A Study of Gimistotug (BGB-A445) in Combination With Other Investigational Agents in Participants With Non-Small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Gimistotug (BGB-A445) in Combination With Other Investigational Agents in Participants With Non-Small Cell Lung Cancer (NCT06029127), a randomized phase 2 study. The primary outcome is Overall Response Rate.

2. Clinical Question

The study’s stated objective was to evaluate anti-tumor activity of gimistotug combined with investigational agents in participants with non-small cell lung cancer. It was a Phase 2, randomized, open-label, multi-arm study. Both combinations contained gimistotug, so the comparison concerns the two treatment combinations rather than gimistotug versus no gimistotug.

Participants were to continue treatment until disease progression, intolerable toxicity, withdrawal of informed consent, or another discontinuation criterion. The primary response assessment was scheduled every 6 weeks for the first 9 months and every 12 weeks thereafter. The maximum follow-up for this endpoint was up to 13.5 months. Those details define the setting and observation period for the response outcome; they do not establish whether each participant completed every scheduled assessment.

The registry record does not report a formal hypothesis for the comparison between combinations or identify one combination as a control for a confirmatory test. The primary endpoint is reported separately by arm, with confidence intervals. Accordingly, the result is best read as a descriptive account of confirmed responses within these cohorts, with a comparison between the posted estimates interpreted in light of their small sample sizes and uncertainty.

3. Trial Design

This was a Phase 2, randomized, open-label, parallel-group study sponsored by BeiGene. The registry record lists non-small cell lung cancer as the condition and describes participants as having been previously treated with an anti-PD-(L)1 antibody. Enrollment was 35. The study began on 2023-10-23, had a primary completion date of 2024-12-31, and is listed as completed.

The two treatment cohorts were gimistotug plus docetaxel and gimistotug plus BGB-15025. In the docetaxel cohort, both drugs were administered by intravenous infusion once every 3 weeks. In the BGB-15025 cohort, gimistotug was administered intravenously once every 3 weeks and BGB-15025 orally once daily during each 3-week cycle. Treatment continued until disease progression, intolerable toxicity, withdrawal of informed consent, or another discontinuation criterion.

Randomization means treatment assignment was described as randomized, but the allocation ratio, sequence-generation method, and concealment procedure are not reported in the registry record. The study was open-label, with no masking. Thus, participants and investigators were not described as blinded to the assigned treatment. This matters when interpreting investigator-assessed response and adverse-event reporting, though it does not by itself establish the direction or size of any bias.

The primary endpoint was overall response rate (ORR), defined using confirmed complete or partial response assessed by investigators according to RECIST v1.1. Response was assessed every 6 weeks for the first 9 months and every 12 weeks thereafter; maximum follow-up was up to 13.5 months. The posted primary endpoint groups contain 21 participants in the gimistotug plus docetaxel cohort and 14 in the gimistotug plus BGB-15025 cohort. These are the analysis counts shown for this outcome, not a claim that every enrolled participant had complete follow-up.

Secondary outcomes included disease control rate, clinical benefit rate, duration of response, treatment-emergent and serious adverse events, drug concentrations, and anti-drug antibodies. The registry record does not report the allocation ratio, a sample-size justification for the trial, or a protocol-based statistical analysis plan.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Intent-to-Treat population included all participants who were enrolled or randomized to a treatment cohort.

5. Baseline Characteristics

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

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate, measured over Response was assessed every 6 weeks for the first 9 months and every 12 weeks thereafter; maximum time on follow-up was up to 13.5 months.

Overall response rate is the percentage of participants whose best overall response was a confirmed complete response or partial response. The registry record defines both categories according to investigator assessment using RECIST v1.1. A complete response requires disappearance of target and non-target lesions, with lymph nodes meeting the stated size criterion, no new lesions, and normalization of tumor marker level where applicable. A partial response requires at least a 30% decrease in the sum of target-lesion diameters from baseline, no new lesions, and the stated conditions for non-target lesions and tumor markers.

ORR is used as a tumor-activity endpoint because it counts participants who meet a specified threshold for tumor shrinkage, rather than combining response with stable disease. It is a binary outcome at the participant level: a person either meets the confirmed CR or PR definition or does not. The reported percentage is therefore directly tied to the response criteria and assessment schedule, rather than representing the amount of shrinkage for every participant.

