Trial of mFOLFOX6 + Trastuzumab + Avelumab in Gastric and Esophageal Adenocarcinomas: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Trial of mFOLFOX6 + Trastuzumab + Avelumab in Gastric and Esophageal Adenocarcinomas (NCT03783936), a na phase 2 study. The primary outcome is Best Objective Response Rate (bORR).

2. Clinical Question

The trial was initially intended as a multicenter, single-arm, open-label Simon's two-stage Phase 2 study. The registry summary says accrual would stop after Stage I enrollment of 18 participants, and that the decision was not due to safety issues. Participants already receiving treatment were to continue until protocol-defined criteria were met. These statements describe the planned conduct and the stated reason for stopping accrual; they do not establish how many participants completed treatment or had an endpoint assessment.

3. Trial Design

The registry identifies this as a Phase 2, single-group study of mFOLFOX6 plus trastuzumab and avelumab for first-line treatment of metastatic HER2-amplified gastric and esophageal adenocarcinomas. It lists the study as completed and records enrollment of 18. Allocation is not applicable, and masking is absent, consistent with the open-label design described in the summary. No parallel control group is registered.

The initial intent was a multicenter Simon's two-stage Phase 2 trial. The summary states that accrual would halt after Stage I, when enrollment reached 18, and explicitly says the decision was not due to safety issues. Participants already on treatment were to continue until protocol-defined criteria were met. The registry record does not explain why the study did not proceed beyond Stage I, beyond stating that the decision was not due to safety issues. It does not report a protocol amendment, the stage-specific decision criteria, or whether the planned design was changed.

The treatment description lists oxaliplatin, leucovorin and 5 fluorouracil as mFOLFOX6, alongside trastuzumab and avelumab. The registered arm is titled Induction and Maintenance; its description refers to cycles 1–9 for induction and specifies a 14-day cycle. The record includes dose and administration details, but does not report participant-level exposure, dose changes or treatment completion.

The primary endpoint was best objective response rate by 24 weeks. Progression-free survival, progression-free survival by iRECIST, overall survival, disease control rate and grade 3–4 treatment-related adverse events are listed as secondary outcomes. Although the record lists six outcome measures, it reports zero analyses. The registry record does not report endpoint estimates, an analysis population, statistical methods used for the outcomes, or a comparison with another regimen. The record’s design supports a descriptive account of outcomes in the enrolled group, not a randomized estimate of comparative treatment effects.

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

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 Best Objective Response Rate (bORR), measured over 24 weeks.

Best objective response rate (bORR) is the percentage of participants whose best recorded tumor response meets the definition of complete response or partial response during the specified assessment period. For this trial, the period was 24 weeks and the response criteria were RECIST 1.1. In RECIST, complete response refers to disappearance of all target lesions, while partial response requires a reduction in the summed longest diameters of target lesions. Stable disease and progressive disease do not qualify as complete or partial response for the bORR definition.

The endpoint is used to describe tumor shrinkage meeting prespecified response criteria. It reduces a series of tumor assessments to a participant-level classification of best response within the assessment window. The registry description says that a complete or partial response needed confirmation by a repeat assessment performed no less than 4 weeks later to count as confirmed; otherwise it would be considered unconfirmed. This distinction matters because an apparent response at an assessment may not be confirmed on subsequent evaluation.

A response percentage describes the share of the analyzed participants who met the endpoint definition; a confidence interval communicates uncertainty around that estimate. The registry lists a 95% confidence interval as the dispersion measure, but does not report the bORR estimate, its confidence interval, or the counts used to calculate it. The record also does not state how participants without an assessment by 24 weeks were handled. Without those details, the registered endpoint definition can be explained, but its observed value and precision cannot be evaluated.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods are described as general concepts where the registry record does not report what the investigators actually used. The distinction matters: an endpoint definition is not the same as a statistical analysis plan, and a planned design does not establish which analyses were ultimately performed.

