Alisertib in Adults With Nonhematological Malignancies, Followed by Alisertib in Lung, Breast, Head and Neck or Gastroesophageal Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Alisertib in Adults With Nonhematological Malignancies, Followed by Alisertib in Lung, Breast, Head and Neck or Gastroesophageal Malignancies (NCT01045421), a randomized phase 1, phase 2 study. The primary outcome is Phase 1: Number of Participants With Dose-Limiting Toxicities (DLTs).

2. Clinical Question

3. Trial Design

This completed study combined a phase 1 dose-escalation portion with a phase 2 portion. Its brief description characterizes it as open-label and multicenter. The official title identifies phase 1 evaluation of MLN8237 in adults with nonhematological malignancies, followed by phase 2 study in lung, breast, head and neck, or gastroesophageal malignancies. Total enrollment is listed as 273. The record lists the study start as 2010-02 and primary completion as 2013-05.

The intervention was alisertib, administered as an enteric-coated tablet. Phase 1 arms report dose groups at 10 mg, 20 mg, 40 mg, 50 mg, and 60 mg, as well as a separate 50 mg pancreatic cancer group. These are not described as comparisons against another treatment. The phase 2 safety groups are labeled for breast cancer, gastric cancer, head and neck squamous cell carcinoma, non-small cell lung cancer, and small cell lung cancer. The record also includes a phase 1 dose-escalation safety group.

There is an apparent design-label tension in the registry record: allocation is listed as randomized, but the model is listed as single group, masking as none, and the reported intervention arms are all alisertib groups. How the allocation label relates to the dose cohorts is not reported in the registry record. A randomization procedure, the phase 2 allocation structure, and a comparator are not described in the posted information.

The primary outcome is a phase 1 count of participants with dose-limiting toxicities. The listed secondary outcomes cover phase 2 progression-free survival, time to disease progression, duration of response, and adverse events, as well as phase 1 pharmacokinetic measures. The record lists 14 outcome measures and 12 posted secondary outcomes. Outcome definitions and results are not all available in the information summarized in the registry record. No protocol-based sample-size justification or detailed statistical analysis plan is reported in the registry record.

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: Participants with a historical diagnosis of relapsed or refractory advanced nonhematological malignancies were enrolled in 1 of the 2 stages, Phase 1 (lead-in alisertib dose escalation stage) and Phase 2 (efficacy and safety assessment stage for alisertib dose determined in Phase 1).

Only the first 6 of 11 groups are shown.

Phase 1

Reasons for not completing this period:

Phase 2

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: DLT-Evaluable population: all participants in the Phase 1 portion of the study who either experienced DLT during Cycle 1 or completed at least 85% of the planned doses of alisertib and had sufficient follow-up data to allow the investigators and sponsor to determine whether DLT occurred.

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 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Phase 1: Number of Participants With Dose-Limiting Toxicities (DLTs), measured over Phase 1: Cycle 1 Day 1 to Cycle 2 Day 21.

The primary endpoint is the number of participants with dose-limiting toxicities in phase 1. A DLT is a treatment-related toxicity that meets specified severity, duration, or treatment-delay criteria during the assessment window from Cycle 1 Day 1 to Cycle 2 Day 21. The registry says toxicity was graded according to National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0, and that the investigator determined whether a toxicity was related to alisertib.

The definition includes, among other events, Grade 4 neutropenia lasting more than 7 days, Grade 4 thrombocytopenia lasting more than 7 days, platelet count below 10,000 cells/mm3, and a delay of more than 7 days in starting the subsequent cycle because of treatment-related toxicity. It also specifies exceptions and qualifications for certain nonhematological toxicities. These rules make the endpoint a defined safety measure rather than a count of every symptom or abnormality.

In dose escalation, a DLT count can inform the observed tolerability of a dose during a restricted early-treatment window. It is not a complete account of later toxicity or overall safety. The endpoint population is also selected: participants had to experience a DLT during Cycle 1 or meet the planned-dose and follow-up criteria described by the registry. The record reports counts by phase 1 group, but does not provide an inferential estimate or uncertainty interval for the primary endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary result is a count outcome: it records the number of participants meeting the DLT definition within each listed phase 1 group. For a count reported alongside a group size, the basic descriptive reading is the number of participants with the event and the number assessed in that group. This kind of result is different from a continuous outcome such as a mean laboratory value or a time-to-event outcome such as progression-free survival.

