An Investigational Immuno-therapy Study of Nivolumab, or Nivolumab in Combination With Ipilimumab, or Placebo in Patients With Extensive-Stage Disease Small Cell Lung Cancer (ED-SCLC) After Completion of Platinum-based Chemotherapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for An Investigational Immuno-therapy Study of Nivolumab, or Nivolumab in Combination With Ipilimumab, or Placebo in Patients With Extensive-Stage Disease Small Cell Lung Cancer (ED-SCLC) After Completion of Platinum-based Chemotherapy (NCT02538666), a randomized phase 3 study. The primary outcome is Overall Survival (OS) of Nivolumab + Ipilimumab Versus Placebo In The Global Population.

2. Clinical Question

3. Trial Design

This was a randomized, multicenter, double-blind, Phase 3 study with a parallel-group design. The registry records quadruple masking. The trial evaluated maintenance therapy after completion of first-line platinum-based chemotherapy in participants with extensive-stage disease small cell lung cancer. Overall enrollment was 907.

The primary outcome compared the global nivolumab plus ipilimumab group with the global placebo group. The study also included nivolumab monotherapy, and the posted secondary outcomes include comparisons involving that arm. The registry describes the combination regimen as nivolumab 1 mg/kg plus ipilimumab 3 mg/kg and the monotherapy regimen as nivolumab 240 mg in the posted outcome groups. Placebo was the control for the primary comparison.

The recorded study start was 2015-10-13, and primary completion was 2018-10-01. The study status is completed. These registry entries describe the design and conduct at a high level, but they do not establish how randomization was generated, which factors were used for stratification, or how the masking was maintained; those details are not reported in the registry record. The trial's original sample-size justification and detailed statistical analysis plan are also 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: 907 participants were randomized and 903 were treated. 2 participants were in both the Global and China population.

Pre-Treatment

Reasons for not completing this period:

Treatment

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants in the nivolumab + ipilimumab and placebo arms in the global population

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: All Randomized Participants

6. Primary Endpoint

The registry lists the primary outcome as Overall Survival (OS) of Nivolumab + Ipilimumab Versus Placebo In The Global Population, measured over From randomization to 400 deaths across the two treatment groups (Nivo+Ipi vs Placebo) (up to approximately 37 months).

The primary endpoint was overall survival, measured from randomization to the date of death. If a participant had not died, the record says that the observation was censored at the last known alive date. Overall survival follow-up continued during blinded study-drug treatment and then every 3 months by in-person or phone contact after a participant discontinued blinded study drug.

Overall survival is a time-to-event endpoint: it accounts for both whether the event occurred and how long participants were followed before the event or censoring. It is often useful in trials because death is a clearly defined event, and the endpoint captures the duration from randomization to that event without requiring a tumor-response judgment. The registry specifies that the primary comparison was followed from randomization to 400 deaths across the two treatment groups, up to approximately 37 months.

The posted results report median overall survival with 95% confidence intervals and a hazard ratio from a stratified log-rank analysis. These statistics answer related but different questions. The median describes the time by which half of the estimated group has experienced the event; the hazard ratio compares event rates over follow-up under its analysis assumptions. Neither statistic alone describes every feature of the survival experience.

7. Results Posted in the Registry

• Global Placebo vs Global Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg: hazard ratio 0.92 (95% CI 0.75–1.12), reported p = 0.369 (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) of Nivolumab Versus Placebo

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Global Placebo vs Global Nivolumab 240 mg: hazard ratio 0.81 (95% CI 0.68–0.97) (as reported in the registry).
  • China Placebo vs China Nivolumab 240 mg: hazard ratio 0.93 (95% CI 0.53–1.66) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Global Placebo vs Global Nivolumab 240 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between China Placebo vs China Nivolumab 240 mg and the control. The interval gives the range of effects that remain compatible with the data.

Overall Survival (OS) of Nivolumab + Ipilimumab Versus Nivolumab

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Global Nivolumab 240 mg vs Global Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg vs China Nivolumab 240 mg: hazard ratio 1.13 (95% CI 0.94–1.36) (as reported in the registry).
  • China Nivolumab 240 mg vs China Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg: hazard ratio 1.00 (95% CI 0.56–1.79) (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 Global Nivolumab 240 mg vs Global Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg vs China Nivolumab 240 mg 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 China Nivolumab 240 mg vs China Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg and the control. The interval gives the range of effects that remain compatible with the data.

