This page separates reported trial results from statistical interpretation. Numerical trial results and trial-specific facts are restricted to the ClinicalTrials.gov record for NCT04487080. The registry provides the official trial record.
1. Trial at a Glance
MARIPOSA is a randomized, parallel, phase 3 treatment trial evaluating amivantamab and lazertinib combination therapy versus osimertinib in locally advanced or metastatic non-small cell lung cancer. The ClinicalTrials.gov record identifies progression-free survival (PFS) as the single registered primary endpoint and report one formal statistical analysis for that endpoint.
| Feature | MARIPOSA |
|---|---|
| Trial name | MARIPOSA |
| NCT ID | NCT04487080 |
| Phase | Phase 3 |
| Condition | Carcinoma, Non-Small-Cell Lung |
| Design | Randomized, parallel |
| Masking | Triple |
| Primary purpose | Treatment |
| Enrollment | 1074 |
| Primary endpoint | Progression-Free Survival (PFS) According to Response Evaluation Criteria in Solid Tumors (RECIST) Version 1.1 as Assessed by Blinded Independent Central Review (BICR) |
| Primary endpoint type | Time-to-event |
| Primary analysis | Log-rank test; hazard ratio from a stratified Cox regression model |
| Hypothesis type | Superiority |
| Trial status | Active, not recruiting |
2. Clinical Question
The central statistical question is whether amivantamab plus lazertinib improves progression-free survival compared with osimertinib plus placebo matching lazertinib in the randomized comparison reported by the registry.
Population
Participants with carcinoma, non-small-cell lung, in the locally advanced or metastatic setting described by the trial's brief title.
Intervention
Amivantamab and lazertinib combination therapy.
Comparator
Osimertinib with placebo matching lazertinib.
Primary question
Does the randomized intervention comparison produce a statistically superior progression-free survival outcome?
3. Trial Design
Amivantamab
- Amivantamab
- Experimental arm
- Serious adverse events: 205 affected of 421 at risk
Osimertinib + placebo matching lazertinib
- Osimertinib
- Placebo
- Active comparator arm
- Serious adverse events: 143 affected of 428 at risk
Lazertinib
- Lazertinib
- Experimental arm
- Serious adverse events: 75 affected of 213 at risk
Randomized parallel trial
- Allocation: randomized
- Design model: parallel
- Primary purpose: treatment
4. Trial Timing and Registry Status
Trial start
The ClinicalTrials.gov record lists September 30, 2020 as the trial start date.
Primary completion
The ClinicalTrials.gov record lists August 11, 2023 as the primary completion date.
Active, not recruiting
the ClinicalTrials.gov record in the ClinicalTrials.gov record reports the status as ACTIVE_NOT_RECRUITING.
The registry record reports enrollment of 1074 participants. The trial is sponsored by Janssen Research & Development, LLC, identified in the ClinicalTrials.gov record as an industry sponsor.
5. Primary Endpoint
| Endpoint | Registry definition / time frame | Analysis |
|---|---|---|
| Progression-Free Survival (PFS) | Progression-Free Survival (PFS) According to Response Evaluation Criteria in Solid Tumors (RECIST) Version 1.1 as Assessed by Blinded Independent Central Review (BICR). From randomization to either disease progression or death whichever occurs first (up to 32.8 months). | Log-rank test; hazard ratio estimated using a stratified Cox regression model |
How PFS is defined
The registry defines PFS as the time from randomization until the date of objective disease progression based on BICR using RECIST version 1.1 or death from any cause in the absence of progression, whichever came first.
For a time-to-event endpoint, participants do not all necessarily experience the event during observation. Participants without documented progression or death at the relevant end of observation contribute follow-up information through censoring. This is one reason survival-analysis methods are used instead of methods designed for a simple binary endpoint.
6. Statistical Methodology
Full analysis set
The primary analysis population was the full analysis set (FAS), which included all randomized participants. Participants were classified according to their assigned treatment arm regardless of the actual treatment received.
