Study of ASTX029 in Subjects With Advanced Solid Tumors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of ASTX029 in Subjects With Advanced Solid Tumors (NCT03520075), a randomized phase 1, phase 2 study. The primary outcome is Phase 1: Number of Participants With Dose Limiting Toxicities (DLTs).

2. Clinical Question

The registry describes a first-in-human, open-label, multicenter Phase 1-2 study assessing safety, pharmacokinetics, pharmacodynamics, and preliminary clinical activity. Its overall population was people with advanced solid malignancies for whom approved or available therapies were not suitable. Phase 1 examined multiple ASTX029 dose levels and formulations, while Phase 2 enrolled cohorts defined by cancer type or molecular features.

The primary endpoint is a safety count, not a comparison of tumor response between ASTX029 and another treatment. The posted Phase 1 results give dose-limiting toxicity counts by cohort during Cycle 1, whose length was 21 days. The record also reports Phase 2 disease control by cohort and unadjusted comparisons of Cohorts B and C with Cohort A. These are different questions: dose-limiting toxicity concerns selected early safety events, while disease control is a later clinical activity measure assessed over follow-up.

The record labels the study randomized, but the reported arms and outcomes are organized by dose-escalation, expansion, and tumor-defined cohorts. How randomization was implemented, and whether it applied to particular cohorts or comparisons, is not reported in the registry record. Consequently, the cohort results should not be read as though the registry describes a straightforward randomized comparison of ASTX029 against a separate control treatment.

3. Trial Design

This is a Phase 1-2 study with an open-label, multicenter description. The record lists randomized allocation, no masking, and a parallel model. It reports enrollment of 192 and a completed status. The start date is 2018-05-07 and the primary completion date is 2024-08-01. The record does not give a detailed account of how the randomized allocation operated across the listed cohorts.

Phase 1A included dose-escalation cohorts using ASTX029 at doses from 10 mg through 280 mg, with powder-in-bottle or tablet formulations. The treatment schedule was once daily on Days 1-21 of each 21-day cycle. Cohorts differed in dose, formulation, and whether dosing was under fed or fasted conditions. The Phase 1B expansion used 200 mg tablets, once daily on Days 1-21 of each 21-day cycle, under fasted conditions. The listed median treatment duration for these Phase 1 cohorts was 1.6 months.

Phase 2 comprised six cohorts, all receiving 200 mg tablets once daily on Days 1-21 of each 21-day cycle. Cohort A involved NRAS-mutant melanoma; Cohort B, KRAS-mutant or KRAS-amplified non-small cell lung cancer; Cohort C, BRAF V600-mutant cancers other than colorectal cancers; Cohort D, BRAF-fusion cancers; Cohort E, gynecological cancers with MAPK-pathway alterations; and Cohort F, tumors with other gene aberrations that upregulate the MAPK signal pathway. The registry does not report a separate control treatment.

The primary endpoint is reported for Phase 1, whereas Phase 2 disease-control results are posted as a secondary outcome. The study’s full protocol, dose-escalation decision rules, randomization procedures, and statistical analysis plan are not reported in the registry record. Those omissions limit how far the design can be reconstructed from the posted information alone.

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: A total of 192 participants were enrolled in the study to receive ASTX029, of which 2 participants died before receiving treatment. The study was conducted in 2 phases: Phase 1 included Cohorts 1-12 in Part A (Dose Escalation) and a single cohort for Part B (Dose Expansion) and Phase 2 included Cohorts A to F.

Only the first 6 of 19 groups are shown.

Phase 1 (Dose Escalation)

Reasons for not completing this period:

Phase 2 (Dose Expansion)

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The safety analysis set included data from all participants who received any amount of study drug.

5. Baseline Characteristics

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

Population: The Safety Analysis Set included data from all participants who received any amount of study drug.

Only the first 6 of 20 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 Cycle 1 (cycle length = 21 days).

The primary endpoint is the number of participants with dose-limiting toxicities during Cycle 1, with a cycle length of 21 days. The safety analysis set included participants who received any amount of study drug. A dose-limiting toxicity was defined as a selected adverse event during the first cycle that met specified severity or clinical criteria.

