Tusamitamab Ravtansine Monotherapy and in Combination in Patients With CEACAM5-positive Advanced Solid Tumors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tusamitamab Ravtansine Monotherapy and in Combination in Patients With CEACAM5-positive Advanced Solid Tumors (NCT04659603), a non-randomized phase 2 study. The primary outcome is Objective Response Rate (ORR).

2. Clinical Question

3. Trial Design

NCT04659603 was an open-label, non-randomized, parallel-cohort Phase 2 trial. The registry lists an enrollment of 50 and a status of terminated. The recorded study start was 2021-03-29, and the primary completion date was 2024-02-16. The reason for termination is not reported in the registry record.

The study included three cohorts. Cohort A enrolled participants with metastatic breast cancer and used tusamitamab ravtansine monotherapy. Cohort B enrolled participants with metastatic pancreatic adenocarcinoma and also used monotherapy. Cohort C enrolled participants with metastatic pancreatic adenocarcinoma and used tusamitamab ravtansine with gemcitabine. Treatment was open-label, and masking was listed as none.

The stated objectives differed by cohort and part. Cohorts A and B and Cohort C Part 2 were intended to assess antitumor activity. Cohort C Part 1 was intended to confirm the recommended dose of tusamitamab ravtansine with gemcitabine. The record also lists objectives concerning safety and tolerability, other efficacy parameters, immunogenicity, and pharmacokinetics.

The primary endpoint timeframe began on the first study-treatment administration, Day 1, and extended up to the maximum treatment exposure recorded for each cohort: 26.1 weeks for Cohort A, 51.6 weeks for Cohort B and 43.3 weeks for Cohort C. These are cohort-level maximum exposure windows, not evidence that every participant had that duration of observation. Details of randomization, a control regimen, treatment discontinuation criteria and the reason the study ended are not reported in the registry record.

4. Participant Flow and Accounting

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

Before assignment: Total 50 participants enrolled to receive tusamitamab ravtansine as single agent treatment (Cohorts A, B) or in combination with gemcitabine (Cohort C). All participants received the same dose in the respective Cohorts. The study was terminated due to the discontinuation of the overall development program of tusamitamab ravtansine by the Sponsor.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All-treated population included all enrolled participants exposed to the study treatment, regardless of the amount of treatment administered.

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 enrolled participants exposed to the study treatment, regardless of the amount of treatment administered.

6. Primary Endpoint

The registry lists the primary outcome as Objective Response Rate (ORR), measured over From date of first study treatment administration (Day 1) up to maximum exposure of treatment: 26.1 weeks (Cohort A), 51.6 weeks (Cohort B), 43.3 weeks (Cohort C).

The primary endpoint is objective response rate (ORR), defined as the percentage of participants with a confirmed complete response or partial response according to RECIST version 1.1. A complete response requires disappearance of all target and non-target lesions and normalization of tumor marker level; pathological lymph nodes must also have a short axis under 10 mm. A partial response requires at least a 30% decrease in the sum of target-lesion diameters, using the baseline sum as the reference.

ORR is a binary outcome at the participant level: each participant either meets the response definition or does not. Expressing the count meeting the definition as a percentage gives a measure of observed tumor response in the analyzed population. A confirmed response definition is intended to distinguish a response that is verified from one based on a single assessment. The registry record does not report the response counts, percentages or confidence intervals, nor does it detail the confirmation schedule.

As a general statistical point, ORR describes response frequency over the defined assessment period; it does not by itself describe how long responses last or how long participants remain free of progression. The listed secondary outcomes include duration of response and progression-free survival, but their results are not reported in the registry record. The endpoint timeframe is specified as exposure-based, and the precise assessment schedule is not reported.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary response measure reported as a percentage of participants. At the participant level, the endpoint classification depends on whether a confirmed complete or partial response is documented under RECIST version 1.1. The percentage summarizes the proportion meeting either response category within the relevant analysis population. The registry record does not report the cohort-level response numerators or denominators, so no ORR can be described from the available record.

The endpoint is accompanied in its definition by a 95% confidence interval. In general, a confidence interval communicates the uncertainty around an estimate: it describes a range of values that are reasonably compatible with the observed data and the method used. It is not a guarantee that the true value lies within the interval, and it is not a range of individual participant outcomes. No ORR interval is reported, and the method used to calculate one is not reported in the registry record.

RECIST version 1.1 supplies the response criteria. A complete response is defined in the record through disappearance of target and non-target lesions, normalization of tumor marker level, and a specified requirement for pathological lymph nodes. A partial response is defined by a decrease in the summed diameters of target lesions relative to baseline. These definitions turn imaging and related clinical assessments into response categories. The record does not describe the assessment schedule, independent review, adjudication procedures or the confirmation rules beyond stating that responses were confirmed.

As a general matter, response percentages depend on the population used as the denominator and on how participants without a usable assessment are classified. The record defines the all-treated population as enrolled participants who were exposed to treatment, regardless of the amount administered. It does not state whether the ORR denominator was exactly that population, whether participants with missing assessments were counted as nonresponders, or whether another missing-data approach was applied. Those details are important for reproducing or interpreting a percentage.

