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 ClinicalTrials.gov record.
1. Trial at a Glance
MONALEESA-2 was a randomized, quadruple-masked, parallel-group phase 3 study of ribociclib plus letrozole versus placebo plus letrozole in postmenopausal women with advanced or metastatic breast cancer. The trial enrolled 668 participants and posted four statistical analyses, including a formal analysis of the primary progression-free survival endpoint.
| Feature | MONALEESA-2 |
|---|---|
| Trial name | MONALEESA-2 |
| NCT identifier | NCT01958021 |
| Brief title | Study of Efficacy and Safety of LEE011 in Postmenopausal Women With Advanced Breast Cancer |
| Phase | Phase 3 |
| Status | Completed |
| Enrollment | 668 |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Lead sponsor | Novartis Pharmaceuticals |
| Sponsor type | Industry |
| Conditions | Advanced, Metastatic Breast Cancer |
| Start | December 17, 2013 |
| Primary completion | January 29, 2016 |
2. Clinical Question
The trial addresses whether adding ribociclib to letrozole improves clinical outcomes compared with letrozole plus placebo in postmenopausal women with advanced or metastatic breast cancer. The registered primary endpoint was progression-free survival by investigator assessment.
Population
Postmenopausal women with advanced or metastatic breast cancer, as specified by the trial's brief title and condition description.
Intervention
Ribociclib plus letrozole.
Comparator
Placebo plus letrozole.
Primary question
Does ribociclib plus letrozole improve progression-free survival compared with placebo plus letrozole?
3. Trial Design
Ribociclib + Letrozole
- Ribociclib
- Letrozole
- Serious adverse events affected 108 of 334 participants in the registry-reported arm-level safety data.
Placebo + Letrozole
- Placebo
- Letrozole
- Serious adverse events affected 62 of 330 participants in the registry-reported arm-level safety data.
4. Analysis Population and Statistical Framework
The posted statistical analyses specify the Full Analysis Set (FAS), including all randomized participants, for the primary progression-free survival analysis and the three secondary analyses reported in the ClinicalTrials.gov record.
| Endpoint | Analysis population | Comparison | Method |
|---|---|---|---|
| Progression-free survival | FAS including all randomized participants | Ribociclib + letrozole vs placebo + letrozole | Log-rank test |
| Overall survival | FAS including all randomized participants | Ribociclib + letrozole vs placebo + letrozole | Log-rank test |
| Overall response rate | FAS including all randomized participants | Ribociclib + letrozole vs placebo + letrozole | Cochran-Mantel-Haenszel test |
| Clinical benefit rate | FAS including all randomized participants | Ribociclib + letrozole vs placebo + letrozole | Cochran-Mantel-Haenszel test |
The registry data identifies two principal statistical method families: the log-rank test for time-to-event outcomes and the Cochran-Mantel-Haenszel test for binary outcomes. The primary hypothesis type and the secondary analyses are described as superiority comparisons.
5. Endpoints
| Endpoint | Registry definition / description | Time frame | Type |
|---|---|---|---|
| Progression Free Survival (PFS) by Investigator Assessment | PFS was defined as the period starting from the date of randomization to the date of the first documented progression or death caused by any reason. In cases where patients did not experience an event, PFS was censored at the date of the last adequate tumor assessment. Clinical deterioration without objective radiological evidence was not considered as documented disease progression. | Up to 23 months | Time-to-event |
| Overall Survival (OS) | Overall survival | Up to approximately 87 months | Time-to-event |
| Overall Response Rate (ORR) by Investigator Assessment | Overall response rate by investigator assessment | Up to 23 months | Binary |
| Clinical Benefit Rate (CBR) by Investigator Assessment | Clinical benefit rate by investigator assessment | Up to 23 months | Binary |
The registry lists one primary endpoint: progression-free survival by investigator assessment. The other three statistical analyses reported in the ClinicalTrials.gov record are secondary endpoints.
