An Open-Label, 2-Cohort, Multicenter, Study of Lenvatinib in Previously Treated Subjects With Unresectable Stage III or Stage IV Melanoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Open-Label, 2-Cohort, Multicenter, Study of Lenvatinib in Previously Treated Subjects With Unresectable Stage III or Stage IV Melanoma (NCT01136967), a non-randomized phase 2 study. The primary outcome is Objective Response Rate (ORR).

2. Clinical Question

3. Trial Design

This was an open-label, multicenter, phase 2 study with two parallel cohorts and nonrandomized allocation. Both cohorts received lenvatinib; cohort membership reflected mutation status and treatment history rather than random assignment. The registry lists enrollment as 182. It does not establish from that figure alone how many participants were included in each endpoint analysis.

Cohort 1 was for participants without the V600E BRAF mutation whose disease had progressed following up to 2 prior systemic anticancer regimens, excluding anti-VEGF regimens. Cohort 2 was for participants with activating BRAF mutations, mainly V600E, whose disease had progressed following BRAF V600E-targeted therapy. Those eligibility histories matter when interpreting cohort results: the groups were not interchangeable, and the registry does not describe them as a head-to-head comparison.

The study began in 2010-08, had primary completion in 2013-04, and is listed as completed. The sponsor was Eisai Inc. The study purpose was to assess objective response rate in previously treated participants with disease progression. Secondary outcomes included progression-free survival, overall survival, disease control rate, clinical benefit rate, adverse events and serious adverse events, changes in clinical biomarkers in whole blood, and plasma lenvatinib concentration. The registry records 8 outcome measures.

Although enrollment and study-completion information are reported, the registry record does not report the protocol’s sample-size justification, a detailed statistical analysis plan, or an interim-analysis plan. The nonrandomized design also means the cohort structure cannot provide the protection against confounding that random treatment allocation can offer.

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: Full Analysis Set (Intent-to-Treat \[ITT\] Analysis Set) included all participants who received at least 1 dose study drug.

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 Objective Response Rate (ORR), measured over From date of treatment start until all participants completed a minimum of 6 cycles (28-day cycles) or discontinued treatment prior to end of Cycle 6 (up to 24 weeks).

The primary endpoint was objective response rate, expressed as the percentage of participants in each cohort with a best overall response of complete response or partial response. The response definition used RECIST v1.1 for target lesions assessed on MRI/CT scans, with independent radiologic review. A complete response required disappearance of all target lesions; any pathological lymph nodes had to have a short axis less than 10 mm. A partial response required at least a 30 percent decrease in the sum of target-lesion diameters, using the baseline sum as the reference.

Responses had to be confirmed by a later assessment at least 4 weeks after the initial assessment. A complete response required a subsequent complete response assessment. A partial response required a subsequent assessment of complete or partial response. This confirmation requirement makes the endpoint more specific than counting an initial scan change alone.

In general, response rate is used to describe the proportion of an assessed group meeting a prespecified tumor-response definition. It is a measure of observed tumor shrinkage under the criteria used; it does not by itself describe how long a response lasts or establish a causal treatment effect. The registry record states the endpoint definition and timeframe but does not report the observed response percentages or confidence intervals.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does objective response rate measure?

Objective response rate is a binary-outcome summary: each participant is classified according to whether the required response was achieved, and the estimate is the percentage meeting that definition. In this study, the qualifying outcomes were complete response and partial response, combined as ORR. The registry record describes a separate percentage for each cohort. It does not report the actual percentages in the information available for this description.

Binary outcomes are straightforward to interpret as proportions, but their precision depends on the number of participants contributing to the analysis and on the observed number meeting the response criteria. The registry lists study enrollment as 182, but does not report the endpoint-specific denominators. It would therefore be inappropriate to infer precision from enrollment alone.

Why use a confidence interval?

The endpoint’s stated dispersion measure is a 95% confidence interval. In general, a confidence interval expresses the uncertainty associated with an estimate under a specified statistical procedure. It gives readers more information than a percentage alone: two estimates can be similar while having different precision. The interval does not account for every source of uncertainty, such as selection effects, measurement limitations, or differences in participant history.

The interval limits are not reported in the registry record information available for this description. The method used to calculate the interval is also not reported. Without those details and the point estimate, the precision of the ORR cannot be evaluated.

How were responses defined?

RECIST v1.1 posted the response framework for target lesions, with MRI/CT assessment and independent radiologic review. Complete response required disappearance of all target lesions, with pathological lymph nodes reduced to a short axis less than 10 mm. Partial response required at least a 30 percent reduction in the sum of target-lesion diameters compared with the baseline sum. The best overall response had to be confirmed: a complete response by a later complete-response assessment, and a partial response by a later complete- or partial-response assessment. Each confirmation had to occur at least 4 weeks later.