Progression Free Survival (PFS) Per BICR

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor; it was not recalculated.
• Result:
  • Global Placebo vs Global Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg: hazard ratio 0.74 (95% CI 0.62–0.88) (as reported in the registry).
  • Global Placebo vs Global Nivolumab 240 mg: hazard ratio 0.66 (95% CI 0.55–0.79) (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Global Placebo vs Global Nivolumab 1 mg/kg + Ipilimumab 3 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Global Placebo vs Global Nivolumab 240 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary outcomes broaden the description beyond the primary combination-versus-placebo overall-survival question. They include overall-survival comparisons involving nivolumab monotherapy and combination therapy, as well as progression-free survival assessed per BICR and tumor-mutational-burden subgroup outcomes. Their estimates do not all convey the same degree of precision or the same direction of association. Some posted global hazard-ratio intervals lie entirely below 1, while other comparisons, including the primary result and the China comparisons, have intervals that span 1.

This broad pattern is relevant context, but it does not replace the primary endpoint. Secondary analyses can help describe other outcomes and comparisons, and may suggest questions for further evaluation. They do not turn a primary result with substantial uncertainty into a definitive finding. Nor should an isolated secondary estimate be treated as confirmatory solely because its interval excludes the null value.

When many outcomes, treatment contrasts, and subgroups are examined, the chance of obtaining apparently notable results by chance increases unless the analysis accounts for multiplicity. The registry record lists secondary outcomes and analyses but does not report a multiplicity adjustment or a hierarchy for interpreting them. Consequently, unadjusted secondary results carry less weight as confirmatory evidence than a prespecified primary analysis with a stated testing strategy. The record also does not report enough detail to determine whether every secondary comparison was prespecified in the same form as posted.

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

• Overall Survival (OS) of Nivolumab Versus Placebo
• Overall Survival (OS) of Nivolumab + Ipilimumab Versus Nivolumab
• Progression Free Survival (PFS) Per BICR
• Overall Survival (OS) in Tumor Mutational Burden (TMB) High and Low Subgroups by TMB Cutoff In The Global Population
• Progression Free Survival (PFS) Per BICR in Tumor Mutational Burden (TMB) High and Low Subgroups by TMB Cutoff In The Global Population

9. Statistical Interpretation

10. Statistical Methods Explained

What does the median measure?

For a time-to-event outcome, the median is the estimated time at which the survival probability reaches the halfway point. It is a feature of the estimated survival distribution, not an arithmetic average of participants' observed times. Some participants may have died before that point, while others may still be alive at later follow-up or may have been censored. The median therefore condenses a survival curve into a single interpretable time, but it does not preserve the curve's full shape.

For the primary outcome, the registry defines death as the event and last known alive date as the censoring point for participants who had not died. Censoring means that the event time is not observed during the available follow-up; it does not mean that the participant experienced the event at the censoring date. Survival methods use the information available up to that point, subject to assumptions about censoring. The registry's description of censoring is available, but a fuller account of assumptions and sensitivity analyses is not reported in the registry record.

What does the hazard ratio add?

A hazard ratio compares the event rate at a given time between groups, conditional on participants having remained event-free up to that time. It is a relative time-to-event measure, not a ratio of median survival times. A value of 1 is the conventional null value: it corresponds to no relative difference in the hazards under the model's interpretation. Values on either side indicate different estimated directions, but the confidence interval and analysis assumptions matter when interpreting that estimate.

A hazard ratio compresses information across follow-up into one number. If the relative hazards change over time, one number may not capture that changing pattern well. The registry record does not report whether proportional-hazards assumptions were assessed or whether other time-varying analyses were used. The posted median estimates and their confidence intervals complement the hazard ratio, but they do not resolve every question about the shape of the survival experience.

Why use a stratified log-rank analysis?

The primary reported method is a stratified log-rank analysis. In general, the log-rank test compares time-to-event experience between groups while accounting for the ordering of observed events and censoring over follow-up. Stratification can preserve comparisons within defined groups rather than treating all participants as if they came from one undifferentiated set. Which factors defined the strata is not reported in the registry record, so the role of any particular stratification factor cannot be evaluated from this record.

The registry reports a hazard ratio and confidence interval alongside the stratified log-rank method. These outputs should be read as the reported sponsor analysis, rather than assuming that the p-value is derived from a different, unreported model. The record does not describe the estimator used for the hazard ratio, its variance calculation, or the details of any adjustment.

How should confidence intervals be read?