The definition includes grade 4 thrombocytopenia of any duration; hematologic toxicity of grade 3 or higher with complications; febrile neutropenia of any duration or grade 4 neutropenia lasting 5 days or more; liver-associated abnormalities; eye disorders of grade 2 or higher; symptomatic grade 2 cutaneous toxicities, including skin rash; and other nonhematologic adverse events of grade 3 or higher, with specified exceptions for grade 3 nausea, vomiting, or diarrhea. It also includes an event that, in the opinion of the Data and Safety Review Committee, would make further dose escalation pose unacceptable risk.

This endpoint is used in early-phase dose escalation because it focuses on serious or otherwise clinically consequential events that may inform evaluation of dose levels. It is not a complete measure of every adverse event, nor does it by itself describe longer-term safety. A count with no events in a cohort means that no participant in that cohort was recorded as meeting this particular definition in the stated timeframe; it does not establish that no other adverse events occurred.

The endpoint is reported as counts, with each cohort’s number of participants also stated. The registry does not report a formal comparison between doses or a statistical model for these counts. The results therefore describe the observed cohort experience rather than a tested difference between treatment strategies.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For the primary endpoint, each cohort’s result is a count of participants with a dose-limiting toxicity during Cycle 1. Counts are useful in early-phase safety reporting because the endpoint is defined by whether a participant met the event criteria, rather than by a continuous measurement. The cohort size gives context for the count, but these small, differently sized cohorts do not support treating the observed counts as precise estimates of risk or as direct dose comparisons. The registry does not report confidence intervals for the primary counts or a statistical test across dose cohorts.

The DLT definition is a clinical classification, not simply any adverse event. It combines event types and severity criteria, and allows the Data and Safety Review Committee to classify an event as dose limiting when it indicates that further escalation could create unacceptable risk. The endpoint therefore depends on the stated criteria and the review process. The registry does not report further adjudication details or the committee’s decisions beyond the results.

The Phase 2 disease-control outcome is presented as a percentage with a 95% confidence interval for each cohort. A percentage conveys the observed proportion meeting the outcome definition; the confidence interval expresses uncertainty around that estimate under the statistical method used. A wider interval generally signals less precision. With small cohorts, individual outcomes can materially influence the percentage, so intervals are important context. The registry does not report the detailed disease-control definition or the method used to calculate its confidence intervals.

The posted pairwise calculations compare Cohort B and Cohort C separately with Cohort A using risk differences. A risk difference is the difference between the observed proportions in two groups; a value below zero means the proportion in the named cohort was lower than in Cohort A for this outcome. The reported estimates are -0.056 for Cohort B versus Cohort A and -0.240 for Cohort C versus Cohort A. Their respective 95% confidence intervals are -0.354 to 0.241 and -0.563 to 0.084. Both intervals include zero, so the data are compatible with no difference as well as with differences in either direction.

The unadjusted p-values for those comparisons are 0.708 and 0.15, respectively. A p-value measures how unusual results at least as incompatible with a null hypothesis would be under specified assumptions; it does not measure the probability that the null hypothesis is true, or the size or clinical importance of an effect. These are unadjusted values, and the registry does not report a multiplicity procedure or a formal analysis plan. They should therefore be treated as exploratory rather than definitive evidence.

The confidence intervals and p-values in these comparisons are recalculated, unadjusted summaries based on posted outcome counts. The registry lists no reported analyses for the secondary outcome. This distinction matters: a calculation based on posted summary data does not substitute for a sponsor’s full analysis, which can use prespecified definitions, participant-level data, covariate adjustments, and handling rules not available in a summary record. The record does not describe such an analysis for the comparisons at issue.

Planning illustrations are also calculated from observed effects and use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They show hypothetical participants per arm for 80% or 90% power at different fractions of the observed effect. Such illustrations describe sample sizes under those assumptions; they are not the observed power of this study, do not establish that the trial was designed around those effects, and are especially sensitive to an unstable observed effect from small cohorts. The registry’s own sample-size justification is not reported in the registry record.

Some common statistical approaches do not fit these posted results. There is no two-arm primary endpoint comparison to analyze, and the DLT count results are not reported with an inferential model. The record does not provide enough information to reconstruct randomization or to determine whether any planned comparisons were prespecified. No estimate should be created by pooling the cohorts, since the doses, formulations, treatment conditions, and tumor populations differ.

For any confidence interval, interpretation depends on the analysis population and outcome definition. The disease-control analysis denominators are reported by cohort, but missing-outcome handling, exclusions, and the analysis population definition for this secondary outcome are not reported in the registry record. The Phase 1 primary outcome does state that the safety analysis set included participants who received any amount of study drug. These population details should not be assumed to apply identically to every secondary endpoint.