What does the DLT count represent?

The DLT count is based on a rule-defined endpoint. The registry specifies toxicity grading, investigator attribution to alisertib, a defined follow-up interval, and criteria based on severity, duration, bleeding, or treatment delay. Applying these criteria creates a consistent endpoint for reporting, but the posted aggregate counts do not show how each event met the definition. The record does not report event-level adjudication details or a separate count for each DLT criterion.

The population definition matters. The DLT-evaluable population includes phase 1 participants who had a DLT during Cycle 1 or completed at least 85% of planned doses and had sufficient follow-up to determine whether DLT occurred. This is not necessarily identical to everyone enrolled in phase 1. The registry gives group denominators with the DLT counts, but does not provide a complete accounting of enrollment, exclusions, or missing assessments for the primary outcome. The relationship between total enrollment and each endpoint population is not reported in the registry record.

Why are the counts descriptive?

A descriptive count reports what was observed in the recorded group. Inferential methods would quantify uncertainty or compare groups under stated assumptions. The registry does not post confidence intervals, p-values, model estimates, or an analysis comparing the dose groups for this primary outcome. Since the phase 1 groups have different listed doses and small group sizes, the raw counts alone do not establish that one dose has a higher or lower underlying DLT frequency than another. The registry also does not describe randomized assignment to these dose groups, despite listing allocation as randomized.

In general, a confidence interval expresses the range of values reasonably compatible with observed data under a statistical model. For event counts in small groups, such intervals can be wide, especially when few or no events are observed. No such intervals are posted for the DLT endpoint, so the precision of the underlying event frequencies cannot be evaluated from the primary results alone. No interval should be inferred from the displayed counts.

How do the other outcomes differ?

The registry lists phase 2 progression-free survival, time to disease progression, and duration of response, which are time-to-event outcomes. General time-to-event analysis accounts for the time until an event and can also accommodate participants whose event has not been observed by the end of follow-up, using censoring methods. The record names these outcomes but does not post their results or analysis details in the information. No survival curves, hazard ratios, or event totals are available in the registry record.

Other listed outcomes concern pharmacokinetics, including maximum observed plasma concentration, time to maximum concentration, area under the concentration-time curve over the dosing interval, and terminal elimination half-life. These measures describe drug concentration over time; they do not answer the same question as a DLT count. The record lists the measures as phase 1 secondary outcomes, but the corresponding numerical results and analysis methods are not reported in the registry record.

What does the planning illustration mean?

Any sample-size values associated with a planning illustration are hypothetical planning calculations based on the observed effect under stated assumptions, not the trial's own design justification and not evidence that the study had a particular probability of detecting an effect. The stated assumptions include 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 fraction of the observed effect. For time-to-event outcomes, the planning quantity is total events, while other outcomes use participants per arm. The primary endpoint as posted has no estimable two-arm comparison, so such a comparison does not supply a trial-specific power analysis for these DLT counts. The registry does not report its original sample-size rationale.

Over-dispersion is a general issue for count outcomes when their variability exceeds what a simple count model assumes. It can make standard errors and confidence intervals too narrow if a model fails to account for the extra variability. The primary outcome is presented as participants with an event, rather than a repeated count of events per participant, and no over-dispersion model is reported. This concept does not change the posted group counts; it is relevant only if a model is fitted and its assumptions are appropriate.

Overall, the registry supports a descriptive reading of the phase 1 counts and the listed safety summaries. It does not provide enough analysis detail to reconstruct adjusted estimates, account for covariates, or assess robustness to alternative statistical assumptions.

