A Phase II Study of Selinexor Plus Cytarabine and Idarubicin in Patients With Relapsed/Refractory Acute Myeloid Leukemia (AML): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase II Study of Selinexor Plus Cytarabine and Idarubicin in Patients With Relapsed/Refractory Acute Myeloid Leukemia (AML) (NCT02249091), a non-randomized phase 2 study. The primary outcome is Number of Participants With CR/CRi = Overall Reponse Rate.

2. Clinical Question

The clinical question is whether this combination can produce a documented remission response during the induction period in patients whose AML is relapsed or refractory. The primary assessment used the best response after selinexor treatment, over one or two induction cycles, with the stated timeframe of 1-2 induction cycles (4 - 8 weeks). The response definitions distinguish marrow findings from recovery of blood counts: CR requires specified marrow and blood-count criteria, while CRi allows incomplete recovery of neutrophils and/or platelets. MLFS is defined by marrow findings and the absence of specified disease features, and was included in the responder group.

This design describes treatment and response within the studied cohorts; it does not provide a randomized comparison against another treatment. The central efficacy question is therefore about the observed response in the treated population, not a causal contrast between assigned treatments. Because the registry record does not report the primary response result, it is not possible to quantify the response rate from that record.

3. Trial Design

This was a Phase II, non-randomized study of selinexor combined with cytarabine and idarubicin in patients with relapsed or refractory AML. The registry record lists a sequential model and no masking. It records enrollment of 42 and a completed status. The start is listed as 2014-09 and primary completion as 2018-07-31. GSO Global Clinical Research BV is identified as the sponsor.

The record describes two cohorts, rather than a randomized treatment comparison. Cohort 1 is identified as selinexor 40 mg/m^2 with cytarabine and idarubicin; Cohort 2 is identified as selinexor 60 mg flat dose with the same chemotherapy drugs. Both descriptions specify cytarabine as a continuous infusion at 100 mg/m² on days 1-7 and idarubicin at 10 mg/m^2 intravenously on days 1, 3, and 5, every 4 weeks. If a second cycle is applied, the description says idarubicin is given only on days 1 and 3. Treatment with selinexor is described as continuing for up to 2 induction cycles. The registry record does not provide a complete selinexor administration schedule in the cohort descriptions.

The primary endpoint timeframe is 1-2 induction cycles (4 - 8 weeks). The registry record does not report a randomized allocation procedure, a concurrent control arm, masking, or a statistical justification for the enrollment target. The cohorts differ in selinexor dose and are not described as randomized. Any numerical difference between cohort outcomes would therefore be a descriptive comparison, potentially affected by differences between the groups as well as by dose.

The record lists eight outcome measures, including the primary response endpoint and secondary measures such as partial remission, transplantation after induction therapy, early death, and several survival outcomes. It records no statistical analyses. The record does not report endpoint results for the primary or secondary outcomes, nor does it describe a prespecified statistical analysis plan.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All patients who have received study medication at least once and for whom post-baseline efficacy data upon treatment was available.

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 Number of Participants With CR/CRi = Overall Reponse Rate, measured over 1-2 induction cycles (4 - 8 weeks).

The primary endpoint is the count of participants with CR or CRi, described as the overall response rate. The registry record also says that participants with MLFS were included among responders. Thus the response grouping is broader than CR alone: it includes complete response, complete response with incomplete blood count recovery, and morphologic leukemia free-state as defined in the record. The endpoint is assessed using the best response after selinexor treatment, after the induction cycle or cycles.

CR is defined using blood counts and marrow findings, including bone marrow blasts below 5%, no Auer rods, and no evidence of extramedullary disease. CRi uses the same criteria, but permits an absolute neutrophil count below 1.0x10^9/L and/or a platelet count below 100x10^9/L. MLFS is defined by bone marrow blasts below 5%, no Auer rods, and no evidence of extramedullary disease. The definitions are attributed in the registry record to recommendations on AML diagnosis and management from an international expert panel on behalf of the European LeukemiaNet.

