A Study of Necitumumab in the First-Line Treatment of Stage IV Squamous Non-Small Cell Lung Cancer (NSCLC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Necitumumab in the First-Line Treatment of Stage IV Squamous Non-Small Cell Lung Cancer (NSCLC) (NCT01763788), a randomized phase 1, phase 2 study. The primary outcome is Phase 1b: Number of Participants With Dose Limiting Toxicities (DLTs).

2. Clinical Question

The study addressed two related but distinct questions. The Phase 1b portion was intended to assess tolerability of the combination and determine a recommended dose for the subsequent Phase 2 portion. Its posted primary endpoint was a count of participants with dose-limiting toxicities during Cycle 1. The Phase 2 portion was intended to evaluate the combination’s efficacy in the same first-line setting; the posted secondary outcomes included objective tumor response and several other clinical and patient-reported measures.

The Phase 2 comparison was between gemcitabine and cisplatin plus necitumumab, abbreviated GC+N in the results, and gemcitabine and cisplatin alone, abbreviated GC. The posted objective response results describe the percentage achieving complete or partial response. Patient-reported outcomes included EuroQol-5D-3L index and visual analog scale scores, as well as the Lung Cancer Symptom Scale. These address different aspects of the study question and should not be treated as interchangeable measures.

The registry record identifies randomization and a parallel design, but provides no detailed account of the randomization procedure. The study was open-label, so participants and study personnel were not masked according to the record. The posted results can describe the observed groups and their precision, but the record does not report enough detail to resolve every question about conduct, analysis populations, or how the Phase 2 endpoints were prioritized.

3. Trial Design

This was an open-label, multicenter Phase 1b/2 study with randomized allocation and a parallel design. The registry record identifies Eli Lilly and Company as sponsor. The study status is completed; the recorded start date was 2013-05-07 and primary completion was 2017-06-28. Overall enrollment was 192 participants.

The two phases had different purposes. Phase 1b used dose escalation of necitumumab with gemcitabine and cisplatin. The study description says this phase was intended to investigate tolerability and determine the recommended dose for Phase 2. The posted Phase 1b cohorts contained 3 and 6 participants for the dose-limiting toxicity result. The registry record lists the primary endpoint population as all enrolled participants who received at least one dose of study drug during Phase 1b.

Phase 2 was randomized and compared gemcitabine and cisplatin plus necitumumab with gemcitabine and cisplatin alone. The posted objective response results include 90 participants in the combination group and 91 in the chemotherapy group. These are the denominators reported for that outcome; they should not be taken as a complete account of participant flow or of the analysis population for every endpoint. Detailed participant flow and the numbers randomized, treated, and analyzed for all outcomes are not reported in the registry record.

Open-label design is relevant when considering outcomes that can be influenced by knowledge of treatment assignment, including clinical assessments and patient-reported measures. The registry record does not describe blinding of response assessment or other safeguards against assessment bias. It also does not report a detailed randomization method, allocation concealment procedure, or protocol deviations. These omissions limit the information available for evaluating the conduct of the comparison, rather than changing the numerical results that are posted.

The treatment descriptions identify necitumumab, gemcitabine, and cisplatin, but the available arm descriptions are incomplete. The full dosing and administration details are not reported in the registry record in the material summarized for this analysis. The trial’s own sample-size justification and detailed statistical analysis plan are also not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Completers included participants who died from any cause and participants who were alive and on study (but off treatment) at study conclusion.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All enrolled participants who received at least one dose of study drug during study Phase 1b.

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 enrolled participants who received at least one dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Phase 1b: Number of Participants With Dose Limiting Toxicities (DLTs), measured over Day 1 to Day 21 in Cycle 1 (Up To 21 days).

The registry-listed primary endpoint was the number of Phase 1b participants with dose-limiting toxicities. It counted participants, rather than counting every toxicity event separately, during Day 1 to Day 21 in Cycle 1, described as up to 21 days. A participant was considered to have a dose-limiting toxicity when a specified event occurred in that window and was considered definitely or probably related to necitumumab and/or gemcitabine-cisplatin chemotherapy.

