S0802 - Topotecan With or Without Aflibercept in Treating Patients With Extensive-Stage Small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for S0802 - Topotecan With or Without Aflibercept in Treating Patients With Extensive-Stage Small Cell Lung Cancer (NCT00828139), a randomized phase 2 study. The primary outcome is Progression-free Survival (PFS).

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group phase II trial sponsored by the National Cancer Institute. The registry record lists an enrollment of 189 and a completed status. The recorded start was 2009-05, and primary completion was 2012-09. Masking was listed as none, so this was an open-label design according to the registry record.

The two treatment strategies were topotecan with ziv-aflibercept and topotecan alone. The listed topotecan schedule was intravenous administration on days 1, 8 and 15, repeated every 21 days for 4 courses. The combination arm also received intravenous ziv-aflibercept on day 1. The arm descriptions mention further treatment for participants with responsive or stable disease after the initial courses, but the registry record does not include the complete description of that subsequent treatment.

The posted PFS results are presented in four platinum-status subgroups: platinum-sensitive participants receiving the combination, platinum-sensitive participants receiving topotecan alone, platinum-refractory participants receiving the combination, and platinum-refractory participants receiving topotecan alone. The subgroup sizes associated with the PFS outcome were 42, 41, 55 and 51, respectively. These are outcome-group counts, not a substitute for the registry’s overall enrollment figure.

The registry record does not report details of the randomization sequence, allocation concealment, eligibility criteria beyond the listed disease context, or the trial’s original sample-size justification. Those omissions limit how fully the design can be assessed from the record alone.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

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 Progression-free Survival (PFS), measured over Disease assessments were performed every 6 weeks, up to 2 years..

The primary endpoint was progression-free survival (PFS), reported as a median in months. It measures the time from registration until the first qualifying event: documented progression, symptomatic deterioration or death from any cause. The endpoint therefore incorporates both disease progression and death, rather than measuring tumor response alone.

For measurable target lesions, progression was defined as a 20% increase in the sum of the longest diameters compared with the smallest sum observed, or baseline if there had been no decrease during therapy. The registry definition also includes unequivocal progression of non-measurable disease in the treating physician’s opinion, appearance of a new lesion or site, and death due to disease without prior documentation of progression or symptomatic deterioration. Assessments were scheduled every 6 weeks, up to 2 years.

In general, PFS is used in trials when the timing of disease progression or death is relevant to the treatment comparison. A median describes the point in the observed time-to-event distribution at which half the events or survival experience fall on either side, subject to the analysis and follow-up methods. The registry record does not report the censoring rules or detailed analysis population used for this endpoint, so those aspects cannot be evaluated from the posted information.

7. Results Posted in the Registry

• Platinum-Sensitive Treated With Topotecan and Ziv-aflibercept vs Platinum Sensitivity Treated With Topotecan Alone: hazard ratio 1.32 (95% CI 0.91–1.92) (as reported in the registry).
• Platinum Refractory Treated With Topotecan + Ziv-aflibercept vs Platinum Refractory Treated With Topotecan Aloine: hazard ratio 1.51 (95% CI 1.08–2.10) (as reported in the registry).

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary outcome is a time-to-event measure. Each participant’s PFS time begins at registration and ends at the first qualifying event under the registry definition. In general, time-to-event methods can use information from participants who have not yet experienced the event by the end of their available follow-up, provided the analysis accounts for this incomplete observation appropriately. The specific handling of such observations, including censoring rules, is not reported in the registry record.

What does a median PFS estimate describe?

A median is a location summary of the PFS distribution. It gives a single estimate of the time at which the distribution reaches its midpoint, but does not describe the full pattern of events over follow-up. Two groups can have similar medians while having different event patterns at other times, or can have different medians while their uncertainty is substantial. The posted medians are reported with 90% confidence intervals, which communicate estimation uncertainty rather than the range of individual participant outcomes.

The median estimates are presented separately for platinum-sensitive and platinum-refractory participants. That separation is clinically relevant to interpreting the reported comparisons because the groups are not combined into one overall treatment estimate. However, the registry record does not say whether platinum status was a planned stratification factor, how it was determined, or whether the subgroup comparisons were prespecified. The separate estimates should not be treated as proof that treatment effects differ by platinum status; an explicit test of interaction is not reported.

How should the hazard ratios be read?

