Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Platinum-based Chemotherapy With or Without Pemetrexed in Patients With NSQ NSCLC: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Platinum-based Chemotherapy With or Without Pemetrexed in Patients With NSQ NSCLC (NCT04524689), a non-randomized phase 2 study. The primary outcome is All Cohorts: Number of Participants With Study Drug-Related Dose-Limiting Toxicities (DLTs).

2. Clinical Question

The safety run-in objective was to assess tolerability and determine recommended doses for the expansion part. The expansion objective was to assess antitumor activity across the described combinations and dose levels. These are distinct aims: early-cycle dose-limiting toxicities inform tolerability and dose selection, while antitumor activity is assessed through separate outcomes. The registry lists progression-free survival, disease control rate, duration of response, and objective response rate among secondary outcomes, with some outcomes limited to particular cohorts.

The study does not answer a randomized comparative question about whether one regimen is better than another. Its design instead organizes observations across non-randomized treatment cohorts. Differences between cohorts could reflect differences in regimen, dose, or participant characteristics, and cannot be interpreted as treatment effects from a randomized comparison. No analyzed primary endpoint result is reported in the registry record, so the clinical question about the observed frequency of dose-limiting toxicities remains unanswered by the posted results.

3. Trial Design

The registry record describes NCT04524689 as an open-label, Phase 2, parallel study with non-randomized allocation and no masking. It lists an enrollment of 57 and a condition of non-squamous non-small-cell lung cancer. The official title specifies CEACAM5-positive expression and advanced or metastatic disease. The study began on 2020-10-26, had a primary completion date of 2024-03-12, and is listed as terminated. The reason for termination is not reported in the registry record.

The treatment cohorts combined tusamitamab ravtansine with pembrolizumab. Some cohorts added cisplatin or carboplatin, and the quadruplet cohorts also included pemetrexed. The registry describes dose levels of 150 mg/m^2 and 170 mg/m^2 for tusamitamab ravtansine. It labels the doublet, triplet, and quadruplet cohorts separately, reflecting whether platinum-based chemotherapy and pemetrexed were included. Treatment was administered intravenously; the listed descriptions state that pembrolizumab was given before tusamitamab ravtansine every 3 weeks. Detailed schedules are not fully available in the registry record.

The safety run-in covered the doublet, triplet, and quadruplet cohorts and was intended to support dose selection for expansion. The expansion part was intended to assess antitumor activity, including participants treated at the recommended dose for expansion from the safety run-in. Because allocation was non-randomized, the cohorts should be read as separate treatment experiences rather than as randomized comparison arms. The registry does not report a randomization procedure, masking, a control regimen, or a protocol-based statistical comparison between cohorts.

The record lists safety outcomes across all cohorts and several efficacy and pharmacokinetic outcomes for selected cohorts. The primary outcome is the number of participants with study-drug-related dose-limiting toxicities during Cycle 1 Day 1 through Cycle 1 Day 21. Although adverse-event counts are posted by cohort, the registry reports no analysis for the primary endpoint. The status of the trial as terminated does not, by itself, explain why the study ended or whether planned objectives were completed.

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 57 participants were enrolled in either Part A \[doublet cohort (tusamitamab ravtansine + pembrolizumab)\], Part B \[triplet cohort (tusamitamab ravtansine + pembrolizumab + platinum-based chemotherapy i.e., cisplatin or carboplatin)\] or Part C \[quadruplet cohort (tusamitamab ravtansine + pembrolizumab + platinum-based chemotherapy + pemetrexed)\] of the study per Investigator's choice.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: DLT-evaluable population included participants who received 1 cycle with at least 80% of the intended dose for each study treatment of the combination during the safety run-in at both doses of tusamitamab ravtansine in Part A (doublet cohort), Part B (triplet cohort) or Part C (quadruplet cohort).

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-treated population included all participants assigned to study treatment and who actually received at least 1 dose of study treatment.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as All Cohorts: Number of Participants With Study Drug-Related Dose-Limiting Toxicities (DLTs), measured over Cycle 1 Day 1 up to Cycle 1 Day 21 (cycle duration=3 weeks).

The primary endpoint counts participants with study-drug-related dose-limiting toxicities, or DLTs, during Cycle 1 Day 1 through Cycle 1 Day 21, a cycle duration of 3 weeks. It is a participant count, not a measure of the total number of events or their severity distribution. The DLT definition includes specified blood-related and non-blood-related abnormalities, as well as other adverse events that the recruiting investigators and sponsor deemed dose-limiting.