A confidence interval expresses uncertainty in an estimated quantity under the statistical procedure used. A narrower interval indicates greater precision on that scale; a wider interval indicates greater uncertainty. For a hazard ratio, an interval spanning 1 is compatible with the null value at the confidence level used, while an interval entirely on one side of 1 would indicate a more clearly directional estimate under that procedure. This is not the same as saying that every value in the interval is equally plausible or that the interval is a prediction range for individual outcomes.

The primary 95% confidence interval spans 1, and the reported p-value is 0.369. The p-value measures how unusual data at least as inconsistent with the null as those observed would be under the null and the test's assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect.

What can randomization and masking contribute?

In general, random assignment helps balance prognostic factors across groups in expectation, supporting a fair comparison of assigned strategies. Masking can reduce the influence of expectations on care, outcome assessment, and reporting. The registry identifies the study as randomized and records quadruple masking, but the operational details of allocation and masking are not reported in the registry record. Randomization does not guarantee perfect balance in a particular study, and masking does not remove all sources of uncertainty.

The trial's registry record gives a high-level method description, not the complete statistical analysis plan. Details such as the strata, assumptions checks, handling of protocol deviations, and prespecified decision rules are not reported in the registry record. The posted estimate should therefore be interpreted as a reported analysis with the limits of methodological detail available in the record.

11. Confidence Intervals and Precision

Precision concerns how tightly the data locate an estimate. The primary median estimates have 95% confidence intervals of 8.18 to 11.01 months for placebo and 8.15 to 10.25 months for nivolumab plus ipilimumab. Those intervals express uncertainty around each group's median; they are not direct confidence intervals for the difference between medians. Overlap or non-overlap between separate group intervals is not, by itself, a formal test of the between-group contrast.

The hazard-ratio estimate is 0.92, and its 95% confidence interval extends from 0.75 to 1.12. Because this interval includes 1, the estimate is compatible with a range of relative effects that includes no difference in hazards under the conventional interpretation. The interval also permits estimates on either side of 1. Its breadth means the data do not pinpoint a single relative effect precisely. The reported p-value, 0.369, is consistent with the uncertainty evident in the interval, but it does not quantify the probability that the treatment groups are the same.

Precision is distinct from clinical importance. A narrow interval can precisely estimate an effect that is small, while a wide interval can leave uncertainty about both magnitude and direction. Assessing importance requires context beyond the statistical interval, and the registry record does not supply a threshold for a clinically important difference. Nor does it report all details needed to assess whether uncertainty estimates account for every design feature.

For secondary comparisons, precision varies across the posted global and China groups. Smaller outcome groups generally have less information available for estimation, but the record's estimates and intervals should be considered directly rather than inferred from group size alone. The global primary comparison remains the central confirmatory result identified by the registry.

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 sample-size display for the primary time-to-event comparison is expressed as total events, not participants per arm. Its planning rows use 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect. For example, planning for 100% of the observed effect lists 4,516 total events for 80% power and 6,046 total events for 90% power. These are total event targets, not enrollment targets.

The planning illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It uses the observed effect as a reference and considers the stated fraction of that effect as the effect to detect. The registry record does not report the trial's original design assumptions or sample-size justification; these planning assumptions are not a substitute for that missing information.

In plain language, 80% power describes the chance of detecting an effect of the assumed size under the planning assumptions if that effect is truly present. The 90% power column represents a higher planned chance of detection under those assumptions. These columns do not say what power the completed trial actually had. The figures are planning illustrations, not observed power.

The curve should be read with the same distinction between assumed effect and required information: a smaller effect is harder to distinguish from random variation and therefore needs more events for a given planned power. Because this is a time-to-event endpoint, the relevant target is the number of events, not simply the number enrolled. The display does not establish that the trial reached any listed target or that its original design was based on these calculations.

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 = 0.920, so ln(HR) = -0.0834.
2. d = 4 × (1.96 + 0.842)² / [ln(0.920)]² = 4,515.7, rounded up to 4,516 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 <- 0.92                              # 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(2538666)
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 = 4,516 (80% power), 6,046 (90% power)
 75% of observed effect: events = 8,028 (80% power), 10,748 (90% power)
 50% of observed effect: events = 18,063 (80% power), 24,182 (90% power)
Simulated power with 4,516 events: 0.809 (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. 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 7,120 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 907 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 <- 9.17          # control-arm median, months (converted from the posted unit)
hr          <- 0.92          # 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: 5028         p (1-sided)   0.0001
 Events: 2039        ~HR at bound   0.8484
    Month: 17    P(Cross) if HR=1   0.0001
              P(Cross) if HR=0.92   0.0338
    IA 2: 67%                   Z   2.5114
      N: 7120         p (1-sided)   0.0060
 Events: 4077        ~HR at bound   0.9243
    Month: 27    P(Cross) if HR=1   0.0060
              P(Cross) if HR=0.92   0.5603
        Final                   Z   1.9930
      N: 7120         p (1-sided)   0.0231
 Events: 6116        ~HR at bound   0.9503
    Month: 60    P(Cross) if HR=1   0.0250
              P(Cross) if HR=0.92   0.9000