What does a response percentage estimate?

For a binary endpoint such as bORR, each participant is classified according to whether the endpoint occurred. The response percentage is the proportion of the analysis population classified as having a complete or partial response under the registered criteria. It is a summary of observed outcomes in the analyzed group, not a measurement of the size of tumor shrinkage for every participant. The registry record reports the endpoint definition and a planned 95% confidence interval, but no response count, denominator or estimate.

A confidence interval describes the statistical uncertainty around an estimated quantity under the assumptions of the method used. A narrower interval generally indicates greater precision, while a wider interval indicates greater uncertainty. The interval does not correct for bias from selection, missing assessments, outcome misclassification or the absence of a control group. No confidence interval for bORR is reported in the registry record, so the precision of the primary endpoint estimate cannot be assessed.

What can a single-group design establish?

A single-group study can characterize outcomes observed among enrolled participants. Without a concurrent comparator, however, it cannot separate an intervention-associated outcome from other explanations, such as the natural course of disease, participant selection or differences in care and assessment. This is a design limitation rather than a problem that can be resolved by calculating a more precise confidence interval. A precise description of one group would remain a description of that group, not a randomized treatment contrast.

The trial was open-label, with masking listed as none. In general, knowledge of treatment assignment can affect subjective decisions, participant reporting or assessment conduct, depending on the outcome and safeguards in place. RECIST response uses imaging-based criteria, but the registry record does not report whether assessments were centrally reviewed, whether reviewers were blinded, or how disagreements were resolved. Those procedures are not reported in the registry record and should not be assumed.

How should the two-stage intent be understood?

The registry summary describes an initial intent to use Simon's two-stage design and says accrual would halt after Stage I enrollment of 18 participants. In general, a two-stage design can allow a study to stop early according to prespecified outcome criteria, limiting further enrollment when results meet a stopping rule. The record does not report the planned response thresholds, the operating characteristics of the design, the stage-specific analysis, or whether a statistical stopping boundary was met. It says the decision to halt accrual was not due to safety issues, but it does not state the statistical reason for not proceeding to another stage.

Enrollment after Stage I is not itself evidence that the stage's efficacy criterion was met or failed. The registry record reports no analyses, and no primary response results. Therefore, no conclusion about a stage decision can be drawn from the enrollment count or the study status. The protocol's design rationale, sample-size justification and interim analyses are not reported in the registry record.

What do the sample-size illustrations represent?

The record describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. For time-to-event outcomes, it specifies total events; for other outcomes, it specifies participants per arm. Such calculations are planning scenarios, not observed study findings or observed power. They describe hypothetical requirements under stated assumptions, rather than evidence that the trial had a particular probability of detecting an effect.

The record does not report the numerical outputs of those illustrations in the registry record for this analysis. It is not possible to discuss a particular projected sample size or to treat the illustration as the trial's own design calculation. In addition, the actual trial was single-group, whereas the stated illustration assumptions include equal allocation between two arms. A hypothetical two-arm planning scenario should not be mistaken for the enrolled trial design.

What remains unknown about analysis?

The registry lists secondary outcomes, but does not report statistical models, confidence intervals or results for them. The number of analyses recorded is zero. The analysis population, handling of participants without evaluable assessments, missing-data assumptions, any adjustment for prognostic factors, and any interim analysis details are not reported in the registry record. General statistical principles cannot fill those reporting gaps. Without an analysis population and methods, a reader cannot determine which participants contributed to each outcome or how uncertainty was estimated.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and several secondary outcomes, including progression-free survival, iRECIST progression-free survival, overall survival, disease control rate and grade 3–4 treatment-related adverse events. When a study evaluates multiple endpoints or conducts repeated statistical tests, the chance of at least one apparently notable finding can increase. Statistical plans may address this through a prespecified hierarchy, adjustment or descriptive interpretation. The registry record does not report whether any multiplicity strategy was planned or used. No outcome analyses are reported, so the record does not establish how multiple outcomes were interpreted.

