This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
MOUNTAINEER-03 is a randomized, open-label, parallel-group phase 3 treatment trial evaluating tucatinib with trastuzumab and mFOLFOX6 versus standard-of-care treatment in first-line HER2+ metastatic colorectal cancer. The registry lists a planned enrollment of 400 participants and one primary efficacy endpoint: progression-free survival assessed by blinded independent central review according to RECIST v1.1.
| Feature | MOUNTAINEER-03 |
|---|---|
| Phase | Phase 3 |
| Condition | Colorectal Neoplasms |
| Study title | A Study of Tucatinib With Trastuzumab and mFOLFOX6 Versus Standard of Care Treatment in First-line HER2+ Metastatic Colorectal Cancer |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Planned enrollment | 400 |
| Number of arms | 2 |
| Recruitment status | Recruiting |
| Start date | 2022-10-24 |
| Primary completion date | 2027-12-31 |
| Lead sponsor | Seagen, a wholly owned subsidiary of Pfizer |
| Sponsor type | Industry |
| ClinicalTrials.gov | NCT05253651 |
2. Clinical Question
The central question is whether a first-line regimen containing tucatinib with trastuzumab and mFOLFOX6 improves progression-free survival compared with standard-of-care treatment in patients with HER2+ metastatic colorectal cancer.
Population
Patients with HER2+ metastatic colorectal cancer in the first-line treatment setting.
Intervention
Tucatinib with trastuzumab and mFOLFOX6.
Comparator
Standard-of-care treatment. The registry lists bevacizumab, cetuximab, oxaliplatin, leucovorin, levoleucovorin, and fluorouracil among the study interventions.
Primary question
Does the randomized intervention strategy change the time from randomization to disease progression or death from any cause?
3. Trial Design
Tucatinib + trastuzumab + mFOLFOX6
- Tucatinib
- Trastuzumab
- mFOLFOX6
Standard-of-care treatment
- Standard-of-care treatment
- Registry-listed study interventions include bevacizumab and cetuximab
- Registry-listed chemotherapy components include oxaliplatin, leucovorin, levoleucovorin, and fluorouracil
The design has an important statistical feature: treatment is randomized but the trial is not masked. Randomization protects the comparison from many forms of allocation bias, whereas the absence of masking means that participants and investigators know treatment assignment. The primary endpoint incorporates blinded independent central review, which provides an independent radiologic assessment of disease progression.
4. Trial Timeline and Status
Study start
The registry lists 2022-10-24 as the study start date.
Recruiting
The ClinicalTrials.gov record identifies MOUNTAINEER-03 as recruiting.
Primary completion date
The registry lists 2027-12-31 as the primary completion date.
The planned timeline matters statistically because the primary endpoint is a time-to-event outcome. The amount of follow-up available for each participant, the accumulation of progression or death events, and the censoring pattern determine how much information is ultimately available for estimating the treatment effect.
5. Primary Endpoint
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Progression-free survival (PFS) per Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST v1.1) by Blinded Independent Central Review (BICR) | The time from the date of randomization to the BICR assessment of disease progression according to RECIST v1.1 or death from any cause | Up to approximately 3 years |
The endpoint has two possible event types: documented disease progression according to RECIST v1.1 or death from any cause. A participant therefore does not need to have radiologically documented progression for the PFS event to occur; death is also an event.
Participants who have neither progression nor death by their relevant follow-up point contribute information until they are censored according to the prespecified analysis rules.
6. Statistical Methodology
The ClinicalTrials.gov record identifies progression-free survival as the primary endpoint and specifies that progression is assessed according to RECIST v1.1 by blinded independent central review. The registry record does not report a statistical test, regression model, event target, alpha level, multiplicity procedure, interim-analysis rule, or prespecified treatment-effect threshold for the primary endpoint.
For a randomized two-arm trial with a primary time-to-event endpoint such as PFS, the standard statistical framework is based on the full distribution of event times rather than simply comparing the proportion of participants who have progressed at one fixed date.
Kaplan-Meier estimation
A Kaplan-Meier analysis would estimate the probability of remaining progression-free over time while appropriately accounting for participants who are censored before experiencing progression or death.
Here, di is the number of events at time ti, while ni is the number of participants at risk immediately before that time.
The resulting survival curve represents the estimated probability of remaining event-free over time. For PFS, the event is the first occurrence of RECIST-defined progression or death.
Log-rank comparison
A log-rank test is a conventional way to compare two survival distributions when the scientific question concerns whether event-time experience differs between randomized treatment groups. Unlike a simple proportion comparison, it incorporates the timing of events throughout follow-up.
Cox proportional-hazards model
A Cox proportional-hazards model is commonly used to estimate a hazard ratio comparing two treatment groups. If the prespecified analysis uses this model, the hazard ratio summarizes the relative instantaneous event rate over the analyzed follow-up under the proportional-hazards model.
