Study Of Irinotecan Hydrochloride (Campto(R)) And Cisplatin Versus Etoposide And Cisplatin In Small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study Of Irinotecan Hydrochloride (Campto(R)) And Cisplatin Versus Etoposide And Cisplatin In Small Cell Lung Cancer (NCT00143455), a randomized phase 3 study. The primary outcome is Overall Survival (OS) for the Full Analysis Population (FAP).

2. Clinical Question

3. Trial Design

NCT00143455 is a completed, randomized, open-label, parallel-group Phase III study sponsored by Pfizer. The registry describes it as a multicentre study in chemotherapy-naive patients with extensive disease small cell lung cancer. Enrollment was 485. The listed study start was 2002-06, and primary completion was 2008-12.

Participants were assigned to irinotecan plus cisplatin or etoposide plus cisplatin. The study was unmasked, so treatment assignment was not concealed from participants and study personnel after randomization. The registry reports results separately for Cohort 1 and Cohort 2, but the outcome estimates and reported primary hazard ratio concern Cohort 2. Cohort 1 enrollment was terminated early under a protocol amendment dated 29 September 2003; the registry notes a limited number of subjects in that cohort. Cohort 1 outcome estimates are not reported in the registry record.

The planned primary outcome was overall survival in the full analysis population. Secondary outcomes listed in the registry included confirmed response, duration of response, time to tumor progression, quality of life measured with the EORTC QLQ-C30, and tumor-related symptoms. The registry lists seven outcome measures and five posted analyses. The original sample-size justification, the detailed allocation procedure, and the trial’s 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.

Cohort 1

Reasons for not completing this period:

Cohort 2

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: FAP = Full analysis population (all treated patients)analyzed in the arm they were assigned by randomization, assuming they had a confirmed small cell lung cancer. The enrollment of Cohort 1 was terminated early per protocol amendment (29 September 2003). With limited number of subjects in Cohort 1, the efficacy analysis was exploratory.

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 Overall Survival (OS) for the Full Analysis Population (FAP), measured over Baseline to date of death (every 3 weeks for up to 6 months on study treatment and every 2 months for a minimum of 13 months post study treatment).

Overall survival measures the time from randomization until death from any cause. It is a time-to-event endpoint: the event is death, and the analysis takes account of how long each participant was followed. A participant not known to have died is censored at the last contact when they were known to be alive. Censoring retains information about the participant’s observed follow-up without treating the last contact as a death.

The registry reports median overall survival for each Cohort 2 treatment group, estimated with the Kaplan–Meier method. A median is a point on the estimated survival distribution, not an average lifespan. The endpoint is useful for comparing how long participants survive after randomization while accounting for censoring, but the medians alone do not describe the full survival pattern or the uncertainty in every part of that pattern.

The registry also reports a Cox proportional hazards comparison. Its hazard ratio summarizes the relative event rate over follow-up under the model, rather than directly expressing a difference in median survival. The confidence intervals and the model-based comparison therefore add information beyond the medians. The registry’s definition uses death from any cause, so the endpoint is not limited to deaths attributed to lung cancer or treatment.

7. Results Posted in the Registry

• Irinotecan + Cisplatin (Cohort 2) vs Etoposide + Cisplatin (Cohort 2): hazard ratio 1.24 (95% CI 0.99–1.54), reported p = 0.056 (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.

Duration of Response (DR)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Cox Proportional Hazard Model); it was not recalculated.
• Result:
  • Irinotecan + Cisplatin (Cohort 2) vs Etoposide + Cisplatin (Cohort 2): hazard ratio 1.35 (95% CI 0.96–1.91), reported p = 0.083 (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 Irinotecan + Cisplatin (Cohort 2) vs Etoposide + Cisplatin (Cohort 2) and the control. The interval gives the range of effects that remain compatible with the data.

