A Trial to Find and Investigate a Safe Dose of a New Substance (BI 754091) for Patients With Solid Tumours: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Trial to Find and Investigate a Safe Dose of a New Substance (BI 754091) for Patients With Solid Tumours (NCT02952248), a non-randomized phase 1 study. The primary outcome is Phase Ia Dose Escalation: Number of Participants With Dose-limiting Toxicities (DLTs) in the First Cycle (3 Weeks).

2. Clinical Question

The trial also set out to assess safety and tolerability, pharmacokinetics, and preliminary antitumour activity. In dose expansion, the objectives included further assessment of safety, efficacy, pharmacokinetics and biomarkers at the recommended Phase 2 dose. These stated objectives describe what investigators planned to evaluate; they do not establish what the results were. The registry lists secondary outcomes, including confirmed objective response, pharmacokinetic measures, progression-free survival and adverse events, but does not provide outcome results for those measures in the record described.

3. Trial Design

This was a completed Phase I trial with non-randomized allocation, no masking and a sequential model. The registry describes an open-label study intended to investigate safety, pharmacokinetics and efficacy of BI 754091 in patients with advanced solid tumours. Enrollment was 110. The recorded start date was 2016-11-21, and primary completion was 2021-04-30.

The design included Phase Ia dose escalation and Phase Ib dose expansion, with an additional Phase Ib dose-escalation cohort. Phase Ia escalation included low, medium and high dose cohorts, identified in the safety results as 80 mg, 240 mg and 400 mg ezabenlimab. Expansion cohorts included patients with seven tumour types at 240 mg, patients with high tumour mutational burden excluding MSI-high tumours, and patients with refractory squamous cell cervical, anal and skin tumours. A further cohort concerned recurrent HPV-positive or HPV-negative vaginal or vulvar squamous cell carcinoma.

The design is not a randomized comparison of treatment groups. Dose and cohort differences therefore cannot be interpreted as if participants had been assigned to comparable groups under randomization. The cohorts differ in dose, tumour eligibility and clinical context. The registry record does not report a protocol-level explanation of the allocation process beyond describing the trial as non-randomized and sequential.

The registry names Boehringer Ingelheim as sponsor and lists the condition as neoplasms. The study is marked completed. Completion describes the study status; it does not mean that all planned outcomes or statistical analyses are reported in the registry record. The record does not report the rationale for the enrollment target, details of any interim analyses, or a complete 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.

Before assignment: All subjects were screened for eligibility prior to participation in the trial. Subjects attended a specialist site which ensured that they (the subjects) strictly met all inclusion and none of the exclusion criteria. Subjects were not to be allocated to a treatment group if any of the entry criteria were violated.

Only the first 6 of 7 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Treated Set (TS): The TS included all participants who received at least 1 dose of ezabenlimab. Phase Ia dose escalation.

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: Treated Set (TS): The TS included all participants who received at least 1 dose of ezabenlimab.

Only the first 6 of 8 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Phase Ia Dose Escalation: Number of Participants With Dose-limiting Toxicities (DLTs) in the First Cycle (3 Weeks), measured over Up to 3 weeks..

The primary endpoint counts participants with dose-limiting toxicities during the first cycle, defined as 3 weeks, in Phase Ia dose escalation. The registry describes these toxicities as graded according to CTCAE Version 4.03 / 5.0. The endpoint population is the Treated Set, which includes participants who received at least 1 dose of ezabenlimab in Phase Ia.

A dose-limiting toxicity endpoint is used in dose escalation to describe clinically important toxicity observed during a defined early treatment period. In general, such information can contribute to assessment of tolerability at the doses studied and to decisions about dose selection for further investigation. It does not by itself summarize every aspect of safety: adverse events outside the endpoint definition or observation period may be relevant to a broader safety assessment.