The posted result is 0.0% in each treatment cohort. That means no confirmed complete or partial responses were recorded in either cohort’s primary endpoint analysis. It does not mean that no participant experienced any tumor change or disease control: those concepts are captured separately in the registry’s secondary outcomes. The confidence intervals indicate how uncertain the underlying response percentages remain with these cohort sizes.

7. Results Posted in the Registry

• Gimistotug + BGB-15025: 0.0% (0/14) versus 0.0% (0/21) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0); unadjusted two-proportion z-test p = n/a.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Disease Control Rate (DCR)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Gimistotug + BGB-15025: 21.4% (3/14) versus 71.4% (15/21) on control; risk difference -50.0 percentage points (95% CI -78.9 to -21.1), risk ratio 0.30 (0.11–0.85); unadjusted two-proportion z-test p = 0.004.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Gimistotug + BGB-15025 than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Clinical Benefit Rate (CBR)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Gimistotug + BGB-15025: 0.0% (0/14) versus 19.0% (4/21) on control; risk difference -19.0 percentage points (95% CI -35.8 to -2.3); unadjusted two-proportion z-test p = 0.083.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Gimistotug + BGB-15025 and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants With Anti-Drug Antibodies to Gimistotug

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Gimistotug + BGB-15025: 15.4% (2/13) versus 11.1% (2/18) on control; risk difference 4.3 percentage points (95% CI -20.1 to 28.7), risk ratio 1.38 (0.22–8.59); unadjusted two-proportion z-test p = 0.726.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Gimistotug + BGB-15025 and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes add context to the primary response result by capturing other aspects of disease status, treatment safety, pharmacology, and immunogenicity. They do not all measure the same construct as ORR. In particular, disease control and clinical benefit are distinct endpoints from confirmed tumor shrinkage, so their patterns need not match the primary endpoint. Pharmacokinetic measures and anti-drug antibodies address exposure and immune response rather than antitumor response directly.

Across the posted secondary comparisons, the estimates vary in direction and precision, and some comparisons are based on very few events. This makes the overall pattern more informative as exploratory description than as a unified test of the two combinations. A secondary endpoint can appear different between cohorts even when the primary endpoint has no observed responses in either; that is not inherently contradictory, because the endpoints define success differently.

The unadjusted p-values reported for selected secondary comparisons do not account for the fact that several endpoints were assessed. When many outcomes are examined, the chance of at least one apparently notable result can increase. The registry record does not report a multiplicity adjustment or a hierarchy specifying which secondary outcomes should be considered confirmatory. Therefore, the unadjusted results carry less evidential weight than a prespecified, multiplicity-controlled analysis would. They should be interpreted with their confidence intervals, event counts, and endpoint definitions rather than as stand-alone declarations of a treatment difference.

The record lists seven secondary outcomes, including adverse events, duration of response, disease control, clinical benefit, drug concentrations, and anti-drug antibodies. It also lists eight outcome measures overall, but reports no sponsor analyses for the secondary endpoints. The absence of reported analyses limits what can be concluded about adjusted comparisons or model-based estimates. The secondary results provide useful descriptive context, but they do not replace the primary endpoint or resolve the uncertainty in the ORR estimates.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Participants With Treatment Emergent Adverse Events (TEAEs) and Serious Adverse Events (SAEs)
• Duration of Response (DOR)
• Disease Control Rate (DCR)
• Clinical Benefit Rate (CBR)
• Serum Concentrations of Gimistotug
• Plasma Concentrations of BGB-15025
• Number of Participants With Anti-Drug Antibodies to Gimistotug

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant either had a confirmed complete or partial response or did not. The registry reports the proportion meeting that definition and a 95% confidence interval for each arm. The posted values are 0.0% for each combination. With no observed response events, the estimate is at the lower boundary of the possible proportion, but the confidence interval still extends upward. A zero observed proportion is not the same as certainty that the population proportion is zero.

What does the confidence interval express?

A confidence interval is a way to communicate sampling uncertainty around an estimate. In general, a 95% confidence interval is produced by a procedure that, under its assumptions, would cover the underlying parameter in 95% of repeated samples. It is not a guarantee that the true value lies within this particular interval, nor is it a range containing 95% of individual participant outcomes. For these response data, the interval is especially useful because the point estimate alone cannot convey how much uncertainty remains when the observed count is zero.