6. Statistical Methodology
Time-to-event analysis
Progression-free survival and overall survival are time-to-event endpoints. Unlike a simple binary outcome, these endpoints retain information about when an event occurred and can also accommodate participants whose event status is not observed by the end of follow-up.
For PFS in MONALEESA-2, the registry definition begins at randomization and ends at the first documented progression or death from any reason. Participants without an event were censored at the date of their last adequate tumor assessment. This makes censoring part of the statistical definition of the endpoint rather than merely an administrative detail.
A survival function represents the probability that the event time exceeds a specified time. In a clinical trial, Kaplan-Meier estimation is a standard way to estimate this function in the presence of right-censored observations.
Log-rank test
The registry reports the log-rank test for both PFS and OS. The log-rank procedure compares the observed pattern of events between treatment groups over follow-up rather than comparing only a single time point.
For the primary PFS endpoint, the analysis therefore asks whether the observed time-to-progression-or-death experience differs between the randomized ribociclib-plus-letrozole group and the placebo-plus-letrozole group.
Hazard ratio
The primary and OS analyses use the hazard ratio as their reported effect measure. A hazard ratio compares the estimated instantaneous event rates between the treatment groups over the analyzed follow-up.
A hazard ratio is a relative time-to-event measure. It is not a probability, not a median survival time, and not the percentage of patients who benefit.
Cochran-Mantel-Haenszel test
The registry reports the Cochran-Mantel-Haenszel test for overall response rate and clinical benefit rate. This is a categorical-data method that can compare binary outcomes while accounting for stratification variables when such strata are part of the analysis specification.
The ClinicalTrials.gov record does not identify the specific stratification variables used for these analyses, so no particular stratification scheme is inferred here.
Full Analysis Set
All four posted analyses identify the Full Analysis Set as including all randomized participants. This is important because analyzing participants according to randomized assignment preserves the treatment comparison established by randomization rather than restricting the primary efficacy comparison only to participants who completed treatment.
7. Primary Result: Progression-Free Survival
The primary endpoint was progression-free survival by investigator assessment, measured for up to 23 months. The analysis used a log-rank test in the Full Analysis Set, including all randomized participants. The comparison was ribociclib plus letrozole versus placebo plus letrozole.
Hazard ratio for progression or death
95% CI: 0.429–0.720 · P = 0.00000329
Two-sided confidence interval · Superiority hypothesis
| Primary endpoint | Result |
|---|---|
| Endpoint | Progression Free Survival (PFS) by Investigator Assessment |
| Time frame | Up to 23 months |
| Analysis population | Full Analysis Set, including all randomized participants |
| Method | Log-rank test |
| Effect measure | Hazard ratio |
| Hazard ratio | 0.556 |
| 95% CI | 0.429–0.720 |
| P-value | 0.00000329 |
| Hypothesis | Superiority |
The hazard ratio of 0.556 means that, within the statistical framework represented by the reported time-to-event analysis, the estimated instantaneous hazard of progression or death was approximately 55.6% as high with ribociclib plus letrozole as with placebo plus letrozole. Expressed as a relative reduction in estimated hazard, this corresponds to approximately a 44.4% lower estimated hazard.
That interpretation does not mean that 44.4% of participants avoided progression, that 44.4% of participants benefited, or that each participant had exactly a 44.4% reduction in risk. A hazard ratio is a relative time-to-event effect measure, not an individual probability.
The 95% confidence interval of 0.429–0.720 describes statistical uncertainty around the estimated hazard ratio. It indicates that the estimate is not being presented as an exact fixed quantity; the interval represents the uncertainty associated with the estimate under the relevant statistical framework.
The P-value of 0.00000329 addresses evidence against the null hypothesis in the reported statistical test. It does not measure the size or clinical importance of the treatment effect. Effect magnitude is conveyed by the hazard ratio and its confidence interval.
