A Study Comparing Nivolumab, Nivolumab in Combination With Ipilimumab and Placebo in Participants With Localized Kidney Cancer Who Underwent Surgery to Remove Part of a Kidney: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Comparing Nivolumab, Nivolumab in Combination With Ipilimumab and Placebo in Participants With Localized Kidney Cancer Who Underwent Surgery to Remove Part of a Kidney (NCT03138512), a randomized phase 3 study. The primary outcome is Disease-Free Survival (DFS) by BICR - Treatment Part A and B.

2. Clinical Question

3. Trial Design

This was a phase 3 randomized, double-blind, quadruple-masked, parallel-group study. The registry lists enrollment of 1641 participants and describes the trial as completed. The recorded start date is 2017-07-07, and primary completion is listed as 2023-09-28. The registry does not report the allocation ratio, the method used to generate the randomization sequence, or the specific people covered by quadruple masking.

The study addressed localized renal cell carcinoma after radical or partial nephrectomy, in participants described as being at high risk of relapse. Treatment Part A compared nivolumab plus ipilimumab with placebo. Treatment Part B included nivolumab plus ipilimumab, nivolumab alone, and placebo. The primary endpoint population was defined differently from the full set of treatment groups: it included Treatment Part A and specified placebo and nivolumab groups in Part B, while the Part B combination group was prespecified to be excluded from the endpoint objective. Group counts in the primary endpoint listing include 405 for Treatment Part A combination, 411 for Treatment Part A placebo, 208 for Treatment Part B placebo, and 411 for Treatment Part B nivolumab. The listed Part B combination count for that endpoint is 0.

The primary outcome was disease-free survival assessed by blinded independent central review. The registry describes the observation period as from randomization until local recurrence, metastasis, or death, whichever came first, up to approximately 72 months. Listed secondary outcomes include overall survival, a five-year overall survival rate, disease-free survival in contemporaneously randomized Part B combination and monotherapy participants, and adverse events. The registry record does not report the original sample-size justification, detailed treatment schedules, or a full protocol-level account of the 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: Part A and B participants are separately randomized and treated.

Randomization

Reasons for not completing this period:

Treatment Part A

Reasons for not completing this period:

Treatment Part B

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants in Treatment Part A (Nivo + Ipi and Placebo Arms) and Treatment Part B (Placebo and Nivo Arms). The Nivo + Ipi Arm was prespecified to be excluded from the endpoint objective for Treatment Part B.

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 Disease-Free Survival (DFS) by BICR - Treatment Part A and B, measured over From randomization to development of local disease recurrence, distance metastasis, or death, whichever came first (up to approximately 72 months).

Disease-free survival was the primary endpoint. It measured the time from randomization to the first of local disease recurrence, metastasis, or death, according to blinded independent central review. Local recurrence included recurrence of the primary tumor in situ or occurrence of a secondary renal cell carcinoma primary cancer. The recorded timeframe extends up to approximately 72 months.

A time-to-event endpoint uses both when an event occurs and the fact that some participants may not have experienced the event by the end of their observed follow-up. This makes it different from a simple count of events at one fixed visit. General explanation: time-to-event analyses can make use of follow-up accumulated over time, while accounting for participants whose event time is not observed during the study period. The registry identifies Kaplan-Meier estimates for the endpoint and reports hazard ratios from Cox proportional-hazards analyses for the primary comparisons.

Disease-free survival is used to summarize recurrence or death after surgery in a single endpoint, with the first qualifying event determining the recorded time. It therefore addresses how long participants remain free of the specified events, rather than measuring tumor response at a single assessment. The endpoint does not tell readers which component event drove an estimate unless those components are separately reported. The registry record does not report median disease-free survival values for the listed primary groups; it gives incomplete median information, with only a lower confidence bound recorded for one listed group. That makes the hazard-ratio comparisons more interpretable than a direct comparison of reported medians, which are not available for these groups.

7. Results Posted in the Registry

• Treatment Part A: Nivo + Ipi vs Treatment Part A: Placebo: hazard ratio 0.95 (95% CI 0.75–1.20), reported p = 0.668 (as reported in the registry).
• Treatment Part A: Placebo vs Treatment Part B: Nivo: hazard ratio 0.93 (95% CI 0.67–1.28), reported p = 0.656 (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.