What is the arm comparison?

The comparison is expressed as a risk difference: the proportion in the gimistotug plus BGB-15025 cohort minus the proportion in the gimistotug plus docetaxel cohort. The posted estimate for the primary endpoint is 0.0, with a confidence interval from 0.0 to 0.0. This is an unadjusted recalculation from the posted arm summaries, not a reported sponsor analysis. Because both observed proportions are zero, the estimate reflects equality in these observed data. It should not be mistaken for a precise estimate of equality in the underlying populations, particularly when the component arm estimates have intervals extending above zero.

For the secondary disease control rate and clinical benefit rate, the posted comparisons also use risk differences and include unadjusted p-values. A p-value measures how compatible the observed data are with a specified null hypothesis under the statistical model; it is not the probability that the null hypothesis is true, and it does not measure clinical importance. The p-values are unadjusted for the several secondary endpoints. The registry record does not report a multiplicity procedure, so they are best treated as exploratory evidence rather than definitive tests.

Other comparisons include relative measures for selected secondary outcomes. A risk ratio compares the proportions between arms, while an odds ratio compares the odds of an event. These measures can be unstable when event counts are small or absent; a confidence interval that is wide signals limited precision. They do not remove the need to consider the arm counts, endpoint definitions, and study design. No model adjustment is reported for these recalculations.

Why do small event counts matter?

With small cohorts and uncommon outcomes, each event can have a large influence on the estimated proportion. When no events are observed, or when only a few occur, conventional large-sample approximations may be less reliable than they are with abundant data. Confidence intervals are therefore important alongside the point estimate. A narrow interval should not be inferred simply because two point estimates match; the arm-specific intervals show the uncertainty more clearly.

For secondary outcomes, the registry also lists planning illustrations based on observed effects for selected comparisons. These are not observed power and are not the trial’s original sample-size justification. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The table gives participants per arm for the listed power targets. Such calculations describe hypothetical planning under those assumptions; they do not show that the present trial had adequate power, nor do they establish that the observed effect is stable enough to plan a future study around without further evidence.

The registry record reports no sponsor analyses for the secondary endpoints. The listed comparisons are recalculated from the posted summary statistics. The record’s discrepancy indicator is false, so there is no flagged conflict between those recalculations and a sponsor analysis. Still, a recalculation from summary data is not a substitute for an individual-participant analysis or a full prespecified model. Details such as covariate adjustment, handling of participants without an evaluable response, and analysis populations beyond the stated endpoint population are not reported in the registry record.

Finally, the clinical endpoint is investigator-assessed in an open-label study. In general, lack of masking can influence assessment or reporting, particularly when a judgment is required. RECIST v1.1 provides structured criteria, but the registry record does not report independent central review. It is therefore appropriate to describe the posted outcome as investigator-assessed, rather than assume an independent blinded review.

11. Confidence Intervals and Precision

The primary response estimates are 0.0% in both cohorts, with 95% confidence intervals of 0.0% to 16.1% for gimistotug plus docetaxel and 0.0% to 23.2% for gimistotug plus BGB-15025. These intervals communicate that the data are compatible with a range of nonzero underlying response rates, despite no recorded confirmed responses. The point estimates are exact descriptions of the observed proportions, but they are not precise estimates of the population response rates.

The intervals are not symmetrical around the point estimates. This is expected for a proportion near its lower boundary: a proportion cannot fall below zero, but it can extend upward. The interval for the cohort with 14 participants has a higher upper endpoint than the interval for the cohort with 21 participants. This is consistent with less precision when there are fewer observations, but it does not imply that the underlying response probability differs between cohorts.

The risk-difference interval for the primary comparison is reported as 0.0 to 0.0 because both observed event proportions are zero in the posted unadjusted comparison. That degenerate interval should be read cautiously alongside the separate arm-level confidence intervals. It reflects the lack of observed events in the summary-based comparison; it should not be interpreted as eliminating uncertainty about the underlying response rates or as evidence of equivalence.