The definition included selected severe blood-related toxicities, Grade 3 or higher nonhematologic toxicity with listed exclusions, and toxicity leading to omission of necitumumab on specified treatment days. The criteria were graded according to NCI-CTCAE version 4.0. These rules make the endpoint a protocol-defined early tolerability measure, not a count of every adverse event experienced over the entire study.

Counting participants with a qualifying toxicity is useful in dose escalation because it provides a direct summary of how often a prespecified safety threshold was reached during the initial cycle. It can inform dose selection alongside other safety and treatment information. It does not, by itself, summarize later toxicities, the full range of adverse events, or treatment benefit. Nor does a low observed count establish that uncommon events cannot occur: estimates based on small cohorts are imprecise.

The posted primary endpoint results were zero participants in Phase 1b Cohort 1, with 3 participants, and zero in Cohort 2, with 6 participants. The Phase 2 outcomes are reported separately and should not be mistaken for results of this Phase 1b primary endpoint. A Phase 2 primary endpoint is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help distinguish what the posted estimates say from what they cannot establish. The study includes a count endpoint in a small dose-escalation phase, percentages and continuous outcomes in a randomized phase, and planning illustrations based on observed effects. Each type of result has a different interpretation.

What does a confidence interval communicate?

In general, a confidence interval expresses the precision of an estimate under the assumptions of the statistical method. A narrower interval indicates greater precision than a wider interval on the same measurement scale. It is not a range that guarantees the true value lies inside it, and it does not mean that every value in the interval is equally plausible. For the Phase 2 response percentages, the posted 95% intervals are 40.8% to 61.4% for the combination group and 12.5% to 29.2% for chemotherapy alone. These intervals reflect uncertainty around the group estimates, while the comparison interval addresses uncertainty in the contrast between groups.

For the continuous quality-of-life comparisons, the intervals for the mean differences include zero. In general, zero represents no mean difference on the scale used. Including zero means the posted data do not rule out no difference under the interval method; it does not establish that there is no difference. A broad interval can remain compatible with meaningfully different effects in either direction. The study’s meaningful-difference thresholds are not reported in the registry record, so the clinical importance of these differences cannot be assessed from the reported intervals alone.

How should p-values be read?

In general, a p-value measures how incompatible observed data are with a null hypothesis, given the chosen model and its assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. The posted unadjusted p-value for objective response is p < 0.001; the unadjusted values for the EuroQol-5D-3L index and visual analog scale comparisons are 0.725 and 0.297. They should be considered alongside the estimated effects and confidence intervals, not used alone.

The record does not describe the detailed statistical tests, model specifications, covariate adjustments, or any prespecified testing hierarchy for these Phase 2 outcomes. Therefore, the results can be described as the posted unadjusted comparisons, but the exact analysis procedures cannot be reconstructed from the registry record.

Why do small cohorts limit a zero event count?

The Phase 1b result is a count of participants meeting a predefined dose-limiting toxicity rule, not a rate estimated with a posted confidence interval. Zero events in a cohort means no qualifying event was observed among the listed participants. In general, small samples have limited ability to detect uncommon outcomes and produce imprecise estimates of event frequency. The Phase 1b cohorts contained 3 and 6 participants, so their counts are descriptive evidence about those participants and the specified observation window, not a precise estimate of broader safety experience.

The endpoint definition also matters. Only events meeting the specified grading, timing, and attribution criteria counted as dose-limiting toxicities. Events outside the Cycle 1 window, events not meeting the severity rules, and adverse events not considered related under the definition do not become primary endpoint events merely because they occurred during the study. The registry record’s separate adverse-event summaries cover a broader safety context.

What are the planning illustrations?