Because PFS includes censoring, interpretation also depends on the endpoint's censoring rules. The registry specifically states that participants without an event were censored at the date of the last adequate tumor assessment. The ClinicalTrials.gov record does not report a separate assessment of the proportional-hazards assumption, so no such assumption check is attributed to this trial record.
8. Secondary Result: Overall Survival
Overall survival was a secondary time-to-event endpoint with a time frame of up to approximately 87 months. The Full Analysis Set was compared using a log-rank test, with hazard ratio as the effect measure.
Hazard ratio for overall survival
95% CI: 0.628–0.932 · P = 0.004
Two-sided confidence interval · Superiority hypothesis
| Secondary endpoint | Result |
|---|---|
| Endpoint | Overall Survival (OS) |
| Time frame | Up to approximately 87 months |
| Analysis population | Full Analysis Set, including all randomized participants |
| Method | Log-rank test |
| Effect measure | Hazard ratio |
| Hazard ratio | 0.765 |
| 95% CI | 0.628–0.932 |
| P-value | 0.004 |
| Hypothesis | Superiority |
The OS hazard ratio of 0.765 corresponds to an estimated instantaneous hazard of death approximately 76.5% as high in the ribociclib-plus-letrozole group as in the placebo-plus-letrozole group, or approximately a 23.5% lower estimated hazard under the reported analysis.
This does not mean that 23.5% of participants were prevented from dying, nor does it describe an absolute difference in survival probability. A hazard ratio summarizes a relative time-to-event comparison.
The 95% confidence interval of 0.628–0.932 expresses uncertainty around the estimated hazard ratio. It is not an interval containing the outcomes of individual participants and does not state that individual treatment effects must fall between those two numbers.
The P-value of 0.004 measures the statistical evidence against the null hypothesis under the reported test. It should not be interpreted as a 0.4% probability that the treatment effect is real, nor as a measure of effect magnitude.
OS is less directly tied to tumor-assessment censoring than PFS because death is the event itself, but time-to-event interpretation still depends on follow-up, censoring, and the analysis framework. The ClinicalTrials.gov record does not provide additional information about proportional-hazards diagnostics or alternative survival models.
9. Secondary Result: Overall Response Rate
Overall response rate by investigator assessment was a secondary binary endpoint with a time frame of up to 23 months. The analysis used the Full Analysis Set and a Cochran-Mantel-Haenszel test.
Statistical comparison
Cochran-Mantel-Haenszel test · Superiority hypothesis
The registry analysis does not provide an effect estimate or confidence interval for this endpoint.
| Endpoint | Result |
|---|---|
| Outcome | Overall Response Rate (ORR) by Investigator Assessment |
| Time frame | Up to 23 months |
| Endpoint type | Binary |
| Analysis population | Full Analysis Set, including all randomized participants |
| Method | Cochran-Mantel-Haenszel test |
| P-value | 0.000155 |
| Hypothesis | Superiority |
| Effect estimate / CI | Not included in the ClinicalTrials.gov record |
The reported P-value of 0.000155 indicates strong statistical evidence against the null hypothesis under the reported Cochran-Mantel-Haenszel test. However, the ClinicalTrials.gov record does not provide the actual response rates, an absolute difference, a risk ratio, an odds ratio, or a confidence interval.
Consequently, the P-value alone cannot tell us how large the difference in response was. Two trials can have similarly small P-values with very different effect sizes, depending on sample size and outcome variability.
The binary nature of ORR also makes it conceptually different from PFS and OS. ORR reduces the response experience to whether a participant meets the prespecified response classification, whereas time-to-event endpoints incorporate the timing of events.
10. Secondary Result: Clinical Benefit Rate
Clinical benefit rate by investigator assessment was another secondary binary endpoint, evaluated for up to 23 months. The analysis used the Full Analysis Set and the Cochran-Mantel-Haenszel test.
