Pembrolizumab (MK-3475) Plus Lenvatinib (E7080/MK-7902) Versus Chemotherapy for Endometrial Carcinoma (ENGOT-en9 / MK-7902-001): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pembrolizumab (MK-3475) Plus Lenvatinib (E7080/MK-7902) Versus Chemotherapy for Endometrial Carcinoma (ENGOT-en9 / MK-7902-001) (NCT03884101), a randomized phase 3 study. The primary outcome is Progression-free Survival (PFS) Based on Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1) as Assessed by Blinded Independent Central Review (BICR) in Mismatch Repair Proficient (pMMR) Participants.

2. Clinical Question

3. Trial Design

LEAP-001 (NCT03884101) was a Phase 3, randomized, open-label, parallel-group study comparing lenvatinib plus pembrolizumab with paclitaxel plus carboplatin as first-line treatment for advanced or recurrent endometrial carcinoma. The registry lists an enrollment of 842 participants and a completed study status. The recorded study start was 2019-04-11, and primary completion was 2023-10-02.

The primary analysis concerned PFS in pMMR participants. The registry lists 320 participants in the lenvatinib plus pembrolizumab group and 322 in the paclitaxel plus carboplatin group for this endpoint. Participants were analyzed according to their randomized assignment. This preserves the randomized comparison as the basis for the treatment-group estimate, even when treatment received later differs from assignment.

The study was not masked. That matters when outcomes or decisions can be influenced by knowledge of treatment assignment. For the primary endpoint, the registry specifies blinded independent central review of RECIST 1.1 disease assessments. This provides an assessment process separate from the treating team, although it does not make the entire trial blinded.

The stated study purpose also included comparing overall survival, and the registry lists secondary outcomes involving response, quality of life, and adverse events. The registry record does not report the original sample-size justification, the full analysis plan, or details of any interim analyses. Amendment 7 allowed eligible participants completing the study to transition to an extension study, if available, for continued treatment with one or both study drugs.

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: Per protocol, response/progression or adverse events (AEs) that occurred during the second course were not counted towards efficacy outcome measures or safety outcome measures, respectively.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized pMMR participants were analyzed according to the treatment arm to which they were randomized.

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 Progression-free Survival (PFS) Based on Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST 1.1) as Assessed by Blinded Independent Central Review (BICR) in Mismatch Repair Proficient (pMMR) Participants, measured over Up to approximately 44 months.

The primary endpoint was PFS in pMMR participants. It measures the time from randomization to the first documented progression or death from any cause, whichever occurred first. Progression was assessed under RECIST 1.1 by blinded independent central review. Under the registry definition, progression included at least a 20% increase in the sum of target-lesion diameters, with an absolute increase of at least 5 mm, or the appearance of one or more new lesions.

Time-to-event endpoints such as PFS retain information about when an event occurs, rather than reducing follow-up to a simple count of participants with an event. A median summarizes the time by which half of the observed group has experienced the endpoint, subject to the survival-analysis method and follow-up available. Comparing medians alone does not fully describe the relative event pattern over follow-up, which is why the registry also reports a hazard ratio.

PFS is used in oncology trials because progression and death are clinically meaningful events that can be observed during follow-up. It is not the same as overall survival: the primary PFS definition includes documented progression as well as death, while overall survival concerns death. The reported PFS timeframe is up to approximately 44 months. The registry record does not report further details about assessment schedules or handling of particular intercurrent events.

7. Results Posted in the Registry

• Lenvatinib + Pembrolizumab vs Paclitaxel + Carboplatin: hazard ratio 1.01 (95% CI 0.83–1.24), reported p = 0.555 (as reported in the registry).

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.

Objective Response Rate (ORR) Based on RECIST 1.1 as Assessed by BICR in pMMR Participants

• 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:
  • Paclitaxel + Carboplatin: 55.2% (175/317) versus 50.9% (162/318) on control; risk difference 4.3 percentage points (95% CI -3.5 to 12.0), risk ratio 1.08 (0.94–1.25); unadjusted two-proportion z-test p = 0.282.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Paclitaxel + Carboplatin and the control. The interval gives the range of effects that remain compatible with the data.

ORR Based on RECIST 1.1 as Assessed by BICR in All Randomized Participants

• 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:
  • Paclitaxel + Carboplatin: 56.0% (233/416) versus 56.0% (234/418) on control; risk difference 0.0 percentage points (95% CI -6.7 to 6.8), risk ratio 1.00 (0.89–1.13); unadjusted two-proportion z-test p = 0.993.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Paclitaxel + Carboplatin and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes spanning objective response, quality of life, and adverse events. Numerical response-rate results are posted for pMMR participants and for all randomized participants. They provide a broader descriptive view of tumor response, but they do not supersede the primary PFS comparison, which concerns time to progression or death in the pMMR population. Response and PFS measure different aspects of outcome and need not move in parallel.

