Lenvatinib/Everolimus or Lenvatinib/Pembrolizumab Versus Sunitinib Alone as Treatment of Advanced Renal Cell Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Lenvatinib/Everolimus or Lenvatinib/Pembrolizumab Versus Sunitinib Alone as Treatment of Advanced Renal Cell Carcinoma (NCT02811861), a randomized phase 3 study. The primary outcome is Progression-free Survival (PFS) by Independent Imaging Review (IIR).

2. Clinical Question

3. Trial Design

NCT02811861 is a randomized, parallel-group, phase 3 trial with no masking. It is titled “Lenvatinib/Everolimus or Lenvatinib/Pembrolizumab Versus Sunitinib Alone as Treatment of Advanced Renal Cell Carcinoma,” and its official title identifies it as the CLEAR trial. Eisai Inc. is listed as the sponsor. The registry records a start date of 2016-10-13 and a primary completion date of 2020-08-28. Its status is active, not recruiting.

The registry reports 1069 enrolled participants and three treatment arms. The primary endpoint results list 357 participants in the lenvatinib-plus-everolimus arm, 355 in the lenvatinib-plus-pembrolizumab arm, and 357 in the sunitinib arm. The primary analysis used the full analysis set, defined in the record as all randomized participants regardless of treatment actually received. The arm counts listed for the adverse-event results differ from the primary endpoint counts; those safety denominators are described separately in the serious adverse-event discussion.

The stated primary purpose was to demonstrate superiority of either lenvatinib combination over sunitinib for progression-free survival. Independent imaging review assessed progression using RECIST 1.1. The registry lists objective response rate, overall survival, adverse events, treatment discontinuation due to toxicity, time to treatment failure due to toxicity, and several health-related quality-of-life measures as secondary outcomes. It lists 13 outcome measures and two analyses. Further details about randomization procedures, allocation concealment, stratification factors, and the original statistical analysis plan are not reported in the registry record.

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: A total of 1417 participants were screened, of which, 1069 were randomized. Of the 1069 randomized, 1047 participants were treated in the study.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The FAS included all randomized participants regardless of the treatment actually received.

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: The full analysis set (FAS) included all randomized participants regardless of the treatment actually received.

6. Primary Endpoint

The registry lists the primary outcome as Progression-free Survival (PFS) by Independent Imaging Review (IIR), measured over From the date of randomization to the date of the first documentation of PD or date of death, whichever occurred first or up to data cutoff date 28 Aug 2020 (up to approximately 46 months).

Progression-free survival is a time-to-event endpoint. In this trial, its clock began at randomization and stopped at the first documentation of progressive disease or death, whichever occurred first. Disease progression was assessed by independent imaging review using RECIST 1.1. The registry defines progression as at least a 20 percent increase or 5 mm increase in the sum of target-lesion diameters, taking the smallest sum recorded since treatment started as the reference, or the appearance of one or more new lesions.

The endpoint captures how long participants remain alive without documented progression, combining two clinically important types of event into one measure. A time-to-event analysis can use information from participants who have not yet had an event by the data cutoff, subject to censoring rules. The detailed censoring rules are not reported in the registry record. The stated timeframe runs from randomization to progression or death, or to the data cutoff date of 28 Aug 2020, up to approximately 46 months.

The registry reports Kaplan-Meier estimates of median progression-free survival. A median gives a time point on the estimated event-free survival curve; it does not describe every participant’s experience or the distribution of all event times. The comparison also reports hazard ratios from stratified log-rank analyses, which summarize relative event rates over follow-up under the assumptions needed for that interpretation.

7. Results Posted in the Registry

• Lenvatinib 18 mg Plus Everolimus 5 mg vs Sunitinib 50 mg: hazard ratio 0.65 (95% CI 0.53–0.80), reported p < 0.001 (as reported in the registry).
• Lenvatinib 20 mg Plus Pembrolizumab 200 mg vs Sunitinib 50 mg: hazard ratio 0.39 (95% CI 0.32–0.49), reported p < 0.001 (as reported in the registry).

