Tepotinib With Gefitinib in Participants With Locally Advanced or Metastatic NSCLC (INSIGHT): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tepotinib With Gefitinib in Participants With Locally Advanced or Metastatic NSCLC (INSIGHT) (NCT01982955), a randomized phase 1, phase 2 study. The primary outcome is Phase 1b: Number of Participants Experiencing at Least One Dose Limiting Toxicity (DLT).

2. Clinical Question

3. Trial Design

NCT01982955, titled INSIGHT, is recorded as a Phase 1b/2, multicenter, randomized, open-label, parallel trial. Its registry status is completed. Enrollment is listed as 88. The recorded study start is 2013-12-23 and primary completion is 2017-12-12. The sponsor is Merck KGaA, Darmstadt, Germany.

The early Phase 1b portion assessed tepotinib with gefitinib at tepotinib doses of 300 mg and 500 mg, with gefitinib 250 mg in each group. The principal Phase 1b endpoint was the count of participants experiencing at least one dose-limiting toxicity in Cycle 1. The registry describes a dose-limiting toxicity set that included participants who experienced a dose-limiting toxicity during Cycle 1 or received at least 80 percent of planned treatment doses during that cycle.

In Phase 2, participants with MET-positive, T790M-negative NSCLC were randomized between tepotinib 500 mg plus gefitinib 250 mg and pemetrexed with cisplatin or carboplatin. The registry summary describes 2:1 randomization for the combination versus chemotherapy. A separate single-arm cohort included participants with MET-positive, T790M-positive NSCLC who received tepotinib plus gefitinib. The randomized and single-arm groups answer different kinds of questions: a randomized comparison can estimate a between-group contrast, while the single-arm results describe outcomes in that cohort without a concurrent randomized comparator.

The study was open-label, so masking was not used. The registry describes 44 outcome measures and 41 posted secondary outcomes. Specific details of the randomization procedure, allocation concealment, and any protocol changes are not reported in the registry record. The Phase 1b primary endpoint is distinct from the later Phase 2 outcomes, which include objective response, electrocardiogram findings, and performance status.

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: A total of 18 participants were enrolled in Phase 1b part of the study and a total of 70 participants were enrolled in phase 2 part of the study. Participants enrolled in phase 1b were not eligible for randomization in phase 2.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Dose Limiting Toxicity (DLT) set included all participants who experienced a DLT during Cycle 1, or received at least 80 percent of all planned doses of treatment during Cycle.

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 Phase 1b: Number of Participants Experiencing at Least One Dose Limiting Toxicity (DLT), measured over Day 1 to Day 21 of Cycle 1 (each cycle is 21 days).

The primary endpoint counts participants who experienced at least one dose-limiting toxicity during Day 1 to Day 21 of Cycle 1, with each cycle defined as 21 days. A dose-limiting toxicity was an event judged related to study treatment by the investigator and/or sponsor under National Cancer Institute Common Terminology Criteria for Adverse Events Version 4.0. The definition included specified severe blood-count toxicities, febrile neutropenia, uncontrolled gastrointestinal toxicity, and other severe non-hematological adverse events, with stated exceptions.

This endpoint is used in early-phase dose assessment because it focuses on treatment-related toxicity during a defined initial treatment period. As a general statistical point, a binary endpoint such as whether a participant experienced at least one event gives a simple count of participants meeting the definition. It does not describe the number of events per participant, their duration, or all aspects of tolerability. Nor does a count by itself establish how precisely the underlying event frequency has been estimated.

The registry reports 0 participants with a dose-limiting toxicity in the 300 mg tepotinib plus gefitinib group and 0 in the 500 mg plus gefitinib group. The group sizes recorded for this endpoint are 3 and 12, respectively. These are Phase 1b results, not the randomized Phase 2 efficacy comparison. The registry record does not report a confidence interval, a formal comparison, or an analysis plan for this primary count.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either experienced at least one qualifying dose-limiting toxicity during the defined Cycle 1 window or did not. The registry reports counts rather than a modeled treatment effect. As a general statistical explanation, binary counts are most informative when read alongside the number of participants assessed and the endpoint definition. A count of zero is an observed result, not proof that the true event probability is zero.