Time to Tumor Progression (TTP)

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Cox Proportional Hazard Model); it was not recalculated.
• Result:
  • Irinotecan + Cisplatin (Cohort 2) vs Etoposide + Cisplatin (Cohort 2): hazard ratio 0.96 (95% CI 0.77–1.21), reported p = 0.754 (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 Irinotecan + Cisplatin (Cohort 2) vs Etoposide + Cisplatin (Cohort 2) and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary time-to-event outcomes show a mixed pattern rather than a consistent extension of the primary overall-survival finding. Duration of response and time to tumor progression are distinct outcomes: the former concerns how long a confirmed response lasts, while the latter concerns time until progression. Their estimates and intervals should be interpreted on their own definitions and in light of their uncertainty, not treated as substitutes for overall survival.

The registry also lists confirmed response, EORTC QLQ-C30 quality of life, and tumor-related symptoms including pain, dyspnea, cough, hemoptysis, weight and analgesic use. Detailed results for those measures are not reported in the registry record. Their listing indicates that the study planned to assess outcomes beyond survival, but it does not establish what those assessments found.

Secondary analyses generally carry less weight than the prespecified primary question because they are additional opportunities to observe apparent differences. When many outcomes are examined, some estimates may look noteworthy by chance even if there is no underlying effect. The registry does not report a multiplicity adjustment or a hierarchy for interpreting secondary tests. Accordingly, secondary results are best treated as supportive or exploratory context and should not overturn the uncertain primary comparison. The available secondary findings do not provide a single, consistent pattern that resolves the overall-survival uncertainty.

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

• Number of Subjects With Overall Confirmed Response
• Duration of Response (DR)
• Time to Tumor Progression (TTP)
• European Organization for Research and Treatment of Cancer Quality of Life (EORTC QLQ-C30)
• Tumor Related Symptoms (Pain, Dyspnea, Cough, Hemoptysis, Weight, and the Use of Opioids and Non-Opioids Analgesics)

9. Statistical Interpretation

10. Statistical Methods Explained

How does the Kaplan–Meier method use follow-up?

Overall survival is observed over time, and not every participant necessarily has a recorded death during follow-up. Kaplan–Meier estimation handles this by incorporating observed event times and censoring times. When a participant’s survival time is censored, their observed follow-up contributes information until the last contact at which they were known to be alive. The method does not assume that a censored participant died at that contact.

The registry defines overall survival as time from randomization to death from any cause. For a participant who did not expire, the record says that survival time was censored at the last known-alive contact. The stated follow-up schedule included contact every 3 weeks for up to 6 months on study treatment and every 2 months for a minimum of 13 months after study treatment. These schedule details describe the planned observation pattern, not necessarily the follow-up duration for every participant.

The Kaplan–Meier estimate produces a survival curve across observed follow-up. The median is the time at which the estimated survival curve reaches the midpoint of its survival distribution. It is a compact summary, but it does not show the entire curve. The registry reports confidence intervals for the medians calculated with the Brookmeyer and Crowley method. The curves and intervals help readers assess the pattern and uncertainty, while the median alone cannot show whether hazards vary over time.

What does the Cox model comparison represent?

The registry reports a Cox proportional hazards model for the comparison of the Cohort 2 groups. This model uses event and follow-up information to estimate a hazard ratio. The hazard is the instantaneous event rate among participants still at risk at a given time, conditional on the model. A hazard ratio compares that rate between the groups. It is not a ratio of median survival times and does not directly state how many months of life are gained or lost.

The proportional hazards model expresses the relative hazard as constant over time. That is a modeling assumption, not a fact established merely by reporting a Cox estimate. Assessment of that assumption and any diagnostics are not reported in the registry record. The registry’s comparison is therefore interpreted as the sponsor’s reported model-based summary, with the limitation that the record does not provide the supporting diagnostics or details of model covariates.

The hazard ratio is reported as 1.236, with a 95% confidence interval from 0.995 to 1.536 and p = 0.056. A hazard ratio of 1 corresponds to equal hazards under the comparison. An interval containing 1 indicates that the data do not exclude equal hazards at the stated interval level. It also allows a range of relative hazards, so the estimate should be considered together with its uncertainty rather than interpreted as a precise fixed effect.

What does the confidence interval add?