Overall Survival (OS) - Treatment Part A and B

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Treatment Part A: Nivo + Ipi vs Treatment Part A: Placebo: hazard ratio 1.26 (95% CI 0.85–1.85), reported p = 0.244 (as reported in the registry).
  • Treatment Part A: Placebo vs Treatment Part B: Nivo: hazard ratio 1.36 (95% CI 0.61–3.07), reported p = 0.450 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Treatment Part A: Nivo + Ipi vs Treatment Part A: Placebo and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Treatment Part A: Placebo vs Treatment Part B: Nivo and the control. The interval gives the range of effects that remain compatible with the data.

Overall Survival (OS) Rate (5 Years) - Treatment Part A and B

• 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:
  • Treatment Part A: Nivo + Ipi: 84.9% (344/405) versus 87.1% (358/411) on control; risk difference -2.2 percentage points (95% CI -6.9 to 2.6), risk ratio 0.98 (0.92–1.03); unadjusted two-proportion z-test p = 0.372.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Treatment Part A: Nivo + Ipi and the control. The interval gives the range of effects that remain compatible with the data.

Disease-Free Survival (DFS) Per BICR in Contemporaneously Randomized Combination and Monotherapy Participants - Treatment Part B

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Treatment Part B: Nivo + Ipi vs Treatment Part B: Nivo: hazard ratio 1.22 (95% CI 0.89–1.67) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Treatment Part B: Nivo + Ipi vs Treatment Part B: Nivo and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the record beyond recurrence-related disease-free survival to include overall survival, a five-year overall survival rate, and safety outcomes. A separate Part B analysis compares disease-free survival in contemporaneously randomized combination and monotherapy participants. These outcomes address related but distinct questions and do not simply replicate the primary endpoint.

Across the posted secondary time-to-event comparisons, the point estimates lie on either side of the neutral hazard-ratio value, and their confidence intervals include 1. The five-year overall survival estimates are reported for the Treatment Part A arms, with confidence intervals, while no such values are listed for the Part B groups. A p-value is not reported for the Part B disease-free survival comparison. Taken together, the secondary findings do not establish a uniform direction across every outcome or treatment contrast. Their uncertainty and the differing populations and outcome definitions make a single summary claim inappropriate.

Secondary endpoints help describe whether patterns are consistent across outcomes, but they usually carry less confirmatory weight when many outcomes and comparisons are assessed. General explanation: testing multiple hypotheses increases the chance that at least one result appears unusual by chance if no adjustment or prespecified hierarchy is used. The registry record does not report a multiplicity procedure or the full hierarchy governing secondary endpoint interpretation. Thus, secondary estimates can be informative and hypothesis-generating, but they should not be treated as independent confirmation of the primary result without knowing how multiplicity was addressed.

Overall survival is clinically distinct from disease-free survival because it concerns death rather than the first of recurrence, metastasis, or death. The five-year rate is a fixed-time summary, whereas the hazard-ratio analyses use event timing over follow-up. Adverse-event outcomes address safety rather than efficacy. Differences in outcome construction, follow-up, and analysis populations mean that results across these endpoints should be compared cautiously rather than combined as if they measured the same effect.

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

• Overall Survival (OS) - Treatment Part A and B
• Overall Survival (OS) Rate (5 Years) - Treatment Part A and B
• Disease-Free Survival (DFS) Per BICR in Contemporaneously Randomized Combination and Monotherapy Participants - Treatment Part B
• Overall Survival (OS) in the Contemporaneously Randomized Combination and Monotherapy Participants - Treatment Part B
• The Number of Participants With Adverse Events up to 30 Days After Last Dose of Study Therapy - Treatment Part A
• The Number of Participants With Adverse Events up to 30 Days After Last Dose of Study Therapy - Treatment Part B
• The Number of Participants With Adverse Events up to 100 Days After Last Dose of Study Therapy - Treatment Part A
• The Number of Participants With Adverse Events up to 100 Days After Last Dose of Study Therapy - Treatment Part B

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is a time-to-event measure, and the registry identifies Kaplan-Meier estimates and Cox proportional-hazards analyses. The principal reported comparisons are hazard ratios with 95% confidence intervals and p-values. Those statistics answer related but distinct questions: the hazard ratio gives a relative summary of event rates, the interval describes uncertainty around that estimate, and the p-value assesses compatibility with a null hypothesis under the analysis model.