A hazard ratio is a relative comparison of event hazards over follow-up, as estimated under a time-to-event model. It is not the same quantity as a difference between medians, and it does not directly express the probability that an individual participant will experience progression or death. Interpretation depends on the event definition, model, treatment coding and assumptions. The registry record reports hazard ratios and confidence intervals but does not name the statistical method or state the coding convention for the treatment contrast. The estimates are therefore best read as the sponsor-reported relative comparisons, without assigning a direction of benefit from the sign or size alone.

In general, a proportional-hazards model assumes that the hazard ratio is sufficiently stable over time for a single relative estimate to be meaningful. If that assumption is not appropriate, a single hazard ratio may be difficult to interpret as a concise description of the difference. The registry record does not report whether proportional hazards were assessed, whether another model was considered, or how model fit was evaluated.

What do the three-month estimates add?

The registry also reports PFS estimates at 3 months, along with 90% confidence intervals. A time-specific estimate answers a different descriptive question from the median: it reports the estimated PFS at a selected time point. It can make an early portion of the outcome distribution more visible, but its interpretation still depends on follow-up and analysis methods. The record does not identify the estimation method or provide the number at risk at that time point, so those details cannot be assessed.

What is not specified about the analysis?

The registry record does not report the statistical analysis plan, the analysis populations, missing-data procedures, interim analyses, adjustment for multiple comparisons, or handling of protocol deviations. It also does not explain the selection of the 90% confidence level. These omissions matter because they affect how planned comparisons, uncertainty and incomplete outcome data should be interpreted. The methods description is therefore limited to the reported endpoint, estimates and general statistical principles; it does not infer a protocol that is not in the registry record.

10. Confidence Intervals and Precision

The confidence intervals show how much uncertainty surrounds the estimates. For the median PFS values, the reported 90% intervals are 1.5 to 2.2 months for the platinum-sensitive combination group, 1.2 to 1.4 months for the platinum-sensitive topotecan-alone group, 1.3 to 1.7 months for the platinum-refractory combination group, and 1.3 to 1.4 months for the platinum-refractory topotecan-alone group. The interval widths vary, so the medians are not all estimated with the same apparent precision.

The 90% confidence intervals for the hazard ratios are 0.91 to 1.92 in the platinum-sensitive comparison and 1.08 to 2.10 in the platinum-refractory comparison. In general, a confidence interval describes uncertainty in an estimate under the method used to construct it. It is not a range containing a known proportion of individual participant outcomes, and it does not guarantee that the true value lies within the displayed bounds.

The platinum-sensitive hazard-ratio interval includes 1, while the platinum-refractory interval lies above 1. The direction of the comparison cannot be confidently inferred without the event and treatment coding convention, which is not reported in the registry record. The record also does not provide p-values for these hazard-ratio results. The intervals are the available basis for describing the precision of these comparisons, and no stronger claim about statistical significance is warranted from the posted information alone.

The 3-month PFS estimates also have 90% confidence intervals: 0.14 to 0.37, 0.07 to 0.27, 0.18 to 0.39 and 0.04 to 0.2 across the four reported groups. They express uncertainty in the time-specific estimates, not variation among participants. In general, narrower intervals indicate greater precision under the analysis method, while interval width is affected by the amount and distribution of outcome information. The registry record does not report the number of events or the number at risk at each follow-up time, which limits further assessment of precision.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations are based on the reported hazard ratios and use total event counts, not participants per arm. The figures therefore describe how many events a future time-to-event study would plan for under the stated assumptions; they are not enrollment totals. For the platinum-sensitive comparison, the illustration gives total event requirements for 80% and 90% power at 100%, 75% and 50% of the observed effect. For the platinum-refractory comparison, the same effect fractions and power levels are shown with their own event requirements.

The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The time-to-event calculations use the Schoenfeld approach. The original trial’s own sample-size assumptions and justification are not reported in the registry record. Smaller effects require more events to detect under the same planning framework, which is why planning for 75% or 50% of the observed effect calls for more events than planning for 100% of the observed effect.

The 80% power column represents the planned probability of detecting the specified effect under the calculation assumptions if that effect is present. The 90% power column uses a higher target probability and consequently requires a larger event count. These figures are planning illustrations based on an observed estimate, not observed power and not evidence about the trial’s probability of success after completion. The actual trial’s event accrual and analysis details are not reported in a way that allows those planning figures to be treated as its design requirements.

12. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Hazard ratio HR = 1.320, so ln(HR) = 0.2776.
2. d = 4 × (1.96 + 0.842)² / [ln(1.320)]² = 407.3, rounded up to 408 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 <- 1.32                              # 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(828139)
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 = 408 (80% power), 546 (90% power)
 75% of observed effect: events = 725 (80% power), 970 (90% power)
 50% of observed effect: events = 1,630 (80% power), 2,182 (90% power)
Simulated power with 408 events: 0.810 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

13. Analyses Reported by the Sponsor

The hazard ratios shown in the posted results are sponsor-reported estimates: 1.32 for the platinum-sensitive comparison and 1.51 for the platinum-refractory comparison. The same estimates appear with their 90% confidence intervals in the reported analyses. The registry record does not report p-values for these comparisons or name the method used to obtain the hazard ratios.

The planning illustrations use these observed hazard-ratio estimates as inputs, then show hypothetical event requirements for detecting the full estimate or a fraction of it. They are not an independent confirmation of the sponsor’s analysis, and they do not replace the sponsor-reported outcome estimates. In general, planning calculations based on an observed effect are sensitive to that estimate; a smaller assumed effect requires more events for the same target power.

The record indicates no discrepancy between the posted sponsor results and the extracted comparison figures. The sponsor’s reported analysis is therefore the source used for interpreting the hazard ratios. The registry record does not provide enough detail to reconstruct the sponsor’s model or independently reproduce the reported inference. Any alternative calculation from summary statistics would be an unadjusted re-calculation and would not supersede the sponsor’s analysis.

14. Multiplicity, Missing Data and Analysis Populations

The record includes a primary PFS outcome, secondary outcomes for overall survival, response rate and grade 3 through 5 adverse events related to study drugs, and six reported analyses. It presents separate PFS estimates for four platinum-status and treatment groups, as well as hazard-ratio comparisons within the platinum-sensitive and platinum-refractory populations. The registry record does not state which analyses were prespecified or whether the subgroup comparisons were intended as confirmatory or exploratory.

When several outcomes, groups or comparisons are examined, the chance of finding an apparently notable result by chance can increase. Statistical approaches to multiplicity may control that risk or define a hierarchy of tests. Whether any such approach was used is not reported in the registry record. The record also does not report a formal treatment-by-platinum-status interaction analysis, so differences between subgroup estimates should not be interpreted as evidence that the treatment effect varies by platinum status.

The analysis populations are not reported in the registry record. It is therefore unclear whether the reported estimates use all randomized participants, participants with an evaluable outcome, or another population. The registry record also does not report how missing outcome data, loss to follow-up, withdrawals, protocol deviations or deaths without prior progression were handled analytically. Although death appears in the PFS definition, the detailed rules for event classification and censoring are not given.

Interim analyses and stopping rules are not reported in the registry record. Nor are the original sample-size justification, statistical analysis plan, confidence-interval construction method, or procedures for adjusting comparisons reported. These are important omissions for assessing whether the reported estimates and uncertainty match the planned analysis. The available results can be described as posted, but the record alone cannot resolve these design and analysis questions.

15. 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: death - disease progression nos (9.2%), platelets (8.7%), death not associated with ctcae term - death nos (6.5%), dehydration (6.5%), dyspnea (shortness of breath) (6.5%).

The adverse-event summary lists 92 participants at risk in the ziv-aflibercept plus topotecan group, of whom 44 are recorded as affected by serious adverse events. In the topotecan group, 87 participants were at risk and 17 were recorded as affected by serious adverse events. The separate other-adverse-event figures list 91 affected in the combination group and 86 in the topotecan group, with at-risk counts of 92 and 87, respectively.

The listed serious terms and percentages include death due to disease progression NOS at 9.2%, platelets at 8.7%, death not associated with a CTCAE term—death NOS—at 6.5%, dehydration at 6.5%, dyspnea at 6.5%, and fatigue at 5.4%. The record does not specify the denominator or detailed coding basis for each term in the summary, so these percentages should be read as posted rather than recalculated or interpreted as rates on a different denominator.

These are descriptive adverse-event counts and terms. They do not by themselves establish that a study drug caused an event. The secondary outcome refers to grade 3 through 5 adverse events related to study drugs, but the registry’s adverse summary does not provide enough detail to treat every listed serious event as drug-related or to infer comparative safety from the counts alone. Death counts are not reported in the adverse group summary.

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

The registry results describe PFS in a randomized phase II comparison of topotecan with or without ziv-aflibercept, presented separately for platinum-sensitive and platinum-refractory participants. The posted medians and hazard ratios provide different summaries of time to progression or death, while the 3-month PFS estimates describe outcomes at a specified time. Confidence intervals show uncertainty around those estimates.