Statistical comparison
Cochran-Mantel-Haenszel test · Superiority hypothesis
The registry analysis does not provide an effect estimate or confidence interval for this endpoint.
| Endpoint | Result |
|---|---|
| Outcome | Clinical Benefit Rate (CBR) by Investigator Assessment |
| Time frame | Up to 23 months |
| Endpoint type | Binary |
| Analysis population | Full Analysis Set, including all randomized participants |
| Method | Cochran-Mantel-Haenszel test |
| P-value | 0.018 |
| Hypothesis | Superiority |
| Effect estimate / CI | Not included in the ClinicalTrials.gov record |
The CBR comparison has a reported P-value of 0.018. Under the registry-reported statistical framework, this provides evidence against the null hypothesis of the reported comparison.
Again, the P-value does not quantify the magnitude of the treatment difference. Because the registry analysis does not include the underlying CBR percentages or a confidence interval, the statistical record cannot support a numerical statement about the absolute size of the difference.
The distinction between statistical evidence and effect magnitude is important: an inferential test answers whether the observed data are inconsistent with the null model, while an effect estimate describes the size and direction of the observed difference.
11. Safety Results
The ClinicalTrials.gov record reports serious adverse events by arm as affected participants divided by the number at risk. These data should be interpreted separately from the efficacy analyses because safety is concerned with adverse-event occurrence rather than time to progression or death.
| Group | Serious adverse events affected | At risk |
|---|---|---|
| Ribociclib + Letrozole | 108 | 334 |
| Placebo + Letrozole | 62 | 330 |
| Crossover to Ribociclib + Letrozole | 1 | 4 |
The ClinicalTrials.gov record therefore identifies 108/334 participants with serious adverse events in the ribociclib-plus-letrozole group and 62/330 in the placebo-plus-letrozole group. It separately identifies a crossover group of 1/4. The ClinicalTrials.gov record does not provide a formal statistical comparison, confidence interval, or P-value for these serious-adverse-event counts.
12. What the Primary Hazard Ratio Means
The primary PFS hazard ratio of 0.556 indicates a lower estimated instantaneous hazard of progression or death in the ribociclib-plus-letrozole group relative to the placebo-plus-letrozole group under the reported analysis. The corresponding relative reduction in estimated hazard is approximately 44.4%.
This is a relative measure. It does not say that 44.4% of participants were progression-free, that 44.4% of participants were cured, or that every patient experienced the same proportional reduction.
The 95% confidence interval of 0.429–0.720 communicates uncertainty around the PFS hazard-ratio estimate. The interval is relatively narrower than an extremely imprecise estimate would be, but its width should still be considered when describing the precision of the treatment effect.
A confidence interval is not a prediction interval for individual patients. It also does not mean that 95% of future individual hazard ratios will fall between 0.429 and 0.720.
The primary P-value of 0.00000329 measures evidence against the null hypothesis for the reported superiority comparison. It is not an effect-size measure. The hazard ratio and its confidence interval are needed to describe the magnitude and precision of the treatment comparison.
13. Why Time-to-Event Endpoints Require Special Interpretation
PFS and OS are not ordinary binary outcomes because participants can enter the analysis at different times and may be followed for different lengths of time. A participant who has not experienced the event by the end of observation is not necessarily equivalent to a participant who was never at risk; instead, the observed follow-up contributes information up to the censoring point.
Timing matters
A time-to-event analysis incorporates when events occur, rather than simply counting how many participants experienced an event by a fixed date.
Censoring matters
For PFS, the registry specifically states that participants without an event were censored at their last adequate tumor assessment.
Hazard is not risk
A hazard ratio compares event hazards. It should not be described as though it were a simple ratio of cumulative probabilities.
Follow-up matters
The registry gives different follow-up windows for PFS, ORR and CBR versus OS, so their results should not be treated as though they came from identical observation periods.
14. Statistical Methods Explained
Why was a log-rank test used for PFS?