Missing-data handling is also not reported. For bORR, potentially relevant details include whether participants without a response assessment by 24 weeks were included in the denominator, how early discontinuation was handled, and how unconfirmed responses were classified in the analysis. The endpoint description explains that responses not confirmed by a repeat assessment no less than 4 weeks later would be considered unconfirmed, but it does not state how such responses entered the primary analysis. The missing-data assumptions and sensitivity analyses are not reported in the registry record.

The analysis populations are not reported either. Enrollment of 18 does not identify the number assessed for bORR, the population used for safety summaries, or the participants included in survival analyses. The registry record does not describe whether analyses were based on all enrolled participants, those who received treatment, or another defined group. It also does not report a statistical analysis plan or explain how protocol deviations and incomplete follow-up were handled.

11. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: tumor hemorrhage (11.1%), cardiac arrest (5.6%), dizziness (5.6%), fall (5.6%), hematoma (5.6%).

The registry adverse-event summary lists 18 participants at risk for serious adverse events and 7 affected. It lists 18 at risk for other adverse events and 18 affected. The summary also records 9 deaths. The serious adverse-event terms with listed percentages are tumor hemorrhage, 11.1%; cardiac arrest, 5.6%; dizziness, 5.6%; fall, 5.6%; hematoma, 5.6%; and lung infection, 5.6%. These entries describe the registry’s safety reporting; they should not be read as a comparative analysis.

The record does not give event timing, duration, recurrence, severity detail beyond the listed categories, or participant-level overlap between event terms. It does not establish that a listed event or death was caused by the study treatment. The listed death count should therefore not be interpreted as a treatment-attributable count. The registry summary also does not provide a comparison group, person-time, or a detailed account of treatment exposure to support comparisons of event rates.

The trial summary states that stopping accrual after Stage I was not due to safety issues. That statement reports the stated reason for the accrual decision; it does not replace the adverse-event summary or establish that no safety concerns occurred. Conversely, the serious-event entries and deaths do not, without attribution and context, establish that treatment caused them.

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

The registry record describes a Phase 2, open-label, single-group study of mFOLFOX6, trastuzumab and avelumab in first-line, metastatic, HER2-amplified gastric and esophageal adenocarcinomas. Its primary endpoint was bORR by 24 weeks under RECIST 1.1. The endpoint definition specifies what would count as a qualifying response and distinguishes confirmed from unconfirmed response. These details define the clinical outcome the trial set out to measure.

The record does not report the observed bORR, its response counts, or its 95% confidence interval. It therefore does not show how frequently qualifying responses were recorded or how precisely that frequency was estimated. The listed enrollment of 18 is not a substitute for an endpoint result or for the endpoint analysis denominator. The registry also does not report how missing response assessments were handled.

Even if an observed response estimate were available, the single-group design would not provide a randomized comparison against another treatment. It could describe outcomes in the enrolled group, but it could not by itself establish that outcomes differed because of the regimen. The absence of masking may also be relevant to assessment procedures; the registry does not report whether outcome review was blinded or centrally assessed.

The secondary outcomes and adverse-event counts provide some context about what the study tracked, but the record reports no analyses for the listed efficacy outcomes. The safety summary includes serious-event counts and deaths, without establishing cause or treatment attribution. The statement that accrual stopped after Stage I for reasons other than safety should not be taken as an efficacy finding or as evidence that the regimen was free of safety concerns.

Accordingly, the record establishes the intended research question, design, endpoint definition and selected safety summaries. It does not establish the magnitude or precision of the primary response outcome, a comparative treatment effect, or a full benefit-harm assessment. The protocol rationale, statistical analysis plan and missing-data methods are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically notable because its registered design and its result reporting answer different questions. The single-group design can describe outcomes among participants treated in the study, while the absence of a concurrent comparator prevents a direct randomized comparison. That distinction remains important even when an endpoint is defined with standardized response criteria and an estimate is accompanied by a confidence interval.