The registry defines the dose-limiting toxicity population as including participants who experienced a dose-limiting toxicity during Cycle 1 or received at least 80 percent of planned doses during that cycle. That definition matters because an analysis population is not automatically identical to all enrolled participants. The registry record does not report the full disposition of participants for this endpoint, a missing-data method, or a confidence interval. Consequently, the posted primary count cannot be expanded into an estimate with a stated precision that the record does not give.

How to read a risk difference

A risk difference compares the proportion meeting a binary outcome in one group with the proportion in another. It is expressed as an absolute contrast, not a relative change. The comparison estimates in the posted results are recalculated, unadjusted summaries from the reported counts and group sizes. They are not reported sponsor analyses; the registry lists no reported analyses for these outcomes. For objective response, the recalculated risk difference is -0.118, with an interval from -0.376 to 0.139. The interval includes zero, indicating that the summary data are compatible with contrasts on either side of zero under the calculation used. The endpoint's arm percentages and the contrast interval should be read together rather than treating the point estimate as exact.

Confidence intervals describe uncertainty in an estimate under the statistical method used. As a general explanation, a wider interval indicates less precision, and an interval that includes the value representing no difference does not establish a clear between-group distinction. The objective-response arm estimates are accompanied by 90% confidence intervals in the registry. The recalculated comparison interval is presented separately; the registry record does not report a sponsor analysis for it. A confidence interval is not a range containing a fixed proportion of individual participant outcomes, nor does it resolve limitations in study design or measurement.

What unadjusted p-values can and cannot say

The posted comparison p-values are unadjusted. As a general statistical explanation, a p-value measures how unusual the observed data, or more extreme data, would be under a specified null model. It is not the probability that a treatment effect is absent, and it does not measure clinical importance. Unadjusted p-values also do not account for the possibility of false-positive findings when many outcomes or comparisons are examined. The registry record does not report a multiplicity adjustment or an inferential hierarchy for the secondary outcomes.

For binary endpoints, sparse counts can make estimates imprecise. As a general matter, when only a few participants meet an outcome definition, a small change in the observed count can materially affect the estimated contrast. This is relevant when interpreting the posted electrocardiogram findings and the performance-status outcome. No dispersion parameter is needed for a binary participant count in the way it may be needed for an over-dispersed count model. General explanation: over-dispersion matters when count data vary more than a model's assumed distribution allows, because standard errors can then be too small; the registry's primary endpoint is posted as a participant count, not as a rate model.

How to read the planning illustrations

The sample-size illustrations use the observed effect or a stated fraction of it, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. They are planning illustrations, not the trial's design justification and not observed power. For objective response, planning for 100% of the observed effect gives 267 participants per arm for 80% power and 357 participants per arm for 90% power. Planning for 75% of the observed effect gives 481 participants per arm for 80% power and 643 participants per arm for 90% power. These values depend on the assumed effect and statistical method; they should not be interpreted as evidence that the trial had a particular probability of detecting an effect. The registry record does not report the trial's own sample-size justification.

Recalculated comparisons are simple unadjusted calculations from posted summary statistics. The discrepancy flag is false, so there is no recorded discrepancy requiring a sponsor analysis to take precedence. Even without a discrepancy, the recalculations remain distinct from a reported trial analysis: the registry does not provide the analysis model, covariate adjustment, or full analysis population for these comparisons.

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: Tepotinib 300 mg + Gefitinib 250 mg", "Phase 1b: Tepotinib 500 mg + Gefitinib 250 mg")
n      <- c(3, 12)
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: Tepotinib 300 mg + Gefitinib 250 mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Phase 1b: Tepotinib 500 mg + Gefitinib 250 mg  12  0  0.0  0.0  26.5  0.0  24.2  0.2  24.7

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists many outcome measures, including 41 posted secondary outcomes, but does not report how multiplicity was handled. It does not report a hierarchy of hypotheses, a family-wise error strategy, or another adjustment for testing multiple outcomes. As a general statistical explanation, when multiple hypotheses are tested, the chance of at least one apparently unusual result can increase, even if no underlying effect exists. An unadjusted p-value for an individual secondary comparison therefore does not by itself account for the full set of analyses that may have been considered.