Confidence intervals quantify sampling uncertainty under the method used to construct them. They do not account automatically for every source of uncertainty, such as endpoint assessment, missing outcome data, or choices in statistical analysis. Those details are not reported in the registry record. In general, interval overlap is not itself a formal test of a difference, and interval width is not a measure of clinical value. Here, the main precision message is that the observed response proportions are zero, while the limited cohort sizes leave meaningful uncertainty about what response rates might be in a broader population.

12. Multiplicity, Missing Data and Analysis Populations

The trial evaluates a primary endpoint alongside several secondary outcomes. The registry record does not report a plan for adjusting secondary comparisons for multiplicity, nor does it state a hierarchy or gatekeeping strategy for interpreting them. Without such a plan, unadjusted p-values should be treated as exploratory. A nominally small p-value among several comparisons can occur by chance, and the record does not supply a multiplicity-adjusted result to show how that concern was handled.

The primary endpoint population is described as the Intent-to-Treat population, including all participants enrolled or randomized to a treatment cohort. The endpoint arm counts are 21 for gimistotug plus docetaxel and 14 for gimistotug plus BGB-15025. Other outcomes can have different denominators: for example, the anti-drug antibody outcome reports 18 and 13 participants in its two cohorts. The record does not explain the reasons for these outcome-specific denominators or how participants without a measurement were handled. No missing-data imputation method is reported in the registry record.

The registry record also does not report whether there were interim analyses, how the sample size was justified for the original trial, or whether response assessment was independently reviewed. The stated maximum follow-up and scheduled assessment intervals describe the planned observation period, but do not establish that all participants had complete follow-up. Details on censoring, missing assessments, sensitivity analyses, and protocol deviations are not reported in the registry record.

The primary endpoint definition identifies the analysis population but does not, by itself, answer every question about implementation. For example, the record does not describe how an unconfirmed response, an unevaluable assessment, or an assessment after treatment discontinuation would be classified. These omissions do not change the posted counts, but they limit independent assessment of the analysis rules and their consistency with a prespecified plan.

As a result, the primary descriptive estimate is interpretable as the posted ORR for the stated population, while inferential claims across the wider outcome set are constrained by unreported analysis details. The lack of reported multiplicity handling and missing-data methods is particularly important when reading secondary comparisons as evidence rather than as exploratory findings.

13. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (7.1%), pneumonia bacterial (7.1%), anaemia (4.8%), febrile neutropenia (4.8%), leukopenia (4.8%).

The posted serious adverse event counts are 8 participants in the gimistotug plus docetaxel cohort and 1 participant in the gimistotug plus BGB-15025 cohort. The adverse-event record also lists other adverse events affecting 20 and 12 participants, respectively. These are participant counts within the stated safety groups; they are not measures of severity beyond the categories named, and the categories may overlap. They should not be added together to create a total.

The posted death counts are 12 for gimistotug plus docetaxel and 7 for gimistotug plus BGB-15025. The registry record does not establish that these deaths were caused by study treatment, nor does it provide a causal attribution in the listed details. A count alone does not show timing, cause, or the relationship to treatment, and the registry record does not report those details.

Serious adverse event terms listed among the most frequent include pneumonia and bacterial pneumonia, each reported as 7.1%, as well as anaemia, febrile neutropenia, leukopenia, and myelosuppression, each reported as 4.8%. A listed term may occur in the same participant as another term, so term percentages should not be summed. These safety data describe events recorded in the study; comparisons are limited by cohort size and the absence of further event-level details in the registry record.

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

The primary result establishes a narrow descriptive fact: the registry reports no confirmed complete or partial responses in either treatment cohort during the primary endpoint assessment period. The two posted ORR estimates are 0.0%, and the confidence intervals show that the underlying response proportions remain uncertain. The result does not prove that the true response rate is zero, that the two combinations have identical effects, or that no participant experienced any other form of disease control.

Nor does a zero ORR result answer every question about these regimens. ORR focuses on confirmed tumor shrinkage meeting RECIST v1.1 criteria. It does not encompass every secondary outcome, including disease control, clinical benefit, duration of response, adverse events, drug concentrations, or anti-drug antibodies. Differences across those endpoints do not change the definition of the primary outcome; they provide separate information that should be considered with its own uncertainty and limitations.

The study compared two combinations that both included gimistotug. It therefore does not isolate the effect of gimistotug relative to no gimistotug, nor does it compare either combination directly with a placebo or a standard-care-only cohort in the registry record. The open-label design and investigator assessment are also relevant context. The RECIST criteria structure response assessment, but masking and independent review are not reported.