In general, confirmation rules reduce the chance that a transient or uncertain imaging change is counted as a response. They also make the endpoint dependent on follow-up assessments being available. The registry record does not report how missing or unevaluable scans, early discontinuations, or incomplete confirmation sequences were handled.

What does the design allow readers to compare?

The study was nonrandomized and open-label, with two parallel cohorts. In general, randomization helps balance measured and unmeasured participant characteristics between treatment groups; it was not used. The cohorts differed by BRAF status and prior treatment context. They can be described separately, but they are not like-for-like control groups for estimating a treatment contrast. Independent radiologic review is part of the endpoint description, but does not turn the study into a randomized comparison.

The registry identifies no analysis records. A detailed statistical plan, including any adjusted analyses or the specific confidence-interval procedure, is not reported in the registry record. Readers should not assume an unreported model or adjustment method.

How should planning calculations be read?

In general, sample-size illustrations estimate the number of participants or events that a hypothetical study might require under stated assumptions. They are not observed power and do not describe the precision or strength of evidence in the completed study. The planning assumptions recorded for this study use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. Time-to-event illustrations count total events; other illustrations count participants per arm.

Those assumptions are for planning illustrations, not a description of the nonrandomized cohort allocation in this trial. The registry record does not report the trial’s own sample-size justification. No sample-size illustration values are included in the record information available for this description, so no planning count can be stated.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort 1 (V600E BRAF Negative)", "Cohort 2 (V600E BRAF Positive)")
n      <- c(93, 89)
events <- c(8, 8)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort 1 (V600E BRAF Negative)  93  8  8.6  3.8  16.2  4.4  16.1  4.5  16.1
ARM  Cohort 2 (V600E BRAF Positive)  89  8  9.0  4.0  16.9  4.6  16.7  4.7  16.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes alongside ORR: progression-free survival, overall survival, disease control rate, clinical benefit rate, safety and tolerability outcomes, change from baseline in clinical biomarkers in whole blood, and plasma lenvatinib concentration. Assessing multiple outcomes can raise a multiplicity issue: in general, the more statistical questions examined, the greater the chance of finding an apparently notable result by chance if each is interpreted independently. A prespecified testing hierarchy or adjustment can address some multiplicity concerns. The registry record does not report whether any such approach was planned or used.

The record identifies the primary endpoint and its response criteria, but it does not report a detailed statistical analysis plan, hypothesis-testing procedure, or interim-analysis plan. It lists no analysis records. Therefore, it is not possible to determine from this information whether secondary outcomes were treated as descriptive, whether formal testing was performed, or how any multiple comparisons were handled.

The endpoint population is described as the Full Analysis Set, also called the ITT Analysis Set, and includes participants who received at least 1 dose of study drug. This definition is important because it specifies who was intended to contribute to the primary analysis. The actual endpoint denominator and the number with evaluable response assessments are not reported in the registry record information.

Missing-data handling is also not reported. In general, missing tumor assessments can affect a response-rate estimate, and the effect depends on why assessments are missing and how the analysis classifies participants without an evaluable response. The record does not state how missing scans, withdrawals, early treatment discontinuations, or unconfirmed responses were treated. The endpoint timeframe allows for treatment completion or discontinuation before the end of Cycle 6, but that description alone does not establish the statistical handling of missing outcomes.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: hypertension (5.4%), nausea (3.4%), abdominal pain (3.4%), deep vein thrombosis (3.2%), pulmonary embolism (3.2%).

The safety fields list participants at risk and participants affected by serious adverse events in each cohort. For Cohort 1, the record lists 93 at risk and 39 affected; for Cohort 2, it lists 89 at risk and 36 affected. These are cohort-level counts, not a comparison against an untreated control group. The corresponding other-adverse-event fields list 93 at risk and 93 affected in Cohort 1, and 89 at risk and 88 affected in Cohort 2. Death counts are not reported in the registry record.

The listed serious adverse-event terms and percentages are hypertension, 5.4%; nausea, 3.4%; abdominal pain, 3.4%; deep vein thrombosis, 3.2%; pulmonary embolism, 3.2%; and pneumonia, 2.2%. These entries describe reported serious-event terms; they do not establish that an event was caused by lenvatinib. The record information does not specify additional details such as event timing, duration, severity grading, or the rules used to select these terms for display.

In general, safety counts should be read alongside the population at risk and the definitions used for event collection. They do not, on their own, provide a controlled estimate of excess risk or establish a causal relationship. The registry describes adverse events and serious adverse events as a safety and tolerability outcome, but comparative safety conclusions are not supported by this nonrandomized design.

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

The study record documents a phase 2 investigation of lenvatinib in previously treated participants with unresectable stage III or stage IV melanoma. It defines a primary response endpoint, the criteria for complete and partial response, and two cohorts based on BRAF status and treatment history. These details clarify what the study set out to measure and how a qualifying response was defined.