The registry definition includes grade 4 neutropenia lasting 7 or more consecutive days; grade 3 to 4 neutropenia with fever of at least 38.5 degrees Celsius on more than 1 occasion, or with documented infection; and grade 3 or higher thrombocytopenia associated with clinically significant bleeding requiring intervention. Non-hematological criteria include grade 4 non-hematologic adverse events, with an exception for symptoms attributed to the underlying disease in the investigator’s judgment, and grade 3 or higher keratopathy. Other adverse events could also qualify if investigators and the sponsor deemed them dose-limiting, regardless of grade.

In general, an early-cycle DLT endpoint is used in a safety run-in to identify tolerability concerns and inform dose selection. It focuses attention on clinically important events occurring during a defined initial treatment period. The registry’s DLT-evaluable population includes participants who received one cycle with at least 80% of the intended dose for each combination treatment during the safety run-in, across the specified cohorts and both tusamitamab ravtansine dose levels. The registry record does not report the number of participants meeting that evaluability definition or their DLT count. Therefore, the endpoint definition is available, but its primary result cannot be interpreted from the posted record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on matching each result to its endpoint, population, and time window. In this study, the primary endpoint is a count of participants with study-drug-related DLTs in the initial treatment cycle. The registry defines a DLT-evaluable population for the safety run-in, but does not report the size of that population or the endpoint count. A definition alone does not establish how often DLTs occurred.

What does a participant count tell the reader?

A participant count records how many people met an endpoint definition. For a binary endpoint such as whether a participant had at least one qualifying event, a count is easiest to interpret alongside the relevant denominator. Counts from groups with different denominators are not directly comparable in magnitude. A proportion can describe the observed frequency within a group, but should not be calculated unless the correct numerator and denominator are available and clearly refer to the same endpoint and population. The registry’s posted adverse-event counts are not the primary DLT count, so they do not fill that gap.

In general, a confidence interval describes the uncertainty around an estimate under a statistical model. It does not guarantee that the true value lies within the interval, nor does it account automatically for design limitations such as non-random allocation. A narrow interval generally indicates greater precision under the model; a wide interval indicates less precision. No confidence interval for the primary endpoint is reported in the registry record. With small cohort counts, estimates are inherently sensitive to individual observations, and uncertainty should be made explicit rather than relying on a bare count.

Why does the non-randomized design matter?

Random allocation is one way to balance participant characteristics between treatment groups, on average, and supports causal comparisons when the study is otherwise appropriately conducted and analyzed. This trial is listed as non-randomized and open-label. Cohort membership is therefore not a randomized assignment. The groups also differ in their treatment combinations and dose levels. A crude difference between groups, even if calculated, would not isolate the effect of changing one drug or dose. The registry record lists the study as non-randomized and explicitly does not present the arms as a like-for-like control comparison.

Open-label conduct means masking was absent. In general, knowledge of assigned treatment can influence care, reporting, or assessment, although the extent depends on the outcome and its assessment procedures. The registry does not report details that would allow a reader to assess how outcome assessment was standardized or whether any independent review was used. No such procedure should be assumed.

How should the safety summaries be read?

The serious adverse-event fields provide a group count of participants affected and a group count at risk. These describe the recorded safety experience in each cohort, but they do not by themselves establish causality. An adverse event may occur during treatment without being caused by treatment. The study also lists deaths, but the posted counts do not state in this information whether each death was treatment-related, due to progression, or due to another cause. The registry’s serious-event term percentages should be read as the listed summaries; the information does not specify the denominators or provide a comparison framework for those percentages.

In general, adverse-event analysis may distinguish events by seriousness, severity, timing, and attribution. These concepts are not interchangeable. Seriousness refers to regulatory or clinical criteria for an event, while severity describes its intensity. A DLT is a protocol-defined category that may use severity grades and additional clinical criteria. The posted safety counts should not be treated as counts of DLTs unless the registry explicitly reports that mapping, which it does not.

What do planning calculations mean?

In general, a sample-size planning illustration asks how many participants or events might be needed under specified assumptions to detect an effect at a chosen significance level and power. Such an illustration is not an observed result, and it does not show the power actually achieved by a completed study. The registry’s planning assumptions describe a two-sided significance level of 0.05, equal allocation, and a normal approximation for the stated fraction of an observed effect. They are described as a planning illustration rather than as the trial’s own design. Because this study was non-randomized, a two-arm planning illustration should not be mistaken for a valid causal comparison between its cohorts. The registry does not report the trial’s sample-size justification or a corresponding primary endpoint analysis.

For time-to-event outcomes, methods generally account for event timing and censoring; a participant who has not experienced an event by the end of observation may contribute information up to that point. The registry lists progression-free survival as a secondary outcome in the doublet and quadruplet cohorts, but does not report its analysis details or results in the record information summarized. It is therefore not possible to assess censoring rules, proportional hazards assumptions, or the precision of a survival estimate from this information.