The available secondary results do not support a single comprehensive conclusion across all listed outcomes. In particular, the registry record lists quality-of-life outcomes but does not include numerical results for them in the results. Safety outcomes have their own denominators and definitions and should be interpreted separately. A finding on one secondary outcome cannot establish a consistent pattern across response, quality of life, and safety.

Secondary analyses commonly have less evidential weight than the primary analysis when many outcomes or comparisons are examined. In general, conducting multiple tests increases the chance that at least one result appears notable by chance. Adjustments or a prespecified testing hierarchy can address this, but the registry record does not report a multiplicity procedure for the outcomes discussed. Therefore, secondary p-values should be treated as exploratory unless the analysis plan establishes otherwise.

The registry also includes unadjusted recalculations from summary statistics. Those are not replacements for the sponsor’s stratified response analyses and should not be used to claim confirmation of the primary time-to-event result. The overall secondary evidence is best read as additional context, with attention to population, endpoint definition, analysis method, and the absence of reported results for some listed outcomes.

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

• Objective Response Rate (ORR) Based on RECIST 1.1 as Assessed by BICR in pMMR Participants
• ORR Based on RECIST 1.1 as Assessed by BICR in All Randomized Participants
• Mean Change From Baseline in the Global Health Status/Quality of Life Score of the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core-30 (QLQ-C30) in pMMR Participants
• Mean Change From Baseline in the Global Health Status/Quality of Life Score of the EORTC QLQ-C30 in All Randomized Participants
• Number of Participants Who Experienced an Adverse Event (AE)
• Number of Participants Who Experienced a Serious Adverse Event (SAE)
• Number of Participants Who Experienced an Immune-Related Adverse Event (irAE)
• Number of Participants Who Discontinued Study Intervention Due to an AE

9. Statistical Interpretation

10. Statistical Methods Explained

For PFS, the registry reports medians with 95% confidence intervals and a hazard ratio from a log-rank analysis. These quantities describe different aspects of the time-to-event data. The median locates a point in each group’s event-time distribution. The hazard ratio summarizes the relative rate of experiencing the event at a given time, under the assumptions of the analysis. Neither quantity is a direct percentage reduction in risk, and the hazard ratio should not be interpreted as a comparison of the two medians.

What does a hazard ratio summarize?

In general, a hazard is the event rate among participants who have not yet experienced the event, considered over follow-up. A hazard ratio compares those rates between groups. A value of 1 corresponds to no relative difference under the model; values below or above 1 indicate a relative difference in opposite directions, according to the stated group order. The reported estimate is 1.01. Its 95% confidence interval, 0.83 to 1.24, includes 1, so the data are compatible with no relative difference as well as with effects in either direction within the interval.

A hazard ratio is not automatically a constant biological or clinical effect throughout follow-up. Its interpretation depends on the time-to-event analysis and assumptions about how relative hazards behave. The registry record identifies the log-rank method but does not report enough detail to assess all modeling choices, proportional-hazards diagnostics, or sensitivity analyses.

What does the log-rank method do?

In general, the log-rank method compares the observed timing of events with the timing expected under a no-difference hypothesis, accounting for the fact that participants can have different lengths of follow-up. Participants without a recorded event by their last follow-up are censored rather than treated as if an event had occurred. This method uses the sequence of events over time, whereas a comparison of medians alone does not use all event-time information. The registry names the method but does not report detailed censoring rules or all implementation details.

Why report medians and confidence intervals?

A median offers a recognizable summary of event timing. The accompanying confidence interval expresses statistical uncertainty around that estimate. In general, a confidence interval is a range produced by a method that, under its assumptions, would cover the underlying parameter at the stated rate in repeated use. It is not a guarantee that the true value lies within the particular interval, nor does it describe the range of individual participants’ outcomes.

The group medians are close in magnitude, and the reported confidence intervals overlap. Overlap is not itself a formal test of the treatment comparison. The hazard ratio and its own confidence interval provide the registry’s direct relative comparison. A reader should avoid inferring statistical significance by visually comparing separate group intervals; the uncertainty of a between-group contrast is evaluated using the contrast’s own analysis.

How should the response-rate analyses be understood?