General statistical context: a response endpoint is useful for describing whether participants meet an agreed clinical and laboratory definition at a specified assessment. A count can be converted into a rate only when its relevant analysis population is clear. The registry record defines the primary population as participants who received study medication at least once and had post-baseline efficacy data available upon treatment. It does not report the number meeting the response definition or a rate, so the primary endpoint cannot be quantified from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical context: methods should match the outcome type, study design, and question. Here, the primary outcome is a binary response status summarized as a count of participants who met the response definition. The registry record identifies the endpoint as a count and names it the overall response rate, but does not report the observed count or rate. No statistical analyses are recorded. The record therefore does not allow assessment of a fitted model, an adjusted effect estimate, a confidence interval, or a hypothesis test for the primary outcome.

What does a response count represent?

A response count represents the number of people in a defined analysis population who meet the endpoint criteria. A rate expresses that count relative to an appropriate denominator. Defining the analysis population matters: the primary endpoint population in the registry record includes those who received study medication at least once and had post-baseline efficacy data available upon treatment. This is not necessarily identical to every enrolled participant. The record does not state how many participants met this primary analysis-population definition, how many lacked post-baseline data, or how missing response assessments were handled. It also does not report a primary response denominator or result.

General statistical context: a confidence interval gives a range of values compatible with an estimate under the statistical model and sampling assumptions. For a binary response rate, interval estimation can communicate uncertainty that a single point estimate cannot. It does not remove bias from selection, missing outcome data, or a non-randomized design. The registry record does not report a confidence interval for the primary endpoint, so its precision cannot be judged from the record.

What does non-randomized allocation imply?

General statistical context: random assignment is one way to make treatment groups comparable, on average, with respect to both measured and unmeasured baseline factors. In a non-randomized study, group membership may be associated with participant characteristics, clinical decisions, calendar time, or other factors. Statistical adjustment can sometimes address measured differences, but it cannot guarantee removal of unmeasured confounding. The registry record identifies a non-randomized sequential model and two cohorts with different selinexor doses. It does not report an adjusted comparison or the covariates that might be used in one. Cohort outcome contrasts should therefore be read descriptively, not as isolated dose effects.

General statistical context: sequential cohorts can also be affected by changes in practice or patient mix over time. The record labels the model sequential, but does not report enough detail to characterize the timing or allocation process beyond that label. It would be inappropriate to infer a specific protocol rationale or claim that cohort differences were controlled.

How should the sample-size illustrations be read?

The planning illustrations use the observed effect as a reference and assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. For time-to-event outcomes, the illustrations give a total number of events using the Schoenfeld approach; for other outcomes, they give participants per arm. These are planning illustrations, not the study's original design or observed power. They do not establish that the trial was designed around those assumptions. The registry record does not report the trial's own sample-size justification. Since the registry record does not report the primary response result or an inferential analysis, these planning assumptions should not be mistaken for evidence about efficacy or the precision of the trial's findings.

General statistical context: over-dispersion is relevant when count data vary more than a chosen probability model expects; ignoring it can make uncertainty appear too small. The primary endpoint is described as a participant count with a binary response definition, not as a repeated event count. The registry record does not report use of an over-dispersed model, and there is no basis to infer that such a model was used.

What can be concluded from the listed analyses?

The record lists zero statistical analyses. It does not report a p-value, a confidence interval, an adjusted estimate, or a model specification for the primary endpoint. General statistical context: a p-value, when reported, assesses compatibility of data with a specified null model; it is not the probability that a treatment works, nor does it communicate the magnitude or clinical importance of an effect. An estimate and its uncertainty, interpreted alongside design limitations, are usually more informative. In the absence of a reported analysis, no such inferential interpretation can be made for the primary response endpoint.

The secondary outcomes include several binary and time-to-event measures. General statistical context: time-to-event analyses account for the time until an event and may also handle participants whose event status is not observed through the end of follow-up. Appropriate interpretation depends on definitions of the event, follow-up, censoring, and analysis population. Those details and results are not reported in the registry record. No survival estimate or comparison should be inferred from the outcome names alone.