Likewise, the registry record does not report the statistical model, confidence-interval method, adjustment strategy, analysis software, or handling of missing outcome data for the primary endpoint. These methods should not be inferred from the endpoint definition or from the presence of descriptive adverse-event counts. The record’s stated details support a descriptive reading of the cohort data, not a reconstruction of an unreported analysis plan.

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("Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab", "Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab", "Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab ", "Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab ", "Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab ", "Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab ")
n      <- c(21, 2, 5, 1, 21, 3)
events <- c(0, 0, 0, 0, 0, 1)
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  Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab  21  0  0.0  0.0  16.1  -0.0  15.5  0.1  15.4
ARM  Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab  2  0  0.0  0.0  84.2  0.0  65.8  0.8  70.8
ARM  Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab   5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab   1  0  0.0  0.0  97.5  0.0  79.3  1.3  84.2
ARM  Tusamitamab Ravtansine 150 mg/m^2 + Pembrolizumab   21  0  0.0  0.0  16.1  -0.0  15.5  0.1  15.4
ARM  Tusamitamab Ravtansine 170 mg/m^2 + Pembrolizumab   3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6

11. Multiplicity, Missing Data and Analysis Populations

The study lists multiple outcomes across different treatment cohorts. These include the primary DLT count, treatment-emergent adverse events, progression-free survival for the doublet and quadruplet cohorts, disease control rate, duration of response, objective response rate for the triplet cohort, and pharmacokinetic outcomes. When several outcomes, cohorts, or dose levels are examined, the chance of at least one apparently noteworthy result can increase. In general, multiplicity procedures can control that risk, but the choice depends on the objectives and planned analyses.

The registry record does not report a multiplicity strategy, a hierarchy among secondary outcomes, or whether any significance levels were adjusted. It also reports no analyses for the outcomes listed in the record. The registry’s sample-size planning assumptions are not a substitute for an analysis plan, and the planning illustration is not evidence of observed power.

Missing-data handling is not reported in the registry record. The record does define a DLT-evaluable population: participants in the safety run-in who received one cycle with at least 80% of the intended dose for each treatment in the combination. It does not report how many participants met this definition, how participants with incomplete dosing were handled in the primary analysis, or whether other populations were used for the secondary endpoints. The analysis populations for those other outcomes are not reported in the registry record.

For time-to-event outcomes, censoring rules and follow-up handling are important; these are not reported in the registry record. The record does not specify how missing assessments, withdrawals, or incomplete outcome measurements were addressed. Without those details and the underlying analysis results, readers cannot assess whether missingness could have influenced estimates or whether the analysis population matched the intended study population.

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: non-cardiac chest pain (100.0%), hyperglycaemia (33.3%), alanine aminotransferase increased (33.3%), pyrexia (16.7%), disease progression (13.6%).

The registry reports serious adverse-event counts by treatment cohort. For the 150 mg/m^2 tusamitamab ravtansine plus pembrolizumab cohort, 9 participants were affected among 23 at risk; for the 170 mg/m^2 doublet cohort, the counts are 0 among 2. The corresponding figures are 1 among 6 and 1 among 1 for the triplet cohorts, and 11 among 22 and 2 among 3 for the quadruplet cohorts, respectively. The record also gives counts for other adverse events and deaths by cohort.

These are descriptive cohort summaries. They should not be interpreted as estimates of comparative treatment effects: allocation was non-randomized, the regimens differ, and some cohorts have few participants at risk. The registry does not report an adjusted safety comparison or uncertainty intervals. The serious-event counts also do not identify which events met the primary DLT definition, which has a specific time window and criteria.

The record lists serious-event terms with percentages, including Non-Cardiac Chest Pain at 100.0%, Hyperglycaemia at 33.3%, and Alanine Aminotransferase Increased at 33.3%. The listed information does not specify enough about the denominators to interpret those percentages as cohort risks or to compare them across treatment groups. Nor do the counts alone establish that an event was caused by treatment. The registry record does not report the attribution or cause for each death in the safety summaries.

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

The registry record establishes the study’s design and endpoint definitions, and it contains cohort-level adverse-event summaries. It does not report a primary endpoint analysis for study-drug-related DLTs. Consequently, the posted record does not establish the observed DLT count or provide an estimate of its precision. The DLT-evaluable population is defined, but the number of participants in that population is not reported in the registry record.

The safety counts can be read as descriptions of events recorded in the listed cohorts. They are not evidence that one regimen is safer than another. Participants were not randomly allocated, and the cohorts differ in dose and in whether platinum-based chemotherapy or pemetrexed was included. Other differences between participants may also affect observed event counts. Without a randomized comparison or an appropriate adjustment strategy, those factors cannot be separated from a treatment effect.