A confidence interval describes the precision of an estimate under the statistical method and its assumptions. A narrower interval indicates greater precision; a wider interval indicates greater uncertainty. The interval is not a range containing a known fixed percentage of individual survival times, and it is not a probability statement that the true effect has a particular chance of falling within the observed interval.

For a median, the interval characterizes uncertainty in the estimated median. For a hazard ratio, it characterizes uncertainty in the estimated relative hazard. The two intervals answer different questions because the median and hazard ratio summarize different features of the time-to-event data. Their interpretation also depends on the population and model used. The registry identifies the primary population as the full analysis population and reports the Cohort 2 comparison; the detailed handling of missing or incomplete records beyond the stated censoring rule is not reported in the registry record.

Why do the methods matter for interpretation?

Time-to-event methods make use of unequal follow-up rather than reducing the outcome to a simple count of deaths. This is important when participants are observed for different lengths of time. Censoring is informative only under assumptions about how follow-up ends; if the chance of being censored is related to future survival in ways not addressed by the analysis, estimates may be affected. The registry describes censoring at last known-alive contact but does not report analyses exploring departures from the assumptions.

The reported estimate is also specific to the trial population and comparison. The full analysis population is defined as treated participants analyzed according to randomized assignment, assuming confirmed small cell lung cancer. The registry does not report additional population definitions or sensitivity analyses for the primary outcome. These boundaries matter: a model estimate summarizes the data as analyzed, not all possible patients, settings, or treatment practices.

11. Confidence Intervals and Precision

The primary median estimates are close in magnitude, but each has uncertainty. The 95% confidence interval for irinotecan plus cisplatin extends from 8.969 to 11.66 months; the interval for etoposide plus cisplatin extends from 8.936 to 11.07 months. Their overlap illustrates that the posted estimates do not sharply separate the survival distributions. Overlap alone is not a formal test of the treatment contrast, which is why the registry also reports a Cox model comparison.

The hazard ratio confidence interval, 0.995 to 1.536, spans the value of 1 and allows both a very small relative hazard in the direction below 1 and a higher hazard in the direction above 1. Its width conveys that the estimated direction and magnitude are not pinned down precisely. The p-value of 0.056 is consistent with this uncertainty; it should not be read as a measure of the probability that the treatment effect is absent.

Confidence intervals express uncertainty under the calculation method and its assumptions. They do not account automatically for all possible sources of bias, model misspecification, missing outcomes, or selective reporting. The registry does not report diagnostics for the proportional hazards assumption or a detailed missing-data sensitivity analysis. Accordingly, the intervals are useful descriptions of statistical precision, but not a complete measure of certainty about the broader clinical interpretation.

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 illustrations are planning calculations based on the observed effect, not the trial’s original design justification. Because the outcome is time-to-event, the entries are total event targets, not participants per arm. The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They consider detecting 100%, 75% or 50% of the observed effect. The original design assumptions are not reported in the registry record.

The 80% power column describes the event target calculated to give an 80% chance of detecting the specified effect under those planning assumptions, if that effect is the true one and the model assumptions hold. The 90% power column expresses the corresponding target for a higher chance of detection. These are prospective planning quantities; they do not tell how likely the trial was to detect its actual effect after the results were observed. This is not observed power.

Smaller effects are harder to distinguish from random variation, so the planning calculations require more events when the target effect is reduced. For overall survival, the total-event targets at 80% power are 700 when planning for 100% of the observed effect, 1,244 for 75%, and 2,798 for 50%. At 90% power, the corresponding total-event targets are 937, 1,665 and 3,745. These are event counts, not arm-specific participant counts. The sample-size curve should be read as a visual representation of how the event requirement changes as the target effect changes, rather than as an account of the trial’s actual enrollment or observed power.

13. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Hazard ratio HR = 1.236, so ln(HR) = 0.2119.
2. d = 4 × (1.96 + 0.842)² / [ln(1.236)]² = 699.3, rounded up to 700 events in total for 80% power.