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

15. Analyses Reported by the Sponsor

The primary result reported in the registry is a stratified log-rank analysis, with a hazard ratio of 0.92, a 95% confidence interval from 0.75 to 1.12, and p = 0.369. This is the sponsor's reported analysis of the primary comparison. It is the appropriate reference for describing the trial's stated primary result.

Any simple re-calculation from posted summary statistics would be unadjusted and would not necessarily reproduce the sponsor's stratified analysis. Such a calculation cannot replace the trial's analysis or its reported p-value. In this record, the discrepancy indicator is false, so there is no flagged disagreement requiring an alternative estimate to take precedence. The sponsor's reported result remains the main inferential result because it is the analysis specified in the registry's posted results.

That distinction does not mean the reported method is fully documented. The registry does not identify the stratification factors or provide all details of the estimator and analysis plan. The estimate, interval, method, and p-value can be stated as reported, while the unavailable methodological details remain unknown rather than being filled in by assumptions.

16. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple outcome measures and analyses, including several overall-survival comparisons, progression-free survival outcomes, and tumor-mutational-burden subgroup analyses. When investigators examine multiple outcomes or contrasts, the probability of at least one apparently unusual result can increase. A prespecified testing hierarchy or multiplicity adjustment can help control the interpretation of that family of tests. Whether such procedures were used is not reported in the registry record.

The record identifies the primary population as all randomized participants in the global nivolumab plus ipilimumab and placebo arms. That statement clarifies the population for the primary comparison, but it does not supply a full account of analysis populations for every outcome. The analysis populations for the secondary endpoints, including whether all randomized participants were included for each one, are not reported in the registry record.

For overall survival, participants who had not died were censored at their last known alive date. The record describes follow-up contacts but does not report how missing or incomplete follow-up was handled beyond that censoring description. It does not state whether sensitivity analyses were conducted for missing outcome information, whether any imputation was used, or how protocol deviations affected inclusion in each analysis.

The registry record also does not report the original statistical analysis plan, the rationale for the sample size, interim analyses, stopping rules, or the factors used for stratification. Those omissions limit assessment of whether the posted analyses follow a prespecified strategy and how multiplicity, missing data, or analysis-population choices may affect interpretation. They should be treated as unknowns, not as evidence that a particular procedure was or was not used.

17. 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: malignant neoplasm progression (18.3%), immune-mediated enterocolitis (8.3%), liver injury (8.3%), hyponatraemia (8.3%), platelet count decreased (8.0%).

The safety record lists participants at risk and participants with serious adverse events and other adverse events for global and China groups. In the global placebo group, 117 participants had serious adverse events among 273 at risk, and 228 had other adverse events among 273 at risk. In the global nivolumab 240 mg group, the corresponding counts were 131 among 279 and 244 among 279. In the global nivolumab 1 mg/kg plus ipilimumab 3 mg/kg group, they were 198 among 278 and 263 among 278.

The record also lists deaths for these groups and the China groups, but these counts alone do not establish that deaths were caused by treatment. Serious adverse-event terms listed among the leading terms include malignant neoplasm progression, immune-mediated enterocolitis, liver injury, hyponatraemia, platelet count decreased, and pleural effusion. The percentages attached to those terms are reported values and should not be read as a complete description of all harms.

These counts describe events recorded in the registry. They are not a randomized comparison adjusted for follow-up or other factors, and the record does not provide enough detail to establish causality. Safety interpretation should distinguish serious events from other adverse events and should not treat every event after treatment as treatment-caused.

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

The primary statistics describe overall survival in the randomized global comparison of nivolumab plus ipilimumab with placebo after first-line chemotherapy. The median estimates are similar in scale, and the reported hazard-ratio interval includes 1. The result therefore does not give a precise estimate that clearly distinguishes the groups under the reported analysis. The confidence interval leaves uncertainty about the relative effect, and the p-value does not establish that the treatments are equivalent.

These results do not prove that no effect exists, nor do they establish a clinically important benefit or harm. A failure to show a clear difference is not the same as demonstrating equivalence or non-inferiority; those conclusions require suitable designs, margins, and analyses, which are not reported for this primary question. Statistical uncertainty also does not determine clinical importance by itself.