A hazard ratio is not a percentage of participants who avoid progression, and it is not the same quantity as a difference between two fixed-time PFS percentages.
Blinded independent central review
The primary endpoint specifies BICR assessment. This means that the endpoint determination is based on an independent central review process rather than relying solely on the treating investigator's assessment. Statistically, this is important because progression is part of the event definition itself: differences in how progression is assessed can otherwise affect the observed timing of PFS events.
7. Planned Analysis
No results have been posted on ClinicalTrials.gov for the primary endpoint. The registry identifies PFS per RECIST v1.1 by BICR as the primary endpoint and specifies a time frame of up to approximately 3 years.
Primary statistical question
Compare the time from randomization to BICR-assessed RECIST v1.1 progression or death from any cause between the randomized treatment strategies.
For this type of endpoint, the usual analysis would describe PFS using Kaplan-Meier estimates and compare the treatment groups using an appropriate time-to-event hypothesis test, commonly a log-rank test. A Cox proportional-hazards model may be used to estimate a hazard ratio and its confidence interval when specified in the statistical analysis plan.
The appropriate analysis population, stratification factors, censoring rules, handling of missing or incomplete assessments, event target, type I error allocation, and any interim-analysis procedure should be determined from the prespecified statistical analysis plan if those details are to be evaluated. The registry record does not report those statistical specifications.
8. Statistical Methods Explained
What exactly is being measured by PFS?
PFS measures the elapsed time from randomization until the first qualifying event: disease progression according to RECIST v1.1 by BICR or death from any cause. This definition makes progression and death competing ways of reaching the same endpoint.
Why use time-to-event methods rather than a simple response percentage?
A response percentage describes whether a particular tumor response criterion was achieved. PFS instead asks how long participants remain free of progression or death. Participants can have different follow-up durations, and survival methods allow those unequal observation times to be incorporated rather than discarding participants who have not yet experienced an event.
What does censoring mean in a PFS analysis?
Censoring occurs when the relevant event has not been observed by the point at which usable follow-up ends under the prespecified rules. A censored participant contributes information about being progression-free up to the censoring time, but does not contribute a known event time beyond that point.
Why does blinded independent central review matter?
Progression is part of the endpoint definition. Independent central review creates a standardized assessment process separate from the treating investigator. This can reduce the influence of treatment knowledge on radiologic progression assessments, although the overall trial itself is open label.
What would a hazard ratio represent?
A hazard ratio would compare the estimated instantaneous rates of progression or death between the randomized groups. An HR below 1 would indicate a lower estimated event rate in the intervention group under the fitted model; an HR above 1 would indicate a higher estimated event rate.
Why is the confidence interval important?
A confidence interval describes statistical uncertainty around an estimated treatment effect. A narrow interval indicates greater precision than a wide interval, all else equal. It does not describe the range of outcomes that individual participants will experience.
Why can a p-value not substitute for the treatment-effect estimate?
A p-value addresses evidence against a specified null hypothesis under the statistical model and assumptions used. It does not describe the magnitude of the treatment effect. For clinical interpretation, the effect estimate, its confidence interval, the absolute time-to-event experience, and the design and analysis assumptions all matter.
9. Analysis Population and Randomization
The registry identifies MOUNTAINEER-03 as randomized, with planned enrollment of 400 participants and two parallel treatment arms. Randomization is the central design mechanism for creating comparable treatment groups in expectation before outcomes are observed.
Why randomization matters
Random allocation helps balance measured and unmeasured prognostic factors between treatment groups in expectation, supporting a causal interpretation of differences in outcomes between the randomized strategies.
Why analysis population matters
Time-to-event results can depend on who is included, how treatment assignment is handled, and how censoring is defined. The registry record does not report the prespecified primary analysis population.
Because the trial is still listed as recruiting, the planned enrollment should not be interpreted as the number of participants who have completed follow-up or the number included in an eventual efficacy analysis.
10. Open-Label Design and Endpoint Blinding
The study is identified as having no masking. Participants and investigators therefore know the assigned treatment. At the same time, the primary endpoint specifies blinded independent central review for radiologic assessment.
| Feature | Statistical relevance |
|---|---|
| Randomized allocation | Supports comparability of treatment groups at baseline in expectation. |
| No masking | Knowledge of treatment assignment can influence behavior, treatment decisions, reporting, or investigator assessments. |
| BICR | Provides an independent central assessment of radiologic progression for the primary endpoint. |
| RECIST v1.1 | Provides a defined framework for determining disease progression. |
The combination of randomization and blinded central review is therefore relevant to interpretation of PFS. Randomization addresses allocation, while BICR addresses an important component of endpoint ascertainment.
