This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Safety and Efficacy of Pembrolizumab (MK-3475) Plus Binimetinib Alone or Pembrolizumab Plus Chemotherapy With or Without Binimetinib in Metastatic Colorectal Cancer (mCRC) Participants (MK-3475-651/KEYNOTE-651) (NCT03374254), a non-randomized phase 1 study. The primary outcome is Percentage of Participants Who Experienced Dose-Limiting Toxicity (DLT).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03374254 |
| Phase | Phase 1 |
| Condition | Metastatic Colorectal Cancer |
| Allocation | Non-Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 116 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Merck Sharp & Dohme LLC |
| Results posted | Yes |
2. Clinical Question
Population
Participants with metastatic colorectal cancer enrolled in a multi-cohort study. The primary-outcome population was participants who received treatment and were DLT evaluable in Part 1.
Intervention
The study assessed pembrolizumab with binimetinib, chemotherapy, or chemotherapy plus binimetinib. Chemotherapy regimens included mFOLFOX7 and FOLFIRI; some cohorts also explored different binimetinib doses.
Comparator
There was no randomized control group. The cohorts received different combinations or doses, but the registry record does not establish them as like-for-like comparative arms.
Primary question
What percentage of participants experienced a dose-limiting toxicity during the specified early treatment period? The stated purpose also included assessing safety and tolerability and establishing a preliminary recommended Phase 2 dose.
The study was designed as an early-phase, multi-cohort evaluation rather than a randomized test of one treatment against another. Its primary outcome focuses on whether prespecified toxicities occurred during an early observation window, not on a direct estimate of comparative treatment benefit. The registry also lists objective response rate by RECIST 1.1 as a secondary outcome, but its numerical results are not reported in the registry record.
The cohort structure matters when interpreting the clinical question. Pembrolizumab was paired with binimetinib alone in Cohort A, with mFOLFOX7 in Cohort B, with mFOLFOX7 and binimetinib in Cohort C, with FOLFIRI in Cohort D, and with FOLFIRI and binimetinib in Cohort E. These are distinct regimens and dose-level groupings, not randomized comparisons. The central question is therefore about early safety and dose selection across the combinations, rather than whether one listed regimen outperformed another.
3. Trial Design
The registry identifies NCT03374254 as a Phase 1 study and describes it as a Phase 1b multi-cohort study. It is marked completed, with an enrollment of 116. The recorded start date is 2018-02-16 and the primary completion date is 2021-09-08. The record describes a parallel model, nonrandomized allocation, and no masking.
Five cohorts evaluated pembrolizumab with different treatment partners. Cohort A paired pembrolizumab with binimetinib, including 30 mg and 45 mg dose groups. Cohort B paired pembrolizumab with mFOLFOX7. Cohort C combined pembrolizumab, mFOLFOX7, and binimetinib. Cohort D paired pembrolizumab with FOLFIRI. Cohort E combined pembrolizumab, FOLFIRI, and binimetinib, with 30 mg and 45 mg groups described in the arm listings. Some cohorts also have Part 1 and Part 2 listings.
The stated purpose was to assess safety and tolerability and establish a preliminary recommended Phase 2 dose for the combinations. That purpose fits an early-phase study: the primary outcome is an early toxicity measure, and the cohort structure explores combinations and dose levels. The registry record does not report a randomized allocation process or a common control regimen. Consequently, differences among cohort results, if reported elsewhere, would not automatically support a causal comparison between treatments.
Descriptions of treatment include pembrolizumab with mFOLFOX7 or FOLFIRI, with or without binimetinib. The listed chemotherapy schedules and drug doses vary by regimen, and the record describes treatment continuing until disease progression or discontinuation in several arm descriptions. Further operational details, including a complete account of dose escalation or cohort assignment, 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: A total of 161 participants were screened, 116 participants were allocated, and 114 received at least 1 dose of study treatment across 14 study sites in 2 countries. This study included 2 parts: Part 1 was a dose finding phase and Part 2 was a dose confirmation phase to further examine safety and explore efficacy. Cohorts A, C, and E did not proceed to Part 2.
