The observed estimate is based on limited numbers, and the interval remains compatible with a range of effects.

The comparison is specific to two schedules of GTI-2040 with HiDAC in refractory or relapsed AML. It cannot answer how either schedule compares with HiDAC alone or with no GTI-2040. The endpoint is the combined CR/CRi response definition at 29-35 days. It does not describe how long responses lasted, survival, or outcomes beyond that assessment window.

The registry's enrollment total is 27, but primary-outcome arm denominators account for 10 Pilot participants and 15 Phase II Arm participants. The reason for this discrepancy is not reported in the registry record. The absence of participant-flow and missing-data details limits a fuller assessment of how representative the posted counts are of all enrolled participants.

The safety summaries also have limits: the record posts affected-participant counts and selected serious-event terms, but does not provide a complete account of event severity, timing, relatedness, or death counts. Thus, the response and safety results should be read as registry-record findings with substantial uncertainty and incomplete reporting, not as a stand-alone basis for treatment decisions or broad conclusions about the regimens.

16. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between comparing treatment components and comparing treatment schedules. Both groups received GTI-2040 with HiDAC, but GTI-2040 was timed differently relative to HiDAC. The resulting contrast addresses that timing difference; it does not isolate the effect of adding GTI-2040 to HiDAC. Clearly identifying the comparator is essential for defining what the estimate can answer.

The study also illustrates why point estimates need uncertainty measures. The observed response counts differ, but the group sizes are small and the confidence intervals span the no-difference values. The p-value alone cannot show the size of effects compatible with the data. Reading the estimate, interval, and sample sizes together gives a more informative account of how much the registry result can establish.

The enrollment total and primary-outcome denominators also highlight the value of participant-flow reporting. Without an explanation for why the outcome counts cover fewer participants than the enrollment total, readers cannot fully assess the analysis population or the possible role of missing outcome information. That is an interpretive limitation distinct from the sampling uncertainty reflected in confidence intervals.

Finally, a small study's observed effect can be an unstable input for planning. The planning illustrations are explicitly based on the observed effect and stated assumptions, so their apparent numerical precision should not be mistaken for certainty about the effect a future study should target. The original sample-size rationale and statistical design details are not reported in the registry record. This example shows why planning assumptions, outcome definitions, missing-data handling, and analysis methods all matter when moving from a registry result to a broader statistical interpretation.

17. 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 trial-specific limitations constrain interpretation. The outcome groups are small: the primary endpoint is reported for 10 participants in Pilot and 15 in the Phase II Arm. The registry reports 27 enrolled participants, but does not explain why the outcome denominators account for fewer participants. Participant flow, exclusions, and the handling of unavailable outcome assessments are not reported in the registry record.

The endpoint is defined as CR or CRi while on study, assessed at 29-35 days. Further details of response assessment, confirmation, or independent adjudication are not reported in the registry record. The posted endpoint therefore gives a count under the stated definition, but does not allow assessment of how response classifications were made or whether assessment methods were consistent across groups.

The registry records randomization and an open-label parallel design, but does not report the method of sequence generation or allocation concealment. It does not report an analysis plan, the original sample-size justification, details of the Simon two-stage procedures, or any interim analysis rules. The unadjusted p-value is posted, but the specific inferential method and any prespecified adjustment are not reported in the registry record.

The safety summary is limited to affected-participant counts and selected serious-event terms. Death counts are unavailable, and event severity, timing, relatedness, and participant-level overlap are not reported in the registry record. Together, these missing details mean that neither the response comparison nor the safety information can be interpreted as a complete account of all enrolled participants and outcomes.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Combination of GTI-2040 and Cytarabine in the Treatment of Refractory and Relapsed Acute Myeloid Leukemia (AML) — NCT00565058.
• Registry record last updated 2015-07-27; status last verified 2015-04.
• Results first posted 2015-07-27.
• 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.

21. Record Summary

NCT00565058 is a completed, randomized, open-label, parallel-group Phase II study of GTI-2040 with HiDAC in refractory or relapsed AML. The groups differed in timing: GTI-2040 was delayed until 24 hours after HiDAC initiation in the Pilot group, and given 24 hours before HiDAC was added in the Phase II Arm. The registry reports 27 enrolled participants.

The primary outcome was the number achieving CR or CRi while on study, assessed at 29-35 days. Posted counts are 3 among 10 Pilot participants and 4 among 15 Phase II Arm participants. The unadjusted p-value is 0.856, and the confidence intervals are broad and include no difference. The posted outcome denominators do not account for the full enrollment total; the reason is not reported in the registry record.

The record also lists treatment-emergent adverse events as a secondary outcome. Safety reporting includes serious-event counts and selected terms, but death counts are unavailable. Details of the statistical plan, missing-data handling, original sample-size rationale, and the registry's Simon two-stage procedures are not reported in the registry record. The results therefore describe an imprecise comparison of two combination schedules, with important reporting limitations.