This is closely aligned with the intention-to-treat principle. Analyzing participants according to randomized assignment preserves the treatment comparison created by randomization and avoids redefining the primary efficacy population according to treatment exposure after randomization.
Log-rank test
The reported comparison method was the log-rank test. The log-rank test is designed for comparing time-to-event distributions between treatment groups while accounting for the timing of events and right censoring.
Rather than comparing only the percentage of participants who have progressed by a single fixed date, the log-rank framework uses the ordering of observed event times and the numbers at risk throughout follow-up.
At each relevant event time, the log-rank framework compares the observed number of events in the treatment groups with the number expected under a common survival experience, accumulating the evidence over follow-up.
Stratified Cox regression
The registry reports that the hazard ratio and its 95% confidence interval were estimated using a stratified Cox's regression model with treatment as the sole explanatory variable.
The ClinicalTrials.gov record identifies the model as stratified but do not specify the variables used to form the strata. Therefore, no particular stratification factors are assigned to the model on this page.
A hazard ratio is a relative time-to-event measure. It summarizes the estimated separation between event hazards under the fitted model; it is not a direct measure of the probability of progression for an individual participant.
Superiority hypothesis
The ClinicalTrials.gov record classifies the hypothesis type as superiority. Thus, the statistical objective was to determine whether the randomized comparison demonstrated evidence of a treatment difference favoring the experimental comparison rather than to establish that two treatments were sufficiently similar within a predefined non-inferiority margin.
Blinded independent central review
The primary endpoint was assessed using Blinded Independent Central Review (BICR) according to RECIST version 1.1. Independent central assessment can reduce the influence of knowledge of treatment assignment on radiologic classification of progression.
7. Primary Result: Progression-Free Survival
The ClinicalTrials.gov record contains one formal statistical analysis, corresponding to the registered primary PFS endpoint. The reported comparison was amivantamab plus lazertinib versus osimertinib plus placebo matching lazertinib.
Hazard ratio for progression or death
95% CI: 0.58–0.85 · P = 0.0002
Two-sided 95% confidence interval; superiority hypothesis.
| Primary endpoint | Experimental comparison | Active comparator | Effect estimate |
|---|---|---|---|
| Progression-Free Survival | Arm A: Amivantamab + Lazertinib | Arm B: Osimertinib + Placebo Matching Lazertinib | HR 0.70 95% CI 0.58–0.85 P = 0.0002 |
A hazard ratio of 0.70 means that, under the fitted stratified Cox model, the estimated instantaneous rate of progression or death in the amivantamab-plus-lazertinib group was approximately 70% of the corresponding estimated rate in the osimertinib-plus-placebo group. Expressed as a simple relative interpretation, this corresponds to an estimated 30% lower instantaneous hazard for progression or death.
The HR does not mean that 30% of participants avoided progression, that every participant experienced exactly a 30% reduction in risk, or that the probability of progression or death was 30% lower at every particular time point.
The 95% CI of 0.58–0.85 describes statistical uncertainty around the estimated hazard ratio under the model and sampling framework. It does not describe the range of individual patient outcomes. The interval also provides information about precision: it is substantially narrower than an interval that would be centered on the same estimate but based on much less information.
The p-value of 0.0002 addresses evidence against the relevant null hypothesis under the specified statistical framework. It does not measure the size of the treatment effect, the probability that the treatment works, or the clinical importance of the effect. Effect size is conveyed by the HR and its confidence interval, while the p-value addresses statistical evidence against the null.
The analysis is based on a time-to-event framework with censoring and a Cox model. Interpretation of a single Cox hazard ratio is most straightforward when the proportional-hazards structure is a reasonable description of the treatment comparison over time. The ClinicalTrials.gov record does not report a separate assessment of that assumption.
What the confidence interval tells us
The interval from 0.58 to 0.85 gives a range of values that represents uncertainty around the estimated HR of 0.70. Because the entire reported interval is below 1, the confidence interval is consistent with a lower estimated hazard of progression or death for the experimental comparison throughout the interval of model-compatible effects represented by this 95% confidence interval.