Only the first 6 of 9 groups are shown.
Overall Study
| Stage | Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | Cohort B Part 1: Pembrolizumab + mFOLFOX7 | Cohort B Part 2: Pembrolizumab + mFOLFOX7 | Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | Cohort D Part 1: Pembrolizumab + FOLFIRI |
|---|---|---|---|---|---|---|
| STARTED | 6 | 16 | 15 | 16 | 11 | 16 |
| Treated | 6 | 14 | 15 | 16 | 11 | 16 |
| COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
| NOT COMPLETED | 6 | 16 | 15 | 16 | 11 | 16 |
Reasons for not completing this period:
| Reason | Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | Cohort B Part 1: Pembrolizumab + mFOLFOX7 | Cohort B Part 2: Pembrolizumab + mFOLFOX7 | Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | Cohort D Part 1: Pembrolizumab + FOLFIRI |
|---|---|---|---|---|---|---|
| Death | 5 | 16 | 13 | 11 | 9 | 12 |
| Lost to Follow-up | 0 | 0 | 0 | 0 | 0 | 0 |
| Sponsor Decision | 1 | 0 | 2 | 3 | 2 | 3 |
| Withdrawal by Subject | 0 | 0 | 0 | 2 | 0 | 1 |
Analysis population for the primary outcome, as defined in the registry: Population based on all participants who received treatment and that were DLT Evaluable in Part 1 of study. These populations included all allocated participants in Part 1 who received at least 1 dose of study intervention according to the study intervention they received.
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.
Only the first 6 of 10 groups are shown.
| Characteristic | Unit | Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | Cohort B Part 1: Pembrolizumab + mFOLFOX7 | Cohort B Part 2: Pembrolizumab + mFOLFOX7 | Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | Cohort D Part 1: Pembrolizumab + FOLFIRI |
|---|---|---|---|---|---|---|---|
| Participants analysed | 6 | 16 | 15 | 16 | 11 | 16 | |
| Age, Continuous | Years | 55.0 (SD 9.3) | 59.4 (SD 11.3) | 55.5 (SD 11.6) | 50.3 (SD 12.0) | 54.8 (SD 14.1) | 56.3 (SD 11.7) |
| Sex: Female, Male: Female | Participants | 1 | 11 | 7 | 3 | 3 | 11 |
| Sex: Female, Male: Male | Participants | 5 | 5 | 8 | 13 | 8 | 5 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 0 | 1 | 1 | 1 | 2 | 3 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 6 | 15 | 13 | 15 | 9 | 13 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 1 | 0 | 0 | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 | 0 | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Asian | Participants | 1 | 1 | 3 | 0 | 0 | 2 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 | 0 | 0 | 0 | 0 | 0 |
| Race (NIH/OMB): Black or African American | Participants | 1 | 0 | 0 | 0 | 2 | 0 |
| Race (NIH/OMB): White | Participants | 4 | 15 | 11 | 15 | 9 | 14 |
| Race (NIH/OMB): More than one race | Participants | 0 | 0 | 1 | 0 | 0 | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 | 0 | 0 | 1 | 0 | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Percentage of Participants Who Experienced Dose-Limiting Toxicity (DLT), measured over Up to approximately first 28 days of treatment.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | 80% Confidence Interval |
| Unit | Percetange of Participants |
| Endpoint classification | not reduced to a single comparison |
The primary outcome is the percentage of participants who experienced dose-limiting toxicity, or DLT. A DLT is a prespecified toxicity judged possibly, probably, or definitely related to study treatment during the initial treatment period. The registry lists different observation windows: the first 21 days for Cohort A and the first 28 days for Cohorts B, C, D, and E. The outcome record describes an 80% confidence interval.