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a time-to-event measure, and the registry names two methods: Kaplan-Meier estimation for progression-free survival and a stratified log-rank test for each comparison. The reported effect measure is a stratified hazard ratio. These methods answer related but distinct questions: Kaplan-Meier estimation describes the event-free experience over time, while the hazard ratio and log-rank test compare the randomized groups.

What does Kaplan-Meier estimation do?

In general, Kaplan-Meier estimation constructs an estimate of the probability of remaining event-free over time. Participants contribute information up to their event or the point at which their follow-up is censored. Censoring means that an event was not observed during the participant’s available follow-up; it does not itself mean that the event did or did not occur later. Standard interpretation relies on assumptions about censoring and follow-up. The record states that Kaplan-Meier methods were used but does not report the detailed censoring rules or diagnostic checks.

The median progression-free survival is read from the estimated curve at the time when the estimated event-free proportion reaches one half. It is a compact summary, not a comparison test, and it can hide differences in the curve’s shape. The registry reports confidence intervals around the arm medians, giving a range of uncertainty for each estimated median. The separate comparison tests use the event-time information rather than comparing the median values by simple subtraction.

What does the stratified log-rank test compare?

In general, a log-rank test compares the event-time patterns between groups across follow-up. A stratified version performs that comparison while accounting for prespecified strata, when those strata are part of the design and analysis. The registry names a stratified log-rank test but does not identify the stratification factors. Its reported p-values are each less than < 0.001. The p-value is evidence against the test’s null hypothesis under its assumptions; it is not a measure of effect magnitude, clinical importance, or the probability that chance alone caused the results.

How should the hazard ratio be understood?

In general, a hazard is the rate at which an event occurs among participants who have not yet experienced it, at a given point in time. A hazard ratio compares those rates between groups. A ratio of 1 corresponds to equal estimated hazards; a value below 1 indicates a lower estimated hazard in the first-listed group under the usual orientation of the comparison. The registry reports 0.65 for lenvatinib plus everolimus versus sunitinib and 0.39 for lenvatinib plus pembrolizumab versus sunitinib.

A hazard ratio is not a ratio of median survival times, nor does it directly state how many months an individual might gain. Its usual single-number interpretation is clearest when the relative hazards are reasonably stable over time. The proportional-hazards assumption, any tests or plots assessing it, and any alternative analyses are not reported in the registry record. The posted summary does not permit those features to be assessed independently.

What do the confidence intervals and p-values add?

A confidence interval expresses uncertainty around an estimated quantity under the statistical model and procedure used. For the hazard ratios, the intervals describe a range of values compatible with the data and method at the stated 95% confidence level. They provide more information about precision and plausible effect sizes than a p-value alone. A p-value does not tell whether an estimate is precise, and a confidence interval does not resolve questions about bias, endpoint measurement, or unreported analysis choices.

The registry reports the hazard ratios and 95% confidence intervals as sponsor analyses using stratified log-rank tests. The median estimates are reported with 95% confidence intervals, but the method used to calculate those median intervals is not stated in the record. No recalculated hazard-ratio analysis is used in the interpretation; the sponsor’s reported estimates are the relevant comparison results. The sample-size planning illustrations are a separate calculation and do not replace the trial’s reported outcome analysis.

What analysis details remain unknown?

The record identifies the full analysis set as all randomized participants regardless of treatment actually received. It does not provide detailed rules for handling missing endpoint assessments, loss to follow-up, treatment switching, or censoring. It also does not report the strata, proportional-hazards diagnostics, interim analyses, or a complete statistical analysis plan. These omissions limit the ability to reproduce the analysis or assess every modeling assumption from the registry record alone. They do not change the fact that the registry reports the named endpoint, methods, estimates, intervals, and p-values.

10. Confidence Intervals and Precision

The arm-specific median estimates have 95% confidence intervals. For lenvatinib plus everolimus, the median is 14.7 months, with an interval from 11.1 to 16.7 months. For lenvatinib plus pembrolizumab, the median is 23.9 months, with an interval from 20.8 to 27.7 months. For sunitinib, the median is 9.2 months, with an interval from 6.0 to 11.0 months. Each interval describes uncertainty in its own arm’s median estimate; comparing overlap between these intervals is not a substitute for the reported treatment comparison.