Importantly, the confidence interval is not a prediction interval for future individual patients. Nor does it establish that the true effect is equally strong at every time point.
Why the p-value should not be read as an effect-size measure
A p-value is influenced by both the magnitude of the observed treatment difference and the amount of statistical information available. A very small p-value therefore does not automatically imply a large clinical effect, just as a larger p-value does not prove that treatments have identical effects.
For this reason, the useful statistical reading of this result combines the HR of 0.70, its 95% CI of 0.58–0.85, the p-value of 0.0002, the randomized analysis population, and the time-to-event design rather than focusing on the p-value alone.
8. Safety Results
The ClinicalTrials.gov record reports serious adverse events by randomized arm as affected participants divided by participants at risk. These data are presented exactly as reported in the registry.
| Arm | Treatment description | Serious adverse events |
|---|---|---|
| Arm A | Experimental; open-label; Amivantamab | 205/421 |
| Arm B | Active comparator; double-blind; Osimertinib | 143/428 |
| Arm C | Experimental; double-blind; Lazertinib | 75/213 |
These are safety counts rather than a formal efficacy comparison. The registry-reported statistical analysis is for PFS in Arm A versus Arm B; it does not provide a formal statistical comparison of serious adverse events among the three arms.
9. Statistical Methods Explained
Why was a log-rank test used?
PFS is a time-to-event endpoint: participants can experience progression or death at different times, while others may remain event-free at the end of observation and therefore be censored. The log-rank test is specifically constructed for comparing survival-type time-to-event distributions while incorporating that timing and censoring information.
Why is the primary analysis based on all randomized participants?
The FAS included all randomized participants and classified them according to assigned treatment regardless of actual treatment received. This preserves the randomized comparison and is consistent with the core intention-to-treat principle. Excluding participants after randomization because of treatment changes can introduce selection into the efficacy analysis.
What does an HR of 0.70 mean?
An HR of 0.70 means the fitted model estimates the instantaneous hazard of progression or death to be 0.70 times that of the comparator. A useful descriptive translation is an estimated 30% lower instantaneous hazard. It is not equivalent to saying that 30% of participants were protected from progression or death.
Why was a stratified Cox model used?
The registry specifically reports a stratified Cox regression model. Stratification allows the baseline hazard to vary across specified strata while estimating the treatment effect through the Cox model. The ClinicalTrials.gov record does not identify the variables used to define those strata, so the page does not infer them.
Why use BICR for the PFS endpoint?
PFS depends partly on identifying objective disease progression. The registered endpoint specifies assessment by Blinded Independent Central Review using RECIST version 1.1. Blinded central assessment provides an assessment framework intended to limit the influence of treatment assignment on progression classification.
Why is the confidence interval as important as the p-value?
The p-value addresses statistical evidence against the null hypothesis, whereas the confidence interval describes uncertainty around the estimated treatment effect. The reported interval of 0.58–0.85 therefore provides information about the range of hazard-ratio values compatible with the statistical analysis, while the p-value of 0.0002 provides the corresponding hypothesis-test evidence.
What does "up to 32.8 months" mean in the endpoint definition?
The registered time frame states that PFS was followed from randomization to progression or death, whichever occurred first, up to 32.8 months. This identifies the registry's stated observation horizon for the endpoint rather than representing a median PFS value.
10. Intention-to-Treat and Censoring
Intention-to-treat principle
The registry-reported primary analysis used the full analysis set, including all randomized participants and classifying them according to assigned treatment. This is important because randomization establishes the basis for the comparison before subsequent treatment exposure or deviations occur.
Randomization
Randomization creates the treatment groups before outcome information is observed, providing the fundamental structure for an unbiased treatment comparison under the trial design.
Analysis assignment
The FAS preserves the randomized assignment even when actual treatment received differs from assignment.
