Clofarabine and Cytarabine in Treating Young Patients With Refractory or Relapsed Acute Myeloid Leukemia or Acute Lymphoblastic Leukemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Clofarabine and Cytarabine in Treating Young Patients With Refractory or Relapsed Acute Myeloid Leukemia or Acute Lymphoblastic Leukemia (NCT00372619), a non-randomized phase 1, phase 2 study. The primary outcome is Overall Response (CR for ALL Patients), (CR + CRp for AML Patients).

2. Clinical Question

3. Trial Design

This was a phase I/II, nonrandomized, parallel-group study sponsored by Children's Oncology Group. The registry lists enrollment of 74 and a completed status. The recorded start is 2007-03 and the primary completion is 2012-08. Masking was none. Because allocation was nonrandomized, differences among the groups cannot be interpreted as if participants had been assigned to comparable groups by chance.

The study's stated purpose included assessing side effects and dose of clofarabine with cytarabine, as well as assessing how the combination performed in young patients with refractory or relapsed AML or ALL. The phase I portion is described as closed to enrollment as of 09/16/09. The arm descriptions separate patients by leukemia type and clofarabine dose, and distinguish feasibility-focused groups from efficacy-focused groups. An arm for acute leukemia of ambiguous lineage is also listed.

The regimen descriptions include clofarabine 40 mg/m² or 52 mg/m², administered intravenously on days 1-5, and high-dose cytarabine administered intravenously on days 1-5. They also describe intrathecal cytarabine on Day 0. Some arm descriptions refer to Cycle 1 as lasting 14-42 days, while other Cycle 2 details are incomplete in the registry record. The allocation process, reasons participants entered particular arms, and any protocol amendments are not reported in the registry record.

The registry lists methotrexate and laboratory biomarker analysis as interventions in addition to clofarabine and cytarabine. Their specific contribution to the described treatment groups is not reported in the registry record. Without randomized allocation or a reported control group, the trial design primarily supports descriptive reporting across its defined groups rather than a controlled estimate of comparative treatment effects.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Only the first 6 of 7 groups are shown.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response (CR for ALL Patients), (CR + CRp for AML Patients), measured over 2 cycles or up to 84 days.

The primary endpoint was the number of participants with an overall response, assessed after 2 cycles or up to 84 days. The response definition differed by leukemia type. For ALL, response was CR: an M1 bone marrow (< 5% blasts), no circulating blasts or extramedullary disease, and recovery of peripheral counts, defined as ANC > 750/μL and platelets > 75,000/μL.

For AML, the endpoint combined CR and CRp. CR required an M1 bone marrow (<5% blasts), no circulating blasts or extramedullary disease, and recovery of peripheral counts, defined as ANC > 1000/uL and platelets > 100,000/uL. CRp required the same marrow and disease criteria and ANC > 1000/uL, with platelet transfusion independence, defined as no platelet transfusions for 1 week.

Response is used as a measure of whether disease criteria for remission are met during a specified assessment period. The endpoint is a count of participants meeting the relevant definition, not a measure of how long response lasts. The registry describes the criteria and assessment window, but does not report how many participants met them. It also does not report endpoint-specific analysis populations or handling of participants whose response status was unavailable.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The interpretation of a trial begins with its allocation design. This study is listed as nonrandomized and parallel. In a randomized comparison, chance assignment is intended to make groups comparable on average with respect to measured and unmeasured characteristics. That protection is not established by the design. As a general statistical point, an observed difference between nonrandomized groups can reflect differences in patient characteristics, treatment, or both. The record does not report an adjustment strategy that would address confounding.

What does the response endpoint count?

The primary outcome is a binary classification at the participant level: a participant either meets the applicable response definition or does not. The registry lists the outcome unit as participants and the parameter as a number. For ALL, the criterion is CR; for AML, it is CR or CRp. These disease-specific definitions use bone marrow, disease-examination, and blood-count criteria. The response assessment is specified as after 2 cycles or up to 84 days. That timing gives the outcome a defined assessment window, but it does not describe response duration after that point.