What does a hazard ratio represent?

General explanation: a hazard is the instantaneous rate of experiencing the event among participants who have not yet experienced it, at a particular point in follow-up. A hazard ratio compares that rate between two groups over time. A value of 1 represents equal hazards under the comparison; a value below 1 indicates a lower estimated hazard in the numerator group, and a value above 1 indicates a higher estimated hazard. The registry labels the reported estimates as Cox proportional-hazards results and gives the group order for each comparison, so direction must be interpreted in that stated order.

A hazard ratio is not a risk ratio at a fixed time, a difference in the number of events, or a difference in median survival time. Its interpretation as one relative summary over follow-up relies on the proportional-hazards model being suitable. The registry record does not report a test or graphical assessment of the proportional-hazards assumption, nor does it provide model diagnostics. That absence does not demonstrate that the assumption failed; it means the record does not document how it was assessed.

How do Kaplan-Meier estimates fit?

General explanation: the Kaplan-Meier method estimates the probability of remaining event-free over time while retaining information from participants whose follow-up ends before an event is observed. Such participants are generally censored at their last known event-free time, subject to the analysis rules. The registry describes the primary endpoint as based on Kaplan-Meier estimates, but details of censoring rules, follow-up completeness, and handling of missing assessments are not reported in the registry record. These details matter because a time-to-event estimate depends on how observation ends and on assumptions about participants no longer followed.

The record lists median disease-free survival as the endpoint parameter, but the median values are not available for most groups. In a survival curve, the median is the time at which the estimated event-free proportion reaches one half, if that point is reached during observed follow-up. General explanation: if the estimated curve does not reach that point, the median may not be estimable within the available follow-up. The listed values do not permit a complete median comparison, so the hazard ratios and their intervals are the available comparative summaries.

What do the confidence intervals and p-values add?

General explanation: a confidence interval gives a range of estimates compatible with the data under the statistical method, at a stated confidence level. It communicates precision as well as direction. An interval that includes 1 for a hazard ratio includes the neutral relative-hazard value, but this does not mean every value in the interval is equally likely or that the groups are equivalent. A p-value is not an effect size and is not a measure of the probability that the null hypothesis is true.

The primary comparisons use 95% confidence intervals and log-rank testing is named as the method in the registry comparison entries. The reported analyses also identify Cox proportional-hazards parameters. The registry does not give further details of test implementation, stratification factors, covariate adjustment, or the exact relationship between the listed log-rank method and the Cox model output. Those details are not reported in the registry record and should not be inferred from the labels alone.

Why can over-dispersion matter?

General explanation: over-dispersion arises when observed variation is greater than a statistical model assumes. It is especially relevant for count outcomes modeled with a distribution that expects variance to be tied closely to the mean. If over-dispersion is ignored in such a model, standard errors and confidence intervals can be too small, making evidence appear more precise than it is. The primary endpoint is time-to-event, and the registry does not identify an over-dispersion model for its analysis. The concept is useful generally, but it should not be treated as a reported feature of this trial's primary analysis.

The registry's analysis population is described for the primary endpoint: randomized Treatment Part A participants and the designated Treatment Part B placebo and nivolumab participants, with the Part B combination group excluded from the endpoint objective. The record does not report the analysis-set rules in greater detail, such as how protocol deviations or missing endpoint assessments were handled. It also does not report interim analyses, the stopping rule, or a complete prespecified statistical analysis plan. These omissions limit independent assessment of the analysis decisions, but do not justify substituting assumptions about what the protocol may have specified.

11. Confidence Intervals and Precision

The primary disease-free survival estimates have confidence intervals that include the neutral hazard-ratio value of 1. For Treatment Part A combination versus placebo, the hazard ratio is 0.95 with a 95% confidence interval from 0.75 to 1.20. For Treatment Part A placebo versus Treatment Part B nivolumab, the hazard ratio is 0.93 with a 95% confidence interval from 0.67 to 1.28. The intervals extend below and above 1, so the reported data remain compatible with different relative-hazard directions. The second interval is wider in absolute range, indicating less precision in that comparison's relative-hazard estimate.