Right censoring
Participants who have not experienced progression or death by the relevant observation endpoint can contribute follow-up information up to their censoring time.
Event timing
Unlike a simple binary endpoint, PFS incorporates when progression or death occurs, not merely whether an event eventually occurred during observation.
Censoring is a critical component of survival analysis. Standard Kaplan-Meier, log-rank, and Cox methods rely on assumptions about the relationship between censoring and the event process. The ClinicalTrials.gov record does not provide a separate diagnostic analysis of censoring, so no additional assumption is attributed specifically to MARIPOSA beyond the standard framework of the reported methods.
11. Hazard Ratio, Proportional Hazards, and Time-to-Event Interpretation
For MARIPOSA, the reported HR of 0.70 is a relative measure of the progression-or-death hazard from the stratified Cox model. It should be interpreted together with the confidence interval and the underlying time-to-event design.
Hazard is not the same as risk
Risk generally refers to the probability of an event over a defined period. Hazard is an instantaneous event rate conditional on remaining event-free up to a particular time. Because these quantities are different, an HR of 0.70 should not automatically be translated into a 30% lower cumulative probability of progression or death at every fixed time point.
The proportional-hazards assumption
The Cox model's conventional hazard-ratio interpretation is especially straightforward when the relative hazards are reasonably stable over time. If hazards cross or the treatment effect changes substantially over follow-up, a single HR can compress a more complicated time-varying pattern into one summary measure.
The ClinicalTrials.gov record reports the stratified Cox model but do not report a separate proportional-hazards diagnostic. Accordingly, the HR of 0.70 should be treated as the reported model-based summary rather than as proof that the proportional-hazards assumption holds at every time point.
12. Statistical Analysis Workflow
This workflow illustrates an important distinction between a hypothesis test and an effect estimate. The log-rank test provides the formal comparison of the time-to-event distributions, while the stratified Cox model provides the reported hazard-ratio estimate and its confidence interval.
| Statistical component | Role in MARIPOSA |
|---|---|
| Randomization | Creates the randomized treatment comparison. |
| Full analysis set | Includes all randomized participants according to assigned treatment. |
| RECIST 1.1 / BICR | Defines and assesses the registered PFS progression component. |
| Log-rank test | Formal comparison of the time-to-event distributions. |
| Stratified Cox regression | Estimates the hazard ratio and its 95% confidence interval. |
| P-value | Quantifies evidence against the relevant null hypothesis under the specified test. |
| Confidence interval | Quantifies uncertainty around the hazard-ratio estimate. |
13. Multiplicity and Additional Analyses
the ClinicalTrials.gov record reports 1 primary endpoint, 16 posted outcome measures, and 1 statistical analysis. The single formal analysis in the ClinicalTrials.gov record is the PFS comparison between Arm A and Arm B.
Because the ClinicalTrials.gov record do not describe a multiplicity-adjustment strategy, endpoint hierarchy, alpha allocation, or interim-analysis plan, none is inferred here. The presence of multiple registered or posted outcome measures does not by itself establish how type I error was controlled across them.
14. Stratified Analysis
The primary statistical analysis states that the hazard ratio and its 95% confidence interval were estimated from a stratified Cox's regression model. The analysis text identifies treatment as the sole explanatory variable.
What stratification does
Stratification allows the underlying baseline hazard to differ across specified strata while estimating a common treatment effect within the Cox framework.
What is not reported
The ClinicalTrials.gov record does not identify the variables or levels defining the Cox-model strata.
That distinction is important. It would be inappropriate to reconstruct specific stratification factors from general knowledge of the trial or from another publication when the ClinicalTrials.gov record does not state them.
