Phase 1/2 Study of APR-246 in Combination With Pembrolizumab in Subjects With Solid Tumor Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 1/2 Study of APR-246 in Combination With Pembrolizumab in Subjects With Solid Tumor Malignancies (NCT04383938), a non-randomized phase 1, phase 2 study. The primary outcome is To Evaluate the Safety of APR-246 in Combination With Pembrolizumab in Subjects With Solid Tumors..

2. Clinical Question

3. Trial Design

This was a phase 1/2, open-label study with non-randomized allocation and a parallel model. The registry lists the study as completed. Enrollment was 40. The listed study start was 2020-06-25, and primary completion was 2022-04-30.

The study description says it included a safety lead-in followed by a phase 2 expansion in specific disease groups. The safety lead-in was for advanced non-CNS primary tumors. The three expansion cohorts were for gastric cancer, bladder cancer, and non-small cell lung cancer. The registry names the arms as Safety Lead In, Expansion 1, Expansion 2, and Expansion 3, respectively.

All four arms are described as receiving APR-246 with pembrolizumab. The safety lead-in description gives APR-246 4.5g/d and pembrolizumab 200 mg IV on day 3 every 21 days. The expansion descriptions give APR-246 4.5 g/d on days 1-4, with pembrolizumab 200 mg IV on day 3 every 21 days. These descriptions identify the planned cohort regimens; the registry record does not report individual adherence or delivered exposure.

The open-label design means the registry does not describe masking of participants or investigators. Because allocation was non-randomized and there is no separate control arm, the cohorts are not a controlled comparison. Their sizes, diagnoses, and clinical circumstances differ, so cohort-to-cohort differences should be treated as descriptive rather than evidence that one regimen or disease group caused a difference in outcomes.

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

Analysis population for the primary outcome, as defined in the registry: Data is presented by cohort/arm. 3 out of the 40 overall patients did not receive treatment on study. 37 total patients were analyzed

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: Baseline characteristics are presented by cohort.

6. Primary Endpoint

The registry lists the primary outcome as To Evaluate the Safety of APR-246 in Combination With Pembrolizumab in Subjects With Solid Tumors., measured over Through study completion, approximately 1 year.

The primary endpoint concerns the frequency of treatment-emergent adverse events and serious adverse events related to APR-246 in combination with pembrolizumab. The registry specifies the outcome unit as participants and the parameter as a number. The timeframe is through study completion, approximately 1 year. The registry also says that outcome data are presented by cohort or arm.

Safety endpoints are used to describe harms observed during treatment and follow-up. Counting participants with an event provides a direct account of how many people in each cohort were recorded as affected; serious adverse events represent a subset of safety outcomes that the registry reports separately. These counts do not by themselves establish that the combination caused an event. Attribution, event timing, severity details, and the clinical circumstances of individual cases matter for interpretation, and those details are not fully reported in the registry record.

The primary endpoint is not an efficacy measure. Although the brief study description mentions preliminary efficacy, the listed primary outcome is safety. The registry record does not report a primary efficacy endpoint or an inferential estimate comparing the combination with a control treatment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a participant count describe?

The primary outcome is expressed as a number of participants with an event. In general, a count answers how many people in a defined group were recorded as experiencing the outcome. It is not the same as a count of all event episodes: one participant may have more than one event, while a participant-level measure may count that person once for a particular outcome. The registry describes the primary measure in participant terms, but the record does not provide all event-level details needed to reconstruct how repeated events were handled.

The cohort-specific figures should be read with their corresponding at-risk counts. For example, the serious adverse-event entries give both the number affected and the number at risk for each cohort. Keeping those paired values together prevents a raw count from being mistaken for a common risk across differently sized groups. No new percentage or pooled estimate is calculated from these entries.

Why does study design affect interpretation?

In general, random allocation helps balance measured and unmeasured characteristics across treatment groups, making a treatment contrast easier to interpret causally. This trial was non-randomized and has no like-for-like control arm. Its four arms are experimental cohorts receiving the combination in different disease settings or study portions. Comparisons across them can be affected by differences in patient characteristics, diagnoses, eligibility, follow-up, or other factors. The registry record does not report a design that controls for those differences.

Open-label means treatment assignment was not masked as described in the registry. In general, knowledge of treatment can affect reporting, assessment, or management of events. That possibility is relevant when reading safety outcomes, especially when outcomes depend on clinical assessment. It does not establish that reporting was biased in this study; the registry record does not report an assessment of such bias.

What do uncertainty intervals and significance tests add?

In general, a confidence interval expresses the statistical uncertainty around an estimate under a stated analysis model. A wider interval indicates less precision, while a narrower interval indicates greater precision. A significance test evaluates compatibility with a specified null hypothesis under its assumptions. Neither one, by itself, establishes clinical importance or causation. The registry record does not report confidence intervals, p-values, or an inferential analysis for the listed safety counts.