The trial’s allocation was non-randomized, and masking was listed as none. In a randomized comparison, assignment by chance is intended to balance measured and unmeasured characteristics on average between groups. Without randomization, cohort differences may reflect differences in participant characteristics, disease setting, treatment selection or other factors, as well as any treatment-related effect. Open-label treatment also means the study was not masked; the record does not describe additional safeguards for outcome assessment. The cohorts should therefore be read as separate study groups rather than as randomized treatment and control arms.

The listed efficacy measures beyond ORR include progression-free survival, disease control rate and duration of response. In general, time-to-event outcomes account for the time until an event or censoring, and their interpretation depends on follow-up and censoring rules. Those analysis details and results are not reported in the registry record. The record also lists safety, laboratory, immunogenicity and pharmacokinetic measures, but it does not report their analysis methods in enough detail to reconstruct the analyses.

The record lists 12 outcome measures and zero analyses. No statistical analysis plan, model specification, covariate adjustment, interim analysis procedure or multiplicity strategy is reported in the registry record. These omissions mean that it is not possible to determine from the record how the stated endpoints were analyzed, how repeated testing across endpoints or cohorts was addressed, or whether the analyses followed a prespecified plan.

As a general concept, over-dispersion matters when observed variability exceeds what a statistical model assumes; a model that ignores it may understate uncertainty. The registry record does not identify a model for the primary binary endpoint, report evidence of over-dispersion, or indicate that such a model was used. It would therefore be inappropriate to infer a particular analysis technique. Likewise, no recalculated estimates or sample-size planning outputs are reported in the registry record. Any planning illustration based on an observed effect would not represent observed power; no observed effect is reported for this endpoint.

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("Cohort A: mBC", "Cohort B: mPAC", "Cohort C: mPAC")
n      <- c(6, 28, 16)
events <- c(0, 1, 5)
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  Cohort A: mBC  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Cohort B: mPAC  28  1  3.6  0.1  18.3  0.6  17.7  0.8  17.8
ARM  Cohort C: mPAC  16  5  31.2  11.0  58.7  14.2  55.6  14.2  56.0

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple objectives and outcomes, including ORR, progression-free survival, disease control rate, duration of response, safety outcomes, immunogenicity and pharmacokinetic measures. It does not report how statistical testing across these outcomes, cohorts or study parts was controlled. No hierarchy of confirmatory and exploratory hypotheses, adjustment procedure or prespecified multiplicity strategy is reported in the registry record. Without such information, a reader cannot determine whether any reported significance tests—if available elsewhere—would account for the number of questions examined. No analyses are reported in this record.

Missing-data handling is also not reported in the registry record. In particular, the record does not state how participants without an adequate tumor assessment, participants who discontinued treatment, or participants with incomplete follow-up were handled for ORR. It gives no imputation method, censoring rules for time-to-event outcomes, or sensitivity analyses for missing outcomes. This matters because the analyzed denominator and assumptions about unobserved outcomes can affect estimates.

The record defines the all-treated population as enrolled participants exposed to study treatment, regardless of the amount of treatment received. It does not explain whether this definition applied to each endpoint analysis or whether other analysis populations were used. Analysis-population details, the prespecified statistical plan and any deviations from it are not reported in the registry record. As a result, the available information does not establish how the planned endpoint analyses were implemented.

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: disease progression (21.4%), pneumonia aspiration (6.2%), urinary tract infection (6.2%), urosepsis (6.2%), ischaemic cerebral infarction (6.2%).

The registry record reports serious adverse event counts by cohort: Cohort A had 6 participants at risk and 0 affected; Cohort B had 28 at risk and 16 affected; Cohort C had 16 at risk and 9 affected. For other adverse events, the corresponding figures were 6 at risk and 6 affected in Cohort A, 28 and 23 in Cohort B, and 16 and 14 in Cohort C. These are participant counts, not measures of event severity beyond the distinction between serious and other adverse events.

The record also lists deaths by cohort: 4 in Cohort A, 21 in Cohort B and 9 in Cohort C. A death count during a trial does not establish that treatment caused the death. The registry information does not provide a causal assessment for each death in these figures. The listed serious-event terms include Disease Progression at 21.4%, and Pneumonia Aspiration, Urinary Tract Infection, Urosepsis, Ischaemic Cerebral Infarction and Gastric Dilatation at 6.2% each. The record does not provide the event-level context needed to interpret those percentages as treatment-caused risks.

As a general point, adverse-event counts are descriptive and depend on the population observed, exposure and follow-up. Cohort sizes and clinical settings differ, and the trial was not randomized; raw counts should not be treated as direct comparative evidence. Further details about timing, duration and attribution are not reported in the registry record.

