A Study in Non Small Cell Lung Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study in Non Small Cell Lung Cancer (NCT01139775), a randomized phase 1, phase 2 study. The primary outcome is Phase 2: Progression-Free Survival Time.

2. Clinical Question

The study’s stated objective also included finding a LY2603618 dose that could be safely combined with standard doses of pemetrexed and cisplatin. The Phase 1 component addressed dose finding and pharmacokinetic outcomes, while the randomized Phase 2 component assessed progression-free survival and other tumor outcomes. The record’s primary endpoint results describe the randomized comparison; they should not be treated as an estimate from the Phase 1 dose-finding group.

3. Trial Design

This was a Phase 1/Randomized Phase 2 study, with randomized allocation, a parallel-group model, and no masking. The record describes participants with Stage IV nonsquamous non-small cell lung cancer receiving first-line palliative treatment. The sponsor was Eli Lilly and Company. The study began in 2011-02, had primary completion in 2013-05, and is listed as completed.

The Phase 1 component evaluated LY2603618 doses from 130 to 275 mg with pemetrexed and cisplatin. Its regimen description gives 21-day cycles, with pemetrexed and cisplatin on Day 1 and LY2603618 on Day 2. The Phase 2 combination group also received pemetrexed and cisplatin with LY2603618, while the comparator received pemetrexed and cisplatin without LY2603618. The record describes the first four cycles as 21-day cycles and says maintenance could continue in the Phase 2 groups, but the comparator arm’s maintenance description is incomplete in the registry record.

The primary analysis included all randomized Phase 2 participants. The posted arm summaries list 39 participants for the combination and 23 for pemetrexed plus cisplatin. The overall enrollment field reports 76 for the study, which includes the study as a whole rather than identifying a randomized Phase 2 analysis count. The registry record does not report the randomization method, allocation concealment, a sample-size justification, or a detailed statistical analysis plan. It also does not report whether an interim analysis was planned or performed.

The study’s status is completed, but the record’s primary endpoint timeframe extends to 6 months after the last participant entered treatment. The stated analysis population and endpoint definitions are important when interpreting the posted estimates: they describe the randomized Phase 2 comparison, not all study participants or the separate Phase 1 dose-finding work.

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: All randomized Phase 2 participants.

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 participants.

6. Primary Endpoint

The registry lists the primary outcome as Phase 2: Progression-Free Survival Time, measured over Randomization up to first date of PD or death from any cause (up to 6 months after the last participant entered treatment).

The primary endpoint is progression-free survival time, measured in months. It is defined as the time from randomization to the first documented objective progressive disease or death from any cause. This combines disease progression and death into an event, so the measure reflects the time until either one is first recorded. Time-to-event endpoints are generally useful when the timing of an event matters and participants may have different lengths of observation.

The registry record summarizes progression-free survival using Kaplan-Meier estimates and reports the median for each Phase 2 group with a 90% confidence interval. The median is the estimated time by which half of the group has experienced the defined event, subject to the time-to-event estimation method. A confidence interval describes uncertainty around an estimate; it is not a range containing every participant’s time to progression or death.

Participants without documented progression by the data cutoff were censored at their last objective progression-free disease assessment. Participants who started subsequent systemic anticancer therapy before progression were censored at their last such assessment before that therapy began. In general, censoring allows participants with incomplete event follow-up to contribute information up to their last qualifying observation. Its interpretation depends on the censoring rules and on assumptions about how censoring relates to the chance of the event.

The record states that the primary timeframe ran from randomization to the first date of progressive disease or death, with follow-up described as up to 6 months after the last participant entered treatment. This endpoint is distinct from tumor response, tumor-size change, and overall survival, which are listed as other outcomes.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Progression-free survival is a time-to-event outcome: each participant contributes a time from randomization until documented progression or death, or until a censoring time if neither event is recorded by the relevant observation point. The event definition combines two possible first events. This is different from analyzing a simple proportion with progression, because the timing of events and the amount of follow-up contribute to the estimate.

What does a Kaplan-Meier median represent?

In general, the Kaplan-Meier method estimates the probability of remaining event-free over time while accounting for censored observations. It uses event times and the participants still under observation at those times. The median is the estimated time at which the event-free probability reaches the halfway point. It is not the arithmetic average of observed times, and it does not describe a typical participant in every other respect. If follow-up is limited or many observations are censored, the estimate’s precision can be affected.