The hazard-ratio interval for lenvatinib plus everolimus versus sunitinib is 0.53 to 0.80 around an estimate of 0.65. For lenvatinib plus pembrolizumab versus sunitinib, it is 0.32 to 0.49 around an estimate of 0.39. Both are reported as 95% confidence intervals, and both lie below 1. They indicate uncertainty around the relative hazard estimate, conditional on the analysis method and its assumptions. They do not capture all uncertainty arising from potential bias, measurement, missingness, or analysis choices that are not described.

Precision should not be confused with importance or certainty about an individual outcome. The confidence intervals convey sampling uncertainty, not a guarantee that the true value lies within a particular interval for this single trial. The record does not report the primary endpoint event counts, detailed follow-up distribution, or interval estimation procedures for the medians, so additional precision diagnostics cannot be assessed from the registry summary.

11. 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 figures are event targets, not participant counts per arm. For the time-to-event calculations, each value is the total number of events across the two comparison groups. The rows consider planning for 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. The calculations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the event requirements use the Schoenfeld method. The effect to detect is the stated fraction of the observed effect.

The 80% power column gives the total events targeted under those assumptions for a design with 80% power to detect the specified effect. The 90% power column gives the corresponding event target for 90% power. In general, power describes the probability of detecting an effect of the specified size if that effect is present and the assumptions used in the calculation hold. These are planning illustrations based on the observed effect, not observed power for the completed trial.

For both comparisons, the scenarios based on smaller fractions of the observed effect require more total events than scenarios based on the full observed effect. This reflects a general planning principle: smaller effects are harder to distinguish from random variation and therefore require more information. These calculations do not establish the original trial’s planned sample size or prove that the trial achieved a particular power. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

12. 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 = 0.650, so ln(HR) = -0.4308.
2. d = 4 × (1.96 + 0.842)² / [ln(0.650)]² = 169.2, rounded up to 170 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 <- 0.65                              # 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(2811861)
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 = 170 (80% power), 227 (90% power)
 75% of observed effect: events = 301 (80% power), 403 (90% power)
 50% of observed effect: events = 677 (80% power), 906 (90% power)
Simulated power with 170 events: 0.806 (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.

13. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 278 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 1,069 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 9.2          # control-arm median, months (converted from the posted unit)
hr          <- 0.65          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
    Analysis               Value Efficacy
   IA 1: 33%                   Z   3.7103
      N: 206         p (1-sided)   0.0001
  Events: 76        ~HR at bound   0.4268
   Month: 18    P(Cross) if HR=1   0.0001
             P(Cross) if HR=0.65   0.0338
   IA 2: 67%                   Z   2.5114
      N: 278         p (1-sided)   0.0060
 Events: 152        ~HR at bound   0.6653
   Month: 28    P(Cross) if HR=1   0.0060
             P(Cross) if HR=0.65   0.5603
       Final                   Z   1.9930
      N: 278         p (1-sided)   0.0231
 Events: 228        ~HR at bound   0.7679
   Month: 60    P(Cross) if HR=1   0.0250
             P(Cross) if HR=0.65   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

14. Analyses Reported by the Sponsor

The registry reports the primary comparisons as stratified hazard ratios from stratified log-rank tests. For lenvatinib plus everolimus versus sunitinib, the sponsor-reported hazard ratio is 0.65, with a 95% confidence interval of 0.53 to 0.80 and p < 0.001. For lenvatinib plus pembrolizumab versus sunitinib, it is 0.39, with a 95% confidence interval of 0.32 to 0.49 and p < 0.001.

No discrepancy is flagged between the registry’s reported comparisons and the analysis summary used for this interpretation. The reported hazard ratios are therefore the results to use; no alternate hazard-ratio recalculation is substituted. The planning illustrations are distinct from these outcome analyses: they use observed effects to illustrate event requirements under stated planning assumptions and do not revise the sponsor’s trial estimates.

The registry does not give the full model specification, the stratification factors, or detailed sensitivity analyses. Consequently, the sponsor’s reported estimates can be described as posted, but their complete analytic implementation cannot be independently reproduced from the record.