11. RECIST v1.1 and Event Definition
The registry defines the primary PFS endpoint using RECIST v1.1. In this trial, the registry's endpoint description links the radiologic progression determination specifically to BICR.
The event occurs at the first qualifying endpoint event after randomization under the prespecified assessment and censoring rules.
This construction is important because progression-free survival is not equivalent to time until tumor shrinkage, objective response, treatment discontinuation, or a particular radiologic measurement. It is specifically defined around progression or death.
12. Censoring and Missing Data
The registry identifies PFS as a time-to-event endpoint but does not report the detailed censoring rules or missing-assessment algorithm. Those rules are important because the timing of an observed progression can depend on how assessments surrounding an incomplete or missed evaluation are handled.
In a typical PFS analysis, censoring rules are specified before the analysis so that the treatment comparison does not depend on ad hoc decisions made after observing the results. The statistical analysis plan would ordinarily define how to handle situations such as a participant remaining progression-free at the end of available follow-up, missing tumor assessments, withdrawal from follow-up, or other events that affect endpoint ascertainment.
13. Proportional-Hazards Assumption
If a Cox proportional-hazards model is used for the primary PFS analysis, the hazard ratio relies on a proportional-hazards framework. In simplified terms, this assumes that the relative event rate between treatment groups is reasonably stable over the relevant time scale.
This does not mean that the absolute hazard must remain constant. Rather, the model assumes that the ratio of the hazards between groups is adequately represented by a common relative effect.
If proportional hazards is reasonable
A single hazard ratio can provide a concise summary of the relative treatment effect over the analyzed period.
If proportional hazards is not reasonable
A single hazard ratio may obscure changes in the relative treatment effect over time, making Kaplan-Meier curves and time-specific measures particularly important for interpretation.
The registry record does not report a proportional-hazards diagnostic or an alternative time-varying analysis.
14. Multiplicity and Interim Analysis
The ClinicalTrials.gov record identifies one registered primary endpoint: PFS. It does not report an alpha-allocation scheme, multiplicity adjustment, interim-analysis schedule, stopping boundary, or other formal type I error-control procedure.
These design details matter because repeated statistical testing can increase the probability of incorrectly rejecting a true null hypothesis if the testing process is not accounted for. When a trial includes planned interim looks, the statistical analysis plan normally specifies how the overall type I error is controlled across those looks.
| Design question | What the registry reports | Statistical importance |
|---|---|---|
| Primary endpoint | PFS | Defines the principal efficacy question. |
| Number of primary endpoints | One registered primary endpoint | Provides a single primary time-to-event target in the registry record. |
| Interim analysis | Not reported in the ClinicalTrials.gov record | Would determine whether alpha spending or other sequential-error control is required. |
| Multiplicity procedure | Not reported in the ClinicalTrials.gov record | Would be relevant if multiple confirmatory hypotheses were tested. |
| Alpha level | Not reported in the ClinicalTrials.gov record | Defines the prespecified type I error threshold for formal testing. |
15. Safety and Other Endpoints
The registry information identifies the study interventions but does not provide posted statistical safety results, serious adverse-event counts by treatment arm, or secondary endpoint results. No formal statistical analyses are posted to ClinicalTrials.gov in the record represented here.
Safety analysis in a randomized treatment trial is normally handled separately from the primary efficacy analysis. Adverse events are generally summarized by treatment exposure, severity, seriousness, and other prespecified categories rather than using a single time-to-event hazard ratio for all safety information.
Because the registry information does not contain arm-specific serious adverse-event results, no numerical safety comparison is presented here.
16. What a PFS Hazard Ratio Would — and Would Not — Mean
If the final analysis reports a PFS hazard ratio below 1, that would indicate a lower estimated instantaneous rate of progression or death in the intervention group relative to the comparator under the fitted survival model.
It would not mean that the same percentage of participants are guaranteed to avoid progression, nor would it directly describe the difference in PFS at any single calendar time.
A 95% confidence interval around the hazard ratio would quantify statistical uncertainty around the estimated relative treatment effect. Its width would provide information about precision, while its position relative to the null value of 1 would inform conventional hypothesis-test interpretation.
Time-specific PFS probabilities and median PFS, if reported, would provide an absolute perspective that a hazard ratio alone cannot provide. The relative and absolute summaries answer different questions and are most informative when interpreted together.