The sample-size illustrations are calculations based on the observed effect, not the trial’s own design justification and not observed power. Their assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They show participants per arm for 80% or 90% power when planning for 100%, 75%, or 50% of the observed effect. For the objective response comparison, the illustrations range from 39 participants per arm for 80% power when planning for 100% of the observed effect to 168 when planning for 50% of it; the corresponding 90% illustrations are 51 and 225. The intermediate 75% planning values are 72 and 96 participants per arm for 80% and 90% power.

For the continuous outcomes, the illustrations are much larger because the observed mean differences are small relative to the reported variability. For the index-score comparison, the 100% observed-effect illustrations are 4,248 participants per arm for 80% power and 5,686 for 90% power. For the visual analog scale comparison, they are 498 and 666 participants per arm, respectively. Planning for 75% or 50% of the observed effect produces still larger illustrations: for the index score, 7,551 and 16,989 for 80% power, and 10,109 and 22,744 for 90%; for the visual analog scale, 885 and 1,990 for 80%, and 1,184 and 2,664 for 90%.

In general, reducing the effect size one wishes to detect increases the sample size required under otherwise fixed assumptions. These calculations are sensitive to the observed effect, which can be unstable, especially when estimates are imprecise. They do not show that the completed study had a particular probability of detecting an effect. The trial’s own sample-size justification and assumptions are not reported in the registry record.

Why does over-dispersion matter?

In general, over-dispersion occurs when observed variability exceeds what a statistical model assumes. It can make standard errors too small and confidence intervals too narrow if ignored. Over-dispersion is especially relevant to some count-data models, but the posted primary endpoint is presented as participant counts and the registry record does not report a model for it. There is no basis in the registry record to claim that an over-dispersion adjustment was used or required for the analyses described.

Finally, the registry record does not describe the handling of missing endpoint values, censoring rules for time-to-event outcomes, or a prespecified statistical analysis plan. General statistical principles can clarify how methods work, but they cannot substitute for those study-specific details.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Phase 1b: Cohort 1", "Phase 1b: Cohort 2")
n      <- c(3, 6)
events <- c(0, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Phase 1b: Cohort 1  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1b: Cohort 2  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0

11. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a statistical analysis plan describing how multiple outcomes were prioritized or how the family of statistical tests was controlled. In general, testing many outcomes increases the chance that at least one result will appear unusual even when there is no underlying effect. A multiplicity procedure can control that chance, but it may also change the threshold or interpretation for individual comparisons. No such procedure is reported for the posted Phase 2 comparisons, which are identified as unadjusted.

The study lists numerous secondary outcomes, including response, progression-free survival, treatment failure, patient-reported measures, and pharmacokinetics. A Phase 2 primary endpoint is not reported in the registry record. Without a reported hierarchy, it is not possible to determine which outcome was intended to carry the main confirmatory claim in Phase 2, or whether secondary outcomes were formally exploratory. The posted unadjusted p-values should therefore not be treated as if they automatically account for all outcomes assessed.

Missing-data handling is also not reported in the registry record. This matters because the posted Phase 2 quality-of-life results use 77 participants in the combination group and 61 in the chemotherapy group, while the objective response results use 90 and 91. Different outcome-specific denominators can arise for many reasons, but the record does not explain the reasons in this case. It does not state whether missing questionnaires were excluded, imputed, or handled with a model, nor does it report sensitivity analyses. The reported means consequently describe the participants included for those measures; the record does not establish how those results relate to participants without observed measurements.

The Phase 1b primary endpoint specifies its population as all enrolled participants who received at least one dose of study drug during Phase 1b. Equivalent analysis-population definitions for the Phase 2 outcomes are not reported in the registry record. Participant flow, exclusions, protocol deviations, and the handling of deaths or other intercurrent events for each endpoint are also not described in the record. These are important for understanding how the analyzed groups correspond to randomized participants.