The listed DLT criteria cover several kinds of harm, including severe hematologic or nonhematologic toxicity, febrile neutropenia, treatment delays or discontinuation due to treatment-related toxicity, missing more than 25% of binimetinib doses because of drug-related adverse events, and specified cardiac, vascular, and eye disorders. The criteria also include Grade 5 toxicity. This breadth makes the endpoint a structured early safety measure rather than a count of one specific symptom or laboratory abnormality.
In early-phase combination studies, DLT is used to help characterize tolerability during a defined initial period and inform dose selection. It does not capture every safety event over the full course of treatment, nor does it by itself measure antitumor benefit. The registry describes the primary population as participants who received at least one dose in Part 1 and were DLT evaluable, according to the intervention they received. Numerical DLT estimates and their confidence intervals are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The available registry information defines the study question and lists safety-event counts, but it does not include a numerical result for the primary DLT outcome. The reported endpoint is a percentage with an 80% confidence interval, yet neither the percentage nor its interval is reported in the registry record. As a result, the primary outcome cannot be interpreted quantitatively from this record alone. The stated study purpose includes selecting a preliminary recommended Phase 2 dose, but the selected dose and the evidence supporting that selection are not reported in the registry record.
The design also limits what can be concluded from comparisons across cohorts. Allocation was nonrandomized, masking was absent, and the cohorts received different drug combinations or dose levels. Cohorts therefore do not function as interchangeable control and treatment groups. Any differences in outcomes could reflect regimen differences, participant differences, or other factors; the registry record does not report an analysis that separates these possibilities. No comparative efficacy conclusion follows simply from the list of arms.
The registry lists objective response rate by RECIST 1.1 as a secondary outcome, but does not report its numerical findings. Thus, the available information does not establish the observed response rate or its precision. Likewise, the trial’s purpose and treatment descriptions should not be mistaken for evidence that any combination was effective or ineffective.
Safety listings give group-level counts for serious events, other events, and deaths. The counts vary across the listed groups, but the registry information does not provide a causal comparison between groups. The death counts should not be interpreted as treatment-related deaths: attribution, follow-up context, and a group-level causal analysis are not reported in the registry record. Serious adverse-event terms are also listed with percentages, but the denominator and calculation details for those percentages are not reported in the registry record.
The enrollment figure describes the trial overall, not the size of each cohort. Cohort-specific counts appear in the adverse-event listings, but those counts should not be treated as the primary endpoint population without confirmation that the definitions and observation periods match. The primary analysis population is specifically described as Part 1 participants who received treatment and were DLT evaluable. Numerical DLT results for that population are not reported in the registry record.
Overall, the record supports a description of an early-phase, nonrandomized investigation of several pembrolizumab combinations and their safety assessment. It does not provide enough primary-outcome or secondary-outcome results to quantify the observed toxicity rate, response rate, or comparative effects. Statistical interpretation should remain limited to the design, definitions, and safety counts that the registry actually reports.
9. Statistical Methods Explained
The trial’s primary outcome is a binary event for each participant: whether a qualifying DLT occurred during the defined observation period. The outcome is then expressed as a percentage within the stated analysis population. In general, a percentage is a sample estimate of the frequency of an event in a population; it is not a guarantee that the same frequency would be observed in another group or setting. The registry record does not report the observed DLT percentage.
What does the confidence interval add?
A confidence interval communicates statistical uncertainty around an estimate. In general, a narrower interval indicates greater precision, while a wider interval indicates less precision. The registry specifies an 80% confidence interval for the primary outcome, but does not report its endpoints or the method used to calculate it. The interval should not be interpreted as a range containing a fixed share of individual outcomes. It concerns uncertainty in the estimated group percentage under the assumptions of the interval procedure.
The confidence level is part of the statistical method, not a measure of how likely the treatment is to work or how safe it is. A confidence interval also does not address sources of uncertainty outside its calculation, such as differences among nonrandomized cohorts, incomplete follow-up, or limits of the endpoint definition. The registry record does not state whether the primary interval was calculated using a particular exact, score-based, or approximate method.
Why define a DLT window?