Objective response rate is a binary outcome: a participant is classified as having or not having a response under the stated response criteria. The registry reports response percentages and 95% confidence intervals for pMMR participants and for all randomized participants. It also lists sponsor analyses using a stratified Miettinen & Nurminen method for differences in percentage. That method estimates a difference between group response proportions while accounting for the stated stratification approach. The registry does not give further details about the strata in the available record.

Unadjusted recalculations based on posted summary statistics are not the same analysis as the sponsor’s stratified method. Differences between their estimates or p-values can arise from the method and its assumptions, not necessarily from an error. The recalculated values should be read as simple, unadjusted comparisons; the sponsor’s reported analyses are the trial’s official analyses. They should not be combined as though they were identical estimates from one model.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variability is greater than a model assumes. It can make standard errors too small if the model fails to account for that extra variability, resulting in confidence intervals that are too narrow and tests that are too optimistic. This is a general statistical issue, not a specific finding established by this registry record. The record does not report whether over-dispersion was assessed for any endpoint or whether a correction was used.

The study’s randomized assignment and analysis of participants according to assigned group are relevant to interpretation. Randomization aims to make groups comparable at assignment, while an intention-to-treat style analysis retains that assignment-based comparison. However, the registry record does not provide the full statistical analysis plan, handling of missing endpoint data, details of protocol deviations, or a complete account of all sensitivity analyses. Those omissions limit independent evaluation of analytic choices.

11. Confidence Intervals and Precision

The primary PFS estimate is accompanied by a 95% confidence interval for each group median and for the hazard ratio. The interval around the lenvatinib plus pembrolizumab median is 8.2 to 11.9 months; the interval around the chemotherapy median is 8.4 to 10.4 months. For the hazard ratio, the interval is 0.83 to 1.24. The values within these intervals are estimates compatible with the data and method at the stated confidence level; they are not a prediction interval for individual outcomes.

The hazard-ratio interval includes 1. That makes the data compatible with no relative difference, as well as with a range of relative differences in either direction. The estimate of 1.01 lies close to 1, but proximity to 1 should not be confused with proof that the treatments have identical effects. The interval is the more useful guide to the precision of the relative estimate.

Confidence intervals for each group’s median do not directly answer the same question as the confidence interval for the between-group hazard ratio. Nor does overlap between separate intervals serve as a substitute for the formal comparison. The registry reports the hazard-ratio interval for the primary comparison, which should guide interpretation of the relative result.

General explanation: wider confidence intervals indicate greater uncertainty about the parameter, while narrower intervals indicate greater precision under the analysis assumptions. Precision depends on factors such as the amount of information and the event pattern. The registry record does not provide enough information to attribute the observed interval width to particular causes or to evaluate all assumptions underlying the intervals.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations show the number of total events for the time-to-event comparison, not participants per arm. Other endpoint illustrations use participants per arm. The rows vary the planning effect from 100% of the observed effect to 75% and 50% of that effect. Smaller effects require more information to detect, so the illustrations show larger required totals as the effect to detect becomes smaller.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect is the stated fraction of the observed effect. For the time-to-event endpoint, event totals are based on the Schoenfeld method. The registry record does not report the original trial’s own design assumptions or sample-size justification, so these calculations should not be mistaken for the basis on which the trial was planned.

The 80% power column describes the number of events or participants the illustration estimates would be needed for an 80% chance of detecting the chosen effect under the assumptions, if that effect were real. The 90% power column gives the corresponding requirement for a higher chance of detection. These are planning concepts, not probabilities that the observed result is true. The values are sensitive to the assumed effect size and method.

These figures are planning illustrations based on the observed effect, not observed power. They do not measure the probability that the trial’s result is correct, and they do not change the interpretation of the confidence interval or sponsor-reported analysis.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 1.010, so ln(HR) = 0.0100.
2. d = 4 × (1.96 + 0.842)² / [ln(1.010)]² = 317,097.4, rounded up to 317,098 events in total for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
hr <- 1.01                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(3884101)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 317,098 (80% power), 424,504 (90% power)
 75% of observed effect: events = 563,729 (80% power), 754,673 (90% power)
 50% of observed effect: events = 1,268,390 (80% power), 1,698,014 (90% power)
Simulated power with 317,098 events: 0.799 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Analyses Reported by the Sponsor

The sponsor’s primary analysis reports a hazard ratio of 1.01 with a 95% confidence interval of 0.83 to 1.24, using a log-rank method. The primary time-to-event interpretation relies on this reported analysis; no separate recalculated hazard ratio is presented. The reported group medians and their intervals are descriptive summaries and do not replace the sponsor’s comparative analysis.