These results should not be used alone to make treatment decisions. They are trial findings for the enrolled population and the recorded follow-up, not individualized medical guidance or proof of benefit or lack of benefit in other settings. The registry record does not report a complete statistical analysis plan, missing-data procedures, or adjusted secondary analyses. Those gaps, together with the limited cohort sizes, make careful description more appropriate than broad claims about comparative effectiveness.

15. Why This Trial Matters Statistically

This trial illustrates why a point estimate should not be read without its uncertainty. When no response events are observed, the estimated ORR is at zero, but the confidence intervals still allow for nonzero underlying response rates. That distinction is central to interpreting small clinical cohorts: the observed data can be stated plainly while acknowledging that population-level precision is limited.

It also shows why endpoint definitions matter. ORR requires a confirmed complete or partial response under RECIST v1.1. Secondary measures of disease control or clinical benefit capture different outcome categories and can therefore present a different pattern. A result on one endpoint cannot be substituted for a result on another, even when both concern tumor status.

The two arms share gimistotug but differ in their partner agent. This design addresses a comparison between combinations, not the contribution of gimistotug itself relative to an untreated or non-gimistotug group. Randomization is a relevant strength for comparing assigned cohorts, while the open-label design, small arm sizes, and unreported analysis details limit the certainty of interpretation.

Finally, secondary comparisons demonstrate the importance of multiplicity and transparent planning. Unadjusted p-values across several outcomes can be useful signals for further study, but they are not equivalent to confirmatory evidence. Planning illustrations based on observed effects are similarly hypothetical and do not establish the original trial’s adequacy or future-study requirements. The statistical value of the record lies in separating what was observed from what remains uncertain.

16. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The primary outcome has no observed response events in either cohort, and the arm sizes are small. Consequently, the estimates are at the lower boundary while the confidence intervals still permit nonzero response proportions. The posted primary comparison interval does not convey the same uncertainty as the separate arm-level intervals, so the arm estimates and their intervals are important context for interpreting the zero risk difference.

The study was open-label and response was investigator-assessed. Although RECIST v1.1 supplies structured criteria, the registry record does not report independent central review or masking of assessors. The extent to which assessment or reporting could be influenced cannot be determined from the record.

The outcome analysis counts differ across endpoints. The registry record defines the primary population but does not explain all outcome-specific denominators, missing assessments, or missing-data handling. It also does not report the trial’s original sample-size justification, interim analyses, detailed allocation procedures, or a full statistical analysis plan. These omissions limit evaluation of the analysis choices and the degree to which the posted comparisons were prespecified.

Several secondary endpoints are listed, but multiplicity adjustment is not reported, and the secondary comparisons are unadjusted recalculations rather than sponsor analyses. Small event counts can make comparisons unstable, while an unadjusted p-value does not account for the number of outcomes examined. The registry record also does not provide enough detail to attribute the listed deaths to treatment or characterize their causes.

Finally, these findings apply to the recorded study population and follow-up. The registry record does not establish how directly they generalize to other patient groups, treatment settings, or longer observation periods.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: A Study of Gimistotug (BGB-A445) in Combination With Other Investigational Agents in Participants With Non-Small Cell Lung Cancer — NCT06029127.
• Registry record last updated 2026-01-05; status last verified 2025-12.
• Results first posted 2026-01-05.
• 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.

20. Record Summary

NCT06029127 is a completed, randomized, open-label Phase 2 study of gimistotug combined with docetaxel or BGB-15025 in participants with non-small cell lung cancer previously treated with an anti-PD-(L)1 antibody. Enrollment was 35. The primary endpoint was investigator-assessed ORR using RECIST v1.1, with responses assessed every 6 weeks for the first 9 months and every 12 weeks thereafter, and maximum follow-up up to 13.5 months.

The posted ORR was 0.0% in each cohort. The arm-specific 95% confidence intervals extend to 16.1% and 23.2%, reflecting uncertainty around the zero observed response estimates. Secondary outcomes provide additional descriptive information, but their unadjusted comparisons should be interpreted cautiously because multiplicity handling and other analysis details are not reported in the registry record. The results describe the posted trial data and do not establish treatment effects beyond this study.