PFS is a time-to-event endpoint, so a method designed for time-to-event data is appropriate. The log-rank test compares the survival experience between treatment groups over the period of observation rather than reducing the endpoint to a single binary outcome. In MONALEESA-2, the registry explicitly reports the log-rank test for PFS.
What does a hazard ratio of 0.556 mean?
A hazard ratio of 0.556 means that the estimated instantaneous hazard in the ribociclib-plus-letrozole group was approximately 55.6% of that in the placebo-plus-letrozole group under the reported analysis. Equivalently, the estimated hazard was approximately 44.4% lower. It does not mean that 44.4% of patients benefited or that the probability of progression was reduced by exactly 44.4% at every time point.
Why is the confidence interval important?
The estimate 0.556 is only one estimate of the treatment effect. The 95% confidence interval of 0.429–0.720 describes uncertainty around that estimate. Reporting the interval prevents the point estimate from being interpreted as though it were known without statistical uncertainty.
Why does the P-value not measure effect size?
The primary P-value of 0.00000329 describes the evidence against the null hypothesis in the specified statistical test. It does not tell us whether the effect is large or small in clinical terms. The hazard ratio provides the relative effect estimate, while the confidence interval provides information about its precision.
Why was the Cochran-Mantel-Haenszel test used for ORR and CBR?
ORR and CBR are binary outcomes in the registry classification. The Cochran-Mantel-Haenszel test is a categorical-data procedure that can compare treatment groups while accounting for strata when a stratified analysis is specified. The registry explicitly reports this method for both secondary endpoints.
Why does the Full Analysis Set matter?
The posted analyses identify the Full Analysis Set as including all randomized participants. Analyzing randomized participants according to treatment assignment maintains the treatment comparison generated by randomization and avoids redefining the efficacy population based solely on subsequent treatment exposure or completion.
15. Primary and Secondary Statistical Evidence
| Endpoint | Role | Type | Method | Effect measure | Result |
|---|---|---|---|---|---|
| Progression Free Survival by Investigator Assessment | Primary | Time-to-event | Log-rank | HR 0.556 (95% CI 0.429–0.720) | P = 0.00000329 |
| Overall Survival | Secondary | Time-to-event | Log-rank | HR 0.765 (95% CI 0.628–0.932) | P = 0.004 |
| Overall Response Rate by Investigator Assessment | Secondary | Binary | Cochran-Mantel-Haenszel | Not reported in registry-reported analysis | P = 0.000155 |
| Clinical Benefit Rate by Investigator Assessment | Secondary | Binary | Cochran-Mantel-Haenszel | Not reported in registry-reported analysis | P = 0.018 |
This table illustrates an important distinction in clinical-trial reporting: a formal hypothesis test may be available without a corresponding effect estimate in the registry's statistical-analysis record. For ORR and CBR, the ClinicalTrials.gov record supports reporting the P-values and methods, but not inventing response percentages or confidence intervals that are not present.
16. Multiplicity, Interim Analysis, and Other Design Issues
The registry-reported MONALEESA-2 data identifies the primary endpoint, three secondary statistical analyses, and superiority hypotheses. It does not provide information about a non-inferiority margin, a prespecified interim-analysis boundary, an alpha-spending procedure, a multiplicity-adjustment strategy, or a detailed endpoint hierarchy.
| Design topic | What the ClinicalTrials.gov record supports |
|---|---|
| Superiority | Reported for the primary PFS analysis and all three registry-reported secondary analyses. |
| Non-inferiority margin | Not applicable to the reported superiority analyses; no non-inferiority margin is reported. |
| Interim analysis | No interim-analysis method or boundary is reported. |
| Multiplicity | No multiplicity-adjustment procedure is reported. |
| Stratification | No specific stratification factors are reported in the ClinicalTrials.gov record. |
| Bayesian methods | No Bayesian analysis is reported. |
| Missing-data imputation | No imputation method is posted on ClinicalTrials.gov for the posted analyses. |
| Crossover | The ClinicalTrials.gov record identifies 1/4 participants in a group labeled crossover to ribociclib + letrozole; the ClinicalTrials.gov record does not provide a detailed crossover-analysis methodology. |
17. Crossover and Interpretation of the Randomized Comparison
The ClinicalTrials.gov record identifies a group labeled Crossover to Ribociclib + Letrozole, with serious adverse events affecting 1 of 4 participants. This establishes that a crossover category is represented in the ClinicalTrials.gov record.
