Comparison of Two Combination Chemotherapy Regimens 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 Comparison of Two Combination Chemotherapy Regimens in Treating Patients With Extensive-Stage Small Cell Lung Cancer (NCT00057837), a randomized phase 2 study. The primary outcome is Proportion of Patients With Objective Response by Solid Tumor Response Criteria (RECIST).

2. Clinical Question

This is a regimen comparison, not a comparison of a treatment against no treatment. The two arms differed in the drugs and their sequence. The registry states that the trial was intended to compare effectiveness, while also noting that which combination was more effective was not known when the study was framed. Random allocation supports a comparative design, but a design alone cannot establish whether an observed difference is reliable. That requires an interpretable analysis, including a treatment contrast and its uncertainty. The posted arm-specific estimates describe observed response proportions; they do not by themselves establish a statistically supported difference or identify which regimen has a higher response probability in the wider population.

3. Trial Design

The registry describes a randomized, open-label, parallel-group phase II study sponsored by the Eastern Cooperative Oncology Group. It compared PET, a sequence of topotecan followed by etoposide and cisplatin with G-CSF, against PIE, a sequence using irinotecan and cisplatin followed by oral etoposide. The registry status is completed. The recorded start date is 2004-07-14, and primary completion is recorded as 2012-01.

The overall enrollment is listed as 140. For the primary response endpoint, each arm has 66 participants in the posted results. The endpoint population is described as treated and eligible patients only. The adverse-event record uses different at-risk counts: 69 for PET and 68 for PIE. These counts differ across recorded outcome populations; the registry record does not explain the participant flow that produced them. They should therefore be read in the context of the particular outcome rather than assumed to be a single common analysis population.

The parallel design assigns participants to one regimen rather than describing a crossover. Randomization can reduce systematic differences between groups at assignment, but the registry record does not report how the random sequence was generated, whether allocation was concealed, or whether other safeguards were used. Masking is listed as none. The record does not report the reasons for that choice or whether any outcome assessment was conducted with masking.

The primary endpoint is the proportion with objective response by RECIST, assessed every 6 weeks while on treatment and then at intervals that depend on time from study entry. Duration of response and overall survival are listed as secondary outcomes. Although the registry lists three outcome measures and two secondary outcomes, it posts no analysis for those secondary endpoints in the information available. The protocol details, planned statistical analysis, and sample-size justification 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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Only treated and eligible patients are included in this analysis.

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 Proportion of Patients With Objective Response by Solid Tumor Response Criteria (RECIST), measured over Assessed every 6 weeks while on treatment, and then every 3 months for patients < 2 years from study entry, every 6 months if patient is 2-3 years from study entry..

The primary endpoint is the proportion of treated and eligible participants with an objective response, defined as complete response plus partial response under RECIST. A complete response requires disappearance of all target and nontarget lesions. A partial response requires at least a 30% decrease from baseline in the sum of the longest diameters of target lesions, alongside the stated conditions concerning nontarget lesions and tumor-marker levels. The registry describes the result as a proportion of participants.

Response is a binary outcome at the participant level: each participant is classified as having an objective response or not having one for this endpoint. The proportion provides a direct description of how often the defined response was observed in the analysis population. It does not describe how long a response lasts, whether survival differs, or the magnitude of change for every participant. Those questions concern other outcomes; duration of response and overall survival are listed as secondary outcomes, but their results are not reported in the posted results.

The registry says response was assessed every 6 weeks while participants were on treatment, with follow-up intervals afterward depending on time since study entry. A proportion is useful for comparing response frequency across trial arms, but interpretation depends on consistent assessment and a clear analysis population. The registry does not provide enough detail to determine how missing or unevaluable response assessments were handled. The endpoint estimates should therefore be understood as the posted results for treated and eligible patients, not as a complete account of every enrolled participant.

7. Results Posted in the Registry

• PIE (Irinotecan/Cisplatin/Etoposide): 0.0% (0/66) versus 0.0% (0/66) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0); unadjusted two-proportion z-test p = n/a.