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 1A: Cohort 1 Dose Escalation", "Phase 1A: Cohort 2 Dose Escalation", "Phase 1A: Cohort 3 Dose Escalation", "Phase 1A: Cohort 4 Dose Escalation", "Phase 1A: Cohort 5 Dose Escalation", "Phase 1A: Cohort 6 Dose Escalation")
n      <- c(3, 3, 3, 5, 10, 6)
events <- c(0, 0, 0, 0, 1, 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 1A: Cohort 1 Dose Escalation  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1A: Cohort 2 Dose Escalation  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1A: Cohort 3 Dose Escalation  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1A: Cohort 4 Dose Escalation  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Phase 1A: Cohort 5 Dose Escalation  10  1  10.0  0.3  44.5  1.8  40.4  2.3  41.3
ARM  Phase 1A: Cohort 6 Dose Escalation  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists numerous outcomes, including the primary Cycle 1 DLT count, pharmacokinetic measures, and secondary clinical outcomes. It also reports Phase 2 disease-control results by tumor cohort and unadjusted comparisons for Cohorts B and C against Cohort A. With multiple outcomes and comparisons, the chance of at least one apparently unusual result can increase if each is interpreted on its own. A multiplicity adjustment or a clearly defined hierarchy can help control this issue, but the registry record does not report such a procedure.

The record does not report a statistical analysis plan, the prespecified status of the secondary comparisons, interim analyses, stopping rules, or a sample-size justification. It also does not report whether any outcome analyses were adjusted for baseline characteristics. These omissions do not prove that no such plans or procedures existed; they mean their details are not reported in the registry record and cannot be evaluated from the posted information.

The primary safety endpoint specifies a safety analysis set: participants who received any amount of study drug. The registry does not similarly describe the analysis population for each secondary outcome. Although cohort denominators are posted for disease control, handling of missing assessments, incomplete follow-up, or participants excluded from an outcome analysis is not reported in the registry record. The record also does not give a complete account of missing-data methods.

These gaps matter because analysis choices can affect estimates and uncertainty, particularly when cohort sizes are limited. For the disease-control comparisons, the posted p-values are unadjusted and the record lists no reported formal analyses. They should not be treated as confirmatory findings. The DLT counts are directly descriptive, but they too are bounded by the stated safety population, timeframe, and endpoint definition.

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: pneumonia (40.0%), anaemia (33.3%), ascites (33.3%), nausea (33.3%), malaise (33.3%).

The serious adverse-event listings report, by cohort, the number of participants at risk and the number affected. These counts describe recorded serious adverse events in the posted safety data; they are distinct from the primary DLT outcome, which uses a specific event definition and is restricted to Cycle 1. A participant can have a serious adverse event that does not meet the DLT criteria, and the DLT count is not a summary of all serious adverse events.

The listed top serious-event terms include pneumonia at 40.0%, and anaemia, ascites, nausea, malaise, and infectious enterocolitis at 33.3% each. The denominator and method underlying these term percentages are not reported in the registry record, so they should be read as the posted percentages rather than converted into counts or compared across cohorts.

Deaths are also listed by cohort. Those figures alone do not establish that deaths were caused by ASTX029, nor do they describe timing or cause. The registry record does not report a causal attribution for these counts in the information summarized for these outcomes. Differences across cohorts may reflect cohort size and participant characteristics as well as treatment exposure. No pooled or comparative safety estimate should be inferred from the separate cohort listings.

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

The primary result describes the number of participants meeting a defined dose-limiting toxicity criterion during the first 21-day cycle in each Phase 1 cohort. It shows two cohorts with one recorded DLT each and no recorded DLTs in the other listed Phase 1 cohorts. This is an account of the posted observations, not proof that a particular dose is safe or unsafe, and not a comparison against a control treatment.

The endpoint does not include every adverse event, and its timeframe is limited to Cycle 1. Later events and events that do not satisfy the DLT definition are outside this primary measure. The separate serious adverse-event listings provide additional safety information, but they do not identify causation from the counts alone.

The Phase 2 disease-control estimates describe outcomes in cohorts selected by tumor type or molecular feature. The unadjusted comparisons with Cohort A do not establish a clear difference for Cohort B or Cohort C, and the confidence intervals indicate uncertainty. Because Cohort A is not a distinct standard-care or placebo group, and because the cohorts are not interchangeable populations, these results do not estimate a treatment advantage over another therapy.