The results do not by themselves establish that one regimen is superior. The hazard-ratio coding convention is not stated, p-values are not reported, and the confidence intervals show varying levels of uncertainty. The primary outcome and subgroup summaries should be interpreted within the limits of the information in the registry record, including the absence of a reported analysis plan, analysis population and multiplicity strategy.

The outcome is PFS, not overall survival, quality of life or response alone. Although overall survival, response rate and adverse events are listed as secondary outcomes, the registry entries do not include results for overall survival or response rate. The adverse-event information is descriptive and does not establish causation or provide a complete account of risks.

More generally, a phase II trial can provide estimates useful for understanding a treatment comparison and planning later studies, but the registry record does not report the original design justification or how these findings were intended to inform subsequent research. The planning event counts are illustrations derived from the observed hazard ratios; they are not evidence that the trial had a particular power or that its results are definitive. The conclusions should remain limited to the outcomes and precision actually reported.

17. Why This Trial Matters Statistically

This trial is statistically notable because its PFS results are separated by platinum sensitivity rather than presented as a single pooled comparison. That organization makes the population context visible, but it also means the results involve multiple groups and comparisons. Without a reported interaction analysis or multiplicity plan, subgroup differences need cautious interpretation.

It also illustrates why medians and hazard ratios should not be treated as interchangeable. A median summarizes a point in the PFS distribution, while a hazard ratio compares event hazards across follow-up under a model. Both can contribute useful information, but each has assumptions and interpretive limits. Reporting confidence intervals alongside both types of estimate helps readers assess uncertainty rather than relying on point estimates alone.

The record provides 90% confidence intervals and no p-values for the hazard-ratio comparisons. This highlights the importance of distinguishing an estimated effect from the precision and inferential framework used to evaluate it. The direction of the hazard ratio cannot be assigned from the record’s treatment labels alone because the coding convention is not described.

Finally, the event-based planning illustrations demonstrate how assumed effect size affects future study requirements. Planning for a smaller fraction of the observed effect requires more events under the stated assumptions. These are prospective planning quantities, not a retrospective measure of trial performance. Together, these features make the record useful for discussing how endpoint summaries, subgroup presentation, uncertainty and planning assumptions shape interpretation of a clinical trial.

18. 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 leaves important trial details unavailable. Eligibility criteria, participant characteristics, randomization procedures, allocation concealment and the basis for platinum-status classification are not reported in the registry record. The treatment descriptions also stop before fully describing the subsequent treatment mentioned for participants with responsive or stable disease.

The statistical analysis plan, analysis populations, missing-data rules, censoring procedures and interim analyses are not reported in the registry record. The method for estimating the hazard ratios and confidence intervals is also not identified, and the event and treatment coding convention for the hazard ratios is not specified. These limitations prevent a full assessment of model assumptions, analysis choices and the handling of incomplete follow-up.

The results are divided into platinum-sensitive and platinum-refractory groups, but the registry record does not say whether these subgroup comparisons were prespecified or provide a formal interaction test. Adjustment for multiple outcomes and comparisons is not reported. The record also does not include p-values for the hazard-ratio comparisons, event counts for PFS, or numbers at risk over follow-up.

Safety information is limited to summary counts and selected serious adverse-event terms. The record does not provide enough detail to infer causality or to assess every adverse event by severity, timing or exposure. Results for the listed secondary outcomes of overall survival and response rate are not included in the registry entries. These gaps mean the reported findings should be read as a limited registry summary rather than a complete clinical and statistical report.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: S0802 - Topotecan With or Without Aflibercept in Treating Patients With Extensive-Stage Small Cell Lung Cancer — NCT00828139.
• Registry record last updated 2017-08-21; status last verified 2017-07.
• Results first posted 2016-04-04.
• 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.

22. Record Summary

NCT00828139 is a completed randomized phase II trial comparing topotecan with ziv-aflibercept against topotecan alone in platinum-treated extensive-stage small cell lung cancer. It was open-label and used a parallel-group design. The record lists enrollment of 189, with a start in 2009-05 and primary completion in 2012-09.

The primary endpoint was PFS, reported as median months with 90% confidence intervals. Results are separated by platinum sensitivity. Sponsor-reported hazard ratios were 1.32 for the platinum-sensitive comparison and 1.51 for the platinum-refractory comparison. The registry record does not report p-values or the statistical method for those comparisons.

The record also lists overall survival, response rate and drug-related grade 3 through 5 adverse events as secondary outcomes. Details needed to fully evaluate the statistical plan, analysis populations, missing data and multiplicity are not reported in the registry record.