15. Multiplicity, Missing Data and Analysis Populations

The registry reports two primary comparisons: lenvatinib plus everolimus versus sunitinib, and lenvatinib plus pembrolizumab versus sunitinib. Both are labeled superiority analyses and both have reported p-values of less than < 0.001. The registry record does not report whether or how the significance level was adjusted to account for testing more than one comparison, nor does it describe a hierarchy, gatekeeping procedure, or other multiplicity strategy. Without that information, the reported p-values should be read as posted test results rather than as a complete account of family-wise error control.

The registry lists multiple secondary outcomes, including overall survival, objective response rate, safety outcomes, treatment discontinuation and treatment failure due to toxicity, and quality-of-life measures. The analysis details for these outcomes are not described in the registry record for the primary endpoint. The record does not report which secondary analyses were confirmatory, whether secondary testing was adjusted for multiplicity, or whether outcomes were tested in a prespecified sequence.

The primary analysis population is the full analysis set: all randomized participants, regardless of the treatment actually received. This supports an analysis based on randomized assignment as defined in the record. Detailed handling of missing imaging assessments, missed visits, withdrawal, loss to follow-up, or treatment discontinuation is not reported. The endpoint definition gives the event criteria and follow-up cutoff, but not the operational censoring rules or sensitivity analyses for missing information.

Interim analyses, stopping rules, and the original statistical analysis plan are not reported in the registry record. Their absence from the record does not demonstrate that such procedures were or were not used; it means they cannot be characterized from the posted information. These unknowns matter when assessing how the two primary tests and the broader set of outcomes were planned and interpreted.

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: pneumonia (4.2%), diarrhoea (3.9%), vomiting (2.8%), acute kidney injury (2.8%), pneumonitis (2.6%).

The safety results report participants with at least one serious adverse event and participants with other treatment-emergent adverse events, alongside the number of deaths. In the lenvatinib-plus-everolimus group, 164 of 355 participants at risk had at least one serious adverse event, and 353 of 355 had another treatment-emergent adverse event; 119 deaths are listed. In the lenvatinib-plus-pembrolizumab group, the corresponding counts are 178 of 352 and 351 of 352, with 78 deaths listed. In the sunitinib group, they are 113 of 340 and 332 of 340, with 99 deaths listed.

These are counts from the posted safety results, not adjusted comparisons between randomized groups. The registry summary does not give the definitions, timing, causes, or treatment attribution for the listed deaths, nor does it provide a formal statistical comparison of these safety counts. The safety denominators differ from the primary endpoint arm counts, so the safety figures should be read using their own listed populations.

The registry lists pneumonia (4.2%), diarrhoea (3.9%), vomiting (2.8%), acute kidney injury (2.8%), pneumonitis (2.6%), and hypertension (2.3%) among the top serious adverse-event terms. The record does not specify which arm-specific denominator applies to each of these percentages in the summary. These terms and percentages complement, but do not replace, the overall serious adverse-event counts.

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

The posted progression-free survival results show longer estimated median times in each lenvatinib-combination arm than in the sunitinib arm. The reported stratified hazard ratios are below 1 for both comparisons, and their 95% confidence intervals also lie below 1. The registry reports p < 0.001 for each comparison and labels both analyses as superiority analyses. These are the trial’s reported statistical results for the specified endpoint and comparisons.

Those results do not establish a direct ranking between the two combination regimens. Each was compared separately with sunitinib; the registry does not report a statistical test comparing the combinations with each other. Nor does a median difference describe the experience of every participant or quantify an individual’s likely outcome. Hazard ratios summarize relative event hazards rather than an absolute number of months gained, and their interpretation depends on model assumptions that are not fully documented in the record.

The primary endpoint concerns documented progression or death, not every aspect of disease status, survival, symptoms, or quality of life. The registry lists overall survival, response, adverse events, and quality-of-life outcomes as secondary measures, but the primary endpoint results alone cannot answer those separate questions. Safety counts also require their own interpretation and do not establish comparative risks without fuller definitions and analysis.