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("Phase 1: MLN8237 10 mg", "Phase 1: MLN8237 20 mg", "Phase 1: MLN8237 40 mg", "Phase 1: MLN8237 50 mg", "Phase 1: MLN8237 60 mg", "Phase 1: MLN8237 50 mg- Pancreatic Cancer")
n      <- c(1, 3, 4, 12, 3, 1)
events <- c(0, 0, 0, 2, 0, 0)
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  Phase 1: MLN8237 10 mg  1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2
ARM  Phase 1: MLN8237 20 mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1: MLN8237 40 mg  4  0  0.0  0.0  60.2  0.0  49.0  0.5  52.2
ARM  Phase 1: MLN8237 50 mg  12  2  16.7  2.1  48.4  4.7  44.8  5.0  45.4
ARM  Phase 1: MLN8237 60 mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1: MLN8237 50 mg- Pancreatic Cancer  1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2

11. Multiplicity, Missing Data and Analysis Populations

The study contains several outcome measures spanning DLTs, adverse events, tumor outcomes, and pharmacokinetic measures. When many outcomes or comparisons are analyzed, the chance of at least one apparently unusual result can increase. General approaches to multiplicity include prespecifying a primary hypothesis, controlling a family-wise error rate, or using a false-discovery-rate procedure. Which, if any, approach was planned for this study is not reported in the registry record.

The registry identifies a phase 1 primary endpoint and lists secondary outcomes, but does not post analyses or a detailed statistical analysis plan. It is therefore not possible to determine whether additional comparisons were prespecified, whether any multiplicity adjustment was planned, or how results across the phase 1 and phase 2 portions were intended to be interpreted together. No adjustment should be assumed.

Missing outcome data also affect interpretation. The DLT-evaluable population has defined eligibility criteria, including sufficient dosing and follow-up for some participants. However, the registry does not report the number of participants excluded from that population, the reasons for exclusion, the amount of missing DLT follow-up, or a sensitivity analysis for missing outcomes. It does not state how withdrawals or incomplete assessments were handled in the posted counts.

Analysis populations for the phase 2 outcomes are not described in sufficient detail to establish whether reported participants were analyzed according to enrollment, treatment received, or another rule. The results for those outcomes are not posted in the information summarized in the registry record. Without analysis-population definitions and missing-data methods, readers cannot assess the potential impact of incomplete follow-up on those outcomes.

These are reporting limitations, not proof that no planning or analysis occurred. The registry record simply does not report the multiplicity procedures, missing-data strategy, or complete analysis-population rules needed to evaluate them.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: neutropenia (100.0%), nausea (100.0%), leukopenia (100.0%), stomatitis (100.0%), flank pain (100.0%).

The serious adverse-event figures count participants reported with one or more serious adverse events in each listed safety group. In the phase 1 dose-escalation group, 13 of 23 participants were reported with a serious adverse event; in the separate phase 1 50 mg pancreatic cancer group, the count was 1 of 1. For phase 2, the counts were 23 of 53 in breast cancer, 30 of 55 in gastric cancer, 19 of 55 in head and neck squamous cell carcinoma, 8 of 26 in non-small cell lung cancer, and 28 of 60 in small cell lung cancer.

These are group-specific descriptive counts. They do not indicate that an event was caused by alisertib, and the registry does not provide a comparative control group for interpreting them. The listed phase 1 and phase 2 groups also differ in their disease categories and sizes, so their counts should not be treated as direct comparisons.

The registry separately lists serious adverse-event terms, each with a posted value of 100.0%: neutropenia, nausea, leukopenia, stomatitis, flank pain, and pancreatic carcinoma. The denominator and the precise interpretation of these term percentages are not reported in the registry record. They should not be read as percentages of all trial participants or as rates for each safety group without that information. Death counts are not reported for the listed groups.

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

The posted results establish that DLTs were recorded in the phase 1 50 mg group and not recorded in the other listed phase 1 dose-group entries. They also document serious adverse events across the phase 1 and phase 2 safety groups. These observations are bounded by the endpoint definitions, the assessment period, and the populations represented in each group.

They do not establish that alisertib is effective or ineffective for any malignancy. Although phase 2 tumor outcomes are listed, their results are not reported in the registry record. There is also no posted comparative treatment-effect estimate for the primary DLT endpoint. The registry does not describe an active control or placebo arm in its intervention and group details.

The counts do not establish that the observed differences among dose groups are statistically significant, that DLT risk rises or falls with dose, or that a specific dose is optimal. The phase 1 groups are small, and the record provides no uncertainty intervals or formal dose-response analysis. A zero count in a group is an observation within that group, not evidence that the underlying risk is zero.