General explanation: confidence intervals express statistical uncertainty around an estimate under the model and analysis used. They are not guarantees that the true effect lies in the interval, and they do not account automatically for every source of uncertainty, such as model misspecification, selective reporting, or multiple testing. A confidence interval containing the neutral value does not demonstrate no effect. It indicates that the data do not rule out that value at the stated confidence level, while also allowing other values within the interval.

Precision is constrained by the number and timing of observed events, not enrollment alone. The registry lists enrollment of 1641, but does not provide the number of primary endpoint events for the reported comparisons in the registry record in this record. The median disease-free survival values are also missing for most groups. As a result, readers cannot use those figures to independently assess event-based information or compare median times across all groups. Follow-up is described as up to approximately 72 months, but the record does not give a detailed distribution of follow-up or completeness.

The secondary time-to-event estimates also have intervals spanning 1, and the Part B contemporaneous disease-free survival comparison has no p-value reported. These features reinforce that precision varies by outcome and comparison. A point estimate should not be treated as a precise statement when its interval is broad or includes materially different effects. Statistical uncertainty is distinct from clinical importance, and the registry does not define a threshold for a clinically important difference.

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 figures illustrate how much information would be needed under stated assumptions to detect a fraction of the observed effect. For time-to-event comparisons, the figures are total events across both groups, not participants per arm. For other outcomes, the figures are participants per arm. The effect scenarios are labeled as 100%, 75%, and 50% of the observed effect. Smaller effects require more information to detect at the same target power, which is why the planning requirements increase as the assumed effect becomes smaller.

The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target is the stated fraction of the observed effect. For time-to-event outcomes, the total event requirements use the Schoenfeld approach. For the other outcomes, the counts are participants per arm. The registry's original trial-design sample-size assumptions and justification are not reported in the registry record; these planning values are illustrations based on the observed effect, not a description of the trial's original design.

The 80% power column means the hypothetical design would have an 80% chance of detecting the specified effect under the assumptions if that effect were the true one. The 90% power column has the corresponding 90% target. These are design targets, not probabilities that the reported finding is true and not a measure of the probability that the trial succeeded. The figures are not observed power. Observed power is not a useful substitute for reading the estimate and its confidence interval after results are available. The actual trial enrollment and the planning counts answer different questions: one is the number enrolled, while the other is an illustrative requirement for a hypothetical comparison under specified assumptions.

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 = 0.950, so ln(HR) = -0.0513.
2. d = 4 × (1.96 + 0.842)² / [ln(0.950)]² = 11,932.9, rounded up to 11,933 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 <- 0.95                              # 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(3138512)
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 = 11,933 (80% power), 15,975 (90% power)
 75% of observed effect: events = 21,215 (80% power), 28,400 (90% power)
 50% of observed effect: events = 47,732 (80% power), 63,900 (90% power)
Simulated power with 11,933 events: 0.802 (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 registry reports the sponsor's primary comparisons as Cox proportional-hazards estimates, with log-rank testing identified in the comparison entries. The reported hazard ratios, 95% confidence intervals, and p-values are the analyses to use when describing the sponsor's results. The comparison records do not flag a discrepancy between those analyses and any simple unadjusted recalculation, so there is no competing recalculated result to privilege over the sponsor's report.

A simple recalculation from posted summary statistics, when available, is not a substitute for the sponsor's analysis. Such a calculation generally cannot reproduce individual-participant time-to-event modeling, the precise analysis population, censoring rules, stratification, covariate adjustment, or other protocol-defined details. It is best understood as a check on a limited summary-statistic calculation, rather than an alternative primary analysis. If a discrepancy were present, the sponsor's own analysis would take precedence in reporting the trial result; no discrepancy is flagged in this record.