The absence of an inferential analysis is particularly important for cohort comparisons. A numerical difference between two descriptive counts is not automatically a statistically reliable difference, and a statistical difference would not by itself show that treatment caused it. No treatment-effect comparison should be inferred from the cohort summaries.

What assumptions underlie the planning illustration?

The registry’s sample-size assumptions describe a planning illustration, not an observed-power calculation. In general, planning calculations ask how much information a future study might need to detect a specified effect under chosen assumptions. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. For non-time-to-event outcomes, the illustration is expressed as participants per arm for 80% power or participants per arm for 90% power; time-to-event illustrations use a total number of events. These are planning scenarios and are not observed trial results.

The record does not report the trial’s own sample-size justification, nor does it report a statistical analysis plan. No independent calculation can fill those gaps. In general, a sample-size calculation depends on the target effect, outcome variability or event frequency, significance level, power target, allocation, and analysis method. Without the trial’s documented planning rationale and complete analysis assumptions, the registry’s design rationale cannot be reconstructed from the posted cohort counts.

Why are safety counts not a causal estimate?

In general, a causal comparison requires a defensible counterfactual: an estimate of what would have happened to similar participants under a different treatment or no treatment. The listed cohorts do not supply that comparison. The safety counts can describe events recorded during the study, but they cannot isolate the contribution of APR-246, pembrolizumab, the underlying conditions, or other clinical factors. Treatment-relatedness is part of the stated primary question, but the record does not provide enough case-level detail to independently assess attribution.

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("Experimental: Safety Lead In", "Experimental: Expansion 1", "Experimental: Expansion 2", "Experimental: Expansion 3")
n      <- c(6, 6, 5, 20)
events <- c(6, 6, 5, 20)
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  Experimental: Safety Lead In  6  6  100.0  54.1  100.0  61.0  100.0  59.0  99.6
ARM  Experimental: Expansion 1  6  6  100.0  54.1  100.0  61.0  100.0  59.0  99.6
ARM  Experimental: Expansion 2  5  5  100.0  47.8  100.0  56.6  100.0  54.1  99.6
ARM  Experimental: Expansion 3  20  20  100.0  83.2  100.0  83.9  100.0  83.9  99.9

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists two outcome measures and does not report statistical analyses for them. It does not describe a multiplicity strategy, such as a hierarchy of tests or an adjustment for testing multiple outcomes. In general, testing several outcomes or subgroups can increase the chance of a false-positive finding if each is assessed separately. Since inferential results are not reported, there is no basis for describing how multiplicity was handled in an analysis.

The record states that 3 out of the 40 overall patients did not receive treatment on study and that 37 total patients were analyzed. It does not explain why those patients did not receive treatment, whether they contributed any outcome data, or how missing observations were handled. It also does not report a detailed analysis-population definition, such as the rules determining inclusion in each safety outcome. These details are important because the population used as a denominator can affect the interpretation of event counts.

In general, missingness can be informative: people with incomplete follow-up may differ from those with complete follow-up, and simply omitting them can change the apparent event frequency. The record does not describe follow-up completeness, censoring, imputation, or sensitivity analyses for missing safety information. It also does not report how participants who stopped treatment or left the study were handled in the endpoint analysis.

Accordingly, the posted counts can be described as registry-reported cohort summaries, but their handling of incomplete or unavailable data cannot be independently evaluated. The statistical analysis plan and further missing-data details 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: pain (20.0%), abdominal pain (20.0%), abdominal pain lower (20.0%), nausea (20.0%), hypotension (20.0%).

The serious adverse-event counts are listed as participants affected and participants at risk for each cohort: Safety Lead In, 1 and 6; Expansion 1, 4 and 6; Expansion 2, 3 and 5; Expansion 3, 7 and 20. Read each pair within its cohort. The groups differ in size, and the study was non-randomized, so raw counts are not a fair basis for ranking cohorts or attributing a difference to treatment.

The registry separately lists other adverse events and deaths. Those categories should not be merged with serious adverse events or interpreted as treatment-related without supporting information. The record does not specify whether the listed deaths were caused by treatment. It also lists Pain, Abdominal Pain, Abdominal pain lower, Nausea, Hypotension, and Pelvic Pain at 20.0% each among serious adverse-event terms, but does not state the denominator or clarify the event-counting convention for these percentages.

These are descriptive registry entries. The record does not report event-level timing, duration, recurrence, severity detail, or uncertainty intervals for the cohort counts. They can describe what was recorded, but they do not establish causation or quantify the precision of a comparative safety effect.

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

The registry describes a completed, open-label phase 1/2 study of APR-246 with pembrolizumab across a safety lead-in and disease-specific expansion cohorts. Its primary endpoint was safety, specifically the frequency of treatment-emergent adverse events and serious adverse events related to the combination. The posted safety counts provide a descriptive account of events recorded by cohort, alongside counts for other adverse events and deaths.