The registry reports the median progression-free survival separately for each randomized Phase 2 group, along with 90% confidence intervals. In general, a confidence interval expresses the uncertainty associated with an estimate under the method used to construct it. It does not mean that the true value has a 90% probability of lying in the particular interval under a frequentist interpretation. The record does not state the confidence-interval construction details beyond its 90% level.

Why do the censoring rules matter?

Censoring is used when the outcome event has not been observed by the participant’s last qualifying assessment or by the analysis cutoff. The record says participants without objective progression by the cutoff were censored at their last objective progression-free assessment. It also specifies censoring at the last qualifying assessment before subsequent systemic anticancer therapy when that therapy began before progression. In general, censoring preserves information up to the point observation ends, but it does not reveal what would have happened afterward. Time-to-event estimates rely on assumptions about censoring, and the registry record does not report diagnostic assessments of those assumptions.

The primary endpoint timeframe is randomization to progressive disease or death, described as up to 6 months after the last participant entered treatment. The record says the analysis population was all randomized Phase 2 participants. It does not provide a separate account of missing endpoint assessments, numbers censored, or reasons for loss of follow-up, so those features cannot be assessed from the stated endpoint information alone.

What does the Bayesian result add?

The record reports a Bayesian posterior probability of 0.96 for the primary comparison. In general, a Bayesian posterior probability expresses uncertainty about a hypothesis or parameter after combining a statistical model with prior information and observed data. Its meaning depends on what was modeled, the prior distribution, and the event or parameter to which the probability refers. The registry record does not state these details or a threshold for interpreting the value. It also does not report a hazard ratio for the primary endpoint. The posterior probability should therefore be quoted as reported, without converting it into a claim about the probability of a particular clinical outcome.

How should the other comparisons be read?

The secondary endpoint recalculations use unadjusted comparisons derived from posted summary statistics. For binary outcomes, a risk difference compares the observed proportions; for a continuous outcome, a mean difference compares the reported means. Confidence intervals describe uncertainty in those recalculated effects. These calculations do not replace the sponsor’s analyses, and they do not adjust for other outcomes or comparisons. A p-value is a measure of how compatible the data are with a specified null model under the test assumptions; it is not the probability that the null hypothesis is true or the probability that a result occurred by chance.

The sponsor’s reported secondary tests used Chi-squared methods for the two binary outcomes and Wilcoxon (Mann-Whitney) for change in tumor size. These methods do not necessarily test exactly the same quantity as a simple unadjusted recalculation from posted arm summaries. Any difference between the reported and recalculated values should be interpreted in light of that distinction, not as evidence that one calculation was performed incorrectly.

What are the planning illustrations?

In general, sample-size calculations describe how many participants a hypothetical study might need under specified assumptions. The registry’s planning illustration uses a two-sided significance level of 0.05, equal allocation, a normal approximation, and an effect set to a stated fraction of the observed effect. The listed values are planning illustrations based on the observed results, not the original trial’s design justification and not observed power. The registry record does not report the assumptions used by the original investigators to select the study enrollment.

For time-to-event outcomes, the planning note describes event totals rather than participants per arm; for other outcomes it describes participants per arm. In general, planning results are sensitive to the assumed effect and analysis conditions. An effect estimated in one study can be uncertain, so a planning illustration based on it should not be read as a guarantee of future precision or as evidence that the completed study was adequately or inadequately powered.

10. Analyses Reported by the Sponsor

For the primary endpoint, the registry reports arm-specific Kaplan-Meier medians and a Bayesian posterior probability of 0.96. It does not report a sponsor hazard ratio or a confidence interval for a comparative progression-free survival effect. The Bayesian result is the sponsor-reported comparative statistic available in the registry record, but the record does not specify the hypothesis or threshold needed to interpret it more fully. There is no conflicting recalculated primary comparison in the posted summary that would replace this reported analysis.

For overall tumor response rate and clinical benefit rate, the sponsor reports Chi-squared p-values of 0.082 and 0.095, respectively. The unadjusted recalculations from the posted summaries give p-values of 0.082 and 0.095. For change in tumor size, the sponsor reports a Wilcoxon (Mann-Whitney) p-value of 0.492, while the unadjusted mean-based recalculation gives 0.127. The Wilcoxon method and a comparison of means address different aspects of the data, so the latter is not a substitute for the sponsor’s test.