For the response-rate outcomes, the sponsor reports stratified Miettinen & Nurminen analyses of the difference in percentage. The unadjusted recalculations use the posted summary counts and do not reproduce that stratified analysis. They are useful as a transparent, simple comparison, but their estimates and p-values need not match the sponsor’s results. The record does not report enough detail to reconstruct the full stratified analysis independently.

There is no identified discrepancy flag for the primary analysis. Even so, the distinct methods should be kept distinct: sponsor-reported results describe the trial’s stated analysis, while recalculated results are unadjusted calculations from summary statistics. Neither should be presented as if it came from the other method. The registry record does not report the complete statistical analysis plan or additional detail needed to resolve every methodological difference.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary PFS endpoint and multiple secondary outcomes, including response, quality of life, and adverse events. Testing several outcomes creates a multiplicity issue: if each is evaluated independently, the chance of at least one apparently unusual result can increase. In general, a prespecified testing hierarchy, adjustment procedure, or clear exploratory designation helps readers understand how to interpret a family of tests.

The registry record does not report a multiplicity adjustment or a complete testing hierarchy for the listed outcomes. It is therefore not possible to determine from this record alone whether the secondary analyses were confirmatory, how error rates were controlled, or whether testing of one outcome depended on results for another. Secondary results should not be treated as equivalent in evidential status to the primary endpoint without that context.

The primary PFS population is identified as randomized pMMR participants analyzed according to randomized treatment arm. The registry also reports response outcomes for pMMR participants and for all randomized participants. Those are different analysis populations and should be kept separate when comparing results. The safety records use still other denominators, so counts from safety summaries cannot be substituted for the endpoint populations.

The registry record does not report full details of missing-data handling, censoring rules, endpoint assessment schedules, or sensitivity analyses for the primary endpoint. It also does not report a complete account of participants excluded from each analysis or the reasons for any exclusions. These are relevant because incomplete follow-up or missing assessments can affect time-to-event and response analyses. The randomized, assignment-based approach is stated for the primary endpoint, but it does not answer every question about missing observations or protocol deviations.

These omissions do not invalidate the posted estimates; they define what can be assessed from the registry record. Independent evaluation of multiplicity, missingness, and analysis populations would require the full protocol and statistical analysis plan, which are not reported in the registry record.

16. 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: intestinal obstruction (5.9%), intra-abdominal haemorrhage (5.9%), diverticulitis (5.9%), urinary tract infection (5.9%), hypokalaemia (5.9%).

The safety records report serious adverse events and other adverse events separately, with denominators that differ from the primary PFS population. For the first-course lenvatinib plus pembrolizumab group, 230 of 420 participants at risk experienced a serious adverse event, and 416 of 420 experienced another adverse event. For paclitaxel plus carboplatin, the corresponding figures were 84 of 411 for serious adverse events and 397 of 411 for other adverse events. The registry also lists a second-course lenvatinib plus pembrolizumab group, with 3 of 17 participants experiencing a serious adverse event and 15 of 17 experiencing another adverse event.

These are counts of participants with events, not counts of distinct events, and the registry’s categories should not be combined. The listed serious-event terms include intestinal obstruction, intra-abdominal haemorrhage, diverticulitis, urinary tract infection, and hypokalaemia, each reported as 5.9%, as well as diarrhoea at 2.6%. The record does not clearly identify the treatment-group attribution for these listed term percentages in the information available.

The registry also records deaths of 233 in the first-course combination group and 253 in the chemotherapy group, with 3 in the second-course group. Those safety-record counts should not be interpreted as a treatment effect or as a substitute for the overall-survival analysis. Differences in group size, follow-up, exposure, and event definitions matter; the registry record does not provide all details needed to interpret these counts comparatively.

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

The primary analysis estimates PFS in randomized pMMR participants. The reported medians were 9.6 months and 10.2 months, and the sponsor-reported hazard ratio was 1.01 with a 95% confidence interval of 0.83 to 1.24. Taken together, these results do not establish a clear difference in PFS between the randomized groups in this analysis. The interval remains compatible with a range of relative effects, so the result is not proof that the treatments are identical.

The result does not answer every question relevant to the trial. It concerns the specified pMMR population and endpoint, not necessarily every participant or every outcome. The registry also lists an overall-survival hypothesis, but the primary result is PFS. No conclusion about overall survival follows automatically from this estimate. Similarly, response rates, quality of life, and adverse events are distinct outcomes and require their own analyses.