Missing-data handling is not reported in the registry record for the outcomes discussed. The record also does not specify whether the randomized analyses used an intention-to-treat population, a per-protocol population, or another analysis population. The primary Phase 1b endpoint does have a stated dose-limiting toxicity set definition, but that does not establish the analysis population used for Phase 2 outcomes. The endpoint-specific denominators should therefore be understood as the reported assessment groups, not as proof of a particular analysis convention.

The registry does not report interim analyses or stopping rules, nor does it provide details of how incomplete outcome assessments were treated. These omissions limit assessment of whether outcome availability could have affected the estimates. General explanation: missing outcomes can affect interpretation if their absence is related to prognosis, treatment, or the outcome itself; the direction and magnitude of such effects cannot be inferred without information on the missingness and the analysis approach.

The posted comparisons are recalculated, unadjusted summaries, and the registry lists no reported analyses for them. They are useful for describing the posted counts and uncertainty but do not replace a prespecified analysis plan. The registry record does not report whether the secondary analyses were prespecified, exploratory, or adjusted for baseline characteristics.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cellulitis (16.7%), pneumonia (16.7%), arthralgia (16.7%), back pain (16.7%), dyspnoea (16.7%).

The registry reports serious adverse event counts by group, alongside numbers at risk, and separately reports deaths. These are descriptive counts. In Phase 1b, serious adverse events were recorded for 4 of 6 at risk in the 300 mg tepotinib plus gefitinib group and 7 of 12 at risk in the 500 mg group. In Phase 2, the corresponding posted counts were 13 of 31 for the combination group, 8 of 23 for chemotherapy, and 5 of 15 for the T790M-positive single-arm cohort.

The registry also records deaths of 1, 3, 23, 20, and 10 across those respective groups. A death count in an adverse-event summary does not by itself establish that the death was caused by treatment. The registry information in the registry record does not specify causes or a causal attribution for those deaths. Serious adverse events are not interchangeable with all other adverse events; the record separately reports other adverse events for the groups.

Serious-event terms listed at 16.7% include cellulitis, pneumonia, arthralgia, back pain, dyspnoea, and dyspnoea exertional. The registry record does not provide additional context for interpreting these terms as comparative treatment effects. Differences in group size, phase, and cohort type also limit direct comparison of raw counts. These figures describe reported events, not a causal estimate of treatment-related risk.

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

The primary finding means that the registry recorded no qualifying dose-limiting toxicity among the participants in either Phase 1b dose group during the defined Cycle 1 observation period. It is a result about that endpoint, population, and time window. It does not mean that no other adverse events occurred: the adverse-event record lists serious and other adverse events. It also does not establish that the true chance of a dose-limiting toxicity is zero, particularly because the endpoint groups were limited in size and no confidence interval is reported.

The Phase 2 secondary findings describe outcomes in randomized groups and in a separate single-arm cohort. The objective-response percentages and recalculated comparison provide an estimate with considerable uncertainty; the interval for the risk difference spans zero. That does not prove the groups are equivalent. Equivalence requires a prespecified margin and an analysis designed to evaluate it, neither of which is reported in the registry record. Nor should an unadjusted p-value be interpreted as a measure of clinical importance.

The posted information does not establish that one treatment strategy works or fails. It does not provide a reported sponsor analysis for the listed comparisons, and it does not report multiplicity handling, missing-data methods, the Phase 2 analysis population, or a Phase 2 primary efficacy result in the registry record. Those limits constrain what can be concluded from the available summaries.

General explanation: a trial result is strongest when the endpoint, population, analysis, and uncertainty are all clear and aligned with the question being asked. Here the early-phase primary endpoint addresses initial dose-limiting toxicity, while the Phase 2 outcomes address response and other measures. The data should therefore be interpreted endpoint by endpoint, without treating an early safety count as evidence of comparative efficacy or treating exploratory secondary outcomes as a definitive overall conclusion.