These are not findings of a discrepancy requiring the recalculations to take precedence. The sponsor’s analyses remain the relevant reported analyses, especially where their method differs from a simple summary-statistic recalculation. The recalculations are useful as transparent, unadjusted checks, but they do not reproduce the full analysis plan or account for any unreported modeling choices. In general, the sponsor’s own analysis takes precedence when a discrepancy is identified; the registry record does not flag a discrepancy for this study.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 16 outcome measures and 13 posted secondary outcomes, as well as five reported analyses. It does not report a multiplicity adjustment, a testing hierarchy, or a prespecified approach to controlling error across outcomes. In general, testing several outcomes without adjustment can make isolated small p-values less persuasive because the analysis involves multiple opportunities to observe a result that appears notable under a null model. This does not establish that the analyses were inappropriate; it means the registry record does not provide enough information to assess how multiplicity was handled.

The primary population is described as all randomized Phase 2 participants. That is useful, but the record does not explicitly state whether all analyses followed an intention-to-treat principle, how participants were assigned for analysis if treatment was discontinued, or how protocol deviations were handled. A named analysis population is not, on its own, a complete account of these decisions.

For progression-free survival, the record gives censoring rules for participants without documented progression by the cutoff and for those who began subsequent systemic anticancer therapy before progression. These rules identify how some incomplete follow-up was treated. The record does not report the number of censored observations, the reasons for incomplete outcome data, missingness for the secondary endpoints, or sensitivity analyses for missing data. It also does not report whether missing assessments were imputed for outcomes other than progression-free survival.

Sample-size justification and interim analyses are not reported in the registry record. The planning illustrations based on observed effects do not supply that missing design information. Without a detailed statistical analysis plan, it is not possible to determine from the registry record which analyses were planned in advance, whether additional analyses were undertaken, or how the sponsor addressed the joint interpretation of primary and secondary findings.

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: pulmonary embolism (12.8%), anaemia (7.1%), ileus (4.5%), nausea (4.5%), vomiting (4.5%).

The safety summary lists participants at risk and participants affected by serious adverse events and other adverse events in each reported group. In Phase 1, the record lists 14 at risk and 1 affected by a serious adverse event, with 14 affected by other adverse events. In Phase 2, the combination group lists 39 at risk and 16 affected by a serious adverse event, with 37 affected by other adverse events. The pemetrexed-and-cisplatin group lists 22 at risk and 6 affected by a serious adverse event, with 22 affected by other adverse events.

These are participant counts as recorded, not estimates of comparative risk adjusted for follow-up or other factors. The registry record does not provide a detailed exposure-adjusted analysis or enough information to establish causal attribution from these counts. Death counts are not reported in the safety summary.

The serious adverse event terms listed include pulmonary embolism, 12.8%; anaemia, 7.1%; ileus, 4.5%; nausea, 4.5%; vomiting, 4.5%; and pyrexia, 4.5%. The registry information does not give further context for these percentages in the stated summary, so they should be read as listed rather than converted into comparative rates. Safety counts describe events recorded during the study and do not, by themselves, establish whether an event was caused by a particular treatment.

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

The primary endpoint results describe progression-free survival in the randomized Phase 2 groups. The combination group’s reported median was 4.7 months, and the pemetrexed-and-cisplatin group’s reported median was 1.5 months. Their 90% confidence intervals express uncertainty in each arm-specific median. The registry also reports a Bayesian posterior probability of 0.96 for the primary comparison, without defining the precise hypothesis or decision threshold behind that value.

These statistics do not establish what happened to every participant, and the medians should not be interpreted as guaranteed durations for individuals. They also do not provide a direct estimate of the between-group difference in progression-free survival on a hazard-ratio scale. No hazard ratio for the primary endpoint is reported in the registry record. The interpretation is bounded by the endpoint definition, the censoring rules, and the stated follow-up timeframe.

The study does not answer every question about treatment. The primary endpoint is not overall survival, and although overall survival is listed as a secondary outcome, its numerical result is not reported in the registry record. Nor do the secondary tumor outcomes answer questions about outcomes beyond their own definitions and timeframes. The safety summary provides event counts and listed serious-event terms, but it does not establish causation or a complete comparative safety profile.