The registry does not report the complete statistical analysis plan, detailed model specifications, or all diagnostics. Readers can therefore describe the estimates as reported without claiming to have independently reproduced the full analysis. The p-values and intervals should be read together: neither alone supplies a complete account of effect direction, magnitude, and precision.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies a primary endpoint and lists secondary endpoints, but it does not report a complete statistical analysis plan or a multiplicity-adjustment procedure. It is therefore not possible to determine from the registry record whether a formal hierarchy, alpha allocation, or another method was used to control the chance of false-positive findings across the set of comparisons. General explanation: when multiple hypotheses are tested, the chance of at least one apparently positive result can rise unless the analysis plan accounts for that multiplicity. This matters especially when secondary outcomes are interpreted as confirmatory evidence.

The record describes the primary population as randomized participants in Treatment Part A and designated placebo and nivolumab groups in Treatment Part B, excluding the Part B combination group from the endpoint objective. It does not report further analysis-population definitions, such as whether the primary analysis used intention-to-treat principles in every detail, how protocol deviations were handled, or whether any per-protocol or other sensitivity populations were analyzed. The population description should therefore be kept to what is stated rather than expanded by assumption.

Missing-data and censoring procedures are not reported in the registry record. For a time-to-event outcome, the rules for censoring participants who have no observed event by their last follow-up can affect estimates. The registry does not give enough information about lost follow-up, withdrawal, missed assessments, or sensitivity analyses to evaluate how robust the results are to those issues. It also does not report an approach for imputing missing data for the fixed-time overall survival rate.

Interim analyses, any data-monitoring boundaries, and stopping rules are not reported in the registry record. This does not establish that they were absent; it means the posted record does not document them. Likewise, the record does not provide enough information to assess whether all outcomes were analyzed exactly as initially planned or whether changes were made during the trial. These reporting gaps limit independent review of multiplicity, missingness, and analysis populations. They do not justify guessing what procedures the protocol may have specified.

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: adrenal insufficiency (3.5%), diarrhoea (2.6%), acute kidney injury (2.1%), hypophysitis (2.0%), colitis (2.0%).

The safety listing gives counts of participants affected by serious adverse events, other adverse events, and deaths for aggregated treatment groups. The serious-event counts are 196 among 608 in the combined Treatment Part A and B nivolumab-plus-ipilimumab group, 49 among 614 in the combined placebo group, and 56 among 408 in the Treatment Part B nivolumab group. The corresponding counts listed for other adverse events are 544 among 608, 469 among 614, and 325 among 408. Death counts are 65, 55, and 22, respectively. These are reported counts with their listed group denominators; no derived rates are calculated.

The registry also lists serious adverse-event terms including adrenal insufficiency at 3.5%, diarrhoea at 2.6%, acute kidney injury at 2.1%, hypophysitis at 2.0%, colitis at 2.0%, and pneumonitis at 1.2%. These percentages are registry entries and should be interpreted in the context of their event definitions and the safety population. The record lists adverse-event outcomes with windows extending up to 30 days and up to 100 days after the last dose, but it does not clarify in this summary how each aggregated safety count maps to those windows. Group sizes differ, follow-up may differ, and the figures alone do not establish causation.

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

The posted primary estimates describe disease-free survival comparisons in the populations defined by the registry. Both primary hazard-ratio confidence intervals include 1, and the point estimates are close to that neutral value. This means the results leave uncertainty about the direction and magnitude of the relative hazard. They do not establish that the treatments are equivalent, and they do not establish that the true effect is exactly zero.

The result also does not answer every question about recurrence after surgery. The endpoint combines local recurrence, metastasis, and death, and the posted summary does not separate the contribution of each component. Median disease-free survival values are not reported for most groups, so the available record does not support a complete comparison in median months. Nor does a hazard ratio directly tell a reader the absolute probability of an event by a particular time.

The findings apply to the recorded trial population and treatment contrasts, not automatically to every person with renal cell carcinoma or every treatment setting. The study involved participants described as having localized disease after radical or partial nephrectomy and at high risk of relapse. The primary endpoint objective excluded the Treatment Part B combination group. Comparisons that cross Treatment Parts should be interpreted with attention to their distinct structure, rather than treated as though every contrast were drawn from one contemporaneous randomization.

Secondary results and safety counts add context, but do not erase uncertainty in the primary estimates. Multiple secondary endpoints cover different questions, and the registry does not report how multiplicity was handled. Safety counts describe reported events in listed groups; they are not, by themselves, proof that a particular treatment caused an event. The record also lacks detail on missing data, censoring, and some analysis decisions. Accordingly, these results support a description of the posted estimates and their limits, not medical advice or a broad claim that a treatment works or does not work.