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

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
hr <- 1.236                              # 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(143455)
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 = 700 (80% power), 937 (90% power)
 75% of observed effect: events = 1,244 (80% power), 1,665 (90% power)
 50% of observed effect: events = 2,798 (80% power), 3,745 (90% power)
Simulated power with 700 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. Analyses Reported by the Sponsor

The registry reports the primary Cohort 2 hazard ratio from a Cox proportional hazards model as 1.236, with a 95% confidence interval from 0.995 to 1.536 and p = 0.056. The stated method and estimate are the sponsor’s reported analysis. For this time-to-event outcome, the posted interpretation uses the sponsor’s hazard ratio rather than substituting a simple recalculation from summary statistics. A median and its interval do not contain enough information to reconstruct the same Cox model comparison.

The record’s discrepancy indicator is false, and no conflicting sponsor result is identified. The sponsor’s own analysis therefore remains the relevant reported comparison. The registry does not report sufficient model detail to independently reproduce the analysis or evaluate every modeling choice. The reported analysis should be read with its confidence interval and the limitations of the registry’s methods description, rather than treated as an independently verified reanalysis.

15. Multiplicity, Missing Data and Analysis Populations

The registry identifies overall survival as the primary outcome and lists several secondary outcomes. It does not report whether secondary comparisons were adjusted for multiplicity, whether a formal testing hierarchy was used, or how the family of outcomes was defined for statistical testing. Without that information, nominal secondary p-values, where posted, should not be treated as if they carried the same confirmatory weight as a single prespecified primary test. The analysis plan for interim looks and any interim-analysis boundaries are not reported in the registry record.

The primary endpoint description states that participants without a recorded death were censored at their last known-alive contact. That is the reported approach to incomplete survival follow-up. Other missing-data procedures, including handling of missing covariates, missing secondary outcomes, or sensitivity analyses for informative censoring, are not reported in the registry record. The stated censoring rule does not by itself demonstrate that censoring was unrelated to future survival.

The full analysis population is described as all treated patients analyzed in the randomized arm, assuming confirmed small cell lung cancer. The registry does not report additional analysis-population definitions or sensitivity analyses that would show how departures from this definition affect the result. It notes that Cohort 1 enrollment was terminated early per protocol amendment on 29 September 2003, and that the number of subjects in that cohort was limited. The reported primary comparison is for Cohort 2; Cohort 1 primary outcome results are not reported in the registry record.

The original sample-size justification and the trial’s original statistical analysis plan are also not reported in the registry record. The planning illustration based on the observed effect is not evidence about the original design decisions. These omissions limit assessment of prespecification, multiplicity control, and robustness, but they do not change the numerical estimates that the registry reports.

16. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: fatigue (17.9%), diarrhoea nos (16.3%), nausea (15.4%), weight decreased (15.4%), dyspnoea nos (14.9%).

The registry reports serious adverse event counts for each listed cohort and treatment group, alongside the number at risk. In Cohort 2, serious adverse events affected 45 of 202 participants at risk in the irinotecan plus cisplatin group and 29 of 203 at risk in the etoposide plus cisplatin group. In Cohort 1, the corresponding figures were 9 of 39 and 4 of 38. These are group-level counts; they should be read with the cohort distinction and the limited Cohort 1 enrollment in mind.

The registry also lists serious adverse event terms, including fatigue, diarrhoea, nausea, weight decreased, dyspnoea and vomiting. The posted percentages describe the listed serious terms and do not, by themselves, establish a causal relationship to treatment or compare overall safety comprehensively. Death counts are not reported in the registry record. The adverse-event figures should not be combined with the survival result as if they were the same outcome, and the registry entry does not provide enough detail to characterize duration, severity grading, or event timing for each participant.

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

The primary result is an estimate of overall survival in the Cohort 2 groups, accompanied by uncertainty. The medians are reported with overlapping 95% confidence intervals, and the Cox hazard ratio interval includes 1. Together, these results do not provide a precise separation of the groups’ survival experience. They also do not establish equivalence: an uncertain estimate is not the same as evidence that the treatments have identical effects.