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("Cohort 1 / Selinexor 40 mg/m^2 in Combination With", "Cohort 2 / Selinexor 60 mg Flat Dose in Combinatio")
n      <- c(27, 15)
events <- c(15, 6)
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  Cohort 1 / Selinexor 40 mg/m^2 in Combination With  27  15  55.6  35.3  74.5  37.3  72.4  37.2  72.5
ARM  Cohort 2 / Selinexor 60 mg Flat Dose in Combinatio  15  6  40.0  16.3  67.7  19.8  64.3  19.8  64.6

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists eight outcome measures. The primary measure is the count of participants with CR/CRi, with MLFS included in the responder group. Secondary measures include partial remission, the percentage transplanted after induction therapy, early death, overall survival, relapse-free survival, event-free survival, and progression-free survival. Having multiple endpoints can create multiplicity: when many hypotheses are tested, the chance of at least one apparently unusual result can increase. General statistical context: studies may address this by naming a primary endpoint in advance, adjusting for multiple testing, or interpreting secondary analyses as exploratory. The registry record does not report a multiplicity adjustment or a plan for interpreting multiple tests.

The record lists zero statistical analyses. It does not report hypothesis tests or endpoint estimates for the primary or secondary outcomes. A named primary endpoint does not by itself reveal whether a multiplicity strategy was specified or applied. The registry record does not report such a strategy.

The primary population is defined as participants who received study medication at least once and for whom post-baseline efficacy data upon treatment were available. This makes the availability of post-baseline efficacy data relevant to inclusion. The record does not state how many participants lacked such data, whether any assessments were missing, or how missing responses were handled. It also does not report whether participants were analyzed in another population, such as all enrolled participants, or how deviations from treatment were addressed.

General statistical context: missing outcome data can affect both the denominator and the estimated response rate. Excluding participants without observed outcomes may produce a different estimate from treating missing responses in another way, and the direction of any effect depends on why data are missing. Without a stated missing-data approach and the relevant counts, readers cannot assess the impact of missingness. Details on missingness, analysis populations beyond the primary definition, and multiplicity are not reported in the registry record.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: bone marrow hypocellular (13.3%), lung infection (13.3%), anemia (6.7%), asystole (6.7%), diarrhea (6.7%).

The registry safety entries distinguish serious adverse events from other adverse events and report participants at risk, participants affected, and deaths by cohort. In Cohort 1, 12 of 27 participants at risk are listed as affected by serious adverse events; in Cohort 2, 8 of 15 are listed. The entries list deaths of 16 in Cohort 1 and 9 in Cohort 2. These are separate registry counts and should not be read as proof that a serious event or death was caused by treatment.

The most frequent listed serious adverse event terms are bone marrow hypocellular and lung infection, each reported at 13.3%. Anemia, asystole, diarrhea, and hemophagocytosis syndrome are each reported at 6.7%. The registry record does not provide the event-level timing, duration, severity grading, or treatment attribution in these summaries. The percentages should be read as recorded; the denominator and details behind each term are not specified in the safety summary.

The other adverse event entries indicate that 27 of 27 participants at risk in Cohort 1 and 15 of 15 in Cohort 2 were affected. General statistical context: adverse-event counts describe events observed and recorded during a specified safety assessment; they do not alone establish causation. Interpretation also depends on exposure, follow-up, definitions, and attribution procedures. Those details are not reported in the registry record, so the entries support description, not a treatment-causality conclusion.

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

The registry record establishes that a Phase II study of selinexor with cytarabine and idarubicin was completed in relapsed or refractory AML, and it defines a primary response endpoint. It does not report the primary response count or rate, nor results for the listed secondary outcomes. It also records no statistical analyses. Therefore, the registry record does not allow readers to quantify the study's efficacy findings or assess their statistical precision.

The endpoint definition is informative: CR, CRi, and MLFS are grouped as responders, with CRi allowing incomplete recovery of blood counts and MLFS defined by marrow findings. But knowing how a response was defined is different from knowing how many participants met that definition. The outcome definition cannot substitute for the missing result.

The two cohorts received different selinexor dose regimens, but allocation was non-randomized and the model was sequential. This means cohort differences, if observed, would not automatically represent an effect of dose. A randomized control comparison is not described. The registry record does not report a statistical adjustment that would support a causal comparison between these cohorts.

Safety entries describe serious adverse events, other adverse events, and deaths by cohort. They should be interpreted as registry-recorded counts, not as proof of causation. The record does not provide detail sufficient to establish whether individual events were related to treatment, or to characterize their timing and duration.