8. Statistical Interpretation

9. Statistical Methods Explained

Binary outcome and arm-specific proportion. Objective response is recorded as a proportion because each participant is categorized according to whether a complete or partial response was observed. The estimate is a summary of the fraction of the analyzed group meeting that definition. It is not a continuous measure of tumor change and does not capture response duration. The posted proportions are 0.697 for PET and 0.576 for PIE. Each is accompanied by a 90% confidence interval: 0.591 to 0.789 for PET and 0.467 to 0.679 for PIE.

What a confidence interval conveys. In general, a confidence interval describes the precision of an estimate under the assumptions of the method used to calculate it. Wider intervals indicate greater uncertainty about the underlying value; narrower intervals indicate more precision. A confidence interval is not a range that contains the true value with certainty, and it is not an interval for an arm-to-arm contrast unless the contrast itself was calculated. The registry labels the endpoint dispersion as a 90% confidence interval but does not report the method used to calculate those intervals. Therefore, a particular formula or set of assumptions for the posted endpoint intervals cannot be established from the record.

Arm-specific estimates are not a comparative test. A randomized trial generally seeks to compare outcomes between assigned groups. That comparison needs a defined contrast, such as a difference in proportions, and an uncertainty estimate for that contrast. A test of a null hypothesis may also be used, depending on the analysis plan. Here, the registry posts arm-specific response estimates but no interpretable between-arm analysis. The zero-valued fields in the comparison entry do not provide a meaningful estimate. A reader should not infer equality, superiority, or a p-value from those fields. The record does not report a sponsor analysis that could be used instead.

Randomization and masking. Random allocation is intended, in general, to make assigned groups comparable with respect to factors known and unknown at baseline, on average. It does not guarantee exact balance in a particular trial, and it does not remove concerns about missing outcomes or analysis choices. The record identifies allocation as randomized and the model as parallel. It lists masking as none. The randomization method, allocation concealment, and any masking of outcome assessment are not reported in the registry record. These omissions limit assessment of possible bias, although they do not by themselves demonstrate that bias occurred.

Planning illustrations. Any sample-size illustrations associated with the results are planning calculations, not observed trial findings and not observed power. The stated illustration assumptions use a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is defined as a fraction of the observed effect. Those calculations are hypothetical and depend on the assumed effect and method. They do not show what the trial was originally designed to detect. The original sample-size justification and design assumptions are not reported in the registry record. No planning values need be interpreted as the number enrolled or the actual amount of information collected.

Over-dispersion and model choice in general. Over-dispersion occurs when variability in observed data is greater than a statistical model assumes. In general, failing to account for over-dispersion can make standard errors too small and intervals too narrow. This is a general statistical consideration, not a documented issue in this trial. The registry does not report a regression model or an over-dispersion assessment for the response endpoint. Because the posted outcome is a participant-level proportion and no model details are recorded, no claim about model fit or over-dispersion is possible.

Outcome timing and follow-up. Response assessments were scheduled every 6 weeks while on treatment, followed by less frequent assessments based on time from study entry. Assessment timing matters because response classification depends on when and how measurements are made. The registry's description establishes the broad schedule but does not report adherence to it, the method of central review, or how discordant assessments were resolved. These details are not reported in the registry record.

Analysis populations and missing observations. The registry states that only treated and eligible patients are included in the primary analysis. The endpoint results show 66 participants in each arm, while total enrollment is 140 and adverse-event at-risk counts are 69 and 68. The record does not give a participant flow account, an intention-to-treat analysis, exclusions by reason, or a missing-data method. In general, an analysis restricted after randomization can differ from an analysis that follows all randomized participants, particularly if treatment or eligibility exclusions relate to outcomes. Whether that concern applies cannot be determined from the posted record.

10. Confidence Intervals and Precision

The arm-specific estimates are reported with 90% confidence intervals. For PET, the response proportion is 0.697, with an interval from 0.591 to 0.789. For PIE, it is 0.576, with an interval from 0.467 to 0.679. The intervals communicate uncertainty around each separate proportion. They do not directly quantify uncertainty in the difference between the regimens.

In general, precision depends in part on the amount of information in the analysis and on the variability of the outcome. The response analysis includes 66 participants per arm. The posted interval endpoints show that the estimates retain appreciable uncertainty; the point estimates should not be treated as exact population values. Since the interval calculation method is not reported in the registry record, its detailed assumptions cannot be assessed from the posted information.

