ZW25 in Women With Endometrial Cancers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for ZW25 in Women With Endometrial Cancers (NCT04513665), a non-randomized phase 2 study. The primary outcome is Overall Response Rate/ORR of Participants.

2. Clinical Question

3. Trial Design

This is a Phase 2 study of ZW25 in women with HER2-overexpressed advanced endometrial cancers and carcinosarcomas. It is classified as non-randomized and unmasked. The registry also labels the model as crossover, but it does not describe the sequence of treatments, the conditions for crossing over, or whether any participant actually crossed over. The record lists ZW25 as the drug intervention and does not identify a separate control treatment.

The planned design had two stages. Stage 1 was described as enrolling participants with recurrent endometrial carcinoma or carcinosarcoma, HER2 overexpression, and 1-2 prior lines of chemotherapy. The registry says 16 participants would be accrued in Stage 1. If 2 or greater responses were seen during that stage, Stage 2 would accrue an additional 9 participants from the same described population. Stage 2 did not accrue; the primary endpoint population note explicitly says no participants accrued to the Stage 2 arm.

The overall enrollment is 16, and the registry status is completed. The recorded study start is 2020-08-12, and the primary completion date is 2024-03-08. These dates describe the study record and do not provide details about when individual participants were treated or assessed. The reason Stage 2 did not accrue is not reported in the registry record. The record does not state whether the Stage 1 response threshold was met, how response assessments were reviewed, or whether any protocol changes occurred.

The design is a stage-based, non-randomized evaluation rather than a randomized comparison of ZW25 with another treatment. Consequently, any outcome description would pertain to participants treated in the study and would not, by itself, estimate a difference against a concurrent control group. The registry does not provide additional design details such as a formal sample-size justification, treatment assignment procedures, or a 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: No participants accrued to Stage 2 arm.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: No participants accrued to Stage 2 arm

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: No participants accrued to Stage 2

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate/ORR of Participants, measured over </= 24 weeks from the start of treatment.

The primary endpoint is the overall response rate, described as the count of participants with a complete response or partial response by RECIST v 1.1 within 24 weeks from the start of treatment. In general, an endpoint defines the outcome used to address a study question. Here, the response definition sets the kinds of observed tumor response that count toward the primary measure, while the stated assessment period specifies when those responses are considered.

Response endpoints are used in clinical studies to describe whether participants meet prespecified criteria for tumor shrinkage or other response classification. The registry names RECIST v 1.1 as the assessment standard but does not describe the imaging schedule, assessment process, reviewer roles, or rules for handling assessments that cannot be classified. Those details are not reported in the registry record.

Although the endpoint title uses the term “rate,” the registered parameter is a count of participants. To interpret a rate, readers would need both the number meeting the response definition and the relevant population denominator. The record does not post the complete-response count, the partial-response count, or their combined count, and it does not provide a numerical response-rate estimate. It therefore does not permit a quantitative assessment of the primary endpoint. The Stage 2 population note says no participants accrued to that stage, so the registry does not provide Stage 2 response results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the endpoint count?

The registry defines the primary outcome as the count of participants achieving complete response or partial response by RECIST v 1.1 within 24 weeks from treatment start. This is a binary outcome at the participant level: each participant either meets the response definition within the stated period or does not. The registered parameter is a count, rather than a continuous measure of tumor change. The registry does not give the observed count or the number assessed for response, so the endpoint cannot be reconstructed from the record.

In general, a binary endpoint can be summarized by counting participants in each outcome category and relating the count of participants meeting the criterion to the population being analyzed. The denominator matters: participants who were not assessed, withdrew, or had an unevaluable result may be handled in different ways under different analysis rules. The rules for these situations are not reported in the registry record. A response count should therefore not be interpreted as a response rate without knowing the analysis population and denominator.

What would an estimate and confidence interval convey?

In general, an estimate describes the response observed in the analyzed study population, while a confidence interval expresses the statistical uncertainty around that estimate. A narrower interval indicates greater precision under the assumptions used for the analysis; a wider interval indicates less precision. An interval does not remove limitations of study design or guarantee that the estimate applies to other populations. The registry does not report a response estimate or confidence interval, and it does not report the method for estimating either.

For a non-randomized study without a separate control treatment, a response estimate would be descriptive of the participants and assessment procedures in that study. It would not directly estimate a difference between ZW25 and another treatment. Comparisons with results from other studies can be affected by differences in eligibility, prior treatment, disease characteristics, outcome assessment, and follow-up. The registry does not provide information allowing those potential differences to be evaluated.

How does the staged accrual plan work?

The registered plan specified 16 participants for Stage 1 and an additional 9 for Stage 2 if 2 or greater responses were seen in Stage 1. This is a conditional accrual rule: progression to the later stage depends on an observed response threshold. Such a rule can govern whether further enrollment is pursued, but the registry record does not report the operating characteristics of the rule, a formal sample-size justification, or the decision made on the basis of Stage 1 results. It also does not state why Stage 2 did not accrue. The lack of Stage 2 participants does not reveal the Stage 1 response count by itself.