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

The record describes a Phase 2 investigation of tusamitamab ravtansine in two disease settings, with monotherapy cohorts and a combination cohort. Its stated primary endpoint, ORR, is a measure of confirmed tumor response under RECIST version 1.1. The listed objectives also include safety and other efficacy measures. However, the registry record does not report primary response estimates, response counts or confidence intervals. It therefore does not establish the observed ORR or how precisely it was estimated.

The study’s non-randomized, open-label design also limits the interpretation of comparisons between cohorts. Cohort A concerns metastatic breast cancer, whereas Cohorts B and C concern metastatic pancreatic adenocarcinoma; Cohorts B and C also differ in whether gemcitabine was included. These groups cannot serve as equivalent randomized controls for one another. Differences in their outcomes, if reported, could not be attributed to treatment alone without accounting for the design and possible differences between participants.

The safety counts describe events recorded in the cohorts, but they do not establish causation. Deaths recorded during the study should not automatically be interpreted as treatment-related deaths. The registry record does not include enough event-level information to establish that relationship or support a causal comparison between cohorts.

More broadly, a response percentage would not alone show how long a response lasted, whether progression was delayed, or how participants experienced treatment. The listed secondary measures address some of these separate questions, but their results are not reported in the registry record. The status of terminated is recorded, but the reason for termination is not reported; that status alone does not explain why the study ended or what consequences it had for the analyses.

The defensible conclusion is limited to the design, intended endpoints and reported safety counts. The registry record does not provide enough efficacy results or analysis detail to characterize treatment effects, precision, or comparative benefit. It also does not support medical advice or an individual treatment decision.

14. Why This Trial Matters Statistically

This trial raises several statistical questions that are important in early-phase clinical research. It studies separate cohorts across metastatic breast cancer and metastatic pancreatic adenocarcinoma, with both monotherapy and combination treatment. Keeping those settings distinct is essential: a cohort’s outcome is meaningful in relation to its own population and treatment context, not as an automatically comparable control for another cohort.

ORR is a clear, clinically interpretable binary endpoint when response criteria and the analysis population are defined. Its interpretation nevertheless depends on confirmed assessments, the denominator, follow-up and handling of missing evaluations. The record specifies RECIST version 1.1 and the all-treated population definition, but leaves important analysis details unreported. This illustrates why an endpoint definition and an endpoint result are different kinds of information.

The registry also makes clear why precision matters. A response estimate without its uncertainty can give an incomplete picture, particularly in cohort-based research. The primary endpoint confidence interval is specified as a measure, but no interval or response estimate is reported. Readers therefore cannot assess the precision of the intended efficacy assessment from the registry record.

Finally, the study’s non-randomized, open-label design requires restraint in interpreting differences across cohorts. The recorded safety counts can describe what was observed, but they are not proof of causation and do not provide a randomized comparison. The trial’s statistical importance lies in the questions its design makes necessary: how outcomes were defined, which participants were analyzed, how uncertainty was quantified, and what conclusions the design can support. Several of those analysis details and the primary results are not reported in the registry record.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record does not report the primary ORR estimates, response counts or confidence intervals. It also does not report cohort-specific denominators for that endpoint, the statistical method used to calculate the interval, or the handling of participants without evaluable assessments. These gaps prevent assessment of both the observed response frequency and its precision.

The design is non-randomized and open-label. There is no randomized control group, and the cohorts differ by disease setting and treatment. Consequently, comparisons across cohorts are not like-for-like estimates of treatment effects. The record does not report whether outcome assessments were independently reviewed or otherwise protected from knowledge of treatment assignment.

The record lists a maximum exposure timeframe for each cohort, but does not provide individual follow-up duration or the assessment schedule. The maximum exposure window should not be mistaken for common follow-up among participants. Missing-data methods, censoring rules, analysis populations for outcomes and sensitivity analyses are not reported in the registry record.

Multiplicity handling, the statistical analysis plan, interim analyses and the reason for termination are also not reported. The record lists safety counts and several serious adverse-event terms, but does not provide sufficient event-level detail to establish treatment attribution. These limitations apply to interpretation of the registry record; they do not establish that unreported procedures or analyses were absent from other study documents.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Tusamitamab Ravtansine Monotherapy and in Combination in Patients With CEACAM5-positive Advanced Solid Tumors — NCT04659603.
• Registry record last updated 2025-10-01; status last verified 2025-09.
• Results first posted 2025-03-25.
• 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

NCT04659603 was an open-label, non-randomized Phase 2 trial of tusamitamab ravtansine alone or with gemcitabine in CEACAM5-positive advanced solid tumors. It enrolled 50 participants and is listed as terminated. The study start date was 2021-03-29, and the primary completion date was 2024-02-16.

Cohort A involved metastatic breast cancer treated with monotherapy; Cohort B involved metastatic pancreatic adenocarcinoma treated with monotherapy; and Cohort C involved metastatic pancreatic adenocarcinoma treated with tusamitamab ravtansine plus gemcitabine. ORR was the listed primary endpoint, defined by confirmed complete or partial response under RECIST version 1.1. The registry record does not report ORR results, confidence intervals, the reason for termination, or key statistical analysis details.