15. Results Scope and What Is Not Reported Here
The ClinicalTrials.gov record supports a full results page for the primary PFS endpoint because a formal statistical analysis is posted. However, only one statistical analysis is provided in the trial data.
| Topic | What the ClinicalTrials.gov record supports |
|---|---|
| Primary PFS | Formal analysis with HR, 95% CI, p-value, analysis population, and method. |
| Secondary efficacy endpoints | The registry profile reports 16 outcome measures overall, but no additional statistical-analysis results are reported in the ClinicalTrials.gov record for this page. |
| Median PFS | Not reported in the ClinicalTrials.gov record. |
| Kaplan-Meier estimates | Not reported. |
| Subgroup estimates | Not reported. |
| Baseline characteristics | Not reported. |
| Formal safety comparison | Not reported; only serious-adverse-event affected/at-risk counts by arm are provided. |
| Multiplicity strategy | Not reported. |
| Interim-analysis strategy | Not reported. |
| Missing-data or imputation strategy | Not reported. |
| Bayesian methods | Not reported in the registry-reported analysis. |
This distinction is central to responsible trial interpretation. A results page should not fill gaps in the registry extract with numbers taken from memory or from an unrelated analysis. The statistical story presented here is therefore limited to what the ClinicalTrials.gov record actually document.
16. Limitations
- Single formal analysis reported: the ClinicalTrials.gov record contain one statistical analysis, corresponding to the registered primary PFS endpoint. Additional outcome measures are not accompanied by formal statistical-analysis results in the ClinicalTrials.gov record.
- Limited effect measures: the registry-reported primary analysis contains a hazard ratio, confidence interval, and p-value, but does not provide median PFS, fixed-time PFS estimates, or Kaplan-Meier estimates.
- Stratification details: the analysis is described as stratified, but the ClinicalTrials.gov record does not identify the variables defining the strata.
- Proportional-hazards interpretation: the reported Cox HR is a model-based summary. The ClinicalTrials.gov record does not report a separate assessment of proportional hazards.
- Censoring: time-to-event analyses depend on censoring assumptions. The ClinicalTrials.gov record does not provide a detailed censoring diagnostic or missing-follow-up analysis.
- Multiplicity: the ClinicalTrials.gov record does not specify how multiple registered or posted outcome measures were handled in the type I error framework.
- Safety inference: serious-adverse-event counts are reported by arm, but no formal comparative safety analysis is reported.
- Generalizability: the registry description identifies a specific trial population and treatment setting. The ClinicalTrials.gov record does not provide enough baseline information to evaluate how representative the enrolled population is of broader clinical populations.
17. Why This Trial Matters Statistically
MARIPOSA is a useful teaching case because its registry-reported primary analysis combines the core elements of randomized time-to-event inference: randomization, a prospectively defined progression-free survival endpoint, central blinded assessment, an intention-to-treat-compatible analysis population, a log-rank comparison, and a stratified Cox hazard ratio.
| Concept | How it appears in MARIPOSA |
|---|---|
| Randomization | The trial uses randomized allocation in a parallel phase 3 design. |
| Blinding | The registry profile reports triple masking; the arm descriptions distinguish open-label and double-blind arms. |
| Intention-to-treat principle | The FAS includes all randomized participants according to assigned treatment. |
| Time-to-event endpoint | PFS is measured from randomization to progression or death, whichever occurs first. |
| BICR | PFS progression is assessed by blinded independent central review using RECIST version 1.1. |
| Log-rank test | Used for the reported primary comparison. |
| Hazard ratio | The reported treatment effect is HR 0.70. |
| Confidence interval | The 95% two-sided CI is 0.58–0.85. |
| P-value | The reported p-value is 0.0002. |
| Stratified Cox model | Used to estimate the HR and its confidence interval. |
| Superiority testing | The registered hypothesis type is superiority. |
| Three-arm design | the ClinicalTrials.gov record reports 3 arms, while the registry-reported primary statistical comparison is Arm A versus Arm B. |
18. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The reported randomized comparison of Arm A versus Arm B produced a PFS hazard ratio of 0.70, with a two-sided 95% CI of 0.58–0.85 and p = 0.0002 under the reported log-rank and stratified Cox analysis framework.