18. Why This Trial Matters Statistically

This trial is statistically instructive because it combines time-to-event outcomes with a complex treatment structure. The primary endpoint uses time from randomization to the first of recurrence, metastasis, or death, rather than a single fixed-time measurement. Interpreting its hazard ratios requires attention to event timing, follow-up, censoring, and the model used. The registry identifies Kaplan-Meier and Cox methods, while leaving some implementation details unreported.

The population definition also matters. Treatment Part A and Treatment Part B do not contribute identically to the primary endpoint objective, and the Part B combination group was prespecified for exclusion from that objective. This highlights why the endpoint population must be read alongside the overall enrollment figure and treatment-arm structure. Enrollment alone does not tell a reader which participants contributed to each comparison.

The confidence intervals illustrate the distinction between a point estimate and precision. Estimates near the neutral hazard ratio can still have intervals spanning effects in either direction. A non-small p-value does not establish equivalence, and an observed estimate should not be converted into a categorical conclusion without considering the interval and study design. Secondary results raise the further issue of how multiple comparisons are interpreted; the registry does not document a multiplicity procedure.

Finally, the record makes clear that a reported result is only as interpretable as its accompanying documentation. Missing median values, limited reporting on analysis assumptions, and absent details about missing-data and censoring procedures restrict independent assessment. These are not reasons to dismiss the trial, but they are reasons to state precisely what the registry reports and to distinguish reported evidence from unreported design or analysis details.

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.

Several limitations are specific to the posted record and trial structure. The primary endpoint is defined for selected groups across Treatment Parts, and the Part B combination group is excluded from the primary endpoint objective. The registry comparison of Treatment Part A placebo with Treatment Part B nivolumab spans treatment parts; the record does not give enough detail to treat it as a contemporaneous randomized comparison or to explain the exact design relationship between those groups.

Median disease-free survival values are missing for most primary groups. The record therefore does not support a complete comparison of median times, and the listed confidence-bound information alone cannot fill that gap. The primary outcome is a composite of local recurrence, metastasis, and death, but separate component results are not included in the endpoint details summarized in the registry record. The timeframe extends up to approximately 72 months, yet detailed follow-up distributions and completeness are not reported.

The registry does not report the original sample-size justification, detailed protocol assumptions, full statistical analysis plan, multiplicity procedure, interim analyses, missing-data rules, or censoring details. It also does not provide model diagnostics or a reported assessment of the proportional-hazards assumption. These limitations affect how fully the analyses can be independently evaluated; they should not be used to infer that any particular procedure was or was not performed.

The safety results are aggregated by broad treatment groups. The posted counts do not provide a complete account of exposure duration, event timing, severity grading, or the relation of each count to the listed 30-day and 100-day safety windows. Differences in group denominators also mean that counts should not be compared as if they were rates. Finally, the listed secondary outcomes differ in population and measurement, and the registry does not state how the family of secondary tests was controlled statistically.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study Comparing Nivolumab, Nivolumab in Combination With Ipilimumab and Placebo in Participants With Localized Kidney Cancer Who Underwent Surgery to Remove Part of a Kidney — NCT03138512.
• Registry record last updated 2024-12-18; status last verified 2024-11.
• Results first posted 2024-12-18.
• 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

The registry describes a completed randomized, double-blind, quadruple-masked, parallel-group phase 3 study enrolling 1641 participants with localized renal cell carcinoma after radical or partial nephrectomy. It compared nivolumab, nivolumab plus ipilimumab, and placebo across Treatment Parts A and B. The primary endpoint was disease-free survival by blinded independent central review, defined by time to local recurrence, metastasis, or death.

The reported primary hazard ratios are 0.95 for Treatment Part A combination versus placebo and 0.93 for Treatment Part A placebo versus Treatment Part B nivolumab; both 95% confidence intervals include 1. Median disease-free survival values are not reported for most groups. Secondary outcomes include overall survival, a five-year overall survival rate, a Part B disease-free survival comparison, and adverse events. Further details on multiplicity, missing data, the original sample-size justification, and several analysis procedures are not reported in the registry record.