Cytarabine With or Without SCH 900776 in Treating Adult Patients With Relapsed Acute Myeloid Leukemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cytarabine With or Without SCH 900776 in Treating Adult Patients With Relapsed Acute Myeloid Leukemia (NCT01870596), a randomized phase 2 study. The primary outcome is Response Rate(CR/CRi) Rate.

2. Clinical Question

This is a phase II trial of a treatment combination in relapsed AML. The study asks a focused question about adding a CHK1 inhibitor to a cytarabine regimen, while retaining cytarabine as the comparator treatment. Because the primary endpoint is a binary response outcome, each participant is counted according to whether the recorded response met the CR/CRi definition. The posted results report six responses in Arm A and nine in Arm B. Those counts are descriptive; interpreting the comparison also requires attention to the number of participants in each arm and to the uncertainty around the estimates.

The trial was registered as NCT01870596 and is marked completed. Its start and primary-completion dates are May 2013 and December 2014. The registry record does not report additional follow-up details or a separate account of when each participant’s response was assessed. The primary outcome timeframe is stated as up to 3 years.

3. Trial Design

This was a randomized, unmasked, parallel-group phase II trial sponsored by the National Cancer Institute. Participants were assigned to one of two arms: cytarabine plus SCH 900776, or cytarabine alone. The registry lists enrollment of 32. The posted arm counts are 14 for Arm A and 18 for Arm B. The record does not report a rationale for the allocation balance or details of the randomization sequence, allocation concealment, or any stratification factors.

In Arm A, cytarabine was infused continuously over 72 hours on days 1–3 and 10–12. SCH 900776 was administered intravenously over 30 minutes on days 2, 3, 11, and 12. Arm B received cytarabine on the same schedule. The registry also lists laboratory biomarker analysis as an intervention of type other, but does not post a secondary outcome measure for it.

The primary outcome is the CR/CRi response count, reported separately by arm for descriptive purposes. The stated outcome timeframe is up to 3 years. CR and CRi follow definitions described in the registry record: CR requires a bone marrow with less than 5% myeloblasts, normal maturation of all cell lines, an ANC of at least 1000/μL and a platelet count of 100,000/μL, along with absence of peripheral blood blasts and identifiable leukemic cells in the bone marrow, clearance of disease-associated cytogenetic abnormalities, and clearance of previously existing extramedullary disease. CRi meets the CR criteria except for residual neutropenia, defined as ANC < 1000/μL.

The trial is marked completed, with a start date of May 2013 and primary completion in December 2014. The registry record does not report the original sample-size justification, the trial’s prespecified statistical analysis plan, or whether interim analyses were planned or performed. No outcome analyses are posted in the registry record beyond the arm-level primary outcome counts and the comparison statistics summarized for that endpoint.

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.

Overall Study

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.

6. Primary Endpoint

The registry lists the primary outcome as Response Rate(CR/CRi) Rate, measured over Up to 3 years.

The primary endpoint counts participants who achieved CR or CRi. It combines complete response with complete response despite incomplete blood count recovery, using the criteria described in the registry record. The endpoint is binary at the participant level: a person either meets the CR/CRi response definition or does not. The results are reported as counts separately for Arm A and Arm B, with the stated timeframe of up to 3 years.

Response outcomes are used in clinical trials because they summarize whether disease findings meet a defined response threshold. Here, the definition includes bone marrow findings, blood counts, absence of blasts, cytogenetic status, and extramedullary disease. CRi allows residual neutropenia while retaining the other CR criteria. That distinction means the combined CR/CRi endpoint includes participants who meet the response criteria but have incomplete neutrophil recovery.

The registry says the rate will be reported for descriptive purposes. Accordingly, the posted count is not by itself a complete account of response duration, time to response, relapse, or other clinical outcomes. Those outcomes are not reported as primary measures in the registry record. Nor does the endpoint description alone establish how missing assessments or unevaluable participants were handled; those details are not reported in the registry record.

7. Results Posted in the Registry

• Arm B (Cytarabine): 50.0% (9/18) versus 42.9% (6/14) on control; risk difference 7.1 percentage points (95% CI -27.6 to 41.9), risk ratio 1.17 (0.55–2.50); unadjusted two-proportion z-test p = 0.688.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant contributes a response count if CR or CRi is recorded. For a binary endpoint, results can be summarized as counts and as proportions or comparative measures. The registry posts arm counts, a risk ratio, a risk-difference estimate, a confidence interval, and an unadjusted p-value. These statistics address different aspects of the comparison, and none should be read in isolation.

What does the risk ratio express?