Confidence intervals are often more informative than a point estimate alone because they show a range of values compatible with the data under the interval's assumptions. They do not mean every value in the interval is equally plausible, and they do not establish that a treatment effect is clinically important. In particular, separate intervals should not be treated as a substitute for an interval around a direct comparison. The registry does not report such a comparative interval or an interpretable test result.

As a result, the response proportions can be described for each arm, but the precision of the between-regimen contrast cannot be judged from the posted results. The registry also does not report endpoint results for duration of response or overall survival, so their precision cannot be assessed.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, objective response, and two secondary outcomes: duration of response and overall survival. It reports three outcome measures but no analyses for the secondary outcomes. The planned statistical approach to handling multiple outcomes or multiple analyses is not reported in the registry record. In general, when many hypotheses are examined, the chance of at least one apparently unusual result can increase unless the analysis plan accounts for multiplicity. Whether or how that issue was addressed in this trial cannot be determined from the record.

The primary endpoint population is described as treated and eligible patients only. The posted endpoint counts are 66 in each arm. Overall enrollment is 140, and the adverse-event at-risk counts are 69 for PET and 68 for PIE. The registry does not report a flow diagram or reasons for the differences among these counts. It also does not state how missing response assessments, unevaluable participants, or withdrawals were handled. The extent and treatment of missing outcome data are therefore not reported in the registry record.

The record does not describe whether an intention-to-treat analysis was performed, how eligibility was determined after randomization, or whether any sensitivity analyses assessed the impact of exclusions or missing data. These details matter because analysis-population decisions can affect the interpretation of a randomized comparison. No conclusion about the direction or magnitude of such an effect can be drawn from the available information.

The sample-size justification, interim analyses, stopping rules, and full statistical analysis plan are not reported in the registry record. The planning illustrations use stated hypothetical assumptions, but they are not evidence of the trial's original design rationale and should not be mistaken for the analysis actually conducted.

12. 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: neutropenia (45.6%), leukopenia (32.4%), transfusion: prbcs (26.1%), diarrhea w/o prior colostomy (25.0%), thrombocytopenia (21.7%).

The adverse-event record reports 48 participants with serious adverse events among 69 at risk in the PET group, and 51 among 68 at risk in the PIE group. It also records other adverse events in 66 of 69 PET participants and 67 of 68 PIE participants. These are counts in their recorded safety populations; they are not the same populations as the primary response analysis.

The registry lists the following leading serious adverse-event terms with their posted percentages: neutropenia, 45.6%; leukopenia, 32.4%; transfusion: PRBCS, 26.1%; diarrhea without prior colostomy, 25.0%; thrombocytopenia, 21.7%; and anemia, 20.3%. The registry does not identify these term percentages by treatment arm in the record, so they should not be assigned to PET or PIE individually.

Safety counts and term percentages describe recorded events; they do not establish that one regimen is safer than the other. A valid comparative safety interpretation would require clearly defined event populations, attribution and severity rules, exposure information, and an analysis of between-arm differences. Those details are not reported in the registry record. Death counts are not posted.

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

The posted response results show an objective-response proportion of 0.697 in the PET arm and 0.576 in the PIE arm, with a 90% confidence interval for each estimate. These values describe treated and eligible participants included in the respective endpoint analyses. They are useful as descriptive summaries of the recorded outcome, but they do not establish a reliable difference between the regimens.

The most important boundary is that no interpretable between-arm statistical analysis is reported. A comparison entry with zero-valued fields cannot be taken as evidence of no difference, and arm-specific confidence intervals are not a substitute for a direct contrast. The registry record therefore does not support a formal conclusion about comparative response. It also does not provide posted results for the listed secondary outcomes of duration of response and overall survival.

The record establishes the trial's randomized, open-label, parallel phase II design and identifies the two treatment sequences. It does not provide enough methodological detail to evaluate randomization procedures, the handling of missing response assessments, or the effects of restricting the primary analysis to treated and eligible participants. Those gaps limit how fully the response estimates can be interpreted.

The response endpoint is not a complete measure of clinical benefit. It records whether a response meeting RECIST criteria was observed; it does not capture its duration or answer the survival question. Nor can the safety counts alone establish a comparative safety profile. The registry's reported data should be read as a limited account of outcomes, not as a complete assessment of all potential benefits and harms.