However, the available data does not provide enough information to determine how crossover was incorporated into the primary efficacy analysis, whether crossover occurred before or after a particular endpoint event, or whether a sensitivity analysis adjusted for crossover. Therefore, the primary PFS and OS results should be interpreted using the reported randomized Full Analysis Set comparison rather than constructing an unreported crossover-adjusted effect.
18. Understanding the Four Posted Statistical Analyses
| Analysis | Question addressed | Why the method fits |
|---|---|---|
| PFS | Do the randomized groups differ in time to first documented progression or death? | PFS is a time-to-event endpoint; the registry reports a log-rank comparison and hazard ratio. |
| OS | Do the randomized groups differ in overall survival? | OS is a time-to-event endpoint; the registry reports a log-rank comparison and hazard ratio. |
| ORR | Do the randomized groups differ in the binary response outcome? | ORR is classified as binary; the registry reports a Cochran-Mantel-Haenszel comparison. |
| CBR | Do the randomized groups differ in the binary clinical-benefit outcome? | CBR is classified as binary; the registry reports a Cochran-Mantel-Haenszel comparison. |
This is a useful example of matching the statistical method to the structure of the endpoint. The same randomized trial can require different inferential methods because PFS and OS contain timing and censoring information, while ORR and CBR are classified as binary outcomes.
19. Important Limitations and Interpretation Issues
- Limited registry detail: The ClinicalTrials.gov record provides the primary effect estimate and confidence interval, but do not provide all components of a full statistical analysis plan.
- Missing binary effect estimates: ORR and CBR have reported P-values and methods, but the ClinicalTrials.gov record does not provide the corresponding response percentages, effect estimates, or confidence intervals.
- Censoring: PFS includes censoring at the date of the last adequate tumor assessment for participants without an event.
- Hazard-ratio interpretation: HRs are relative time-to-event measures and should not be interpreted as absolute risk reductions or individual treatment probabilities.
- Proportional hazards: The ClinicalTrials.gov record does not report a formal assessment of the proportional-hazards assumption, so the hazard ratios should not be supplemented with an unreported assumption check.
- Multiplicity: The ClinicalTrials.gov record does not specify a multiplicity-adjustment strategy across the primary and secondary endpoints.
- Interim monitoring: No interim-analysis boundary or alpha-spending method is included in the ClinicalTrials.gov record.
- Stratification: The methods include the Cochran-Mantel-Haenszel test, but the ClinicalTrials.gov record does not identify the specific strata used.
- Safety interpretation: Serious-adverse-event counts are reported separately from efficacy analyses and should not be treated as though they were time-to-event efficacy estimates.
- Crossover: A crossover group is represented in the ClinicalTrials.gov record, but its quantitative influence on the efficacy estimates cannot be determined from the provided information.
20. Why This Trial Matters Statistically
MONALEESA-2 is a useful teaching example because the ClinicalTrials.gov record brings together two major classes of clinical-trial endpoints and two corresponding statistical approaches. The primary endpoint is a time-to-event outcome analyzed with a log-rank test and summarized with a hazard ratio, while ORR and CBR are binary outcomes analyzed with the Cochran-Mantel-Haenszel test.