In general, statistical evidence is not the same as a treatment recommendation. The results are observations from a particular randomized Phase 2 study with a Phase 1 component, in the population and treatment setting described by the record. They should not be generalized automatically to other populations, regimens, settings, or follow-up periods. The registry record also leaves important methodological details unreported, including the full analysis plan and multiplicity handling. Those limits affect how confidently the posted estimates can be interpreted, but they do not change the values as recorded.

14. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between a randomized comparison and a collection of arm-level summaries. Random allocation in the Phase 2 component makes the two treatment groups relevant to a comparative question, but the primary endpoint is reported as separate Kaplan-Meier medians and a Bayesian posterior probability rather than a conventional hazard ratio with a confidence interval. Readers need to distinguish the descriptive arm estimates from a direct treatment-effect estimate.

The endpoint also illustrates why time-to-event methods are used when participants may have different follow-up and some have no recorded event by the cutoff. Censoring rules determine how those participants contribute information. Here, the registry specifies censoring for assessments ending without documented progression and for subsequent systemic anticancer therapy begun before progression. Those details help define the result, while the lack of reported censoring counts and sensitivity analyses limits further assessment.

The study also demonstrates that secondary endpoints need not agree in their statistical meaning. Tumor response, clinical benefit, tumor-size change, and progression-free survival summarize different features of the observed outcomes. The sponsor’s Wilcoxon test for tumor-size change and an unadjusted recalculation based on means are not interchangeable analyses. Similarly, unadjusted secondary p-values do not automatically account for the number of outcomes examined.

Finally, planning illustrations based on observed effects are not a substitute for a prospective sample-size rationale. They can show how assumed effect size relates to hypothetical planning requirements, but they do not reveal the design assumptions used before the trial or the study’s observed power. The registry record does not report the original sample-size justification, making that distinction especially important when reading planning information alongside completed-trial results.

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.

Several limitations are specific to the information reported for this study. The registry describes a Phase 1/Randomized Phase 2 design, so the full enrollment field and randomized Phase 2 arm summaries refer to different parts of the study. The record identifies the primary analysis population as all randomized Phase 2 participants, but does not provide a complete accounting of analysis populations for every secondary endpoint.

The Phase 2 regimen description records a change around 25 Oct 2012 in the combination arm’s LY2603618 schedule. The record does not explain the reason for that change or provide a separate outcome analysis by regimen version. The comparator’s maintenance-therapy description is incomplete in the registry record, which limits assessment of the full treatment schedules from this information alone.

The primary endpoint results include arm medians and confidence intervals, plus a Bayesian posterior probability, but no primary hazard ratio or comparative confidence interval. The registry does not describe the Bayesian model, prior assumptions, or decision threshold. Censoring rules are stated, but censoring counts, missing-data details, and sensitivity analyses are not reported in the registry record.

The record does not report a complete statistical analysis plan, multiplicity handling, an original sample-size justification, or interim-analysis information. Numerical results are also not reported for some listed outcomes, including overall survival and the pharmacokinetic outcomes. The safety summary provides affected-participant counts and selected serious adverse event terms, but not exposure-adjusted comparisons or causal attribution. These gaps constrain interpretation of the posted findings without changing what the registry reports.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study in Non Small Cell Lung Cancer — NCT01139775.
• Registry record last updated 2018-05-18; status last verified 2018-05.
• Results first posted 2018-05-18.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT01139775 is a completed Phase 1/Randomized Phase 2 study of LY2603618 with pemetrexed and cisplatin in non-small cell lung cancer. The randomized Phase 2 comparison assessed pemetrexed plus cisplatin with or without LY2603618. It used a parallel-group design and no masking.

The primary endpoint was progression-free survival from randomization to documented objective progressive disease or death from any cause, analyzed with Kaplan-Meier estimates. The registry reports arm-specific medians and a Bayesian posterior probability for the primary comparison. Secondary outcomes include overall survival, tumor response, tumor-size change, clinical benefit, and pharmacokinetic measures, although numerical results are not reported for every listed outcome.

The record also reports serious and other adverse event counts. Important analysis details—including multiplicity handling, a full statistical analysis plan, and the original sample-size justification—are not reported in the registry record.