The record reports overall enrollment of 16, consistent with the number described for Stage 1, and says no participants accrued to the Stage 2 arm. It does not provide a participant flow account, such as numbers screened, treated, assessed, or included in each analysis. Those details are not reported in the registry record.

What do the sample-size assumptions mean?

The planning illustration is described as using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It is also described as targeting an effect equal to a stated fraction of the observed effect. These are assumptions for a planning illustration, not a description of the actual study design: the trial is non-randomized and no two-arm control comparison is listed. The illustration is not the trial’s own design or sample-size justification.

In general, a significance level is a threshold used in a specified statistical test, and a normal approximation is a mathematical method for approximating a sampling distribution. Equal allocation describes a planning scenario that assigns the same planned number of participants to each comparison group. A target effect fraction represents a planning effect smaller than the reference observed effect, but the values for those fractions and the resulting sample-size estimates are not reported in the registry record. Since the primary response result is not posted, the observed effect underlying the illustration cannot be checked from this record. These planning estimates are not observed power, and they do not establish the precision or performance of the completed study.

Why does dispersion matter in general?

For outcomes measured as counts, the variability of the count can affect standard errors and the reliability of model-based uncertainty estimates. Over-dispersion occurs when the variability is greater than a model assumes. If it is ignored, uncertainty can be understated and statistical conclusions can be too confident. Appropriate handling depends on the outcome, data structure, and analysis model. The primary outcome in this study is registered as a count of participants meeting a binary response criterion, and no dispersion estimate or count-model analysis is reported. A more detailed statistical model cannot be inferred from the endpoint label.

What analysis details are unavailable?

The registry lists no analyses for the primary outcome. It does not report a statistical test, model, confidence-interval method, covariate adjustment, subgroup analysis, sensitivity analysis, or interim analysis. It also does not state how missing response assessments would be handled, which participants would be included in the primary analysis, or how protocol deviations would be treated. These are not details that can be inferred from the phase, endpoint name, or staged design. The posted record therefore permits explanation of the endpoint and general statistical principles, but not a reproduction or evaluation of an outcome analysis.

10. Multiplicity, Missing Data and Analysis Populations

The registry identifies one primary outcome measure, overall response rate, and lists no analyses. A registered endpoint does not by itself show how many statistical tests, subgroup comparisons, or secondary outcome analyses were planned or performed. A multiplicity adjustment controls the risk of false-positive findings when multiple statistical hypotheses are tested; whether such an adjustment is needed depends on the analysis plan and the claims being made. The planned testing strategy, any secondary endpoints, and any multiplicity procedures are not reported in the registry record.

Missing outcome data can affect both the estimated response and its precision. For example, the interpretation may depend on whether participants without an evaluable response are counted as nonresponders, excluded, or handled using another method. These are general possibilities, not descriptions of this study’s methods. The registry does not report missing response counts, reasons for missingness, an imputation method, or a sensitivity analysis.

The record also does not define the primary analysis population beyond the endpoint population note that no participants accrued to Stage 2. It does not state whether the primary analysis includes all enrolled participants, only treated participants, or only participants with an evaluable assessment. The numbers included in any response analysis are not reported in the registry record. Without these details and the outcome numerator and denominator, neither the response estimate nor the effect of missing assessments can be evaluated.

The absence of these reporting details does not show that analyses or missing-data procedures were not undertaken. It means that they are not reported in the registry record and cannot be assessed from the posted information.

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: abdominal pain (12.5%).

The registry’s Stage 1 safety summary reports 16 participants at risk for serious adverse events and 2 participants affected. It reports 16 participants at risk for other adverse events and 16 affected. The same Stage 1 summary lists 8 deaths. Stage 2 has zero participants at risk and zero affected in the posted summary, consistent with the record stating that no participants accrued to that stage.

These are counts of participants in the registry’s safety groups. A serious adverse event is a category based on seriousness, while “other adverse events” is reported separately in the record. The figures do not state that an event was caused by ZW25. They also do not provide the timing, duration, severity details, exposure time, or event-level relationship to treatment. The 8 deaths should therefore be read as deaths recorded in the Stage 1 safety summary, not as deaths attributed to the drug.

Abdominal pain is listed as the top serious term at 12.5%. The record does not provide additional detail about the number of abdominal pain events, whether the percentage refers to participants or events, or how the term was classified. The safety information is descriptive and has no concurrent control group for comparison. It cannot establish comparative risk or causation.

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

The registry record shows that a Phase 2 study of ZW25 enrolled 16 participants and was marked completed. Its primary endpoint was defined as the number of participants with complete or partial response by RECIST v 1.1 within 24 weeks from treatment start. Its staged plan specified further accrual if 2 or greater responses were seen in Stage 1, but no participants accrued to Stage 2.