The information available in the registry record for this description does not report the observed ORR estimates or their confidence intervals. It therefore cannot establish the size or precision of the response observed in either cohort. The study’s objective, eligibility criteria, and endpoint definition are not outcome results. Nor should the listed enrollment figure be treated as the denominator for the primary analysis without endpoint-specific reporting.

Even if separate cohort results are reported elsewhere in the registry, the nonrandomized design does not support interpreting one cohort as a control for the other. Cohorts differed in BRAF mutation status and prior treatment context. Any observed difference could reflect those differences or other factors, as well as treatment response. The record does not present a randomized comparison against another treatment.

In general, an ORR summarizes the proportion meeting response criteria during the defined assessment period. It does not, by itself, describe response duration, progression-free survival, overall survival, or quality of life. Those are distinct outcomes; progression-free survival and overall survival are listed as secondary outcomes, but their results are not reported in the information available for this description.

The safety fields provide cohort-level counts and selected serious adverse-event percentages. They do not by themselves establish causation or quantify a comparative risk against a control group. Overall, the record supports a careful account of the study question and design, but not a conclusion about treatment efficacy or comparative safety.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of how a clinical question, endpoint definition, and study design must be read together. ORR is a clear, interpretable binary endpoint, and the RECIST v1.1 criteria specify how complete and partial responses were classified. The required confirmation assessment and independent radiologic review add detail to the measurement process.

At the same time, the design sets limits on the comparisons that can be made. The cohorts were nonrandomized and defined by mutation status and treatment history. They are useful for describing outcomes in their respective participant groups, but they do not form a randomized treatment-versus-control comparison. The open-label structure and distinct cohort histories should remain part of any interpretation of results.

The record also illustrates why endpoint estimates and confidence intervals matter. A response percentage alone describes the observed proportion; its interval gives information about precision. Neither can be assessed from an endpoint definition or study enrollment figure. The registry information available for this description does not report the cohort response estimates or interval limits, so statistical interpretation must stop short of judging the response magnitude.

Finally, the listed secondary outcomes raise questions about analysis planning and multiplicity, while missing-data rules affect how a response endpoint is formed. The registry record does not report those methods. Recognizing what is documented—and what remains unreported—is an essential part of reading clinical trial evidence without overstating what the design or posted data can establish.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

A key limitation is that the registry record information available for this description does not include the primary ORR estimates or their confidence intervals. The endpoint definition is detailed, but without the observed results it is not possible to assess response magnitude or precision. The number analyzed for ORR is also not reported; the listed enrollment of 182 should not be substituted for an endpoint denominator.

The nonrandomized design limits causal interpretation and does not provide a like-for-like control comparison. The two cohorts had different BRAF mutation status and prior treatment contexts, including different treatment histories. Consequently, cohort outcomes should be interpreted as group-specific descriptions rather than direct evidence of a treatment difference.

The study was open-label, and the registry record does not provide further detail about procedures that might address potential assessment or conduct biases. It does specify independent radiologic review for the primary endpoint, but does not report how review disagreements were handled. The missing-data rules for absent or unevaluable scans, early discontinuation, and unconfirmed responses are not reported in the registry record.

A detailed statistical analysis plan, sample-size justification, interim-analysis plan, confidence-interval method, and multiplicity strategy are not reported. The record lists no analysis records. These omissions prevent a full assessment of the prespecified inferential approach. The registry also does not report the death counts in the safety fields. These limitations concern what can be established from the record; they do not determine what other study documents may contain.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: An Open-Label, 2-Cohort, Multicenter, Study of Lenvatinib in Previously Treated Subjects With Unresectable Stage III or Stage IV Melanoma — NCT01136967.
• Registry record last updated 2019-11-13; status last verified 2016-11.
• Results first posted 2016-03-16.
• 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.

19. Record Summary

The registry record describes a completed, open-label, multicenter, nonrandomized phase 2 study of lenvatinib in previously treated participants with unresectable Stage III or Stage IV melanoma. Enrollment is listed as 182. The parallel cohorts were defined by V600E BRAF status and prior treatment history.

The primary endpoint was objective response rate by cohort, based on confirmed complete or partial response under RECIST v1.1 with independent radiologic review. The registry information available for this description does not report the observed ORR estimates or confidence intervals. It lists progression-free survival, overall survival, disease control rate, clinical benefit rate, safety, biomarker change, and plasma lenvatinib concentration among the secondary outcomes.

The record gives some cohort-level adverse-event counts and selected serious-event percentages, but does not report death counts. It does not report the detailed analysis plan, missing-data approach, or multiplicity strategy. Cohort results should not be treated as a randomized comparison.