Clinical interpretation
The ClinicalTrials.gov record establishes a statistically reported difference in the registered PFS endpoint. They do not, by themselves, provide median PFS, fixed-time PFS rates, subgroup effects, or a complete benefit-risk assessment.
The distinction is important. Statistical evidence addresses the compatibility of the observed data with a null hypothesis under a specified analysis. Clinical interpretation requires additional information about the magnitude, duration, patient population, treatment burden, safety, and other outcomes. The ClinicalTrials.gov record does not provide enough information to make all of those assessments.
19. Understanding the Three-Arm Structure
the ClinicalTrials.gov record reports 3 arms, while the registry-reported primary statistical analysis compares Arm A (Amivantamab + Lazertinib) with Arm B (Osimertinib + Placebo Matching Lazertinib).
| Arm | Registry description in the ClinicalTrials.gov record | Role in registry-reported primary analysis |
|---|---|---|
| Arm A | Experimental; open-label; Amivantamab | Included in primary PFS comparison |
| Arm B | Active comparator; double-blind; Osimertinib | Included in primary PFS comparison |
| Arm C | Experimental; double-blind; Lazertinib | Not included in the registry-reported formal primary statistical comparison |
This is a useful reminder that the number of randomized arms does not automatically tell us which pairwise comparisons constitute a primary hypothesis. The statistical analysis definition posted on ClinicalTrials.gov for MARIPOSA explicitly identifies the Arm A versus Arm B comparison.
20. What the PFS Result Does — and Does Not — Mean
The reported HR of 0.70 is a model-based relative measure of the hazard of progression or death. A value below 1 indicates a lower estimated hazard in the experimental comparison relative to the active comparator under the fitted model.
The HR is not a statement that 70% of participants remained progression-free, nor that 30% of participants were spared progression or death. Absolute event probabilities require time-specific survival estimates, which are not included in the ClinicalTrials.gov record.
The 95% CI of 0.58–0.85 quantifies uncertainty around the estimated HR. It does not represent the distribution of treatment effects across individual patients.
The reported p-value of 0.0002 indicates strong statistical evidence against the relevant null hypothesis within the stated testing framework. It is not a measure of effect magnitude or clinical importance.
21. Related Statistical Tutorials
Learn more about the methods used in this trial:
22. Related Statistical Calculators
23. Sources
- ClinicalTrials.gov: NCT04487080 — MARIPOSA.
- PubMed: PMID 40923797.
- PubMed: PMID 40617394.
- PubMed: PMID 38924756.
- PubMed: PMID 34911336.
Continue through the Clinical Biostats statistical pathway
Use the trial as a practical starting point for learning how randomized time-to-event endpoints connect study design, survival analysis, hazard ratios, confidence intervals, and hypothesis testing.
24. Record Summary
MARIPOSA provides a clear example of a randomized phase 3 time-to-event analysis. The ClinicalTrials.gov record describes a 1074-participant, three-arm, randomized parallel trial with a registered primary endpoint of PFS assessed according to RECIST version 1.1 by BICR. The registry-reported formal analysis used the full analysis set, a stratified log-rank test, and a stratified Cox regression model.
For the reported Arm A versus Arm B comparison, the estimated PFS hazard ratio was 0.70, with a two-sided 95% CI of 0.58–0.85 and p = 0.0002. The statistical interpretation is that the fitted model estimated a lower instantaneous hazard of progression or death for the amivantamab-plus-lazertinib comparison relative to the osimertinib-plus-placebo comparison. The result should be interpreted as a model-based relative time-to-event estimate rather than as an absolute probability or a statement about every individual participant.
The ClinicalTrials.gov record provides serious-adverse-event affected/at-risk counts for all three arms, but no formal comparative safety analysis. Likewise, the ClinicalTrials.gov record do not establish median PFS, fixed-time PFS estimates, subgroup effects, multiplicity procedures, interim-analysis methods, or missing-data methods. Keeping those distinctions explicit is essential when translating a registry record into a statistical teaching page.