The registry states that this endpoint was intended to support assessment of the maximum tolerated dose. The trial objectives also refer to determining a recommended Phase 2 dose. Those are objectives, not reported conclusions. The number of participants with dose-limiting toxicities is not reported in the registry record. Consequently, the endpoint’s observed frequency and its contribution to dose selection cannot be evaluated from the posted primary result information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Interpretation depends on separating the trial’s design, its planned measures and its reported results. The design was open-label, non-randomized and sequential. The primary outcome was defined as a participant count, rather than a continuous measurement or time-to-event estimate. The registry record lists the endpoint and its observation window, but does not report the count observed or a statistical analysis for it.

What does a count endpoint tell us?

In general, a count of participants with an event is a direct description of how many people met an outcome definition in a specified population and period. It is not automatically a rate, a risk difference, or a comparison between treatments. To interpret a count fully, readers need the relevant analysis population, the event definition, the observation period and the number assessed. The record defines the Treated Set as participants who received at least 1 dose of ezabenlimab in Phase Ia, and defines the primary observation period as the first cycle, 3 weeks. The observed primary count is not reported in the registry record.

Why does the study design matter?

In general, random allocation helps make treatment groups comparable with respect to measured and unmeasured characteristics, subject to chance variation. Here, allocation is described as non-randomized, and there is no masking. Cohorts are defined by dose and by tumour eligibility. Differences between their observed safety counts can therefore reflect differences in patient mix, disease context, exposure or other factors, as well as any potential dose-related pattern. A descriptive contrast between cohorts does not isolate a causal treatment effect. The registry itself flags that the arms are not a like-for-like control comparison.

The sequential model and dose-escalation structure are relevant to the purpose of the study. In general, dose escalation gathers safety information as investigation proceeds across dose levels. It is not equivalent to random assignment across fixed treatment groups. The registry record does not report enough detail to reconstruct escalation rules, cohort decision thresholds or the conduct of any interim review. Those details should not be inferred from the cohort labels alone.

How should statistical uncertainty be understood?

In general, an estimate describes the observed data, while a confidence interval conveys the range of values compatible with those data under a specified statistical model and procedure. A wide interval indicates limited precision; an interval does not remove the design limitations or establish causality. No confidence interval for the primary endpoint is reported in the registry record. Because the primary result itself is not reported, there is no posted primary estimate whose precision can be assessed.

Small cohorts can produce imprecise descriptions because each event can have substantial influence on the observed count. This is a general statistical consideration, not a claim about an unreported estimate in this trial. The safety cohort figures include small Phase Ia groups and larger expansion groups, but the groups also differ in eligibility and study phase. Numerical comparisons across them should not be treated as controlled dose comparisons.

What do the posted analysis details establish?

The registry lists zero analyses. That means the record does not contain a posted analysis entry for the outcomes. It does not establish whether analyses were or were not conducted outside the registry; such information is not reported in the registry record. The registry also does not report a complete statistical analysis plan, model specifications, handling of missing outcome data, or procedures for adjusting for multiple comparisons.

In general, multiplicity arises when many outcomes, groups or hypotheses are examined. Without a prespecified hierarchy or adjustment strategy, the chance of obtaining apparently notable findings by chance can increase across a collection of tests. The listed secondary outcomes cover several domains, but no inferential results or multiplicity procedure are reported. Therefore, there is no basis in this record for assessing whether any secondary finding was selected from a larger set or how statistical error was controlled.

The record describes a sample-size planning illustration based on an observed effect, with a two-sided significance level of 0.05, equal allocation and a normal approximation. Such planning quantities are not observed trial outcomes and are not observed power. The registry record does not report the trial’s own sample-size justification. A planning illustration cannot fill in the missing primary result or establish how the trial was designed to make its primary conclusion.

Finally, safety counts and efficacy outcomes answer different questions. General safety summaries describe adverse events recorded in specified groups; they do not alone determine whether an event was caused by treatment. Efficacy outcomes such as objective response or progression-free survival require their own definitions, follow-up and analysis. Although these outcomes are listed, their result estimates are not reported in the registry record. Keeping these distinctions explicit prevents a design description or a safety count from being mistaken for a complete statistical result.