A risk ratio compares the proportion meeting the outcome in one group with the proportion in the other. A value of 1 represents equal proportions; values above or below 1 indicate a relative contrast in opposite directions, depending on the comparison orientation. The posted risk ratio is 1.167. Its confidence interval is 0.545 to 2.497, so the interval includes 1. This indicates that the estimate is imprecise: the posted data are compatible with a range of relative effects, including no relative difference. The interval does not identify which value within that range is the true effect.

What does the risk difference express?

A risk difference compares the outcome proportions by subtracting one group’s proportion from the other. A value of zero represents no absolute difference. The registry posts a risk-difference interval extending from a negative to a positive value, which likewise includes the no-difference value. This conveys that the direction of the absolute contrast is uncertain. The exact risk-difference point estimate and interval endpoints are not repeated in this explanation because the registry’s posted values have more than three decimal places.

What does the p-value tell the reader?

The unadjusted p-value is 0.688. In general, a p-value describes how unusual data at least as discordant with a specified null hypothesis would be if that hypothesis and the test assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. A p-value alone also does not indicate whether the estimate is precise. Here, the confidence interval and the small arm-level counts are important companions to the p-value. The registry record does not report the complete test specification or a prespecified inferential framework for interpreting this value.

Why do the counts matter?

The response counts are six in Arm A and nine in Arm B, while the respective arm sizes are 14 and 18. Counts are directly interpretable as numbers of participants meeting the endpoint, but raw counts are affected by group size. Comparative measures use the group sizes to express the contrast. When the data contain few observations, the estimates can be sensitive to individual outcomes and uncertainty can remain substantial. That general statistical consideration is relevant to interpreting these posted counts; it does not establish a particular cause for the observed uncertainty.

What is the role of the confidence interval?

In general, a confidence interval is a range generated by a statistical procedure that reflects uncertainty in an estimate. A confidence interval is not a guarantee that the true effect lies within its endpoints, and its width depends on the data and method used. For the posted risk ratio, the interval crosses the null value of 1 and includes both lower and higher relative effects. This is consistent with the registry’s unadjusted p-value not showing a clear statistical distinction. The interval’s breadth makes the point estimate a poor stand-alone summary of what the trial can establish.

What analysis details are missing?

The registry record does not report a full statistical analysis plan, the method used to calculate the confidence intervals, or whether any covariates were included. The comparison is labeled unadjusted, so it should not be described as controlling for participant characteristics. The record does not specify the analysis population, handling of missing outcome assessments, or how participants without an evaluable response were treated. No separate secondary outcome analyses are posted. These omissions limit the ability to reproduce the analysis or assess whether the posted comparison matches a prespecified plan.

General methods can explain how binary outcomes and uncertainty are interpreted, but they cannot fill those gaps in the trial record. The appropriate reading stays close to the posted information: the response counts are descriptive, the risk-ratio estimate is uncertain, and the unadjusted p-value does not establish a clear difference. The registry does not report enough methodological detail to support more specific claims about analysis conduct.

10. Confidence Intervals and Precision

The registry’s posted risk-ratio estimate is 1.167, and its 95% confidence interval runs from 0.545 to 2.497. The interval includes 1, the value corresponding to no relative difference. It also extends to values on either side of that null value, indicating uncertainty about both the magnitude and direction of the relative contrast. The result should not be summarized by the point estimate alone.

In general, a confidence interval expresses the precision of an estimate under a statistical method. A narrower interval indicates less statistical uncertainty than a wider interval, while a broad interval indicates that multiple effect sizes remain compatible with the observed data under that method. The interval is broad. It does not show that all effects within its range are equally plausible, nor does it identify a single effect as established.

The registry also posts a risk-difference interval that includes zero. Zero is the null value for an absolute difference. Its inclusion is consistent with uncertainty about whether the response proportions differ in either direction. The exact risk-difference estimate and interval endpoints have more than three decimal places in the registry record and are therefore not reproduced.

The posted group counts are six responses among 14 participants in Arm A and nine among 18 participants in Arm B. With counts of this size, individual outcomes can have a noticeable influence on the estimated comparison. That general feature of small binary-outcome datasets helps explain why the estimates should be treated cautiously, but it does not permit a more precise conclusion than the interval supports.

The interval describes statistical uncertainty for the posted comparison; it does not account for every possible source of uncertainty. The registry record does not report enough detail to assess missing outcome data, analysis populations, or the methods used to form the interval. Precision should therefore be understood in relation to the posted result, not as a complete assessment of all limitations.

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 planning figures are participants per arm, not total events. They are planning illustrations based on the observed effect, rather than the trial’s original design assumptions. The illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to a stated fraction of the observed effect. The registry record does not report the assumptions used for the trial’s original sample-size design.

The columns for 80% and 90% power describe the number of participants per arm the planning illustration would require to have the stated probability of detecting the chosen effect, under the listed assumptions. Power is a property of a design calculation for a specified effect and analysis; it is not a measure of whether the posted trial result is correct. These figures are not observed power.