For an early safety endpoint, a defined observation window makes clear which events count toward the dose-limiting assessment. Here, the registry specifies the first 21 days for Cohort A and the first 28 days for Cohorts B, C, D, and E. The outcome is therefore tied to early treatment and to cohort-specific timing. It should not be read as a complete count of every toxicity that may occur during all subsequent treatment.
The criteria include events of different types and levels of severity, and some require a relationship to study treatment. In general, an endpoint that combines several event types can serve a dose-finding purpose, but its meaning depends on the exact criteria, attribution process, and evaluability rules. Those criteria are partly described in the record; further details of adjudication and the operational definition of DLT are not reported in the registry record.
How should the cohort structure affect comparisons?
Randomization is a method for reducing systematic differences between groups at the point of assignment. This trial is listed as nonrandomized, with no masking. Its cohorts differ in treatment combinations, dose levels, or study part. In general, such groups may differ for reasons beyond the treatment being received, so a numerical contrast between them would not have the same interpretation as a randomized comparison. The registry explicitly flags that the arms are not a like-for-like control comparison.
The parallel model indicates that the record organizes participants into treatment groups rather than describing a crossover design. It does not make those groups statistically comparable by itself. The record does not report an adjusted comparative model or a prespecified statistical plan that would support a causal interpretation of differences across the cohorts.
What do the planning assumptions mean?
The record describes planning illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It states that the effect to detect is a specified fraction of the observed effect. These are planning assumptions, not observed trial outcomes. Sample-size illustrations are not observed power, and they do not show that the completed study had a particular probability of detecting an effect. The numerical planning estimates and the observed effect used for those illustrations are not reported in the registry record.
In general, a sample-size calculation asks how many participants would be needed under assumptions about the effect, variability, allocation, and statistical method. Changing those assumptions changes the result. The two-arm equal-allocation setup is a simplified planning framework and should not be confused with this study’s multi-cohort, nonrandomized structure. It is not evidence that the actual cohort assignments followed equal allocation, or that the registry used the calculation to design the trial.
The record reports zero analyses, and does not provide detailed inferential methods for the primary or secondary outcomes. The statistical test, model, handling of missing observations, and approach to multiple outcomes are not reported in the registry record. Without those details and the numerical outcome results, there is no basis to reconstruct an adjusted treatment effect or assess its precision.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists a primary DLT outcome and objective response rate by RECIST 1.1 as a secondary outcome, and records two outcome measures. It does not report a multiplicity strategy. In general, when several outcomes or comparisons are examined, the chance of at least one apparently notable result can increase; methods such as prespecifying a primary comparison or adjusting statistical thresholds may address that issue. Whether any such method was used is not reported in the registry record.
Missing-data methods are also not reported in the registry record. The record describes the primary population as participants who received treatment and were DLT evaluable in Part 1, including those who received at least one dose according to the intervention received. It does not explain how participants who were not evaluable, had incomplete observation, or discontinued early were handled in the reported endpoint calculation. The evaluability criteria beyond the described population statement are not reported in the registry record.
Analysis populations for the secondary outcome and adverse-event listings are not fully detailed in the information available. In general, the choice of population can change an estimate: an analysis of everyone assigned to a group is not necessarily the same as one restricted to participants who started treatment or met evaluability criteria. The record’s explicit DLT population should therefore not be assumed to apply automatically to every other outcome.