For a binary endpoint, a count can be converted into a proportion only when both the count and the relevant denominator are known. The registry record does not report primary response counts or the number assessed for response in each group. No response proportions can therefore be interpreted from the record. A denominator based on overall enrollment should not be assumed to be the endpoint denominator, because the record does not say which enrolled participants were included in the response analysis.

What can group comparisons establish?

General statistical interpretation distinguishes description from causal comparison. Descriptive results report what was observed in each group. A causal comparison asks what would have happened under alternative treatments for comparable participants. Randomization can strengthen that comparison by balancing group characteristics in expectation, but this study is explicitly nonrandomized. The arm categories also differ in leukemia type and purpose, including feasibility and efficacy assessments. Directly contrasting their outcome counts would not provide a like-for-like estimate of a dose effect.

Even for a nonrandomized study, adjusted analysis can sometimes address measured differences if appropriate variables and methods are available. The registry record does not report an analysis plan, covariate adjustment, or adjusted estimates. It also lists 0 analyses. Thus, there is no reported statistical adjustment to interpret, and no basis in the registry record for claiming that group differences have been separated from differences in participant characteristics.

How should uncertainty be understood?

In general, a confidence interval describes a range of estimates compatible with the data and the method used, reflecting sampling uncertainty. A narrow interval indicates greater precision than a wide interval, while an interval that spans substantially different values leaves greater uncertainty about the effect. A p-value is a measure calculated under a statistical model and a null hypothesis; it is not the probability that a treatment works or that the null hypothesis is true. The record reports no primary response confidence interval or p-value. The response result's precision and statistical compatibility with a null comparison therefore cannot be assessed.

Small group counts can make estimates unstable and confidence intervals wide; this is a general statistical consideration, not a calculated assessment of these response groups. Group-specific outcome denominators and response counts are not reported, so response precision cannot be quantified. Safety counts are available, but these do not supply the missing efficacy estimate or its uncertainty.

What do planning calculations represent?

General sample-size calculations estimate the amount of information that might be needed under stated assumptions. The planning illustration associated with the posted material uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. These are planning assumptions, not a description of the trial's own design. The registry record does not report a sample-size justification or state that those assumptions guided enrollment.

Such calculations should not be read as observed power. Observed power is not a substitute for a confidence interval or a direct interpretation of the observed result. Here, the primary response effect itself is not reported in the registry record, and the study was nonrandomized rather than an equal-allocation randomized comparison. Planning illustrations can therefore be read only as hypothetical design scenarios under their stated assumptions, not as evidence about the trial's response findings.

Why can over-dispersion matter?

As a general concept, over-dispersion occurs when outcome variability exceeds what a statistical model assumes. If that feature is ignored, standard errors may be too small and confidence intervals may be too narrow. Over-dispersion is relevant to some count-data models, but this registry record does not report a count-model analysis or identify over-dispersion as a concern. It should not be assumed to have been assessed in this study.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Clofarabine 40 mg/m² to Assess Feasibility in ALL ", "Clofarabine 40 mg/m² to Assess Feasibility in AML ", "Clofarabine 52 mg/m² to Assess Feasibility in ALL ", "Clofarabine 52 mg/m² to Assess Efficacy in ALL Pat", "Clofarabine 52 mg/m² to Assess Feasibility in AML ", "Clofarabine 52 mg/m² to Assess Efficacy in AML Pat")
n      <- c(8, 2, 3, 10, 7, 39)
events <- c(1, 2, 0, 2, 2, 19)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Clofarabine 40 mg/m² to Assess Feasibility in ALL   8  1  12.5  0.3  52.7  2.2  47.1  2.8  48.2
ARM  Clofarabine 40 mg/m² to Assess Feasibility in AML   2  2  100.0  15.8  100.0  34.2  100.0  29.2  99.2
ARM  Clofarabine 52 mg/m² to Assess Feasibility in ALL   3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Clofarabine 52 mg/m² to Assess Efficacy in ALL Pat  10  2  20.0  2.5  55.6  5.7  51.0  6.0  51.8
ARM  Clofarabine 52 mg/m² to Assess Feasibility in AML   7  2  28.6  3.7  71.0  8.2  64.1  8.5  65.1
ARM  Clofarabine 52 mg/m² to Assess Efficacy in AML Pat  39  19  48.7  32.4  65.2  33.9  63.8  33.8  63.9