For planning at 100% of the observed effect, the illustration gives 765 participants per arm for 80% power and 1,023 participants per arm for 90% power. At 75% of the observed effect, it gives 1,356 participants per arm for 80% power and 1,814 participants per arm for 90% power. At 50% of the observed effect, it gives 3,041 participants per arm for 80% power and 4,070 participants per arm for 90% power. Smaller effects require more participants to detect under the same assumptions, which is why the planning figures increase as the effect fraction decreases.

The curve represents the same planning relationship between the effect size being targeted and the number of participants required under these assumptions. These calculations are illustrations based on the observed estimate, not evidence that the trial was designed with these sample sizes or assumptions. The record does not report the original sample-size justification, so the planning numbers should not be treated as a reconstruction of the trial’s design.

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 = 0.429; treatment proportion p₁ = 0.500; average p̄ = 0.464.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7053 = 1.3824.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7035 = 0.5921.
4. n = (1.3824 + 0.5921)² / (0.0714)² = 764.1, rounded up to 765 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 <- 0.428571428571429                          # control proportion
p1 <- 0.5                          # 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(1870596)
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 = 765 (80% power), 1,023 (90% power)
 75% of observed effect: n per arm = 1,356 (80% power), 1,814 (90% power)
 50% of observed effect: n per arm = 3,041 (80% power), 4,070 (90% power)
power.prop.test: 764.1 
Simulated power at n = 765 per arm: 0.793 (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 measure and no posted secondary outcome measures. It also reports zero posted analyses. The record therefore does not provide a collection of outcome-specific analyses that can be assessed for multiplicity, nor does it report a multiplicity adjustment or hierarchy of statistical tests. Whether the trial protocol included additional outcomes or planned analyses is not reported in the registry record.

Multiplicity arises in general when several statistical tests or outcomes are examined. Repeated testing can increase the chance of observing an apparently unusual result by chance, so a prespecified plan may describe how such testing is handled. The registry record does not state whether any multiplicity procedure was planned or used. Since the primary outcome is described as being reported for descriptive purposes, the posted comparison should not be interpreted as a complete confirmatory analysis with a known multiplicity strategy.

Handling of missing data is also not reported in the registry record. The record does not say whether every randomized participant had an evaluable response, how unevaluable participants were classified, or whether missing assessments were imputed. It does not specify whether the analysis used an intention-to-treat population, a per-protocol population, or another analysis population. The arm-level outcome counts alone cannot resolve these questions.

Randomization is reported, but the registry does not provide details about the randomization process or allocation concealment. It also does not describe any exclusions after randomization or deviations from assigned treatment. These details matter when interpreting which participants contributed to the outcome summaries and whether the analysis preserved the randomized comparison.

The outcome definition describes CR and CRi criteria, but the registry record does not report who assessed response, whether assessments were independently reviewed, or how assessments were timed within the stated timeframe of up to 3 years. These omissions do not change the posted counts, but they limit evaluation of endpoint ascertainment and reproducibility. The record also does not report a complete statistical analysis plan, so the prespecified status of the comparison statistics cannot be established.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: sepsis (11.1%), cardiac arrhythmia (11.1%), acute kidney injusry (7.1%), death multi system organ failure (5.6%).

The serious adverse event summary reports two affected participants among 14 at risk in Arm A and six among 18 at risk in Arm B. For other adverse events, the registry reports 14 affected among 14 at risk in Arm A and 18 affected among 18 at risk in Arm B. These are participant counts from the safety summaries. They should be read separately from the primary CR/CRi outcome and should not be added together; the categories may represent different event groupings, and the record does not provide enough detail to establish overlap.

The registry lists serious-event terms with percentages: Sepsis, 11.1%; Cardiac arrhythmia, 11.1%; Acute Kidney Injusry, 7.1%; and death multi system organ failure, 5.6%. The record does not clearly identify an arm for each listed term or give the denominators and event counts underlying these percentages. They should not be attributed to a particular arm without that information.

Deaths are not reported in the registry record as a separate group count. The safety summaries do not by themselves establish a comparative safety effect. Adverse-event definitions, attribution to treatment, exposure duration, severity grading, and event-level detail are not reported in the registry record, limiting interpretation of the counts and percentages.

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

The result describes the CR/CRi response counts recorded for the two randomized arms: six in the cytarabine-plus-SCH 900776 arm and nine in the cytarabine-alone arm. The posted risk ratio is 1.167, with a 95% confidence interval from 0.545 to 2.497, and the unadjusted p-value is 0.688. The interval includes the null value, and the registry characterizes the outcome reporting as descriptive. Together, these details indicate that the posted data do not yield a precise estimate of the relative response contrast.