The registry does not report a complete statistical analysis plan, interim-analysis procedures, or methods for handling censoring or follow-up time. These details are not reported in the registry record. Their absence prevents assessment of whether the published outcome definitions and analyses followed a prespecified plan.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | 1 | 6 | 16.7% |
| Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | 7 | 14 | 50.0% |
| Cohort B Part 1: Pembrolizumab + mFOLFOX7 | 5 | 15 | 33.3% |
| Cohort B Part 2: Pembrolizumab + mFOLFOX | 8 | 16 | 50.0% |
| Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | 7 | 11 | 63.6% |
| Cohort D Part 1: Pembrolizumab + FOLFIRI | 6 | 16 | 37.5% |
| Cohort D Part 2: Pembrolizumab + FOLFIRI | 2 | 16 | 12.5% |
| Cohort E Part 1: Pembrolizumab + FOLFIRI + Binimetinib 30 mg | 5 | 9 | 55.6% |
| Cohort E Part 1: Pembrolizumab + FOLFIRI + Binimetinib 45 mg | 7 | 11 | 63.6% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | 5 of 6 |
| Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | 16 of 16 |
| Cohort B Part 1: Pembrolizumab + mFOLFOX7 | 13 of 15 |
| Cohort B Part 2: Pembrolizumab + mFOLFOX | 12 of 16 |
| Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | 9 of 11 |
| Cohort D Part 1: Pembrolizumab + FOLFIRI | 13 of 16 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | 6 of 6 |
| Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | 14 of 14 |
| Cohort B Part 1: Pembrolizumab + mFOLFOX7 | 15 of 15 |
| Cohort B Part 2: Pembrolizumab + mFOLFOX | 16 of 16 |
| Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | 11 of 11 |
| Cohort D Part 1: Pembrolizumab + FOLFIRI | 16 of 16 |
Most frequent serious adverse event terms
| Event term | Organ system | Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | Cohort B Part 1: Pembrolizumab + mFOLFOX7 | Cohort B Part 2: Pembrolizumab + mFOLFOX | Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | Cohort D Part 1: Pembrolizumab + FOLFIRI |
|---|---|---|---|---|---|---|---|
| Small intestinal obstruction | Gastrointestinal disorders | 0 of 6 | 1 of 14 | 0 of 15 | 1 of 16 | 2 of 11 | 1 of 16 |
| Pyrexia | General disorders | 0 of 6 | 0 of 14 | 0 of 15 | 0 of 16 | 2 of 11 | 1 of 16 |
| Infusion related reaction | Injury, poisoning and procedural complications | 0 of 6 | 0 of 14 | 1 of 15 | 2 of 16 | 0 of 11 | 0 of 16 |
| Abdominal pain | Gastrointestinal disorders | 0 of 6 | 0 of 14 | 0 of 15 | 1 of 16 | 1 of 11 | 0 of 16 |
| Ascites | Gastrointestinal disorders | 0 of 6 | 1 of 14 | 1 of 15 | 0 of 16 | 0 of 11 | 0 of 16 |
| Colitis | Gastrointestinal disorders | 0 of 6 | 0 of 14 | 1 of 15 | 0 of 16 | 1 of 11 | 0 of 16 |
| Diarrhoea | Gastrointestinal disorders | 0 of 6 | 0 of 14 | 0 of 15 | 1 of 16 | 0 of 11 | 1 of 16 |
| Vomiting | Gastrointestinal disorders | 0 of 6 | 1 of 14 | 0 of 15 | 0 of 16 | 1 of 11 | 0 of 16 |
| Pneumonia | Infections and infestations | 0 of 6 | 1 of 14 | 0 of 15 | 0 of 16 | 0 of 11 | 1 of 16 |
| Pleural effusion | Respiratory, thoracic and mediastinal disorders | 0 of 6 | 1 of 14 | 0 of 15 | 0 of 16 | 1 of 11 | 0 of 16 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Cohort A Part 1: Pembrolizumab +Binimetinib 30 mg | Cohort A Part 1: Pembrolizumab +Binimetinib 45 mg | Cohort B Part 1: Pembrolizumab + mFOLFOX7 | Cohort B Part 2: Pembrolizumab + mFOLFOX | Cohort C Part 1: Pembrolizumab + mFOLFOX7 + Binimetinib 30 mg | Cohort D Part 1: Pembrolizumab + FOLFIRI |
|---|---|---|---|---|---|---|---|
| Diarrhoea | Gastrointestinal disorders | 5 of 6 | 8 of 14 | 11 of 15 | 8 of 16 | 10 of 11 | 13 of 16 |
| Fatigue | General disorders | 4 of 6 | 10 of 14 | 11 of 15 | 11 of 16 | 7 of 11 | 9 of 16 |
| Nausea | Gastrointestinal disorders | 1 of 6 | 7 of 14 | 11 of 15 | 12 of 16 | 9 of 11 | 9 of 16 |