17. Statistical Interpretation of the Trial Design
The trial's statistical architecture is centered on a randomized comparison of two treatment strategies and a single primary time-to-event endpoint. This is a natural design when the scientific question concerns whether one treatment strategy changes the time until progression or death rather than merely whether an individual tumor response occurs.
| Concept | How it appears in MOUNTAINEER-03 |
|---|---|
| Randomization | Participants are assigned randomly to two treatment arms. |
| Parallel design | The registry identifies the design model as parallel. |
| Open-label treatment | The registry identifies masking as none. |
| Primary time-to-event endpoint | PFS from randomization to BICR-assessed RECIST v1.1 progression or death. |
| Independent assessment | BICR is incorporated into the primary endpoint definition. |
| Defined response framework | RECIST v1.1 is specified for progression assessment. |
| Planned follow-up window | Primary endpoint time frame is up to approximately 3 years. |
18. Limitations
- No posted efficacy results: the ClinicalTrials.gov record represented here does not contain posted statistical analyses or numerical primary endpoint results.
- Recruiting status: the registry identifies the study as recruiting, so the planned enrollment of 400 should not be interpreted as a completed efficacy population.
- Incomplete statistical specification: the registry information does not report the primary statistical test, Cox-model specification, event target, alpha level, interim-analysis plan, or multiplicity procedure.
- Censoring rules: detailed rules for handling incomplete follow-up and missing assessments are not reported in the registry information.
- Open-label design: treatment assignment is not masked, which can influence some aspects of participant or investigator behavior and assessment.
- Endpoint ascertainment: BICR provides an independent radiologic assessment, but the overall trial remains open label.
- Generalizability: the trial is specifically focused on first-line HER2+ metastatic colorectal cancer, so results from the eventual study population would need to be interpreted in relation to those eligibility characteristics.
19. Why This Trial Matters Statistically
MOUNTAINEER-03 provides a useful teaching example of how a modern randomized oncology trial can be structured around a time-to-event endpoint while separating treatment allocation from endpoint assessment.
| Statistical concept | Application in this trial |
|---|---|
| Randomized comparison | Two treatment strategies are compared using randomized allocation. |
| Time-to-event analysis | The primary endpoint measures time from randomization to progression or death. |
| Kaplan-Meier estimation | A standard framework for estimating the PFS distribution while accounting for censoring. |
| Hazard ratio | A conventional relative treatment-effect measure for time-to-event data when a Cox model is appropriate. |
| Confidence interval | Quantifies uncertainty around the estimated treatment effect. |
| Log-rank testing | A conventional method for comparing survival distributions between randomized groups. |
| RECIST v1.1 | Provides the framework for determining radiologic progression. |
| BICR | Provides an independent central assessment of the radiologic component of PFS. |
| Censoring | Allows participants without an observed event during available follow-up to contribute partial time-to-event information. |
| Open-label design | Creates a distinction between knowledge of treatment assignment and independent assessment of the primary radiologic endpoint. |
20. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The primary question is a comparison of two randomized time-to-event distributions. The appropriate treatment-effect summary would quantify how progression-free survival differs between the randomized groups while accounting for censoring.
Clinical interpretation
The clinical question concerns whether adding tucatinib and trastuzumab to mFOLFOX6 changes the course of first-line HER2+ metastatic colorectal cancer relative to standard-of-care treatment. That question ultimately depends on the mature efficacy and safety results.
These two perspectives should remain distinct. Statistical significance, if established, describes evidence under a prespecified statistical framework; clinical importance depends on the magnitude and durability of the effect, absolute outcomes, safety, treatment burden, and the population in which the evidence was generated.
21. Planned Analysis Framework
Because the primary endpoint is PFS, a complete eventual statistical report would ordinarily contain several complementary pieces of information rather than a single number.
These measures are complementary. A hazard ratio summarizes relative event rates, while a Kaplan-Meier curve and time-specific estimates show how the absolute probability of remaining progression-free evolves over time.
22. Sources
- ClinicalTrials.gov: MOUNTAINEER-03, NCT05253651.
- PubMed: PubMed record associated with the MOUNTAINEER-03 evidence base.
Continue through Clinical Biostats
Explore statistical tutorials, calculators, and additional clinical-trial analyses covering the methods used to design and interpret randomized studies.
23. Record Summary
MOUNTAINEER-03 is a randomized, parallel-group phase 3 treatment trial with planned enrollment of 400 participants. It evaluates tucatinib with trastuzumab and mFOLFOX6 versus standard-of-care treatment in first-line HER2+ metastatic colorectal cancer. Its registered primary endpoint is progression-free survival, defined as the time from randomization to BICR-assessed disease progression according to RECIST v1.1 or death from any cause, with a time frame of up to approximately 3 years.
From a statistical perspective, the central feature of the trial is the use of a randomized treatment comparison with a time-to-event primary endpoint and independent central assessment of radiologic progression. The eventual interpretation should combine the estimated PFS distribution, relative treatment effect, confidence interval, absolute time-specific outcomes, censoring behavior, analysis population, and prespecified type I error framework rather than relying on a single statistical quantity.