The listed secondary outcomes indicate the topics the study was designed to assess, including antitumor activity and pharmacokinetics. Their presence in the registry does not mean results are available. The registry record does not report analyzed results for these outcomes in the information summarized. It also does not report a reason for the terminated status, so termination should not be interpreted as evidence of a particular efficacy or safety finding.

In general, a Phase 2 study can provide descriptive evidence that helps characterize treatment experience and inform later research. That general role does not make non-randomized cohort differences causal, nor does a planning calculation establish that an effect was observed. Here, the planning assumptions are explicitly described as an illustration, not as the trial’s own design. Any sample-size illustration based on an observed effect should not be read as observed power.

The responsible conclusion is limited: the registry record describes a non-randomized study of several combinations and dose levels, gives some cohort-level safety counts, and defines a first-cycle DLT endpoint. It does not provide the primary endpoint result or enough analysis detail to support a comparative conclusion. No medical decision or treatment recommendation follows from these registry data alone.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of why the design and endpoint must be read together. It explores several treatment combinations and dose levels in a safety run-in and expansion structure. The primary endpoint is a protocol-defined first-cycle DLT count, whereas the posted adverse-event summaries cover broader safety categories. Distinguishing those measures prevents a reader from treating all reported adverse events as if they were primary endpoint events.

The non-randomized parallel design also matters. Cohort-level counts can describe what was recorded in each treatment group, but the groups are not a randomized control comparison. Differences in the combinations, dose levels, and participant composition can complicate interpretation. This is particularly important where some at-risk counts are small: a count can change substantially in meaning depending on its denominator, and a descriptive contrast does not remove confounding.

The registry record also illustrates the difference between a study objective and a reported result. It describes an objective to assess tolerability and determine recommended doses, and an expansion objective to assess antitumor activity. Yet the primary DLT analysis is not reported in the record. Objectives and endpoint definitions establish what was intended to be studied; they do not establish what was observed.

Finally, termination status should be kept separate from statistical interpretation. Without a reported reason, it cannot be treated as evidence about the treatment’s performance or the study’s data. Careful reading therefore requires attention to the outcome definition, evaluable population, cohort design, missing analysis details, and limits on causal comparison. Those principles apply broadly to early-phase studies with multiple non-randomized cohorts.

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 in this registry record. The study is non-randomized and open-label, with parallel cohorts receiving different dose levels and treatment combinations. There is no randomized comparator, so cohort differences cannot be treated as like-for-like treatment effects. The record does not report a statistical adjustment that would address differences between groups.

The primary DLT endpoint is carefully defined, including its first-cycle window and evaluability requirements, but its analysis is not reported. The record does not give the number of participants meeting the DLT-evaluable definition or the number with DLTs. The broader adverse-event counts cannot replace that missing primary result. Some cohort counts are small, limiting the precision and interpretability of descriptive summaries.

The registry lists multiple secondary outcomes across different cohorts, but analysis results and methods are not reported in the information summarized. A multiplicity approach, missing-data strategy, time-to-event censoring rules, and analysis populations beyond the stated DLT-evaluable population are not reported in the registry record. The record also does not report the reason for termination, the trial’s sample-size justification, or a full statistical analysis plan.

Safety data require further caution. Counts and listed serious-event terms do not by themselves establish treatment causality, and the record does not clarify the cause or attribution of each death. The available serious-event term percentages lack sufficient denominator context for comparative interpretation. Together, these limitations mean that the registry information supports descriptive reporting, but not a reliable comparative efficacy or safety conclusion.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Tusamitamab Ravtansine (SAR408701) in Combination With Pembrolizumab and Platinum-based Chemotherapy With or Without Pemetrexed in Patients With NSQ NSCLC — NCT04524689.
• Registry record last updated 2025-10-29; status last verified 2025-09.
• Results first posted 2025-03-17.
• 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

NCT04524689 is listed as an open-label, non-randomized, parallel Phase 2 study of tusamitamab ravtansine with pembrolizumab, with platinum-based chemotherapy and, in some cohorts, pemetrexed. The population is described as CEACAM5-positive advanced or metastatic non-squamous NSCLC. The registry lists enrollment of 57, a start date of 2020-10-26, primary completion on 2024-03-12, and status as terminated.

The primary outcome is the number of participants with study-drug-related DLTs during Cycle 1 Day 1 through Cycle 1 Day 21. The record defines the DLT-evaluable population but does not report a primary endpoint analysis. Cohort-level adverse-event counts are posted, but the non-randomized design does not support direct causal comparisons between regimens. The reason for termination and the analysis details for missing data, multiplicity, and secondary outcomes are not reported in the registry record.