| Concept | How it appears in MONALEESA-2 |
|---|---|
| Randomization | The trial uses randomized allocation with 668 enrolled participants. |
| Parallel design | The trial is classified as a parallel-group study with 2 arms. |
| Quadruple masking | The registry classifies the trial as quadruple masked. |
| Full Analysis Set | The posted analyses include all randomized participants in the FAS. |
| Time-to-event endpoint | PFS is the registered primary endpoint; OS is a secondary endpoint. |
| Log-rank test | Used for the PFS and OS statistical analyses. |
| Hazard ratio | Reported for both PFS and OS. |
| Confidence interval | Reported for the primary PFS HR and secondary OS HR. |
| Binary endpoint | ORR and CBR are classified as binary outcomes. |
| Cochran-Mantel-Haenszel test | Used for ORR and CBR. |
| Superiority testing | Reported hypothesis type for all four registry-reported statistical analyses. |
| Censoring | Specified for participants without a PFS event at their last adequate tumor assessment. |
21. Reading the Results as a Statistical Story
The primary PFS analysis gives the clearest numerical picture in the ClinicalTrials.gov record. The hazard ratio of 0.556, together with its 95% confidence interval of 0.429–0.720 and P-value of 0.00000329, provides three complementary pieces of information: the estimated direction and magnitude of the relative treatment effect, the uncertainty around that estimate, and the strength of evidence against the null hypothesis under the specified test.
The OS analysis uses the same general time-to-event framework but is a secondary endpoint and has a different reported effect estimate: HR 0.765, with a 95% CI of 0.628–0.932 and P = 0.004. The two hazard ratios therefore should not be collapsed into a single overall treatment-effect number.
ORR and CBR add binary-outcome evidence. Their P-values, 0.000155 and 0.018, respectively, provide inferential information, but the registry analysis does not provide the underlying percentages or effect estimates. A statistically careful summary therefore reports those P-values without inventing a numerical treatment difference.
22. Results Summary
| Endpoint | Effect | 95% CI | P-value | Interpretive role |
|---|---|---|---|---|
| Primary PFS | HR 0.556 | 0.429–0.720 | 0.00000329 | Primary time-to-event evidence |
| Secondary OS | HR 0.765 | 0.628–0.932 | 0.004 | Secondary time-to-event evidence |
| Secondary ORR | Not reported | Not reported | 0.000155 | Binary-outcome evidence |
| Secondary CBR | Not reported | Not reported | 0.018 | Binary-outcome evidence |
23. Related Tutorials
Learn more about the methods used in this trial:
24. Related Calculators
25. Sources
- ClinicalTrials.gov: NCT01958021 — MONALEESA-2.
- PubMed record: PMID 42334791.
- PubMed record: PMID 38107828.
- PubMed record: PMID 37673211.
- PubMed record: PMID 36861085.
- PubMed record: PMID 36800111.
Continue with the statistical methods
Explore the underlying survival-analysis, categorical-data, clinical-trial, and statistical-inference concepts used to interpret randomized trial evidence.
26. Record Summary
MONALEESA-2 provides a useful example of how a randomized phase 3 clinical trial can combine different statistical methods according to endpoint structure. Its registered primary endpoint, progression-free survival by investigator assessment, is a time-to-event outcome analyzed with a log-rank test and reported as a hazard ratio of 0.556 with a 95% confidence interval of 0.429–0.720 and P = 0.00000329. Overall survival was analyzed similarly, with HR 0.765, 95% CI 0.628–0.932, and P = 0.004. Overall response rate and clinical benefit rate were analyzed as binary endpoints using the Cochran-Mantel-Haenszel test, with reported P-values of 0.000155 and 0.018, respectively.
The most important statistical lesson is that these numbers answer different questions. Hazard ratios quantify relative time-to-event effects; confidence intervals describe uncertainty around those estimates; P-values quantify evidence against a null hypothesis under a specified test; and binary-outcome analyses require their own effect measures for understanding the magnitude of a difference. A careful interpretation therefore preserves the distinction between statistical evidence, effect size, uncertainty, endpoint definition, and analysis population.