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 Ia Dose Escalation: 80 mg Ezabenlimab", "Phase Ia Dose Escalation: 240 mg Ezabenlimab", "Phase Ia Dose Escalation: 400 mg Ezabenlimab")
n      <- c(3, 3, 3)
events <- c(0, 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 Ia Dose Escalation: 80 mg Ezabenlimab  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase Ia Dose Escalation: 240 mg Ezabenlimab  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase Ia Dose Escalation: 400 mg Ezabenlimab  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes alongside the primary dose-limiting toxicity endpoint. These include confirmed objective response, maximum measured concentration, area under the concentration-time curve, dose-limiting toxicities through end of treatment, progression-free survival, adverse events, serious adverse events and laboratory abnormalities. The registry record does not report a multiplicity strategy, such as a testing hierarchy or adjustment procedure. It also does not report statistical analysis results for these outcomes.

In general, testing or examining multiple outcomes can make isolated findings difficult to interpret unless the analysis plan explains how the family of questions is handled. This concern is especially relevant when outcome domains, cohorts and time frames differ. Without a reported plan and results, it is not possible to assess which findings were confirmatory, how error rates were controlled, or whether any analyses were exploratory.

Missing-data methods are not reported in the registry record. The primary population is defined as the Treated Set, including participants who received at least 1 dose of ezabenlimab in Phase Ia. The record does not describe how missing endpoint assessments, early discontinuations, or incomplete follow-up were handled. It also does not describe analysis populations for the listed expansion outcomes. These omissions limit assessment of who contributed to each result and whether missing observations could affect interpretation.

The registry lists zero analyses. This does not establish that no analyses were conducted; it means that analysis details are not reported in the registry record. The record also does not report a complete statistical analysis plan, interim-analysis details, or the trial’s sample-size justification. Those aspects cannot be reconstructed from the listed objectives and endpoint definitions.

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: malignant neoplasm progression (66.7%), small intestinal obstruction (33.3%), arterial injury (33.3%), fistula (33.3%), foot amputation (33.3%).

The safety record gives serious adverse event counts by cohort, alongside the number at risk. In Phase Ia, the 80 mg cohort reports 2 serious adverse event cases among 3 at risk; the 240 mg cohort reports 1 among 3; and the 400 mg cohort reports 2 among 3. In Phase Ib expansion, the seven-tumour cohort reports 10 among 30, the high-TMB cohort 9 among 27, the cervical/anal/skin cohort 16 among 31, and the VSCC cohort 6 among 13.

These counts describe the listed cohorts. They should not be read as a randomized comparison: allocation was non-randomized, and the expansion cohorts have different eligibility descriptions. The record also lists deaths by cohort: 1, 1 and 2 in the Phase Ia groups, and 16, 10, 17 and 9 in the Phase Ib groups in the same cohort order. These figures do not, by themselves, attribute deaths to treatment.

The listed serious adverse event terms include malignant neoplasm progression at 66.7%, and small intestinal obstruction, arterial injury, fistula and foot amputation at 33.3% each; pneumonia is listed at 9.7%. The registry record does not report the denominators or participant-level overlap for these term percentages in the registry record described. A percentage for a term should not be treated as a treatment-related risk without the relevant definitions and attribution information.

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

The record establishes that this was a completed, non-randomized Phase I investigation of BI 754091 in advanced solid tumours, with dose escalation and expansion. It establishes the stated purpose of assessing dose-limiting toxicities in the first cycle and describes additional safety, pharmacokinetic, efficacy and biomarker objectives. It also provides selected cohort-level safety counts.

The record does not establish the observed primary endpoint result. The number of participants with dose-limiting toxicities in Phase Ia is not reported in the registry record. It therefore does not show the primary endpoint’s observed frequency or document how that result informed maximum tolerated dose or recommended Phase 2 dose selection. The presence of endpoint and objective descriptions is not equivalent to reporting their results.

The available safety figures are descriptive and do not support causal conclusions. The groups were not randomized, and their dose levels and tumour eligibility differ. Counts across cohorts cannot be interpreted as treatment effects or direct comparisons. The listed deaths are counts, not evidence of cause. The record does not supply the causal attribution details needed to interpret them as treatment-related outcomes.