General statistical context: an observed response estimate, when reported, is a description of the studied population and is subject to sampling uncertainty and design limitations. A confidence interval can describe some of that uncertainty but cannot correct for non-random allocation or unreported missing data. Since neither the primary estimate nor its interval is reported in the registry record, the strength and precision of any response finding cannot be judged. These limitations do not show that the combination is ineffective; they mean the registry record does not provide the information needed to draw an efficacy conclusion.

14. Why This Trial Matters Statistically

This study is statistically notable as a Phase II investigation of a drug combination in relapsed or refractory AML, with a response endpoint defined using CR, CRi, and MLFS criteria. Clear endpoint definitions matter because they establish which clinical and laboratory outcomes count toward the response measure. Here, the registry record also identifies the population intended for the primary assessment: participants who received study medication at least once and had post-baseline efficacy data available.

The design is important to interpretation. The two cohorts are non-randomized and sequential, and differ in selinexor dose. This structure can describe outcomes under each cohort's regimen, but it does not provide the protection against confounding that randomization can offer. The study should not be treated as a randomized dose comparison. Understanding that distinction is essential when readers encounter any between-cohort contrast.

The record also illustrates a difference between a registered outcome and a reported result. The primary endpoint and its timeframe are specified, but the response count, response rate, and statistical analyses are not reported in the registry record. Consequently, the endpoint's definition can be discussed, while its observed magnitude and precision cannot be evaluated from that record. This distinction helps prevent protocol descriptions from being mistaken for findings.

Finally, the safety entries provide counts by cohort and a list of serious adverse event terms. These can be described, but their meaning depends on event definitions, exposure, follow-up, and attribution details that are not reported in the registry safety summary. The study's statistical value for readers is therefore partly methodological: it shows why design, endpoint definitions, analysis populations, missing-data handling, and transparent results reporting all matter when interpreting clinical trial evidence.

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.

The most direct limitation for interpreting efficacy is that the registry record does not report the primary response count or rate, despite defining CR/CRi and including MLFS in the responder group. It also does not report results for the listed secondary outcomes. No statistical analyses are recorded, so there are no reported estimates, confidence intervals, or tests with which to assess effect magnitude or precision.

The study is non-randomized and sequential, with two selinexor dose cohorts. This limits causal interpretation of any cohort contrast: group differences could reflect factors other than dose. The registry record does not report a randomized control group or an adjusted comparison. The cohort descriptions also do not provide a complete selinexor administration schedule.

The primary analysis population requires receipt of study medication and post-baseline efficacy data. The registry record does not report the number meeting that definition, the extent of missing efficacy data, or how missing outcomes were handled. It does not describe analysis populations beyond the stated primary population, or report how treatment discontinuations and protocol deviations were addressed.

The record does not report the trial's sample-size justification, a multiplicity strategy, a detailed statistical plan, or the assumptions used for any primary endpoint analysis. The planning illustrations are not the original trial design and are not observed power. Safety summaries provide counts and selected serious event terms, but do not report event-level timing, duration, attribution, or the denominator behind each listed percentage. These limitations constrain both efficacy interpretation and detailed safety assessment.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Phase II Study of Selinexor Plus Cytarabine and Idarubicin in Patients With Relapsed/Refractory Acute Myeloid Leukemia (AML) — NCT02249091.
• Registry record last updated 2021-08-25; status last verified 2021-08.
• Results first posted 2021-08-25.
• 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.

19. Record Summary

The registry record describes a completed, non-randomized Phase II study of selinexor with cytarabine and idarubicin in relapsed or refractory AML. Enrollment is listed as 42. The model is sequential, masking is none, and the two cohorts are distinguished by selinexor 40 mg/m^2 or 60 mg flat dose. The primary endpoint is the number of participants with CR/CRi, with MLFS included among responders, assessed after one or two induction cycles.

The registry record does not report the primary response result, secondary outcome results, or statistical analyses. Safety entries report serious adverse events, other adverse events, and deaths by cohort, alongside selected serious adverse event terms and percentages. The non-randomized cohort structure and missing outcome results limit efficacy interpretation; the safety entries are descriptive and do not establish treatment causation.