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary response outcome and secondary outcomes covering safety and tolerability measured by CTCAE v3.0, and correlation of apoptosis-specific gene expression. It reports 3 outcome measures and 0 analyses. The record does not describe a statistical analysis plan or explain how multiple outcomes, dose groups, leukemia types, or response definitions were handled in formal inference. A multiplicity adjustment, hierarchy among outcomes, or prespecified testing sequence is not reported in the registry record.

Multiplicity is a general concern when many outcomes or comparisons are examined: the chance of a false-positive finding can increase if each is assessed independently without an appropriate plan. This is relevant context for interpreting a study with several groups and outcome domains, but it does not establish that a particular analysis was performed or that any result is false. No inferential analysis results are listed, so there are no reported p-values to evaluate for multiplicity.

Missing-data procedures are also not reported in the registry record. The record does not state how participants with unavailable response assessments, early discontinuation, or incomplete follow-up were handled. It does not say whether missing outcomes were excluded, counted as nonresponse, imputed, or addressed in another way. These choices can affect a binary response estimate, so the absence of a stated method limits reproducibility and interpretation.

Analysis populations are not reported in the registry record. In particular, the record does not identify whether the response endpoint was analyzed among all enrolled participants, participants who began treatment, or another defined group. The enrollment figure should not be treated as the response denominator without an explicit statement. Interim analyses and any stopping rules are not reported in the registry record; no assumptions about them can be made.

12. 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: febrile neutropenia (50.0%), hypercalcemia (50.0%), hyperglycemia (50.0%), hypermagnesemia (50.0%), hypoalbuminemia (50.0%).

The adverse-event results report, for each listed group, the number at risk and the number affected by serious adverse events, as well as the number at risk and affected by other adverse events. For example, the group titled Clofarabine 40 mg/m² to Assess Feasibility in ALL Patients lists 8 at risk and 5 affected by serious events, and 8 at risk and 8 affected by other events. The group titled Clofarabine 52 mg/m² to Assess Efficacy in AML Patients lists 40 at risk and 14 affected by serious events, and 40 at risk and 38 affected by other events. These are group-specific counts, not randomized comparisons.

The registry lists febrile neutropenia, hypercalcemia, hyperglycemia, hypermagnesemia, hypoalbuminemia, and hypokalemia among the top serious adverse-event terms, each at 50.0%. The denominator and whether participants could have more than one listed term are not reported in the registry record. A term percentage should not be treated as a proportion of all enrolled participants unless its denominator is specified.

Counts across groups should not be combined or interpreted as evidence that one dose is safer: the groups differ in disease and study purpose, and allocation was nonrandomized. The registry does not report exposure-adjusted rates or deaths for the listed groups. Safety summaries describe reported events and do not establish causation.

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

The record describes a phase I/II investigation of clofarabine with cytarabine in young patients with refractory or relapsed leukemia, including groups focused on feasibility and efficacy. It specifies the primary response endpoint and the criteria used to classify response in ALL and AML. Those definitions make clear that response was assessed differently by disease type and that the AML endpoint included CRp as well as CR.

What the registry record does not establish is the frequency of response. It contains no reported primary response counts, rates, confidence intervals, or p-values. It therefore cannot show how large a response estimate was, how precisely it was estimated, or whether it differed from a control outcome. The absence of these results in the record is not evidence that no responses occurred; it means the primary response findings are not reported there.

The design also limits what can be inferred from comparisons among groups. The study was nonrandomized, and group descriptions differ by leukemia type, dose, and feasibility or efficacy purpose. A difference in outcomes among such groups would not by itself identify a causal effect of dose or treatment. There is no randomized control comparison reported in the registry.

The adverse-event counts describe participants affected in the listed groups, while the serious-event term percentages are reported without a denominator explanation. They provide safety information as recorded, but they do not substitute for response results or establish that a treatment caused an event. The registry also does not report exposure-adjusted safety rates, deaths, missing-data methods, or analysis populations.