These data do not establish that the combination caused any particular event. The cohort design is non-randomized, there is no control group, and the patient groups represent different disease settings. The record does not provide adjusted comparisons or the detailed clinical information needed to separate treatment-related events from events due to illness or other factors. The death counts are not identified as treatment-related in the posted outcome listing.

The safety summaries also do not answer whether the combination is effective. Although the study description mentions preliminary efficacy, the primary outcome listed is safety, and the registry record does not report an efficacy comparison or inferential analysis. The absence of a control group would also limit causal interpretation of any uncontrolled outcome.

Precision is not quantified with confidence intervals or other uncertainty estimates. The recorded counts should therefore be understood as descriptive observations from the registry, not as definitive estimates of comparative risk. The record further notes that 3 of the 40 overall patients did not receive treatment on study, while 37 total patients were analyzed; reasons and missing-data handling are not reported.

In general, these limitations do not make descriptive safety information meaningless. They define what kind of conclusion it supports: an account of reported events in the studied cohorts, rather than a causal estimate, a comparison against another treatment, or a conclusion about efficacy.

14. Why This Trial Matters Statistically

This trial is statistically notable for the distinction between a safety-focused early-phase study and a controlled treatment comparison. Its phase 1/2 structure includes a safety lead-in followed by disease-specific expansion cohorts, but the non-randomized allocation means the arms do not serve as equivalent comparison groups. That design distinction is essential when reading cohort-level event counts.

The registry also makes the analysis population issue visible: 3 of the 40 overall patients did not receive treatment on study, and 37 total patients were analyzed. Without reasons for non-treatment or explicit endpoint-specific population rules, readers cannot fully reconstruct the analysis denominator or evaluate the impact of missing information.

Another statistical lesson is that a list of counts is not a complete uncertainty assessment. The registry reports event counts by cohort but does not report confidence intervals, p-values, or adjusted comparisons. In general, such estimates help readers judge precision and distinguish descriptive variation from evidence of a reliable group difference. Their absence means interpretation should stay with the descriptive summaries.

The record’s planning assumptions are also conceptually separate from observed results. Planning illustrations depend on assumptions about allocation, significance level, power, and effect size; they do not demonstrate that the completed study had a particular power or that a treatment effect was established. Distinguishing design calculations from observed evidence helps prevent a planning scenario from being mistaken for a trial finding.

Overall, the trial highlights why endpoint definition, analysis population, design, and uncertainty all matter alongside event counts. These features determine the scope of what the registry data can support.

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 gives limited detail about the safety analysis. It does not report the statistical analysis plan, endpoint-specific inclusion rules, confidence intervals, or inferential comparisons. The stated primary endpoint includes events related to the combination, but the posted counts do not provide case-level information for independently evaluating relatedness.

The four cohorts are non-randomized and differ by study stage and tumor group. They therefore cannot be treated as interchangeable arms. Without a randomized control group, observed events cannot be separated from the effects of the underlying disease, other care, or participant characteristics. The registry does not report adjusted analyses to address these differences.

The record notes that 3 of the 40 overall patients did not receive treatment on study and that 37 total patients were analyzed, but it does not report why treatment was not received, how much follow-up was available, or how missing outcome information was handled. The at-risk counts for individual safety outcomes are listed, but the registry does not fully explain their relationship to the analysis population.

The adverse-event term list gives percentages without stating the denominator or clarifying the counting convention. The death counts are not accompanied by attribution or cause-of-death details. Event timing, duration, recurrence, and severity information are also not sufficiently reported in the posted summaries to support a detailed safety assessment.

Finally, the study description mentions preliminary efficacy, but the primary outcome listed is safety, and no efficacy analysis is reported in the registry record. These limitations narrow interpretation to descriptive safety information and prevent stronger causal or comparative conclusions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Phase 1/2 Study of APR-246 in Combination With Pembrolizumab in Subjects With Solid Tumor Malignancies — NCT04383938.
• Registry record last updated 2025-05-13; status last verified 2025-05.
• Results first posted 2025-05-13.
• 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

NCT04383938 is a completed, open-label, non-randomized phase 1/2 study of APR-246, also called eprenetapopt, in combination with pembrolizumab for subjects with solid tumor malignancies. Enrollment was 40. The study included a safety lead-in and expansion cohorts for gastric cancer, bladder cancer, and non-small cell lung cancer.

The primary endpoint assessed treatment-emergent adverse events and serious adverse events related to the combination through study completion, approximately 1 year. The registry reports safety counts by cohort and states that 3 of the 40 overall patients did not receive treatment on study, while 37 total patients were analyzed.

The design has no randomized control group, and the registry record does not report inferential comparisons, uncertainty intervals, a statistical analysis plan, or detailed missing-data methods. The posted counts are descriptive and do not establish causation or comparative efficacy.