In general, a prespecified analysis plan and transparent missing-data rules help readers assess whether reported estimates answer the original study questions. Without those details, the numerical results remain interpretable as posted estimates, but uncertainty about selection, multiplicity, and analysis choices should remain part of their interpretation.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: tumour pain (33.3%), febrile neutropenia (6.7%), pneumonia (6.7%), decreased appetite (3.3%), lung infection (3.3%).

The registry record reports serious adverse-event summaries for the Phase 1b cohorts and the Phase 2 groups. For Phase 1b, serious adverse events affected 1 participant in Cohort 1, among 3 at risk, and 0 participants in Cohort 2, among 6 at risk. In Phase 2, serious adverse events affected 32 participants among 90 at risk in the GC+N group and 22 among 91 at risk in the GC group. Deaths recorded in the Phase 2 safety summaries were 63 in GC+N and 74 in GC. These are separate safety counts and should not be treated as estimates of treatment effect without further context.

The summaries also report other adverse events affecting all participants at risk in each group. This broad category is not interchangeable with serious adverse events. A count of participants affected does not describe the number of events per participant, their timing, duration, severity distribution, or attribution to treatment. The registry record does not provide the context needed to determine why deaths occurred or whether they were considered treatment-related.

Serious adverse-event figures are descriptive and should be read alongside the study design and the available participant denominators. Phase 1b cohorts are small and serve a different purpose from the randomized Phase 2 groups. The record’s safety counts do not by themselves establish comparative risk, and the reported summaries should not be interpreted as clinical advice.

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

The posted results describe what was observed in this study, not a universal prediction for every participant or setting. In Phase 1b, no participant in either of the two small cohorts met the primary endpoint definition for a dose-limiting toxicity during the specified first-cycle period. This is a narrow statement about a defined endpoint and window. It does not mean that no adverse events occurred, that later-cycle toxicity was absent, or that the combination has no safety risks.

In Phase 2, the objective response percentages and their confidence intervals describe the observed randomized groups. The unadjusted comparison is statistically distinct from the continuous quality-of-life comparisons, whose intervals include zero. Neither the p-values nor the intervals alone establish the clinical importance of an outcome. The registry record does not report thresholds for meaningful change on the patient-reported scales, a Phase 2 primary endpoint, or adjusted analyses that would place all secondary findings within a single confirmatory framework.

General statistical interpretation distinguishes evidence of a difference from proof of a difference, and absence of strong evidence from proof of no difference. A wide interval may include both small and larger effects. An unadjusted p-value does not communicate how large or important an effect is, and a confidence interval does not account for unreported sources of bias or missing data. The open-label design and incomplete reporting of analysis details are relevant when assessing how confidently the results can be generalized beyond the recorded groups.

The study record also describes a completed trial with overall enrollment of 192, but the denominators vary by outcome. Those outcome-specific counts should be retained when reading each estimate rather than treating every result as if it came from the same analysis population. Additional details about participant flow, missingness, endpoint procedures, and the statistical plan are not reported in the registry record.

Accordingly, the results support a careful description of the reported outcomes and their statistical uncertainty. They do not, on their own, settle broader questions about comparative treatment effects, the safety profile across longer exposure, or how the findings apply beyond the study population. No medical recommendation follows from these statistics alone.

14. Why This Trial Matters Statistically

This trial illustrates why a clinical study can require different statistical questions at different stages. Phase 1b used dose escalation and a narrowly defined first-cycle toxicity endpoint. Phase 2 used randomization to compare a combination regimen with chemotherapy alone and recorded outcomes ranging from tumor response to patient-reported measures. The results therefore need to be interpreted by phase and endpoint rather than compressed into one overall success-or-failure statement.

The small Phase 1b cohorts make a useful teaching example of the difference between an observed count and a precise risk estimate. A zero count is straightforward to report, but the small denominator means it cannot establish that the event is rare or absent in a larger population. The explicit endpoint window and toxicity criteria also show why endpoint definitions matter: what qualifies as an event, when it is counted, and how relatedness is assessed shape the result.