14. Why This Trial Matters Statistically

INSIGHT illustrates why trial phase and endpoint purpose matter in statistical interpretation. Its primary endpoint is an early dose-limiting toxicity count from Phase 1b. The randomized Phase 2 portion addresses a different question in a defined MET-positive, T790M-negative population, with chemotherapy as the comparator. The single-arm T790M-positive cohort adds descriptive information but cannot supply the same comparative evidence as randomization.

The record also demonstrates how a point estimate can look more definite than the underlying evidence. A risk difference compresses two group results into one contrast, but its interval communicates the uncertainty around that contrast. For objective response, the recalculated estimate is accompanied by an interval spanning zero. The interval, group sizes, and absence of a reported sponsor analysis are essential context for reading that estimate.

Multiplicity is another important statistical issue. A registry can contain many outcomes, and unadjusted secondary analyses should not be treated as if each had been the sole primary question. When the testing hierarchy and adjustment strategy are not reported, selective emphasis on one secondary result can misrepresent the broader evidence. The posted results are best read as endpoint-specific descriptions rather than as a single conclusive test.

Finally, planning illustrations help show how assumed effect size and desired power relate to sample size, but they are not retrospective evidence of power and do not reveal the original design rationale. The registry record does not report the trial's own sample-size justification. Keeping planning calculations, observed results, and prespecified analyses distinct is central to transparent clinical-trial interpretation.

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 specific to the registry information constrain interpretation. The primary endpoint is reported for Phase 1b and measures qualifying toxicity during Cycle 1; it does not capture the full safety experience over treatment. The two Phase 1b endpoint groups contain 3 and 12 participants, and the record provides no confidence intervals or formal comparison for their primary counts.

The randomized Phase 2 results cover a defined MET-positive, T790M-negative population, while the T790M-positive cohort is single-arm. Outcomes across those populations should not be compared as though they came from one randomized contrast. The chemotherapy description includes pemetrexed with cisplatin or carboplatin, so the comparator encompasses alternative platinum regimens rather than a single stated regimen.

The registry lists numerous outcomes, but it does not report a multiplicity strategy, missing-data approach, interim analyses, analysis populations for the Phase 2 outcomes, or a sponsor-reported analysis for the posted comparisons. The recalculated summaries are unadjusted and do not account for information unavailable in the posted aggregate counts. The record also does not report the trial's sample-size justification, so planning illustrations based on observed effects should not be mistaken for the original design rationale.

Open-label conduct means masking was not used, as recorded in the registry. The registry entries do not establish whether this affected assessment or reporting. The available summary also does not provide enough detail to assess outcome-assessment completeness, protocol deviations, baseline comparability, or the reasons for discontinuation. These are limitations of what can be determined from the registry record, not evidence that a particular bias occurred.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Tepotinib With Gefitinib in Participants With Locally Advanced or Metastatic NSCLC (INSIGHT) — NCT01982955.
• Registry record last updated 2022-11-08; status last verified 2022-10.
• Results first posted 2020-07-23.
• 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

NCT01982955 is the completed INSIGHT Phase 1b/2 randomized, open-label, parallel trial of tepotinib plus gefitinib compared with pemetrexed and cisplatin or carboplatin in a Phase 2 population with MET-positive, T790M-negative locally advanced or metastatic NSCLC. A separate single-arm cohort included participants with MET-positive, T790M-positive disease. The registry lists enrollment of 88, a study start of 2013-12-23, and primary completion on 2017-12-12.

The primary endpoint was the number of Phase 1b participants with at least one dose-limiting toxicity during Day 1 to Day 21 of Cycle 1. The posted count was 0 in each of the 300 mg and 500 mg tepotinib dose groups. Secondary outcomes include objective response, electrocardiogram findings, and performance status. The registry record does not report sponsor analyses for the listed comparisons, multiplicity handling, missing-data methods, or the trial's sample-size justification. Recalculated comparisons and planning illustrations should be interpreted as unadjusted summaries and hypothetical planning aids, not as reported trial analyses or observed power.