No treatment recommendation follows from these registry results. They report response estimates and selected safety information, while leaving important comparative analyses and secondary outcome results unavailable. Any broader conclusion would require the relevant prespecified analysis and fuller outcome reporting, which are not reported in the registry record.

14. Why This Trial Matters Statistically

This trial illustrates why a randomized design and a descriptive outcome table are not enough to establish a treatment difference. Randomization defines a comparative framework, but readers also need a reported contrast and its uncertainty. The registry gives arm-specific response proportions and confidence intervals, yet no interpretable between-arm analysis. The distinction is central: two estimates can be compared descriptively, but a statistical claim about their difference requires analysis directed at that difference.

It also shows why analysis populations must be explicit. The study enrollment is 140, the primary endpoint counts are 66 in each arm, and safety populations have separate at-risk counts. These counts may reflect different outcome-specific populations, but the record does not describe participant flow or explain exclusions. Without that information, readers cannot fully assess who contributed to each result or how missing outcomes were addressed.

The response endpoint and the listed secondary outcomes answer different questions. A response proportion summarizes whether a defined tumor response was recorded. Duration of response and overall survival would address other aspects of outcome, but their results are not posted. A single endpoint cannot stand in for all dimensions of a trial's results.

Finally, planning assumptions and observed results must remain distinct. Sample-size illustrations based on an observed effect are hypothetical planning tools, not evidence of the trial's original design rationale or observed power. The original justification is not reported in the registry record. Taken together, these features make the study a useful example of how to separate design description, descriptive results, statistical inference, and missing reporting when reading a clinical trial record.

15. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves several trial-specific limitations unresolved. It does not report the random sequence-generation method or allocation concealment, so those aspects of the randomized design cannot be evaluated. Masking is listed as none, but the record does not say whether response assessment was independently reviewed or otherwise protected from knowledge of treatment assignment.

The primary response analysis includes treated and eligible patients, with 66 participants reported in each arm. Overall enrollment is 140, while safety at-risk counts are 69 for PET and 68 for PIE. The registry does not explain the participant flow, eligibility exclusions, treatment discontinuations, or reasons that outcome populations differ. It also does not report how missing or unevaluable response assessments were handled. These gaps prevent a full assessment of selection and missing-data concerns.

Most importantly for comparative interpretation, the registry reports arm-specific response proportions and confidence intervals but no interpretable between-arm analysis. The zero-valued comparison entry cannot be treated as a meaningful estimate. No sponsor analysis or p-value is posted to clarify the comparison. The confidence interval method is not stated, and the original sample-size justification and statistical analysis plan are not reported.

Duration of response and overall survival are listed as secondary outcomes, but their results are not posted. The safety record gives counts and a list of serious adverse-event terms, but does not provide arm-specific term percentages or a comparative safety analysis. The registry also does not report death counts. These limitations concern what can be concluded from the registry record; they do not establish that the unreported analyses were not conducted elsewhere.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Comparison of Two Combination Chemotherapy Regimens in Treating Patients With Extensive-Stage Small Cell Lung Cancer — NCT00057837.
• Registry record last updated 2023-07-05; status last verified 2023-06.
• Results first posted 2013-02-13.
• 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.

19. Record Summary

NCT00057837 is a completed randomized, open-label, parallel phase II study sponsored by the Eastern Cooperative Oncology Group. It enrolled 140 participants with extensive-stage small cell lung cancer and compared PET, a topotecan/etoposide/cisplatin sequence with G-CSF, against PIE, an irinotecan/cisplatin/etoposide sequence. The recorded start date is 2004-07-14, and primary completion is listed as 2012-01.

The primary outcome was objective response by RECIST among treated and eligible patients. The registry reports proportions of 0.697 for PET and 0.576 for PIE, each with a 90% confidence interval, and 66 participants per arm. No interpretable inferential comparison is posted. Duration of response and overall survival are listed as secondary outcomes, without posted analysis results.

Safety information includes serious and other adverse-event counts for separate at-risk populations and a list of serious adverse-event terms. The registry record does not report the original sample-size justification, detailed statistical plan, participant flow, or missing-data methods.