The randomized Phase 2 comparison demonstrates the value of reporting both group estimates and uncertainty. Response percentages are accompanied by confidence intervals, while the continuous patient-reported outcomes include mean differences and intervals. Together, these show that a p-value alone is insufficient: effect magnitude, precision, measurement scale, and relevance all matter. The different endpoint denominators also highlight why readers need clear analysis-population and missing-data reporting.

The record is also a reminder that unadjusted secondary findings need cautious interpretation when multiple outcomes are listed. Without a reported Phase 2 primary endpoint, analysis hierarchy, or multiplicity approach, readers cannot determine which result was intended to support a primary confirmatory conclusion. The study-specific sample-size rationale is not reported, while the posted planning illustrations are based on observed effects and should not be mistaken for the original design.

For biostatisticians and clinical researchers, the central lesson is to distinguish descriptive results, comparative estimates, and planning calculations. Each has a valid but different role, and each depends on transparent reporting of its assumptions and population.

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 trial-specific reporting gaps limit interpretation beyond the numeric estimates. The registry record identifies randomization, an open-label design, and parallel groups, but does not report the randomization sequence method, allocation concealment, or any blinding of outcome assessment. This is particularly relevant to outcomes involving clinical judgment or participant-reported responses. The record also does not describe protocol deviations or how they were handled.

The Phase 1b primary endpoint has very small cohort denominators, and its scope is limited to qualifying toxicities in Cycle 1. The posted result is not accompanied by an uncertainty interval. The endpoint definition sets thresholds and timing, but it does not summarize all safety events or later exposure. The Phase 2 safety summaries provide participant counts, but the registry record does not supply sufficient information about event timing, duration, attribution, or cause of death to support detailed comparative safety conclusions.

For Phase 2, the response comparison has posted percentages and confidence intervals, but the Phase 2 primary endpoint and detailed statistical analysis plan are not reported in the registry record. The comparisons are unadjusted, and no multiplicity strategy is reported. Results for several listed secondary outcomes are not reported numerically in the registry record. This limits any synthesis across the full outcome set.

The quality-of-life analyses have different denominators from the response analysis. Reasons for missing observations, methods for handling missing data, the analysis populations, and sensitivity analyses are not reported. The record also does not state a threshold for clinically meaningful change on the EuroQol measures. Therefore, the statistical estimates alone cannot determine whether scale changes were meaningful to participants or how missing responses might affect the interpretation.

Finally, the sample-size illustrations use observed effects and specified planning assumptions; they do not document the trial’s original sample-size justification. The full treatment descriptions are incomplete in the record available for this analysis. These limitations do not invalidate the posted numbers, but they constrain how fully the study’s design, analyses, and broader implications can be assessed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Necitumumab in the First-Line Treatment of Stage IV Squamous Non-Small Cell Lung Cancer (NSCLC) — NCT01763788.
• Registry record last updated 2019-10-25; status last verified 2018-12.
• Results first posted 2019-03-22.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

This completed study evaluated necitumumab with gemcitabine and cisplatin in first-line treatment of Stage IV squamous non-small cell lung cancer. It had Phase 1b dose escalation and a randomized, open-label Phase 2 comparison against gemcitabine and cisplatin alone. Overall enrollment was 192.

The registry-listed primary endpoint was the Phase 1b count of participants with dose-limiting toxicity during Cycle 1. No such events were recorded among 3 participants in Cohort 1 or 6 in Cohort 2. Phase 2 posted objective response percentages of 51.1% for GC+N and 20.9% for GC, with an unadjusted p < 0.001. Posted EuroQol-5D-3L comparisons had confidence intervals that included zero.

The Phase 2 primary endpoint, detailed statistical plan, multiplicity handling, and missing-data methods are not reported in the registry record. Secondary results should be interpreted with those reporting limits in mind.