The primary endpoint is clearly specified as best objective response by 24 weeks, with RECIST 1.1 criteria and a confirmation requirement. Such definitions help make the intended outcome interpretable: they identify the response categories that qualify and the assessment window. But a well-defined endpoint does not ensure that an estimate is available. The registry record reports no primary response result, no confidence interval and no analysis population, so neither the event proportion nor its precision can be judged.

The Stage I accrual statement also illustrates why design descriptions must be separated from outcome evidence. The initial intent was a Simon's two-stage Phase 2 study, yet accrual was to halt after Stage I enrollment of 18. The summary says this was not due to safety issues, but does not report the decision criteria or statistical results that explain the course of the study. Enrollment alone does not reveal whether an efficacy boundary was met.

Finally, the registry distinguishes a clinical endpoint program from the results actually reported. It lists survival, disease control and safety outcomes, and includes selected serious adverse-event figures, but reports zero analyses. This limits the ability to examine efficacy, uncertainty, missingness or consistency across outcomes. Statistically, the key lesson is that endpoint definitions, design labels and safety counts are informative context, but they do not replace analyzed outcome estimates or a transparent account of how those estimates were produced.

14. 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 central limitation is that the registry record reports no primary endpoint estimate, response count or confidence interval, despite specifying bORR by 24 weeks as the primary outcome. It also reports zero analyses. This prevents assessment of the observed response proportion, its precision, or its consistency with the endpoint definition.

The study is single-group and open-label, with no concurrent control. Thus, even a reported response proportion would not directly estimate a causal contrast against another treatment. The registry record does not report whether response assessments were centrally reviewed or blinded. It also does not provide details about participant characteristics, eligibility criteria, treatment adherence, or modifications, limiting assessment of who contributed to the findings and how care was delivered.

The summary says accrual would stop after Stage I enrollment of 18 and that the decision was not due to safety issues. The record does not give the Stage I decision rules, clarify why there was no further accrual, or report whether planned statistical criteria were evaluated. The protocol rationale and sample-size justification are not reported in the registry record. The sample-size assumptions described for planning illustrations do not establish the trial's own design calculation or observed power.

For the primary endpoint, the record does not report the analysis denominator, missing-assessment rules, or how unconfirmed responses entered the analysis. It does not state the analysis populations or give statistical methods for the listed secondary outcomes. The adverse-event summary includes counts and selected terms, but lacks reported attribution, timing and exposure context. The death count is not linked to a reported cause or treatment relationship. These gaps limit both efficacy interpretation and safety attribution.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Trial of mFOLFOX6 + Trastuzumab + Avelumab in Gastric and Esophageal Adenocarcinomas — NCT03783936.
• Registry record last updated 2023-12-19; status last verified 2023-11.
• Results first posted 2023-12-19.
• 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.

18. Record Summary

The registry record describes a completed Phase 2, single-group, open-label study of mFOLFOX6 plus trastuzumab and avelumab for first-line, metastatic, HER2-amplified gastric and esophageal adenocarcinomas. Enrollment is listed as 18. The initial intent was a multicenter Simon's two-stage trial, and the summary says accrual would halt after Stage I; it states that this decision was not due to safety issues.

The primary endpoint was bORR by 24 weeks under RECIST 1.1, with confirmed response requiring a repeat assessment no less than 4 weeks later. The registry lists survival, disease control and adverse-event outcomes as secondary measures. It reports no analyses and no primary response estimate or confidence interval.

The safety summary lists 7 participants with serious adverse events among 18 at risk, 18 with other adverse events among 18 at risk, and 9 deaths. Event attribution and detailed context are not reported. The record supports interpretation of the intended design and registered outcomes, but not a quantified primary efficacy result or a randomized treatment comparison.