The record describes a randomized study, but it does not explain how randomization was implemented among the listed cohorts. It therefore does not support treating every between-cohort contrast as randomized. The registry also does not report the complete statistical plan, missing-data handling for secondary outcomes, or multiplicity adjustment. Those limitations constrain interpretation of the secondary results.

Overall, the posted outcomes are descriptive evidence from a Phase 1-2 study of ASTX029. They do not by themselves establish comparative effectiveness, a causal effect on disease control, or a comprehensive characterization of safety. The relevant interpretation is limited to the endpoint definitions, timeframes, populations, and cohort results actually reported.

14. Why This Trial Matters Statistically

This study illustrates why early-phase safety endpoints should be read at the level of their definition and timeframe. A dose-limiting toxicity is a selected event in the first cycle, not a synonym for any adverse event and not a complete account of long-term safety. The distinction is important when interpreting a cohort with no recorded DLTs alongside separate reports of serious or other adverse events.

The study also demonstrates the challenges of interpreting a multi-cohort design. Phase 1 cohorts varied by dose, formulation, and food condition; Phase 2 cohorts were defined by tumor type or molecular alteration. Those features make cohort-specific reporting useful, but they also make direct comparisons difficult. The registry’s randomized label does not resolve this issue because the allocation procedure is not reported in sufficient detail to reconstruct which contrasts were randomized.

For clinical researchers, the record highlights the difference between a primary safety count and secondary activity outcomes. Disease-control rates may help describe preliminary activity, but cohort-level percentages with broad confidence intervals are imprecise, and unadjusted comparisons across several cohorts are vulnerable to chance findings. Their interpretation depends on prespecification, analysis populations, missing-data methods, and multiplicity handling, details that are not reported in the registry record.

The posted results therefore offer a useful example of separating description from inference. Counts and percentages summarize what was recorded; they do not automatically demonstrate dose effects or comparative treatment effects. Statistical interpretation requires attention to the study structure and the limits of the available reporting, rather than treating every numerical contrast as a causal finding.

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 registry record does not provide the full protocol, statistical analysis plan, randomization procedure, or dose-escalation decision rules. Although the study is labeled randomized, the posted cohort descriptions do not identify a distinct comparator treatment or explain how allocation was conducted. This limits interpretation of comparisons as randomized evidence.

The primary endpoint is confined to the first cycle and to the defined DLT criteria. The cohort counts are small and vary in size, and the record gives no confidence intervals or formal comparison for these counts. The dose, formulation, and fed or fasted conditions also differ across some Phase 1 cohorts, complicating comparisons between them.

The Phase 2 cohorts differ by tumor type or molecular feature and are not equivalent populations. The confidence intervals for disease control are broad, and the posted comparisons with Cohort A are unadjusted. No multiplicity procedure or prespecified status for those comparisons is reported. The record does not report detailed disease-control criteria, missing-data handling, or the analysis population for each secondary endpoint.

While the record lists serious adverse-event counts, term percentages, and deaths, details needed to interpret causality, timing, and the denominator for the term percentages are not reported in the registry record. Numerical results for the listed pharmacokinetic outcomes are also not included in the information summarized for these outcomes. These limits mean the posted data cannot serve as a complete account of all safety, pharmacokinetic, or clinical activity findings.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study of ASTX029 in Subjects With Advanced Solid Tumors — NCT03520075.
• Registry record last updated 2025-07-03; status last verified 2025-06.
• Results first posted 2025-07-03.
• 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

NCT03520075 is a completed, open-label Phase 1-2 study of ASTX029 in advanced solid tumors. The record describes a multicenter study with randomized allocation, no masking, a parallel model, and enrollment of 192. Phase 1 included dose-escalation and expansion cohorts; Phase 2 included six cohorts defined by tumor type or molecular features.

The primary endpoint was the number of participants with dose-limiting toxicities during Cycle 1, defined as 21 days. One participant was reported with a DLT in Phase 1A Cohort 5 and one in Cohort 12; the other listed Phase 1 cohorts had no recorded DLTs. Phase 2 disease-control rates and selected unadjusted comparisons with Cohort A were also posted. The registry does not report a detailed statistical analysis plan, randomization procedures, or multiplicity methods, so the cohort results are primarily descriptive.