| Vomiting | Gastrointestinal disorders | 2 of 6 | 7 of 14 | 4 of 15 | 10 of 16 | 7 of 11 | 7 of 16 |
| Constipation | Gastrointestinal disorders | 1 of 6 | 4 of 14 | 7 of 15 | 10 of 16 | 7 of 11 | 6 of 16 |
| Neuropathy peripheral | Nervous system disorders | 0 of 6 | 1 of 14 | 7 of 15 | 12 of 16 | 8 of 11 | 2 of 16 |
| Stomatitis | Gastrointestinal disorders | 0 of 6 | 2 of 14 | 7 of 15 | 7 of 16 | 4 of 11 | 7 of 16 |
| Dermatitis acneiform | Skin and subcutaneous tissue disorders | 4 of 6 | 5 of 14 | 3 of 15 | 3 of 16 | 8 of 11 | 2 of 16 |
| Anaemia | Blood and lymphatic system disorders | 1 of 6 | 2 of 14 | 4 of 15 | 5 of 16 | 3 of 11 | 9 of 16 |
| Decreased appetite | Metabolism and nutrition disorders | 1 of 6 | 4 of 14 | 2 of 15 | 6 of 16 | 4 of 11 | 7 of 16 |
Serious events with the highest arm-level rate: small intestinal obstruction (18.2%), pyrexia (18.2%), intestinal obstruction (16.7%), infusion related reaction (12.5%), large intestine perforation (11.1%).
The adverse-event listings give counts at risk and counts affected for serious events and other events, plus death counts, across the listed treatment groups. For example, the Cohort A Part 1 pembrolizumab plus binimetinib 30 mg listing records 1 serious-event affected participant among 6 at risk, 6 affected by other events among 6 at risk, and 5 deaths. These are descriptive group counts, not evidence that the treatment caused the events or deaths.
The serious-event terms listed with percentages include Small intestinal obstruction and Pyrexia, each at 18.2%; Intestinal obstruction at 16.7%; Infusion related reaction at 12.5%; and Large intestine perforation and Vomiting, each at 11.1%. The denominator and calculation details for these percentages are not reported in the registry record, so they should not be treated as rates calculated from the group at-risk counts without further information.
Counts across groups should not be compared as if the groups were randomized or matched. The cohorts differ in combinations, doses, or study parts, and the registry does not provide a causal analysis. Death counts do not by themselves identify cause, timing, or treatment relationship; those details are not reported in the registry record. Serious adverse events also differ from the primary DLT endpoint, which has its own criteria, treatment-attribution requirement, and early observation window.
12. What the Result Does, and Does Not, Mean
The record establishes that this was a completed early-phase study of pembrolizumab combinations in metastatic colorectal cancer, with safety and tolerability and preliminary dose selection as stated aims. It defines DLT as the primary outcome and specifies the assessment windows and a broad set of qualifying toxicities. These details clarify what the study was designed to assess.
However, the primary outcome’s numerical percentage and 80% confidence interval are not reported in the registry record. The selected preliminary recommended Phase 2 dose is also not reported in the registry record. The secondary objective response rate is listed, but its result is not reported in the registry record. Therefore, the record does not establish the observed DLT frequency, the precision of that estimate, the response rate, or the basis for any dose-selection decision.
The safety listings provide descriptive counts and event terms, but they do not establish that a treatment caused a particular event or death. Because treatment allocation was nonrandomized and the regimens differ across cohorts, comparisons among groups should not be interpreted as randomized treatment effects. The record does not support a conclusion that one combination is safer or more effective than another.