The hazard ratio points above 1 in the comparison order reported by the registry, while its confidence interval allows values close to 1 and values farther above it. The p-value should be interpreted as a feature of the test under its assumptions, not as the probability that one regimen is better or worse. Statistical uncertainty is only one part of interpretation. The registry does not report a clinical threshold for an important survival difference, so the practical importance of the observed estimates cannot be determined from the statistical result alone.

The findings apply to the analyzed trial population and the regimens as studied. They should not be generalized automatically to other patient groups or treatment settings. Cohort 1 was terminated early and has no primary outcome estimates reported in the registry record; the reported primary comparison is specifically for Cohort 2. Secondary measures add context but are not a replacement for the primary endpoint, especially where full results or multiplicity procedures are not reported.

These data describe trial results; they do not constitute individualized medical advice. They do not, on their own, settle which regimen is appropriate for a particular patient or establish a comprehensive benefit–risk judgment. Such conclusions would require broader clinical context and evidence beyond the posted statistical summaries.

18. Why This Trial Matters Statistically

This trial is statistically instructive because its primary endpoint is time to an event, with censoring, rather than a simple outcome measured once for every participant. Kaplan–Meier estimation allows survival information to contribute through observed follow-up, and the Cox model summarizes a relative hazard. These methods answer related but different questions: the medians describe points on the estimated survival distributions, while the hazard ratio summarizes a model-based relative event rate.

The result also demonstrates why an estimate should be read together with its interval. The point estimate alone suggests a direction, but the confidence interval includes the no-difference value and permits a range of effect sizes. The reported p-value similarly does not erase uncertainty or convey the clinical importance of the estimate. The relevant conclusion is not determined by a single threshold in isolation.

The cohort structure matters as well. The registry describes early termination of Cohort 1 enrollment and reports the primary comparison for Cohort 2. Treating the cohorts as though they offered equally informative, fully reported comparisons would obscure what the record actually contains. Finally, the planning illustration makes clear that detecting smaller time-to-event effects requires more events under the stated assumptions. That illustration is useful for understanding future study planning, but it is not a reconstruction of this trial’s original design or a measure of its observed power.

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 record does not report the original sample-size justification, detailed randomization procedures, or the original statistical analysis plan. The reported Cox comparison is not accompanied by model diagnostics, covariate details, or an assessment of the proportional hazards assumption. These omissions restrict independent evaluation and reproduction of the analysis.

The primary outcome results are reported for Cohort 2. Cohort 1 enrollment was terminated early per protocol amendment, and its primary outcome estimates are not reported in the registry record. The study’s total enrollment should therefore not be read as the size of the primary Cohort 2 comparison. The available summaries do not describe the distribution of follow-up or the number of deaths supporting the primary estimate.

The record states how survival was censored but does not report sensitivity analyses for potentially informative censoring or other missing-data procedures. It also does not report whether multiplicity was controlled across secondary outcomes or whether interim analyses were conducted. Detailed findings for several listed secondary outcomes are not reported in the registry record. These limitations affect how fully the posted results can be audited and contextualized; they do not justify replacing the reported estimates with unsupported assumptions.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study Of Irinotecan Hydrochloride (Campto(R)) And Cisplatin Versus Etoposide And Cisplatin In Small Cell Lung Cancer — NCT00143455.
• Registry record last updated 2010-02-18; status last verified 2010-02.
• Results first posted 2010-01-21.
• 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

NCT00143455 is a completed randomized, open-label, multicentre Phase III trial comparing irinotecan plus cisplatin with etoposide plus cisplatin in chemotherapy-naive patients with extensive disease small cell lung cancer. The primary endpoint was overall survival, defined from randomization to death from any cause. The reported primary comparison is for Cohort 2, using Kaplan–Meier medians and a Cox proportional hazards model.

The registry reports a hazard ratio of 1.236, with a 95% confidence interval from 0.995 to 1.536 and p = 0.056. The estimate is uncertain, and the interval includes 1. Cohort 1 enrollment ended early under a protocol amendment, and its primary results are not reported in the registry record. Secondary and safety information provides additional context, with important limitations in completeness and reported methods.