The registry also does not report outcome estimates for the listed secondary efficacy and pharmacokinetic measures, or a statistical analysis plan. There is no reported inferential result that would support claims about a treatment effect, comparative efficacy, or statistical significance. General statistical concepts such as confidence intervals and multiplicity remain relevant, but the missing estimates and analysis details prevent trial-specific evaluation of precision or error control.

Accordingly, the posted record is useful for understanding the study structure, the endpoint definitions and the scope of the safety summaries. It is not a complete account of the trial’s primary or secondary results. No conclusion about clinical benefit or lack of benefit follows from these omissions or from the descriptive safety counts alone.

14. Why This Trial Matters Statistically

This trial illustrates why early-phase study design and result reporting need to be read together. Dose escalation focuses on safety observations across dose levels, while expansion cohorts may focus on particular tumour types or characteristics. Those aims are reflected in the registry’s Phase Ia and Phase Ib descriptions. The resulting cohorts are not interchangeable: dose and eligibility differ, so their event counts do not form a controlled comparison.

The primary endpoint is specific in its event definition, population and time frame. It counts participants with dose-limiting toxicities in the first cycle, defined as 3 weeks, among the Phase Ia Treated Set. That specificity helps clarify what the trial set out to measure. But without the observed count, readers cannot assess the central endpoint from the registry record. This demonstrates that a well-described outcome definition is not a substitute for a posted result.

The registry also highlights the difference between a broad objective and an estimable result. Safety, pharmacokinetics, efficacy and biomarkers are stated aims, and several secondary outcomes are named. Yet no analyses are listed and the secondary outcome results are not reported in the record described. Statistical interpretation depends on the actual estimates, their uncertainty and the analysis population, not only on the list of intended measures.

For biostatisticians and clinical researchers, the record is a reminder to distinguish descriptive safety summaries from causal comparisons, and planned assessments from completed analyses. It also underscores why reporting analysis methods, missing-data handling and multiplicity procedures matters when a study includes several outcomes and clinically distinct cohorts. Those details are not reported in the registry record, so conclusions must remain bounded by the information that is available.

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 statistical interpretation is that the registry record does not report the primary endpoint result: the number of participants with dose-limiting toxicities during the first cycle in Phase Ia. The record lists secondary outcomes but does not report their results. It therefore does not permit evaluation of the primary safety question or the stated efficacy and pharmacokinetic objectives.

The design was non-randomized, open-label and sequential. Cohorts differed by dose, study phase and tumour eligibility, and there was no like-for-like control comparison. Cohort-level safety figures are therefore descriptive and cannot isolate the effect of dose or treatment from differences in participant and disease characteristics. The record does not provide a causal attribution for the reported deaths or serious adverse events.

Analysis methods are limited in the registry record. It lists zero analyses and does not report a complete statistical analysis plan, interim-analysis details, missing-data procedures, multiplicity handling, or the trial’s sample-size justification. The record defines the primary Treated Set, but does not report analysis populations for each secondary outcome or explain how incomplete assessments were handled.

Safety term percentages are reported, but the corresponding denominators and participant-level overlap are not reported in the registry record. The registry’s enrollment figure does not, by itself, establish the analysis population for any outcome. These limits make it important not to infer unreported results, precision, causality or comparative effects from the design description and safety summaries alone.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Trial to Find and Investigate a Safe Dose of a New Substance (BI 754091) for Patients With Solid Tumours — NCT02952248.
• Registry record last updated 2026-01-27; status last verified 2025-12.
• Results first posted 2026-01-27.
• 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

NCT02952248 is a completed, open-label, non-randomized Phase I trial sponsored by Boehringer Ingelheim. It enrolled 110 patients with advanced solid tumours and included sequential dose-escalation and dose-expansion cohorts. The recorded start date was 2016-11-21, with primary completion on 2021-04-30.

The primary endpoint was the number of participants with dose-limiting toxicities during the first cycle, defined as 3 weeks, in the Phase Ia Treated Set. The registry record does not report that outcome count. It lists secondary safety, pharmacokinetic and efficacy outcomes, but does not report their results or statistical analyses. Cohort-level serious adverse event and death counts are available, but the non-randomized design does not support treating cohort differences as controlled comparisons.