These findings are bounded by the population, treatments, endpoint definition, follow-up period, and analysis recorded for the trial. Details about missing data, censoring, multiplicity, stratification factors, and the original statistical plan are not reported in the registry record. The results should therefore be understood as the posted estimates and tests, not as a complete account of every uncertainty or as medical advice.

18. Why This Trial Matters Statistically

This trial illustrates why time-to-event results are often reported with more than one statistic. Median progression-free survival gives a time-based summary within each arm, while a hazard ratio compares event rates over follow-up. Confidence intervals add information about uncertainty around both kinds of estimate, and the log-rank p-values address evidence against the comparison’s null hypothesis. These quantities are related but are not interchangeable.

The design also illustrates the importance of keeping treatment comparisons distinct. Two regimens were each compared with the same active comparator, but that does not create a direct randomized comparison between the regimens. Reading the two hazard ratios side by side may invite a ranking, yet the registry does not report a test of that question. A clear distinction between the prespecified comparisons and untested contrasts helps prevent overinterpretation.

Finally, the record shows why statistical conclusions depend on more than a small p-value. The listed estimates can be interpreted alongside their confidence intervals, but important analysis details—including multiplicity handling, missing-data procedures, and stratification factors—are not reported. The planning illustrations also concern hypothetical event targets under stated assumptions, not the trial’s observed power or original design rationale. Keeping these categories separate supports a more careful reading of the evidence.

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 registry summary leaves several trial-specific details unresolved. It names a stratified log-rank test but does not identify the stratification factors or describe the proportional-hazards assessment. It reports Kaplan-Meier medians but does not state the method used to construct their confidence intervals. Detailed endpoint censoring rules, imaging schedule, missing-assessment procedures, and sensitivity analyses are not reported in the registry record. These omissions limit independent reproduction and assessment of the analysis.

The record reports two superiority comparisons but does not state how multiplicity across the primary comparisons was handled. It also does not report interim-analysis procedures or the complete original statistical analysis plan. The reported p-values and confidence intervals can be stated as posted, while the implications of unreported analysis choices remain uncertain.

The sunitinib dosing description contains an internal schedule inconsistency: it states 4 weeks on treatment followed by 2 weeks off, and also places that schedule in each 21-day cycle. The record does not resolve this. The safety results use denominators different from those listed for the primary endpoint, and the summary does not explain the difference. Details on the listed deaths, including cause and attribution, are not reported. The serious adverse-event term percentages are presented without arm-specific denominators in the summarized record.

Finally, the registry’s results are tied to independent imaging review of progression using RECIST 1.1 and to the stated data cutoff. The record does not give complete participant characteristics or detailed follow-up distributions in the registry record used for these results, limiting assessment of generalizability and the amount of follow-up contributing to each estimate.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Lenvatinib/Everolimus or Lenvatinib/Pembrolizumab Versus Sunitinib Alone as Treatment of Advanced Renal Cell Carcinoma — NCT02811861.
• Registry record last updated 2026-02-23; status last verified 2026-02.
• Results first posted 2021-09-24.
• 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

NCT02811861 is a randomized, open-label, parallel-group phase 3 trial of lenvatinib plus everolimus, lenvatinib plus pembrolizumab, and sunitinib in first-line advanced renal cell carcinoma. The registry reports 1069 enrolled participants and identifies progression-free survival by independent imaging review as the primary endpoint. The primary analysis population includes all randomized participants regardless of treatment actually received.

The posted median progression-free survival estimates are 14.7 months for lenvatinib plus everolimus, 23.9 months for lenvatinib plus pembrolizumab, and 9.2 months for sunitinib. The reported hazard ratios versus sunitinib are 0.65 and 0.39, respectively, with 95% confidence intervals below 1 and p < 0.001 for each comparison. These results describe the two separate comparisons; a direct comparison of the combination regimens is not reported.

The registry also lists secondary efficacy, safety, toxicity, and quality-of-life outcomes. Several details needed for a full assessment of analysis assumptions and design—including multiplicity handling, missing-data procedures, and the original sample-size justification—are not reported in the registry record.