This result does not establish that adding SCH 900776 changes the CR/CRi response outcome, and it does not establish that the treatments have the same effect. The uncertainty is compatible with a range of relative effects, and the record does not supply a formal confirmatory analysis plan. A non-significant p-value should not be treated as proof of equivalence or as proof that a meaningful difference is absent.

The primary endpoint is a defined response measure, not a complete account of disease course. The registry record does not report response duration, time to response, relapse outcomes, or other longer-term efficacy measures as posted outcomes. Although the endpoint timeframe is up to 3 years, the record does not give additional details about assessment timing or follow-up completeness. No conclusions about outcomes outside the posted CR/CRi measure can be drawn from these results.

The posted safety figures also have limits. They report participants affected by serious and other adverse events, but event attribution, exposure time, and complete arm-specific event details are not reported. The figures should not be used on their own to claim that one regimen is safer or more harmful.

In general, a randomized comparison can help assess differences between assigned groups, but interpretation still depends on the quality and completeness of outcome measurement, the analysis population, and the analysis plan. Several of those details are not reported in the registry record. The most defensible interpretation is therefore limited to the descriptive response counts, the uncertain posted comparison, and the safety counts as recorded.

16. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison should be interpreted through both the observed counts and their uncertainty. The arm-level response counts are six and nine, while the arm sizes are 14 and 18. Looking only at the number of responses would leave out the difference in group size. The posted comparative estimate accounts for group size, but its confidence interval remains broad and crosses the null value. The point estimate alone is not a stable summary of the evidence.

The record also makes an important distinction between a descriptive endpoint and a confirmatory claim. The CR/CRi outcome is stated to be reported separately by arm for descriptive purposes. The presence of a p-value in a registry comparison does not, by itself, establish that the trial was designed or analyzed as a definitive hypothesis test. The original analysis plan and sample-size justification are not reported in the registry record.

The planning illustration underscores the relationship between the effect targeted and the information a study needs. Under its stated assumptions, smaller fractions of the observed effect require more participants per arm. These figures are planning calculations, not evidence about the trial’s original design or about the certainty of its observed result. They also cannot substitute for a prespecified clinical and statistical rationale.

Finally, the registry’s omissions matter to reproducibility. It reports randomization and the response definition, but not the detailed analysis population, missing-data handling, response-assessment procedures, or a full statistical plan. Those are not merely technical details: they determine how the posted estimates should be interpreted and how independently assessable the analysis is. The trial is therefore a useful example of reading a compact registry result with attention to design, endpoint, and precision, while keeping conclusions within what the record documents.

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.

The posted arm counts are limited in size: 14 participants in Arm A and 18 in Arm B. The response counts are six and nine, so the comparison rests on a small number of binary outcomes. The registry’s risk-ratio interval, 0.545 to 2.497, is broad and includes the null value. This limits precision and leaves considerable uncertainty about the relative contrast.

The trial was unmasked. The registry record does not explain whether response assessment involved independent review, whether assessors knew treatment assignment, or how potential assessment bias was addressed. The CR/CRi criteria are described, but the process for applying them is not reported in the registry record.

The primary endpoint is described as descriptive, and the registry does not report the original sample-size justification, a full statistical analysis plan, or interim-analysis details. It also does not report the analysis population, missing-data handling, protocol deviations, or whether all randomized participants had evaluable outcome assessments. These gaps restrict assessment of the robustness and reproducibility of the comparison.

The endpoint timeframe is up to 3 years, but the registry record does not give response timing, follow-up completeness, or other outcome details such as response duration or relapse. The posted results therefore address a limited outcome rather than the full course of disease.

Safety reporting is also limited. The registry includes counts for serious and other adverse events, but detailed event listings, attribution, exposure duration, and the basis for the listed serious-event percentages are not reported. Death counts are not reported in the registry record. These omissions prevent a detailed comparative safety interpretation.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Cytarabine With or Without SCH 900776 in Treating Adult Patients With Relapsed Acute Myeloid Leukemia — NCT01870596.
• Registry record last updated 2016-09-01; status last verified 2016-07.
• Results first posted 2016-09-01.
• 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

NCT01870596 is a completed, randomized, unmasked, parallel-group phase II trial sponsored by the National Cancer Institute. It compared cytarabine plus the CHK1 inhibitor SCH 900776 with cytarabine alone in adults with relapsed AML. The registry lists enrollment of 32, with 14 participants in Arm A and 18 in Arm B. The trial started in May 2013 and reached primary completion in December 2014.

The primary outcome was the descriptive CR/CRi response count, with six responses in Arm A and nine in Arm B. The posted risk ratio is 1.167, its 95% confidence interval is 0.545 to 2.497, and the unadjusted p-value is 0.688. The record also posts serious and other adverse-event summaries. It does not report a full statistical analysis plan, original sample-size rationale, analysis population, missing-data approach, or additional outcome analyses.