In general, trial conclusions depend on the endpoint definition, the participants included in analysis, the comparison design, and the uncertainty around estimates. Because key outcome results and analysis details are absent from the registry record, conclusions should be limited to the study's stated design, endpoint definitions, and posted adverse-event summaries.

14. Why This Trial Matters Statistically

This trial illustrates why endpoint definitions matter in clinical research. The primary outcome was not simply a generic remission label: it used different criteria for ALL and AML, and the AML definition included CRp. The marrow, disease, and blood-count requirements determine which participants qualify as responders. Clear definitions support consistent classification, while the disease-specific criteria also mean that results across the diagnostic groups would need careful interpretation.

The record also illustrates the distinction between a stated endpoint and a reported result. A registry can describe what investigators intended to measure without reporting the resulting outcome counts or analyses. Here, the response criteria and assessment window are specified, but response results and inferential analyses are not reported. Statistical interpretation requires both a defined outcome and data that show how participants fared under that definition.

Design is equally important. The nonrandomized parallel groups are organized by leukemia type, dose, and study purpose, rather than by randomized allocation to comparable treatments. This makes simple between-group contrasts vulnerable to differences in who entered each group. The arm labels should not be mistaken for a controlled dose comparison.

Finally, the safety summaries show why counts need context. Group sizes vary, the listed serious-event terms lack a stated denominator, and exposure-adjusted rates are not reported. Without those details, the counts are descriptive rather than a basis for ranking groups. These features make the record useful for understanding the questions, definitions, and reporting limitations of a complex early-phase study, while limiting claims about comparative effects or precision.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this trial affect interpretation. First, the primary response counts and rates are not reported in the registry record. Although the response definition and assessment window are detailed, there is no endpoint estimate or uncertainty measure to assess. The registry also does not identify the response analysis population or provide endpoint-specific denominators.

Second, allocation was nonrandomized. Groups differ by leukemia type, clofarabine dose, and whether the arm assessed feasibility or efficacy. The record does not report a control group or an adjustment strategy for differences between participants. Accordingly, the groups do not support a like-for-like comparative estimate from the information reported.

Third, some regimen descriptions are incomplete. Several arm descriptions end after Cycle 1 or begin a Cycle 2 description without specifying the full regimen. The registry lists methotrexate and laboratory biomarker analysis, but does not describe their roles in the arm regimens. The eligibility criteria, participant characteristics, and reasons for assignment to particular groups are also not reported in the registry record.

Fourth, the adverse-event data are descriptive counts. The registry does not report exposure time, exposure-adjusted event rates, or deaths for the listed groups. It does not explain the denominator for the serious-event term percentages or whether a participant could contribute to multiple terms. Missing-data handling and analysis populations are not reported.

Finally, the record lists 0 analyses and does not provide a statistical analysis plan, multiplicity strategy, sample-size justification, or inferential results. Interim analyses and stopping rules are not reported in the registry record. These omissions limit assessment of precision, robustness, and reproducibility.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Clofarabine and Cytarabine in Treating Young Patients With Refractory or Relapsed Acute Myeloid Leukemia or Acute Lymphoblastic Leukemia — NCT00372619.
• Registry record last updated 2017-06-05; status last verified 2017-02.
• Results first posted 2014-02-03.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry record describes a completed, nonrandomized phase I/II parallel study sponsored by Children's Oncology Group, with enrollment of 74. It lists clofarabine with cytarabine in young patients with refractory or relapsed AML or ALL, as well as an ambiguous-lineage arm. Clofarabine dose groups include 40 mg/m² and 52 mg/m²; the phase I portion is described as closed to enrollment as of 09/16/09.

The primary endpoint was overall response after 2 cycles or up to 84 days, defined as CR for ALL and CR or CRp for AML. The registry describes the response criteria but does not report primary response results, confidence intervals, p-values, or analyses. Safety counts and serious-event terms are listed by group. Analysis populations, missing-data procedures, multiplicity methods, and deaths are not reported in the registry record.