Likewise, the serious adverse-event counts do not by themselves establish causality, comparative safety, or the full burden of adverse events. They summarize participants reported with at least one serious event in group-specific denominators. The registry includes additional counts for other treatment-emergent adverse events, but these do not replace detailed event-level data, timing, severity, or attribution. The listed serious-event terms lack reported denominator detail.

The study record is useful for identifying the design, outcome definitions, group counts, and gaps in posted analyses. It cannot support conclusions that require unreported phase 2 results, detailed statistical methods, or a well-defined comparison. The appropriate reading is descriptive: the registry records the stated counts, with limited information about their precision and no posted inferential analysis for the primary endpoint.

14. Why This Trial Matters Statistically

This trial illustrates a common distinction in early-phase research: a safety endpoint can be recorded as a straightforward count, while the interpretation depends on population rules, toxicity definitions, follow-up, and cohort structure. The registry defines DLTs using severity and duration criteria and specifies a DLT-evaluable population. Those details are necessary to understand what the count includes and what it leaves out.

The phase 1 counts also show why small cohorts call for caution. Zero events in a small group do not establish a zero underlying risk, and a count in one cohort cannot by itself demonstrate a dose-response relationship. Precision estimates and a prespecified analytic approach would help characterize uncertainty, but neither is posted for this primary endpoint.

The record also highlights the value of consistency between design labels and descriptions. It lists randomized allocation, yet identifies a single-group model and reports alisertib dose groups without a described comparator. The registry does not explain this relationship. Clear reporting of how allocation was conducted and what groups were intended to compare would be important for interpreting any treatment-effect analysis.

Finally, the study spans distinct phases and outcome types. A phase 1 DLT count, phase 2 time-to-event endpoint, pharmacokinetic measure, and serious adverse-event summary answer different questions. Keeping these outcome types and populations separate prevents descriptive safety information from being mistaken for efficacy evidence or a comparative treatment effect. Missing analysis details do not negate the posted observations, but they limit what statistical conclusions can be drawn from them.

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 main limitation for interpreting the primary endpoint is that the registry posts only group counts, without confidence intervals, a hypothesis test, or a formal dose-response analysis. Several listed groups have small denominators. The primary result is defined for a DLT-evaluable population, but a participant flow and a full account of exclusions or incomplete assessments are not reported in the registry record.

The endpoint is restricted to DLTs during Cycle 1 Day 1 to Cycle 2 Day 21. It does not, by itself, describe adverse events outside that window or provide a complete account of later safety. The record supplies the DLT criteria, but does not describe the individual events that made up the reported count or provide event-level attribution detail beyond the stated investigator assessment.

The registry's design labels are not fully reconciled in the posted description: allocation is listed as randomized, while the model is single group and the listed arms contain alisertib dose groups. The randomization procedure, comparator structure, and relation between these labels are not reported in the registry record. This limits assessment of whether any group differences have a causal interpretation.

For serious adverse events, the posted counts are group-specific and descriptive. A denominator and interpretation for the listed 100.0% serious-event terms are not reported, and deaths are not reported for the listed safety groups. The phase 2 efficacy and pharmacokinetic outcome results are also not included in the record information. Statistical methods for those outcomes, multiplicity handling, missing-data procedures, and the original sample-size justification are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Alisertib in Adults With Nonhematological Malignancies, Followed by Alisertib in Lung, Breast, Head and Neck or Gastroesophageal Malignancies — NCT01045421.
• Registry record last updated 2016-08-15; status last verified 2016-07.
• Results first posted 2016-08-15.
• 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

NCT01045421 is a completed, open-label, multicenter phase 1 and phase 2 study of MLN8237 (alisertib) in adults with nonhematological malignancies. The registry lists enrollment of 273, a study start of 2010-02, and primary completion in 2013-05. Phase 1 involved dose escalation; phase 2 included disease-specific groups. The record lists randomized allocation but a single-group model and does not explain the relationship between those design labels.

The primary phase 1 outcome was the number of participants with DLTs during Cycle 1 Day 1 to Cycle 2 Day 21. Posted counts range from 0 of 1 to 2 of 12 across the listed phase 1 entries. Serious adverse-event counts are also posted for phase 1 and phase 2 groups. No inferential analysis is posted for the primary outcome. Phase 2 progression and response outcomes, analysis methods, missing-data procedures, and the trial's sample-size justification are not reported in the registry record.