Those design details do not answer the primary clinical question. The registry does not post the response count, response rate, confidence interval, or statistical analysis. It also does not say whether the Stage 1 response threshold was met. No response outcome can be inferred from the fact that Stage 2 did not accrue. The posted record therefore does not establish a particular response result.

Even if a response estimate were available, a result from this non-randomized study without a listed control treatment would describe the outcomes observed among the participants assessed under the study conditions. It would not by itself show what would have happened to comparable participants under another treatment. Differences in participant characteristics, prior treatment, assessment, and follow-up can complicate comparisons across studies. The crossover label cannot be used to make a comparative interpretation because the registry does not explain crossover procedures or report whether they occurred.

The safety counts indicate that adverse events and deaths were recorded in the Stage 1 safety group. They do not identify the cause of individual events or deaths, and the record does not provide comparative safety data. A serious-event count is not a complete description of all safety dimensions, and an event term listed as the top serious term does not establish a causal relationship to treatment.

The sound conclusion is limited: the record documents a completed, non-randomized Phase 2 study with 16 enrolled participants, a specified response endpoint, no Stage 2 accrual, and the posted Stage 1 safety counts. It does not provide enough primary outcome information to quantify response or enough comparative information to estimate a treatment effect.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of why a registered endpoint and an observed result are different things. The endpoint description gives a clear response definition—complete response or partial response by RECIST v 1.1 within 24 weeks from treatment start—but the registry does not post the response count or its analysis. Without those results, the central endpoint cannot be quantified. The study’s enrollment and completion status cannot substitute for the missing outcome estimate.

The two-stage plan also illustrates how accrual rules depend on an intermediate response threshold. Stage 1 was planned for 16 participants, and Stage 2 would add 9 if 2 or greater responses were seen. Because the record reports no Stage 2 accrual but does not state the Stage 1 response count or the reason for stopping accrual, readers cannot determine which decision path occurred. This highlights the value of reporting both the prespecified decision rule and the observed data that triggered a decision.

The trial’s non-randomized design matters for interpretation. A descriptive response outcome can characterize participants treated in the study, but a treatment comparison requires an appropriate counterfactual. No separate control treatment is listed, so causal comparisons cannot be drawn from the design details in the registry record. Likewise, a crossover classification without details is not enough to establish that crossover generated a valid within-participant comparison.

The safety counts have a similar interpretive boundary: they describe participants in the Stage 1 safety group but do not establish causation or comparative risk. Together, these features demonstrate why complete reporting of outcome counts, denominators, analysis populations, uncertainty estimates, missing-data handling, and safety attribution is important for statistical interpretation.

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 main limitation for statistical interpretation is that the registry does not report the primary response result. It gives the endpoint definition but no complete-response count, partial-response count, combined count, response-rate estimate, confidence interval, or test result. The population note states that no participants accrued to Stage 2, but the record does not explain why or report whether the Stage 1 response threshold was met. The missing outcome information prevents a quantitative interpretation of efficacy.

The study is non-randomized and does not list a separate control treatment. Therefore, the record does not support a like-for-like comparison of ZW25 with another treatment. The registry labels the model crossover but does not describe the crossover procedure or report whether participants crossed over, limiting interpretation of that design label.

Other details needed to assess the response result are not reported in the registry record, including the analysis population, the response assessment schedule, handling of unevaluable or missing assessments, and the method for calculating uncertainty. The registry does not provide a statistical analysis plan, formal sample-size justification, or information about interim analyses. These absences prevent assessment of whether the reported endpoint and any analysis were implemented as intended.

The safety summary is limited to participant counts and a top serious term. It does not provide timing, duration, exposure, detailed event counts, treatment attribution, or a control group. The counts should not be interpreted as causal effects. Finally, the sample-size assumptions describe a planning illustration based on a normal approximation and equal allocation between two arms, which do not match the listed non-randomized study structure. The resulting illustration cannot be treated as the trial’s actual design justification or observed power.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: ZW25 in Women With Endometrial Cancers — NCT04513665.
• Registry record last updated 2025-05-22; status last verified 2024-03.
• Results first posted 2024-10-21.
• 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

NCT04513665 is a completed Phase 2 study of ZW25 in women with HER2-overexpressed advanced endometrial cancers and carcinosarcomas. It is listed as non-randomized, unmasked, and crossover, with Memorial Sloan Kettering Cancer Center as sponsor. The registry records 16 participants enrolled, a start date of 2020-08-12, and a primary completion date of 2024-03-08.

Stage 1 was planned for 16 participants with recurrent endometrial carcinoma or carcinosarcoma, HER2 overexpression, and 1-2 prior lines of chemotherapy. Stage 2 was conditional on 2 or greater responses and would have accrued an additional 9 participants; no participants accrued to Stage 2. The primary endpoint was the count of complete or partial responses by RECIST v 1.1 within 24 weeks from treatment start. The response results and analysis are not reported in the registry record.

The Stage 1 safety summary lists 2 participants with serious adverse events, 16 with other adverse events, and 8 deaths. Abdominal pain is listed as the top serious term at 12.5%. The record does not establish event causation.