The primary endpoint concerns time from randomization to death, with censoring for participants not known to have died by the last known alive date. It does not directly measure every aspect of disease control, quality of life, or treatment burden. Progression-free survival and other outcomes provide different information, but they do not answer the same question as overall survival. Their interpretation is also tempered by the number of secondary analyses and the absence of reported multiplicity procedures.

The trial was randomized and masked according to the registry, which supports a structured comparison of assigned groups. Still, the posted results are population-level estimates, not predictions for an individual. They do not show that every participant had the same outcome, and they cannot by themselves identify which clinical or biological factors explain variation in survival. No treatment recommendation follows from these statistics alone.

19. Why This Trial Matters Statistically

This trial is statistically informative because it evaluates a time-to-event outcome in a randomized, masked, parallel-group Phase 3 study and reports both medians and a hazard ratio with a confidence interval. The combination-versus-placebo comparison asks a defined question about overall survival after chemotherapy, while the use of censoring allows participants with incomplete event follow-up to contribute information up to their last known alive date.

The result also illustrates why a point estimate should not be read alone. A hazard ratio of 0.92 is accompanied by an interval that spans 1, and the reported p-value is 0.369. The point estimate gives one summary of the data; the interval communicates that the estimate is uncertain. Reading both prevents overstatement in either direction, including treating a non-significant result as proof of no difference.

The multiple secondary outcomes make the trial a useful example of the distinction between primary and exploratory evidence. Secondary estimates can add context, but without a reported multiplicity plan, they warrant more cautious interpretation. The record also illustrates why reporting the analysis population, censoring rules, and statistical plan matters: those details help readers judge the assumptions behind the estimates.

Finally, the sample-size display is a planning illustration based on the observed effect, not a record of the trial's original design or a measure of achieved power. Distinguishing planning calculations from trial design and from observed results is essential when interpreting time-to-event evidence. The registry provides important outcome information, but the missing design and analysis details limit a complete statistical audit.

20. 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 record gives limited detail about how the primary analysis was implemented. Although it names a stratified log-rank method, it does not identify the stratification factors, explain the hazard-ratio estimator, or report checks of time-to-event model assumptions. The original statistical analysis plan and sample-size justification are not reported in the registry record, so the posted planning illustration should not be taken as the trial's original design calculation.

The primary analysis has a confidence interval that spans the null value, leaving uncertainty about the direction and magnitude of the relative effect. The record's primary follow-up description refers to 400 deaths across the two treatment groups, up to approximately 37 months, while the outcome record also describes follow-up up to approximately 73 months for secondary overall-survival outcomes. These are different recorded timeframes and should not be treated as identical analysis windows.

Several secondary comparisons and subgroup outcomes are posted, but the record does not describe multiplicity control or a testing hierarchy. This makes confirmatory interpretation of individual secondary results more difficult. The record also does not report full missing-data procedures, sensitivity analyses, or analysis populations for every secondary outcome.

Safety entries provide event counts for specified groups, but these alone do not establish causation or permit a complete assessment of comparative risk. The registry record does not provide a full account of exposure-adjusted safety rates or all details needed to assess follow-up differences. These limitations concern the information available in the record; they should not be read as proof that unreported procedures were not performed.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: An Investigational Immuno-therapy Study of Nivolumab, or Nivolumab in Combination With Ipilimumab, or Placebo in Patients With Extensive-Stage Disease Small Cell Lung Cancer (ED-SCLC) After Completion of Platinum-based Chemotherapy — NCT02538666.
• Registry record last updated 2023-01-05; status last verified 2022-12.
• Results first posted 2019-10-22.
• 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.

24. Record Summary

NCT02538666 is a completed randomized, multicenter, double-blind, Phase 3 parallel-group study of maintenance therapy after first-line chemotherapy for extensive-stage disease small cell lung cancer. Overall enrollment was 907, and the record identifies quadruple masking.

The primary outcome was overall survival for the global nivolumab plus ipilimumab and placebo groups. Median overall survival was 9.17 months for combination therapy and 9.56 months for placebo. The reported stratified log-rank hazard ratio was 0.92, with a 95% confidence interval from 0.75 to 1.12 and p = 0.369.

The registry also posts secondary survival outcomes, subgroup analyses, and safety counts. The primary interval spans 1, and important details about multiplicity, the original sample-size rationale, and some analysis procedures are not reported in the registry record. The reported estimates should therefore be read with their uncertainty and documentation limits in mind.