The randomized design helps support a comparison between assigned treatment groups, and central blinded review was used for the primary disease assessments. The open-label design remains relevant to outcomes and processes that may be influenced by awareness of assignment. The available registry record does not give enough detail to assess every aspect of follow-up, missing data, multiplicity, or protocol implementation.

Statistical significance, clinical importance, and precision are separate questions. A p-value addresses compatibility with a specified null model under the analysis assumptions; a confidence interval communicates uncertainty about the estimated effect; and clinical importance requires context beyond either statistic. The posted numbers alone do not establish how a treatment choice should be made for an individual or group. They describe the trial’s comparative evidence for its stated endpoint and population.

18. Why This Trial Matters Statistically

LEAP-001 illustrates why a clinical-trial result should be read through its endpoint, population, and analysis method rather than through a single number. The primary endpoint is time-to-event PFS in pMMR participants, while the response-rate outcomes use binary classifications and broader analysis populations. These endpoints answer related but different questions, and their results cannot be treated as interchangeable.

The trial also illustrates the value of separating a group median from a comparative estimate. The medians describe event timing within each group; the hazard ratio summarizes the relative time-to-event comparison. The hazard-ratio confidence interval includes 1, so the primary result does not establish a clear relative difference. Reporting both the estimate and interval allows readers to consider direction and uncertainty rather than relying only on whether a test crosses a threshold.

Open-label conduct alongside blinded central review is another useful design distinction. Masking of participants and study personnel was absent, while the primary endpoint assessment was assigned to blinded independent review. These design features do not eliminate all potential sources of bias, but they clarify which part of the assessment process was blinded.

Finally, the registry separates efficacy, safety, and quality-of-life outcomes, but does not report every detail needed to evaluate multiplicity, missing data, or the complete analysis plan. That makes transparent interpretation important: the primary result can be described, while uncertainty about secondary testing and unreported analytic details should remain explicit. Planning illustrations based on observed effects are also distinct from the trial’s original design justification.

19. 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 primary comparison is reported for pMMR participants, not for the full enrolled population. Although the registry also reports response results for all randomized participants, those are secondary outcomes and do not broaden the primary PFS inference. The endpoint sample sizes and safety denominators differ, which limits simple cross-comparison of counts.

The trial was open-label. Blinded independent central review was specified for the primary RECIST assessment, but the record does not report the full process for resolving assessment differences, assessment frequency, or all safeguards against bias. It also does not report detailed censoring and missing-data rules or sensitivity analyses, so the robustness of the primary estimate to alternative assumptions cannot be assessed from this record alone.

The registry record does not report the original sample-size justification, full statistical analysis plan, multiplicity strategy, or details of interim analyses. These omissions restrict evaluation of the intended error control, monitoring, and prespecified analysis choices. The planning illustrations based on the observed effect are not a substitute for the original design rationale.

Secondary outcomes span response, quality of life, and safety, but the registry does not provide numerical results for every listed outcome in the available record. The available response analyses use different methods for sponsor-reported and unadjusted recalculated results; these should not be conflated. Safety counts also require caution because exposure, denominators, follow-up, and event definitions are not fully characterized in the registry summaries. These limitations define what can be concluded from the posted record; they do not by themselves show that the reported analyses are incorrect.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Pembrolizumab (MK-3475) Plus Lenvatinib (E7080/MK-7902) Versus Chemotherapy for Endometrial Carcinoma (ENGOT-en9 / MK-7902-001) — NCT03884101.
• Registry record last updated 2026-02-02; status last verified 2026-01.
• Results first posted 2024-10-16.
• 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.

23. Record Summary

NCT03884101, LEAP-001, was a completed Phase 3 randomized, open-label, parallel-group study of lenvatinib plus pembrolizumab versus paclitaxel plus carboplatin for first-line treatment of advanced or recurrent endometrial carcinoma. Enrollment was 842. The primary endpoint was PFS assessed by blinded independent central review in randomized pMMR participants.

The reported median PFS was 9.6 months for lenvatinib plus pembrolizumab and 10.2 months for paclitaxel plus carboplatin. The sponsor-reported hazard ratio was 1.01, with a 95% confidence interval of 0.83 to 1.24. This estimate does not establish a clear PFS difference, while its interval allows uncertainty in both directions.

The registry also lists response, quality-of-life, and safety outcomes. The results should be interpreted with attention to endpoint populations, distinct analysis methods, open-label conduct, and details not reported in the registry record, including the full statistical plan and multiplicity approach.