More generally, a study’s purpose is not itself a result. An early-phase investigation can be designed to characterize tolerability and guide later testing, but the design description alone cannot show whether a regimen met those aims. Likewise, a listed secondary endpoint does not demonstrate a favorable or unfavorable outcome when its numerical result is absent.
The findings that can be described from this record are therefore bounded: it documents the study structure, outcome definitions, and adverse-event listings. It does not supply the numerical primary or secondary efficacy results needed for a fuller statistical assessment. No medical conclusion about the combinations follows from the information.
13. Why This Trial Matters Statistically
This study is statistically informative as an example of how early-phase combination research differs from a conventional randomized comparison. Its multi-cohort structure evaluates several combinations and dose groupings, while the primary endpoint focuses on early treatment-related toxicity. Those design choices align with the stated aims of assessing safety and tolerability and establishing a preliminary recommended Phase 2 dose.
The endpoint definition highlights why timing and population rules matter. DLT is assessed within a specified early window, and the registry gives different windows for Cohort A and the other cohorts. The outcome also depends on treatment attribution and DLT evaluability. A percentage calculated from this population would answer a targeted early safety question, not summarize every adverse event during the full course of treatment.
The nonrandomized design is equally important for interpretation. Multiple treatment groups may offer descriptive information about distinct combinations, but their differences are not automatically treatment effects. Without random assignment, a common comparator, and reported analytical methods, a contrast among groups cannot isolate the effect of adding or changing a drug.
Finally, this record illustrates the difference between a registered question and a reported result. It specifies the primary endpoint and lists adverse-event counts, but the primary DLT estimate, its confidence interval, and the secondary response result are not reported in the registry record. Statistical transparency depends not only on naming outcomes, but also on reporting their estimates, uncertainty, analysis populations, and methods. Those missing results limit what can be concluded from the record.
14. 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 principal limitation for statistical interpretation is the absence of numerical results for the primary DLT outcome. Although the record specifies an 80% confidence interval, neither its estimate nor its bounds are reported in the registry record. The secondary objective response rate is listed without numerical findings, and the preliminary recommended Phase 2 dose is not reported in the registry record.
The nonrandomized design and different treatment regimens prevent a like-for-like comparison among cohorts. The record does not report a randomized control group, masking, or an adjusted comparative analysis. Cohort differences should therefore be treated descriptively rather than as evidence of relative treatment effects.
Several details needed to assess the analyses are not reported in the registry record, including a complete statistical analysis plan, missing-data procedures, multiplicity handling, and details of DLT evaluability beyond the stated population description. The statistical methods used for the confidence interval and secondary outcome are not reported in the registry record.
The adverse-event listings also have limits. The record gives group counts and selected serious-event terms with percentages, but does not report the denominator or calculation details for those percentages. It does not provide the context needed to interpret death counts as treatment-related outcomes. The safety listings should not be equated with the primary DLT endpoint, which has a distinct definition and observation period.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Safety and Efficacy of Pembrolizumab (MK-3475) Plus Binimetinib Alone or Pembrolizumab Plus Chemotherapy With or Without Binimetinib in Metastatic Colorectal Cancer (mCRC) Participants (MK-3475-651/KEYNOTE-651) — NCT03374254.
- Registry record last updated 2024-11-21; status last verified 2024-11.
- Results first posted 2024-10-15.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT03374254 is a completed, nonrandomized Phase 1 study described as a Phase 1b multi-cohort investigation in metastatic colorectal cancer. The record lists enrollment of 116 and evaluates pembrolizumab with binimetinib, mFOLFOX7, FOLFIRI, or combinations of chemotherapy and binimetinib. It records no masking and a parallel model.
The primary outcome is the percentage of participants experiencing DLT during the specified early treatment period, with an 80% confidence interval. The registry also lists objective response rate by RECIST 1.1 as a secondary outcome. Numerical primary and secondary results are not reported in the registry record. Adverse-event counts and selected serious-event terms are listed, but they do not support causal comparisons between the nonrandomized cohorts.